Effect of Ischemic Preconditioning of the Myocardium on cAMP
Abstract Reduction of cAMP has been implicated in the protection of ischemic preconditioning (IP), but until now, this possibility has not been directly addressed. In this study, we found that in the in vivo rabbit heart, 10 to 30 minutes of sustained regional ischemia was accompanied by a nearly twofold rise in cAMP levels. This increase in cAMP was attenuated when sustained ischemia was preceded by IP induced with a single cycle of transient ischemia and reperfusion (TI/R) and prevented when ischemia was preceded by three cycles of TI/R. The mechanism of cAMP reduction by IP does not involve activation of protein kinase C (PKC), since the PKC inhibitor polymyxin B (24 mg/kg) did not raise cAMP levels during sustained ischemia in IP hearts. Furthermore, this effect is also not mediated by reduced responsiveness of the β-adrenergic effector pathway, since both nonischemic hearts and hearts subjected to three cycles of TI/R exhibited similar increases in cAMP in response to 5 μg/kg isoproterenol. However, propranolol (0.75 mg/kg) abolished the rise in cAMP levels observed during sustained ischemia in control hearts but did not reduce cAMP levels further in IP hearts. These data indicate that the ischemia-induced rise in cAMP levels in control hearts was mediated by activation of the β-adrenergic receptor. Taken together with data demonstrating that β-adrenergic responsiveness was not affected by IP, these data support the conclusion that the lack of elevation in cAMP levels observed during sustained ischemia in IP hearts is mediated by an attenuation of norepinephrine release. To examine whether the protection of IP against necrosis was mediated by the lack of elevation in cAMP levels, we determined whether the infarct size–limiting effect of IP could be blocked by NKH477, an activator of adenylyl cyclase. Four groups of rabbits were subjected to 30 minutes of in vivo regional ischemia and 90 minutes of reperfusion. Control hearts (n=10) had 53.6±5.5% infarction of the area at risk. IP with three cycles of transient ischemia limited infarct size to 3.2±1.3% (n=13, P<.0001). NKH477 (45 μg/kg) increased average cAMP levels in IP hearts during sustained ischemia to levels similar to those in untreated control hearts. However, NKH477 did not block IP (50.2±7.7% of the area at risk was infarcted in the control+NKH477 group [n=10] versus 10.0±5.9% in the IP+NKH477 group [n=7], P<.05). Therefore, we conclude that although IP lowers cAMP levels during sustained ischemia, this effect is not necessary for its protection against necrosis, since raising cAMP does not block this protection of IP.
Ischemic preconditioning (IP) refers to the phenomenon whereby one or more periods of transient myocardial ischemia protect the heart against injury produced by a subsequent, more prolonged period of myocardial ischemia and reperfusion. The signal transduction pathways involved in producing this phenomenon are currently under investigation; but as yet, their importance remains incompletely understood. One signal transduction pathway that has been implicated in IP1 2 but has received little experimental attention is the cAMP pathway. In the present study, we show that IP induced with three cycles of transient ischemia and reperfusion prevents the nearly twofold rise in cAMP that occurs with sustained ischemia.
One goal of the present study was to investigate the mechanism by which IP prevents the ischemia-induced elevation in cAMP levels. Three possibilities were studied. The first was that activation of PKC reduces cAMP levels in IP. Evidence suggesting that PKC activation plays an important role in the protection of IP has recently emerged.3 4 There is considerable “cross talk” between the cAMP and PKC signal transduction pathways, with PKC exerting multiple effects on both the cAMP-generating and -degrading pathways in the heart, some of which are consistent with a reduction of cAMP levels.5 To investigate the importance of PKC activation in the cAMP-lowering effect of IP, we determined whether this effect could be blocked with the specific PKC inhibitor polymyxin B.
The second possibility was that the episodes of transient ischemia in IP hearts result in the liberation of norepinephrine, as well as other mediators, which then act to desensitize the β-adrenergic effector pathway and prevent the rise in cAMP during the subsequent sustained ischemia. We focused on the β-adrenergic effector pathway, because the β-adrenergic receptor can desensitize fairly rapidly,6 a requirement for a role in IP. Rapid desensitization of the β-adrenergic receptor occurs through uncoupling of this receptor from Gs proteins by multiple mechanisms, including phosphorylation of this receptor by the β-adrenergic receptor kinase and the subsequent binding of β-arrestin,7 activation of cAMP-dependent kinases,6 and activation of PKC.5 We reasoned that if the β-adrenergic signal transduction pathway was desensitized by IP, then myocardium that had been ischemically preconditioned would exhibit an attenuated rise in cAMP in response to β-adrenergic receptor stimulation with isoproterenol compared with cAMP levels in nonpreconditioned myocardium.
The third possibility was that events triggered by transient ischemia and reperfusion in IP hearts attenuate norepinephrine release during sustained ischemia and that this, in turn, leads to diminished β-receptor activation and, consequently, reduced cAMP formation. Schömig8 has shown that in nonpreconditioned rat hearts, periods of myocardial ischemia of >10-minute duration can increase local norepinephrine concentrations within the ischemic myocardium to 100 to 1000 times normal plasma concentrations. Furthermore, this group recently demonstrated that norepinephrine release during sustained ischemia is greatly reduced by IP.9 We reasoned that if reduced stimulation of the β-adrenergic receptor by norepinephrine was responsible for preventing the rise in cAMP levels in IP hearts, then β-blockade with propranolol should reduce cAMP levels in control hearts subjected to sustained ischemia to levels similar to those observed in IP hearts.
Another goal of the present study was to determine whether the lack of elevation in cAMP levels in IP hearts was necessary for its protection against necrosis. Several lines of evidence are consistent with this hypothesis. First, even in the absence of ischemia, elevations in cAMP have been shown to produce myocardial necrosis.10 11 High levels of cAMP are also known to increase sarcolemmal calcium entry12 and to increase the activation of cardiac lipases,13 effects that are known to aggravate ischemic injury. Furthermore, Thornton and colleagues have demonstrated that the protective effect of IP against necrosis can be mimicked by the administration of either adenosine A1 receptor agonists14 or muscarinic M2 receptor agonists.15 In cardiac myocytes, both these receptors are coupled to adenylyl cyclase through Gi proteins,16 and stimulation of these receptors inhibits cAMP formation.17 Finally, the protection of IP against necrosis can be blocked by inactivation of Gi,15 an effect that would also be expected to increase cAMP. To test the hypothesis that IP is mediated by a reduction in cAMP levels, we treated hearts with NKH477, an activator of adenylyl cyclase, to determine whether this could block the protection of IP against necrosis in our in vivo model of regional ischemia and reperfusion in the rabbit heart.
