A1 or A3 Adenosine Receptors Induce Late Preconditioning Against Infarction in Conscious Rabbits by Different Mechanisms
Abstract—We investigated whether activation of A1 or A3 adenosine receptors (ARs) induces late preconditioning (PC) against infarction in conscious rabbits using the selective AR agonists 2-chloro-N6-cyclopentyladenosine (CCPA) and N6-3-iodobenzyladenosine-5′-N-methylcarboxamide (IB-MECA). In vitro radioligand binding and cAMP assays demonstrated CCPA to be ≈200- to 400-fold selective for the rabbit A1AR and IB-MECA to be ≈20-fold selective for the rabbit A3AR. We observed that (1) pretreatment of rabbits 24 hours earlier with CCPA (100 μg/kg IV bolus) or IB-MECA (100 or 300 μg/kg) resulted in an ≈35% to 40% reduction in the size of the infarct induced by 30 minutes of coronary artery occlusion and 72 hours of reperfusion compared with vehicle-treated rabbits, whereas pretreatment with the selective A2AAR agonist CGS 21680 (100 μg/kg) had no effect; (2) the delayed cardioprotective effect of CCPA, but not that of IB-MECA, was completely blocked by coadministration of the highly selective A1AR antagonist N-0861; (3) inhibition of nitric oxide synthase (NOS) with Nω-nitro-l-arginine during the 30-minute occlusion abrogated the infarct-sparing action of CCPA but not that of IB-MECA; and (4) inhibition of ATP-sensitive potassium (KATP) channels with sodium 5-hydroxydecanoate during the 30-minute occlusion blocked the cardioprotective effects of both CCPA and IB-MECA. Taken together, these results indicate that activation of either A1ARs or A3ARs (but not A2AARs) elicits delayed protection against infarction in conscious rabbits and that both A1AR- and A3AR-induced cardioprotection involves opening of KATP channels. However, A1AR-induced late PC uses an NOS-dependent pathway whereas A3AR-induced late PC is mediated by an NOS-independent pathway.
- adenosine receptors
- ischemia/reperfusion injury
- myocardial infarction
- ATP-dependent potassium channels
- nitric oxide synthase
Although it is well established that stimulation of adenosine receptors (ARs) can induce early preconditioning (PC), it is controversial whether ARs can elicit the late phase of PC.1 2 Studies by Baxter et al1 first demonstrated in an open-chest, anesthetized rabbit model that administration of the A1AR agonist 2-chloro-N6-cyclopentyladenosine (CCPA) limited infarct size after a 30-minute coronary artery occlusion produced 24 to 72 hours later. On the other hand, we found in conscious rabbits that a 24-hour pretreatment with CCPA did not attenuate myocardial stunning.2 One interpretation of these results is that stimulation of ARs is capable of inducing late PC against infarction, but not late PC against myocardial stunning. This hypothesis is not implausible, because myocardial infarction and myocardial stunning are pathophysiologically distinct.3 4 However, since the studies by Baxter et al1 were performed in open-chest, anesthetized rabbits whereas our studies were performed in conscious rabbits, differences in the model systems that were used is another potential explanation for the disparate results. Given the lack of uniformity of the studies described above, the ability of ARs to induce late PC remains uncertain.
Of the four subtypes of ARs known to exist (A1, A2A, A2B, and A3), it is generally thought that the A1AR subtype is responsible for mediating the cardioprotective effects of adenosine.5 There is increasing evidence, however, that A3ARs may also exert cardioprotection.6 7 8 9 10 Evidence in support of this hypothesis includes the observations that early PC is not blocked by selective A1AR antagonists but is blocked by nonselective antagonists at high concentrations capable of blocking the A3AR, suggesting that the A3AR can also induce early PC.6 9 Furthermore, activation of A3ARs has been suggested to reduce injury in several different models of ischemia/reperfusion injury.7 8 10 Based on these results, it has been hypothesized that A3ARs are also expressed in ventricular cardiomyocytes and that they provide protection via a similar mechanism as A1ARs. The role of A3ARs in late PC, however, is unknown.
