PD 81,723, an Allosteric Enhancer of the A1 Adenosine Receptor, Lowers the Threshold for Ischemic Preconditioning in Dogs
PD 81,723 (PD) acts allosterically to increase agonist binding to A1 adenosine receptors and to enhance functional A1 receptor–mediated responses in the heart and other tissues. To determine if PD lowers the threshold for ischemic preconditioning (PC), pentobarbital-anesthetized dogs were subjected to 60 minutes of left anterior descending coronary artery (LAD) occlusion and 3 hours of reperfusion. Ischemic PC was produced by either 2.5 or 5 minutes of LAD occlusion 10 minutes before the 60-minute occlusion. PD (100 μg/kg total dose, 5 to 50 μmol/L in coronary arterial blood) or vehicle was infused intracoronarily for 17.5 minutes before the 60-minute occlusion period in non-PC dogs or in dogs preconditioned with 2.5 minutes of ischemia. Myocardial infarct size was determined by triphenyltetrazolium staining and expressed as a percentage of the area at risk. Compared with the control group (26.3±3.6%, mean±SEM), infarct size was not significantly affected by 2.5 minutes of PC alone (23.4±4.2%) or by PD alone (26.5±1.7%) but was decreased by PD+PC (14.6±1.7%, P<.05) or by a longer period (5 minutes) of PC alone (12.5±3.3%). The intravenous administration of the selective antagonist of A1 adenosine receptors, 8-cyclopentyl-1,3-dipropylxanthine (1 mg/kg), or the ATP-sensitive K+ channel blocker, glibenclamide (0.3 mg/kg), for 15 minutes before PD+PC blocked the protection (23.6±2.3% or 25.9±3.3%, respectively). None of the compounds studied affected systemic hemodynamics, collateral blood flow, or AAR. To determine which subtypes of canine adenosine receptors were affected by 10 μmol/L PD, radioligand binding studies were conducted using membranes derived from COS-7 cells expressing recombinant canine receptors and agonist radioligands. PD enhanced the binding of [125I]N6-4-amino-3-iodobenzyladenosine (125I-ABA) to A1 receptors by increasing the t1/2 for dissociation by 2.18-fold, but PD had no effect on the dissociation kinetics of 125I-ABA from A3 receptors or [125I]-[2-(4-amino-3-iodo-phenyl)ethylamino]adenosine from A2A receptors. Glibenclamide at concentrations up to 10 μmol/L had no effect on the binding of radioligands to recombinant canine A1, A2A, or A3 receptors. These data suggest that PD reduces the amount of time required for ischemia to produce preconditioning by enhancing adenosine binding to its A1 receptor. Glibenclamide prevents the protection afforded by A1 receptor activation by a mechanism not involving adenosine receptor blockade.
- ischemic preconditioning
- adenosine receptors
- ATP-sensitive K+ channels
- allosteric enhancers
Ischemic PC, which was first described by Murry et al1 in 1986, has been widely studied by a number of laboratories. Although the mechanism by which brief periods of ischemia protect the heart from a more prolonged period of ischemia is still not fully understood, it is known that PC must be maintained for a minimum period of time (threshold) to be effective. In anesthetized dogs, a single 5-minute coronary artery occlusion followed by a 10-minute reperfusion period is sufficient to produce ischemic PC.2 On the other hand, Yao and Gross3 have recently shown that 3 minutes of PC was not sufficient to reduce infarct size in anesthetized dogs. Similarly, Miura et al4 demonstrated that 2 minutes of PC was not sufficient to reduce infarct size in anesthetized rabbits unless the rabbits were pretreated with the adenosine uptake inhibitor dipyridamole. They concluded that stimulation of adenosine receptors by endogenous adenosine mediates the infarct size–reducing effect afforded by ischemic PC. Consistent with this finding, adenosine administration has been shown to mimic the effect of ischemic PC to reduce infarct size,5 and it is likely that activation of the A1 adenosine receptor is, at least in part, involved in this phenomenon.6 Endogenous adenosine has also been shown to mediate ischemic PC in swine.7
PD 81,723 is an allosteric enhancer of the A1 adenosine receptor.8 This compound has been found to increase radioligand binding and functional activation of A1 receptors in rat brain and heart, guinea pig heart, and human myocardium and in CHO cells stably transfected with recombinant human A1 receptors.8 9 10 11 12 The enhancer has been shown to stabilize agonist receptor–G protein complexes.12 Thus, in the present study, we wished to determine whether PD 81,723 lowers the threshold for ischemic PC in intact dogs. Second, by use of radioligand analysis with recombinant adenosine receptors, we determined whether PD 81,723 acts selectively on A1 receptors or whether it also influences other adenosine receptor subtypes. Furthermore, since the A1 adenosine receptor has been shown to be coupled to the KATP channel via a Gi protein13 and opening of myocardial KATP channels results in ischemic PC,14 we determined whether the effect of PD 81,723 is also blocked by the sulfonylurea KATP antagonist glibenclamide. Finally, since glibenclamide blocks responses to adenosine in several tissues in a manner resembling a competitive receptor antagonist,15 we determined whether glibenclamide competes for radioligand binding to recombinant canine A1, A2A, and A3 adenosine receptors.
