Journal Logo

Article

Antiischemic and Antiarrhythmic Activities of Some Novel Alinidine Analogs in the Rat Heart

Challinor-Rogers, Joanne L.; Rosenfeldt, Franklin L.; Du, Xiao-Jun; McPherson, Grant A.

Author Information
Journal of Cardiovascular Pharmacology: April 1997 - Volume 29 - Issue 4 - p 499-507
  • Free

Abstract

Kobinger et al. (1) first proposed that alinidine (the n-allyl derivative of clonidine) possessed antiischemic properties after initial studies of acute coronary occlusion in cats. Alinidine reduced both severity of ischemia and myocardial oxygen consumption (MVO2). This was initially attributed to the bradycardic action of alinidine, as a reduction in heart rate will decrease myocardial oxygen demand and increase myocardial oxygen supply simply by prolonging diastole. In addition to protective effects on the myocardium, alinidine possesses antiarrhythmic properties, as evidenced by a reduction in the incidence of ventricular fibrillation, mortality, and infarct size in rats (2-4) and the reversion of tachycardias and prevention of ventricular ectopic beats in dogs (5).

Although bradycardia was initially believed responsible for the antiischemic activity of alinidine, subsequent studies using paced hearts (6,7) demonstrated a protective effect of alinidine on the myocardium in the absence of any changes in heart rate. It thus became clear that the mechanism responsible for the beneficial effects of alinidine was more complicated than that predicted from initial studies. In addition, the mechanism by which alinidine produces antiarrhythmic effects remains unclear.

We recently found that alinidine and a number of related compounds are also antagonists of KATP channel openers such as cromakalim in vascular smooth muscle, possessing actions similar to those of another KATP channel antagonist, glibenclamide (8-10). KATP channel antagonism by alinidine and its analogs is not related to the bradycardic activity of the compounds (11), as compounds have been produced that are selective for one of these actions.

We developed several compounds related to alinidine but with markedly different pharmacologic activity in terms of their bradycardic/KATP channel antagonist action, and we wished to determine whether these novel compounds possessed dissimilar antiischemic or antiarrhythmic actions or both. The choice of compounds to be tested and the concentrations used were based on previous characterization (10,11). Alinidine was chosen as the reference compound, equipotent as both a KATP channel antagonist and bradycardic agent; TH92:20 and TH92:4 as more potent and selective bradycardic agents; and TH91:21 and TH91:22 as more potent and selective KATP channel antagonists. The isolated working rat-heart model was used to determine the antiischemic actions of the alinidine analogs, and an in situ model of ischemia-and reperfusion-induced arrhythmias was used to investigate the antiarrhythmic actions of the same compounds. In addition to characterizing the cardioprotective activity of these compounds, we also performed some studies investigating a possible role for KATP channel antagonism in the antiischemic and antiarrhythmic actions of these compounds. To this end we also tested glibenclamide, a known KATP channel blocker, at a concentration that produced similar antagonism of KATP channels as the more potent alinidine analogs (10,11), for comparison of cardioprotective effects in the same experimental models.

METHODS

Isolated working rat-heart preparation and protocols

Male WKY rats aged 11-16 weeks (336 ± 4 g) were anesthetized with 4% halothane in oxygen. Heparin (200 IU) was administered i.v. and the heart rapidly excised and transferred to ice-cold heparinized buffer. The heart was immediately mounted on the isolated working rat-heart apparatus, described by Taegtmeyer et al. (12), by aortic cannulation, and perfused. Once the aorta had been cannulated and perfusion commenced, the left atrium was also cannulated. The perfusate was a standard Krebs-Henseleit solution (in mM): NaCl, 118; KCl, 4.7; CaCl2, 2.5; Ca.EDTA, 0.5; MgSO4, 1.2; KH2PO4, 1.2; NaHCO3, 25; and glucose 10 (37°C, gassed with 95% O2/5% CO2; pO2, 400-500 mm Hg; pCO2, 36-41 mm Hg; pH 7.4).

In brief, the isolated working rat-heart apparatus effectively allows three distinct modes of operation.

  1. Nonworking, which is the same as the classic Langendorff preparation, in which the myocardium is perfused via the coronary arteries and the heart contracts spontaneously but performs no external hydraulic work. In this mode, oxygenated perfusion medium is delivered to the heart via an aortic cannula.
  2. Working mode, in which oxygenated perfusate is introduced into the left atrium via an angled cannula. In this mode, the perfusion medium enters the left ventricle through the mitral valve and is ejected into the aorta, the heart thus performing external work. Thus this is essentially a left heart preparation, with inflow into the left atrium and outflow from the aorta.
  3. Global no-flow ischemia, in which the aortic cannula is clamped to cut off coronary perfusion. The treatment solution is introduced into the heart via the aortic cannula immediately before the ischemic insult, and during the ischemic period, the heart is surrounded by the solution in a water bath maintained at 37°C.

