Abstract
Objective: Brain death (BD) abolishes the infarct-limiting effect of ischemic preconditioning (IP) in rabbits. We wished to define the role of the norepinephrine storm in this observation. Methods: Rabbits were randomized into six groups of ten animals each. In control group (CTRL), anaesthetized rabbits were subjected to 30 min left coronary marginal branch occlusion and 90 min reperfusion. In CTRL+IP group, anaesthetized rabbits were preconditioned with a 5-min ischemia and 5-min reperfusion sequence before coronary occlusion. In CTRL+NE+IP group, anaesthetized rabbits received a 10 μg/kg norepinephrine injection 90 min before IP. In BD group, rabbits were subjected to 90 min of BD before coronary occlusion. In BD+IP group, brain-dead rabbits were preconditioned before coronary occlusion. In BD+LA+IP group, rabbits received an intra-arterial bolus injection of an alpha and beta adrenoreceptor blocking agent (labetolol, 1 mg/kg) prior to brain death induction and subsequent preconditioning. BD was induced by rapid inflation of an intracranial balloon. At termination of the experiment, left ventricular volume (LVV), myocardial volume at risk (VAR) and infarct volume (IV) were determined with methylene blue and tetrazolium staining, and measured using planimetry. Results: LVV was not significantly different among groups. Myocardial VAR/LVV was not significantly different between groups (CTRL, 22.5 ± 6.9%; CTRL+IP, 23.3 ± 2.2%; CTRL+NE+IP, 25.9 ± 12.7%; BD, 19.9 ± 4.8%; BD+IP, 21.7 ± 3.1%; BD+LA+IP, 23.4 ± 5.8 %; P = NS). IV/VAR was significantly reduced in the CTRL+IP group as compared with CTRL and CTR+NE+IP groups (12.2 ± 1.2 vs. 49.7 ± 1.2 and 49.3 ± 4.7%; P < 0.0001). There was no significant difference in IV/VAR between BD and BD+IP groups. In contrast, IV/VAR was reduced in BD+LA+IP compared to BD and BD+IP groups (13.9 < 5.4 vs. 50.0 < 1.4 and 49.6 < 1.5%; P < 0.001). Conclusion: The loss of infarct-limiting effect of IP in brain-dead rabbits is related to the massive release of norepinephrine that occurs as a consequence of BD.
1 Introduction
Myocardial preservation for heart transplantation relies mainly on cardioplegic arrest and organ storage in a cold preservation solution [1]. However, this exogenous approach provides only an incomplete protection against ischemia-reperfusion injury and early graft failure remains one of the most frequent causes of death after heart transplantation [2].
Ischemic preconditioning (IP) has been described as a rapid, adaptive response of myocardium to a brief ischemic insult, slowing down the rate of cell death during a subsequent period of prolonged ischemia [3]. This concept has since been subsequently extended to include the protection against other deleterious effects of the ischemia-reperfusion sequence, such as post-ischemic contractile dysfunction, dysrhythmias and endothelial dysfunction. Although there is a strong evidence that the combination of IP and cardioplegia affords no additional protection to routine cardiac surgery, IP might find an elective indication during cardiac allograft cold storage where prolonged ischemic times are to be anticipated [4]. Recently, several investigators have reported that IP provides additional myocardial protection during long term hypothermic cardioplegic arrest[5,6]. Similarly, IP has been shown to provide effective protection during long term preservation of lung [7] and liver [8]. However, all of these studies were performed on organs taken from anaesthetized animals. Therefore, the clinical relevance of these studies remains questionable. One has to consider that in the clinical setting of organ transplantation, donor organs are harvested from brain-dead patients. It is now well established that brain death is a complex pathophysiologic condition which leads to the progressive deterioration of donor organs [9]. Tissue injury is mainly due to the noxious effects of endogenous catecholamine release and of the major endocrine and metabolic changes that occur as a consequence of brain death [9]. We have recently shown that the cytoprotective effect of IP is abolished after acute brain death in rabbit compared to anesthetized animals [10]. We hypothesized that the endogenous norepinephrine release occurring after brain death could interfere with the preconditioning response, rendering the heart unresponsive to the IP stimulus. The present study was designed to evaluate whether norepinephrine release insult is responsible for this observation. We have deliberately restricted our study to the infarct-limiting effect of IP since other end-points such as reduction of post-ischemic contractile dysfunction, dysrhythmias or endothelial dysfunction may not be manifestations of IP as originally defined.