The present study was undertaken to address these issues and to further characterize the cardioprotective actions of AR stimulation. The results demonstrate, for the first time, that activation of A1 and A3ARs induces late PC against infarction via different mechanisms.
Materials and Methods
Radioligand Binding Assays and cAMP Assays
Binding assays were performed with membranes prepared from HEK 293 cells expressing recombinant rabbit A1ARs or A3ARs using N6-(4-amino-3-[125I]iodobenzyl)adenosine-5′-N-methylcarboxamide ([125I]AB-MECA) as the radioligand. The rabbit A1AR cDNA used for transfections was as described by Bhattacharya et al,11 and the rabbit A3AR cDNA was cloned from a rabbit brain cDNA library. Additional binding assays with [125I]AB-MECA were performed with membranes prepared from rabbit brain and spleen. Inhibition of isoproterenol-induced cAMP accumulation by adenosine agonists was performed on HEK 293 cells expressing rabbit A1ARs or A3ARs using radioimmunoassay.
Reverse Transcriptase–Polymerase Chain Reaction (RT-PCR) Analysis
mRNA expression of A1ARs and A3ARs was determined by RT-PCR of total RNA obtained from rabbit heart or 100 to 200 ventricular rabbit cardiomyocytes isolated by enzymatic digestion.
Studies in Conscious Rabbits
New Zealand White rabbits (2.2 to 2.9 kg; Myrtle’s Rabbitry, Thompson Station, Tenn) were instrumented with a balloon occluder around a major branch of the left coronary artery for occlusion and reperfusion and bipolar ECG leads on the chest wall. In some studies (groups I through IV; see below), a Doppler thickening crystal was sutured to the epicardial surface of the region at risk to measure wall thickening (WTh). The rabbits were allowed to recover for a minimum of 10 days after surgery. All animal experiments conformed to the guidelines established by the University of Louisville.
The experimental protocol is depicted in Figure 1⇓. All rabbits were subjected to 30 minutes of coronary artery occlusion and 3 days of reperfusion. At the end of the experiments, infarct size and the area-at-risk size were measured by dual staining with phthalo blue dye and triphenyltetrazolium hydrochloride. Rabbits were randomly assigned to 12 treatment groups. Group I (control group) received 1 mL of vehicle as an IV bolus 24 hours before the occlusion. In groups II, III, and IV, CCPA (100 μg/kg), CGS-21680 (100 μg/kg; A2A receptor agonist), or IB-MECA (100 μg/kg or 300 μg/kg) were given as boluses 24 hours before the coronary occlusion. Groups V and VI were treated 24 hours before the occlusion with N-0861 (7.5 mg/kg IV bolus followed immediately by an intravenous infusion of 0.3 mg · kg–1 · min–1, which was maintained for 3 hours; total dose, 55.5 mg/kg); 5 minutes after the bolus injection of N-0861, the rabbits were treated with either CCPA (100 μg/kg IV bolus) or IB-MECA (300 μg/kg IV bolus). Groups VII, VIII, and IX were given Nω-nitro-l-arginine (L-NA) (13 mg/kg as an IV bolus) and groups X, XI, and XII were given 5-hydroxydecanoate (5-HD) (5 mg/kg IV bolus) 5 minutes before the 30-minute occlusion in rabbits pretreated 24 hours earlier with either CCPA (100 μg/kg IV bolus; groups VII and X), IB-MECA (300 μg/kg IV bolus; groups VIII and XI), or vehicle (groups IX and XII).
An expanded Materials and Methods section can be found in an online data supplement available at http://www.circresaha.org.
Pharmacological Characterization of Rabbit A1ARs and A3ARs
Because the pharmacology of A3ARs differs markedly among species,12 preliminary studies were performed to characterize rabbit A1ARs and A3ARs using radioligand binding analysis and functional assays.
Specific binding of [125I]AB-MECA to HEK 293 cell membranes expressing either rabbit A1ARs or A3ARs fit best to a one-site binding model. The Kd and Bmax values for the rabbit A1AR were 6.18±1.92 nmol/L and 2675±569 fmol/mg membrane protein (n=3), respectively, and for the rabbit A3AR were 0.39±0.11 nmol/L and 1033±433 fmol/mg membrane protein (n=3), respectively. No specific binding of [125I]AB-MECA was observed to nontransfected cells (data not shown).
