Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The neurohypophysial hormone AVP produced by magnicellular neurons in the supraoptic and paraventricular nuclei of the hypothalamus is in a form bound to the specific peptide neurophysin II in the neurosecretory granula. It is stored in the posterior pituitary and released into the circulation on a decrease in blood pressure or an increase in osmolality. AVP has been shown to play important physiological roles in vasoconstriction and antidiuresis and to exert its effect through binding to specific receptors coupled to distinct second messengers. These AVP receptors have been classified according to the second messenger system to which they are coupled, and at least three AVP receptor subtypes (V_1A, V_1B and V₂) have been identified. Recently, these three AVP receptor subtypes have been cloned and found to belong to the family of seven membrane-spanning receptors that signal through G protein (Thibonnier et al., 1994; Sugimoto et al., 1994; Birnbaumer et al., 1992). AVP activates PLC-mediated hydrolysis of polyphosphoinositides via the V_1A and V_1B receptors to generate two second messengers, inositol-1,4,5-triphosphate, which induces an increase of free intracellular calcium from the endoplasmic reticulum, and 1,2-diacylglycerol, which activates protein kinase C (Michell et al., 1979). The V_1A receptors have been shown to be present in vascular smooth muscle cells (Serradeil-Le-Gal et al., 1995), hepatocytes (Howl et al., 1991), platelets (Thibonnier and Roberts, 1985) and mesangial cells (Jard et al., 1987) and mediate contraction, proliferation and hypertrophy of cells, platelet aggregation and hepatocyte glycogenolysis. The V_1B receptors are located in the anterior pituitary where they stimulate corticotropin release (Jard et al., 1986). In contrast, the V₂ receptors stimulate adenylate cyclase, which results in the production of cAMP. The V₂ receptors are located in the medullary portion of the kidney where they control free water and urea reabsorption (Butlen et al., 1978).

AVP may play a role in several disease conditions, including heart failure, hypertension, hyponatremia and syndrome of inappropriate antidiuretic hormone secretion; AVP antagonists may be useful in the treatment of these diseases (Fujisawa et al., 1993; Naitoh et al., 1994; Laszlo et al., 1991). Although several potent peptide AVP antagonists have been studied for therapeutic use, their partial agonistic activity effects and lack of oral bioavailability have severely limited clinical and therapeutic investigation (Manning and Sawyer, 1986). Recently, orally effective nonpeptide AVP antagonists have been discovered (Serradeil-Le Gal et al., 1993; Yamamura et al., 1991, 1992). However, all of these are selective antagonists of either V_1A or V₂ AVP receptors.

YM087 (fig. 1) is a newly synthesized potent nonpeptide antagonist of both AVP V_1A and V₂ receptors (Yatsu et al., 1997). In this study, we examined the biochemical and pharmacological properties of YM087 by receptor binding, second messenger and in vivo assays.

View larger version (K): [in this window] [in a new window]   Fig. 1.   Chemical structure of YM087, 4-[(2-methyl-1,4,5,6-tetrahydro-imidazo[4,5-d][1]benzazepin-6-yl)-carbonyl]-2-phenylbenzanilide monohydrochloride.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. The radioligands [³H]AVP and [³H]OT with a specific activity of 80 and 50 Ci/mmol, respectively, were obtained from DuPont-New England Nuclear (Boston, MA). AVP and OT were obtained from Peptide Institute Inc. (Osaka, Japan) and d(CH₂)₅Tyr(Me)AVP (Kruszynski et al., 1980) and dDAVP (Zaoral et al., 1967) were from Sigma Chemical Co. (St. Louis, MO). YM087, OPC-21268 and OPC-31260 were synthesized at Yamanouchi Pharmaceutical Co. (Ibaraki, Japan). These nonpeptide antagonists were initially dissolved in dimethyl sulfoxide at 10² M and diluted to the desired concentration with the assay buffer. The final concentration of dimethyl sulfoxide in the assay buffer did not exceed 1%, at which [³H]AVP/OT binding was not affected. Diethylstilbestrol dipropionate was obtained from Sigma. Fura 2-AM was obtained from Dojindo Laboratories (Kumamoto, Japan) and EGTA, ionomycin and IBMX were from Wako Pure Chemicals (Osaka, Japan). The cAMP assay kit was obtained from Amersham (Buckingham, UK). DMEM, phosphate-buffered saline, FCS, antibiotics (penicillin and streptomycin) and trypsin-EDTA were from Gibco (Grand Island, NY). BSA was purchased from Nacalai Tesque Inc. (Kyoto, Japan). Reagent for protein assay was purchased from Bio-Rad Laboratories (Richmond, CA).