2 Material and methods
2.1 Animals
Adult New Zealand white rabbits weighing from 3 to 4 kg were used. All animals received humane care in compliance with the ‘Guide for the Care and Use of Laboratory Animals’ published by the US National Institutes of Health (NIH Publication No.85-23, revised 1996).
2.2 General surgical preparation
Rabbits were premedicated with 0.5 mg/kg intramuscular acetylpromazine. Venous access was obtained through a 24-Gauge IV catheter placed in a marginal ear vein. Anesthesia was induced and maintained using intravenous ketamine (20 0 and 120 mg/kg/h, respectively). After general anesthesia and heparinization (500 UI/kg), animals were ventilated through a tracheotomy with an endotracheal tube at a rate of 60 strokes per min and 60% oxygen fraction. The arterial pH, the partial O2 and CO2 pressures were measured after equilibration. Body temperature was maintained at 38°C throughout the experiments by an automatically adjusted heating lamp connected to an intrarectal temperature catheter. A 20-Gauge catheter connected to a HP 1290 C pressure transducer (Universal Quartz Transducer, Hewlett Packard) was placed in the left femoral artery for continuous monitoring of systemic mean arterial pressure (MAP) and blood sample withdrawal. A similar catheter was introduced into the left ventricle through the right common carotid artery. Left ventricular contractility was assessed by left ventricular developed pressure (LVDP, calculated as the difference between left ventricular end-systolic and end-diastolic pressures), and by the maximal positive and negative rates of rise of left ventricular pressure (+dP/dt and −dP/dt).
2.3 Experimental protocol
The experimental protocol is summarized in Fig. 1 . Rabbits were randomized into six experimental groups of ten animals each. In control group (CTRL), anesthetized rabbits were subjected to 30 min of coronary occlusion and 90 min of reperfusion without any pre-treatment. In CTRL+IP group, anesthetized rabbits were preconditioned with 5-min of ischemia and 5-min of reperfusion before the 30-min coronary occlusion. In CTRL+NE+IP group, anesthetized rabbits were subjected to an intra arterial bolus of 10 μg/kg norepinephrine. After 90 min, animals were preconditioned with 5-min of ischemia and 5-min of reperfusion before the 30-min coronary occlusion. In brain dead group (BD), rabbits were subjected to 90 min of brain death before 30 min of coronary occlusion and 90 min of reperfusion. In BD+IP group, brain dead rabbits were preconditioned with 5-min of ischemia and 5-min of reperfusion before the 30-min coronary occlusion. In BD+LA+IP group, animals had 1 mg/kg intra-arterial bolus of labetalol, an alpha and beta adrenoreceptor blocking agent, 1 min before brain death induction, then procedured as in BD+IP group.
Schematic of experimental protocol for examining the effects of brain death (BD) on the cardioprotective effects of ischemic preconditioning (IP). The IP protocol consisted of a 5 min occlusion of the coronary branch, followed by 5 min reperfusion period.
