Possible Involvement of Cyclic Adenosine Monophosphate–independent Mechanism in the Positive Chronotropic Effect of Norepinephrine in the Isolated Guinea Pig Right Atrium

All experiments were performed in conformity with the Guiding Principles for Research Involving Animals and Human Beings of Yokohama City University School of Medicine (Yokohama, Japan).

Male guinea pigs weighing 250–350 g were used. After the midline thoracotomy was performed during anesthesia with diethyl ether, the heart was rapidly excised and placed in a dissection dish filled with oxygenated Krebs-Henseleit solution of the following composition: 119.0 mm NaCl, 2.5 mm CaCl², 4.8 mm KCl, 1.2 mm MgSO⁴, 1.2 mm KH²PO⁴, 24.9 mm NaHCO³, and 10.0 mm glucose.

The spontaneously beating right atrium was carefully dissected and mounted vertically in a 10-ml double-walled glass chamber filled with Krebs-Henseleit solution, gassed continuously with 95% O²and 5% CO²(pH 7.4), and maintained at 36°C. The lower end of the right atrium was fixed on a hook, and the upper end was connected by a silk thread to an isometric force-displacement transducer (model UL-10GR, Minebea, Nagano, Japan) through a preamplifier (model AS2103, Sanei-Sokki, Tokyo, Japan). Diastolic tension was adjusted to 0.5 g. Spontaneously beating rate was counted on a chart recording of its developed tension. The preparation was equilibrated for at least 60 min before the experiment began.

The concentration–response curves for the positive chronotropic effect of norepinephrine and forskolin were determined in a cumulative manner by increasing their concentrations in steps of 0.5 molar log unit. The difference between the maximum beating rate produced by the agents and the basal beating rate was taken as 100%, and the percent changes in the beating rate produced by various concentrations of the agents were plotted. When examining the influence of an α¹-adrenoceptor antagonist, a β-adrenoceptor antagonist, phosphodiesterase inhibitors, or ion channel inhibitors on the chronotropic effects of the agents, the concentration–response curves were constructed 30 min after the application of each test compound. Because the phosphodiesterase inhibitors and ion channel inhibitors changed the spontaneously beating rate, the beating rate in the presence of each compound was taken as the basal beating rate. The concentrations of norepinephrine or forskolin producing half-maximal response (EC⁵⁰) in the absence and presence of various test compounds were obtained from log-probit plots of individual response as a function of concentration. The following equation was used to obtain pK^Bvalue:

where [A] is the EC⁵⁰of an agonist in the presence of an antagonist, [A]⁰is the EC⁵⁰of an agonist in the absence of an antagonist, and [B] is a concentration of an antagonist.

To make effects of the test compounds on the chronotropic effects of norepinephrine and forskolin comprehensible, the concentration–response curves depicted on the basis of absolute values of the beating rate were inset into the corresponding figures.

The left atrium was dissected and mounted vertically in the glass chamber as previously described, except that the bath was kept at a temperature of 30°C. The resting tension applied to the preparation was adjusted to 1 g. The atrium was paced electrically by rectangular pulses of 0.5 Hz in frequency, 3 ms in duration, and 1.2 times the threshold voltage, delivered by a pair of spiral platinum electrodes connected to an electric stimulator (SEN-7203, Nihon-Kohden, Tokyo, Japan) through an isolation unit (SS-201J, Nihon-Kohden). The preparation was equilibrated for at least 60 min before the experiment began.

The concentration–response curve for the inotropic effect of norepinephrine was determined in a cumula-tive manner by increasing its concentration in steps of 0.5 log molar unit. The difference between the maximum contractile force produced by norepinephrine and the basal contractile force was taken as 100%. When examining the influence of amrinone on the inotropic effect of norepinephine, the concentration–response curves were constructed 30 min after the application of amrinone. Because amrinone increased the contractile force, that in the presence of amrinone was taken as the basal contractile force.

The following compounds were used: norepinephrine, atenolol, prazosin, amrinone, (±) verapamil, forskolin, 3-isobutyl-1-methylxanthine (IBMX), and CsCl (Sigma Chemical Co., St. Louis, MO). Atenolol, forskolin, and amrinone were dissolved in dimethyl sulfoxide, and further dilution was made with distilled water. All other chemicals were dissolved in distilled water.

Values are given as mean ± SD. Comparisons of values obtained from the concentration–response curves were made by one-way analysis of variance followed by Bonferroni t  test. P < 0.05 was considered significant. The EC⁵⁰values were obtained directly from the concentration–response curves depicted using the Kaleida Graph software program (Synergy Software, Reading, PA).

