Tsalikakis_kolletis_european Journal Of Pharmacology

Available online at www.sciencedirect.com

European Journal of Pharmacology 580 (2008) 241 – 249 www.elsevier.com/locate/ejphar

Effects of dual endothelin receptor blockade on sympathetic activation and arrhythmogenesis during acute myocardial infarction in rats Theofilos M. Kolettis a,⁎, Giannis G. Baltogiannis a,1 , Dimitrios G. Tsalikakis b,2 , Alexandros T. Tzallas b , Maria G. Agelaki a,1 , Andreas Fotopoulos c , Dimitrios I. Fotiadis b , Zenon S. Kyriakides d a Departments of Cardiology, University of Ioannina, 1 Stavrou Niarxou Avenue, 45110 Ioannina, Greece Departments of Computer Sciences, University of Ioannina, 1 Stavrou Niarxou Avenue, 45110 Ioannina, Greece c Departments of Nuclear Medicine, University of Ioannina, 1 Stavrou Niarxou Avenue, 45110 Ioannina, Greece d Red Cross Hospital, 1 Erythrou Stavrou Street, 115 26 Athens, Greece

b

Received 28 June 2007; received in revised form 11 October 2007; accepted 3 November 2007 Available online 13 November 2007

Abstract The effects of dual (ETA and ETB) endothelin receptor blockade on ventricular arrhythmogenesis during acute myocardial infarction are not well defined. We randomly allocated Wistar rats to bosentan (100 mg/kg daily, n = 24), a dual endothelin receptor antagonist, or vehicle (n = 23). After 7 days of treatment, myocardial infarction was induced by permanent coronary ligation. Ventricular tachyarrhythmias were evaluated for 24 h following ligation, using a miniature telemetry electrocardiogram recorder. Action potential duration was measured from monophasic epicardial recordings and sympathetic activation was assessed by heart rate variability and catecholamine serum level measurements. Compared to controls (1012 ± 185 s), bosentan (59 ± 24 s) markedly decreased (P b 0.00001) the total duration of ventricular tachyarrhythmias during the delayed (1–24 h) phase post-ligation, with a modest effect during the early (0–1 h) phase (132 ± 38 s, versus 43 ± 18 s, respectively, P = 0.053). Treatment did not affect infarct size or total mortality. Action potential duration at 90% repolarization prolonged in controls (from 93.1 ± 4.7 ms to 117.6 ± 6.9 ms), displaying increased temporal dispersion (from 4.14 ± 0.45 ms to 10.42 ± 2.51 ms, both P b 0.001), but was preserved in treated animals. Bosentan decreased norepinephrine, but increased epinephrine levels 24 h post-ligation. Low frequency spectra of heart rate variability, an index of net sympathetic tone, were lower in bosentan-treated rats. Dual endothelin-1 receptor blockade decreases ventricular tachyarrhythmias during myocardial infarction without reperfusion, by preventing repolarization inhomogeneity. Diverse treatment effects on sympathetic activation may ameliorate the antiarrhythmic action. © 2007 Elsevier B.V. All rights reserved. Keywords: Autonomic nervous system; Catecholamines; Endothelin-1; Myocardial infarction; Ventricular tachyarrhythmias

1. Introduction Myocardial infarction remains a leading cause of death worldwide. Ventricular tachyarrhythmias, namely ventricular tachycardia and ventricular fibrillation, account for a substantial

⁎ Corresponding author. Tel.: +30 265 1097227; fax: +30 265 1097053. E-mail address: [email protected] (T.M. Kolettis). 1 Supported, in part, by a grant from the Cardiovascular Research Institute, Ioannina and Athens, Greece. 2 Supported, in part, by a grant from the national General Council for Research and Technology, Athens, Greece. 0014-2999/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.ejphar.2007.11.002

proportion of acute mortality, with rates remaining stable over the past years (Zheng et al., 2002). Several mechanisms are operative in the genesis of ischemic ventricular tachyarrhythmias, including the accumulation of neurohumoral substances in the myocardium (Clements-Jewery et al., 2005). Endothelin-1 rises markedly during acute myocardial infarction (Loennechen et al., 2001; Stewart et al., 1991) and the magnitude of this rise correlates with prognosis (Yasuda et al., 1990). Endothelin-1 plays a role in arrhythmogenesis, as indicated by a decrease in the early incidence of post-infarction ventricular tachyarrhythmias after endothelin receptor blockade (Raschack et al., 1998; Baltogiannis et al., 2005). The effects of endothelin-1 are mediated via activation of two (ETA and ETB)

242

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249

different G-protein coupled receptors (Molenaar et al., 1993), but the role of ETB receptors on arrhythmogenesis during acute myocardial infarction remains controversial. Antiarrhythmic actions of dual receptor blockade were found in some studies (Douglas et al., 1998; Geshi et al., 1999), but not in others (Richard et al., 1994). Moreover, the selective endothelin ETB receptor agonist sarafotoxin was reported to decrease ischemic ventricular tachyarrhythmias in rats (Crockett et al., 2000), but in isolated rabbit hearts ETB receptor stimulation had no detectable electrophysiological effects either during normal, or during ischemic conditions (McCabe et al., 2005). Given the important pathophysiologic role of endothelin-1 in ischemic heart disease, there is a growing pharmacological research on the effects of endothelin receptor blockade in this setting. To this end, both selective and dual endothelin receptor blockade have been advocated. Although most of the detrimental effects of endothelin-1 appear to be mediated by the ETA receptor, selective antagonists increase endothelin-1 levels and may adversely affect left ventricular function and pulmonary vascular remodeling (Hu et al., 1998; Nguyen et al., 1998; Fraccarollo et al., 2002). In contrast, dual endothelin receptor blockade was reported to ameliorate ventricular dilatation and dysfunction and to decrease endothelin-1 expression (Clozel et al., 2002; Fraccarollo et al., 1997; Oie et al., 1998). Moreover, in contrast to selective ETA receptor blockade, dual endothelin receptor blockade has been shown to decrease mortality in the chronic phase of experimental myocardial infarction in rats (Mulder et al., 1997). Therefore, the purpose of the present study was to examine the effects of bosentan, a dual endothelin receptor peptide antagonist, (a) on the infarct size and (b) on ventricular tachyarrhythmias. All previously available data on the effects of endothelin-1 receptor blockade on ventricular tachyarrhythmias are limited to the 60–120 min following myocardial infarction generation. An important novelty of the present work is the extension of the observation time window; in this study, we examined the incidence of ventricular tachyarrhythmias, occurring during a 24-h observation period after permanent coronary occlusion in the conscious rat. Moreover, the effects of endothelin-1 receptor blockade on catecholamine release are unclear. To provide further insight into possible pathophysiologic mechanisms, we examined the effects of bosentan on the monophasic action potential and on indices of sympathetic activation. 2. Materials and methods

