A comparison of an A₁ adenosine receptor agonist (CVT-510) with diltiazem for slowing of AV nodal conduction in guinea-pig

Abstract

1. The purpose of this study was to compare the pharmacological properties (i.e. the AV nodal depressant, vasodilator, and inotropic effects) of two AV nodal blocking agents belonging to different drug classes; a novel A₁ adenosine receptor (A₁ receptor) agonist, N-(3(R)-tetrahydrofuranyl)-6-aminopurine riboside (CVT-510), and the prototypical calcium channel blocker diltiazem.

2. In the atrial-paced isolated heart, CVT-510 was approximately 5 fold more potent to prolong the stimulus-to-His bundle (S–H interval), a measure of slowing AV nodal conduction (EC₅₀=41 nM) than to increase coronary conductance (EC₅₀=200 nM). At concentrations of CVT-510 (40 nM) and diltiazem (1 μM) that caused equal prolongation of S–H interval (∼10 ms), diltiazem, but not CVT-510, significantly reduced left ventricular developed pressure (LVP) and markedly increased coronary conductance. CVT-510 shortened atrial (EC₅₀=73 nM) but not the ventricular monophasic action potentials (MAP).

3. In atrial-paced anaesthetized guinea-pigs, intravenous infusions of CVT-510 and diltiazem caused nearly equal prolongations of P–R interval. However, diltiazem, but not CVT-510, significantly reduced mean arterial blood pressure.

4. Both CVT-510 and diltiazem prolonged S–H interval, i.e., slowed AV nodal conduction. However, the A₁ receptor-selective agonist CVT-510 did so without causing the negative inotropic, vasodilator, and hypotensive effects associated with diltiazem. Because CVT-510 did not affect the ventricular action potential, it is unlikely that this agonist will have a proarrythmic action in ventricular myocardium.

Introduction

Pharmacological and molecular cloning studies have led to the identification of four distinct cell surface adenosine receptor subtypes: A₁, A_2A, A_2B, and A₃ (Shryock & Belardinelli, 1997). In the heart, mRNAs for all four adenosine receptor subtypes have been detected using the sensitive reverse transcription–polymerase chain reaction, (RT–PCR, Dixon et al., 1996). The mRNA levels for the A_2B and A₃ receptors are very low, whereas A₁ and A_2A mRNA levels are somewhat higher (Dixon et al., 1996). The localization and/or function(s) of A_2B and A₃ receptors in the heart remain to be elucidated. In comparison, it is well established that a number of important cardiac functions are modulated by the A₁ and A_2A receptor subtypes (Shryock & Belardinelli, 1997; Belardinelli et al., 1995). The A₁ adenosine receptor (A₁ receptor) mediates the actions of adenosine to (1) slow heart rate (negative chronotropic effect), (2) slow impulse conduction through the AV node (negative dromotropic effect), (3) depress atrial, but not ventricular contractility (negative inotropic effect), and (4) attenuate the cardiac stimulatory effects of catecholamines (anti-β-adrenergic effect). The A_2A adenosine receptor (A_2AAdoR) mediates the coronary vasodilation caused by adenosine (Poucher et al., 1995; Belardinelli et al., 1998).

The A₁ receptor-mediated negative dromotropic effect of adenosine is the basis for the use of this nucleoside for the treatment of paroxysmal supraventricular tachycardias (PSVT, Belardinelli & Lerman, 1990; Lerman & Belardinelli, 1991). On the other hand, the A_2A receptor-mediated coronary vasodilator effect of adenosine led to its use as a pharmacological coronary vasodilator in combination with myocardial radionuclide imaging for the detection of ischemic heart disease (Verani et al., 1990; Iskandrian et al., 1991). These and a number of other potential beneficial actions of adenosine mediated by A₁ and A_2A receptors are in part responsible for the considerable effort devoted by both academia and the pharmaceutical industry to develop potent and selective agonists for these two receptor subtypes.

CVT-510, N-(3(R)-tetrahydrofuranyl)-6-aminopurine riboside (Figure 1), was specifically developed to exploit the A₁ receptor-mediated negative dromotropic effect for the purposes of treating AV nodal re-entrant arrhythmias. In the present report, we describe the pharmacological properties of CVT-510, a new, potent, and selective agonist of the A₁ receptor subtype and compare them to that of the calcium channel blocker diltiazem.

Figure 1.

Chemical structure N-(3(R)-tetrahydrofuranyl)-6-aminopurine riboside (CVT-510).

Methods

Isolated perfused heart preparation

Guinea-pigs (n=46) of either sex weighing 250–300 g were anaesthetized with methoxyflurane and killed by cervical dislocation. The hearts were quickly removed and rinsed in ice-cold Krebs-Henseleit solution. The aorta was cannulated and the heart retrogradely perfused at a constant flow of 10 ml min⁻¹ with modified Krebs-Henseleit solution containing (mM): NaCl 117.9, KCl 4.8, CaCl₂ 2.5, MgSO₄ 1.18, KH₂PO₄ 1.2, Na₂EDTA 0.5, ascorbic acid 0.14, glucose 5.5, pyruvic acid (sodium salt) 2.0, and NaHCO₃ 25. The K-H solution was oxygenated with 95% oxygen and 5% CO₂, pH 7.4, and temperature maintained at 36°C.