Materials and Methods
We used New Zealand White rabbits of either sex for the present study (weight range, 3 to 4 kg). Rabbits were treated in accordance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health.
To induce anesthesia, we cannulated the marginal ear vein of the right ear for administration of sodium pentobarbital (28 mg/kg) and heparin (285 U/kg). During the experiment, anesthesia was maintained with sodium pentobarbital (25 mg/kg per hour). Anticoagulation was maintained with intravenous heparin (100 U/kg per hour). We administered both drugs diluted in 5% dextrose (25 mL/h). After the induction of surgical anesthesia, we performed a tracheotomy. We ventilated the rabbits with 100% oxygen by using a respirator (Harvard Apparatus). Respiratory rate was set at 35 breaths per minute, and tidal volume was 18 to 24 mL. We made arterial blood gas determinations periodically throughout the experiment and adjusted the tidal volume to keep Pco2, pH, and Po2 within the physiological range. We used a catheter inserted in the left carotid artery to monitor blood pressure and heart rate. For infusion of drugs, we cannulated the marginal ear vein of the left ear.
To expose the heart, we performed a left thoracotomy through the fourth intercostal space and opened the pericardium. We then passed a 2-0 polypropylene suture around the left main coronary artery and passed the ends of the suture through a flanged piece of polyethylene tubing. In experiments involving the creation of ischemia, we accomplished it by pushing the flange against the coronary artery to produce a snare and clamping it in position with a mosquito clamp. We confirmed the presence of ischemia by observing cyanosis distal to the occlusion site. Reperfusion was achieved by releasing the snare. If fibrillation occurred during the course of the experiment, a defibrillator (Burdick Co) was used.
cAMP Determination Study
cAMP Levels in Control and IP Hearts
After the completion of surgery, animals were stabilized for 30 minutes. Rabbits were then assigned randomly into the control and IP groups. Rabbit hearts in the IP group subsequently received three cycles of 5-minute regional ischemia and 10 minutes of reperfusion. Hearts in animals in the control group had an additional 45-minute stabilization period. Control and three-cycle IP hearts were harvested for cAMP determination at the following time points: immediately before the sustained ischemia (ie, after the 75-minute stabilization period in the control group and after three cycles of transient ischemia/reperfusion in the IP group) and after 10, 20, and 30 minutes of sustained ischemia (n≥6 per time point per group). An additional group of hearts was harvested after only the initial 30-minute stabilization period (n=5) to serve as the true baseline value for all groups. As expected, there was no difference in cAMP levels in these hearts and those harvested after the 75-minute stabilization period; consequently, the cAMP data from these two groups of hearts were pooled to generate a larger true-baseline group (n=16).
To examine the effect of transient ischemia on cAMP levels, two additional groups of hearts were harvested at the end of the first and third episodes of 5-minute transient ischemia in hearts assigned to the IP protocol (n=6 per group). The group harvested after the first episode of 5-minute transient ischemia also provided the cAMP levels for 5 minutes of sustained ischemia in the control group, since both groups were subjected to a stabilization period followed by 5-minute ischemia. To examine whether the number of cycles used to induce IP affects the cAMP response during the subsequent sustained ischemia, another group of hearts (n=10) was harvested after being subjected to a single cycle of IP (5-minute regional ischemia and 10-minute reperfusion) followed by 10 minutes of sustained ischemia.
For the drug-treated groups described below, the experiments were conducted in a manner identical to those described above for the untreated control and three-cycle IP hearts.
Inhibition of PKC
We treated control hearts (n=8) and hearts subjected to three cycles of IP (n=7) with a bolus dose of the PKC inhibitor polymyxin B (Sigma Chemical Co) diluted in normal saline (24 mg/kg IV) and administered 5 minutes before a 10-minute period of sustained ischemia, at which time hearts were harvested for determination of cAMP. Recently, we have confirmed that this dose of polymyxin B administered 5 minutes before 30-minute ischemia and 90-minute reperfusion blocks the protective effect of IP on necrosis in our in vivo regional ischemia/reperfusion rabbit model (authors’ unpublished data, 1995).
Responsiveness of the β-Adrenergic Effector Pathway
We infused control hearts and hearts subjected to three cycles of IP with the β-adrenergic receptor agonist isoproterenol (diluted in normal saline [5 μg/kg IV for 5 minutes]) at the end of the 75-minute stabilization period (control group) and after three cycles of transient ischemia and reperfusion (IP group) (n=7 per group). After isoproterenol infusion, we harvested the hearts for determination of cAMP. In both groups, these cAMP measurements were made in the myocardial region supplied by the left main coronary artery (ie, the reperfused myocardium in IP hearts and virgin myocardium in control hearts) and also outside this region.
Blockade of the β-Adrenergic Receptor
We treated control hearts (n=5) and hearts subjected to three cycles IP (n=4) with a bolus dose of the β-adrenergic receptor blocker propranolol (diluted in normal saline [0.75 mg/kg IV]) administered 9 minutes before a 10-minute period of sustained ischemia, at which time hearts were harvested for determination of cAMP.
All hearts were immersed in liquid nitrogen immediately after harvesting. We chose to harvest hearts subjected to one cycle of IP and hearts treated with polymyxin B and propranolol at 10 minutes of ischemia because the variability in cAMP measurements within the untreated control and three-cycle IP groups to which these groups were to be compared was lowest at this time point and, consequently, the statistical power was highest.
In experiments involving ischemia or ischemia and reperfusion, we delineated the area at risk from the remaining myocardium during the in vivo experiment immediately after inducing ischemia. We did this by inserting four 2-0 polypropylene suture threads onto the epicardial surface of the heart at the borderline between the area at risk and the myocardium outside this area. After coronary artery occlusion, this borderline is easily distinguishable, because there is a sharp cutoff between the cyanotic area at risk and the normally perfused myocardium adjacent to this area. For parallelism, we placed the sutures identically in the myocardial perfusion territory supplied by the left main coronary artery in hearts that were never subjected to ischemia (ie, true baseline hearts and nonischemic hearts treated with isoproterenol).