Hearts were initially perfused in the nonworking mode for 15 min (to allow time for the heart to adjust to the extracorporeal conditions), switched to the working mode for 15 min, and then hearts were infused for 1 min with 10 ml of the test solution and subjected to a 25-min period of normothermic (37°C) global noflow ischemia. The heart was then reperfused in the nonworking mode for 15 min, and a second work period of 15 min commenced. Each drug solution was tested on at least six hearts.

"Arterial" and "venous" partial pressures of oxygen were determined from perfusate samples by using a Model 1306 Blood Gas Analyser (Allied Instrument Laboratories). Coronary and aortic flow were determined from timed collections of effluent. Aortic systolic and diastolic pressure were measured by a pressure transducer attached by a sidearm to the aortic cannula and recorded (with heart rate) on a model 7D polygraph (Grass Instrument Company). Left atrial filling pressure was maintained at 15 cm H2O (11 mm Hg), and the aortic overflow pressure was maintained at 100 cm H20 (74 mm Hg).

A number of parameters were used to monitor heart function. The rate of performing external cardiac work (power, P) in mJ/s/g wet weight was calculated as follows: Equation (1) where A is peak aortic pressure (mm Hg), LAP is left atrial pressure (11 mm Hg), CO is cardiac output (ml/min), and W is heart wet weight (g). MVO2 in μl O2/min/g wet weight was calculated as follows: Equation (2) where PaO2 and PvO2 (mm Hg) are arterial and venous partial pressures of oxygen measured across the circulation, 760 is the barometric pressure (mm Hg), and CF is coronary flow (ml/min). The mechanical efficiency of the heart (E) in mJ/μl was calculated by dividing the cardiac power by MVO2. Equation (3) Absolute postischemic values were converted to percentage recovery of preischemic values for each parameter to allow comparison of different drug treatments with the control group.

In situ model of ischemia- and reperfusion-induced arrhythmias

Male WKY rats aged 11-16 weeks (313 ± 7 g) were anesthetized with pentobarbitone (60 mg/kg, i.p.). After an i.v. injection of heparin (200 IU), the heart was exposed and the ascending aorta cannulated, beginning Langendorff perfusion in situ. The perfusion medium was a modified Krebs-Henseleit solution (in mM): NaCl, 125; KCl, 4; CaCl2, 1.85; MgCl, 1.05; NaHCO3, 25; NaH2PO4, 0.5; Ca.EDTA, 0.027; and glucose 11 (37°C, gassed with 95% O2/5% CO2; pH 7.4). It should be noted that this is not an anesthetized rat model. The only difference between this isolated perfused heart model and the typical Langendorff model is that the heart is kept in situ (13). Coronary flow rate was controlled by an Ismatac peristaltic pump and maintained at 5 ml/min/g of heart weight, with coronary perfusion pressure maintained at 32-38 mm Hg, measured by pressure transducer (model P23; Gould Statham) connected to a sidearm of the perfusion line.

After a 15- to 20-min stabilization period, the test solution was infused for a period of 8 min. The left coronary artery was then ligated 2-3 mm from its origin to induce local ischemia for 25 min. The success of occlusion of the coronary artery was ensured by a ≥25% increase in coronary perfusion pressure, and during this period, aortic flow was reduced to maintain perfusion pressure at preischemic levels. At the end of the ischemic period, the ligature was released and the heart was perfused at preischemic flow rates for a period of 3 min. Drug infusion was continued throughout the experiment.

Drugs (or vehicle) were infused through a sidearm in the aortic cannula by using an infusion pump (Harvard Apparatus). Coronary perfusion pressure, ventricular pressure (derived from a microtip pressure transducer; model SPR-249, Millar Instruments), and epicardial ECG (recorded with an electrode positioned on the ischemic region of the left ventricle, with the other lead attached to the metal aortic cannula) were constantly recorded throughout the experiment on a model 7 polygraph (Grass Instrument), with paper speed during coronary artery occlusion set at 25 mm/s. Pacing electrodes were attached to the right atrium and the ventricular wall just below the ascending aorta. Hearts were stimulated 70 beats/min faster than initial heart rates, by using square-wave pulses at 2-4 V and 2-ms duration.

Arrhythmias induced by coronary artery occlusion were defined according to the Lambeth Conventions (14). The severity of the arrhythmias was quantitated by using gaussian distributed arrhythmia scores, adapted from Curtis et al. (15) for ischemia-induced arrhythmias and from McLennan et al. (16) for reperfusi on-induced arrhythmias.

Drugs

Drugs that were generously donated were alinidine (Boehringer Ingelheim) and glibenclamide (Hoechst). The alinidine analogs TH91:21, TH91:22, TH92:4, and TH92:20 were synthesised by Theresa Hay at the Department of Chemistry, Monash University, according to methods previously described (17,18). Figure 1 shows the structures of alinidine and the novel analogs used in our study. All analogs were synthesised as bases.

FIG. 1
FIG. 1:
Structures of alinidine and novel analogs used in this study.