Declaration of Helsinki and with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the United States National Institutes of Health (Publication No. 85-23, revised 1996). The study protocol was approved by the local state authority. 2.2. Drug administration Bosentan was a kind gift from Actelion Pharmaceuticals Ltd., Allschwil, Switzerland. Fresh preparations were made every day as a micro-suspension in 5% arabic gum, containing bosentan (100 mg/kg body weight) and were administered by gavage once daily for 7 days. This dosage has been shown to produce significant and sustained pharmacologic effects (Fraccarollo et al., 1997; Mulder et al., 1997). The control group received the same quantity of vehicle for the same time period. 2.3. Implantation of telemetry transmitter Six days after randomization, a continuous electrocardiogram telemetry transmitter (Dataquest, Data Sciences International, Transoma Medical, Arden Hills, Minnesota, USA) was implanted in the abdominal cavity, after slight modification of the previously described technique (Opitz et al., 1995). Under brief ether anesthesia, the animals were intubated, mechanically ventilated (ventilator model 7025, Ugo Basile, Comerio, Italy) and anesthetized with 2% isoflurane. The use of this agent allows the rats to regain consciousness rapidly (within 2–3 min) after cessation of anesthesia. All animals were housed in cages, placed on a receiver continuously capturing the electrocardiogram signal, independent of animal activity. 2.4. Monophasic action potential recordings A monophasic action potential probe (model 200, EP Technologies, Sunnyvale, California, USA) was placed on the epicardium, exerting mild, constant pressure, as previously described (Franz, 1999). The signal was amplified (preamplifier model 300, EP Technologies, Sunnyvale, California, USA), filtered at 50 Hz with a digital notch filter and for ranges b0.05 Hz and N 500 Hz with a band pass filter. The recordings were stored in a personal computer, equipped with an analog-todigital converter (BNC 2110, National Instruments Corporation, Dallas, Texas, USA). Monophasic action potential signals were recorded from the lateral left and right ventricular walls immediately prior to and 24 h after myocardial infarction generation.

2.1. Experimental design 2.5. Generation of myocardial infarction The study was conducted in 56 Wistar rats of similar age and weight (20 ± 1 weeks old, 200–250 g, respectively). The rats were randomized in 1:1 fashion to receive either bosentan or vehicle. This sample size enables an approximately 90% power to detect a 50% decrease in ventricular tachyarrhythmias. The animals were housed in individual cages, under optimal laboratory conditions (controlled temperature, humidity and light/ dark cycles) and were given water and standard rat chow ad libitum. The research was conducted in accordance with the

Seven days after randomization, myocardial infarction was generated as previously described (Pfeffer et al., 1979), by an operator blinded to treatment assignment. The left coronary artery was ligated using a 6–0 suture, placed between the pulmonary artery cone and the left atrial appendage. A six-lead electrocardiogram was obtained and ST-segment elevation was considered proof of induced MI. In a further group of 6 animals, a sham operation was performed, by encircling but without

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249

243

We analyzed 5-min continuous electrocardiogram recordings, from which non-sinus beats were excluded and the mean value of RR intervals was used. Heart rate was calculated at baseline, at the 5th, 30th and 60th min post-ligation and hourly thereafter.

ment. The significance of premature ventricular contractions, couplets and triplets is debated (Opitz et al., 1995), hence such count was omitted. Ventricular tachycardia and ventricular fibrillation episodes were recorded and the duration of each episode was measured using the time-scale provided by the software. According to the guidelines provided by the Lambeth Conventions for determination of experimental arrhythmias, ventricular tachycardia was defined as 4 or more consecutive premature ventricular contractions and ventricular fibrillation as a signal with indistinguishable QRS deflections (Walker et al., 1988). Despite these guidelines, separation of ventricular tachycardia and ventricular fibrillation is often difficult (Opitz et al., 1995), hence we report these two arrhythmias collectively, as ‘ventricular tachyarrhythmias’. Since different mechanisms are thought to be operative for ventricular tachyarrhythmias occurring during various time periods post-infarction (ClementsJewery et al., 2005), we recorded the number and duration of ventricular tachyarrhythmia episodes for each hourly interval post-ligation. We used two previously established methods for arrhythmia analysis (Curtis and Walker, 1988; Opitz et al., 1995). In the first method (Opitz et al., 1995), we calculated the duration of ventricular tachyarrhythmias for each hourly interval, as the sum of durations of each episode recorded during this interval. The distribution of the hourly duration of ventricular tachyarrhythmias is reported for the entire (24-h) observation period. To provide insight into possible mechanisms, ventricular tachyarrhythmias episodes were separated in those occurring in the early post-ligation period (phase I, i.e. during the first h) and in the late post-ligation period (phase II, i.e. from the second h to the end of the recording or to spontaneous death) (Clements-Jewery et al., 2005). Censoring effects, due to differences in mortality rates and timing, may confound the results (Opitz et al., 1995), hence the hourly duration of ventricular tachyarrhythmias was normalized to survival time (i.e. the time at risk of tachyarrhythmia occurrence). In the second method, a simpler quantification provided by the arrhythmia score was used (Curtis and Walker, 1988). A score of 2 was given for one spontaneously reverting ventricular tachyarrhythmia episode and 3 for two or more episodes, with a total combined duration of b60 s. A score of 4 was given for episodes with a combined total duration of 60–119 s, 5 for ventricular tachyarrhythmias of a combined duration of N 119 s, 6 for fatal ventricular tachyarrhythmias starting at b 15 min post-ligation, 7 for fatal ventricular tachyarrhythmias starting between 4 min and 14 min, 8 for fatal ventricular tachyarrhythmias between 1 and 3 min, and 9 for fatal episodes starting b1 min post-ligation. The arrhythmia score was calculated for four post-ligation time periods, i.e., 0–1 h, 1–6 h, 6–12 h, and 12–24 h. This method also accounts for differences in mortality rates and timing, by giving a score of 9 for the time period(s) following death.