Animals. Male/Female Wistar rats (250-300 g) were used as indicated. All animals were housed in communal cages and maintained on a 12-hr light/dark cycle with food and water available ad libitum. All experiments were performed under the regulations of the Animal Ethical Committee of Yamanouchi Pharmaceutical Co., Ltd.

Radioligand binding studies. Rats were sacrificed by decapitation, and their liver, kidney, pituitary and uterus were quickly removed. All subsequent preparative steps were carried out at 4°C. Liver plasma membranes were prepared by a modification of the method of Nakamura et al. (1983). Kidney medulla plasma membranes were prepared by the method of Campbell et al. (1972). Anterior pituitary plasma membranes were prepared by the method of Lutz-Bucher and Koch (1983). Uterine plasma membranes were prepared by the method of Pettibone et al. (1990) from uterine tissue isolated from female Wistar rats treated with diethylstilbestrol dipropionate at 0.3 mg/kg i.p. 18-24 hr before isolation. Membrane preparations from A10 and LLC-PK₁ cells were prepared as described previously (Sugimoto et al., 1994). Protein was determined by the method of Bradford with BSA as standard (Bradford, 1976). For saturation binding studies, plasma membrane preparations were incubated with various concentrations of [³H]AVP/[³H]OT (0.1-3.0 nM). For competition studies, radioligands (0.5 nM) were added to each membrane preparation and were incubated with various concentrations of compounds in 250 µl of assay buffer containing 50 mM Tris-HCl, pH 7.5, 5 mM MgCl₂ and 0.1% BSA. The binding reactions were initiated by the addition of the plasma membrane preparations and incubations were for 60 min at 25°C, which allowed equilibrium to be established. After incubation, the reaction was terminated by addition of 3 ml of ice-cold Tris buffer (50 mM Tris-HCl, pH 7.5, and 5 mM MgCl₂) followed immediately by rapid filtration through 96-well GF/B UniFilter Plates with a MicroMate Cell Harvester (Packard Instrument Company, Rockville, MD) (Holland et al., 1994). The filters were rinsed twice and the radioactivity retained on the filters was counted with TopCount Microplate Scintillation Counter (Packard Instrument Company). Nonspecific binding was determined with 1 µM unlabeled AVP/OT. Specific binding was calculated as the total binding minus nonspecific binding. The concentration of test compound that caused 50% inhibition (IC₅₀) of the specific binding of [³H]AVP/OT was determined by regression analysis of displacement curves. The inhibitory dissociation constant (K_i) was calculated from the following formula: where [L] is the concentration of radioligand present in the tubes and K_d is the dissociation constant of radioligand obtained from the Scatchard plot (Cheng and Prusoff, 1973). Data were analyzed with the GraphPad PRISM (GraphPAD Software, Inc., San Diego, CA).