Alinidine and the alinidine analogs were made up in 100% methanol as stock solutions of 10 mM. Glibenclamide was made up in 100% ethanol as a stock solution of 1 mM. Dilutions were made in buffer. The final concentrations for alinidine and alinidine analogs were 10 μM and for glibenclamide, 1 μM. The choice of these concentrations was based on previous characterization of the compounds (10,11). Thus the alinidine analogs were used at a concentration of 10 μM, which produced significant KATP channel antagonism in vascular smooth muscle. Glibenclamide was used at a concentration that produced antagonism of vascular KATP channels similar to that of the more potent alinidine analogs, TH91:21 and TH91:22 (10,11).

Statistics

Values given are mean ± SEM for normally distributed data or median (range) for data not normally distributed (arrhythmia data), with the number of experiments specified. The difference between two dependent groups was assessed by using a paired t test. Differences between two independent groups were assessed by using an unpaired t test. Groups of data undergoing multiple comparisons were tested for normality (Kolmogorov-Smirnov test) and equal variance (Levene Median test). For normally distributed data, multiple comparisons of independent groups were made using a one-way analysis of variance (ANOVA) with post-ANOVA Dunnett's test (comparing a number of groups with a control group). Data not normally distributed were compared by using a Kruskal-Wallis one-way ANOVA on ranks with post-ANOVA Dunn's test (comparing a number of groups with a control group). The incidence of arrhythmias was compared by using a Fisher Exact test. Sigmastat (Jandal Scientific) was used to perform these tests. Finally, one-sample analysis was used to compare postischemic function (as a percentage of preischemic values) to 100%. Statview (Abacus Concepts Inc.) was used to perform this test.

RESULTS

Antiischemic properties of the alinidine analogs in the isolated working rat heart

In the vehicle control group, power (P) after the 25-min ischemic insult recovered only 46 ± 8% (n = 14) of preischemic values. In hearts treated with TH91:21 (10 μM), recovery was significantly improved (83 ± 3%; n = 6; p < 0.05; Table 1). Myocardial oxygen consumption (MVO2) recovered 75 ± 7% in the vehicle control group, with a trend to increased recovery (96 ± 6%; n = 6) in the TH91:21 group (see Table 1). Similarly, efficiency (E) was reduced after ischemia in the control group, with recovery to only 58 ± 8% (n = 14) of preischemic levels. Recovery of efficiency was also significantly improved (p < 0.05; Table 1) in hearts treated with 10 μM TH91:21 (87 ± 4%; n = 6).

TABLE 1
TABLE 1:
Pre- and postischemic values for power (P), myocardial oxygen consumption (MVO2), and efficiency (E), for each treatment group, and percentage change in each group after ischemia

Postischemic heart rate did not differ significantly from preischemic values in each treatment group. In addition, there was no effect of any of the drug treatments on the recovery of heart rate after ischemia, when compared with the vehicle control group (Table 1).

All other compounds tested were without effect on MVO2, power (P) and efficiency (E) in this series of experiments: alinidine (10 μM), the alinidine analogs TH91:22 (10 μM) and TH92:20 (10 μM), and glibenclamide (1 μM).

The time for the heart to arrest after the induction of ischemia and the time taken for the heart to beat regularly after cessation of ischemia also were calculated. It was found that in the control group, hearts arrested after 556 ± 88 s (n = 14) of normothermic global no-flow ischemia and took 253 ± 95 s (n = 14) to beat regularly in normal sinus rhythm on reperfusion. There was a tendency for treatment with TH91:21 (10 μM) to reduce the time to arrest and the time taken to resume regular sinus rhythm, with values of 294 ± 38 s (n = 6) and 63 ± 8 s (n = 6), respectively, although these differences were not statistically significant when all treatment groups were compared with the control. It was noted also that all hearts treated with TH91:21 began beating in spontaneous sinus rhythm immediately on reperfusion.

Antiarrhythmic properties of alinidine analogs against ischemia- and reperfusion-induced arrhythmias

Paced hearts.

1. Ischemia-induced arrhythmias. The ability of the alinidine analogs to reduce arrhythmic events occurring in response to regional ischemia was assessed in paced hearts. Vehicle-treated rat hearts served as controls. Hearts were paced at 261 ± 10 beats/min, with no significant difference between the pacing heart rate of any of the treatment groups compared with the control (data not shown). In addition, there was no difference between the control group or any of the treatment groups with respect to the increase in coronary perfusion pressure (34 ± 3% in the control group) produced on coronary artery occlusion (p > 0.05, one-way ANOVA), indicating a similar size of ischemic area induced.

The overall incidence of ventricular premature beats (VPBs) was not significantly different in any of the drug-treatment groups compared with the vehicle-treated control hearts (100% in all groups); however, the total number of VPBs occurring during ischemia was significantly reduced by treatment with either TH91:21 or TH91:22 (10 μM;Table 2).