Materials and Methods
General Preparation of Dogs
All experiments conducted in the present study were in accordance with the Position of the American Heart Association on Research and Animal Use, adopted in 1984 by the American Heart Association, and the guidelines of the animal care committee of the Medical College of Wisconsin. The Medical College of Wisconsin is accredited by the American Association of Laboratory Animal Care.
Adult mongrel dogs of either sex, weighing 19.0 to 28.0 kg, were fasted overnight, anesthetized with a combination of sodium barbital (200 mg/kg) and sodium pentobarbital (15 mg/kg), and ventilated with room air supplemented with 100% oxygen. Atelectasis was prevented by maintaining an end-expiratory pressure of 5 to 7 cm H2O with a trap. Arterial blood pH, Pco2, and Po2 were monitored at selected intervals by an AVL automatic blood gas system and maintained within normal physiological limits (pH 7.35 to 7.45; Pco2, 30 to 35 mm Hg; and Po2, 85 to 100 mm Hg) by adjusting the respiration rate and oxygen flow or by intravenous administration of 1.5% sodium bicarbonate if necessary. Body temperature was maintained at 38±1°C with a heating pad. Aortic blood pressure and left ventricular pressure were monitored by inserting a double-pressure transducer–tipped catheter (model PC 771, Millar) into the aorta and left ventricle via the left carotid artery. Left ventricular dP/dt was recorded by electronic differentiation of the left ventricular pressure pulse, and heart rate was determined by a tachometer. The right femoral vein and artery were cannulated for drug administration and for analyses of blood gases and myocardial tissue blood flow, respectively. A left thoracotomy was performed at the fifth intercostal space, the lung was carefully retracted, and the pericardium was incised. The heart was then suspended in a pericardial cradle. A proximal portion of the LAD distal to the first diagonal branch was isolated from surrounding tissue, and a calibrated electromagnetic flow probe (Statham SP 7515) was placed around the vessel. A flowmeter (Statham 2202) was used to measure blood flow. A mechanical occluder was placed distal to the flow probe such that there were no branches between the flow probe and the occluder. The occluder was used to zero the flow probe (20 minutes before the 60-minute experimental occlusion, the LAD was occluded for 10 seconds), to occlude the vessel, and to reperfuse the myocardium. If the basal heart rate was <150 bpm, the heart was paced at that rate with rectangular pulses of 4-millisecond duration and a voltage twice threshold via bipolar electrodes clipped to the left atrial appendage. Pacing was not used in the few animals with initial rates of >150 bpm. Hemodynamic variables, heart rate, and coronary blood flow were monitored and recorded by a polygraph throughout the experiment. The left atrium was cannulated via the appendage for radioactive microsphere injection.
Animals were assigned to one of seven groups, and all dogs were subjected to 60 minutes of LAD occlusion followed by 3 hours of reperfusion (Fig 1⇓). Approximately 17.5 minutes before the 60-minute occlusion, either PD 81,723 (100 μg/kg) or an equivalent volume of vehicle (0.5 mL polyethylene glycol and Emulphor EL 620) was administered intracoronarily distal to the occlusion site and continued to the time of occlusion in the group administered PD 81,723 (PD group) and the control group, respectively. This dose of PD 81,723 was chosen so that ≈5 to 50 μmol/L of the drug was present in the coronary circulation during coronary perfusion and occlusion, since previous studies9 10 11 12 have shown that 5 to 50 μmol/L of PD 81,723 produces an optimal enhancement of A1 receptor binding for adenosine or selective A1 agonists, whereas at concentrations above 50 μmol/L, PD 81,723 produces a competitive blockade. In the next two groups, 2.5′ PC was elicited before the longer occlusion, and the same dose of PD 81,723 (PD+PC group) or vehicle (vehicle PC group) was given 5 minutes before 2.5′ PC and continued to the time of the longer occlusion. In two other groups, either 1 mg/kg of the selective adenosine A1 receptor antagonist DPCPX (D+PD+PC group) or 0.3 mg/kg of the KATP channel antagonist glibenclamide (G+PD+PC group) was administered intravenously ≈15 minutes before intracoronary PD 81,723 infusion. These doses of DPCPX and glibenclamide have been shown by our laboratory to block ischemic PC in dogs without affecting infarct size in the absence of PC.14 16 For purposes of comparison with the 2.5′ PC group, a 5′ PC group was also included. In all groups, hemodynamics, blood gas analyses, and myocardial blood flow measurements were performed at 30 minutes into the 60-minute occlusion period. After reperfusion, hemodynamics were measured every hour, and myocardial blood flow was determined at the end of the 3-hour reperfusion period. Finally, at the end of the experiment, the hearts were electrically fibrillated, removed, and prepared for infarct size determination and regional myocardial blood flow measurements.