Hearts were electrically paced at a cycle length of 290–300 ms (unless otherwise indicated) using a bipolar electrode placed on the right atrium or ventricle. The stimulator, an interval generator (Model 1830, WPI, Sarasota, FL, U.S.A.) delivered stimuli through a stimulus isolation unit (Model 1880, WPI) as square wave pulses of 3 ms in duration and at least twice the threshold intensity. The stimulus-to-His bundle (S–H) interval was used as index of AV nodal conduction time and was measured on-line during constant atrial pacing from the His bundle electrogram using a Gateway 4 DX-33  MHz microcomputer with a DT-208 1A interface board (Axon Instruments, Burlingame, CA, U.S.A.). In some experiments, for convenience, the stimulus-to ventricle (S–V) interval was measured. Because adenosine and A₁ receptor agonists do not prolong His-bundle-to-ventricle (H–V) interval, prolongation of S–V interval accurately reflects changes in atrium-to-His-bundle (A–H interval), i.e., proximal AV nodal conduction time.

Coronary perfusion pressure was measured using a pressure transducer that was connected to the aortic cannula via a T-connector. Coronary perfusion pressure (in mmHg) was monitored throughout an experiment and recorded on a 4-channel Gould strip-chart recorder (Gould RS3400, Cleveland, OH, U.S.A.). Coronary conductance (in ml min⁻¹ mmHg⁻¹) was calculated as the ratio of coronary perfusion rate (10 ml min⁻¹) and perfusion pressure (mmHg). Isovolumetric ventricular pressure (LVP) was measured using a saline-filled latex balloon placed in the left ventricle and connected to a pressure transducer (Gould P231D). The volume of the balloon was adjusted and held constant to maintain an LV diastolic pressure of 4–8 mmHg. Pressures were recorded on a 4-channel strip chart recorder. The dP dt⁻¹_max was measured offline by using the Snapshot storage program (Version 3.21D, HEM Data Corp., Southfield, MI, U.S.A.) and taken as an index of left ventricular contractility.

Monophasic action potentials (MAP) were recorded using a pressure contact silver-silver chloride electrode (Langendorff Probe, EP Technologies, Inc., Sunnyvale, CA, U.S.A.) placed on the surface of the left antrium and/or left ventricle. The signals were amplified and filtered by an isolated biological amplifier (IsoDam, WPI) and displayed in real time on a digital oscilloscope (2201 Tektronix, Inc., Beaverton, OR, U.S.A.). Signals were considered adequate if they were stable for at least 5 min and the amplitudes of the MAPs exceeded 10 mV. The data were digitized at 2 kHz by a DT-2801A digitizing board (Data Translation, Marlboro, MA, U.S.A.) and stored using the Snapshot data acquisition program (Snapshot Storage Scope, HEM Data Corp., Southfield, MI, U.S.A.) for later analysis. The durations of the atrial and ventricular monophasic action potentials were measured at 90% repolarization (MAPD₉₀), using the Snapshot program. After completion of dissection and instrumentation, the hearts were allowed to equilibrate for 30 min before the experiments were begun. Whenever the baseline and post-intervention (washout) values differed by more than 15%, the data were discarded. Approximately 60–70% of the experiments conformed with this criterion.

Anaesthetized guinea-pig preparation

Guinea-pigs (n=25) of either sex weighing 300–350 g were anaesthetized with 0.75% isoflurane administered through a nose cone, with oxygen making up the balance. After a 20 min induction, a transverse incision was made in the neck region to expose the trachea, jugular veins, and carotid arteries. The trachea was catheterized using a 14-gauge catheter and the animal ventilated with 0.75% isoflurane/oxygen at 60 strokes min⁻¹ and a tidal volume of 2.5 ml using a Harvard Apparatus (Southnatik, MA, U.S.A.) rodent ventilator. In previous experiments isoflurane at the concentration used in the present study was found to provide full anaesthesia, assessed by foot pinching and touching of the cornea, without requiring the use of muscle relaxant or any other anaesthetic agent. The left jugular vein and carotid artery were catheterized with 24-gauge catheters for administration of drugs and for recording of arterial blood pressure, respectively. In experiments involving electrical pacing of the atrium or ventricle, the right jugular vein was catheterized with a custom-made electrocatheter. ECG leads were attached to record in the L-2 configuration. A non-depolarizing neuromuscular blocker, vecuronium bromide (0.5 mg), was administered intravenously (i.v.), after all surgical procedures and instrumentation were completed, to prevent muscular fasciculation from interfering with the ECG recording. Anaesthetic depth was determined by observing autonomic responses such as heart rate and blood pressure. The L-2 of the ECG and arterial blood pressure were monitored throughout the experiments and recorded on a 2-channel Gould strip-chart recorder. After completion of surgery and instrumentation, a 20-min equilibration period was allowed before the experimental protocols were commenced. During the surgical and experimental procedure, the animal temperature was maintained at 37°C using a warm water blanket. Identical to the isolated heart experiments, whenever the baseline and post-intervention (washout) values differed by more than 15%, the data were discarded. Approximately 75% of this series of experiments met this criterion.