We freeze-dried the hearts and then used the polypropylene sutures as a guide to obtain a section of ≈100 mg of myocardium from within the center of the area at risk (or myocardial perfusion territory of the left main coronary artery in the case of hearts that had never been subjected to ischemia). A similar volume of tissue was obtained from virgin left ventricular myocardium remote from this region. cAMP determinations were made in both of these regions in all hearts. Great care was taken to avoid tissue sampling from the border zone.
We then pulverized and weighed the tissues and homogenized them in 6% trichloroacetic acid by using a Polytron PT 10/35 (setting at 7) for three 10-second bursts while keeping the homogenate on ice. We then centrifuged the homogenate at 1000g for 10 minutes at 4°C. We collected the supernatant and washed it in ether at five times the supernatant volume. This ether-washing step was repeated four times. The supernatant was then quick-frozen in liquid nitrogen, freeze-dried, and stored at −70°C until the time of assay. The radioimmunoassay procedure used for cAMP determination was similar to that of Steiner et al,18 as supplied by Amersham Diagnostics. In initial experiments, it was determined that recovery of [3H]cAMP (Sigma) in the freeze-dried supernatant was 85.9±0.8%. Values were not corrected for recovery, since there was little variability between measurements. We assayed all samples in duplicate; the coefficient of variation for the duplicate measurements was 5.1%. All cAMP assays were performed in a blinded fashion. cAMP measurements are expressed in nanomoles per gram dry weight.
Necrosis Determination Study
All experiments were of 3-hour and 15-minute duration and began with a 30-minute stabilization period following surgery. After the stabilization period, hearts in the control groups received an additional 45-minute stabilization period followed by 30 minutes of regional ischemia and 90 minutes of reperfusion. After the 30-minute stabilization period, hearts in the IP groups received three cycles of 5-minute ischemia and 10-minute reperfusion before 30-minute ischemia and 90-minute reperfusion.
We studied four groups of rabbits. Groups 1 (n=10) and 2 (n=13) were the untreated control and IP groups, respectively, the protocols for which are described above. Groups 3 (n=10) and 4 (n=7) were the NKH477-treated control and IP groups, respectively. The protocol for these NKH477-treated groups was identical to that for the untreated control and IP groups, except that these animals were given 45 μg/kg of NKH477 diluted in normal sterile saline 5 minutes before the 30-minute ischemia. We infused NKH477 directly into the left atrium over a 3-minute infusion period. To confirm that this dose of NKH477 increased cAMP levels within the area at risk during the 30-minute period of ischemia, we prepared rabbits in a manner identical to the control+NKH477 and IP+NKH477 groups and harvested the hearts at 10, 20, and 30 minutes of sustained ischemia (three hearts per time point per group) for determination of cAMP.
Infarct and Area-at-Risk Determination
At the end of the experiment, we excised the hearts, mounted them quickly onto a Langendorff apparatus, and perfused them with normal saline to wash out the blood. We then resnared the left main coronary artery at the in vivo occlusion site and subsequently perfused the heart with fluorescent particles (Duke Scientific) suspended in normal saline at a perfusion pressure of 75 mm Hg. With this technique, the fluorescent particles lodge in all myocardium that is not part of the area at risk, while myocardium within the area at risk remains free of particles. After delineation of the area at risk, we removed the hearts from the perfusion apparatus and sliced the ventricles transversely into 3-mm-thick slices. We then incubated the slices at 37°C for 10 minutes in a 1.25% TTC solution made up with 0.2 mol/L Tris buffer (pH 7.4). TTC stains viable myocardium brick red, whereas areas of necrosis appear a lighter tan or brown color. After TTC staining, we placed transparent sheets of acetate over each slice and traced its outline, as well as the outlines of the necrotic areas, onto the acetates. We then examined the slices under ultraviolet light to see the fluorescent particles, and the region of nonfluorescence (area at risk) was also traced onto the acetates. These areas were digitized with a digitizing tablet interfaced with a personal computer (with an Intel 486 microprocessor) and analyzed using Sigma Scan software (Jandel Scientific).
All cAMP and hemodynamic data are expressed as mean±SEM. In the drug-treated groups used for cAMP determination, we evaluated hemodynamic responses to infused drugs immediately before and 5 minutes after drug administration by using paired t tests. We used repeated-measures ANOVA to compare hemodynamic variables in the four groups of the necrosis study. We analyzed the cAMP data by using factorial ANOVA. The Student-Newman-Keuls test was used to determine significant differences between groups.
Because of the large reductions in necrosis produced by IP in our model, several of the IP (both untreated and NKH477-treated) hearts had <5% necrosis of the area at risk. As a result, the assumptions for parametric testing (ie, normal distributions and equal variances between groups being compared) were not met; thus, standard ANOVA techniques were not used for analysis of these data. We used the Mann-Whitney rank sum test, a nonparametric test, to compare infarct sizes between groups. Since the necrosis data did not meet the assumptions for parametric testing, the necrosis data are described in terms of median values for each group, and box plots are used to describe the key characteristics of the distribution. To compare the incidence of fibrillation between groups, we used Fisher’s exact test. Statistical significance was taken as P<.05.
cAMP Determination Study
There were no differences in baseline hemodynamics in any of the groups of the cAMP study (data not shown). In the drug-treated groups, there were no differences in heart rate or arterial blood pressure between the control and IP groups, either before or after drug administration for any of the drugs used. Therefore, these values were pooled for statistical analysis of drug effects. Isoproterenol increased the heart rate by 18±4% of the initial value (P<.001) and reduced the mean arterial blood pressure by 17±3% (P<.01). Propranolol reduced the heart rate by 17±2% (P<.0001) and also produced an 8±3% reduction in the mean arterial pressure (P<.05). Polymyxin B administration reduced the heart rate by 12±1% (P<.0001) and produced profound hypotension, reducing the blood pressure by 50±3% of the initial value (P<.0001).
cAMP Levels in Control and IP Hearts
All cAMP levels presented in Figs 1 to 5 show the levels present within the area at risk in all hearts that were subjected to ischemia or ischemia and reperfusion. In hearts that had never been made ischemic (ie, true baseline hearts and control hearts treated with isoproterenol), the cAMP levels in the myocardial perfusion territory of the left main coronary artery are given (ie, the area usually occupied by the area at risk in experiments involving ischemia).