TABLE 2
TABLE 2:
Arrhythmic events observed during 25-min ischemia and 3-min reperfusion in paced and unpaced hearts

The total duration of ventricular tachycardia (VT) was significantly reduced by treatment of the hearts with either TH91:21 or TH91:22 (10 μM;Table 2), as was the overall incidence of VT, with 9-10% of hearts treated with 10 μM TH91:21 or TH91:22 exhibiting episodes of VT, compared with 91% of hearts in the vehicle-treated control group. The incidence of VT was also significantly reduced by treatment of the hearts with glibenclamide (1 μM;Table 2). The time of onset of VT was not significantly affected by any of the treatment regimens used in this series of experiments when compared with the control group (data not shown).

The overall incidence and total duration of ventricular fibrillation (VF) was significantly reduced by treatment of the hearts with alinidine (10 μM), TH91:21 (10 μM), or TH91:22 (10 μM), although the time of onset of the first episode of VF, when occurring, was similar in all groups (data not shown).

The severity of ischemia-induced arrhythmias was quantified by using a gaussian distributed arrhythmia score, taking into account all arrhythmic events. It was found that TH91:21 and TH91:22 (10 μM) significantly reduced the severity of the arrhythmias. In the control group, arrhythmias scored 4.4 ± 0.4 (n = 21) compared with 0.9 ± 0.3 (n = 10) or 0.5 ± 0.3 (n = 11) in the TH91:21 and TH91:22 treatment groups, respectively (Table 2).

It was shown that there was no significant effect of TH92:4 (10 μM) or TH92:20 (10 μM) on ischemia-induced arrhythmias in paced hearts.

2. Reperfusion-induced arrhythmias. The ability of the alinidine analogs to reduce arrhythmic events occurring in response to reperfusion also was assessed. There was no significant difference in the number of VPBs or the total duration of VT in response to reperfusion in any of the treatment groups when compared with the control, although the incidence of VT was significantly reduced from 87% of control hearts to only 27% of hearts treated with TH91:22 (10 μM;Table 2). Similar to results obtained for ischemia-induced arrhythmias, there was no significant effect of any of the drug treatments, when compared with the control, on the onset of VT (data not shown).

Total duration of VF in the control group was significantly reduced (see Table 2) in hearts treated with TH91:21 (10 μM), TH91:22 (10 μM) or TH92:4 (10 μM). The incidence of sustained VF was also reduced by these same compounds, although only TH91:22 produced a significant (p < 0.01) reduction in the overall incidence of VF (both sustained and transient episodes). Treatment of the hearts with glibenclamide also led to a significant (p < 0.05) reduction in the incidence of VF (Table 2). There was no significant effect of any treatment on the onset time of VF (data not shown).

TH91:21 (10 μM) and TH91:22 (10 μM) significantly reduced the severity of arrhythmias occurring in response to reperfusion (p < 0.05). The severity was 8.3 ± 0.2 (n = 15) in the control group, compared with 1.1 ± 0.5 (n = 10) in the TH91:22-treated group or 5.2 ± 0.7 (n = 10) in the TH91:21-treated group (Table 2).

There was no significant effect of alinidine (10 μM), TH92:4 (10 μM), or TH92:20 (10 μM) against reperfusion-induced arrhythmias in paced hearts.

Unpaced hearts. In these experiments, heart rate was allowed to vary spontaneously. Heart-rate responses were monitored at 1-min intervals during infusion and reperfusion periods and at 5-min intervals during ischemia. All values were then expressed as a percentage of initial heart rate, which was similar in all treatment groups (192 ± 16 beats/min). The increase in perfusion pressure produced by coronary artery occlusion was also similar in all treatment groups (33 ± 2%).

In the control group, heart rate remained stable during the 8-min infusion period (102 ± 3% of initial values at 1-min drug infusion vs. 100 ± 4% of initial values at 8-min drug infusion); however, on coronary artery ligation, there was a significant decrease in heart rate (to 79 ± 5% of initial values; Fig. 2). This response was similar in hearts treated with glibenclamide (1 μM), with no significant difference between the two groups at any time point. Infusion of the alinidine analogs or alinidine produced significant bradycardic responses during both drug infusion and ischemia (Fig. 2).

FIG. 2
FIG. 2:
Heart-rate responses with vehicle treatment (closed circles) in comparison with the effects of alinidine, glibenclamide, and the alinidine analogs (open symbols as indicated) in unpaced hearts during both drug infusion and ischemia after coronary artery occlusion. Standard error bars have been omitted for clarity. The standard error was <15% for each data point. Results are the mean of nine to 14 experiments.

1. Ischemia-induced arrhythmias. The total number of VPBs induced by coronary artery occlusion in unpaced hearts was significantly (p < 0.05) reduced in groups treated with TH91:21 (10 μM), TH91:22 (10 μM), or TH92:20 (10 μM; see Table 2), and there was also a tendency (although not statistically significant) for alinidine to reduce this number. The incidence of VPB was not significantly affected by any of the treatment regimes (data not shown). The total duration and incidence of VTs occurring during ischemia were significantly reduced (p < 0.05; Table 2) in hearts treated with TH91:21 or TH91:22, and there was also a tendency for treatment with TH92:20 or alinidine to reduce the duration of VT. There was no effect of any of the treatments on the time of onset of VT (data not shown). There was no VF in response to coronary artery occlusion in any group of unpaced hearts.