All experimental procedures and protocols for handling animals were approved by the University of Florida Institutional Animal Care and Use Committee, and conform with NIH guidelines for the use of animals.

Binding studies

Membrane preparation

Porcine brain (forebrain and striatum) were minced and then homogenized in ice cold Tris-HCl buffer (50 mM), pH 7.4. Homogenates were filtered through cotton gauze and centrifuged at 48,000×g for 15 min. The membrane pellets were washed twice by resuspension in fresh buffer and centrifugation. Final pellets were resuspended in Tris-HCl buffer (50 mM), pH 7.4, and frozen at −80°C until use.

Binding assays

The effect of CVT-510 on binding to A₁ and A_2A-adenosine receptors of porcine forebrain and striatum membranes, respectively, were determined. Assays for A₁ and A_2A receptors were carried out by using the A₁ receptor antagonist [³H]CPX and the A_2A receptor agonist [³H]CGS 21680. Membranes were treated with adenosine deaminase (2 U ml⁻¹) for 20 min at room temperature prior to and during radioligand binding assays. Membranes (0.2–0.7 mg), adenosine deaminase, and the indicated radioligand were incubated for 3 h in a 300 μl volume of Tris-HCl buffer (50 mM) (pH 7.4). Assays were carried out in triplicate at room temperature. After the incubation period, bound and free radioligand were diluted by the addition of ice-cold Tris-HCl buffer (5 ml), and immediately separated by vacuum filtration of assay contents onto Whatman GF/C filters and washing of trapped membranes with Tris-HCl buffer (20 ml). Filter disks containing membrane-bound radioactivity were placed in 4 ml Scintiverse (Fisher Scientific, Pittsburgh, PA, U.S.A.), and the radioactivity was quantified by a liquid scintillation counter. Specific binding of [³H]CPX and [³H]CGS 21680 was defined as membrane binding displaced in the presence of CPT (10 μM) and R-PIA (10 μM), respectively.

The apparent potencies of CVT-510 to compete with the specific binding of [³H]CPX (1 nM) to forebrain membranes, and of [³H]CGS 21680 (2 nM) to striatal membranes were determined. Membranes were incubated with the indicated concentrations of radioligand and increasing concentrations of the A₁ receptor agonist CVT-510 for 3 h in a total volume of 300 μl. Assays were terminated as described above. Binding parameters (e.g., IC₅₀, K_i, Hill Coefficients) from the competition assays were determined with the computer program LIGAND (Biosoft, Cambridge, U.K.). The K_i for displacement of each labelled ligand by CVT-510 was calculated using the Cheng/Prusoff equation: K_i, Agonist= IC₅₀, Agonist/(1+[Ligand]/K_d, ligand).

Experimental protocols

Negative dromotropic and coronary vasodilator effects of CVT-510 and diltiazem in the guinea-pig isolated heart

The effects of progressively higher concentrations of CVT-510 (0.5–3000 nM) and diltiazem (0.001–30 μM) to prolong the S–H interval (A₁ receptor-mediated effect), and to increase coronary conductance (A_2A receptor-mediated effect) were determined. The data were used to obtain concentration-response relationships for the S–H interval prolongations and for the increases in coronary conductance caused by CVT-510 and diltiazem. In these hearts, coronary perfusion pressure and S–H intervals were measured simultaneously.

In a separate series of experiments, the effects of the A₁ receptor antagonist CPX (60 nM) and of the A_2A receptor antagonist ZM241,385 (100 nM) on the S–H interval prolongation and coronary vasodilator effects of CVT-510 were determined. Hearts were exposed to single concentrations of CVT-510 for up to 30 min. Once steady-state effects to CVT-510 were achieved (5–10 min after onset of infusion), either CPX or ZM241,385 was infused into the perfusate for 15–20 min, and the effects of the antagonists on the responses of the hearts to CVT-510 recorded. This was followed by washout (40–60 min) of all drugs, and the post-intervention measurements were made.

The effects of CVT-510 and diltiazem on left ventricular contractility

Isolated perfused hearts paced at a constant atrial cycle length of 300 ms were instrumented for continuous recording of LVP, dP dt⁻¹_max, S–V interval (i.e., negative dromotropic effect) and coronary perfusion pressure. After an equilibration period of 30 min, control measurements were obtained. CVT-510 or diltiazem was infused at rates (based on concentration-response curves) that caused equal prolongations of the S–V interval (perfusate concentrations of 40 nM and 1 μM, respectively). The steady-state effects of CVT-510 and diltiazem on LVP, dP dt⁻¹_max, and coronary conductance were determined.

Effect of CVT-510 on atrial and ventricular monophasic action potential (MAP)

Isolated perfused hearts were instrumented for simultaneous recording of atrial and ventricular MAP. Concentration-response relationships for CVT-510 to shorten atrial and ventricular MAP were determined. The protocol was similar to that described above for negative dromotropic effect and increase in coronary conductance of CVT-510. After a 30-min equilibration period, control measurements were made of atrial and ventricular MAPD₉₀. CVT-510 was infused into the perfusion line to achieve the desired perfusate concentrations (1–3000 nM). The effect of each concentration of CVT-510 on atrial and ventricular MAPD₉₀ was measured once the response had reached steady state (5–10 min after onset of infusion).

