We tested the hypothesis that selective adenosine A3-receptor stimulation reduces postischemic contractile dysfunction through activation of ATP-sensitive potassium (KATP) channels. Isolated, buffer-perfused rat hearts (n = 8/group) were not drug pretreated (control) or were pretreated with adenosine (20 μM), 2-chloro-N 6-(3-iodobenzyl)-adenosine-5′-N-methyluronamide (Cl-IB-MECA; A3 agonist, 100 nM), Cl-IB-MECA + 8-(3-noradamantyl)-1,3-dipropylxanthine (KW-3902; A1 antagonist, 5 μM), Cl-IB-MECA + glibenclamide (Glib; KATP-channel blocker, 0.3 μM), or Glib alone for 12 min before 30 min of global normothermic ischemia followed by 2 h of reperfusion. After 2 h of reperfusion, left ventricular developed pressure (LVDP, %baseline) in control hearts was depressed to 34 ± 2%. In hearts pretreated with Cl-IB-MECA, there was a statistically significant increase in LVDP (50 ± 6%), which was reversed with coadministration of Glib (37 ± 1%). Control hearts also showed similar decreases in left ventricular peak positive rate of change in pressure (dP/dt). Therefore, the A3 agonist significantly attenuated postischemic cardiodynamic injury compared with the control, which was reversed by Glib. Cumulative creatine kinase (CK in U/min) activity was most pronounced in the control group (10.4 ± 0.6) and was significantly decreased by Cl-IB-MECA (7.5 ± 0.4), which was reversed by coadministration of Glib (9.4 ± 0.2). Coronary flow was increased during adenosine infusion (160% of baseline) but not during Cl-IB-MECA infusion. Effects of Cl-IB-MECA were not reversed by the specific A1 antagonist KW-3902. We conclude that cardioprotection afforded by A3-receptor stimulation may be mediated in part by KATP channels. Cl-IB-MECA may be an effective pretreatment agent that attenuates postischemic cardiodynamic dysfunction and CK release without the vasodilator liability of other adenosine agonists.
- myocardial protection
- purinergic receptors
- adenosine 5′-triphosphate-sensitive potassium channels
protective strategies to reduce myocardial ischemia-reperfusion injury can potentially lead to better recovery of the postischemic heart, decrease postoperative complications, and lower patient mortality and morbidity. Postischemic myocardial injury is mediated during both the ischemic interval and the myocardial reperfusion and is mediated by ionic dyshomeostasis, edema formation, neutrophils, and oxygen-free radical production. Adenosine, a purine nucleoside, has been shown to delay the onset of ischemic contracture (18), reduce the rates of ATP catabolism and intracellular H+ and Ca2+ accumulation during ischemia (17, 18), attenuate myocardial stunning (17, 25, 40), and reduce infarct size (31). Cardioprotective effects of adenosine are exerted when it is administered as a pretreatment either before ischemia (30, 31) or before reperfusion (23, 31). The cardioprotective effects of pretreatment with adenosine administered before ischemia may be mediated by adenosine A1-receptor activation (17, 18,31), whereas protection during reperfusion may be mediated by adenosine A2-receptor activation (23, 31).
Although adenosine is beneficial to myocardial preservation, its untoward effects produced by A1- (bradycardia) and A2a-mediated (vasodilation, hypotension) actions have impeded the routine clinical use of these agents. A recently described adenosine-receptor subtype, designated adenosine A3 (20, 43), could potentially provide myocardial protection without systemic untoward effects. The adenosine A3receptor shares sequence similarity yet has specific pharmacological characteristics not typical of adenosine A1 or A2 receptors.
Activation of the adenosine A3receptor has been proposed to enhance the release of inflammatory mediators from mast cells (7, 10, 35), lower blood pressure (4, 6), depress locomotor activity (13), either induce (16) or reduce apoptosis (38), and protect against cerebral ischemia (36). To date, only a few studies have described the cardioprotective effects of adenosine A3-receptor activation in models of myocardial ischemia-reperfusion (3, 29), as well as its ability to mimic or induce myocardial preconditioning (1, 19, 34, 37). It is plausible that the cardioprotective effects afforded by the activation of the adenosine A3receptor may be mediated via activation of ATP-sensitive potassium (KATP) channels, because the adenosine A3 receptor is similar to the adenosine A1 receptor in receptor transduction through an inhibition in adenylate cyclase (43) and stimulation in protein kinase C (1). The potential involvement of KATP channels by the activation of the adenosine A3 receptors remains undefined.
