Mechanisms of Isoflurane-induced Myocardial Preconditioning in Rabbits 

This study was conducted according to the standards of the American Physiologic Society and with the approval of our hospital's Animal Welfare Committee. New Zealand white rabbits (weight, 3.2-3.5 kg) were sedated with 70 mg/kg intramuscular ketamine. Surgical anesthesia was maintained by continuous intravenous infusion of 10 mg/ml propofol (Diprivan; Zeneca Pharmaceuticals, Wilmington, DE) started at 0.5-1 mg [middle dot] kg (^-1)[middle dot] min^-1. [12,13]During administration of 100% oxygen by mask, a tracheostomy was performed and the trachea was intubated. Ventilation was controlled using a positive-pressure respirator (model 309-0612; Ohio Medical Products, Madison, WI) and an inspiratory fraction of oxygen of 1.0. A carotid artery was isolated and cannulated with a 22-gauge catheter to measure blood pressure and to sample arterial blood. The ventilation rate was adjusted periodically to achieve normocapnia, and small doses of NaHCO³(44 mEq/50 ml) were administered when necessary to maintain physiologic blood pH (at 7.35-7.45). A three-lead surface electrocardiographic lead system was attached to the anterior chest wall, and electric activity of the heart was monitored using a Grass Instruments model 7D recorder (Quincy, MA). Intermittent recording of the electrocardiogram was used to calculate heart rate and to confirm ST-segment changes during ischemia. Arterial blood gases were measured using a Radiometer ABL 2 Acid-Base Laboratory (Copenhagen, Denmark). The end-tidal carbon dioxide tension (P^CO(2)) was monitored using a calibrated Puritan-Bennet anesthetic agent monitor (Wilmington, MA). A warming blanket maintained core body temperature between 38 and 39 [degree sign]C.

Each animal was administered a single intravenous dose of pancuronium (0.4 mg) just before median sternotomy to prevent muscle retractions during electrocautery and therefore to minimize bleeding. After exposing the heart by performing a median sternotomy, the pericardium was opened and a 2-0 polyester suture was passed quickly around the anterolateral coronary artery at approximately the midpoint of its epicardial course. The ends of the suture were passed through a 6-cm piece of plastic tubing to form a snare. [14]The position of the suture varied somewhat among animals as a result of variations in coronary anatomy, and, therefore, a range of areas at risk was obtained in each experimental group. The coronary artery was occluded when necessary by tightening the snare and then clamping the tube with a hemostat. Regional cyanosis and ST-segment elevation confirmed myocardial ischemia. Reperfusion was achieved by releasing the snare and was confirmed by visual observation of reactive hyperemia.

After we placed the snare around the anterolateral coronary artery, we assigned the rabbits to one of seven groups in which the pretreatment was varied. All rabbits in all groups underwent the same 30 min of coronary artery occlusion, followed by 3 h of reperfusion (Figure 1). The first group (control) received no pretreatment. The second group, an ischemia-preconditioned group, received a 5-min coronary occlusion followed by 15 min of reperfusion immediately before the subsequent 30 min of ischemia. Three additional groups of rabbits all received identical isoflurane pretreatment (15 min end-tidal isoflurane, 1.1%, or 0.5 to 0.6 MAC in rabbits)[15]followed by a 15-min washout period; however, these three isoflurane groups varied in what inhibitor was administered to test cellular mechanisms. In an isoflurane group, no inhibitor was administered and only isoflurane was given, as described before. In an isoflurane-plus-glyburide group, rabbits were pretreated with 0.33 mg/kg intravenous glyburide, [16]25 min before administration of isoflurane. In an isoflurane-plus-SPT group, rabbits received 7.5 mg/kg intravenous SPT [16]25 min before isoflurane. Two additional control groups were included to complete the experimental series. In a glyburide control group, rabbits were administered 0.33 mg/kg glyburide but no isoflurane. Finally, in an SPT control group, rabbits were administered 7.5 mg/kg SPT but no isoflurane. In the groups that did not receive glyburide, all rabbits were also administered a control injection of the glyburide vehicle solution (0.25% dimethyl sulfoxide in normal saline) 55 min before the 30-min occlusion (Figure 1).

We drew lots to assign to groups all the rabbits in the first five described groups (control, ischemia preconditioned, isoflurane, isoflurane plus glyburide, and isoflurane plus SPT). These experiments were performed first. To complete the entire study, two additional control groups were added (glyburide control and SPT control). Rabbits were assigned to one of these experiments by chance on each day after placement of the coronary snare. This second series of experiments was performed after the first series.