Prolongation of P–R interval in response to i.v. boluses of CVT-510 or adenosine in anaesthetized guinea-pigs

After completion of surgery and instrumentation, a 20-min equilibration time was allowed before starting the administration of CVT-510 (0.1–1.5 mγ kg⁻¹) or adenosine (25–100 μg kg⁻¹). The ECG (L-2) was monitored throughout the experiments. After obtaining a baseline recording of the ECG, boluses of CVT-510 or adenosine were injected into the left jugular vein and followed by saline flushes. Increasing doses of CVT-510 and adenosine were given 15 and 5 min apart, respectively. The longest P–R intervals (peak effect) were determined after each bolus injection of agonist.

Effect of CVT-510 and NECA on mean arterial blood pressure in ventricle-paced anaesthetized guinea-pigs

After post-surgical equilibration and recording of baseline ECG and mean arterial blood pressure (MABP), each animal in this series received an i.v. infusion of either CVT-510 (5.3 μg kg⁻¹ min⁻¹) or NECA (2.6 μg kg⁻¹ min⁻¹). Once a steady-state 2 : 1 AV block was achieved, the right ventricle was intermittently paced at a rate approximately equal to the intrinsic baseline atrial rate using an electrocatheter positioned in the right ventricle. Intermittent pacing was used to demonstrate the reproducibility of the effect of ventricular pacing on blood pressure during the 2 : 1 A-V block caused by either CVT-510 or NECA. Ventricular rates and arterial blood pressures were monitored throughout an experiment.

Effect of continuous infusions of CVT-510 and diltiazem on P–R interval, heart rate, and mean arterial blood pressure in anaesthetized guinea-pigs

After post-surgical equilibration and recording of baseline ECG and MABP, each animal in this series received an i.v. infusion of either CVT-510 (0.75–1.0 μg kg⁻¹ min⁻¹) or diltiazem (200–400 μg kg⁻¹ min⁻¹). For both CVT-510 and diltiazem, a rate of infusion was chosen that did not cause second-degree AV block, but was sufficient to cause a significant prolongation of the P–R interval. In this series of experiments, hearts were allowed to beat spontaneously.

Data analysis

All values are reported as means±s.e.mean. Significance of differences among values in experiments with multiple treatment groups was determined by analysis of variance followed by Student-Newman-Keuls testing. A value of P<0.05 was considered to indicate a statistically significant difference. Coronary conductance and MAP concentration-response data were fitted to a 4-parameter logistic dose response equation by use of nonlinear regression (Table Curve, Jandel Scientific Software, Corp., Chicago, IL, U.S.A.): y=a+b/(1+[(x/c)^d], where y, x, a, b, c, and d denote level of response, concentration of agonist, extrapolated minimum response, extrapolated maximum response, EC₅₀ value, and the apparent Hill coefficient, respectively. Concentration-response data for S–H and P–R interval prolongation were analysed as previously described (Dennis et al., 1992). The value of the longest stable S–H or P–R interval immediately prior to the onset of AV block was considered the maximum negative dromotropic effect and used for calculation of EC₅₀ values. For all variables (e.g., coronary conductances, S–H interval prolongation, etc.), the EC₅₀ values were obtained for each individual concentration-response curve, and were used for subsequent statistical analysis.

Chemicals

6-(N-3′-(R) tetrahydrofuranyl)-amino-purine ribose adenosine (CVT-510) was provided by CV Therapeutics (Palo Alto, CA, U.S.A.). 8-cyclopentyl-1,3-dipropylxanthine (CPX), 5′-N-ethylcarboxamidoadenosine (NECA), and diltiazem were purchased from Research Biochemicals (Natick, MA, U.S.A.). ZM241,385 (4-(2-[7-amino-2-(2-furyl[1,2,4]-triazolo[2,3-a[1,3, 5]triazin-5-yl-aminoethyl)phenol) was a gift from Dr Simon M. Poucher (Zeneca Pharmaceuticals, Cheshire, U.K.). Stock solutions of CVT-510, CPX, NECA, and ZM241,385 were prepared in dimethylsulfoxide (DMSO, Sigma Chemical Company, St. Louis, MO, U.S.A.) and diluted with saline immediately prior to experiments. Diltiazem was dissolved in saline. Isoflurane was puchased from Ohmeda Caribe, Inc. (Guayama, PR, U.S.A.), and vecuronium bromide from Marsam Pharmaceuticals (Cherry Hill, NJ, U.S.A.).

Results

Negative dromotropic and coronary vasodilatory effect of CVT-510 and diltiazem

In isolated perfused hearts (n=4–6) paced at a constant atrial cycle length of 290–300 ms, CVT-510 and diltiazem caused a concentration-dependent increase of the S–H interval and the coronary conductance (Figures 2 and 3, respectively). The potencies for CVT-510 and diltiazem to prolong the S–H interval and to increase coronary conductance are summarized in Table 1. CVT-510 was ∼5 fold more potent to prolong the S–H interval than to increase coronary conductance (Figure 2, Table 1). In comparison, the calcium channel blocker diltiazem showed very little selectivity between prolonging the S–H interval and increasing coronary conductance (Figure 3, Table 1). In addition, as illustrated in Figure 4, the prolongation of S–H interval ([CVT-510]=60 nM), but not the increase in coronary conductance ([CVT-510]=1 μM), was completely blocked by the A₁ receptor-selective antagonist CPX (60 nM), whereas the increase in coronary conductance caused by CVT-510, but not the prolongation of S–H interval, was completely blocked by ZM241,385 an A_2A receptor-selective antagonist (n=4).