There was a significant (p < 0.01) reduction in the severity scores of the arrhythmias in unpaced hearts from 3.2 ± 0.4 (n = 14) in the control group to 0.4 ± 0.3 (n = 11) and 0.6 ± 0.5 (n = 11) in the groups treated with TH91:21 and TH91:22, respectively (Table 2).

Glibenclamide (1 μM) was without significant effect on ischemia-induced arrhythmias in unpaced hearts.

2. Reperfusion-induced arrhythmias. Interestingly, in unpaced hearts, the total number of VPBs was significantly (p < 0.05) increased by alinidine (10 μM), TH91:21 (10 μM), or TH91:22 (10 μM). TH91:21 and TH91:22 also produced a significant increase in the incidence of VPBs (Table 2), but this is likely to be because the majority of control hearts developed VF soon after reperfusion.

The total duration of reperfusion-induced VT in unpaced hearts was not significantly affected by any of the alinidine analogs tested, although alinidine and TH92:20 (10 μM) were found significantly (p < 0.05) to reduce the incidence of VT under these conditions (Table 2). TH91:22 (10 μM) delayed the time of onset of VT to 34 s (range, 20-162 s) when compared with control values of 2 s (range, 0-45 s) and had a tendency to reduce the incidence of VT (p = 0.08; Table 2).

In unpaced hearts, both TH91:21 (10 μM) and TH91:22 (10 μM) significantly decreased (p < 0.01) the duration and incidence of reperfusion-induced VF (Table 2), although having no effect on the onset of VF (data not shown). Alinidine significantly reduced the incidence of VF (p < 0.01) in unpaced hearts, and there was a tendency for the duration of VF to be reduced, although this was not statistically significant (Table 2).

Finally, in line with other results, the severity of reperfusion-induced arrhythmias was significantly reduced from 7.5 ± 0.8 (n = 14) in the control group to 3.0 ± 0.5 (n = 10), 2.9 ± 0.8 (n = 11) and 1.4 ± 0.6 (n = 11) by treatment with alinidine, TH91:21, and TH91:22, respectively (Table 2).

Glibenclamide (1 μM) had no significant effect on reperfusion-induced arrhythmias in unpaced hearts.

DISCUSSION

This study is the first to demonstrate that TH91:21 (the butyl derivative of alinidine), at the tested concentration of 10 μM, is a potent antiischemic agent in the isolated working rat-heart model and antiarrhythmic agent against both ischemia- and reperfusion-induced arrhythmias in an in situ isolated rat-heart model. TH91:22 (the pentyl derivative of alinidine) also was shown to possess potent antiarrhythmic activity.

The ability of alinidine to antagonize smooth-muscle KATP channels was identified in 1989 (8), 10 years after the bradycardic activity of alinidine was discovered (1). Recently we tested a number of compounds related to alinidine (including TH91:21 and TH91:22) for both bradycardic activity and blockade of vascular KATP channels (10,11). TH91:21 and TH91:22 were selected for our study because they possessed the most potent KATP channel-blocking activity (∼pKB = 6.2). Conversely TH92:20, the cyclic derivative of alinidine, and TH92:4 were chosen because they were the most potent (∼pD2 = 6.1) and selective bradycardic agents (11).

The finding that the two most potent antiarrhythmic agents, TH91:21 and TH91:22, are also the most potent KATP channel antagonists (19) may suggest that their antiarrhythmic activity relies on blockade of the KATP channel. However, glibenclamide [1 μM, chosen because previous studies have shown that this concentration produces a similar blockade of vascular KATP channels as does 10 μM TH91:21 and TH91:22 (10,11)] had less marked antiarrhythmic effects against arrhythmias in paced hearts and no significant antiarrhythmic activity in unpaced hearts, suggesting that KATP channel antagonism is at least not the only mechanism involved in the antiarrhythmic actions of the alinidine analogs. The results with glibenclamide are in line with those of other studies that demonstrated antiarrhythmic activity of glibenclamide in different models. For example, glibenclamide (1 μM) reduced both arrhythmias and K+ efflux in Langendorff perfused rat hearts subjected to either coronary artery occlusion or global ischemia (20). Tosaki et al. (21) also demonstrated, in an isolated working rat-heart model, that glibenclamide (3 μM) was an effective antiarrhythmic agent, although not all studies reported antiarrhythmic effects of glibenclamide (22). However, whereas KATP channel blockade may be involved in the antiarrhythmic action of alinidine, it is likely that at the concentration of TH91:21 and TH91:22 tested (10 μM), KATP channel blockade alone does not account for the activity of these compounds.