Infarct Size Determination
At the end of the 3-hour reperfusion period, the LAD was cannulated. To determine the anatomic AAR and the nonischemic area, 5 mL of Patent blue dye and 5 mL of saline were injected at equal pressure into the left atrium and LAD, respectively. The heart was then immediately fibrillated and removed. The left ventricle was dissected and sliced into serial transverse sections 6 to 7 mm in width. The nonstained ischemic area and the blue-stained normal area were separated, and both regions were incubated at 37°C for 15 minutes in 1% TTC (Sigma Chemical Co) in 0.1 mol/L phosphate buffer adjusted to pH 7.4. The TTC stains the noninfarcted myocardium a brick red color, indicating the presence of a formazan precipitate that results from the reduction of TTC by dehydrogenase enzymes in viable tissue. After storage overnight in 10% formaldehyde, infarcted and noninfarcted tissues within the AAR were separated and determined gravimetrically. Infarct size was expressed as a percentage of the AAR.
Regional Myocardial Blood Flow Measurement
Regional myocardial blood flow was measured by the radioactive microsphere technique as previously described in this laboratory.14 16 Microspheres were administered at 30 minutes into the prolonged 60-minute occlusion period as well as at the end of reperfusion. Carbonized plastic microspheres (15-μm diameter, New England Nuclear) labeled with 141Ce or 95Nb were suspended in isotonic saline with 0.01% Tween-80 added to prevent aggregation. The microspheres were sonicated for 5 minutes and vortexed for another 5 minutes before injection. One milliliter of the microsphere suspension (2 to 4×106 spheres) was given via the left atrial catheter and flushed by 5 mL of saline. A reference blood flow sample was withdrawn from the right femoral artery at a constant rate of 9.4 mL/min starting 30 seconds before the microsphere injection and continuing for 3 minutes. On the following day, the tissue slices were sectioned into subepicardium, midmyocardium, and subendocardium of nonischemic (three pieces) and ischemic (five pieces) regions. Transmural pieces were obtained from the center of several transverse sections used to determine the AAR and were at least 1 cm from the perfusion boundaries as indicated by Patent blue dye. All samples were counted in a gamma counter (Tracor Analytic 1195) to determine the activity of each isotope in each sample. The activity of each isotope was also determined in the reference blood flow samples. Myocardial blood flow was calculated by using a preprogrammed computer to obtain the true activity of each isotope after correction for overlap of the two isotopes in individual samples, and tissue blood flow was calculated from the following equation: Qm=Qr×Cm/Cr, where Qm is myocardial blood flow (milliliters per minute per gram of tissue), Qr is the rate of withdrawal of the reference blood flow (9.4 mL/min), Cr is the activity of the reference blood flow sample (counts per minute), and Cm is the activity of the tissue sample (counts per minute per gram). Transmural blood flow was calculated as the weighted average of the three layers in each region.
Dogs were excluded if (1) heartworms were found after the animals were killed, (2) transmural collateral blood flow was >0.16 mL/min per gram, (3) heart rate was >180 bpm at the beginning of the experiment, or (4) more than three consecutive attempts were needed to convert ventricular fibrillation with low-energy DC pulses (15 to 20 J) applied directly to the heart.
Radioligand Binding Studies
PD 81,723 (also called LY 202472) was synthesized at Eli Lilly and Co, R-PIA was obtained from Research Biochemicals, and all cell culture reagents were obtained from GIBCO BRL. Canine A1 and A2A receptor cDNAs were kind gifts of Dr Gilbert Vassart (Brussels, Belgium), and the A3 receptor was previously cloned as described by our laboratory.17 125I-ABA18 and 125I-APE19 were synthesized as described previously. All other reagents were purchased from Sigma.
Expression of Adenosine Receptors in COS-7 Cells
The canine A1, A2A, and A3 adenosine receptor cDNAs were subcloned into the mammalian expression vector CLDN-10B (a gift of Dr Mitch Reff, SKF Laboratories). The A1 adenosine receptor cDNA (RDC7, 1.3 kb) was introduced between the Kpn I–HindIII sites, and the A2A receptor cDNA (RDC7, 2.4 kb) was subcloned between the Kpn I–EcoRV sites within the polycloning region of CLDN. The A3 adenosine receptor cDNA was inserted into an EcoRI site of a modified version of CLDN, with additional restriction sites placed into the polycloning cassette. The vectors were individually transformed into Escherichia coli (JM109, Promega), and large quantities of plasmid DNAs were isolated (Maxi-Preps, Promega). The plasmid DNA was subsequently introduced into COS-7 cells using the DEAE-dextran method.20 Transfected cells were grown in DMEM with 10% fetal calf serum, 100 U/mL penicillin G, and 100 μg/mL streptomycin on 150-cm2 cell culture plates for 60 hours.