2.10. Arrhythmia analysis

2.11. Mortality

The stored electrocardiographic tracings were analyzed independently by two of the authors, blinded to treatment assign-

Tachyarrhythmic death was defined as ventricular asystole, immediately preceded by an episode of ventricular

ligating the left coronary artery. After the procedure, electrocardiogram recording was continued for 24 h or until spontaneous death. No resuscitation attempts were allowed at any time during the study. 2.6. Catecholamine level measurements Twenty-four h post-ligation, the animals were re-anesthetized and the internal jugular vein was exposed. Blood was collected by venous puncture, centrifuged and the serum was stored at − 20 °C. Levels of epinephrine and norepinephrine were measured using radioimmunoassay (BioSource Europe S.A., Nivelles, Belgium). Subsequently, the site of previous thoracotomy was re-opened for monophasic action potential recordings. 2.7. Infarct size The rats were sacrificed using a lethal dose of potassium chloride and infarct size was measured as described previously (Ytrehus et al., 1994). The heart was excised, frozen (in − 20 °C for one hour), hand-cut in five 2 mm-slices that were incubated (in triphenyltetrazolium chloride for 15 min at 37 °C) and fixed (in 10% formalin for 20 min). After scanning (Scanjet 4570c/ 5500c, Hewlett-Packard, Palo Alto, California, USA), the areas of infarcted and non-infarcted myocardium were measured from both slice sides and averaged. Planimetry was performed using a previously validated software program (Image Tool, University of Texas, USA). The volumes of infarcted and noninfarcted myocardium (defined as measured areas multiplied by slice thickness) were calculated for each slice and summed. Infarct size (expressed as a percentage) was defined as the ratio between infarcted and total left ventricular volume. 2.8. Analysis of monophasic action potentials Arrhythmic recordings with unstable resting potential or with electrical artifacts were excluded. The software utilized for monophasic action potential analysis, developed and validated at our Institution (Tsalikakis et al., 2003), permits signal acquisition and off-line analysis. Fifty sinus beats per tracing were analyzed and the action potential duration at 90% of repolarization was measured at baseline and 24 h post-ligation. The standard deviation of the action potential duration at 90% of repolarization was calculated for each tracing, as a measure of beat-to-beat variation, indicating electrical alternans (Franz, 1999). 2.9. Sinus heart rate

244

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249

Fig. 1. Study protocol. Footnote: ECG = electrocardiogram, HRV = heart rate variability, MAP = monophasic action potential, MI = myocardial infarction.

tachyarrhythmia and bradyarrhythmic death as ventricular asystole, immediately preceded by bradycardia (b200 bpm) associated with complete atrioventricular block. 2.12. Heart rate variability Heart rate variability for the assessment of cardiac autonomic status in conscious rats has been described previously in detail (Kruger et al., 1997). We used 5-min electrocardiographic segments after exclusion of non-sinus beats. Time domain parameters included the S.D. of RR intervals, the root mean square of S.D. and the coefficient of variance (S.D./RR). In the frequency domain, the fast Fourier transform power spectrum was calculated using the Welch periodogram, by dividing the time series into a constant number of segments, overlapping by 50%. After application of a Hanning window and subtracting the mean value, the segment periodogram was calculated and the power spectra of all segments were averaged. Low-frequency (N0.5 Hz b 0.8 Hz) and high-frequency (N 0.8 Hz) bands were calculated. The percentage of peak power in each band and their ratio are reported. The study protocol is depicted in Fig. 1. 2.13. Statistical analysis All values are given as mean ± S.E.M. Mortality rates were compared using two-tailed Fisher's exact test, while infarct size was compared using Student's t-test. Differences in heart rate and action potential duration were compared using the analysis of variance for repeated measures, followed by Duncan's multirange test. The non-parametric Mann–Whitney U-test was used for comparisons of catecholamine concentrations, as well as for ventricular tachyarrhythmias frequencies and durations. Differences in the hourly incidence of ventricular tachyarrhythmias were assessed using the Kruskal–Wallis analysis of variance by ranks. Statistical significance was defined at an alpha level of 0.05.

and were excluded. Thus, the final animal study population consisted of 52 rats, of which 24 (223 ± 3 g) were allocated bosentan and 23 (224 ± 3 g) were allocated vehicle, while 5 rats (224 ± 4 g) were sham-operated. During the 24-h period postligation, 3 animals in the bosentan group (12.5%, all during phase I), and 8 (34.7%, 3 during phase I and 5 during phase II) in the control group had fatal ventricular tachyarrhythmias (P = 0.093). Two (8.3%, 1 during phase I and 1 during phase II) further rats in the bosentan group and 1 (4.3%, during phase II) further rat in the control group died of bradyarrhythmia. The overall mortality did not differ significantly between the two groups (P = 0.21). 3.2. Infarct size and sinus heart rate Infarct size was comparable between groups (37.9 ± 1.5% in bosentan-treated and 39.1 ± 1.4% in control rats, P = 0.57). Sinus heart rate increased significantly over time (F = 5.93, P b 0.0001, Fig. 2), but without differences between groups (F = 0.91, P = 0.38). An increase was also present in shamoperated animals (F = 2.0, P = 0.007), probably attributable to the procedure, with sinus heart rate returning to baseline values after the third h of recording. 3.3. Number and duration of ventricular tachyarrhythmia episodes Apart from scarce premature ventricular contractions, no ventricular tachyarrhythmias were recorded in sham-operated animals. Significantly (P = 0.0017) fewer ventricular tachyarrhythmia episodes were found during phase II in bosentantreated rats (46 ± 20) than in controls (248 ± 92). In contrast, the number of phase I ventricular tachyarrhythmia episodes did not differ (P = 0.178) between the bosentan-treated (7 ± 2) and the control (19 ± 4) groups. The mean duration of each episode was almost identical in the two groups (14.1 ± 7.6 s and 14.1 ± 4.8 s, respectively). 3.4. Hourly duration of ventricular tachyarrhythmias during phases I and II When both phases were considered together, the hourly duration of ventricular tachyarrhythmias was shorter

3. Results 3.1. Animal population and mortality Fifty animals were chosen, of which 26 were randomized to bosentan and 24 to vehicle; 6 further animals were shamoperated. Of the 56 animals, 4 died during the procedure

Fig. 2. Sinus heart rate. Footnote: No differences in sinus heart rate (in beats per minute, bpm) were present between bosentan-treated rats and controls.