A stabilization period of 20-25 min was allowed after the surgical preparation was complete, before the experimental protocols were begun. The duration of the experiments was standardized across all seven groups, as depicted in Figure 1. Anticoagulation was achieved in each rabbit using 1,500 units beef lung heparin before the initial coronary artery occlusion. Ventricular fibrillation, if it occurred, was reversed using direct mechanical stimulation: An index finger was flicked directly against the right ventricular side of the fibrillating heart one to three times to achieve defibrillation. Failure to convert to an organized rhythm after three attempts was defined as intractable fibrillation. In all three groups, 3 h of reperfusion followed the 30-min coronary artery occlusion. At the end of 3 h, the heart was stopped by anesthetic overdose and excised to measure IS and area at risk (AR).

In the three isoflurane-pretreated groups, propofol infusion was interrupted briefly during isoflurane administration to maintain a nearly constant level of anesthesia and to avoid hypotension and possible hypotension-induced myocardial ischemia that could theoretically lead to inadvertent myocardial ischemic preconditioning. [17]Seven to 8 min generally were necessary to achieve the target end-tidal isoflurane concentration of 1.1% isoflurane. After the end-tidal level of 1.1% was reached, the isoflurane concentration was held constant for 15 min, after which the propofol infusion was restarted and isoflurane administration was discontinued. No measurable end-tidal isoflurane could be detected 4-6 min after the isoflurane was discontinued.

Mean arterial pressure was maintained at more than 55 mmHg during ischemia and reperfusion by optimizing the ventricular preload. This was achieved by tilting the surgery Table orby infusing 0.9% saline. In addition, the rate of propofol infusion was varied between 0.89-1.4 mg [middle dot] kg^-1[middle dot] min^-1during reperfusion to minimize hypotension in those rabbits with large infarctions.

Hemodynamics were measured at end expiration at various times during the experimental protocol. Blood pressure was monitored continuously using a calibrated strain-gauge transducer (Abbott Critical Care Systems, North Chicago, IL) and a Grass Instruments model 7D polygraph. Heart rate was calculated from electrocardiography and averaged over several heartbeats during each measurement period. The rate-pressure product was calculated as the mean arterial pressure times the heart rate.

The AR was identified by reoccluding the coronary artery of the excised heart and perfusing the aortic trunk with a solution containing fluorescent microspheres. [14]After perfusing the heart with microspheres, each heart was frozen in liquid nitrogen and sliced into 6 to 10 2-mm-thick sections. The AR was defined as the area of the myocardial slice not illuminated by microspheres when viewed under ultraviolet light. To discriminate between infarcted and viable myocardium, heart slices were bathed in the vital stain triphenyltetrazolium chloride for 15 to 20 min at 37 [degree sign]C and then submerged in 10% formalin to stop the staining process. After slices were flattened in a plastic press at 4 [degree sign]C for approximately 24 h, the AR, area of infarct, and total myocardial area on both sides of each slice were traced onto acetate sheets. These traces were enlarged by overhead projection and digitized using a flat bed scanner (model Arcus II; Agfa Division, Miles, Wilmington, MA) in conjunction with Adobe Photoshop software (Adobe Systems, Mountain View, CA). The AR, area of infarct, and total myocardial area were measured using computerized planimetry (NIH Image 1.60, a public domain software), and mean values for each slice were calculated by averaging the values obtained for the two sides. The mean area of infarct and AR for each slice were divided by the total myocardial area, and the result was multiplied by slice mass to obtain approximate values in grams for the area of infarct and AR. The ratio of area of infarct to AR, the proportion of ischemic myocardium that became infarcted in each heart, was calculated from these data.

