INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

ISCHEMIC PRECONDITIONING (IP) as a cardioprotective phenomenon has been widely explored in various experimental models since 1986 (23). The preconditioning was classically induced by single or repetitive brief episodes of regional ischemia (RI) in the heart itself, and it protected against deleterious effects of a subsequent, prolonged ischemia in the same area. The benefits described were a decrease in myocardial damage (15, 18, 24, 37), improvement of functional recovery (34), and antiarrhythmic effects (10, 25, 31).

Cardioprotection by ischemia at distance within the heart itself was induced by brief occlusion of one of the coronary arteries followed by prolonged occlusion of another coronary artery (28).

However, to the best of our knowledge, it is not yet known whether preconditioning effects, such as a decrease in the reperfusion-induced arrhythmia, could only be obtained by previous brief ischemic episodes of the heart itself or also by extracardiac ischemia, i.e., ischemia of another organ or of a remote part of the body. Gho et al. (9) presented data on IP with transient renal ischemia in the rat; the beneficial effect was manifested by a decrease in myocardial infarct size.

We hypothesized that preconditioning might occur as a result of a neurohumoral stress reaction, which is independent of the location of the ischemia in the whole body. The aims of this study were 1) to examine the possibility of an antiarrhythmic effect due to brief ischemia of a remote part of the body (as a model of preconditioning) before sustained ischemia of the heart and 2) to investigate the role of norepinephrine (NE) and prostacyclin [prostaglandin I₂ (PGI₂)] in mediating cardiac preconditioning with limb ischemia (LI).

Our findings suggest that brief LI followed by a period of recirculation of the limb preconditions against arrhythmia induced by sustained ischemia in the isolated rat heart. The neurotransmitter NE appears to be involved in this antiarrhythmic action.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

The investigation conforms with the Guide for the Care and Use of Laboratory Animals [DHHS Publication No. (NIH) 85-23, Revised 1985].

Preparation. Male Sprague-Dawley rats weighing 250-350 g received 1,000 U heparin and 6 mg/100 g of pentobarbital sodium intraperitoneally. The hearts were rapidly excised, arrested in ice-cold heparinized saline, and immediately mounted on the Langendorff isolated heart apparatus. A modified Krebs-Henseleit solution consisting of (in mM) 118 NaCl, 25 NaHCO₃, 11.1 glucose, 4.9 KCl, 2.7 CaCl₂, 1.2 MgSO₄, and 0.5 NaEDTA was equilibrated with 95% O₂-5% CO₂. The whole system was heated to 37°C by means of water jacketing. The heart was suspended in a sealed chamber containing a silicon funnel for collection and volumetric measurement of the coronary effluent. The perfusate column height was kept at 75-80 cmH₂O, producing a coronary perfusion pressure of ~50 mmHg. A stainless steel electrode was inserted into the right ventricular epicardium for bipolar electrogram recording relative to the aortic cannula. A 30-min stabilization period was used in each group before any intervention on the isolated heart. The hearts were allowed to beat spontaneously during the entire experiment.

Preparation for the experiments with RI consisted of the introduction of a 4-0 silk thread underneath the left descending coronary artery. The artery was approached close to its origin and usually above any visible bifurcations. RI was produced by surgical ligation over a tiny Teflon tubing [6-mm piece from an intravenous catheter (Luer 1614R)]. The artery was compressed without applying traction, and the heart could remain closed in the heart chamber under constant experimental conditions. Reperfusion was initiated by untying the ligature, thus allowing repetition of the same procedure during the subsequent ischemia. LI was produced in anesthetized rats by placing a thin elastic tourniquet (1-mm diameter) around the upper third of the hind extremity in tight position, closed by a clamp to stop the arterial blood supply in the leg. The skin temperature was measured in the second interdigital space of this leg with an electrothermometer. The tourniquet was paced for 10 min and subsequently removed for 10 min of recirculation of the blood.

During the ischemic period, the skin of the leg changed color and the temperature decreased by 6.7 ± 0.2°C. After recirculation, the skin color returned to rose and the temperature increased to baseline. A rapid thoracotomy and excision of the heart for Langendorff perfusion followed as usual.

Cardiac rhythm was continuously displayed on a personal computer monitor. Paper recordings were obtained at baseline, at the onset of ischemia, and at reperfusion. The coronary perfusion pressure was recorded through a Senso Nor 840 transducer at the level of the aortic cannula. Coronary flow was measured volumetrically.