As shown in Fig 1⇓, control hearts underwent a large increase in cAMP levels within the area at risk at 10 minutes of regional ischemia, and this increase persisted throughout the 30-minute ischemic period. In contrast to the substantial increases in cAMP observed in control hearts, cAMP levels in hearts subjected to three cycles of IP were not significantly higher than baseline (nonischemic) values at 10, 20, or 30 minutes of ischemia. Immediately before sustained ischemia, hearts subjected to three cycles of IP had levels of cAMP that were modestly but significantly lower than the levels in control hearts (3.1±0.2 nmol/g dry wt in IP hearts versus 3.9±0.2 nmol/g dry wt in control hearts, P<.01). There were no differences in cAMP levels between control and IP hearts in myocardium remote from the area at risk either before or at any time during the 30-minute sustained ischemia.
cAMP levels measured after 5-minute ischemia (which corresponds to both the end of the first 5 minutes of transient ischemia in IP hearts and 5 minutes of sustained ischemia in control hearts) measured 4.4±0.2 nmol/g dry wt. In IP hearts, cAMP levels measured at the end of the third 5-minute transient ischemia measured 4.4±0.4 nmol/g dry wt. The pooled cAMP levels measured at these two time points were significantly higher than in true baseline hearts (P<.05).
As shown in Fig 2⇓, IP induced with only a single cycle of 5-minute transient ischemia and 10-minute reperfusion also reduced cAMP levels during a subsequent 10-minute period of sustained ischemia (P<.001 versus control). However, cAMP levels in hearts preconditioned with only a single cycle of IP were still significantly higher than those levels measured at this time point in hearts subjected to three cycles of IP (P<.05 versus three cycles of IP).
Inhibition of PKC
As shown in Fig 3⇓, the administration of the PKC inhibitor polymyxin B tended to lessen the increase in cAMP that occurred at 10 minutes of regional ischemia in control hearts. This trend did not, however, achieve statistical significance at the .05 level (P=.10). Despite this effect, polymyxin B–treated IP hearts still had significantly lower levels of cAMP compared with the polymyxin B–treated control hearts (P<.05). Polymyxin B–treated IP hearts also did not have cAMP levels that were different from untreated IP hearts (P=NS). In the myocardium outside the area at risk, cAMP levels in the polymyxin B–treated control and IP hearts were not different from each other (4.7±0.5 in control+polymyxin B–treated hearts versus 4.1±0.5 nmol/g dry wt in IP+polymyxin B–treated hearts), nor were they different from the myocardium outside the area at risk in the untreated control and IP hearts at 10 minutes of ischemia (4.3±0.3 in untreated control hearts versus 3.7±0.2 nmol/g dry wt in untreated three-cycle IP hearts).
Responsiveness of the β-Adrenergic Effector Pathway
As shown in Fig 4⇓, the β-adrenergic receptor agonist isoproterenol significantly increased cAMP levels in the region previously made ischemic by three cycles of transient ischemia and reperfusion and also in the same myocardial territory of control hearts that were never ischemic. However, there was no difference in the cAMP response to isoproterenol between the two groups. Furthermore, when the cAMP response to isoproterenol in the area previously made ischemic by three cycles of IP was compared with myocardium outside this area in the same heart, no difference was detected (7.0±1.2 nmol/g dry wt in the area previously made ischemic versus 7.5±0.9 nmol/g dry wt in nonischemic myocardium). It is interesting to note that the cAMP levels observed within the area at risk at 10 minutes of ischemia in control hearts (9.5±0.8 nmol/g dry wt) was significantly higher than cAMP levels achieved with this high dose of isoproterenol in the pooled control and IP hearts (7.3±0.7 nmol/g dry wt, P=.05).
Blockade of the β-Adrenergic Receptor
As shown in Fig 5⇓, administration of the β-adrenergic receptor blocker propranolol completely abolished the increase in cAMP levels that occurred at 10 minutes of regional ischemia in control hearts but did not reduce cAMP levels in IP hearts. In the myocardium outside the area at risk, cAMP levels in propranolol-treated control and IP hearts were not different from each other (3.7±0.3 in control+propranolol-treated hearts versus 3.8±0.4 nmol/g dry wt in IP+propranolol-treated hearts), nor were they different from the myocardium outside the area at risk in untreated control and IP hearts at 10 minutes of ischemia (4.3±0.3 in untreated control hearts versus 3.7±0.2 nmol/g dry wt in untreated three-cycle IP hearts).
Necrosis Determination Study
The hemodynamics recorded in the necrosis study are presented in the Table⇓. NKH477 significantly increased heart rate at 15 and 30 minutes of sustained ischemia in both the control and IP groups. End-reperfusion hemodynamics were not different among these four groups. Fig 6⇓ shows the infarct size expressed as a percentage of the area at risk in all groups. The median value for infarct size in the untreated control group was 61.9% of the area at risk. Three cycles of transient ischemia and reperfusion significantly protected the heart from infarction: the median necrosis value was only 0.3% of the area at risk (P<.0001 using the Mann-Whitney rank sum test). IP induced with three cycles of transient ischemia and reperfusion was also highly effective in reducing necrosis in the presence of NKH477 (median necrosis, 0.6% of the area at risk in IP+NKH477-treated group versus 62.9% of the area at risk in control+NKH477-treated group, P<.05). There was no detectable difference in infarct size between the untreated control and control+NKH477-treated groups.
Infarct sizes expressed as mean±SEM for the four groups in this necrosis study were as follows: 53.6±5.5% (of the area at risk) in the untreated control group, 3.2±1.3% in the untreated IP group, 50.2±7.7% in control+NKH477-treated group, and 10.0±5.9% in the IP+NKH477-treated group. There were no significant differences in ventricular area or area at risk between any of the four groups of this study.
Before initiating the infarct-size study, we confirmed that the dose of NKH477 used (45 μg/kg) increased cAMP levels within the area at risk during the 30-minute period of regional ischemia in both the control and IP groups. Control and IP hearts treated with NKH477 were harvested for cAMP measurements at 10, 20, and 30 minutes of ischemia. As shown in Fig 7⇓, average cAMP levels (obtained by averaging cAMP levels measured at these three ischemic time points) were higher in the area at risk in IP hearts treated with NKH477 than in untreated IP hearts (P<.05) but were not different from the untreated control hearts. NKH477 also significantly increased cAMP levels in the control+NKH477-treated group (P<.05 versus untreated control group). There was no difference between control and IP hearts in the percentage increase in cAMP by NKH477 during ischemia (66±34% in control hearts and 96±24% in IP hearts, P=NS).