The Ki values of competing ligands are shown in the Table⇓. The salient findings of these studies can be summarized as follows: (1) CCPA is a potent and selective agonist for recombinant rabbit A1ARs (377-fold); (2) IB-MECA is a potent A3AR agonist; however, it is only 21-fold selective versus the A1AR; (3) like rodent A3ARs,13 the rabbit A3AR is resistant to blockade by xanthine antagonists and no selective antagonists were identified; (4) the xanthine and adenine antagonists had high affinity for the A1AR, therefore useful A1 selective antagonists were identified including CPX (7750-fold) and N-0861 (1400-fold); and (5) the A2AAR agonist CGS 21680 and the A2AAR antagonist ZM 241385 had low affinity for both A1ARs and A3ARs.
We next measured [125I]AB-MECA binding to membranes prepared from rabbit lung and spleen to determine whether we could detect endogenous expression of A1ARs and A3ARs and to confirm that endogenous rabbit ARs expressed in tissues are pharmacologically similar to recombinant rabbit receptors expressed heterologously in HEK 293 cells. Previous studies have demonstrated that high levels of A3AR transcript exist in spleen,12 and that the brain is a rich source of A1ARs. These studies could not be performed in heart, because the level of expression of all of the adenosine receptor subtypes is too low to be detected accurately in heart tissue with agonist radioligands. For these studies, binding to A1ARs was defined by specific binding displaced by 500 nmol/L CPX, A3AR binding was defined by the difference in specific binding displaced by 500 nmol/L CPX and 500 nmol/L MRS 1220, and A2AAR binding was defined by specific binding displaced by 100 nmol/L ZM 241385. As shown in Figure 2⇓, incubation of rabbit brain or spleen membranes with ≈0.3 nmol/L [125I]AB-MECA resulted in 90±1% and 74±1% specific binding, respectively. In brain tissue, ≈75% of the specific binding sites were A1, ≈12% were A3, and ≈13% were A2A. In contrast, ≈25% of the specific binding sites in spleen were A1, ≈70% were A3, and <5% were A2A. These data demonstrate that (1) [125I]AB-MECA labels multiple AR subtypes in rabbit brain and spleen, and (2) the majority of [125I]AB-MECA binding in rabbit brain is to A1ARs whereas the majority of binding in rabbit spleen is to A3ARs.
Using rabbit brain membranes as a source of A1ARs and rabbit spleen membranes as a source of A3ARs, we next performed competition binding assays to compare the affinity of CCPA and IB-MECA for endogenously expressed rabbit ARs. [125I]AB-MECA was included at a concentration of 6 nmol/L for assays of brain membranes and 0.3 nmol/L for assays of spleen membranes, ie, concentrations equivalent to its predicted Kd value for A1ARs and A3ARs, respectively. In addition, 100 nmol/L ZM 241385 was included in all of the assays to inhibit binding to A2AARs, and 500 nmol/L CPX was included in assays of spleen tissue to selectively block A1ARs. Both CCPA and IB-MECA concentration dependently competed for [125I]AB-MECA binding in rabbit brain and spleen tissues (Figure 2⇑). In brain tissue, the CCPA and IB-MECA competition binding data fit best to a two-site binding model; the high-capacity binding site reflected binding to the A1AR. In spleen tissue when 500 nmol/L CPX and 100 nmol/L ZM 241825 were added to the assays, the data fit best to a single-site binding model reflecting binding to the A3AR. The IC50 values of CCPA were calculated to be 0.32±0.06 nmol/L for the high-capacity binding site in brain (A1AR) and 50.1±18.1 nmol/L for the binding site in spleen (A3AR). The IC50 values for IB-MECA were calculated to be 0.78±0.26 nmol/L for the single binding site in spleen (A3AR) and 10.1±2.2 nmol/L for the high-capacity binding site in brain (A1AR). These data demonstrate that CCPA binds selectively to endogenously expressed rabbit A1ARs (≈160-fold selective) and that IB-MECA binds potently and with moderate selectivity (≈13-fold selective) to endogenously expressed rabbit A3ARs. These results are similar to those obtained with CCPA and IB-MECA in binding studies using recombinant rabbit ARs expressed in HEK 293 cells (see Table⇑).