Measurement of [Ca⁺⁺]_i. A10 cells from the thoracic aorta of rats were obtained from the American Tissue Culture Collection (Rockville, MD) (Kimes and Brandt, 1976). The cells were grown at 37°C in 75-cm² culture flasks under a humidified 95% air/5% CO₂ atmosphere in DMEM supplemented with 10% FCS and antibiotics. Cells were subcultured every 7 days in culture flasks with 0.25% trypsin-0.01% EDTA, whereas media were renewed every 3 days. Serum-deprived monolayer cultures of cells were grown on coverglasses and assayed 2 days later. Cell monolayers (2 × 10⁵ cells/cm²) were loaded in serum-free DMEM with fura 2-AM (1 µM/coverglass) for 30 min at 37°C. The permeable acetoxymethyl ester of fura 2 is hydrolyzed by cellular esterases on entering the cells to the fura 2 form, which is relatively impermeable and becomes trapped inside the cells. The monolayers of cells were washed, then transferred to serum-free, fura 2-free medium and incubated for an additional 30 min at 37°C. The loaded monolayers were stored in Krebs-Henseleit-HEPES buffer (containing 130 mM NaCl, 5 mM KCl, 1.25 mM CaCl₂, 0.8 mM MgSO₄, 5.5 mM glucose, 20 mM HEPES, and 0.1% BSA, pH 7.4). The coverglass was placed in a quartz cuvette containing 2 ml Krebs-Henseleit-HEPES buffer and maintained at 37°C with continuous stirring. When thermal equilibrium was reached, the fluorescence signal was recorded with CAF-110 spectrofluorometer (Japan Spectrometer Co., Tokyo, Japan) with 340/380 nm excitation and 500 nm emission wavelength. After recording the base-line signal for 3 min, AVP was added to the cuvette to stimulate the mobilization of intracellular calcium in the presence or absence of antagonists. Fluorescence measurements were converted to [Ca⁺⁺]_i by determining maximal fluorescence (F_max) with the nonfluorescent Ca⁺⁺ ionophore, ionomycin (25 µM), after which minimal fluorescence (F_min) was obtained by adding 3 mM EGTA. From the ratio of fluorescence at 340 and 380 nm, the [Ca⁺⁺]_i was determined by use of the following equation: [Ca²⁺]_i (nM) = K_d × [(R  R_min)/(R_max  R)] × b. The term b is the ratio of fluorescence of fura 2 at 380 nm in zero and saturating Ca⁺⁺. K_d is the dissociation constant of fura 2 for Ca⁺⁺, assumed to be 224 nM.

Measurement of cAMP production. LLC-PK₁ cells were purchased from the American Type Culture Collection (Rockville, MD) (Hull et al., 1976) and cultured with DMEM containing 10% FCS and antibiotics. Cell monolayers were grown in 12-well culture plates to 90 to 95% confluence and incubated in DMEM with 0.5 mM IBMX and 0.1% BSA containing vehicle or various concentrations of AVP and/or antagonists for 10 min at 37°C. At the end of incubation, the cell monolayers were washed three times with phosphate-buffered saline followed by lysis with boiling 0.5 mM sodium acetate, pH 6.2, containing 0.2 mM IBMX. Extracts were then boiled for 3 min and kept at 40°C before determination of cAMP with a competitive protein binding assay kit (Tovey et al., 1974).

Effects in pithed rats. Rats were anesthetized with ether and pithed by inserting a steel rod via the orbit and foramen magnum down into the whole length of the spinal canal. Immediately after pithing, the rats were bilaterally vagotomized at the neck, and artificial ventilation with room air was started with a rodent respirator at a frequency of 50 cpm and a volume of 1 ml/100 g b.wt. A catheter was inserted into the carotid artery and femoral vein for recording of arterial blood pressure and intravenous drug administration, respectively. Rats were kept warm at 37°C by means of a thermostat-controlled heating board. For i.v. injection, YM087 was dissolved in dimethylformamide, and other drugs were dissolved in saline. After stabilization of blood pressure, antagonists or vehicle was given (0.5 ml/kg i.v.) 5 min before the injection of AVP (30 mU/kg i.v.). The dose of antagonist causing a 50% inhibition of the pressor response to AVP (ID₅₀) was calculated.

Effects in dehydrated conscious rats. Rats were deprived of drinking water for 16 to 20 hr to stimulate endogenous AVP secretion. Antagonists or vehicle was administered intravenously and spontaneously voided urine was collected for a 2-hr period with use of a metabolic cage. The dose causing an increase in urine volume by 3 ml after antagonist dosing (ED₃) was determined.