Membrane Preparation and Radioligand Binding
Crude membrane preparations were made from transfected COS-7 cells and used for radioligand binding analysis. Cells were initially washed in phosphate-buffered saline (pH 7.4), homogenized in 10 mmol/L EDTA, 10 mmol/L sodium HEPES (pH 7.4), and 0.1 mmol/L benzamidine, and then centrifuged at 20 000g for 20 minutes. Pellets were resuspended and washed twice in HE buffer (10 mmol/L sodium HEPES and 1 mmol/L EDTA, pH 7.4) and were resuspended in the same buffer with 10% (wt/vol) sucrose at a membrane protein concentration of 5 mg/mL. Membranes were frozen in aliquots at −20°C. For radioligand binding analysis, triplicate aliquots of cell membranes were diluted to 1 mg of protein/mL and incubated in 0.1 mL with 2.5 (dissociation assays) or 5 mmol/L (competition assays) MgCl2, 2.5 U/mL adenosine deaminase, and 0.1 to 1 nmol/L 125I-ABA or 125I-APE for 3 hours. 125I-ABA was used as a radioligand for A1 and A3 receptors,17 18 and 125I-APE was used as a radioligand for A2A receptors.19 The bound and free radioligands were separated by rapid filtration through Whatman GF/C filters using a Brandel cell harvester. Nonspecific binding was measured using 5 μmol/L nonradioactive I-ABA or I-APE. To measure dissociation kinetics, radioligand was bound to equilibrium (3 hours) in a water bath (at 21°C or 37°C), and then 20 μmol/L R-PIA was added alone or in combination with 10 μmol/L PD 81,723 for various times before filtration. Data were fit to the following biexponential decay equation:where B1 and B2 are specific binding at equilibrium to two binding sites, B is total binding at time t after the addition of R-PIA, κ−1 and κ−2 are the dissociation rate constants, and NS is nonspecific binding. The values for t1/2 of dissociation were calculated by dividing 0.693 by the rate constant. Three separate experiments were performed in triplicate for each receptor subtype.
Competition by Glibenclamide
The ability of glibenclamide to compete for radioligand binding was determined in equilibrium binding assays. Glibenclamide was added to test tubes with radioligands and incubated for 3 hours before filtration and receptor binding assays, which were performed as described above (n=3, each receptor subtype).
In the radioligand binding studies, differences in dissociation rate constants were compared by Student's t test by paired analysis. Differences between groups in hemodynamics at various time points were compared by using a two-way ANOVA for time and treatment with repeated measures and Fisher's least significant difference test if significant F ratios were obtained. Differences in tissue blood flow, AAR, or infarct size were compared by a one-way ANOVA and Fisher's least significant difference test. ANCOVA was used to determine whether the relationship between transmural collateral blood flow and infarct size differed between the control group and drug-treated or preconditioned groups. All values are expressed as mean±SEM. Differences between groups were considered significant at P<.05.
Mortality and Exclusions
Fifty-three dogs were initially used in the present study. Six dogs were excluded because transmural collateral blood flow was >0.16 mL/min per gram (two dogs in the D+PD+PC group and one each in the control, PD, vehicle PC, and PD+PC groups), and one dog in the control group was excluded because more than three consecutive attempts were needed to convert ventricular fibrillation with DC pulses. Therefore, 46 dogs were used for data analysis: eight dogs for the control and D+PD+PC groups and six each for the remaining groups.
The hemodynamic data are shown in Table 1⇓. There were no significant differences in heart rate, mean arterial blood pressure, rate-pressure product, left ventricular dP/dt, and LAD blood flow at baseline between groups, except for the mean arterial blood pressure in the vehicle PC group. In all groups, there were no significant differences in systemic hemodynamics between groups throughout the experiment, except for the baseline mean arterial blood pressure in the 5′ PC group.
Myocardial Infarct Size
Myocardial infarct size and AAR data are shown in Fig 2⇓. The anatomic AAR expressed as a percentage of the left ventricle was not significantly different between groups (Fig 2A⇓): control group, 29.5±2.4% (mean±SEM); PD group, 32.9±1.1%; vehicle PC group, 32.7±3.3%; PD+PC group, 35.3±3.6%; D+PD+PC group, 31.6±1.7%; G+PD+PC group, 35.6±1.6%; and 5′ PC group, 29.3±0.7%. Myocardial infarct size expressed as a percentage of the AAR (Fig 2B⇓) was significantly reduced in the PD+2.5′ PC group (14.6±1.7%, P<.05) and 5′ PC group (12.5±3.3%, P<.01) compared with the control group (26.3±3.6%). There were no significant differences in the infarct size in the PD group (23.4±4.2%), vehicle PC group (26.5±1.7%), D+PD+2.5′ PC group (23.6±2.3%), and G+PD+2.5′ PC group (25.9±3.3%) compared with the control group.