The glyburide solution was prepared by dissolving 400 mg glyburide (Sigma Chemical Company, St. Louis, MO) in 1 ml dimethyl sulfoxide (Sigma), diluting 0.25 ml of this mixture with 99 ml normal saline, and adding 1 N NaOH by drops until the solution cleared. The volume of the final solution was adjusted to achieve a concentration of 1 mg glyburide/ml. To make the control vehicle solution, 0.25 ml dimethyl sulfoxide was added to 1 ml NaOH, 1 N, and 99 ml normal saline. A solution of SPT was obtained by adding 75 mg SPT (Research Biochemicals International, Boston, MA) to 10 ml saline, 0.9%, and heating this mixture until complete dissolution. The fluorescent microspheres were prepared by adding 120 mg ZnS microspheres, 1- to 20-[micro sign]m, (Duke Scientific, Palo Alto, CA) to 30 ml dextran 40, 10%, containing 0.5 ml Tween 80 (Sigma). Triphenyltetrazolium chloride was prepared by dissolving 2 g 2,3,5-triphenyltetrazolium chloride in 200 ml phosphate buffer, 90 mM, (pH 8.5 to 8.6) at 37 [degree sign]C. [14] 

Values in tables, figures, and the text are expressed as the mean +/− SD. The effect of pretreatment regimen on IS was tested by analysis of covariance using the AR as the covariate. Area at risk, the fraction of the risk zone infarcted (IS:AR), and hemodynamic data were compared in all groups using one-way analysis of variance followed by the Dunnett test for comparison versus control or using the Student-Newman-Keuls test, as appropriate. Differences in hemodynamic measurements were tested among groups and within each group. Statistical analyses were performed using Systat 5.0 (Systat, Evanston, IL) and Statview 4.0 (Abacus Concepts, Berkeley, CA) software. For all statistical analyses, the fiducial limit of significance was chosen as 5%.

Experiments were performed on 64 rabbits: 13 in the control group, 10 in the ischemia-preconditioned group, 9 in the isoflurane group, 10 in the isoflurane-plus-glyburide group, 8 in the isoflurane-plus-SPT group, and 7 each in the glyburide-control and SPT-control groups. Five rabbits died of ventricular fibrillation shortly after coronary occlusion (three in the control group, one in the preconditioning group, and one in the isoflurane group). Two rabbits, both in the isoflurane-plus-glyburide group, were excluded because of intractable hypotension that began during coronary occlusion.

Areas at risk were comparable among the seven treatment groups (Table 1). Myocardial IS, expressed as a percentage of AR (IS:AR), was 30.2 +/− 11% in the control group and was reduced significantly (to 8.3 +/− 5%) by ischemic preconditioning (Figure 2). Preexposure to 15 min of isoflurane (1.1 MAC) also reduced myocardial IS substantially (to 13.4 +/− 8.2%), an effect size that was not significantly different from the protective effect of ischemic preconditioning. The mean IS was 30 +/− 9.1% in the isoflurane-plus-glyburide group and 29.2 +/− 12.6% in the isoflurane-plus-SPT group. Thus, previous administration of glyburide or SPT abolished the myocardial protective effects of isoflurane and returned IS to values comparable to those in the control group. Neither glyburide nor SPT alone increased IS (IS:AR = 33.9%+/− 7.6% and 31.8%+/− 12.7%, respectively; P = not significant).

As previously noted, in the three isoflurane-pretreated groups, propofol infusion was interrupted briefly during administration of isoflurane to maintain a nearly constant level of anesthesia and to avoid hypotension. There were no significant differences in heart rate, mean arterial pressure, or rate-pressure product among the seven treatment groups at baseline or immediately before coronary occlusion (Table 2). Immediately before coronary occlusion, mean arterial pressure, heart rate, and the rate-pressure product were similar in all groups. After 30, 60, or 180 min of reperfusion, significant differences in hemodynamic values also were not distinguishable in the seven treatment groups, despite the significant differences among groups in IS. We attributed this, in part, to our experimental protocol, which called for preload adjustments and changes in the propofol infusion rate to maintain blood pressure with an acceptable range in all groups during the reperfusion period.

As noted before, five rabbits died of intractable ventricular fibrillation (three in the control group, one in the preconditioned group, and one in the isoflurane group). In addition, short periods (< 90 s) of transient, nonlethal ventricular fibrillation were recorded in three rabbits in the control group, in one rabbit in the preconditioning group, in one rabbit in the isoflurane-plus-glyburide group, in one rabbit in the glyburide-control group, and in one rabbit in the SPT-control group. The overall incidence of ventricular fibrillation (lethal plus nonlethal) was not different among the seven groups using chi-squared analysis (P > 0.25).