2.4 Induction and validation of brain death
We reproduced the experimental brain death model in the rabbit previously described and validated by Biswas et al [11]. After 10 min of equilibration, a craniotomy was performed on the midline of the skull at the fusion of the fronto-parietal plates. A 10-Gauge Foley catheter was introduced into the subdural space. Brain death was induced by rapid inflation of the balloon catheter with 4 ml of saline solution, aiming to acutely increase intracranial pressure. The balloon was kept inflated throughout the experiment. All anaesthetic agents were then withheld and animals received no hemodynamic or pharmacological support after brain death induction. Brain death was evident when pupillary reflexes and spontaneous respiration became absent, and cerebral electrical activity ceased. Electroencephalographic changes were recorded by means of two unipolar electrodes planted in the temporo-parietal regions of the scalp and connected to a MP 100 data acquisition system using the Acknowledge® 3.0 software (Biopac Systems, Cerom SARL, Paris, France). Electroencephalographic recordings undulating between −2 and +2 μV were considered as isoelectric. Hemodynamic measures were recorded at baseline and at 1, 10, 30, 60 and 90 min after intracranial balloon inflation. Arterial blood samples for catecholamine level determination were taken at baseline, and at 1 and 90 min after intracranial balloon inflation (and after 1 min of bolus injection in CTRL+NE+IP group). Blood samples were centrifuged in a cooling centrifuge at 3000 rev./min for 15 min. Plasma was then removed and stored at −80°C until analysis. Plasma catecholamines norepinephrine and epinephrine were analyzed by high-performance liquid chromatography with electrochemical (coulometric) detection [12].
2.5 Preconditioning protocol and ischemia-reperfusion injury
After 90 min of brain death or anesthesia, the heart was exposed by means of a thoracotomy in the fourth intercostal space, and the pericardium was opened. A prominent marginal branch of the circumflex coronary artery was identified. A 6-0 polypropylene suture was passed around the vessel and the ends of the suture were threaded through a small vinyl tube to make a snare. The coronary branch was occluded by pulling the snare, which was then fixed by clamping the tube. Myocardial ischemia was confirmed by regional cyanosis of the myocardial surface. The preconditioning protocol consisted of a 5-min occlusion of the coronary artery, followed by a 5-min reperfusion period. Thereafter, the coronary branch was occluded for 30 min and then reperfused for 90 min. Preliminary studies performed in our laboratory have shown that this preconditioning protocol significantly reduces infarct size after 30-min coronary occlusion in anesthetized rabbits. Hemodynamic measures were recorded after 10 and 30 min of ischemia, and after 10, 30, 60 and 90 min of reperfusion.
2.6 Determination of myocardial infarct size
After the reperfusion period, hearts were rapidly explanted and a cannula was secured in the ascending aorta. The coronary branch was re-occluded with the snare and the aorta was perfused retrograde with a solution containing 1% methylene blue. The anatomic area at risk was demarcated by the absence of methylene blue dye. The left ventricle was then dissected free of aorta, pulmonary artery, atria and right ventricle and frozen at −20°C for 24 h before being cut transversely into approximately 2-mm thick slices. Each slice was stained by incubation for 20 min at 37°C in 50 ml of phosphate-buffered 1% triphenyl tetrazolium chloride (TTC, Laboratoire Sigma Aldrich, St Quentin Fallavier, France). TTC staining has been shown to demarcate viable tissue by reacting with myocardial dehydrogenase enzymes to form a brick red stain [13]. TTC negative regions, representing irreversibly injured myocardium, appeared as pale yellow. The heart slices were mounted in a glass press, which compressed the slices into uniform 2-mm thickness. Areas of infarct and risk zone were determined by computer-assisted planimetry (Perfect-Image software, Version 4.2, Clara Vision). Infarct and risk zone volumes were then calculated by multiplying each area by the slice thickness and summing the products.
2.7 Statistical analysis
All results were expressed as mean±SD. Intragroup hemodynamic and hormonal changes were assessed by one-way analysis of variance for repeated measures followed by Dunnet's test. Histologic results were analyzed by one way ANOVA followed by Tukey's post-hoc correction for multiple comparisons. A P value of less than 0.05 was considered significant. All statistical procedures were performed by use of GraphPad software (GraphPad Software, Inc. San Diego, USA).
3 Results
3.1 Validation of brain death
Intracranial balloon inflation rapidly abolished pupillar reflexes and spontaneous respiration in all animals of the BD, BD+IP and BD+LA+IP groups. Electroencephalographic changes were characterized by immediate cessation of cerebral activity after intracranial balloon inflation.