Fig 3⇓ shows the relationship between transmural collateral blood flow measured at 30 minutes of occlusion and infarct size expressed as a percentage of the AAR. In all groups, there was an excellent inverse correlation (r=.614 to .914) between these two variables. However, the regression line of the “preconditioned” groups (5′ PC and PD+2.5′ PC groups) was shifted downward when compared with that of the control groups by ANCOVA (P<.05). There were no statistically significant differences between any of the other groups and the control group by ANCOVA.
Regional Myocardial Blood Flow
Transmural collateral blood flow data in nonischemic (left circumflex coronary artery) and ischemic (LAD) areas are summarized in Table 2⇓. There were no significant differences in transmural collateral blood flow between groups compared with the control group except for the flows in the nonischemic region in the G+PD+2.5′ PC group at 30 minutes of occlusion and in the 5′ PC group at 3 hours of reperfusion.
Radioligand Binding Studies
To determine whether PD 81,723 enhances binding to canine A1 adenosine receptors and to other canine adenosine receptor subtypes (A2A and A3), the effect of PD 81,723 on the dissociation kinetics of radioligand binding to recombinant adenosine receptors was examined, and the results are shown in Fig 4⇓. Rates were determined at 37°C for the A1 and A2A receptors but at 21°C for A3 receptors, since binding to this receptor was quickly lost at higher temperatures, possibly because of the activation of a protease or kinase. For all three receptors, dissociation occurs in two kinetically distinct components. The first component, probably reflecting low-affinity binding to uncoupled receptors, was too rapid to measure accurately by the filtration method used in these studies. PD 81,723 (10 μmol/L) significantly increased the t1/2 of the slow high-affinity component of 125I-ABA binding by 2.18-fold from 5.80±0.70 to 12.6±0.69 minutes (n=3, P<.05 by paired analysis). PD 81,723 had no effect (n=3, each group) on the dissociation kinetics of 125I-ABA from A3 receptors (without PD 81,723, 18.5±0.71 minutes; with PD 81,723, 16.5±0.74 minutes) or 125I-APE from A2A receptors (without PD 81,723, 9.0±0.70 minutes; with PD 81,723, 7.1±0.70 minutes).
Since glibenclamide completely blocked the protection produced by the combination of PC plus PD 81,723, we reasoned that this sulfonylurea might act by preventing adenosine from binding to A1 adenosine receptors. Therefore, competition studies were also performed to determine whether glibenclamide competes for radioligand binding to recombinant canine adenosine receptors. As shown in Fig 5⇓, glibenclamide had no influence on agonist binding to the three adenosine receptor subtypes at concentrations as high as 10 μmol/L (n=3, each receptor subtype).
The major findings of the present study are as follows: (1) PD 81,723 lowers the threshold for ischemic PC, although it had no effect on infarct size by itself; (2) the infarct-limiting effect of PD 81,723 is blocked by the selective A1 adenosine receptor antagonist, DPCPX, and by the KATP channel antagonist, glibenclamide; (3) 10 μmol/L PD 81,723 significantly decreased the slow component of 125I-ABA dissociation from canine A1 adenosine receptors by 2.18-fold in binding studies, whereas PD 81,723 had no effect on the dissociation kinetics of 125I-ABA or 125I-APE from A3 or A2A adenosine receptors, respectively, which demonstrates that PD 81,723 selectively acts on A1 receptors in dogs; and (4) glibenclamide is not a competitive adenosine receptor antagonist at concentrations as high as 10 μmol/L.
Since a number of previous studies have shown that at least a single 5- or 10-minute period of coronary occlusion followed by 10 to 60 minutes of reperfusion is required to produce the beneficial effect of ischemic PC in the canine heart, we initially tested a single 2.5′ PC or 5′ PC stimulus and found that 2.5 minutes of LAD occlusion was not long enough to reduce infarct size, whereas 5′ PC resulted in a significant reduction in infarct size in this canine model. In agreement, Miura et al4 reported that a single 2-minute period of coronary artery occlusion with a 5-minute period of reperfusion followed by a subsequent longer occlusion was insufficient to reduce infarct size in rabbit hearts. Recently, results from our laboratory3 also showed that a single 3-minute period of coronary artery occlusion before a subsequent 60-minute period of occlusion was not sufficient to produce the beneficial effect of PC in dogs. On the other hand, Ovize et al21 demonstrated that 2.5 minutes of coronary occlusion resulted in a significant reduction in infarct size in anesthetized dogs subjected to 60 minutes of occlusion and 4.5 hours of reperfusion. Thus, although there are small differences in the duration of brief coronary artery occlusion necessary to produce ischemic PC, the results of most studies suggest that 3 to 5 minutes of coronary artery occlusion is the threshold of time necessary to produce ischemic PC in dogs and rabbits. Our data indicate that the presence of PD 81,273 during a brief normally subthreshold period of ischemia reduces the threshold necessary to elicit a PC-like effect in dogs.