Figure 2.

Concentration-response relationship for the A₁ receptor-mediated negative dromotropic (S–H interval prolongation) and the A_2A receptor-mediated coronary vasodilator (increase of coronary conductance) effects of CVT-510 in guinea-pig isolated perfused hearts. Symbols and error bars indicate means±s.e.mean of single determinations from each of four hearts. Second-degree AV block occurred in all hearts at a concentration of CVT-510⩾40 nM.

Figure 3.

Concentration-response relationship for the calcium channel blocker diltiazem to prolong the S–H interval and increase coronary conductance in guinea-pig isolated perfused hearts. Symbols and error bars indicate means±s.e.mean of single determinations from each of 4–6 hearts.

Table 1.

The potencies of CVT-510 and diltiazem to prolong the S–H interval and increase coronary conductance

Figure 4.

The antagonism of the A₁-receptor-mediated increase in S–H interval (A) and the A_2A receptor-mediated increase in coronary conductance (B) by the A₁ receptor-selective antagonist CPX (60 nM) and by the A_2A receptor-selective antagonist ZM241,385 (100 nM), respectively, during infusion of CVT-510 (60 nM) in the guinea-pig isolated perfused heart. Bars represent the means±s.e.mean of single determinations from each of four hearts. ^*Denotes a significant difference from control (P<0.05).

The effects of CVT-510 and diltiazem on left ventricular contractility

In this series of experiments (n=4), the effects of CVT-510 on left ventricular contractility and coronary artery conductance were determined and compared to those of diltiazem. Isolated perfused hearts were exposed to concentrations of CVT-510 (40 nM) or diltiazem (1 μM) that caused near equal prolongations of stimulus to ventricle (S–V) interval. Both CVT-510 and diltiazem significantly prolonged the S–V interval (i.e. negative dromotropic effect), but only diltiazem reduced LVP and dP dt⁻¹_max (i.e. negative inotropic effect). Figure 5 summarizes the results of four such experiments. CVT-510 (40 nM) significantly increased the S–V interval from a control of 47±2 to 59±2 ms but did not affect left ventricular pressure, (LVP), dP dt⁻¹_max or coronary conductance (Figure 5A and B). In the same hearts, diltiazem (1 μM) prolonged the S–V interval by 12±2 ms, an increase not significantly different from that caused by CVT-510 (11±1 ms). However, in contrast to CVT-510, the increase in S–V interval caused by diltiazem was accompanied by a 45±3% reduction in LVP and an increase in coronary conductance of 36±3% (Figure 5A and B). Thus, for the same magnitude of slowing of AV nodal conduction, CVT-510 had no effect on left ventricular contractility or coronary conductance whereas diltiazem significantly depressed left ventricular contractility and increased coronary conductance.

Figure 5.

Comparison of the effects of CVT-510 (40 nM) and diltiazem (1 μM) on left ventricular developed pressure (LVP, A) and coronary conductance (B) in guinea-pig isolated perfused heart for the same prolongation of the S–V interval (∼10 ms). Bars represent the means±s.e.mean of single determinations from each of four hearts. Numbers above bars are means±s.e.mean of dP dt⁻¹_max for each data set. ^*Denotes significant differences from both 2 : 1 A–V Block and Washout (P<0.05).

Shortening of atrial but not ventricular monophasic action potential duration by CVT-510

This series of experiments was carried out to determine the effects of CVT-510 on the atrial and ventricular MAP of guinea-pig isolated hearts (n=11 and 5 for atrial and ventricular MAP, respectively). The baseline atrial and ventricular MAPD₉₀ values were 48±3 and 155±2 ms, respectively. CVT-510 shortened the atrial MAP in a concentration-dependent manner (Figure 6), but had no effect on ventricular MAP. CVT-510 maximally shortened the atrial MAP to 21±1 ms. The concentrations that caused 50% (EC₅₀ value) and maximal shortening (E_max value) of the atrial MAP were 126 and 300 nM, respectively. In contrast, concentrations as high as 300 nM caused no shortening of the ventricular MAP.

Figure 6.

Differential effect of CVT-510 on atrial and ventricular monophasic action potential durations (MAPD). CVT-510 (1–300 nM) shortened the atrial MAPD in a concentration-dependent manner but had no effect on the ventricular MAPD. Values are means±s.e.mean of each of 5–11 hearts. Atrial and ventricular MAPD were measured at 90% of repolarization (MAPD₉₀).