Data analysis. Experimental results are expressed as the mean ± S.E.M. or the mean with 95% confidence limits. All experiments were repeated at least three times, and comparable results were obtained. The EC₅₀ and IC₅₀ values were estimated from the concentration-response curves by a nonlinear regression program GraphPad PRISM (GraphPAD Software, Inc., San Diego, CA).

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Radioligand binding studies. Saturation experiments of increasing concentrations of [³H]AVP binding to rat liver (V_1A), pituitary (V_1B) and kidney (V₂) plasma membranes showed that specific binding was saturable. Scatchard analysis of these data gave linear plots consistent with the presence of a single class of high-affinity binding site. The apparent dissociation constant (K_d) was 1.07 ± 0.18 nM (liver), 0.28 ± 0.05 nM (pituitary) and 1.43 ± 0.36 nM (kidney), and maximum binding capacity (B_max) was 538 ± 82.0 fmol/mg protein (liver), 75.9 ± 6.41 fmol/mg protein (pituitary) and 90.5 ± 18.0 fmol/mg protein (kidney). In addition, saturation binding experiments with [³H]OT were performed in rat uterus plasma membranes. A single class of binding sites was identified with K_d = 0.72 ± 0.11 nM and B_max = 290 ± 62.8 fmol/mg protein. In the absence or presence of YM087 (0.3, 1.0, 3.0 nM), [³H]AVP saturation binding experiments were performed in rat liver and kidney plasma membranes. In both preparations, increasing concentrations of YM087 caused successive decreases in the slopes of the curves consistent with an increase in equilibrium dissociation constant (K_d) without a reduction in receptor density (B_max) (fig. 2). The K_d value of rat liver membranes was 0.43 ± 0.11 nM for control and 2.14 ± 0.27 nM for YM087 (3.0 nM), and the K_d value for rat kidney membranes was 0.88 ± 0.12 nM for control and 3.80 ± 0.73 nM for YM087 (3.0 nM). The B_max value of rat liver membranes was 165 ± 14.2 fmol/mg protein for control and 179 ± 13.0 fmol/mg protein for YM087 (3.0 nM), and the B_max value of rat kidney membranes was 59.1 ± 5.45 fmol/mg protein for control and 63.9 ± 10.2 fmol/mg protein for YM087 (3.0 nM). From these saturation experiments, the calculated K_i values for YM087 were 1.29 ± 0.24 nM (V_1A) and 2.84 ± 1.15 nM (V₂) with use of the equation K_d app = K_d × (1 + C/K_i), where K_d app is the apparent K_d in the presence of YM087 and C is the concentration of YM087. These results were consistent with the directly measured K_i values of 0.48 nM (V_1A) and 3.04 nM (V₂) obtained from competition experiments with [³H]AVP. YM087 and several AVP receptor agonists and antagonists were tested for their ability to compete for specific [³H]AVP/OT binding (fig. 3). The inhibition constants (K_i) of AVP agonists and antagonists are shown in table 1. YM087 inhibited specific binding of [³H]AVP to V_1A and V₂ receptors in a concentration-dependent manner, with K_i values of 0.48 ± 0.07 nM and 3.04 ± 1.51 nM, respectively. The Hill coefficients (n_H) of YM087 were close to unity, which suggested a single-site competitive model. The affinity of YM087 for V_1A receptors was 49 times higher than that of OPC-21268 (K_i = 23.5 ± 4.39 nM) whereas that for V₂ receptors was 14 times higher than that of OPC-31260 (K_i = 42.3 ± 14.3 nM). On the contrary, YM087 exhibited much lower affinity for OT receptors, with K_i values of 44.4 ± 13.1 nM, and did not reduce specific binding to V_1B receptors at all (K_i >100 µM). The specificity of the effects of YM087 at AVP receptors was examined in 27 receptor binding screens. YM087 had no affinity (IC₅₀ > 10,000 nM) for adenosine, adrenergic, dopamine, -aminobutyric acid, histamine, serotonin, muscarinic, benzodiazepine, angiotensin II, endothelin, neuropeptide Y, TRH (thyrotropin releasing hormone), VIP (vasoactive intestinal peptide), ANF (atrial natriuretic factor), CRF (corticotropin releasing factor), EGF (epidermal growth factor), forskolin, phorbol ester, inositol triphosphate, estradiol and testosterone receptors (data not shown).