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249

245

Table 1 Changes in action potential duration and temporal variability over time Bosentan LV a APD90 b baseline (ms) LV a APV90 b 24 h post-MI c (ms) Beat-to-beat variability LV a APD b baseline (ms) Beat-to-beat variability LV a APD b 24 h post-MI c (ms) a b c

Fig. 3. Hourly distribution of ventricular tachyarrhythmias post-coronary ligation. Footnote: The horizontal arrow on the x axis indicates significant differences between bosentan and controls. MI = myocardial infarction, VT = ventricular tachycardia, VF = ventricular fibrillation.

(P b 0.00001) in the bosentan group (92 ± 28 s), compared to controls (1013 ± 179 s). This was due to a decrease (P b 0.00001) in the hourly duration of phase II ventricular tachyarrhythmias in the bosentan group (59 ± 24 s), compared to controls (1012 ± 185 s). During phase I, there was a trend (P = 0.053) towards a decrease in bosentan-treated rats (43 ± 18 s), compared to controls (132 ± 38 s). Similar results were obtained after normalization for survival times. For phase I, the trend (P = 0.082) persisted, in favor of bosentan (168 ± 120 s) compared to controls (253 ± 101 s). For phase II, values remained significantly (P b 0.00001) shorter in bosentan-treated rats (3 ± 1 s), than in controls (104 ± 28 s). 3.5. Distribution of the hourly duration of ventricular tachyarrhythmias The results over the entire 24-h period are shown in Fig. 3. Values of ventricular tachyarrhythmia hourly duration were significantly shorter in the bosentan group between the second and eleventh h post-ligation. 3.6. Arrhythmia score Arrhythmia score was similar between 0–1 h (phase I, P = 0.26) and 12–24 h (P = 0.23) post-ligation in the two groups. Differences in favor of bosentan were found between 1–6 h (P = 0.000495) and 6–12 h (P = 0.001073), as shown in Fig. 4.

d

LV: left ventricular. APD90: action potential duration at 90% of repolarization. MI: myocardial infarction. P b 0.05 versus baseline.

3.7. Monophasic action potential duration Of the 33 survivors, analyzable signals were available from 23 animals. No variance (F = 2.12, P = 0.17) was present in right ventricular action potential duration, or in the beat-to-beat variability of the action potential duration (F = 1.52, P = 0.23). There was a significant (F = 8.58, P = 0.0137) variance in left ventricular action potential duration, due to an increase (P = 0.00063) in values in the control group 24 h post-ligation, while values remained constant (P = 0.44) in the bosentan group (Table 1). Furthermore, a statistical variance (F = 6.61, P = 0.018) was found in beat-to-beat variability of left ventricular action potential duration, due to an increase (P = 0.00051) from baseline values in the control group. Beat-to-beat variability remained unchanged (P = 0.52) in the bosentan group (Table 1). 3.8. Catecholamine levels Norepinephrine levels 24 h post-ligation were lower (P b 0.00001) in bosentan-treated rats (10.2 ± 5.4 μg/l), compared to controls (74.0 ± 1.8 μg/l). However, an opposite effect was found in epinephrine levels, being higher (P = 0.00631) in the bosentan (20.0 ± 4.9 μg/l), compared to the control group (5.2 ± 0.6 μg/l).

Table 2 Frequency domain parameters of heart rate variability post-ligation

Low frequency (% 3 min 30 min 90 min 4.5 h 8.5 h 24 h

Control

Bosentan

Sham

of peak) 21.8 ± 3.5 35.2 ± 8.4 23.3 ± 4.8 32.4 ± 6.3 28.2 ± 5.3 33.3 ± 6.1

11.9 ± 1.7 a 36.5 ± 7.1 b 18.0 ± 3.4 18.6 ± 2.3 a 22.9 ± 3.9 18.1 ± 3.4 a

6.3 ± 2.9 a 10.3 ± 3.4 15.3 ± 2.5 11.3 ± 1.4 a 3.9 ± 0.8 11.5 ± 2.0 a

38.9 ± 4.7 41.2 ± 3.5 46.5 ± 3.7 41.3 ± 3.7 46.0 ± 3.9 55.2 ± 4.5

29.2 ± 5.9 26.4 ± 4.2 42.5 ± 4.7 33.7 ± 3.4 17.9 ± 7.1 29.5 ± 2.8

High frequency (% of peak) 3 min 42.5 ± 5.3 30 min 41.9 ± 5.3 90 min 41.0 ± 3.9 4.5 h 44.6 ± 7.5 8.5 h 54.0 ± 10.3 24 h 39.8 ± 9.9 Fig. 4. Arrhythmia score. Footnote: The horizontal arrow on the x axis indicates significant differences between bosentan and controls.

Control

94.1 ± 3.6 93.1 ± 4.7 97.6 ± 7.2 117.6 ± 6.9 d 3.46 ± 0.32 4.14 ± 0.45 4.46 ± 0.54 10.42 ± 2.51 d

a b

P b 0.05 versus controls. P b 0.05 versus 3 min.

246

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249

Fig. 5. Heart rate variability. Footnote: Cardiac autonomic tone expressed by the ratio of low frequency (LF) to high frequency (HF) spectra in heart rate variability analysis. Small asterisk indicates a trend (P = 0.07) and bigger asterisk a significant difference between bosentan and controls.