Prolongation of P–R interval in response to i.v. boluses of CVT-510 or adenosine in anaesthetized guinea-pigs

In anaesthetized closed-chest guinea-pigs atrial paced at a rate of 330–360 beats min⁻¹, i.v. boluses of CVT-510 or adenosine caused dose-dependent prolongations of the P–R interval (Figure 7). The potencies (EC₅₀ values) for CVT-510 and adenosine to prolong the P–R interval were 0.9±0.1 (n=3) and 75±4 (n=4) μg kg⁻¹, respectively. The longest P–R interval recorded before second-degree AV block occurred was 105±6 ms for CVT-510, and 115±3 ms for adenosine.

Figure 7.

Dose-response relationship for the effect of intravenous injections of CVT-510 or adenosine to prolong the P–R interval in atrial-paced (330–360 beats min⁻¹) anaesthetized guinea-pigs. Symbols and error bars represent means±s.e.mean of single determinations from each of 3–4 animals.

Effect of CVT-510 and NECA on mean arterial blood pressure

To determine if the decrease in mean arterial blood pressure (MABP) observed during second-degree AV block caused by CVT-510 is due to peripheral vasodilation or due to its negative dromotropic action, the effect of CVT-510 on MABP was measured in anaesthetized guinea-pigs during intermittent right ventricular pacing. In these experiments, hearts were allowed to beat spontaneously or were subjected to ventricular pacing at the same rate as the control atrial rate to overcome the slowing of ventricular rate caused by the second-degree AV block. As illustrated in Figure 8, a continuous intravenous infusion of CVT-510 (5.3 μg kg⁻¹ min⁻¹) caused second-degree AV block and a marked reduction in ventricular rate which was accompanied by a decrease in MABP. During continuous infusion of CVT-510, immediately upon initiation of right ventricular pacing, the MABP returned to control levels. Upon cessation of ventricular pacing, the MABP markedly fell again. Similar results were obtained in three additional experiments, the data for which are summarized in Figure 9. CVT-510 (5.3 μg kg⁻¹ min⁻¹) caused a stable 2 : 1 AV block that resulted in a reduction of ventricular rate from a control of 302±6 to 102±4 beats min⁻¹ (Figure 9A). The decrease in ventricular rate was accompanied by a decrease in MABP from a control of 52±3 to 30±2 mmHg. A value of 53.1±4.2 mmHg for MABP of resting, awake guinea-pigs was reported by Brown et al. (1989). Thus, the control values of MABP recorded in the present study are within the normal range for this species. Regardless, ventricular pacing completely reversed the decrease in MABP, raising it to 58±3 mmHg. In contrast, the hypotension observed during second-degree AV block caused by the nonselective adenosine receptor agonist NECA was not fully reversed by ventricular pacing. NECA (2.6 μg kg⁻¹ min⁻¹) caused a decrease in the ventricular rate from a control of 303±3 to 86±5 beats min⁻¹ (Figure 9B). The decrease in ventricular rate was accompanied by a decrease in MABP from a control of 71±3 to 32±2 mmHg. In contrast to CVT-510, ventricular pacing failed to reverse the decrease in MABP caused by NECA.

Figure 8.

Representative electrocardiogram (ECG lead 2) and mean arterial blood pressure (MABP) from an anaesthetized guinea-pig heart recorded before (control, t=0), during, and after (washout, t=35 min) a continuous intravenous infusion of CVT-510. The heart was allowed to either beat spontaneously (‘Spontaneous') or was subjected to right ventricular pacing (V-pacing), at the same rate as the control atrial rate. At 10 min of infusion of CVT-510 (t=10 min), a stable 2 : 1 atrioventricular (AV) block occurred and this was associated with a decrease in MABP. Subsequently, (t=12 min and onward) during intermittent V-pacing, the MABP rose to control levels.

Figure 9.

(A) Complete reversal by ventricular pacing of hypotension associated with 2 : 1 AV block caused by CVT-510 in anaesthetized guinea-pigs. Summary of data showing that 2 : 1 AV block caused by a continuous intravenous infusion of CVT-510 is associated with a significant reduction in mean arterial blood pressure (MABP). Pacing of the right ventricle (Ventricular Pacing) at the same rate as the control atrial rate, reversed the decrease in MABP observed during CVT-510-induced 2 : 1 AV block. At 30–40 min after stopping the infusion of CVT-510 (Washout), MABP was not different from baseline. Values are means±s.e.mean of single determinations from each of four guinea-pigs. (B) Partial reversal by ventricular pacing of hypotension associated with 2 : 1 AV block caused by the nonselective adenosine receptor agonist NECA in anaesthetized guinea-pigs. Summary of data showing that 2 : 1 AV block caused by continuous intravenous infusion of NECA is associated with significant reduction of mean arterial blood pressure (MABP). Pacing of the right ventricle (Ventricular Pacing) at the same rate as the control atrial rate only partially reversed the decrease in MABP observed during NECA-induced 2 : 1 AV block. At 30–40 min after stopping the infusion of NECA (Washout), MABP returned to baseline. Values are means±s.e.mean of single determinations from each of three guinea-pigs. ^*Denotes significant differences from both 2 : 1 A-V Block and Washout (P<0.05).