Norepinephrine increased the spontaneously beating rate of the guinea pig right atria in a concentration-dependent manner (fig. 1A). The agent at a concentration of 100 μm maximally increased the beating rate by 159 ± 15 beats/min from the basal rate of 193 ± 18 beats/min (n = 6). Atenorol at a concentration of 1 μm shifted rightward the concentration–response curve in a parallel manner without affecting the maximum response of the atria to norepinephrine. The EC⁵⁰values of norepinephrine in the absence and presence of atenolol were 0.88 ± 0.55 μm and 5.93 ± 2.64 μm, respectively (n = 6). The pK^Bvalue of atenolol at a concentration of 1 μm was 6.76 (170 nm). Norepinephrine also increased the beating rate by 150 ± 17 beats/min (n = 6) with the EC⁵⁰value of 0.61 ± 0.20 μm in the presence of 0.3 μm prazosin. These values were not significantly different from those in the absence of prazosin. The direct adenylate cyclase activator forskolin also increased the beating rate in a concentration-dependent manner, with an EC⁵⁰value of 0.32 ± 0.16 μm (n = 8;fig. 1B). The maximum increase in the beating rate of 137 ± 11 beats/min produced by forskolin at a concentration of 10 μm was not significantly different from that produced by norepinephrine.

The normalized concentration–response curves for the positive chronotropic effect of forskolin in the presence of amrinone are shown in figure 2A. Amrinone at concentrations of 10 and 100 μm increased the spontaneously beating rate by 13 ± 4 beats/min (n = 12) and 37 ± 6 beats/min (n = 10), respectively. Amrinone accelerated the rate of an increase in the beating rate produced by forskolin, resulting in the leftward shift of the normalized concentration–response curves. The maximum increases in the beating rate produced by forskolin plus 10 μm (146 ± 15 beats/min; n = 6) and 100 μm (155 ± 8 beats/min; n = 4) amrinone from the predrug beating rate were not significantly different from that produced by forskolin alone. In contrast to forskolin, the rate of an increase in the beating rate produced by norepinephrine was not accelerated by amrinone even at a concentration of 100 μm (fig. 2B). Norepinephrine plus amrinone increased maximally the beating rate from the predrug basal rate by 151 ± 15 beats/min (n = 6), which was not significantly different from that produced by norepinephrine alone. Figure 3also shows the conflicting influences of the nonselective phosphodiesterase inhibitor IBMX on the positive chronotropic effects of forskolin and norepinephrine. IBMX at a concentration of 3 μm accelerated the rate of an increase in the beating rate produced by forskolin, resulting in the leftward shift of the normalized concentration–response curve. On the other hand, IBMX did not affect the concentration–response curve for the positive chronotropic effect of norepinephrine. IBMX at a concentration of 3 μm itself increased the spontaneously beating rate by 21 ± 6 beats/min (n = 9). Because norepinephrine activates both α¹-, and β-adrenoceptors, activation of α¹-adrenoceptor might interfere the potentiating effect of the phosphodiesterase inhibitors. However, in the presence of 0.3 μm prazosin, the normalized concentration–response curve for the positive chronotropic effect of norepinephrine in the amrinone-treated atria was superimposable on that in the amrinone-untreated atria (n = 6 for each group, data not shown).

The influences of CsCl, an I^finhibitor, and verapamil, an I^Ca(L)inhibitor, on the positive chronotropic effects of forskolin and norepinephrine are shown in figures 4 and 5. CsCl at a concentration of 2 mm itself decreased the spontaneously beating rate by 56 ± 11 beats/min (n = 12), whereas verapamil at a concentration of 0.2 μm resulted in a decrease of 27 ± 19 beats/min (n = 9). In the presence of CsCl, forskolin and norepinephrine increased the beating rate by 87 ± 11 beats/min (n = 6) and 82 ± 9 beats/min (n = 6), respectively. However, the influences of CsCl on the normalized concentration–response curves of norepinephrine and forskolin were conflicting. CsCl tended to potentiate the positive chronotropic effect of forskolin at lower concentrations and suppress that of the agent at higher concentrations, resulting in the cross of the two concentration–response curves at a forskolin concentration of around 0.5 μm. However, as a whole, CsCl did not affect significantly the normalized concentration–response curve for forskolin (fig. 4A). On the other hand, it inhibited significantly the normalized concentration–response curve for norepinephrine (fig. 4B). As in the case of CsCl, although the concentration–response curve for forskolin in the presence of verapamil crossed the control response curve at the similar point, verapamil did not affect the positive chronotropic effect of forskolin as a whole (fig. 5A). In contrast, verapamil inhibited significantly the positive chronotropic effect of norepinephrine (fig. 5B). In the presence of verapamil, forskolin and norepinephrine increased the beating rate by 149 ± 10 beats/min (n = 4) and 124 ± 16 beats/min (n = 5), respectively.