It also should be noted that although KATP channel blockade may produce antiarrhythmic effects, KATP channel antagonists such as glibenclamide are reported either to worsen the metabolic alterations produced by ischemia, decreasing ATP content of the heart and increasing lactate content (23) or to have no significant effect on postischemic recovery (24-26). This is in line with the results of our study, which show that although glibenclamide possesses some antiarrhythmic activity in the in situ model of ischemia- and reperfusion-induced arrhythmias, there was no beneficial effect of glibenclamide on postischemic recovery of the isolated working rat heart. Thus although KATP channel blockade may play a role in the antiarrhythmic activity of the alinidine analogs, it is unlikely that it plays any part in the antiischemic actions observed.

The contribution of the bradycardic action of the alinidine analogs to their beneficial antiischemic and antiarrhythmic action also remains unclear. From the results of our study, combined with those of previous studies (1,2), it appears that the bradycardic action of the alinidine analogs is ultimately beneficial, as heart rate is an important factor in the risk of arrhythmias. As noted earlier, the antiischemic activity of alinidine in studies using paced hearts (6,7) demonstrated a protective effect of alinidine on the myocardium in the absence of any changes in heart rate. In this study, compounds were tested in perfused rat hearts with or without controlling the heart rate by pacing. Thus the contribution of any bradycardic action of these agents to overall activity could be determined. It was observed in our study that the antiarrhythmic actions of TH91:21 and TH91:22 (10 μM) were evident in both cases. These results support the idea that bradycardia is not the only mechanism through which these imidazolidine-based compounds produce cardioprotection. Further support of this hypothesis arises from the fact that alinidine, TH91:21, and TH91:22 are all equipotent in terms of bradycardic activity (11). That TH91:21 and TH91:22 are active where alinidine is not provides additional evidence that the activity of these compounds is not simply a result of their bradycardic activity.

In our series of experiments, we found that alinidine (10 μM) had no significant effect on postischemic recovery of heart function when values of power, myocardial oxygen consumption, and efficiency were compared with control values. The lack of antiischemic effect of alinidine in this series of experiments is in contrast with some previous studies that demonstrated antiischemic effects of alinidine, both in vitro (7,27) and in vivo (1-4). However, only one group used a Langendorff perfused rat-heart model (7,27), and the doses of alinidine tested were higher (6-200 mM) than that used in our study. These investigators examined only metabolic changes because of the inability to measure functional parameters with their model, and as such, it is hard to compare their results with the findings from our study.

It is not unreasonable to predict an antiarrhythmic action of the alinidine analogs as a result of a significant myocardial protective effect. For example, several studies have reported that KATP channel openers have antiischemic actions, reducing the severity of ischemia/reperfusion injury of the myocardium both in vitro and in vivo, and a few of these studies also reported antiarrhythmic effects of these compounds (28-30). However, this may be a simplistic view, given that other studies reported proarrhythmic effects of the KATP channel openers (28,31-32), coupled with antiischemic effects. In our study, although TH91:21 produced both potent antiischemic and antiarrhythmic effects, TH91:22, which was equally as potent at producing antiarrhythmic effects at the concentration studied (10 μM), had no significant effect on postischemic recovery of the isolated working rat heart. Collectively, these results indicate a more complex relationship between myocardial protection and antiarrhythmic activity.

A number of basic structure-activity trends developed for the antiarrhythmic activity of the alinidine analogs. Lengthening of the N-allyl side chain from 3C (alinidine) to 4C (TH91:21) or 5C (TH91:22) led to a significant increase in antiarrhythmic potency. The double bond in the terminal position of the side chain may affect the activity of the analogs against reperfusion-induced arrhythmias, as TH91:21 (which lacks this double bond) was less potent than TH91:22 (which retains the double bond) in reducing the severity of reperfusion-induced arrhythmias in paced hearts. Thus the length and saturation of the carbon side chain are important in the antiarrhythmic activity of these compounds. The development of structure-activity relations for the antiischemic actions of the alinidine analogs was limited by the small number of analogs tested. We can conclude only that the modifications to the alinidine molecule made to form TH91:21 (lengthening the N-allyl side chain to 4C and removing the double bond in the terminal position of the side chain) produce a compound significantly more potent than alinidine in protecting the isolated working rat heart from ischemic damage.

These studies were restricted by the limited amount of these novel analogs available for testing. We could test each compound at only one concentration (10 μM). Despite this limitation, it is clear that we found two compounds (TH91:21 and TH91:22) that, on a molar basis, are much more active antiarrhythmic agents than is alinidine itself. Although our results suggest that KATP channel antagonism and bradycardic actions might be partly responsible for the actions of these compounds, more extensive studies on TH91:21 and TH91:22 are required to determine their actual potency and whether their activity as cardioprotective agents results from enhanced activity of these compounds on sites of action already known for alinidine (e.g., bradycardia and KATP channel antagonism) or the combination of these activities. Another possibility is that these compounds have some other unidentified actions leading to the cardioprotection. In line with this hypothesis, it has been suggested that the antiarrhythmic action of alinidine may be related to a restriction of current through anion-selective channels (33,34).