Effects of CVT-510 and diltiazem on P–R interval, heart rate, and arterial blood pressure

In this series of experiments in anaesthetized guinea-pigs (n=4), both CVT-510 (0.75–1 μg kg⁻¹ min⁻¹) and diltiazem (200–400 μg kg⁻¹ min⁻¹) slowed the heart rate from a control of 290±9 to 193±6 and 185±9 beats min⁻¹, respectively, and prolonged the P–R interval from a control of 66±3 to 94±2 and 84±2 ms, respectively. However, for similar prolongations of P–R interval, diltiazem, but not CVT-510, caused a significant decrease in MABP (Figure 10). Whereas CVT-510 tended to slightly elevate MABP, diltiazem reduced MABP from 50±1 to 27±4 mmHg (Figure 10).

Figure 10.

Summary of data comparing the effects of CVT-510 (0.75–1.0 μg kg⁻¹ min⁻¹) and diltiazem (200–400 μg kg⁻¹ min⁻¹) given as intravenous infusions on P–R interval, heart rate, and mean arterial blood pressure (MABP) in anaesthetized guinea-pigs. Hearts were allowed to beat spontaneously. Control values are given in the results section. Bars represent means±s.e.mean of changes from baseline (horizontal line at 100%) of single determinations from each of six hearts (three for CVT-510, three for diltiazem).

Binding of CVT-510 to A₁- and A_2A-adenosine receptors

Competition by CVT-510 of both [³H]CPX (1 nM) and [³H]CGS21,680 (4 nM) specific binding to pig forebrain (A₁ receptor) and striatum (A_2A receptor), respectively, was concentration-dependent and complete. However, CVT-510 was 356 fold more potent in competing with the binding of the A₁ receptor radioligand [³H]CPX than with the binding of the A_2A receptor radioligand [³H]CGS21,680. The binding affinities (K_i, dissociation constant) for CVT-510 were 6.5 nM and 2315 nM for the forebrain A₁ receptor and striatum A_2A receptor, respectively (n=2).

Discussion

The most important finding of this study is that CVT-510 is a potent agonist of the A₁ adenosine receptor with sufficient functional selectivity to cause slowing of AV nodal conduction (i.e. negative dromotropic effect) at concentrations that do not cause significant coronary vasodilation, depress left ventricular function (i.e. negative inotropic effect), or cause systemic hypotension. In this report, the selectivity of CVT-510 was shown both at the receptor (binding studies) and at the functional level (guinea-pig isolated hearts and anaesthetized guinea-pigs). In guinea-pig isolated perfused hearts, CVT-510 caused vasodilation at concentration 5 fold greater than those required to cause slowing of AV nodal conduction (Figure 2). In anaesthetized guinea-pigs, CVT-150 had no effect on mean arterial blood pressure (MABP) when given at a dose that caused first degree AV block (Figure 10) or even at a dose that caused second degree AV block (Figures 8 and 9). On the other hand, the calcium channel blocker diltiazem, at concentrations that cause equal negative dromotropic effects as CVT-510, significantly increased coronary conductance and decreased left ventricular pressure in the isolated heart (Figure 5), and decreased MABP in anaesthetized guinea-pigs (Figure 10). Thus, in comparison to diltiazem, CVT-510, and presumably other A₁AdoR-selective agonists, may offer advantages over calcium channel blockers in the treatment of supraventricular tachycardias (SVT) in which the AV node is part of the re-entrant circuit.

Selectivity of CVT-510

The actions of CVT-510 to slow AV nodal conduction and cause coronary vasdodilation (albeit at higher concentrations) are most likely mediated by A₁ and A_2A receptors, respectively. This conclusion is based on the following findings: (1) the negative dromotropic effect of CVT-510 is attenuated by the A₁ receptor-selective antagonist CPX, but not by the A_2A receptor-selective antagonist ZM241,385; (2) the coronary vasodilation caused by CVT-510 is attenuated by the A_2A receptor-selective antagonist ZM241,385, but not by the A₁ receptor-selective antagonist CPX; (3) CVT-510 competed with [³H]CPX binding to pig forebrain membranes; and (4) the binding of the A_2A receptor agonist radioligand [³H]CGS21680 to pig striatum was competitively antagonized by CVT-510.

The role of receptor reserve in determining functional selectivity of CVT-510

CVT-510 was found to have a 356 fold greater affinity for A₁ receptors (pig forebrain) than for A_2A receptors (pig striatum). Yet in functional assays (i.e., guinea-pig isolated perfused hearts), CVT-510 was approximately 5 fold more potent in prolonging the S–H interval (A₁ receptor response) than for increasing coronary conductance (A_2A receptor response). The difference between binding and functional selectivity can be explained by the presence of a large receptor reserve for A_2A receptor-mediated coronary vasodilation relative to the receptor reserve present for A₁ receptor-mediated effects (Shryock et al., 1998; Morey et al., 1998). For example, for the adenosine receptor agonist R-PIA to cause half-maximal shortening of atrial action potential duration (A₁ receptor-response, Morey et al., 1998), and to increase coronary conductance (A_2A receptor-response, unpublished) the fractional receptor occupancies were 30 and 10%, respectively. That is, there was a greater receptor reserve for R-PIA to cause vasodilation than to shorten the atrial action potential. However, it should be noted that the magnitudes of receptor reserves for A₁ receptor- and A_2A receptor-mediated responses are both agonist and tissue-dependent. Therefore, whether the difference in receptor reserves for CVT-510 to cause coronary vasodilation and slow AV nodal conduction is sufficient to explain the difference between binding and functional selectivity remains to be determined.