3.2 Hemodynamic data
Evolution of MAP is illustrated in Fig. 2 . Rabbits of the CTRL, CTRL+IP and BD+LA+IP groups exhibited a moderate but significant reduction in MAP during the study period. In contrast, major changes in MAP were observed in rabbits of the BD, BD+IP and CTRL+NE+IP groups. Intracranial balloon inflation, as well as norepinephrine bolus induced an immediate and significant increase in MAP. Thereafter, MAP decreased rapidly and remained significantly below baseline values until the end of the study period. Other hemodynamic parameters are shown in Table 1 . The data are presented at baseline, immediately after intracranial balloon inflation, at the end of the preconditioning sequence just before the application of the 30-min coronary branch occlusion, at the end of the 30-min coronary branch occlusion and after 90 min of reperfusion. The evolution of heart rate, left ventricular developed pressure and of the maximal positive and negative rates of rise of left ventricular pressure paralleled that of MAP.
Evolution of mean arterial pressure (mmHg) in control (CTRL), control plus preconditioning (CTRL+IP), control plus norepinephrine plus preconditioning (CTRL+NE+IP), brain death (BD), brain death plus preconditioning (BD+IP) and brain death plus labetalol plus preconditioning (BD+LA+IP) groups. The ischaemic preconditioning (IP) protocol consisted of a 5 min occlusion of the coronary branch, followed by 5 min reperfusion period. Data are expressed as mean±SD. *P < 0.05 vs. baseline values of the corresponding group.
Hemodynamic dataa,b
3.3 Plasma catecholamines
Plasma catecholamines levels remained stable throughout the study period in rabbits of the CTRL, CTRL+IP and BD+LA+IP groups (Table 2) . In contrast, a significant increase of plasma norepinephrine levels was observed 1 min after intracranial balloon inflation in rabbits of the BD and BD+IP groups and 1 min after bolus in CTRL+NE+IP group. By 90 min after intracranial balloon inflation or bolus, norepinephrine levels had returned to baseline levels. There were no significant changes in plasma epinephrine levels during the study period in these three groups.
Plasmatic norepinephrine and epinephrine levels at baseline, 1 and 90 min after brain death inductiona,b
3.4 Infarct size
The effects of IP on infarct size are shown in Fig. 3 . Infarct volume, expressed as a percentage of volume at risk, was significantly reduced in the CTRL+IP group as compared to CTRL and CTRL+NE+IP groups (respectively, 12.21 ± 1.21 vs. 49.72 ± 1.68 and 49.26 ± 4.70%; P < 0.001). There was no significant difference in infarct volume, expressed as a percentage of volume at risk, between the BD and the BD+IP groups. In contrast, it was significantly reduced in BD+LA+IP group compared to BD and BD+IP groups (respectively, 13.91 P < 0.001 5.38 vs. 49.96 P < 0.001 1.36 and 49.60 P < 0.001 1.52%; P < 0.001). There was no significant difference between CTRL+IP and BD+LA+IP groups. Similarly, infarct volume, expressed as a percentage of left ventricular volume, was significantly reduced in the CTRL+IP group as compared with the CTRL and CTRL+NE+IP groups (respectively, 2.92 ± 0.44 vs. 11.11 ± 3.12 and 12.53 ± 6.03%; P < 0.001). There was no significant difference in infarct volume, expressed as a percentage of left ventricular volume, between the BD and BD+IP groups. However, it was significantly reduced in BD+LA+IP group compared to BD and BD+IP groups (respectively, 3.25 ± 1.55 vs. 10.05 ± 2.51 and 10.77 ± 1.67%; P < 0.001). There was no significant difference between CTRL+IP and BD+LA+IP groups. The left ventricular volume was not significantly different among the six experimental groups (CTRL, 6.56 ± 1.12 cm3; CTRL+IP, 5.75 ± 0.73 cm3; CTRL+NE+IP, 5.33 ± 0.90 cm3; BD, 5.78 ± 0.78 cm3; BD+IP, 6.18 ± 0.92 cm3; BD+LA+IP, 5.45 ± 0.93 cm3; P = NS). Furthermore, the myocardial volume at risk, expressed as a percentage of left ventricular volume, was not significantly different between groups (CTRL, 22.49±6.93%; CTRL+IP, 23.32±2.16%; CTRL+NE+IP, 25.91±12.68%; BD, 19.91±4.75%; BD+IP, 21.70±3.12%; BD+LA+IP, 23.40±5.82%; P=NS).