Norepinephrine increased the contractile force of the electrically driven left atria in a concentration-dependent manner. In contrast to the positive chronotropic effect of norepinephrine, amrinone at a concentration of 100 μm accelerated the increase in the force of contraction produced by norepinephrine, as shown in figure 6. Amrinone itself increased the contractile force by approximately 23%.

In the current study we demonstrated that β-adrenoceptor–mediated positive chronotropic effect was not potentiated by the phosphodiesterase inhibitors amrinone and IBMX, whereas forskolin-produced positive chronotropic effect was potentiated by both agents. The I^finhibitor CsCl and the I^Ca(L)inhibitor verapamil decelerated the rate of an increase in the beating rate produced by norepinephrine. On the other hand, neither CsCl nor verapamil affected significantly the normalized concentration–response curve for the positive chronotropic effect of forskolin. These results suggest that the underlying mechanism of the positive chronotropic effect of the β-adrenergic activation is different from that of the adenylate cyclase activator forskolin.

Cardiac sympathetic activation produces both positive inotropic and chronotropic effects via β-adrenoceptors in the cardiac myocytes and the sinoatrial node cells. The adenylate cyclase–cyclic AMP system has been thought to have a pivotal role in the β-adrenoceptor–mediated positive chronotropic effect as well as the positive inotropic effect. 9,14–16Supporting this notion, forskolin produces the positive chronotropic effect, as shown in the current study. Intracellular application of cyclic AMP also produces the positive chronotropic effect. 17,18 

Inasmuch as cyclic AMP is degenerated by phosphodiesterase, phosphodiesterase inhibitors could theoretically not only mimic but also potentiate the cyclic AMP–mediated responses. In fact, it has been demonstrated that the positive inotropic effect of isoprenaline on human and rabbit ventricular cells, and the antimigratory effect of forskolin on rat aortic vascular smooth muscle cells, are potentiated by phosphodiesterase inhibitors. 19–21If the positive chronotropic effect of β-adrenergic stimulation would also be mediated by cyclic AMP, the effect should be potentiated by the phosphodiesterase inhibition. However, the phosphodiesterase III inhibitor amrinone did not potentiate the positive chronotropic effect of norepinephrine, although amrinone not only increased the beating rate by itself but also potentiated the positive chronotropic effect of forskolin. Because there are several types of phosphodiesterase, 22cyclic AMP increased by β-adrenergic stimulation in the guinea pig sinoatrial node cell might not be degenerated by phosphodiesterase III. To elucidate this possibility, we also examined the influences of the nonspecific phosphodiesterase inhibitor IBMX on the positive chronotropic effects of norepinephrine and forskolin. Consistent with the result obtained from amrinone, IBMX potentiated the effect of forskolin but not the effect of norepinephrine. In addition, amrinone could potentiate the positive inotropic effect of norepinephrine in the guinea pig left atria. Shahid and Rodger 20also reported that both amrinone and IBMX potentiated the positive inotropic effect of isoprenaline in the rabbit papillary muscles but not the positive chronotropic effect of the agent in the rabbit right atria.

There are at least three possible ways to interpret the current results. First, the positive chronotropic effect of norepinephrine is independent of β-adrenoceptors. Second, the cyclic AMP increased by the β-adrenergic activation is compartmentalized in the pacemaker cell, as proposed by Jurevicius and Fischmeister, 23and the phosphodiesterase inhibitors used are hard to access this compartment. Third, cyclic AMP generation is causally unrelated to the positive chronotropic effect of norepinephrine. As shown in figure 1, the β¹-adrenoceptor selective antagonist atenolol shifted rightward the concentration–response curve for the positive chronotropic effect of norepinephrine. The pK^Bvalue of 6.8 for atenolol at a concentration of 1 μm indicated the involvement of β¹-adrenoceptors in the positive chronotropic effect of norepinephrine in the current experiment. 24Recent reports have demonstrated the involvement of α¹-adrenoceptor in regulating heart rate. 25,26However, prazosin did not modify the positive chronotropic effect of norepinephrine in the current study, indicating that α¹-adrenoceptor plays a minor role in controlling the heart rate during the current experimental condition. Amrinone did not potentiate the positive chronotropic effect of norepinephrine even in the presence of prazosin. Therefore, the α¹-adrenoceptors activated by norepinephrine are also unrelated to the inability of amrinone for potentiating the positive chronotropic effect of norepinephrine.