Effect of CVT-510 on atrial and ventricular action potential durations

The A₁ receptor-mediated cellular electrophysiological and contractile effects of adenosine and A₁ receptor agonists in atrial and ventricular myocytes (Belardinelli et al., 1995) can fully account for the effects of CVT-510 presented in this report.

It is well established that adenosine activates the inward-rectifying potassium current, I_KAdo, in guinea-pig atrial but not ventricular myocytes (Belardinelli et al., 1995; Song & Belardinelli, 1996). This differential effect of adenosine on atrial and ventricular myocytes could be explained by the lower density of inward-rectifying potassium channels (i.e. I_K,(Ach,Ado)) in guinea-pig ventricular versus atrial myocytes (Koumi & Wasserstrom, 1994). Regardless, the shortening of the atrial but not the ventricular action potential by CVT-510 is in keeping with the activation by adenosine of I_KAdo found in atrial but not ventricular myocytes. On the other hand, in both atrial and ventricular myocytes, adenosine causes little or no decrease in basal L-type calcium current (Belardinelli et al., 1995; Song & Belardinelli, 1996). Thus, the finding that CVT-510 did not decrease LVP and ventricular dP dt⁻¹ is consistent with reports that adenosine does not affect resting membrane potential, action potential duration and amplitude, or twitch amplitude in ventricular myocytes of various mammalian species, including humans (Belardinelli et al., 1995; Song & Belardinelli, 1996). Similar to adenosine and adenosine receptor agonists, diltiazem reduces action potential duration and contractile force in guinea-pig atrial myocytes (Nabata, 1977). However, unlike adenosine and adenosine receptor agonists, diltiazem shortens the action potential duration and decreases contractile force in guinea-pig ventricular myocytes (Nakajima et al., 1975). In keeping with this latter study (Nakajima et al., 1975) are our findings that diltiazem significantly reduced LVP and dP dt⁻¹ in guinea-pig isolated perfused hearts.

The shortening of atrial action potential and refractory period by adenosine facilitates the induction of atrial flutter or fibrillation (Kabell et al., 1994; DiMarco et al., 1990). In fact, adenosine has been reported to induce transient atrial fibrillation in 2–5% of patients (Kabell et al., 1994; DiMarco et al., 1990). By the same mechanism, that is shortening of the atrial action potential (Figure 6), CVT-510 could also facilitate the induction of atrial flutter or fibrillation. However, CVT-510 is 3 fold more potent in prolonging the S–H interval (EC₅₀=41 nM) than in shortening the atrial action potential (EC₅₀=126 nM). Thus, it is possible that a significant negative dromotropic effect of CVT-510 can be achieved at a concentration that does not significantly shorten the atrial action potential and refractory period. The differential potencies of adenosine and other A₁ receptor agonists to lengthen the S–H interval and to shorten the atrial action potential are summarized in Table 2. For example, the potency of adenosine to prolong the S–H interval is 1.1 fold greater than its potency to shorten the atrial action potential (Raatikainen et al., 1996). The basis for this differential potency of adenosine and A₁ receptor agonists to slow AV nodal conduction and shorten the atrial potential can be explained by the greater A₁ receptor reserve of the AV node than of atrial myocardium. This conclusion is based on the observation that inactivation of 40% of A₁ receptors with the irreversible antagonist FSCPX (Srinivas et al., 1997) caused a significant attenuation of the maximal shortening of atrial potential but did not reduce the maximal S–H interval prolongation caused by CVT-510 (unpublished observation from our laboratory). This is indicative of a larger A₁ receptor reserve in the AV node than in the atrium.

Table 2.

Potency of adenosine and A1 receptor agonists to prolong S–H interval and to shorten atrial action potential

Therapeutic implications

Paroxysmal supraventricular tachycardias (PSVT) affect 570,000 people with 89,000 new cases diagnosed each year (Orejarena et al., 1998). At present, the most commonly used agent for rapid conversion of SVTs to sinus rhythm is adenosine (Langan, 1997). The limitations of adenosine for the long-term treatment of SVTs are owed to its ultra-short half-life in the circulation (∼10 s) and to its vasodilatory effects. In acute cases where adenosine fails to restore normal sinus rhythm or cases in which long term management is required, calcium channel blockers such as verapamil or diltiazem are indicated (Langan, 1997). However, the utility of calcium channel blockers for the treatment of SVTs can be limited by their vasodilatory and negative inotropic actions. As shown in this study, diltiazem at concentrations that cause similar negative dromotropic effects as the A₁ receptor agonist CVT-510, significantly depresses ventricular contractility, increases coronary conductance, and causes hypotension.

In summary, the data reported here suggest that A₁ receptor-selective agonists such as CVT-510 may circumvent the problems associated with both adenosine and calcium channel blockers in the acute and long-term management of SVTs involving the AV node as part of the re-entrant circuit.

Abbreviations

LVP

left ventricular developed pressure

MABP

mean arterial blood pressure

MAPD₉₀

monophasic action potential duration at 90%repolarization

S–H

stimulus-His-bundle interval