Bar graph comparing the effects of ischaemic preconditioning on myocardial infarct volume. All hearts underwent 30 min of regional ischemia followed by 90 min of reperfusion. The ischaemic preconditioning (IP) protocol consisted of a 5 min occlusion of the coronary branch, followed by 5 min reperfusion period. IP significantly reduced infarct volume in CTRL+IP group, but not in CTRL+NE+IP group. Similarly, IP significantly reduced infarct volume in BD+LA+IP group, but not in BD+IP group. Volume at risk of infarction was not significantly different among the six groups. Data are expressed as mean±SD. *P < 0.05 CTRL+IP and BD+LA+IP vs. CTRL, CTRL+NE+IP, BD and BD+IP groups. IV, infarct volume; VAR, volume at risk; LVV, left ventricular volume.
4 Discussion
Our group has recently reported that the myocardial cytoprotective effects of IP were abolished in brain dead rabbits [10]. However, the mechanism by which brain death abolishes the myocardial protective effects of IP was not evaluated in our previous study. The complexity of the mechanism of IP and that of brain death makes it difficult to determine which factors are directly responsible for the observed loss of protection of IP in brain dead rabbits.
Although the exact mechanisms underlying protection from IP are still incompletely understood, there is good evidence that it is a receptor-mediated effect [14]. Norepinephrine (NE), acting through α1-adrenoreceptors has been proposed as a putative mediator of IP. The role of α1-adrenergic stimulation by catecholamines in cardiac preconditioning is supported by the fact that the beneficial effects of IP are completely eliminated by previous reserpine-induced depletion of neuronal NE stores and by selective α1-adrenergic blocade [15–17]. Furthermore, exogenous administration of NE simulates the protective effects of IP [16]. Stimulation of α1-adrenoreceptors induces a rise of intracellular cyclic AMP with subsequent increase of intracellular calcium concentrations and activation of protein kinase C [18]. Protein kinase C has repeatedly been implicated in the biochemical cascade of IP in rabbits and other species and might act through opening of ATP-dependent potassium channels or enhanced release of adenosine via activation of ecto-5′-nucleotidase [14].
It is now well recognized that brain death is a complex pathophysiologic condition with major hemodynamic, endocrine and metabolic modifications [9]. In the present study, we reproduced a validated brain death model in the rabbit [11]. The hemodynamic modifications observed after brain death induction corroborated the findings of previous studies working on experimental brain death in the rabbit [11] or other animal species. Like others, we could distinguish two successive phases. The first, lasting approximately 10 min from the moment of intracranial balloon inflation, was characterized by a hyperdynamic state with significant increases of mean arterial pressure, heart rate, and inotropic and lusitropic states. Thereafter, a significant drop of mean arterial pressure, heart rate, and inotropism and lusitropism below baseline values characterized the second phase. These hemodynamic changes were associated with a significant increase of plasmatic NE levels immediately after intracranial balloon inflation. In contrast with others, we did not notice a significant drop of NE levels after 90 min of BD in comparison to baseline values. Meanwhile, no significant changes of plasma epinephrine levels were observed. The massive increase in circulating catecholamine levels results in a sudden and major increase in myocardial work load and oxygen consumption against increased peripheral vascular resistances [9]. Although the occurrence of intrinsic myocardial damage as a consequence of brain death remains debated [19], the high levels of circulating catecholamines have been shown to increase influx of calcium into myocardial cells [20], resulting in subsequent activation of enzymes such as lipase, protease, endonuclease, nitric oxyde synthase and mitochondrial damage [9]. We have recently reported that brain death induces myocardial apoptosis in rabbit hearts [21]. NE has been shown to stimulate apoptosis in adult rat cardiac myocytes, acting via the β-adrenergic pathway and activation of protein kinase A [22]. Furthermore, accumulation of catecholamine oxidative products and catecholamine-induced increase in plasma free fatty acids may alter cell membrane permeability. Other endocrine and metabolic modifications associated with brain death occur at a later phase. Thus, Imai et al. have shown in an experimental brain death model in the rabbit similar to ours, that plasma free triiodothyronine levels decreased significantly below antemortem values only 4 h after brain death induction [23]. Furthermore, they observed no significant changes in antidiuretic hormone, cortisol and insulin levels during the 6 h study period [23]. Since in our study hearts were explanted after 90 min of brain death, these endocrine changes are unlikely to have contributed to the observed results. Because of these hormonokinetics and the overlap between the cellular pathways of IP and brain death, NE appeared as the most likely candidate responsible for the observed abolition of the effects of IP after brain death.