Therefore, we tested the hypothesis that stimulation of the adenosine A3 receptor with the highly selective A3 agonist 2-chloro-N 6-(3-iodobenzyl)-adenosine-5′-N-methyluronamide (Cl-IB-MECA) would reduce postischemic dysfunction and morphological injury in a neutrophil-free isolated perfused rat heart model via activation of KATPchannels.
MATERIALS AND METHODS
Experimental preparation. The animals in the investigation were handled in full accordance with theGuide for the Care and Use of Laboratory Animals2PO4, 1.2 MgSO4, 118 NaCl, 2.04 CaCl2, 0.05 EDTA tetrasodium salt, 25 NaHCO3, and 11 dextrose. The perfusate was pH balanced to 7.4 after full aeration and was filtered through a 5-μm magna nylon membrane (Fisher Scientific, Pittsburgh, PA). The perfusate temperature was maintained at 37°C with a circulating water-bath heater (Precision Scientific, Chicago, IL) and was continuously oxygenated with 95% O2-5% CO2. The heart was suspended in a water-jacketed chamber to maintain the temperature at 37°C.
All hearts underwent a 25-min stabilization period. During this period, the left atrium was excised and a latex fluid-filled balloon on the end of a PE-100 catheter was inserted into the left ventricle across the mitral annulus. The balloon was inflated to achieve an end-diastolic pressure between 7 and 10 mmHg, after which no further alterations in balloon volume were made. During the stabilization period, two electrodes were placed in the right ventricle, the sinoatrial node was crushed, and the heart was paced at 250 beats/min with a Grass S9 stimulator (Grass Medical Instruments, Kings Park, NY).
To measure coronary perfusion pressure, a solid-state blood pressure transducer (Kent Scientific, Litchfield, CT) was connected to a vertical column of perfusate, mounted in line with and proximal to the aortic cannula. The pressure transducer was positioned at the level of the heart, and coronary perfusion pressure was maintained at >70 mmHg. Coronary flow was continuously measured with an in-line ultrasonic flow probe and T101 ultrasonic blood flow meter (Transonic Systems, Ithaca, NY).
The left ventricular balloon pressure and the aortic perfusion pressure were monitored with fluid-filled pressure transducers (TRN050 blood pressure transducer) and connected to a TRN005 base instrument amplifier (Kent Scientific, Litchfield, CT). The pressure and flow signals were sampled by a personal computer with an analog-to-digital converter (Data Translation, Marlboro, MA). The data were captured, stored, and analyzed with SPECTRUM cardiovascular acquisition and analysis software (Wake Forest University, Winston-Salem, NC).
Experimental design. Animals were randomized to 6 groups (n = 8 in each group) according to pretreatment drug administration:1) KHB (control),2) adenosine (20 μM),3) Cl-IB-MECA (100 nM),4) Cl-IB-MECA + 8-(3-noradamantyl)-1,3-dipropylxanthine (KW-3902; a selective A1 antagonist, 5 μM),5) Cl-IB-MECA + glibenclamide (a KATP-channel blocker, 0.3 μM), and 6) glibenclamide alone (0.3 μM). Hearts received treatment with adenosine or Cl-IB-MECA for 12 min without a washout period before the onset of ischemia via a syringe infusion pump 22 (Harvard Apparatus, South Natick, MA) in a side port located just proximal to the aortic cannulation site. In groups receiving the A1 antagonist KW-3902 and the KATP-channel blocker glibenclamide, the antagonists were administered 2 min before the infusion of Cl-IB-MECA, after which both Cl-IB-MECA and the respective antagonist were infused for the 12 min of drug treatment. This assured that the antagonist was present at the time of A3-receptor agonist infusion. After stabilization and drug administration, all hearts underwent 30 min of normothermic (37°C) global ischemia by the ceasing of all buffer perfusion to the heart. Global myocardial ischemia was initiated immediately after drug administration without a “washout” interval. Pacing was stopped at the onset of ischemia and was resumed 5 min after initiation of reperfusion. Static intraventricular balloon pressure was monitored via the left ventricular balloon pressure transducer throughout the ischemic interval to assess contracture. Reperfusion was initiated by the resumption of buffer perfusion to the heart. Hemodynamic and cardiodynamic measurements were obtained before and after drug administration and after 15, 30, 60, 90, and 120 min of reperfusion. Coronary effluent was collected before and after drug administration and at 15, 30, 60, 90, and 120 min of reperfusion. Samples were stored in a −70°C freezer until used for creatine kinase (CK) analysis.