Protocol. Experiments were performed in the following groups of rats: group I, 30-min stabilization, 30-min RI, and 15-min reperfusion (control rats; prolonged RI; n = 18); group II, 30-min stabilization, a brief 5-min episode of RI, and 15-min reperfusion followed by sustained 30-min RI and 15-min reperfusion [preconditioning with RI in the isolated heart; a group of "classic preconditioning" that was previously reported (1) and is presented here only for reference; n = 12]; and group III, 10-min LI, 10-min recirculation of the limb, excision of the heart, 30-min stabilization perfusion of the isolated heart, 30-min RI, and then 15-min reperfusion (preconditioning with in vivo LI was performed in anesthetized rats; n = 20).

In addition, we carried out experiments to investigate the potential role of NE in the antiarrhythmic effect of preconditioning with LI. In group A, reserpine (0.15 mg/kg diluted in glacial acetic acid-saline 1:50) was administered intraperitoneally 24 h before the usual protocol for group III (depletion of NE by reserpine; n = 14). In group B, vehicle (1 ml of glacial acetic acid-saline 1:50) was administered intraperitoneally instead of reserpine, followed by the same experimental protocol as for groups III and A (control to the above reserpinized animals; n = 10). In group C, NE (20 nM, freshly prepared in buffer saline) was infused for 2 min into the isolated heart after 18 min of stabilizing perfusion, followed by another 10 min of perfusion, 30 min of RI, and 15 min of reperfusion (administration of NE; n = 10). In group D, an ₁-adrenergic blocker, prazosin (200 mg/kg), was injected intravenously 2 min before LI. After 20 min of stabilization in the isolated heart system, prazosin (5 mM) was administered in the perfusion system for 5 min (3). This was followed by 5 min of normal perfusion, 30-min RI, and 15-min reperfusion. The wet heart weight was determined at the end of the experiments.

NE, reserpine, and prazosin hydrochloride were obtained from Sigma Chemical.

Biochemical assays. Samples were collected from the coronary effluent at baseline (after stabilization, just before sustained ischemia), at minute 30 of RI, and minutes 1 and 3 of the reperfusion period. The samples were taken on ice and refrigerated at 20°C for the biochemical assays.

NE in the coronary effluent was determined with high-performance liquid chromatography with electrochemical detection (11). Production of PGI₂ was assessed by measurement of its stable metabolite 6-ketoprostaglandin F₁ (6-keto-PGF₁). The DuPont ³H radioimmunoassay kit was used (22).

Data analysis. All data are presented as means ± SE. Coronary flow and secretion of PGI₂ or NE are expressed per minute per gram wet heart weight. PGI₂ or NE overflow was calculated as a percent change relative to basal secretion before sustained ischemia to reduce the intragroup variability. The extent of RI was defined by a percent reduction in coronary flow compared with baseline.

Arrhythmia during either ischemia or reperfusion was categorized as ventricular tachyarrhythmia (TA). TA was defined as tachycardia or fibrillation. The distinction between tachycardia and fibrillation on a "per animal" basis was difficult in this model because of numerous intermediate grades in the morphology of TA and a tendency to interconvert from one type to another. TA persisting until the end of the correspondent ischemia or reperfusion is referred to as sustained TA. Analyses were carried out in accordance with the Lambeth Conventions (35).

Categorical data, i.e., TA incidence, were compared with Fisher's test. Other results were compared with analysis of variance with Duncan intergroup comparisons or repeated measures as required. Statistical significance was accepted at a two-sided P of <0.05.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Coronary flow was significantly higher in the preconditioned hearts before sustained ischemia (P < 0.05; Table 1). Coronary artery occlusion caused an average reduction of 38 ± 2% in coronary flow without a marked difference among the groups.

In group II (classic IP) at 1 and 3 min of reperfusion, the coronary flow increased significantly in comparison to the control value.

There were no significant differences in the average heart rate during the stabilizing perfusion (Table 2). After the onset of ischemia, it decreased relative to the preischemic period (P < 0.05). Heart rate had a tendency to recover during reperfusion but only in group II did it return to the basal level.

Arrhythmias. All the hearts suffered from ventricular extrasystoles during ischemia and at the onset of the reperfusion period. Ischemic TAs developed in some of the hearts. The onset of the arrhythmia was between minutes 14 and 17 of ischemia and was usually self limited and short term. Reperfusion TAs (RTAs) developed 30-90 s after the release of the coronary ligature and were self limited. The hearts resumed normal sinus rhythm after episodes of TA terminated and remained in that rhythm until the end of the experiment. In two hearts (in groups I and A), the ischemic TA deteriorated into ventricular fibrillation that remained sustained until the end of the experimental protocol.