Incidence of Fibrillation
In the untreated control and IP groups, 1 of 10 and 2 of 13 rabbits, respectively, developed ventricular fibrillation during sustained ischemia. All 3 of these rabbits were successfully defibrillated after one attempt. NKH477 significantly increased the incidence of ventricular fibrillation in the control group (7 of 10 or 70% of the rabbits fibrillated, P<.05 versus untreated control group by Fisher’s exact test) and also in the IP+NKH477-treated group (4 of 7 or 57% of the rabbits fibrillated, P<.05 versus untreated IP group). There was no significant difference in the incidence of fibrillation between the NKH477-treated control and NKH477-treated IP groups.
The most important observations of the present study were as follows: (1) cAMP levels were increased nearly twofold during regional myocardial ischemia (ie, <30 minutes) in rabbit hearts. (2) This increase in cAMP levels was attenuated when ischemia was preceded by a single cycle of transient ischemia and reperfusion and prevented when ischemia was preceded by three cycles of transient ischemia and reperfusion. (3) Inhibition of PKC with polymyxin B did not block the cAMP-lowering effect of IP. (4) Stimulation of the β-adrenergic receptor with isoproterenol produced similar increases in cAMP in nonischemic myocardium and in myocardium subjected to three cycles of transient ischemia and reperfusion. (5) Blockade of the β-adrenergic receptor with propranolol abolished the increase in cAMP that occurred with sustained ischemia in control hearts but did not reduce cAMP levels in IP hearts. (6) Activation of adenylyl cyclase with NKH477 increased cAMP levels during sustained ischemia but did not increase necrosis in control hearts or block the protection of IP against necrosis.
Possible Mechanisms of Reduced cAMP in IP
Activation of PKC
Evidence has emerged suggesting that activation of PKC is the final effector pathway responsible for the protection of IP against necrosis3 and postischemic dysfunction.4 Activation of PKC is also known to produce several effects on the cAMP-generating and -degrading pathways. However, the net effect of PKC activation on myocardial cAMP levels is difficult to predict from available studies, because one of these effects would be expected to decrease cAMP (eg, uncoupling of adenylyl cyclase–coupled receptors from Gs protein), whereas others would be expected to increase cAMP (eg, ablation of Gi function19 and direct sensitization of adenylyl cyclase20 ). In the present study, we demonstrated that the PKC inhibitor polymyxin B did not raise cAMP levels during sustained ischemia in IP hearts. This suggests that the cAMP-lowering effect of IP is not mediated by a PKC-dependent process. Although we found that the dose of polymyxin B used produced a 50% reduction in mean arterial blood pressure, we do not believe that this influenced our conclusion, since the extent of hypotension was similar between control and IP groups. Furthermore, polymyxin B did not affect cAMP levels in nonischemic myocardium outside the area at risk.
In control hearts, polymyxin B tended to blunt the rise in cAMP levels observed during sustained ischemia. One possible explanation for this effect is provided by data from Strasser et al,21 who demonstrated that in the isolated rat heart, adenylyl cyclase is itself sensitized between 5 and 20 minutes of sustained myocardial ischemia and that this sensitization was completely prevented by preperfusing these hearts with either the PKC inhibitors polymyxin B or staurosporine. Thus, in the present study, polymyxin B may have reduced the rise in cAMP in control hearts during ischemia by preventing the PKC-mediated sensitization of adenylyl cyclase.
The dose of polymyxin B used to investigate the effect of PKC on cAMP levels in the present study was selected because this dose blocks the necrosis protection of IP induced with a single cycle of 5-minute ischemia and 10-minute reperfusion in our in vivo model of regional ischemia and reperfusion (authors’ unpublished data, 1995) as well as in a previous study in the rabbit.3 Although we presume that blockade of the antinecrosis effect of IP was mediated by the inhibition of PKC by polymyxin B, we must point out that we did not directly demonstrate that PKC was inhibited in our necrosis study. Therefore, it is possible that the blockade of necrosis protection by polymyxin B was mediated by a mechanism other than through its inhibitory effect on PKC.
Reduced Responsiveness of the β-Adrenergic Effector Pathway
Our results show that control and IP hearts had a similar increase in cAMP in response to the β-adrenergic receptor agonist isoproterenol and thus argue against our hypothesis that the β-adrenergic effector pathway is desensitized by transient ischemia and reperfusion. Further evidence against β-adrenergic receptor desensitization in IP comes from the work of Iwase et al,22 who compared the changes in myocardial β-adrenergic receptor number and the functional activities of Gs and adenylyl cyclase during sustained ischemia in control and IP rabbit hearts. These investigators showed that IP did not affect the sarcolemmal density of β-adrenergic receptors, although it prevented the early reductions in Gs and adenylyl cyclase activity. Furthermore, in reconstituted cell membranes prepared from the sarcolemma of control and IP hearts that had undergone sustained ischemia, these investigators22 found that cAMP production in response to isoproterenol was actually enhanced in IP hearts. Thus, their data indicate that not only did IP not desensitize β-adrenergic signal transduction in ischemia, it was found to preserve it. Thus, our findings and those of Iwase et al both suggest that the β-adrenergic signal transduction pathway is not desensitized by IP.