cAMP Assays With HEK 293 Cells
CCPA and IB-MECA were found to inhibit isoproterenol-stimulated cAMP accumulation in HEK 293 cells transfected with recombinant rabbit A1ARs and A3ARs (Figure 3⇓). The EC50 values of CCPA and IB-MECA were calculated to be 0.05±0.03 and 59.0±12.2 nmol/L, respectively, for HEK 293 cells expressing rabbit A1ARs and 10.3±4.11 and 0.14±0.08 nmol/L, respectively, for HEK 293 cells expressing rabbit A3ARs. CGS 21680 weakly acted on rabbit A1ARs (EC50=821±98.2 nmol/L) and A3ARs (EC50=116±24.6 nmol/L). These results demonstrate that CCPA and IB-MECA are functional agonists of rabbit A1ARs and A3ARs. None of the agonists influenced isoproterenol-induced increases in cAMP levels in nontransfected HEK 293 cells (data not shown).
The cAMP-lowering effects of CCPA in HEK 293 cells transfected with the A1AR were surmountably blocked by the addition of 10 or 50 μmol/L N-0861 (Figure 3⇑; Kd=45±19 nmol/L by Schild regression analysis; Figure 3⇑). N-0861 did not antagonize the inhibitory effects of IB-MECA in HEK 293 cells transfected with the A3AR. Thus, N-0861 acts as a competitive and selective antagonist of A1ARs.
Detection of A1AR and A3AR Message by RT-PCR
No bands corresponding to the A1AR or A3AR were detected in heart tissue or isolated ventricular cardiomyocytes after RT-PCR, electrophoretic separation of the reactions through agarose gels, and staining with ethidium bromide. However, as shown in Figure 4⇓, distinct bands corresponding to the A1AR (310 bp) and A3AR (415 bp) were detected in both heart tissue and isolated cardiomyocytes after Southern blotting of the RT-PCR reactions and probing with A1AR- or A3AR-specific cDNA radioprobes. These results suggest that both A1AR and A3AR mRNAs are expressed in whole heart tissue as well as ventricular cardiomyocytes. Note that no bands were detected in negative control samples in which water was included in the RT-PCR reactions instead of RNA, excluding the possibility that the reactions were contaminated with foreign DNA.
Delayed Cardioprotection by CCPA and IB-MECA in Conscious Rabbits
Of the 129 rabbits instrumented for the studies of myocardial infarction, 26 were excluded because of ventricular fibrillation, technical problems, or a small risk region (<10% of left ventricular weight; see Table 1 in the online data supplement, available at http://www.circresaha.org).
Hemodynamic Variables on Day 1
Hemodynamic data for groups II, III, IVa, IVb, V, and VI on day 1 are presented in Figure 5⇓. At baseline, heart rate and mean arterial blood pressure were similar in all of the treatment groups (which ranged from 241 to 255 bpm and 74 to 82 mm Hg, respectively). In group II, administration of CCPA (100 μg/kg) produced a transient reduction in heart rate and mean arterial blood pressure (maximal decreases of 24% and 14%, respectively). In group III, administration of 100 μg/kg CGS 21680 increased heart rate by 29% and reduced mean arterial blood pressure by 17%. In group IVb, administration of IB-MECA at 300 μg/kg caused an 11% decrease in heart rate but did not cause any appreciable changes in mean arterial blood pressure. However, IB-MECA at a dose of 100 μg/kg (group IVb) did not influence either hemodynamic parameter. In group V and VI in which 100 μg/kg CCPA and 300 μg/kg IB-MECA were administered in animals treated concurrently with the A1AR antagonist N-0861, there were no appreciable changes in heart rate or blood pressure at any time, indicating that N-0861 completely blocked the hemodynamic actions of CCPA and IB-MECA.