The results of our study may have clinical application in the treatment of myocardial ischemia and infarction and the prevention of sudden arrhythmic death. Currently available antiarrhythmic agents primarily target arrhythmias occurring during ischemia and are essentially ineffective against arrhythmias that occur after reperfusion (35). Thus the ability of the alinidine analogs TH91:21 and TH91:22 to reduce reperfusion-induced VF is an important finding. The alinidine analog TH91:21 may also have another advantage. The treatment of ischemic heart disease has two main aims: to minimize the loss of functional myocardium and to reduce the risk of lethal arrhythmias. Drugs that aim to minimize the loss of functional myocardium by offering myocardial protection do not also produce antiarrhythmic effects, making treatment of this condition complex. TH91:21 produces both antiarrhythmic and antiischemic effects, a novel and beneficial pharmacologic profile. Further development of these analogs of alinidine may lead to more simple and effective treatment and management of cardiovascular disorders such as ischemic heart disease and to cardioplegic agents for use during cardiac surgery.

In summary, we identified two novel alinidine analogs that display marked cardioprotective activity. TH91:21 (10 μM) is a potent antiischemic agent in the isolated working rat heart subject to normothermic global ischemia, and both TH91:21 (10 μM) and TH91:22 (10 μM) are potent antiarrhythmic agents, significantly reducing both ischemia- and reperfusion-induced arrhythmias. At present the mechanism of action responsible for the antiischemic and antiarrhythmic effects of TH91:21 and TH91:22 is unclear, although it would appear at this stage that their bradycardic or KATP channel antagonist activities play only a partial role.

Acknowledgment: It is a pleasure to thank Mr. Simon Keily for his assistance with this project. This project was made possible only through the synthesis of the alinidine analogs by Ms. Theresa Hay and the contributions of Professor W. R. Jackson, Dr. A. Patti, and Ms. E. Campi from the Department of Chemistry, Monash University. This work was supported by an Institute grant from the NH&MRC (Australia).