3.9. Heart rate variability No significant differences were found in the time domain (data not shown). In the frequency domain, the low frequency spectrum and the ratio low frequency to high frequency spectra were lower in bosentan-treated animals, at 3 min, 4.5 h and 24 h post-ligation (Table 2, Fig. 5). 4. Discussion 4.1. Main findings and comparison with previous studies We examined the effects of bosentan, a dual endothelin receptor peptide antagonist, on arrhythmogenesis in the rat model of permanent coronary ligation. This model is well characterized, as the rat exhibits a high frequency of ventricular tachyarrhythmias post-ligation (Opitz et al., 1995), with a time course resembling that observed in humans post-myocardial infarction (Clements-Jewery et al., 2005; Fenoglio et al., 1979). We report a reduction in ventricular tachyarrhythmias in bosentan-treated animals. Previous relevant studies have yielded conflicting results (Douglas et al., 1998; Geshi et al., 1999; Richard et al., 1994). Apart from species differences, diversity in pharmacological agents used and protocol designs, the explanation for the diversity in the previously reported results may be threefold: First, the presence or absence of reperfusion. Several differences exist with respect to the mechanisms and temporal distribution between ischemia- and reperfusion-induced ventricular tachyarrhythmias. To our knowledge, the present study is the first to investigate the antiarrhythmic potential of dual endothelin receptor blockade in an in vivo model of permanent coronary ligation. Our results are clinically relevant, given recent reports emphasizing the substantial proportion of acute myocardial infarction patients currently not being treated with reperfusion (Rosengren et al., 2004). Second, the arrhythmogenic potential of endothelin-1 was previously examined after exogenously administered endothelin-1 in some studies (Szokodi et al., 1998) and from the actions of endogenous endothelin-1 in others (Alberola Aguilar et al., 2000; Raschack et al., 1998). However, it is been shown that these actions are substantially different (Sharif et al., 1998), suggesting that only the effects of endogenous endothelin-1 are

clinically relevant. Third and foremost, all previously published studies reported only on early (phase I) ventricular tachyarrhythmias in anesthetized animals. Indeed, in our experiments, we observed only a modest reduction in ventricular tachyarrhythmias in bosentan-treated rats during this time frame. A major strength of our study is the utilization of the telemetry recording system, which facilitates the prolongation of the observation period without the confounding effects of anesthesia. We report a prominent antiarrhythmic effect of bosentan during phase II (between the second and eleventh h postligation). The difference in the antiarrhythmic effect of bosentan between the two phases is further underscored by the fact that all tachyarrhythmic deaths were observed during phase I in treated rats. During myocardial infarction, endothelin-1 production is increased in rats (Loennechen et al., 2001) and humans (Yasuda et al., 1990), but its exact time course is not well defined (Stewart et al., 1991). Progressive rises in endothelin-1 may explain the more pronounced antiarrhythmic effect of bosentan observed during phase II in our experiments. 4.2. Antiarrhythmic mechanism of bosentan An ongoing controversy exists, whether the arrhythmogenic effect of endothelin-1 is direct, a consequence of ischemia, or a combination of both (Duru et al., 2001). Previous studies demonstrating a reduction in infarct size after bosentan administration have utilized the ischemia/reperfusion model (Wang et al., 1995). In our setting of permanent coronary ligation, no difference in infarct size was found, and similar results were reported in an identical model, after acute administration of the dual endothelin receptor antagonist tezosentan (Clozel et al., 2002). The decrease in the incidence of ventricular tachyarrhythmias without a decrease in infarct size, points towards a direct antiarrhythmic effect of bosentan. Although direct arrhythmogenic actions of endothelin-1 were demonstrated long ago (Yorikane et al., 1991), the underlying mechanisms remain unclear (Duru et al., 2001). In vitro, endothelin-1 progressively increased the occurrence of spontaneous calcium transients, via stimulation of the inositol 1,4,5triphosphate pathway, leading into trains of events (Proven et al., 2006). In vivo, early afterdepolarizations were increasingly observed after low dose endothelin-1 intracoronary infusion (Merkely et al., 1998). These findings provide additional explanation for the pronounced antiarrhythmic effects of bosentan during phase II in our study and allow the generation of hypotheses on the underlying electrophysiological mechanisms. Endothelin-1 prolongs the action potential duration (Yorikane et al., 1991) and may produce focal ventricular tachyarrhythmias via triggered activity (Merkely et al., 1998). On the other hand, the ionic changes on cell membrane of ischemic cardiomyocytes display spatial inhomogeneity (Fenoglio et al., 1979), facilitating reentry (Clements-Jewery et al., 2005). Although the differences between endothelin-1- and ischemiainduced ventricular tachyarrhythmias have been clearly delineated, an additive or synergistic effect is likely (Becker et al., 2000). Our study suggests that the preservation of action

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249

potential duration after bosentan treatment may have aborted triggered activity, as well as ischemic reentrant mechanisms. 4.3. Are endothelin ETB receptors involved in arrhythmogenesis? Most studies previously reporting an antiarrhythmic action after endothelin receptor blockade have used selective ETA receptor agents (Raschack et al., 1998; Duru et al., 2001). In the present study, the reduction in ventricular tachyarrhythmias in bosentan-treated rats was similar to previous results after selective ETA receptor blockade in an identical experimental setting (Baltogiannis et al., 2005). Taken together, these findings suggest that the antiarrhythmic effects of endothelin-1 receptor blockade are mainly attributed to ETA receptor blockade and ETB receptors are not substantially involved in ventricular arrhythmogenesis. This conclusion gains importance, given the scarcity of in vivo studies and concurs with data from in vitro experiments (Alexiou et al., 1998; Kelso et al., 1998; McCabe et al., 2005). In isolated working rabbit hearts, the endothelin ETB receptor agonist sarafotoxin had a neutral effect on ventricular tachyarrhythmias, while ETA receptor activation decreased the ventricular effective refractory period and facilitated the induction of ventricular fibrillation (McCabe et al., 2005). In a similar setting, endogenous endothelin-1 displayed a direct arrhythmogenic action that was partially suppressed by ETA receptor antagonists, while ETB receptor blockade was ineffective (Alexiou et al., 1998). 4.4. Endothelin-1 and sympathetic activation The importance of sympathetic activation in modulating the occurrence of ischemic ventricular tachyarrhythmias is well established (Corr and Gillis, 1978). Similarly, the link between endothelin-1 and the sympathetic system was reported long ago (Wennmalm et al., 1989), but their interactions during myocardial infarction and the resultant effects on arrhythmogenesis remain unclear. Early catecholamine release contributes to the genesis of early and delayed after depolarizations, triggered activity and ventricular tachyarrhythmias (Schomig et al., 1991). Endothelin-1 is involved in the adrenal gland function, but ETA and ETB receptors may have diverse actions in this regard (Nagayama et al., 2000). In an in vitro model, catecholamine output was inhibited by selective ETA receptor blockade, while the ETB receptor antagonist BQ-788 was ineffective. However, the inhibitory effect of ETA receptor blockade was abolished by pretreatment with BQ-788 (Nagayama et al., 2000). These observations may explain some of our findings. Although bosentan blunted the very early sympathetic response, ETB receptor blockade might have negated the expected inhibition of the main catecholamine surge, caused by ETA receptor blockade. In other words, during phase I, bosentan had a rather neutral effect on central catecholamine release, resulting in almost identical heart rate variability indices of sympathetic activation 30 min post-ligation. In keeping with previous findings (Crockett et al., 2000), this may explain the modest antiarrhythmic effect observed during this phase. During phase