Heart Rate on Day 2
There were no significant differences in heart rate among any of the groups during the 30-minute coronary occlusion or during the 72-hour reperfusion period, except for an ≈20% decrease in heart rate in all of the groups of rabbits treated with L-NA (see Table 2 in the online data supplement).
Region at Risk and Infarct Size
There were no significant differences among the groups with respect to the weight of the region at risk (which ranged from 15.8% to 20.1% of the left ventricle; see Table 3 in the online data supplement), indicating that this important determinant of infarct size is similar among the treatment groups. However, the average infarct size was 33% smaller in group II (CCPA 100 μg/kg group), 44% smaller in the group IVa (IB-MECA 100 μg/kg group), and 40% smaller in group IVb (IB-MECA 300 μg/kg group) compared with the control group (38.3±4.1%, 33.7±4.9%, and 31.8±2.9% versus 56.4±4.0% of the risk region, respectively), indicating that both CCPA and IB-MECA (the latter even at a hemodynamically inert dose of 100 μg/kg) elicited protection against infarction 24 hours later (Figure 6⇓). The average infarct size in group III (CGS 21680 100 μg/kg group; 61.4±5.5%) did not differ from the control group, indicating that activation of A2AARs does not produce late PC against infarction. In group V (CCPA 100 μg/kg+N-0861 group), infarct size (59.8±3.8%) was indistinguishable from that measured in the control group and significantly larger than that measured in the CCPA group (Figure 6⇓), indicating that N-0681 completely blocked the infarct-sparing effect of CCPA pretreatment. In group VI (IB-MECA 300 μg/kg+N-0861), infarct size (41.8±1.9%) was significantly smaller than that of the control group and was not significantly larger than that observed in the IB-MECA–pretreated group (Figure 6⇓). Thus, the same dose of N-0861 that blocked the infarct-sparing effect of CCPA failed to block the infarct-sparing effect of IB-MECA, implying that activation of A3ARs alone is sufficient to produce late PC against infarction.
Having found that both CCPA and IB-MECA induced late PC against infarction, we examined the effect of a nitric oxide synthase (NOS) inhibitor and an ATP-sensitive potassium (KATP) channel blocker on the cardioprotective responses. In group VII (CCPA 100 μg/kg+L-NA), infarct size (57.7±4.1%) was significantly greater than in group II (CCPA 100 μg/kg) and essentially indistinguishable from controls (Figure 7⇓), indicating that L-NA abrogated the protective effect of CCPA. However, in group VIII (IB-MECA 300 μg/kg+L-NA) infarct size (36.6±3.4%) was significantly smaller than the control group and similar to that in group IVb (IB-MECA 300 μg/kg), indicating that L-NA failed to block the protective effect of IB-MECA pretreatment. In groups X (CCPA 100 μg/kg+5-HD) and XI (IB-MECA 300 μg/kg+5-HD), infarct size (54.8±5.8% and 55.0±4.4%) was greater than in groups II (CCPA 100 μg/kg) and IVb (IB-MECA 300 μg/kg), respectively, and essentially indistinguishable from the control group (Figure 7⇓). Thus, administration of 5-HD on day 2 completely blocked the protection provided by pretreatment with either CCPA or IB-MECA. In groups IX (L-NA) and XII (5-HD), infarct size did not differ significantly from that in the control group (Figure 7⇓), indicating that administration of L-NA or 5-HD did not affect infarct size in vehicle-treated myocardium (55.0±4.4% and 52.0±5.5%, respectively).
The recovery of regional contractile function in the ischemic-reperfused region was measured in groups I through IV. In keeping with the infarct size data, the recovery of systolic WTh was improved in rabbits pretreated with 100 μg/kg CCPA, 100 μg/kg IB-MECA, or 300 μg/kg IB-MECA after 72 hours of reperfusion compared with that measured in the control group (3.5±3.9%, 5.5±9.4%, and 7.3±8.3% of baseline, respectively, versus −16.2±6.4% of baseline; P<0.05). In the CGS 21680-treated group, the recovery of systolic WTh at 72 hours was not significantly improved (−15.4±4.2% of baseline).