This study confirms that pretreatment with a clinically relevant dose of isoflurane protects the myocardium from subsequent ischemia, even after isoflurane is discontinued subsequent ischemia, even after isoflurane is discontinued and washed out. The degree of protection offered is substantial and is approximately equivalent to the protection offered by a 5-min period of preconditioning ischemia. This protective, preconditioning-like effect of isoflurane is reduced by concurrent administration of glyburide, a K^ATPchannel inhibitor, or by SPT, a nonspecific adenosine receptor antagonist. Together, these findings suggest not only that both adenosine receptor activation and K^ATPchannel activation play a role in the underlying mechanisms of anesthetic-induced preconditioning, but also that volatile anesthetic-induced preconditioning shares some of the same fundamental underlying mechanisms as ischemia-induced preconditioning.

The current study confirms and extends previous findings that preconditioning-like effects are induced by isoflurane or other volatile anesthetics. Using a dog model of myocardial ischemia and reperfusion, Kersten et al. [3]found that preadministration of 1 MAC isoflurane, followed by 5 min of washout, reduced the IS:AR ratio by 53%. This effect was equal to the protective effects generated by multiple transient occlusions, and it was inhibited by glyburide. The isoflurane-induced protection lasted at least 30 min after isoflurane was discontinued. Cope et al. [2]used an isolated rabbit heart model to evaluate the myocardial protective effects of three different volatile anesthetics (halothane, enflurane, and isoflurane), which were administered at 2 MAC for only 5 min, followed by 10 min of washout. Preadministration of all three volatile anesthetics caused a significant preconditioning-like effect, decreasing IS after 30 min of ischemia to approximately 20% of that seen in controls. Thus, the protection induced by all three anesthetics was substantial. The protection induced by halothane was tested further in isolated hearts and was inhibited by SPT, a nonspecific blocker of adenosine receptors, and by chelerythrine, an inhibitor of PKC.

In a previous study from our laboratory, [4]a rabbit model was used to identify and measure the preconditioning effects of isoflurane administration in vivo. Isoflurane, 1.1%(0.5 MAC in rabbits), [15]induced a significant preconditioning effect, although the degree of myocardial protection (30% reduction in IS:AR) was substantially less than that found in the current study (a 60% reduction) or in the dog study of Kersten et al. [3](a 53% reduction). We have no certain explanation for the difference in the degree of protection found in our previous study, but subtle differences in experimental methods and laboratory personnel may be important. We also speculate that the absolute size of the rabbit heart may be important: Smaller rabbits were used in our previous study (2.6-3 kg) compared with the current study (3.2-3.5 kg).

Many studies have addressed the role of adenosine receptors and receptor subtypes in ischemic preconditioning, [9,10,18-20]and A1 and A3 receptor subtypes appear to be involved as mediators. In addition, although several studies have suggested a role for adenosine receptors in the cardioprotection conferred by isoflurane when it is administered during ischemia, only one previous study [2]addressed the role of adenosine in anesthetic-induced preconditioning. Our findings are consistent with those of Cope et al., [2]who found that halothane-induced was negative by SPT treatment. In addition, the current study extends the work of Cope et al. [2]by showing that the adenosine-dependent effects of isoflurane preconditioning apply in situ and ex vivo in the isolated crystalloid-perfused heart. Because SPT in the doses we used in the current study is a nonspecific blocker for adenosine receptors, we could not draw conclusions regarding the adenosine receptor subtypes that are implicated. Previous work by Kersten et al., [8]however, showed that when isoflurane is administered during ischemia, part of the cardioprotective effect is mediated by adenosine A1 receptors. The cardioprotective effects of isoflurane in that model also were accompanied by decreases in endogenous adenosine release. The current study supports a role for adenosine in isoflurane-induced myocardial preconditioning and extends the findings of Kersten et al. [8]by showing that adenosine plays a role in the cardioprotective effects of isoflurane in vivo, even after the anesthetic has been discontinued. The specific adenosine receptor subtypes involved remain to be determined.

Many investigators have found evidence strongly linking the phenomenon of myocardial preconditioning to mechanisms that open the cell membrane K^ATPchannel. [9,21-24]The opening of K^ATPchannels even has been proposed as a final common pathway in myocardial preconditioning. The anesthetic-induced preconditioning effect also can be inhibited by glyburide, as shown in the recent study of Kersten et al. [3]and in the current study. This confirms that the opening of K^ATPchannels plays an important role in volatile anesthetic-induced preconditioning and in classic ischemic preconditioning. However, the exact, and probably multiple, mechanisms by which isoflurane affects K^ATPremain unclear.