A large body of evidence suggests that adenosine can protect the heart against ischemia/reperfusion injury5 22 23 24 25 and that activation of A1 adenosine receptors mimics and blockade of these receptors abolishes the effect of ischemic PC to reduce infarct size.6 16 Liu et al6 showed that in isolated rabbit hearts intracoronary administration of adenosine for 5 minutes followed by 10 minutes of a drug-free period was as effective as ischemic PC to limit infarct size. PD 81,723 has been previously shown to increase agonist binding to A1 adenosine receptors and enhance the functional effects of these agonists in tissues derived from rats,9 10 guinea pigs,11 and humans.12 According to those in vitro studies, the maximum effect of PD 81,723 should be obtained at concentrations of ≤20 μmol/L,9 12 and the action of this agent is highly selective for the adenosine receptor.9 10 PD 81,723 has been shown to have no effect on agonist binding to δ-opiate, M2-muscarinic, or α2-adrenergic receptors.9
The selectivity of PD 81,723 has been shown to extend to the A1 subtype of adenosine receptor. In previous studies, it has been shown that PD 81,723 enhances the binding of agonist radioligands to rat, guinea pig, and human A1 adenosine receptors.9 10 11 12 In the present study, we show that PD 81,723 also enhances binding of an agonist, 125I-ABA, to canine A1 adenosine receptors. Although not unexpected, such a finding was not ensured, inasmuch as the ligand binding characteristics of canine A1 adenosine receptors are quite different from the A1 receptors of other species.26 It also has been reported that PD 81,723 is selective in its enhancement for A1 over A2A receptors on the basis of radioligand binding to rat brain.9 Such adenosine receptor subtype selectivity also has been demonstrated functionally in the guinea pig heart, since PD 81,723 enhances the negative dromotropic effect of adenosine mediated by A1 receptors but does not enhance coronary vasodilation, which is mediated by A2A receptors. In the present study, we have shown that the absence of an effect of PD 81,723 on A2A receptors can be extended to the canine species; this extrapolation is based on the lack of effect of PD on 125I-APE binding to recombinant A2A adenosine receptors and on the failure of PD 81,723 to influence hemodynamic responses in the dog. Effects of PD on the A3 subtype of adenosine receptors have not been previously investigated. It seemed possible that PD 81,723 might influence A3 adenosine receptors, since of the four subtypes of adenosine receptors, only the A1 and A3 subtypes couple primarily to Gi/Go, whereas A2A and A2B receptors couple to Gs. Also, A1 receptors are structurally more similar to A3 receptors than to A2A or A2B receptors.27 For these reasons, our new finding that PD 81,723 does not enhance agonist binding to recombinant canine A3 receptors implies even greater specificity of the enhancer actions of PD for a distinct binding domain unique to A1 receptors than does the absence of enhancer effect on A2A receptors, reported previously.
In the context of the present study, the failure of PD 81,723 to enhance agonist binding to A3 adenosine receptors is informative for another reason. It has been postulated that the effects of adenosine to produce PC might be mediated by A3 adenosine receptors.28 29 Although our data do not rule out the involvement of A3 adenosine receptors in the ischemic PC response, the observation that PD 81,723 enhances PC, coupled with the observation that PD 81,723 does not enhance binding of an agonist to A3 adenosine receptors, implies that ischemic PC in the dog is mediated, at least in part, by A1 adenosine receptors. Also consistent with this conclusion is the finding that the A1-selective antagonist DPCPX blocked protection produced by PC plus PD 81,723.
We found that pretreatment with glibenclamide blocks the myocardial protection afforded by PC plus PD 81,723. Since 2.5 minutes of PC alone did not precondition, but PC plus PD 81,723 was protective, and since PD is selective for A1 adenosine receptors in binding assays, we conclude that the protection observed with PC plus PD 81,723 is mediated by A1 adenosine receptors. Furthermore, since the protection afforded by PC plus PD 81,723 is absent in hearts pretreated with glibenclamide, we conclude that the component of ischemic PC mediated by A1 adenosine receptors is mediated via a mechanism that is prevented by the sulfonylurea. This effect of glibenclamide clearly is not due to blockade of adenosine receptors, since in our binding studies it was found that glibenclamide is not an antagonist of adenosine receptors (Fig 5⇑). Glibenclamide most likely antagonizes the effects of PD 81,723 by blocking KATP channels, which supports our original hypothesis that activation of KATP channels is an important step in the mechanism of ischemic PC.14