247

II, the role of sympathetic activation becomes more complex, due to norepinephrine accumulation into the axoplasm and in non-exocytotic local metabolic release (Schomig et al., 1991). These changes, occurring to a different extent in various myocardial regions (Li et al., 2004), lead to increased norepinephrine in the interstitium of ischemic myocardium and in the plasma due to spillover (Lameris et al., 2000; Li et al., 2004). In our experiments, we found a marked decrease in serum norepinephrine levels 24 h post-ligation in treated rats, indicating another antiarrhythmic mechanism of bosentan. Moreover, heart rate variability indices of sympathetic activation were decreased in treated rats, 4.5 and 24 h post-ligation. These findings are in accordance with recent evidence that both endothelin receptors are located in the left ventricular sympathetic nerve varicosities and modulate norepinephrine release (Isaka et al., 2007); in isolated ischemic guinea pig hearts, exogenously applied endothelin-1 increased norepinephrine release and ventricular tachyarrhythmias in a dose dependent manner. These effects were attenuated by selective ETA, as well as by dual receptor blockade (Isaka et al., 2007). These findings, when examined together with our results, indicate that the link between endothelin-1 and the sympathetic system is via activation of the ETA receptor, but dual endothelin receptor blockade confers a decrease in sympathetic activity during the delayed phase post-coronary occlusion. We report increased epinephrine levels 24 h post-ligation in bosentan-treated rats. The explanation for this diverse effect of bosentan on catecholamines may be twofold: First, endothelin1 has positive inotropic effects providing short-term support during myocardial infarction; inhibition of this support might have caused additional sympathetic activation. Nonetheless, the hemodynamic response after endothelin-1 receptor blockade is complex and conflicting results have been reported (Clozel et al., 2002; Nguyen et al., 2001). Second, bosentan, at the dosage used in the present study, produces prominent and sustained vasodilatory effects (Fraccarollo et al., 1997; Mulder et al., 1997). Thus, a reflex sympathetic activation may occur (Gelzer et al., 2004) and may account for the comparable sinus heart rate seen in our experiments. However, the net balance after the diverse effects of bosentan on endogenous catecholamines appears to be a decrease in sympathetic activation, as evidenced by heart rate variability indices. Importantly, the dose of bosentan used in the present study is high when compared to doses used clinically. Such high dosage was selected to compensate for the higher metabolism in rats and has been previously shown to produce maximal pharmacologic effects (Fraccarollo et al., 1997; Mulder et al., 1997). Nonetheless, our results indicate that future studies should evaluate the effects of lower pharmacological dosages of endothelin receptor blockade in the post-myocardial infarction setting. Such a strategy may maintain the antiarrhythmic properties locally on ventricular myocardium, but may avoid the untoward peripheral effects. 4.5. Limitations of the study We feel that our work improves current understanding on the mechanisms of arrhythmogenesis during myocardial infarction.

248

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249

However, three limitations may be apparent. First, the assessment of sympathetic activation would have been more accurate if catecholamine measurements prior to coronary ligation were included. Second, although our sample size was optimal for the assessment of the antiarrhythmic effects of bosentan, our study was underpowered to detect differences in mortality. Third, ventricular remodeling, as well as the development of collateral circulation, both correlating with ventricular tachyarrhythmias, were not examined in our study; however, we feel that such indices as early as 24 h post-ligation, would have added little value to our study. 4.6. Conclusions In the rat model of permanent coronary ligation, bosentan decreases phase II ventricular tachyarrhythmias. Prevention of action potential prolongation and its temporal variability as well as decreased norepinephrine levels are potential mechanisms. Further studies should examine whether these effects are maintained with lower dosage, avoiding systemic effects. The interplay between endothelin-1 and sympathetic activation during ischemia/reperfusion and their effects on ventricular arrhythmogenesis is also subject for future research. Acknowledgements This work was supported, in part, by the national General Council for Research and Technology (project: PENED01ΕD511), Athens, Greece and by the Cardiovascular Research Institute, Ioannina and Athens, Greece. We wish to thank Actelion Pharmaceuticals Ltd., Allschwil, Switzerland, for providing bosentan and Boston Scientific Hellas for providing the monophasic action potential probe and preamplifier. Tzihad Albouharali, MD, performed all radioimmunoassay measurements. Panagiotis Lekkas and Anastasia Alevizatou assisted during the experiments. Eleni Goga, MSc offered invaluable help as a research coordinator. References Alberola Aguilar, A.M., Revert, F., Moya, A., Beltran, J., Garcia, J., San Martin, E., Sancho, S., Such, L., 2000. Intravenous BQ-123 and phosphoramidon reduce ventricular ectopic beats and myocardial infarct size in dogs submitted to coronary occlusion and reperfusion. Gen. Pharmacol. 35, 143–147. Alexiou, K., Dschietzig, T., Simsch, O., Laule, M., Hundertmark, J., Baumann, G., Stangl, K., 1998. Arrhythmogenic effects induced by coronary conversion of pulmonary big endothelin to endothelin: aggravation of this phenomenon in heritable hyperlipidemia. J. Am. Coll. Cardiol. 32, 1773–1778. Baltogiannis, G.G., Tsalikakis, D.G., Mitsi, A.C., Hatzistergos, K.E., Elaiopoulos, D., Fotiadis, D.I., Kyriakides, Z.S., Kolettis, T.M., 2005. Endothelin receptor-A blockade decreases ventricular arrhythmias after myocardial infarction in rats. Cardiovasc. Res. 67, 647–654. Becker, R., Merkely, B., Bauer, A., Geller, L., Fazekas, L., Freigang, K.D., Voss, F., Senges, J.C., Kuebler, W., Schoels, W., 2000. Ventricular arrhythmias induced by endothelin-1 or by acute ischemia: a comparative analysis using three-dimensional mapping. Cardiovasc. Res. 45, 310–320. Clements-Jewery, H., Hearse, D.J., Curtis, M.J., 2005. Phase 2 ventricular arrhythmias in acute myocardial infarction: a neglected target for therapeutic