View larger version (K): [in this window] [in a new window]   Fig. 2.   Scatchard plots of [3H]AVP binding to rat liver (A) and kidney (B) plasma membranes in the absence () or presence of YM087 at concentrations of 0.3 nM (), 1 nM () and 3 nM (). Binding assays were performed as described under "Materials and Methods." Results represent data from five experiments performed in duplicate.

View larger version (K): [in this window] [in a new window]   Fig. 3.   Displacement of the specific binding of [3H]AVP/OT to plasma membranes prepared from rat liver (V1A), kidney (V2), pituitary (V1B) and uterus (OT). The unlabeled compounds added to the binding assay are as follows: , AVP; , OT; , YM087; , OPC-21268; , OPC-31260. Specific binding of [3H]AVP/OT is expressed as the percentage of the control binding. Results are representative data from three to seven experiments performed in duplicate. The combined results of all experiments are summarized in table 1.

Measurement of [Ca⁺⁺]_i. The binding of [³H]AVP to A10 cell plasma membranes was saturable, Scatchard analysis of the binding data revealed a linear plot demonstrating a single class of AVP binding sites with a K_d of 0.65 ± 0.10 nM and a B_max of 558 ± 83.8 fmol/mg protein. The ability of several AVP/OT agonists and antagonists to displace [³H]AVP from A10 cell membranes was examined. Based on the K_i values, the ranking order of [³H]AVP binding inhibition was as follows: YM087 > d(CH₂)₅Tyr(Me)AVP > AVP > OPC-21268 > OT > OPC-31260 (table 2). YM087 showed high affinity for A10 cell V_1A receptors with a K_i value of 0.24 ± 0.03 nM. The affinity of YM087 to this V_1A receptor was 62 times greater than that of OPC-21268 (K_i = 14.9 ± 5.07 nM), and this affinity profile agreed well with those obtained with expressed V_1A receptors from rat liver. An increase of [Ca⁺⁺]_i is the hallmark of V_1A receptor activation. The basal [Ca⁺⁺]_i in the A10 cells was 61.1 ± 6.26 nM. AVP induced an increase in [Ca⁺⁺]_i in a concentration-dependent manner, with an EC₅₀ value of 5.29 (2.54-11.1) nM (fig. 4A). At low concentrations of AVP, the time to attain the increase in [Ca⁺⁺]_i was slower, and this lag phase was abolished at higher concentrations. In this experiment, the threshold concentration at which AVP evoked observable increases in [Ca⁺⁺]_i was approximately 0.3 nM, and the maximal effect was attained at a concentration of 30 to 100 nM. YM087 strongly and concentration-dependently inhibited the increase in [Ca⁺⁺]_i stimulated by AVP (10⁷ M) with an IC₅₀ value of 1.16 (0.68-1.96) nM (fig. 4B). YM087 did not change the basal value of [Ca⁺⁺]_i when tested alone (up to 100 µM). The inhibitory potency of YM087 was 31 times higher than that of OPC-21268 [IC₅₀ = 35.4 (27.7-45.2) nM]; this potency profile was similar to its binding affinity.

View larger version (K): [in this window] [in a new window]   Fig. 4.   Effect of YM087 on AVP-induced [Ca++]i increases in vascular smooth muscle cells. (A) Concentration-response curve for AVP-induced changes in [Ca++]i. (B) Inhibitory effect of YM087 on the response to a submaximal concentration of AVP (100 nM). Values are mean ± S.E.M. of three to eight independent experiments.