Our laboratory first suggested that KATP channels may be involved in ischemic PC on the basis of the finding that glibenclamide, at a low dose that is not proischemic, blocks the protective effect of ischemic PC in dogs.14 This finding has been reproduced in several other laboratories and in several different species.30 31 32 33 Subsequently, we found that glibenclamide also blocks the protective effect of adenosine receptor agonists.16 These findings, in combination with the studies of Kirsch et al13 and Ito et al,34 which demonstrated by patch-clamp methodology that KATP channels are activated after stimulation of adenosine A1 receptors though a Gi protein pathway, suggest that adenosine mediates ischemic PC by activating A1 adenosine receptors, thereby opening KATP channels. According to this hypothesis, however, it is difficult to explain the “memory” associated with PC, since activation of a K+ channel is an acute short-lived event, whereas PC persists for up to 1 to 2 hours.1 There are two potential explanations for this apparent discrepancy. First, recent studies suggest that PKC may also be a cellular mediator in ischemic PC.35 Since PKC may remain active for a prolonged period of time or substrate proteins may not be dephosphorylated immediately, the memory of PC may be explained by prolonged protein phosphorylation. In support of this hypothesis, Liu et al36 have recently shown in rabbit ventricular myocytes that there is an important synergistic interaction between adenosine, PKC, and the KATP channel, which suggests that these three components may be necessary for PC to occur. Furthermore, in pancreatic beta cells, phorbol esters activate the KATP channel, which causes membrane hyperpolarization.37 Thus, in the context of PC, PKC may phosphorylate the KATP channel, or perhaps an accessory protein that influences the KATP channel, such that the channel opens more readily in response to a subsequent ischemic insult.36 On the other hand, recent work of Przyklenk et al38 has suggested that PKC is not involved in ischemic PC in the canine heart. A second possible explanation regarding the KATP channel and the memory of PC comes from the recent studies of Niroomand et al.39 These investigators found that the responsiveness of receptor-mediated Gi protein activation is markedly reduced (38%) within the first 5 minutes of myocardial ischemia in dog hearts. However, in preconditioned hearts, Gi protein responsiveness was actually increased for at least 60 minutes after a single PC stimulus. Thus, communication between adenosine receptors and KATP channels, as well as other effector proteins, such as PKC, may be maintained or increased in preconditioned hearts. Finally, since PD 81,723 and PC have been shown to possess similar effects to increase adenosine receptor coupling to G proteins,11 39 it will be of interest in future studies to determine whether PD 81,723 added during a prolonged occlusion period can produce myocardial protection in the absence of a PC stimulus.
Selected Abbreviations and Acronyms
|2.5′ PC||=||2.5 minutes of LAD occlusion followed by 10 minutes of reperfusion|
|5′ PC||=||5 minutes of LAD occlusion followed by 10 minutes of reperfusion|
|AAR||=||area at risk|
|KATP channel||=||ATP-sensitive K+ channel|
|LAD||=||left anterior descending coronary artery|
|PKC||=||protein kinase C|
This study was supported by National Institutes of Health grants HL-08311 and HL-37942. We would like to thank Anna Hsu and Jeannine Moore for their excellent technical assistance and Carol Knapp for outstanding secretarial assistance.
- Received January 19, 1996.
- Accepted April 10, 1996.
Murry CE, Jennings RB, Reimer KA. Preconditioning with ischemia: a delay of lethal cell injury in ischemic myocardium. Circulation. 1986;74:1124-1136.
Mizumura T, Nithipatikom K, Gross GJ. Bimakalim, an ATP-sensitive potassium channel opener, mimics the effects of ischemic preconditioning to reduce infarct size, adenosine release, and neutrophil function in dogs. Circulation. 1995;92:1236-1245.
Yao Z, Gross GJ. Activation of ATP-sensitive potassium channels lowers threshold for ischemic preconditioning in dogs. Am J Physiol. 1994;267(Heart Circ Physiol 36):H1888-H1894.
Miura T, Ogawa T, Iwamoto T, Shimamoto K, Iimura O. Dipyridamole potentiates the myocardial infarct size-limiting effect of ischemic preconditioning. Circulation. 1992;86:979-985.
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.
Liu Y, Thornton J, Van Winkle DM, Stanley AWH, Olsson RA, Downey JM. Protection against infarction afforded by preconditioning is mediated by A1 adenosine receptors in rabbit heart. Circulation. 1991;84:350-356.
Bruns RF, Fergus JH, Coughenour LL, Courtland GG, Pugsley TA, Dodd JH, Tinney FJ. Structure-activity relationships for enhancement of adenosine A1 receptor binding by 2-amino-3-benzoylthiophenes. Mol Pharmacol.. 1990;38:950-958.
Bruns RF, Fergus JH. Allosteric enhancement of adenosine A1 receptor binding and function by 2-amino-3-benzoylthiophenes. Mol Pharmacol.. 1990;38:939-949.
Mudumbi RV, Montamat SC, Bruns RF, Vestal RE. Cardiac functional responses to adenosine by PD 81,723, an allosteric enhancer of the adenosine A1 receptor. Am J Physiol. 1993;264:H1017-H1022.
Amoah-Apraku B, Xu J, Pelleg A, Bruns RF, Belardinelli L. Selective potentiation by an A1 adenosine receptor enhancer of the negative dromotropic action of adenosine in the guinea pig heart. J Pharmacol Exp Ther. 1993;266:611-617.