Table 3 compares the incidence and duration of TA in the experimental groups. There were no statistically significant differences among groups I-III in the number of TA episodes during the ischemic period.

The frequency of RTA was significantly different between group I (control) and the two preconditioned groups. In group I, 67% of the hearts had RTA compared with 8% in group II and 10% in group III. Thus there was a pronounced protection against RTA by IP with RI of the heart itself as well as by brief LI preceding the prolonged ischemia of the heart.

Reserpine (group A) abolished, at least in part, the antiarrhythmic effect of LI preconditioning. Forty-three percent of the hearts in this group showed TA incidences during reperfusion. Intraperitoneal administration of vehicle instead of reserpine (group B) had no influence on the preconditioning effect of LI either during ischemia or at reperfusion (20% TA incidence during ischemia and 10% RTA).

Pharmacological preconditioning with NE during heart perfusion completely prevented the appearance of TA during 30-min RI and decreased the RTA incidence to 20%.

The ₁-adrenoceptor blocker prazosin did not prevent the antiarrhythmic effect of preconditioning with LI; no arrhythmias were recorded during either ischemia or reperfusion.

PGI₂. The levels of the PGI₂ metabolite in the coronary effluent collected at baseline were 637 ± 50, 717 ± 117, and 860 ± 84 pg · min¹ · g wet wt¹ in groups I-III, respectively. Although the baseline levels in groups I (control) and II (classic preconditioning) do not differ significantly, the values obtained after LI (group III) are significantly higher than the control values (P < 0.05). In all experiments, prolonged RI (30 min) was associated with a reduction in the release of 6-keto-PGF₁. When the data are analyzed percentagewise in reference to the baseline before sustained ischemia (Fig. 1), the decrease in PGI₂ after 30 min of ischemia is significantly less in group II (classic preconditioning) than in the other two groups. Increased release of PGI₂ was found in all groups at minute 1 of reperfusion but was especially pronounced in group II.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Release of 6-ketoprostaglandin F1 (6-keto-PGF1; a stable metabolite of prostacyclin) in perfused isolated rat hearts. Each point represents mean ± SE of measurements obtained in 10-15 experiments at a given time point. * P < 0.05 vs. control and limb ischemia-preconditioned hearts. ** P < 0.05 vs. regional ischemia-preconditioned hearts.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Release of norepinephrine in perfused isolated rat hearts. Each point represents mean ± SE of measurements obtained in 10-15 experiments at a given time point. * P < 0.05 vs. control.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The usual definition of IP is the beneficial effect of one or several episodes of ischemia of the heart against the damage induced by a subsequent, prolonged period of cardiac ischemia. An episode (usually 3-5 min) of RI induced by coronary ligation either in vivo or in the isolated heart is protective against the deleterious electrophysiological, biochemical, and mechanical effects of a longer ischemia within the same region of the heart (2, 24).

Ischemia at distance in the heart itself (ligation of a different coronary artery) might also be cardioprotective, and this interesting phenomenon has only been partially documented to date (28).

Gho et al. (9) described the reduction in infarct size in experiments with 15 min of occlusion of the left renal artery and 10 min of reperfusion followed by 1 h of left coronary artery occlusion and 3 h of reperfusion in rats. Liauw et al. (16) reported significant tissue sparing in the contralateral paired muscle after a subsequent 5 h of ischemia and 48 h of reperfusion. Liauw et al. supposed that such an influence may occur not only in the contralateral muscle but also in different organs and tissues of the body.

We hypothesized that cardioprotection might appear not only within the same organ but that brief ischemia at distance, i.e., ischemia of the limb followed by reperfusion, might also precondition other organs, e.g., the heart. Such a cardioprotective effect of preconditioning in general should be mediated via neurohumoral factors that subsequently work at a distance on the organ in danger.

One of the cardioprotective properties of preconditioning induced by RI of the heart, well documented in various species and in experimental models, is the antiarrhythmic effect of this phenomenon (10, 31). Therefore, we used a model of the isolated heart in which protection is manifested by attenuation of reperfusion arrhythmia to explore 1) the possibility of obtaining the antiarrhythmic effect by an extracardiac IP and 2) the effect of this type of preconditioning, i.e., LI, on the cardiac release of endogenous, physiologically active substances such as PGI₂ and NE.