Studies with inhibitors of K^ATPsuch as glyburide suggest that the net effect of isoflurane in vivo is to open the K^ATPchannel. [5,25]Yet the only direct evidence available to date suggests that the direct action of isoflurane on the channel is to induce closure. [26]Isoflurane apparently induces the channel to be less sensitive to the inhibitory effects of ATP, thus allowing channel opening at higher physiologic ATP concentrations. [26]This alteration in channel sensitivity may be caused directly by isoflurane binding or its solvent-like properties, or it may be caused indirectly by other intracellular signaling cascades, such as adenosine-related mechanisms, G-protein-related mechanisms, or PKC-mediated mechanisms.

Ischemic preconditioning and anesthetic-induced preconditioning share many common traits; that is, K^ATPchannels, adenosine receptors, and possibly PKC activation play roles in both forms of preconditioning. Although the results of our current study suggest strongly that the ability of isoflurane to precondition the myocardium is mediated by the opening of K (^ATP) channels and the activation of adenosine receptors, these findings do not allow us to elucidate precisely the underlying mechanism through which isoflurane interacts with K^ATPchannels and adenosine receptors.

At least two intracellular signaling pathways mediate interactions between adenosine receptors and K^ATPchannels and therefore are candidates for determining the preconditioning effects of isoflurane. First, as shown in isolated cardiomyocytes, adenosine receptors are coupled to and activate K^ATPchannels through a membrane-bound G protein. [27]When directly examined in other models, however, volatile anesthetic interactions with G-protein signaling pathways were inhibitory, [28,29]stimulatory, [30]or absent. [31]We do not know whether volatile anesthetics activate the G-protein pathway that has been shown to regulate K^ATPchannels.

A second pathway by which volatile anesthetics might activate K (^ATP) channels is through the activation of PKC. In some animal models, PKC activation induced a degree of myocardial protection similar to that seen after ischemic preconditioning, [32-34]although PKC activation does not provide protection in all models. [35]Pharmacologic inhibition of PKC also diminishes or eliminates the effectiveness of myocardial preconditioning. [33,34,36]Published reports of the effects of volatile anesthetics on PKC-dependent signaling have shown that PKC-dependent pathways are stimulated [37,38]or inhibited [37,39,40]by volatile anesthetics. Evidence also suggests that PKC-mediated processes may be affected differently, perhaps even oppositely, [37,41]by halothane and isoflurane.

In the current study, the incidence of ventricular fibrillation was highest in the control group, although this difference was not significant (P > 0.25). This result is consistent with the results of Haessler et al., [42]who found that myocardial preconditioning, although effective in reducing infarction, did not effectively reduce arrhythmias.

One possible drawback of our study is that propofol was not administered in the same dose in all groups; it was discontinued briefly during isoflurane administration to avoid anesthetic-induced hypotension, which conceivably could lead to ischemia and ischemia-induced preconditioning. It is unlikely that this experimental design led to any significant changes in our findings. If anything, we would have expected that the discontinuation of propofol during isoflurane administration might lead to a slight bias against finding protective effects of isoflurane. Propofol alone has mild myocardial protective effects. [43] 

The cardioprotective effects of isoflurane shown in the current study almost certainly were induced by a direct pharmacologic effect of isoflurane and not by secondary or indirect effects, as has been reviewed before. [4]Our experimental design did not allow significant isoflurane-mediated hemodynamic effects, and we avoided other potential inadvertent causes of preconditioning, such as hypoxia. [44,45] 

In addition to the remaining questions regarding the exact intracellular mechanisms leading to anesthetic-induced preconditioning, several other important questions remain to be answered. What, for instance, are the important differences, if any, between anesthesia-induced preconditioning and ischemia-induced preconditioning? Are the effects of ischemic preconditioning and anesthetic-induced preconditioning generally additive or competitive? What is the optimal timing and dosage for inducing the anesthetic preconditioning effect? By what mechanisms does the anesthetic-induced preconditioning effect interact with classic ischemia-induced preconditioning?

Preadministration of isoflurane, before a prolonged episode of myocardial ischemia, reduces myocardial IS in rabbits. This cardioprotective effect of isoflurane is mediated through adenosine receptors and K^ATPchannels, mechanisms that are shared by classic ischemic preconditioning. The precise cellular mechanisms mediating anesthetic-induced preconditioning remain to be elucidated.

The authors thank John Rukkila for help in preparing the manuscript and Robert Slocum, B.S., for technical assistance.