Kirsch GE, Codina J, Birnbaumer L, Brown AM. Coupling of ATP sensitive K channels to A1 receptors by G proteins in rat ventricular myocytes. Am J Physiol. 1990;259:H820-H826.
Gross GJ, Auchampach JA. Blockade of ATP-sensitive potassium channels prevents myocardial preconditioning in dogs. Circ Res. 1992;70:223-233.
Niiya K, Uchida S, Tsuji T, Olsson RA. Glibenclamide reduces the coronary vasoactivity of adenosine receptor agonists. J Pharmacol Exp Ther. 1994;271:14-19.
Auchampach JA, Gross GJ. Adenosine A1 receptors, KATP channels, and ischemic preconditioning in dogs. Am J Physiol. 1993;264:H1327-H1336.
Auchampach JA, Caughey GA, Linden J. Molecular cloning and pharmacological characterization of the canine A3 adenosine receptor from BR mastocytoma cells. Mol Pharmacol. In press.
Linden J, Patel A, Sadek S. [125I]Aminobenzyladenosine, a new radioligand with improved specific binding to A1 adenosine receptors in heart. Circ Res. 1985;56:279-284.
Luthin DR, Olsson RA, Thompson RD, Linden J. Characterization of two affinity states of adenosine A2a receptors with a new radioligand, 2-[2-(4-amino-3-[125I]iodophenyl)ethylamino]-adenosine. Mol Pharmacol.. 1995;47:307-313.
Ovize M, Przyklenk K, Hale SL, Kloner RA. Preconditioning does not attenuate myocardial stunning. Circulation. 1992;85:2247-2254.
Zhao ZQ, McGee DS, Nakanishi K, Toombs CF, Johnston WE, Ashar MS, Vinten-Johansen J. Receptor-mediated cardioprotective effects of endogenous adenosine are exerted primarily during reperfusion after coronary occlusion in the rabbit. Circulation. 1993;88:709-719.
Yokota R, Fujiwara H, Miyamae M, Tanaka M, Yamasaki K, Itoh S, Koga K, Yabuuchi Y, Sasayama S. Transient adenosine infusion before ischemia and reperfusion protects against metabolic damage in pig hearts. Am J Physiol. 1995;268:H1149-H1157.
Tucker AL, Robeva AS, Taylor HE, Holeton D, Bockner M, Lynch KR, Linden J. A1 adenosine receptors: two amino acids are responsible for species differences in ligand recognition. J Biol Chem.. 1994;269:27900-27906.
Liu GS, Richards SC, Olsson RA, Mullane K, Walsh RS, Downey JM. Evidence that the adenosine A3 receptor can mediate preconditioning's protection in isolated rabbit hearts. Cardiovasc Res.. 1994;28:1057-1061.
Armstrong S, Ganote CE. Adenosine receptor specificity in preconditioning of isolated rabbit cardiomyocytes: evidence of A3 receptor involvement. Cardiovasc Res.. 1994;28:1049-1056.
Schultz R, Jochen R, Heusch G. Involvement of ATP-dependent potassium channels in ischemic preconditioning in swine. Am J Physiol. 1994;27:H1341-H1352.
Grover GJ, Sleph PG, Dzwonczyk S. Role of myocardial ATP-sensitive potassium channels in mediating preconditioning in the dog heart and their possible interaction with adenosine A1 receptors. Circulation. 1992;86:1310-1316.
Kitakaze M, Komamura K, Chujo M, Mori H, Hori M. Role of ectosolic 5′-nucleotidase activation in the infarct size-limiting effect mediated by the opening K+ channels: implication for ischemic preconditioning. Circulation. 1994;90(pt 2):I-108. Abstract.
Toombs CF, Moore TL, Shebuski RJ. Limitation of infarct size in the rabbit by ischaemic preconditioning is reversible with glibenclamide. Cardiovasc Res. 1993;27:617-622.
Ytrehus K, Liu Y, Downey JM. Preconditioning protects ischemic rabbit heart by protein kinase C activation. Am J Physiol. 1994;266:H1145-H1152.
Liu Y, Gao WD, O'Rourke B, Marban E. Synergistic modulation of ATP-sensitive K+ currents by protein kinase C and adenosine: implications for ischemic preconditioning. Circ Res.. 1996;78:443-454.
De Weille JR, Schmid-Antomarchi H, Fosset M, Lazdunski M. Regulation of ATP-sensitive K+ channels in insulinoma cells: activation by somatostatin and protein kinase C and the role of cAMP. Proc Natl Acad Sci U S A.. 1989;86:2971-2975.
Przyklenk K, Sussman MA, Simkhovich BZ, Kloner RA. Does ischemic preconditioning trigger translocation of protein kinase C in the canine model? Circulation. 1995;92:1546-1557.
Niroomand F, Weinbrenner C, Weis A, Bangert M, Schwencke C, Marquetant R, Beyer T, Strasser RH, Ku¨bler 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.