The results presented herein demonstrate that, in conscious rabbits, activation of either A1ARs or A3ARs (but not A2AARs) can induce late PC against infarction acting via distinct mechanisms. To the best of our knowledge, this is the first investigation to demonstrate that ARs can induce late PC in a conscious rabbit model. This is also the first investigation to demonstrate that the A3AR subtype, in addition to the A1AR subtype, is capable of triggering the development of late PC, that this effect of the A3AR is mediated by KATP channels, and that NOS plays a differential role in A1AR- versus A3AR-induced late PC.
The ability of AR agonists to induce late PC is controversial. Studies in barbital-anesthetized, open-chest rabbits support the concept that the A1AR agonist CCPA can induce late PC against myocardial infarction,1 whereas studies in conscious rabbits suggest that CCPA cannot induce late PC against myocardial stunning.2 One of the goals of the present investigation was to gain insights into this apparent discrepancy. Specifically, we sought to distinguish between two basic possibilities: (1) ARs can induce late PC against infarction but not late PC against myocardial stunning, or (2) the ability of AR agonists to induce late PC is dependent on the system in which they are studied. Our results provide evidence that the former hypothesis is correct. Using the same conscious rabbit model in which we previously observed that CCPA did not induce late PC against myocardial stunning,2 we found that the same dose of CCPA induced late PC against myocardial infarction (Figure 6⇑). To the best of our knowledge, this is the first identification of a late PC stimulus that provides selective protection against a specific type of ischemic injury. Our results imply that there are important differences between the mechanism of late PC against myocardial stunning and late PC against infarction.
A second major goal of the present investigation was to determine whether the A3AR is capable of inducing late PC against infarction. Our approach was to compare the effects of CCPA to those of the recently characterized A3AR agonist IB-MECA.14 We found that IB-MECA, at a dose of 300 μg/kg, produced reductions in infarct size that were equivalent in magnitude to those elicited by CCPA (Figure 6⇑), suggesting that the A3AR is also capable of inducing late PC against infarction. These results, however, must be interpreted with caution, because IB-MECA at a dose of 300 μg/kg produced a modest decrease in heart rate (Figure 5⇑). This finding, coupled with the results of our in vitro radioligand binding studies showing that IB-MECA is only 13- to 21-fold more potent at binding to rabbit A3ARs compared with rabbit A1ARs (Table⇑ and Figure 2⇑), raises the possibility that IB-MECA may have induced late PC not by interacting with A3ARs, but rather through nonspecific interactions with the A1AR. This possibility cannot be addressed using an A3AR-selective inhibitor because no antagonist is currently available that selectively inhibits rabbit A3ARs (Table⇑). Therefore, we performed two additional sets of experiments. First, we administered IB-MECA or CCPA in the presence of the A1AR antagonist N-0861 given at a dose predicted to produce blood levels (20 to 50 μmol/L)15 that do not block the A3AR. The finding that N-0861 blocked the actions of CCPA against infarction but not those of IB-MECA (Figure 6⇑) demonstrates that IB-MECA is capable of eliciting delayed cardioprotection through a mechanism independent of the A1AR, thereby implicating the involvement of the A3AR. Second, we administered IB-MECA at a lower dose (100 μg/kg), which had no effect on heart rate (Figure 5⇑), indicating that it did not activate A1ARs. The finding that this lower dose of IB-MECA exerted a robust delayed cardioprotective action (Figure 6⇑) further corroborates the conclusion that activation of A3ARs in itself induces late PC, independent of A1AR activation. The possibility that IB-MECA produced late PC by acting through A2AARs can be excluded because the selective A2AAR agonist CGS 21680 did not reduce infarct size (Figure 6⇑). We can also exclude the possibility that IB-MECA acted through A2BARs, because it has extremely low affinity for this receptor subtype.16 Thus, our results indicate that, similar to early PC, the late phase of PC can also be induced by activation of A3ARs.