The sinus rate is determined by the spontaneously firing rate of the action potential of the pacemaker cell in the sinoatrial node. β-Adrenoceptor agonists as well as an increase in intracellular cyclic AMP are thought to increase the spontaneously firing rate by means of activation of the I^fchannels and the I^Ca(L)channels. 13,17,27Therefore, we examined influences of the I^fchannel inhibitor CsCl 28and the Ca²⁺channel inhibitor verapamil on the positive chronotropic effects of norepinephrine and forskolin. CsCl and verapamil shifted the normalized concentration–response curve for norepinephrine to the rightward direction. On the other hand, neither CsCl nor verapamil affected significantly the normalized concentration–response curve for forskolin. Even if cyclic AMP that were increased by norepinephrine would be compartmentalized, as far as cyclic AMP were relevant to the positive chronotropic effect of norepinephrine, both CsCl and verapamil should modulate the positive chronotropic effects of norepinephrine and forskolin in a similar manner. However, this is not the case in the current results. Taken together, the current results suggest that cyclic AMP is causally unrelated to the underlying mechanism of the positive chronotropic effect of norepinephrine in the guinea pig heart. It should be mentioned, however, that a decisive conclusion cannot be drawn until direct evidence is obtained, such as no increase in cyclic AMP in the pacemaker cell after treatment with norepinephrine or norepinephrine-produced increase in the ion channel activities in the membrane-delimited model.

Yatani et al.  29,30reported that the β-adrenoceptor–coupled guanosine 5′-triphosphate binding protein Gs could activate both I^Ca(L)and I^fchannels directly, although cyclic AMP–dependent activation of both channels had already been demonstrated. 16,31They also showed the time course of the isoproterenol-produced increase in I^Cain isolated guinea pig cardiomyocytes. 32Isoproterenol increased it in a biphasic manner, with a fast time constant of 150 ms and a slow time constant of 36 s. Forskolin and IBMX increased the I^Caslowly, with time constants of 56 and 150 s, respectively. Therefore, they concluded that the fast response of the myocyte to isoproterenol was caused by the direct modulation of Gs on I^Ca(L)channels, and the slow response was mediated by the second messenger cyclic AMP. 32Isoproterenol also activated I^frapidly, with time constant of 800 ms in cell-free, inside-out patch configuration. 30These direct modulations of β-adrenoceptor–activated Gs on I^Ca(L)and I^fchannels are considered convenient for the sympathetic nervous system for beat-to-beat regulation of heart rate. 33,34The current study might demonstrate these direct interactions of the β-adrenoceptor–Gs ion channels, which have been shown using patch clamp experiments, in the functional experiments.

However, it should be noted that there are some limitations to extrapolate the current data to the clinical setting. We used the atrial tissues and counted the number of the muscle contraction per minute instead of counting the number of the action potential in the pacemaker cells. In addition, we used the isolated tissues. Although isolation of a tissue or a cell makes complicated problem simple, much caution should be used in extrapolating the data to a whole body, which is regulated in a complicated manner by the nervous system and hormonal control. We also have to consider the species difference. The basic heart rate of the guinea pig is much faster than that of humans, suggesting the existence of a delicate difference as to the intracellular regulation on the firing of the action potential in the pacemaker cells between humans and the guinea pig.

Use of the phosphodiesterase III inhibitors amrinone and milrinone is an important therapeutic approach, especially combined with catecholamines, in patients with low cardiac output syndrome. 2The combination of the phosphodiesterase III inhibitors with catecholamines increases cardiac contractility synergistically, whereas the increase in the heart rate is minimal. 1,8These features seem in accord with the ideal inotropic management for perioperative low cardiac output, 1because an increase in heart rate jeopardizes the balance of myocardial oxygen demand and supply. 35The current study might explain, at least partially, the reason for the deferential effects of the phosphodiesterase III inhibitors on the contractility and heart rate observed in the clinical setting. Inasmuch as the heart is innervated by the sympathetic nervous system, the cardiac β-adrenoceptors are activated to a various extent by endogenous norepinephrine. Therefore, the phosphodiesterase III inhibitors could potentiate synergistically the positive inotropic effect of endogenous and exogenous catecholamines, minimally affecting those chronotropic effects.

In conclusion, amrinone and IBMX potentiate the positive chronotropic effect of forskolin. However, the agents did not potentiate the chronotropic effect of norepinephrine. Both CsCl and verapamil shifted rightward the normalized concentration–response curve for the positive chronotropic effect of norepinephrine without affecting the concentration–response curve for that of forskolin. These results strongly suggest that cyclic AMP is irrelevant to the positive chronotropic effect of β-adrenoceptor activation in the guinea pig heart.

The authors thank Noritsugu Tohse, M.D., Ph.D. (Professor and Chairman, Department of Physiology, Sapporo Medical College, Sapporo, Japan), for constructive criticism.