REFERENCES

1. Kobinger W, Lillie C, Pichler L. N-Allyl derivative of clonidine, a substance with specific bradycardic action at a cardiac site. Naunyn Schmiedebergs Arch Pharmacol 1979;306:255-62.
2. Harron DWG, Brezina M, Lillie C, Kobinger W. Antifibrillatory properties of alinidine after coronary artery occlusion in rats. Eur J Pharmacol 1985;110:301-8.
3. Reichhalter R, Lillie C, Kobinger W. Influence of alinidine on survival and infarct size in rats with coronary artery occlusion. Naunyn Schmiedebergs Arch Pharmacol 1988;337:R68.
4. Reichhalter R, Lillie C, Kobinger W. Effects of alinidine on survival and infarct size in rats with coronary artery occlusion. Eur J Pharmacol 1988;157:75-81.
5. Allen JD, Harron DWG, Shanks RG, Wilson R. Effects of alinidine (N-allyl clonidine, ST 567) on experimental cardiac arrhythmias. Br J Pharmacol 1981;73:199P.
6. Hanet C, Pouleur H, Hue L, et al. Effects of alinidine on metabolic response to high-demand myocardial ischemia. J Cardiovasc Pharmacol 1989;13:405-11.
7. Streller I, Walland A. Anti-ischemic effects of alinidine in paced isolated rat hearts. Basic Res Cardiol 1980;85:71-7.
8. McPherson GA, Angus JA. Phentolamine and structurally related compounds selectively antagonise the vascular actions of the K+ channel opener, cromakalim. Br J Pharmacol 1989;97:941-9.
9. McPherson GA, Angus JA. Characterisation of responses to cromakalim and pinacidil in smooth and cardiac muscle by use of selective antagonists. Br J Pharmacol 1990;100:201-6.
10. Challinor JL, McPherson GA. Evidence that imidazol(id)ine- and sulphonylurea-based antagonists of cromakalim act at different sites in the rat thoracic aorta. Clin Exp Pharmacol Physiol 1993;20:567-75.
11. Challinor-Rogers JL, Hay TKC, McPherson GA. Comparison of the cromakalim antagonism and bradycardic actions of a series of novel alinidine analogues in the rat. Naunyn Schmiedebergs Arch Pharmacol 1994;350:158-66.
12. Taegtmeyer H, Hems R, Krebs HA. Utilisation of energy providing substrates in the isolated working rat heart. Biochem J 1980;186:701-11.
13. Du XJ, Anderson KE, Jacobsen A, Woodcock EA, Dart AM. Suppression of ventricular arrhythmias during ischemia-reperfusion by agents inhibiting Ins(1,4,5)P3 release. Circulation 1995;91:2712-6.
14. Walker MJA, Curtis MJ, Hearse DJ, et al. The Lambeth Conventions: guidelines for the study of arrhythmias in ischemia, infarction, and reperfusion. Cardiovasc Res 1988;22:447-55.
15. Curtis MJ, MacLeod BA, Walker MJA. Models for the study of arrhythmias in myocardial ischemia and infarction: the use of the rat. J Mol Cell Cardiol 1987;19:399-419.
16. McLennan PL, Abeywardena MY, Charnock JS. Dietary fish oil prevents ventricular fibrillation following coronary artery occlusion and reperfusion. Am Heart J 1988;116:709-11.
17. Stähle H, Pook K-H. Synthesis in the 2-amino-imidazoline series. Liebigs Ann Chem 1971;751:159-67.
18. Stähle H, Daniel H. Chemistry, pharmacology and structure-activity with a new type of imidazolidines exerting a specific bradycardic action at a cardiac site. J Med Chem 1980;23:1217-22.
19. Challinor JL, McPherson GA. Comparison of the K+ channel block and bradycardic actions of a series of alinidine analogues. Clin Exp Pharmacol Physiol 1992;21:12.
20. Kantor PF, Coetzee WA, Carmeliet EE, Dennis SC, Opie LH. Reduction of ischemic K+ loss and arrhythmias in rat hearts: effect of glibenclamide, a sulfonylurea. Circ Res 1990;66:479-85.
21. Tosaki A, Szerdahelyi P, Das DK. Reperfusion-induced arrhythmias and myocardial ion shifts: a pharmacological interaction between pinacidil and cecletine in isolated rat hearts. Basic Res Cardiol 1992;87:366-84.
22. Pasnani JS, Ferrier GJ. Differential effects of glyburide on premature beats and ventricular tachycardia in an isolated tissue model of ischemia and reperfusion. J Pharmacol Exp Ther 1992;262:1076-84.
23. Kamigaki M, Ichihara K, Abiko Y. Enhancement of ischemic myocardial metabolic derangement by glibenclamide. Jpn J Pharmacol 1994;65:121-9.
24. Iwamoto T, Miura T, Urabe K, Itoya M, Shimamoto K, Iimura O. Effect of nicorandil on post-ischemic contractile dysfunction in the heart: roles of its ATP sensitive K+ channel opening property and nitrate property. Clin Exp Pharmacol Physiol 1993;20:595-602.
25. Warltier DC, Auchampach JA, Gross GJ. Relationship of severity of myocardial stunning to ATP dependent potassium channel modulation. J Cardiovasc Surg 1993;8(suppl):279-83.
26. Toombs CF, Moore TL, Shebuski RJ. Cardioprotection with U-89,232 is not reversible with glibenclamide: evidence of a novel anti-ischemic agent derived from cromakalim. Pharmacology 1994;49:96-104.
27. Walland A, Streller I, Lang U. Antiischemic actions of alinidine in paced rat hearts. J Mol Cell Cardiol 1987;19(suppl III):S106.
28. Antzelevitch C, Di Diego JM. Role of K+ channel activators in cardiac electrophysiology and arrhythmias. Circulation 1992;85:1627-9.
29. Carlsson L, Abrahamsson C, Drews L, Duker G. Anti-arrhythmic effects of potassium channel openers in rhythm abnormalities related to delayed repolarization. Circulation 1991;85:1491-500.
30. D'Alonzo AJ, Darbenzio RB, Hess TA, Sewter JC, Sleph PG, Grover GJ. Effect of potassium on the action of the K channel modulators cromakalim, pinacidil, or glibenclamide on arrhythmias in isolated perfused rat heart subjected to regional ischemia. Cardiovasc Res 1994;28:881-7.
31. De La Coussaye JE, Eledjam, J-J, Bruelle P, et al. Electrophysiologic and arrhythmogenic effects of the potassium channel agonist BRL38227 in anaesthetized dogs. J Cardiovasc Pharmacol 1993;22:722-30.
32. Tosaki A, Szerkahelyi P, Engelman RM, Das DK. Potassium channel openers and blockers: do they possess proarrhythmic or antiarrhythmic activity in ischemic and reperfused rat hearts? J Pharmacol Exp Ther 1993;267:1355-62.
33. Millar JS, Vaughan Williams EM. Pacemaker selectivity: influence on rabbit atria of ionic environment and of alinidine, a possible anion antagonist. Cardiovasc Res 1981;15:335-50.
34. Millar JS, Vaughan Williams EM. Anion antagonism: a fifth class of antiarrhythmic action? Lancet 1981;1:1291-3.
35. Heidbuchel H, Tack J, Vannesta L, Ballet A, Ector H, Van de Werf F. Significance of arrhythmias during the first 24 hours of acute myocardial infarction treated with alteplase and effect of early administration of a beta-blocker or a bradycardic agent on their incidence. Circulation 1994;89:1051-9.
Keywords:

Alinidine; Ventricular arrhythmias; Myocardial ischemia; Rat; KATP channel antagonists

© Lippincott-Raven Publishers