Our experiments demonstrated a protective antiarrhythmic effect of a preceding brief LI, similar to the IP of the heart, against RTAs (Table 3). In a previous study, Arad et al. (1) described a possible relationship between PGI₂ release during RI and the protection from reperfusion arrhythmia in classically preconditioned isolated hearts. In those series, the protective effect was abolished by pretreatment with aspirin. In the present study, we did not find any relationship between the antiarrhythmic effect of remote IP and the pattern of PGI₂ release. However, it appeared that the pattern of NE release, i.e., an increase at baseline but a decrease at reperfusion, might be related to the protective effect of LI (Fig. 2).

PGI₂ and NE are both biologically active substances known to play a role in cardiac arrhythmogenesis (7, 8). PGI₂ was found to have protective effects against ischemic and reperfusion arrhythmia (26) and was described as a possible mediator of the antiarrhythmic effect of preconditioning (33). Other investigators (14, 17) could not confirm the role of these products in preconditioning. Our data suggest that PGI₂ might be involved in classic preconditioning but is not involved in the actions of LI preconditioning on arrhythmogenesis. The role of catecholamines in ischemia and reperfusion-induced arrhythmia was extensively discussed in the literature (8, 27, 29). Although catecholamines are associated with arrhythmogenesis of the ischemic myocardium in rats (30), some investigators (19) demonstrated that catecholamines facilitate experimental ventricular defibrillation. The role of catecholamines in the phenomenon of IP is still controversial (3, 36).

We demonstrated that NE per se protects against both ischemia and reperfusion arrhythmia (Table 3). These results were in agreement with the data published by Banerjee et al. (3) in isolated rat hearts with global ischemia and by Bankwala et al. (4) in an in vivo rabbit model of RI. Furthermore, our experiments showed that prior depletion of catecholamine stores by reserpine partially abolished the antiarrhythmic effect of preconditioning of the heart with LI. These experiments confirm our hypothesis regarding the role of NE in LI preconditioning of the heart. It appears that preconditioning with LI induces NE release due to a systemic stress reaction, and this process involves adrenergic nerve termini of the heart.

Banerjee et al. (3) and Toombs et al. (32) prevented preconditioning by an ₁-adrenergic-receptor blocking agent, prazosin, and suggested that ₁-adrenoceptors are involved in the mechanism of classic IP. We obtained with prazosin the same depression of TA after LI preconditioning. It seems therefore that ₁-adrenoceptors are not involved in LI preconditioning. The difference in the results might be caused by different models (global vs. regional ischemia), different species (in vivo rabbit vs. in vitro rat hearts), or different aspects of protection (functional recovery or infarct size as opposed to antiarrhythmic effect in our study). Bugge and Ytrehus (6) and Mitchell et al. (21) recently claimed that IP is protein kinase C dependent but not via stimulation of -adrenergic receptors in the isolated rat heart.

Note that the antiarrhythmic effect achieved by IP of the heart (group II in the present study) did not appear to be related to endogenous NE release. Although exogenous NE infusion mimicked the preconditioning effect, depletion of catecholamine stores by reserpine only partially blocked the antiarrhythmic effect of LI preconditioning. We hypothesize that several mediators and mechanisms of action might be involved in the preconditioning process. Our study appears to demonstrate that NE is involved in the mechanism of preconditioning with LI but does not clarify the exact mechanism.

Meerson and Malyshev (20) showed that adaptation to repeated stress reduces the incidence of ventricular fibrillation during a subsequent acute infarction. Our results seem to support the theory that stress reaction and not damage per se is responsible for the antiarrhythmic effect of preconditioning. It appears that the ischemic-reperfused tissue elicits a stress reaction independently of its location.

The possibility that stress proteins are involved in the mechanism of this protection should also be considered. Yellon et al. (37) and Black and Lucchesi (5) demonstrated the role of stress proteins in cardioprotection. Liauw et al. (16) received a different pattern of heat shock protein expression 48 h after the first and before the onset of subsequent ischemia of the contralateral muscle. The authors (16) suggest that this cytoprotective effect may be generalized to other tissues. Other authors (12, 13) were able to demonstrate rapid expressions of various stress proteins. However, we have not yet examined this particular mechanism in our model.

In conclusion, brief ischemia of an extremity, similar to preconditioning of the heart with RI in the heart itself, protects against RTA. LI is associated with an increase in the baseline level of NE and a reduced release of NE during reperfusion. The cardioprotective effect obtained by a preceding in vivo brief LI might be a result of a nonspecific stress response. One of the humoral mediators of this response appears to be NE, whereas other possible mediators remain to be identified.