antiarrhythmic drug development and for safety pharmacology evaluation. Br. J. Pharmacol. 145, 551–564. Clozel, M., Qiu, C., Qiu, C.S., Hess, P., Clozel, J.P., 2002. Short-term endothelin receptor blockade with tezosentan has both immediate and long-term beneficial effects in rats with myocardial infarction. J. Am. Coll. Cardiol. 39, 142–147. Corr, P.B., Gillis, R.A., 1978. Autonomic neural influences on the dysrhythmias resulting from myocardial infarction. Circ. Res. 43, 1–9. Crockett, T.R., Sharif, I., Kane, K.A., Wainwright, C.L., 2000. Sarafotoxin 6c protects against ischaemia-induced cardiac arrhythmias in vivo and in vitro in the rat. J. Cardiovasc. Pharmacol. 36, S297–S299. Curtis, M.J., Walker, M.J.A., 1988. Quantification of arrhythmias using scoring systems: an examination of seven scores in an in vivo model of regional myocardial ischemia. Cardiovasc. Res. 22, 656–665. Douglas, S.A., Nichols, A.J., Feuerstein, G.Z., Elliott, J.D., Ohlstein, E.H., 1998. SB 209670 inhibits the arrhythmogenic actions of endothelin-1 in the anesthetized dog. J. Cardiovasc. Pharmacol. 31 (Suppl 1), S99–S102. Duru, F., Barton, M., Luescher, T.F., Candinas, R., 2001. Endothelin and cardiac arrhythmias: do endothelin antagonists have a therapeutic potential as antiarrhythmic drugs? Cardiovasc. Res. 49, 272–280. Fenoglio Jr., J.J., Karagueuzian, H.S., Friedman, P.L., Albala, A., Wit, A.L., 1979. Time course of infarct growth toward the endocardium after coronary occlusion. Am. J. Physiol. 236, H350–H356. Fraccarollo, D., Hu, K., Galuppo, P., Gaudron, P., Ertl, G., 1997. Chronic endothelin receptor blockade attenuates progressive ventricular dilation and improves cardiac function in rats with myocardial infarction: possible involvement of myocardial endothelin system in ventricular remodeling. Circulation 96, 3963–3973. Fraccarollo, D., Galuppo, P., Bauersachs, J., Ertl, G., 2002. Collagen accumulation after myocardial infarction: effects of ETA receptor blockade and implications for early remodeling. Cardiovasc. Res. 54, 559–567. Franz, M.R., 1999. Current status of monophasic action potential recording: theories, measurements and interpretations. Cardiovasc. Res. 41, 25–40. Gelzer, A.R., Attmann, T., Radicke, D., Nydam, D., Candinas, R., Lutter, G., 2004. Effects of acute systemic endothelin receptor blockade on cardiac electrophysiology in vivo. J. Cardiovasc. Pharmacol. 44, 564–570. Geshi, E., Nomizo, A., Arata, Y., Nakatani, M., Katagiri, T., 1999. Effect of nonselective endothelin blockade, TAK-044, on the ischemic cellular injury of rat heart. Basic Res. Cardiol. 94, 94–101. Hu, K., Gaudron, P., Schmidt, T.J., Hoffmann, K.D., Ertl, G., 1998. Aggravation of left ventricular remodeling by a novel specific endothelin ET(A) antagonist EMD94246 in rats with experimental myocardial infarction. J. Cardiovasc. Pharmacol. 32, 505–508. Isaka, M., Kudo, A., Imamura, M., Kawakami, H., Yasuda, K., 2007. Endothelin receptors, localized in sympathetic nerve terminals of the heart, modulate norepinephrine release and reperfusion arrhythmias. Basic Res. Cardiol. 102, 154–162. Kelso, E.J., Spiers, J.P., McDermott, B.J., Scholfield, C.N., Silke, B., 1998. Receptor-mediated effects of endothelin on the L-type Ca+ current in ventricular cardiomyocytes. J. Pharmacol. Exp. Ther. 286, 662–669. Kruger, C., Kalenka, A., Haunstetter, A., Schweizer, M., Maier, C., Ruhle, U., Ehmke, H., Kubler, W., Haass, M., 1997. Baroreflex sensitivity and heart rate variability in conscious rats with myocardial infarction. Am. J. Physiol. 273, H2240–H2247. Lameris, T.W., de Zeeuw, S., Alberts, G., Boomsma, F., Duncker, D.J., Verdouw, P.D., Veld, A.J., van den Meiracker, A.H., 2000. Time course and mechanism of myocardial catecholamine release during transient ischemia in vivo. Circulation 101, 2645–2650. Li, W., Knowlton, D., Van Winkle, D.M., Habecker, B.A., 2004. Infarction alters both the distribution and noradrenergic properties of cardiac sympathetic neurons. Am. J. Physiol, Heart Circ. Physiol. 286, H2229–H2236. Loennechen, J.P., Stoylen, A., Beisvag, V., Wisloff, U., Ellingsen, O., 2001. Regional expression of endothelin-1, ANP, IGF-1, and LV wall stress in the infarcted rat heart. Am. J. Physiol, Heart Circ. Physiol. 280, H2902–H2910. McCabe, C., Hicks, M.N., Kane, K.A., Wainwright, C.L., 2005. Electrophysiological and haemodynamic effects of endothelin ETA and ETB receptors in normal and ischaemic working rabbit hearts. Br. J. Pharmacol. 146, 118–128.