CK determination. CK (Sigma Diagnostics, St. Louis, MO) activity was calculated spectrophotometrically from the effluent collected at the aforementioned time points. Briefly, effluent samples were mixed in a cuvette with 1 ml of freshly prepared reagent. The solutions were maintained at a constant temperature (37°C) and incubated for 3 min. The δ-absorbances at 0- and 30-s intervals for a period of 120 s were used to calculate CK activity (expressed as U/l). In accordance with the coronary flow at each specific time point of collection, CK activity per minute of flow was determined. The sum of the individual CK values per minute at each time point is represented as the cumulative CK activity.
Statistical analysis. The data were analyzed by a repeated-measures two-way ANOVA for group and time (within group) differences. Tukey’s or Student-Newman-Keuls methods were used to determine pairwise differences between groups. AP value <0.05 was considered statistically significant. Data are expressed as means ± SE.
Materials. Adenosine (Calbiochem, La Jolla, CA) was dissolved with KHB to obtain a final perfusate concentration of 20 μM. The selective A3 agonist Cl-IB-MECA was reconstituted in DMSO to a concentration of 10 mM and diluted to achieve a final heart concentration of 100 nM. The final heart perfusate concentration of DMSO was 0.01 nM. The A1-selective adenosine antagonist KW-3902 (Kyowa Hakko Kogyo) was dissolved in 100 μl ethyl alcohol and 50 μl of 1.0 N NaOH first and then diluted with 0.9% saline to achieve a final heart concentration of 5 μM. This concentration for KW-3902 has previously been shown to be a specific adenosine A1-receptor antagonist (41). The specific KATP-channel antagonist glibenclamide (Sigma, St. Louis, MO) was dissolved in 0.5 ml of ethanol, 0.5 ml of 1 N NaOH, and 0.5 ml of polyethylene glycol 400 to achieve a final heart concentration of 0.3 μM. This concentration has previously been shown to be specific for KATP-channel blockade (9).
Left ventricular developed pressure.The contractile state of the isolated rat heart was assessed by measuring the left ventricular developed pressure (LVDP), as measured by a fixed-volume balloon located within the left ventricle. The LVDP was defined as the peak systolic pressure minus the peak end-diastolic pressure. All baseline LVDPs were comparable among groups before and after drug administration (Fig. 1,A andB). After 2 h of reperfusion, control hearts exhibited a 34 ± 2% recovery in LVDP from baseline values. In contrast, LVDP was significantly greater after pretreatment with adenosine and Cl-IB-MECA (54 ± 3 and 50 ± 6% recovery from baseline, respectively; Fig.1 A). To assess if the preservation of LVDP after pretreatment with Cl-IB-MECA was attributable to activation of adenosine A1receptors, the A1 antagonist KW-3902 was coadministered with Cl-IB-MECA. Rat hearts pretreated with Cl-IB-MECA + KW-3902 had a 55 ± 2% recovery from baseline LVDP values after 2 h of reperfusion (Fig.1 B), which was not statistically different from treatment with Cl-IB-MECA alone. To determine whether recovery of function with Cl-IB-MECA involved activation of KATP channels, the KATP-channel blocker glibenclamide was coadministered with Cl-IB-MECA. The protective effect of Cl-IB-MECA on LVDP after 2 h of reperfusion was completely reversed by coinfusion of glibenclamide (37 ± 1% recovery from baseline, Fig.1 B) to values not different from those for control and glibenclamide alone. The administration of glibenclamide alone did not significantly alter LVDP compared with control (36 ± 4%, Fig. 1 B).
Peak positive rate of change in pressure. The changes in developed pressure over time were accompanied by similar changes in peak positive rate of change in pressure (dP/dt max; Fig.2, A andB). All baseline dP/dt max were comparable among groups before and after drug administration. After 2 h of reperfusion, control hearts recovered to only 34 ± 2% of baseline dP/dt maxvalues. In hearts pretreated with adenosine and Cl-IB-MECA, dP/dt max recovery compared with baseline values was significantly greater (56 ± 3 and 52 ± 6% recovery, respectively) than that in control hearts (Fig.2 A). The cardioprotection associated with administration of Cl-IB-MECA was not reversed with the addition of the A1 antagonist KW-3902 (56 ± 2% recovery after 2 h of reperfusion, Fig.2 A). In contrast, the coadministration of the KATPantagonist glibenclamide with Cl-IB-MECA reduced recovery of dP/dt max at the end of reperfusion to 37 ± 2% of baseline value (Fig.2 B), which was comparable with that for control hearts. The administration of glibenclamide alone did not significantly alter dP/dt max compared with control (36 ± 4%, Fig.2 B). With glibenclamide alone, dP/dt maxafter 2 h of reperfusion was not significantly different from control (34 ± 2%).