Although current data indicate that A3ARs can induce both phases of PC, in recent in vivo studies we have observed that mice with genetic disruption of A3ARs exhibit infarcts that are smaller than those in wild-type mice,17 raising the interesting possibility that A3ARs may actually play an injurious role during acute myocardial ischemia in this species. These observations are not in conflict with the idea that A3ARs can trigger PC, since the role of A3ARs in modulating injury during acute myocardial ischemia is distinct from their role in eliciting PC before ischemia. It is also important to keep in mind that the findings obtained with A3AR knockout mice need to be confirmed with the use of A3AR-selective antagonists, because chronic disruption of these receptors may produce compensatory changes in other genes and/or signaling pathways resulting in protection from ischemic injury. In addition, there are marked differences in the properties and tissue expression of A3ARs among species,12 such that observations in mice should not be extrapolated to rabbits.
We also investigated the mechanisms by which CCPA and IB-MECA induce the delayed cardioprotection against infarction. We have previously found that both the anti-stunning and anti-infarct effects of ischemia-induced late PC are mediated by increased activity of NOS, specifically, the inducible isoform of NOS (iNOS).18 Additional studies have demonstrated that the cardioprotective effects of CCPA-induced late PC can be blocked by inhibitors of the KATP channel and are absent in iNOS knockout mice.19 20 Based on these observations, we hypothesized that late PC induced by both CCPA and IB-MECA is also the result of increased NOS activity and increased function of KATP channels. We found that the nonselective NOS inhibitor L-NA and the KATP channel antagonist 5-HD completely abrogated the protective effects of CCPA when they were administered immediately before the 30-minute coronary occlusion on day 2 (Figure 7⇑). On the other hand, IB-MECA–induced late PC was not blocked by L-NA but was completely abrogated by 5-HD (Figure 7⇑). It appears, therefore, that CCPA-induced late PC against infarction involves a mechanism similar to that of ischemia-induced late PC18 ; that is, cardioprotection is due to enhanced production of NO via induction of NOS and enhanced function of KATP channels. IB-MECA–induced late PC also appears to utilize a mechanism requiring KATP channels; however, upregulation of NOS is not a necessary component. Thus, the results suggest that the mechanisms by which A1ARs and A3ARs induce late PC against infarction involve different pathways that ultimately converge on the KATP channel.
Elucidation of the mechanisms by which ARs modulate PC requires knowledge of the cell types within the heart that express A1ARs and A3ARs. It is well established that A1ARs coupled to inhibition of adenylyl cyclase via Gi/o proteins are expressed in cardiomyocytes in both atria and ventricles. These receptors classically are known to counteract the positive inotropic actions of catecholamines.21 Recent data indicate that cardiac A1ARs induce delayed PC via signaling pathways involving protein kinase C, tyrosine kinases, and mitogen-activated/stress-activated protein kinases.22 With regard to the A3AR, however, little is known regarding the specific cell types within the heart that express this receptor subtype. Strickler et al10 recently demonstrated that A3ARs are expressed in cultured embryonic chicken cardiomyocytes and that activation of these receptors produces protection against cell death induced by simulated ischemia and reperfusion. Thus, it is possible that cardiomyocytes also express A3ARs, which may induce PC via similar signaling pathways as the A1AR. This hypothesis is supported by the results of the present investigation in which we were able to detect A3AR mRNA in adult rabbit cardiomyocytes by RT-PCR analysis (Figure 4⇑). The cell types within the heart that express A3ARs may not be limited to cardiomyocytes, however. For example, in other tissues, the A3AR is known to be expressed in resident leukocytes such as macrophages and mast cells.12 A3ARs have also been suggested to be expressed in vascular smooth muscle cells and endothelial cells.23 Based on these observations, it remains possible that A3AR agonists may elicit PC through the release of mediators from nonmyocytic cells. Clearly, additional studies of A3ARs in the heart are warranted.