Measurement of cAMP production. The binding of [³H]AVP to LLC-PK₁ cell plasma membranes was saturable. Scatchard analysis of the binding data revealed a linear plot demonstrating a single class of AVP binding sites with a K_d of 0.69 ± 0.30 nM and a B_max of 450 ± 108 fmol/mg protein. In displacement studies using LLC-PK₁ cell membranes, the affinity of several AVP/OT receptor agonists and antagonists was determined. AVP and dDAVP, a selective V₂ agonist, showed high affinities, whereas OT and d(CH₂)₅Tyr(Me)AVP, a selective V_1A antagonist, were less potent (table 2). YM087 showed high affinity for LLC-PK₁ cell V₂ receptors, with a K_i value of 0.38 ± 0.09 nM. The affinity of YM087 to this V₂ receptor was 16 times higher than that of OPC-31260 (K_i = 6.10 ± 3.67 nM). AVP induced a concentration-dependent increase in cellular cAMP production with an EC₅₀ value of 18.6 (10.3-33.5) nM (fig. 5A). YM087 strongly and concentration-dependently inhibited 100 nM AVP-induced increase in cAMP production in LLC-PK₁ cells with an IC₅₀ value of 17.3 (11.3-26.5) nM (fig. 5B). No change in basal cAMP production occurred when YM087 was tested alone (up to 100 µM). The inhibitory potency of YM087 was 41 times higher than that of OPC-31260 (IC₅₀ = 711 (331-1530) nM) and this potency profile was well correlated with its binding affinity.

View larger version (K): [in this window] [in a new window]   Fig. 5.   Effect of YM087 on AVP-induced production of cellular cAMP in renal epithelial cells. (A) Concentration-response curve for AVP-induced production of cellular cAMP; (B) inhibitory effect of YM087 on AVP-induced production of cellular cAMP. Values are mean ± S.E.M. of six independent experiments.

Effects in rats in vivo. In pithed rats, intravenous administration of YM087 (0.003-0.03 mg/kg) significantly inhibited an AVP-induced increase in blood pressure in a dose-dependent manner with an ID₅₀ value of 0.013 ± 0.001 mg/kg. In contrast, it had no significant effect on the increases induced by angiotensin II (0.1 µg/kg i.v.) and norepinephrine (2 µg/kg i.v.) (fig. 6). Moreover, intravenous administration of YM087 (0.01-0.3 mg/kg) significantly increased urine volume with an ED₃ value of 0.028 (0.019-0.039) mg/kg and reduced urine osmolality in a dose-dependent manner in dehydrated conscious rats (fig. 7).

View larger version (K): [in this window] [in a new window]   Fig. 6.   Inhibitory effects of intravenous administration of YM087 on the pressor responses induced by AVP (), angiotensin II () and norepinephrine () in pithed rats. YM087 was given 5 min before the injection of AVP (30 mU/kg i.v.), angiotensin II (0.1 µg/kg i.v.) and norepinephrine (2 µg/kg i.v.). The resting diastolic blood pressure was 50.0 ± 0.81 mm Hg (n = 16), and the peak pressor response induced by each pressor agent was 129 ± 1.8 mm Hg (AVP), 102 ± 2.5 mm Hg (angiotensin II) and 119 ± 3.1 mm Hg (norepinephrine), respectively. Values represent the mean ± S.E.M. of five to six experiments.

View larger version (K): [in this window] [in a new window]   Fig. 7.   Urine volume and urine osmolality in dehydrated conscious rats over a 2-hr collection period after intravenous administration of YM087 and vehicle. Values are mean ± S.E.M. of five experiments. Statistical comparison by one-way analysis of variance followed by Dunnet's test. ** P < .01 compared with the vehicle values.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