T.M. Kolettis et al. / European Journal of Pharmacology 580 (2008) 241–249 Merkely, B., Geller, L., Toth, M., Kiss, O., Kekesi, V., Solti, F., Vecsey, T., Horkay, F., Tenczer, J., Juhasz-Nagy, A., 1998. Mechanism of endothelininduced malignant ventricular arrhythmias in dogs. J. Cardiovasc. Pharmacol. 31 (Suppl 1), S437–S439. Molenaar, P., O'Reilly, G., Sharkey, A., Kuc, R.E., Harding, D.P., Plumpton, C., Gresham, G.A., Davenport, A.P., 1993. Characterization and localization of endothelin receptor subtypes in the human atrioventricular conducting system and myocardium. Circ. Res. 72, 526–538. Mulder, P., Richard, V., Derumeaux, G., Hogie, M., Henry, J.P., Lallemand, F., Compagnon, P., Macé, B., Comoy, E., Letac, B., Thuillez, C., 1997. Role of endogenous endothelin in chronic heart failure: effect of long-term treatment with an endothelin antagonist on survival, hemodynamics, and cardiac remodeling. Circulation 96, 1976–1982. Nagayama, T., Kuwakubo, F., Matsumoto, T., Fukushima, Y., Yoshida, M., Suzuki-Kusaba, M., Hisa, H., Matsumura, Y., Kimura, T., Satoh, S., 2000. Role of endogenous endothelins in catecholamine secretion in the rat adrenal gland. Eur. J. Pharmacol. 406, 69–74. Nguyen, Q.T., Cernacek, P., Calderoni, A., Stewart, D.J., Picard, P., Sirois, P., White, M., Rouleau, J.L., 1998. Endothelin A receptor blockade causes adverse left ventricular remodeling but improves pulmonary artery pressure after infarction in the rat. Circulation 98, 2323–2330. Nguyen, Q.T., Cernacek, P., Sirois, M.G., Calderone, A., Lapointe, N., Stewart, D.J., Rouleau, J.-L., 2001. Long-term effects of nonselective endothelin A and B receptor antagonism in postinfarction rat. Importance of timing. Circulation 104, 2075–2081. Oie, E., Bjonerheim, R., Grogaard, H.K., Kongshaug, H., Smiseth, O.A., Attramadal, H., 1998. ET-receptor antagonism, myocardial gene expression, and ventricular remodeling during CHF in rats. Am. J. Physiol. 275, H868–H877. Opitz, C.F., Mitchell, G.F., Pfeffer, M.A., Pfeffer, J.M., 1995. Arrhythmias and death after coronary artery occlusion in the rat. Continuous telemetric ECG monitoring in conscious, untethered rats. Circulation 92, 253–261. Pfeffer, M.A., Pfeffer, J.M., Fishbein, M.C., Fletcher, P.J., Spadaro, J., Kloner, R.A., Braunwald, E., 1979. Myocardial infarct size and ventricular function in rats. Circ. Res. 44, 503–512. Proven, A., Roderick, H.L., Conway, S.J., Berridge, M.J., Horton, J.K., Capper, S.J., Bootman, M.D., 2006. Inositol 1,4,5-trisphosphate supports the arrhythmogenic action of endothelin-1 on ventricular cardiac myocytes. J. Cell Sci. 119, 3363–3375. Raschack, M., Juchelka, F., Rozek-Schaefer, G., 1998. The endothelin-A antagonist LU 135 252 supresses ischemic ventricular extrasystoles and fibrillation in pigs and prevents hypoxic cellular decoupling. J. Cardiovasc. Pharmacol. 31 (Suppl 1), S145–S148. Richard, V., Kaeffer, N., Hogie, M., Tron, C., Blanc, T., Thuillez, C., 1994. Role of endogenous endothelin in myocardial and coronary endothelial injury

249

after ischaemia and reperfusion in rats: studies with bosentan, a mixed ETA– ETB antagonist. Br. J. Pharmacol. 113, 869–876. Rosengren, A., Wallentin, L., Gitt, K., Behar, S., Battler, A., Hasdai, D., 2004. Sex, age, and clinical presentation of acute coronary syndromes. Eur. Heart J. 25, 663–670. Schomig, A., Haass, M., Richardt, G., 1991. Catecholamine release and arrhythmias in acute myocardial ischaemia. Eur. Heart J. 12, 38–47 (Suppl F). Sharif, I., Kane, K.A., Wainwright, C.L., 1998. Endothelin and ischaemic arrhythmias-antiarrhythmic or arrhythmogenic? Cardiovasc. Res. 39, 625–632. Stewart, D.J., Kubac, K.B., Costello, K.B., Cernacek, P., 1991. Increased plasma endothelin-1 in the early hours of acute myocardial infarction. J. Am. Coll. Cardiol. 18, 38–43. Szokodi, I., Horkay, F., Merkely, B., Solti, F., Geller, L., Kiss, P., Selmeci, L., Kekesi, V., Vuolteenaho, O., Ruskoaho, H., Juhasz-Nagy, A., Toth, M., 1998. Intraepicardial infusion of endothelin-1 induces ventricular arrhythmias in dogs. Cardiovasc. Res. 38, 356–364. Tsalikakis, D.G., Fotiadis, D.I., Kolettis, T., Michalis, L.K., 2003. Automated system for the analysis of heart monophasic action potentials. Comput. Cardiol. 30, 339–342. Walker, M.J., Curtis, M.J., Hearse, D.J., Campbell, R.W., Janse, M.J., Yellon, D.M., Cobbe, S.M., Coker, S.J., Harness, J.B., Harron, D.W., 1988. The Lambeth conventions: guidelines for the study of arrhythmias in ischaemia infarction, and reperfusion. Cardiovasc. Res. 22, 447–455. Wang, Q.D., Li, X.S., Lundberg, J.M., Pernow, J., 1995. Protective effects of non-peptide endothelin receptor antagonist bosentan on myocardial ischaemic and reperfusion injury in the pig. Cardiovasc. Res. 29, 805–812. Wennmalm, A., Karwatowska-Prokopczuk, E., Wennmalm, M., 1989. Role of the coronary endothelium in the regulation of sympathetic transmitter release in isolated rabbit hearts. Acta Physiol. Scand. 136, 81–87. Yasuda, M., Kohno, M., Tahara, A., Itagane, H., Toda, I., Akioka, K., Teragaki, M., Oku, H., Takeuchi, K., Takeda, T., 1990. Circulating immunoreactive endothelin in ischemic heart disease. Am. Heart J. 119, 801–806. Yorikane, R., Koike, H., Miyake, S., 1991. Electrophysiological effects of endothelin-1 on canine myocardial cells. J. Cardiovasc. Pharmacol. 17 (Suppl 7), S159–S162. Ytrehus, K., Liu, Y., Tsuchida, A., Miura, T., Liu, G.S., Yang, X.M., Herbert, D., Cohen, M.V., Downey, J.M., 1994. Rat and rabbit heart infarction: effects of anesthesia, perfusate, risk zone, and method of infarct sizing. Am. J. Physiol 267, H2383–H2390. Zheng, Z.J., Croft, J.B., Giles, W.H., Mensah, G.A., 2002. Sudden cardiac death in the United States, 1989 to 1998. Circulation 104, 2158–2163.