Recently, Niroomand et al23 have reported that in canine myocardium, Gi proteins were sensitized during reperfusion following a single 5-minute period of transient ischemia and that this sensitization was maintained during a subsequent period of ischemia. Furthermore, these investigators demonstrated that this sensitization of Gi led to an increase in Gi-mediated inhibition of adenylyl cyclase in response to isoproterenol. Although enhanced inhibition of adenylyl cyclase by Gi could offer a potential explanation for the lack of elevation in cAMP levels with IP, we believe that this is unlikely. In our β-adrenergic responsiveness experiments, the cAMP levels measured in response to isoproterenol reflect not just the cAMP production that is mediated through the β-adrenergic receptor/Gs/adenylyl cyclase effector pathway but, rather, the summation of all influences that alter cAMP levels in response to β-adrenergic receptor stimulation. Thus, if three cycles of transient ischemia and reperfusion had increased the activation of Gi, this effect would also have been detected as a lesser rise in cAMP levels in response to isoproterenol in IP hearts compared with nonischemic hearts. Since this was not observed, these data would argue against an important role for the activation of Gi in the cAMP-lowering effect of IP. Adenosine receptor stimulation is also not likely to be responsible for the lack of elevation in cAMP levels in IP hearts, because in pilot experiments we determined that the nonspecific adenosine receptor blocker 8-p-sulfophenyl-theophylline, given at a dose of 10 mg/kg 5 minutes before each of the three episodes of transient ischemia, also did not raise cAMP levels during sustained ischemia in IP hearts (authors’ unpublished data, 1994).
Another possible explanation for the lack of elevation in cAMP levels in IP hearts is the increased activation of phosphodiesterases. However, two pieces of evidence argue against this possibility. First, as mentioned above, the cAMP levels measured in response to isoproterenol reflect the summation of all influences that alter cAMP levels in response to β-adrenergic receptor stimulation. Therefore, if phosphodiesterase activity was increased by three cycles of transient ischemia and reperfusion, this would also have been detected as an attenuated increase in cAMP in response to isoproterenol in these IP hearts compared with nonischemic hearts, which was not observed. Second, if phosphodiesterase activity was higher in IP hearts during sustained ischemia, we would have expected IP hearts to have exhibited a lesser rise in cAMP after the administration of the adenylyl cyclase activator NKH477 compared with control hearts. However, cAMP levels during ischemia rose an average of 96% and 66% in IP and control hearts treated with NKH477, respectively. Since the increase in cAMP levels after the administration of NKH477 was clearly not reduced in IP hearts, these data also do not support a role for phosphodiesterase activation in the cAMP-lowering effect of IP.
In the present study, we investigated the cAMP response to a single high dose of isoproterenol. Although the generation of a dose-response curve using lower doses of isoproterenol may have allowed us to detect subtle differences between the control and IP groups, there was no trend toward reduced responsiveness to isoproterenol in hearts subjected to three cycles of transient ischemia and reperfusion. Therefore, we are confident of our conclusion that reduced responsiveness of the β-adrenergic effector pathway cannot account for the lack of elevation in cAMP levels observed with IP.
Reduced Stimulation of the β-Adrenergic Receptor
Several lines of evidence support the conclusion that the high levels of cAMP we observed with sustained ischemia were caused by increased norepinephrine release leading to increased β-adrenergic receptor activation and, furthermore, that the lack of elevation in cAMP levels in IP hearts was mediated by reduced β-adrenergic receptor stimulation by norepinephrine release.
With respect to the increase in cAMP levels that occurred with sustained ischemia (control hearts), our data demonstrate that β-receptor blockade with propranolol completely abolished this increase in cAMP levels, thereby establishing that β-adrenergic receptor activation was responsible for this rise. Previous studies have also reported that β-adrenergic receptor blockade can either prevent24 or reduce25 ischemia-induced increases in cAMP. In addition, in rats with severely depleted myocardial norepinephrine in association with cardiac hypertrophy, increases in cAMP levels do not occur with ischemia.26
In further support of the conclusion that the increased cAMP levels we observed with myocardial ischemia were caused by increased norepinephrine release, Schömig8 has reported that periods of myocardial ischemia of >10 minutes result in extremely high extracellular concentrations of norepinephrine in the ischemic myocardium (100 to 1000 times normal plasma concentrations). This release of norepinephrine (called nonexocytotic release) does not depend on local sympathetic activation but, rather, on local metabolic factors. These factors cause the uptake1 carrier, which normally functions to reuptake norepinephrine into the sympathetic nerve terminal, to reverse its normal transport direction and produce efflux of norepinephrine from the cytoplasm of the nerve terminals to the extracellular space.8 The time frame during which this nonexocytotic norepinephrine release has been observed (between 10 and 40 minutes of ischemia) is consistent with the time frame of increased cAMP levels (between 10 and 30 minutes of ischemia) in the control hearts in the present study. Before 10 minutes of ischemia, norepinephrine release has been shown to be much less than that occurring with longer ischemic times and to occur through the normal exocytotic process.8 This observation is also consistent with the modest increase in cAMP we observed after 5 minutes of ischemia.
There are also multiple pieces of evidence suggesting that the lack of elevation in cAMP levels in IP hearts is mediated by reduced β-adrenergic receptor activation secondary to reduced norepinephrine release. First, in contrast to the dramatic reductions in cAMP we achieved with β-adrenergic receptor blockade in control hearts, β-blockade had no effect on cAMP levels in IP hearts, thereby supporting the notion that β-adrenergic receptor–mediated cAMP production was already minimal in these hearts. Since our β-adrenergic receptor responsiveness experiments and the study by Iwase et al22 have demonstrated that the β-adrenergic receptor signal transduction pathway is not desensitized by IP, a logical explanation for the lack of elevation in cAMP levels in IP hearts during sustained ischemia is through reduced stimulation of the β-adrenergic receptor by reduced norepinephrine release. Furthermore, Seyfarth et al9 have recently demonstrated that IP results in dramatic reductions in norepinephrine release during sustained ischemia in the isolated rat heart. It is also interesting to note that Seyfarth et al have demonstrated a clear dose dependence, with multiple cycles of IP producing a greater attenuation of norepinephrine release during sustained ischemia compared with a single cycle of IP. We have also observed a similar dose dependence with respect to cAMP levels: three cycles of IP completely prevented the rise in cAMP that occurs with sustained ischemia, whereas a single cycle resulted in only an attenuation of the rise.
Although the mechanism responsible for the reduction of norepinephrine release during sustained ischemia by IP is not known, it should be stressed that this effect is not through depletion of norepinephrine stores by transient ischemia. In fact, Banerjee et al27 report that <3% of the total norepinephrine stores in the rat heart are released by brief periods of myocardial ischemia. Further evidence that argues against the operation of a norepinephrine depletion mechanism in IP comes from the work of Miyazaki and Zipes,28 who demonstrated that IP protects the dog heart against sympathetic denervation during the first hour of coronary occlusion and preserves sympathetic reflexes.