In conclusion, the present study demonstrates that activation of ARs induces a long-lasting cardioprotective effect in conscious rabbits, suggesting that AR agonists could be used to maintain patients in a protracted preconditioned state. Our results further demonstrate that selective agonists of either A1ARs or A3ARs can induce a late PC effect. Because A3AR agonists can provide protection with minimal hemodynamic effects, these results imply that targeting the A3AR could be a novel and useful approach to the protection of the ischemic myocardium. Finally, the present study elucidates the molecular mechanisms of A1AR- and A3AR-induced late PC by identifying common and differential roles for the KATP channel and NOS, respectively, in theses two forms of delayed cardioprotection.
This study was supported in part by National Institutes of Health Grants R01 HL60051 (J.A.A.), HL-43151 (R.B.), and HL-55757 (R.B.); by American Heart Association (AHA) National Center Grant 9630083N (J.A.A.); by AHA Kentucky Affiliate Grant KY 9804558 (H.T.), and by the Jewish Hospital Research Foundation, Louisville, Ky.
Original received March 3, 2000; resubmission received October 12, 2000; revised resubmission received January 17, 2001; accepted January 17, 2001.
This manuscript was sent to James T. Willerson, Consulting Editor, for review by expert referees, editorial decision, and final disposition.
- © 2001 American Heart Association, Inc.
Baxter GF, Marber MS, Patel VC, Yellon DM. Adenosine receptor involvement in a delayed phase of myocardial protection 24 hours after ischemic preconditioning. Circulation. 1994;90:2993–3000.
Maldonado C, Qiu Y, Tang XL, Cohen MV, Auchampach JA, Bolli R. Role of adenosine receptors in late preconditioning against myocardial stunning in conscious rabbits. Am J Physiol. 1997;273:H1324–H1332.
Bolli R, Marban E. Molecular and cellular mechanisms of myocardial stunning. Physiol Rev. 1999;79:609–634.
Armstrong S, Ganote CE. Adenosine receptor specificity in preconditioning of isolated rabbit cardiomyocytes: evidence of A3 receptor involvement. Cardiovasc Res. 1994;28:1049–1056.
Tracey WR, Magee W, Masamune H, Oleynek JJ, Hill RJ. Selective activation of adenosine A3 receptors with N6-(3-chlorobenzyl)-5′-N-methylcarboxamidoadenosine (CB-MECA) provides cardioprotection via KATP channel activation. Cardiovasc Res. 1998;40:138–145.
Auchampach JA, Rizvi A, Qiu Y, Tang XL, Maldonado C, Teschner S, Bolli R. Selective activation of A3 adenosine receptors with N6-(3-iodobenzyl)adenosine-5′-N-methyluronamide protects against myocardial stunning and infarction without hemodynamic changes in conscious rabbits. Circ Res. 1997;80:800–809.
Liu GS, Richards SC, Olsson RA, Mullane K, Walsh RS, Downey JM. Evidence that the adenosine A3 receptor may mediate the protection afforded by preconditioning in the isolated rabbit heart. Cardiovasc Res. 1994;28:1057–1061.
Feoktistov I, Biaggioni I. Adenosine A2B receptors. Pharmacol Rev. 1997;49:381–402.
Guo Y, Bolli R, Bao W, Wu WJ, Black RG Jr, Murphree SS, Salvatore CA, Jacobson MA, Auchampach JA. Targeted deletion of the A3 adenosine receptor confers resistance to myocardial ischemic injury and does not prevent early preconditioning. J Mol Cell Cardiol. In press.
Bernardo NL, Okubo S, Maaieh MM, Wood MA, Kukreja RC. Delayed preconditioning with adenosine is mediated by opening of ATP-sensitive K+ channels in rabbit heart. Am J Physiol. 1999;277:H128–H135.
Belardinelli L, Shryock JC, Song Y, Wang D, Srinivas M. Ionic basis of the electrophysiological actions of adenosine on cardiomyocytes. FASEB J. 1995;9:359–365.
Dana A, Skarli M, Papakrivopoulou J, Yellon DM. Adenosine A1 receptor induced delayed preconditioning in rabbits: induction of p38 mitogen-activated protein kinase activation and Hsp27 phosphorylation via a tyrosine kinase– and protein kinase C–dependent mechanism. Circ Res. 2000;86:989–997.