In the present study, we have demonstrated that preconditioning hearts with one 5-min episode of coronary occlusion followed by 5 min of reperfusion significantly reduces infarct size after 30 min normothermic ischemia in anesthetized rabbits, but failed to reduce infarct size in brain dead rabbits. These findings corroborate those of our previous study [10] and suggest that the infarct-limiting effect of IP is lost, or that its threshold is increased, in hearts taken from brain dead rabbits. Furthermore, we have shown that pharmacological inhibition of the cardiovascular effects of the catecholaminergic storm by administration of labetolol prior to intracranial balloon inflation restored the cytoprotective effects of IP. Labetolol has been previously shown to inhibit cardiovascular changes following brain death in pigs [24]. Finally, intra-arterial administration of NE to anesthetized rabbits reproduced the hemodynamic effects observed in brain dead rabbits and abolished the cardioprotective effects of IP. The dose of NE that was administered in our study was chosen to reproduce the hemodynamic effects observed in brain dead rabbits. As in brain dead animals hemodynamic modifications result from endogenous catecholamine release and from direct neural stimulation [9], rabbits in the CTRL+NE+IP required significantly higher levels of circulating NE than brain dead rabbits in order to display similar hemodynamic changes. Overall, our results strongly suggest that NE release during intracranial balloon inflation is responsible for the absence, or increased threshold, of effects of IP in brain dead rabbits. This finding is in contradiction with previous studies showing that NE is a potential trigger for IP [15–18]. This apparent bivalent behaviour of NE is probably related to the different dosages of NE used in various experimental protocols. Indeed, the dosage used in the present study was approximately 10 times higher than that used by others [16]. We believe that administration of low or massive doses of NE might have different effects IP.
Our study does not provide data to support the mechanism by which NE abolishes the protective effects of IP. Brain death has previously been shown to alter the density and sensitivity of myocardial adrenoreceptors [25]. However, in a recent study performed in rabbits, brain death had no significant effect on levels of messenger RNAs encoding α1-adrenoreceptors or β1-adrenoreceptors [26]. Alternatively, the massive release of NE during brain death could interact with intracellular pathways of IP, rendering them unresponsive to a subsequent preconditioning stimulus. In a recent study, Takizawa et al. have shown that α1-adrenoceptor stimulation partially inhibits ATP-sensitive potassium currents in guinea pig ventricular cells [27]. Thus, α1-adrenoceptor mediated inhibition of ATP-sensitive potassium channels may partly offset the cardioprotective effects of IP in brain dead animals. However, the exact mechanisms remain to be determined.
5 Conclusion
In summary, the present study demonstrates that brain death abolishes the infarct-size-reducing effect of IP in rabbits. This loss of protection was related to the massive endogenous release of NE that occurs as a consequence of brain death. Blockade of alpha- and beta-adrenoreceptors before brain death induction restored the cytoprotective effects of IP in brain dead rabbits. These findings underscore the importance of preventive measures in organ donors, if the beneficial effects of IP are to be exploited for prolonged organ preservation.