Although our evidence (which indicates that the lack of elevation in cAMP levels during sustained ischemia in IP hearts is mediated by a suppression of norepinephrine release) is indirect, these data taken together with previous studies by Schömig,8 Seyfarth et al,9 and Miyazaki and Zipes28 argue strongly that IP acts to protect the sympathetic nerves and either reduces or delays the nonexocytotic release of norepinephrine, which in turn prevents the intracellular elevation in cAMP that normally accompanies sustained ischemia. Our β-receptor responsiveness data do not support the hypothesis that the cAMP-lowering effect of IP is mediated through increased inhibition of adenylyl cyclase through Gi.
Role of Low cAMP Levels in the Necrosis Protection of IP
In untreated hearts, IP using three cycles of transient ischemia and reperfusion resulted in a >90% reduction in necrosis in our model, which, to our knowledge, is the greatest reduction in necrosis ever reported with IP. The infarct size–limiting effect of IP was still highly evident in the presence of high levels of cAMP during sustained ischemia, indicating that a reduction in cAMP levels was not necessary for its protection. Significant elevations in cAMP also did not produce any detectable increase in necrosis in control hearts, an effect that might have been expected if these high levels of cAMP aggravated ischemic injury.
In the present study, the adenylyl cyclase activator NKH477 was used to increase cAMP levels. NKH477 is a derivative of the drug forskolin but differs from forskolin in that it is water soluble, longer acting, and more than twice as efficacious in stimulating cardiac adenylyl cyclase activity.29 The dose of NKH477 used in the present study was selected on the basis of its ability to increase cAMP levels within the area at risk during the 30 minutes of sustained ischemia in IP hearts to levels that were similar to levels in untreated control hearts. This dose of NKH477 also significantly increased heart rate throughout the 30-minute period of ischemia, a known effect of elevated cAMP levels.
In a model identical to that used in the present study, we have previously shown that infusion of the adenylyl cyclase activator forskolin throughout the three cycles of transient ischemia and reperfusion in IP hearts also did not block the protection of IP against necrosis, even though it significantly increased cAMP levels.30 This indicates that the reduction in cAMP levels observed in IP hearts during the transient reperfusion period before the sustained ischemia was also not necessary for its protection against necrosis.
Could the Reduction in cAMP Levels in IP Be Responsible for Its Antiarrhythmic Effect?
In addition to protecting against necrosis, IP has also been shown to protect against arrhythmias in several species,31 including the conscious rabbit.32 A large body of experimental evidence also suggests that high levels of cAMP are causally linked to the occurrence of arrhythmias and that reducing cAMP levels with drugs such as β-adrenergic receptor blockers have antiarrhythmic effects.33 Therefore, the possibility exists that the absence of a rise in cAMP levels we observed in IP hearts was causally linked to its antiarrhythmic effect. The present study was not designed to address the issue of arrhythmias, because anesthetized rabbits are not prone to them in the absence of some arrhythmia-inducing intervention. Thus, on the basis of the infrequent occurrence of ventricular fibrillation in both the untreated control and IP hearts, it is not possible to say whether or not IP protected against arrhythmias in our model. However, we did note that elevating cAMP levels with NKH477 significantly increased the occurrence of fibrillation in both the control and IP groups, thereby suggesting that high levels of cAMP are proarrhythmic in our model. Interestingly, the increase in fibrillation in NKH477-treated control hearts was not accompanied by increased necrosis. Previous work, reported by Opie et al34 has also demonstrated that isoproterenol, given at a dose that significantly increases cAMP, increased fibrillation without increasing infarct size in a porcine regional ischemia/reperfusion model.
Implicit in the proposition that IP may protect against arrhythmias through a reduction of cAMP (but that this is not involved in its protection against necrosis) is the notion that IP protects against arrhythmias and necrosis by separate specific mechanisms. This has recently been suggested by Cohen et al,32 who observed that preconditioning protocols used to produce protection against necrosis did not protect against arrhythmias and, furthermore, that protection against arrhythmias can be achieved without protection against necrosis. Further evidence suggesting that the protection of IP against arrhythmias and necrosis may be mediated by different pathways comes from the work of Speechly-Dick et al,35 who showed that in the rat, the specific PKC inhibitor chelerythrine completely blocked the protection of IP against necrosis, whereas it was unable to block the protection of IP against arrhythmias. Furthermore, these investigators found that the specific PKC agonist 1,2-dioctanoyl-sn-glycerol mimicked the protection of IP against necrosis but was unable to protect against arrhythmias.
Also noteworthy is that the protection of IP against necrosis has been suggested to be an all-or-none phenomenon, with one cycle of transient ischemia conferring similar protection to that observed with multiple cycles.36 However, Lawson and Hearse31 have shown that the protection of IP against arrhythmias shows a definite dose dependence, with a cumulative increase in protection against arrhythmias when the number of cycles used to produce IP is increased from one to three. As mentioned above, the ability of IP to suppress cAMP levels and norepinephrine release during sustained ischemia also shows a similar dose dependence. Thus, the extent to which IP protects against arrhythmias may depend on the extent to which norepinephrine release and, subsequently, myocardial intracellular cAMP levels are reduced by the transient cycles of ischemia and reperfusion.
In summary, IP prevents the rise in cAMP levels that occurs during sustained ischemia. This effect is not mediated by activation of PKC or by reduced responsiveness of the β-adrenergic effector pathway but most likely through an attenuation of norepinephrine release. Furthermore, the antinecrotic effect of IP is not mediated by the lack of elevation in cAMP levels during sustained ischemia. However, the cAMP-lowering effect of IP may well be involved in its protection against arrhythmias. Further studies designed specifically to address the issue of arrhythmias are needed to define the precise relation between the prevention of norepinephrine release, the reduction of intracellular cAMP levels, and the antiarrhythmic effect of IP.
Selected Abbreviations and Acronyms
|Gi protein||=||inhibitory G protein|
|Gs protein||=||stimulatory G protein|
|PKC||=||protein kinase C|
This study was supported by Ontario Heart and Stroke Foundation grant T-2687. Dr Sandhu was the recipient of an Ontario Ministry of Health Fellowship. Dr Wilson was the recipient of a Career Investigator Award of the Ontario Heart and Stroke Foundation. We thank Dr A.K. Sen for advice. We gratefully acknowledge Nippon Kyauka Pharmaceuticals for providing NKH477.