We investigated the biochemical and pharmacological properties of YM087 with radioligand binding and biochemical/pharmacological techniques. We first evaluated the affinity of YM087 for AVP receptors in radioligand binding assay with use of rat liver (V_1A), anterior pituitary (V_1B) and kidney medulla (V₂) plasma membranes. YM087 concentration-dependently inhibited the [³H]AVP binding for V_1A and V₂ receptors with K_i values of 0.48 nM and 3.04 nM, respectively. The affinity of YM087 for V_1A receptors was 49 times higher than that of OPC-21268 (a selective V_1A antagonist), and for V₂ receptors was 14 times higher than that of OPC-31260 (a selective V₂ antagonist). These findings suggest that YM087 possesses potent and balanced binding affinity for both V_1A and V₂ receptors with K_i values similar to those of AVP. In addition, the selectivity of YM087 for V_1A and V₂ receptors relative to other receptors was demonstrated by its lack of activity in other radioligand binding assays. To determine whether YM087 interacts reversibly or irreversibly with AVP receptors, we investigated [³H]AVP saturation binding in the presence or absence of YM087 with rat liver and kidney medulla plasma membranes. In this experiment, YM087 concentration-dependently increased the K_d values of the radioligand for these receptors without changing the B_max values, which indicated that YM087 interacts reversibly and competitively with both receptors.

Several functional studies in vitro were performed to characterize the nature of the interaction of YM087 with V_1A and V₂ receptors. Recently, these V_1A and V₂ receptors were cloned, and it was indicated that cDNA of both receptor subtypes encodes proteins with seven putative transmembrane domains and a structure similar to rodopsin and other G-protein-coupled receptors (Thibonnier et al., 1994; Sugimoto et al., 1994). Furthermore, it was reported that the V_1A and V_1B receptors are coupled to phosphatidylinositol hydrolysis, which mobilizes intracellular Ca⁺⁺, and that the V₂ receptors are coupled to adenylate cyclase-cAMP production system (Birnbaumer et al., 1992).

A10 cells are an established smooth muscle cell line derived from rat embryo aorta. These cells express a V_1A receptor subtype which couples the activation of PLC (Kimes et al., 1976). The intracellular mechanisms underlying the vascular action of AVP via V_1A receptors have been explained by the activation of PLC and the subsequent mobilization of intracellular Ca⁺⁺ and activation of protein kinase C. In the present experiments with A10 cells, AVP increased [Ca⁺⁺]_i in a concentration-dependent manner with an EC₅₀ value of 5.29 nM. Under the same experimental conditions, YM087 potently antagonized the increase in [Ca⁺⁺]_i induced by AVP, with an IC₅₀ value of 1.16 nM. This is consistent with the K_i value of YM087 (0.48 nM for V_1A receptors) obtained from [³H]AVP-binding studies. In the absence of AVP, furthermore, YM087 did not increase [Ca⁺⁺]_i in A10 cells, which indicates a lack of agonistic activity for V_1A receptors.

LLC-PK₁ cells are a porcine-kidney-derived epithelial cell line which expresses 8-lysine vasopressin receptors of the V₂ type coupled to the activation of adenylate cyclase (Hull et al., 1976). This established cell line has been used extensively as a cellular model system in the study of the V₂ receptor adenylate cyclase system (Jans et al., 1989). In the present experiments with LLC-PK₁ cells, AVP concentration-dependently stimulated intracellular cAMP production with an EC₅₀ value of 18.6 nM. Under the same experimental conditions, YM087 concentration-dependently inhibited the production of cAMP induced by AVP, with an IC₅₀ value of 17.3 nM. In the absence of AVP, furthermore, YM087 did not stimulate cAMP production in LLC-PK₁ cells, which indicates that YM087 possesses no agonist activity for V₂ receptors. These results suggest that YM087 is a novel imidazobenzazepine-derived compound which possesses V_1A and V₂ dual antagonistic activities without agonistic activity for these receptors.

To evaluate the antagonistic effect of YM087 in vivo, we examined the effects of intravenous YM087 administration in rats. In pithed rats, intravenous administration of YM087 dose-dependently inhibited an AVP-induced increase in blood pressure with an ID₅₀ value of 0.013 mg/kg. In contrast, it did not alter pressor responses to angiotensin II and norepinephrine. On the other hand, in dehydrated conscious rats, intravenous administration of YM087 dose-dependently increased urine volume and decreased urine osmolality. These results suggest that YM087 exerts potent antagonistic effects on both V_1A and V₂ receptors in vivo.