- Received June 29, 1995.
- Accepted September 25, 1995.
- © 1996 American Heart Association, Inc.
Reimer KA, Vander Heide RS, Murry CE, Jennings RB. Role of altered energy metabolism in ischemic preconditioning. In: Przyklenk K, Kloner R, Yellon D, eds. Ischemic Preconditioning: The Concept of Endogenous Cardioprotection. Norwell, Mass: Kluwer Academic Publishing; 1994:75-103.
Szilvassy Z, Ferdinandy P, Bor P, Jakab I, Lonovics J, Koltai M. Ventricular overdrive pacing-induced anti-ischemic effect: a conscious rabbit model of preconditioning. Am J Physiol. 1994;266:H2033-H2041.
Ytrehus K, Liu Y, Downey JM. Preconditioning protects ischemic rabbit heart by protein kinase C activation. Am J Physiol. 1994;266:H1145-1152.
Mitchell MB, Meng X, Ao L, Brown JM, Harken AH, Banerjee A. Preconditioning of isolated rat heart is mediated by protein kinase C. Circ Res. 1995;76:73-81.
Roth NS, Campbell PT, Caron MG, Lefkowitz RJ, Lohse MJ. Comparative rates of desensitization of β-adrenergic receptors by the β-adrenergic receptor kinase and the cyclic AMP-dependent protein kinase. Proc Natl Acad Sci U S A. 1991;88:6201-6204.
Lohse MJ, Benovic JL, Codina J, Caron MG, Lefkowitz RJ. b-Arrestin: a protein that regulates beta-adrenergic receptor function. Science. 1990;248:1547-1550.
Schömig A. Catecholamines in myocardial ischemia: systemic and cardiac release. Circulation. 1990;82(suppl 2):II-13-II-22.
Seyfarth M, Münch G, Schreieck J, Kurz T, Richardt G, Schömig A. Release of norepinephrine is suppressed by preconditioning in rat ischemic hearts. Circulation. 1994;90(suppl 1):I-108. Abstract.
Thornton JD, Liu GS, Olsson RA, Downey JM. Intravenous pretreatment with A1-selective adenosine analogues protects the heart against infarction. Circulation. 1992;85:659-665.
Houslay MD, Bushfield M, Kilgour E, Lavan B, Griffiths S, Pyne NJ, Tang EK-Y, Murphy GJ. Glucagon activates two distinct signal transduction systems in hepatocytes, which leads to the desensitization of G-protein-regulated adenylate cyclase, the phosphorylation and inactivation of Gi-2 and the phosphorylation and stimulation of a specific cyclic AMP phosphodiesterase. In: Konijn T, Houslay M, Van Haastert P, eds. Activation and Desensitization of Transducing Pathways. Berlin, Germany: Springer-Verlag; 1990;44:63-83. NATO ASI Series.
Strasser RH, Braun-Dullaeus R, Walendzik H, Marquetant R. α1-Receptor–independent activation of protein kinase C in acute myocardial ischemia: mechanisms for sensitization of the adenylyl cyclase system. Circ Res. 1992;70:1304-1312.
Iwase T, Murakami T, Tomita T, Miki S, Nagai K, Sasayama S. Ischemic preconditioning is associated with a delay in ischemia-induced reduction of β-adrenergic signal transduction in rabbit hearts. Circulation. 1993;88:2827-2837.
Niroomand F, Weinbrenner C, Weis A, Bangert M, Schwencke C, Marquetant R, Beyer T, Strasser RH, Kübler W, Rauch B. Impaired function of inhibitory G proteins during acute myocardial ischemia of canine hearts and its reversal during reperfusion and a second period of ischemia: possible implications for the protective mechanism of ischemic preconditioning. Circ Res. 1995;76:861-870.
Ogawa K, Nakamura T. Effect of propranolol on cyclic nucleotides of ischemic dog heart. In: Tajuddin M, Bhatia B, Siddiqui H, Rona G, eds. Advances in Myocardiology. Baltimore, Md: University Park Press; 1980;2:449-4455.
Corr PB, Witkowski FX, Sobel BE. Mechanisms contributing to malignant dysrhythmias induced by ischemia in the cat. J Clin Invest. 1978;61:109-119.
Shahab L, Wollenberger A, Krause EG, Genz S. Changes in catecholamine mobilization and cyclic AMP levels in normal and hypertrophied rat heart. In: Oliver M, Julian D, Donald K, eds. Effect of Acute Ischemia on Myocardial Function. London, England: Churchill Livingstone; 1972:97-107.
Banerjee A, Locke-Winter C, Rogers KB, Mitchell MB, Brew EC, Cairns CB, Bensard DD, Harken AH. Preconditioning against myocardial dysfunction after ischemia and reperfusion by an α1-adrenergic mechanism. Circ Res. 1993;73:656-670.
Miyazaki T, Zipes DP. Protection against autonomic denervation following acute myocardial infarction by preconditioning ischemia. Circ Res. 1989;64:437-448.
Sandhu R, Diaz RJ, Wilson GJ. Raising cAMP levels using forskolin pretreatment does not block ischemic preconditioning. J Mol Cell Cardiol. 1994;26:136. Abstract.
Lawson CS, Hearse DJ. Preconditioning and ischemia- and reperfusion-induced arrhythmias. In: Przyklenk K, Kloner R, Yellon D, eds. Ischemic Preconditioning: The Concept of Endogenous Cardioprotection. Norwell, Mass: Kluwer Academic Publishers; 1994:19-40.
Cohen MV, Yang X-M, Downey JM. Conscious rabbits become tolerant to multiple episodes of ischemic preconditioning. Circ Res. 1994;74:998-1004.
Lubbe WF, Podzuweit T, Opie LH. Potential arrhythmogenic role of cyclic adenosine monophosphate (AMP) and cytosolic calcium overload: implications for prophylactic effects of beta-blockers in myocardial infarction and proarrhythmic effects of phosphodiesterase inhibitors. J Am Coll Cardiol. 1992;19:1622-1633.
Speechly-Dick ME, Mocanu MM, Yellon DM. Protein kinase C: its role in ischemic preconditioning in the rat. Circ Res. 1994;75:586-590.
Li GC, Vasquez JA, Gallagher KP, Lucchesi BR. Myocardial protection with preconditioning. Circulation. 1990;82:609-619.