AVP exerts a variety of biological effects such as vascular resistance control through V_1A receptors and regulation of water excretion through V₂ receptors. It is generally assumed that AVP is involved in several conditions such as heart failure, hyponatremia and hypertension through the V_1A and V₂ receptors (Fujisawa et al., 1993; Naitoh et al., 1994). The development of a nonpeptide AVP antagonist seems essential to the elucidation of the physiological and pathophysiological roles of AVP and to the investigation of the novel therapeutic approach to these diseases. Recently, several orally effective nonpeptide AVP antagonists have been reported, namely the V_1A selective antagonists OPC-21268 (Yamamura et al., 1991) and SR 49059 (Serradeil-Le Gal et al., 1993) and the V₂ selective antagonist OPC-31260 (Yamamura et al., 1992). However, these selective antagonists are not sufficient to inhibit the biological activities of AVP via the peripheral V_1A and V₂ receptors.

The vasoconstrictive effect of AVP is even more potent than that of angiotensin II. Plasma concentrations of AVP are elevated in patients with essential hypertension (Os et al., 1986). In the present study, YM087 completely inhibited the cellular effects of AVP in vascular smooth muscle cells, which suggests that YM087 may have a beneficial effect in the treatment of hypertension. The presence of dilutional hyponatremia has been observed to be a marker of severe decompensated CHF (Lee and Packer, 1986; Levine et al., 1982). The treatment of hyponatremia is particularly frustrating because saluretic diuretics such as furosemide tend to stimulate further the inappropriate release of AVP, leading to further retention of free water and aggravation of hypoosmolality. By increasing free water clearance with minimal saluretic effect, a V₂ antagonist potentially offers an additional therapeutic option for this condition. In this study, YM087 showed potent inhibition of V₂ receptor-mediated effects and would cause a concurrent decrease in preload because of profuse diuresis. The hemodynamic consequences of this contraction in fluid volume may offer an added benefit to CHF patients. That is, CHF is characterized hemodynamically by a raised total peripheral resistance, which helps to maintain blood pressure, sometimes at the expense of cardiac output. This increase in total peripheral resistance is caused by activation of several vasoactive neurohumoral factors, including the renin-angiotensin-aldosterone system, the sympathetic nervous system and AVP, which seem to participate in this process and contribute to the hemodynamic and metabolic alterations characteristic of CHF (Creager et al., 1986). Elevated levels of AVP have been reported in experimental models of heart failure (Kim et al., 1990) and in patients with decompensated heart failure (Szatalowicz et al., 1981; Goldsmith et al., 1986). Nicod et al. (1985) have reported that the V_1A antagonist (d(CH₂)₅Tyr(Me)AVP) produces hemodynamic improvement with a transient decrease in systemic vascular resistance and an increase in cardiac output in CHF patients in whom plasma AVP levels were elevated. Mulinari et al. (1990) have shown that administration of the dual V_1A/V₂ antagonist (d(CH₂)₅-D-Tyr(Et)VAVP) in rats with markedly impaired left ventricular function produces significant hemodynamic changes and profound water diuresis, without much change in systemic blood pressure. The activation of a series of neural and hormonal mechanisms in heart failure produces the fluid retention and vasoconstriction which dominates the syndrome. So important are these neurohumoral mechanisms that therapeutic interventions directed at the compensatory changes have proved far more successful in the long term in improving cardiac function. AVP has potent dual vasoconstrictor and fluid-retaining actions and may therefore play an important role in the pathophysiology of heart failure.

In summary, the present results in vitro receptor binding and second messenger assays indicate that YM087 is a dual AVP V_1A and V₂ receptor antagonist with high affinity and potency. YM087 should therefore prove to be a novel and valuable tool that can be used to define the roles of AVP in various physiological and pathophysiological processes. To the extent that AVP is involved in the etiology and maintenance of various diseases such as heart failure, hyponatremia and the syndrome of inappropriate antidiuretic hormone secretion, YM087 may be a valuable therapeutic agent in the treatment of these chronic disorders.