Left ventricular end diastolic pressure. The diastolic state of the isolated rat heart was assessed by measuring the left ventricular end diastolic pressure (LVEDP) in a fixed-volume balloon located within the left ventricle (Fig. 3, Aand B). Baseline LVEDPs were comparable among groups before and after drug administration. After 2 h of reperfusion, LVEDP in control hearts increased by 687 ± 30% from baseline values. Increases in LVEDP after 2 h of reperfusion were significantly less in hearts pretreated with adenosine and Cl-IB-MECA (347 ± 49 and 359 ± 47% increase, respectively) compared with control hearts (Fig. 3 A). The lower LVEDP with Cl-IB-MECA was not reversed by the A1 antagonist KW-3902 (384 ± 33% increase from baseline, Fig.3 B). However, the coadministration of glibenclamide reversed the decrease in LVEDP seen with Cl-IB-MECA alone to 585 ± 26% of baseline value (Fig.3 B). The administration of glibenclamide alone resulted in no change in LVEDP (668 ± 32%, Fig. 3 B) compared with control.
Coronary perfusate flow. Coronary perfusate flow as measured by an in-line flow probe is summarized in Table 1. All baseline coronary flow measurements (before drug infusion) were comparable among groups. Administration of 20 μM of adenosine significantly increased flow to 160% of its predrug value. However, pretreatment with Cl-IB-MECA (100 nM) did not produce significant coronary vasodilation (9% increase in flow from baseline values). When a greater concentration of Cl-IB-MECA (1 μM) was administered, there was a significant increase (142%) in coronary flow (17 ± 1.2 ml/min) compared with its predrug baseline value (12 ± 0.4 ml/min). After 2 h of reperfusion, coronary perfusate flow (%baseline) was depressed in control hearts to 28 ± 3% of baseline. At the end of reperfusion, coronary perfusate flow as compared with control was increased after pretreatment with Cl-IB-MECA (41 ± 3%), but only in the adenosine-treated group (47 ± 3%) was the flow significantly greater than the control group at the end of reperfusion.
Cumulative effluent CK. CK activity was calculated from coronary effluent, which was collected at the aforementioned time points (Fig. 4). The sum of the individual CK values per minute at each time point is represented as the cumulative CK. In control hearts, cumulative CK concentration (U/min) averaged 10.4 ± 0.6. Adenosine pretreatment decreased cumulative effluent CK activity by 26% relative to the control group (Fig. 4). Cl-IB-MECA was also associated with a similar decrease in CK activity (28%) relative to control. The coinfusion of KW-3902 did not alter the effects of Cl-IB-MECA alone compared with control hearts. However, when glibenclamide was coadministered with Cl-IB-MECA, CK values were significantly greater than those from Cl-IB-MECA administration alone and were not significantly different from values in control hearts (Fig. 4). Therefore, the coadministration of glibenclamide reversed the effect of Cl-IB-MECA on CK release. There was no significant reduction in cumulative CK values in hearts treated with glibenclamide alone.
In this study, we examined the cardiodynamic effects of a selective adenosine A3 agonist, Cl-IB-MECA, in cell-free perfused rat hearts. Our model is devoid of cellular and other humoral components (e.g., neutrophils, complement, histamine), and our study specifically tests the neutrophil-independent cardioprotection afforded by Cl-IB-MECA. We found that pretreatment with a highly selective A3 agonist attenuated postischemic systolic and diastolic dysfunction and CK release. Pretreatment with Cl-IB-MECA for 12 min before 30 min of normothermic global ischemia without an intervening washout period resulted in a significant improvement in recovery of LVDP and dP/dt max compared with control hearts after 2 h of reperfusion. In addition, LVEDP and effluent cumulative CK activity over the course of ischemia and reperfusion were significantly reduced in Cl-IB-MECA-treated hearts compared with control hearts. We have also shown that the cardioprotection afforded by pretreatment with A3 agonists were blocked by the KATP-channel antagonist glibenclamide, suggesting involvement of KATP-channel activation. Incomplete reversal of reductions in CK release may suggest that mechanisms other than KATP-channel activation may be the primary mechanism involved in protection against reversible injury. Furthermore, these cardioprotective effects of Cl-IB-MECA did not result in coronary vasodilation. Our study suggests that selective stimulation of the adenosine A3 receptor attenuates postischemic dysfunction after myocardial ischemia involving the activation of KATP channels.
The physiological role of A3receptors may be very different from A1 and A2a subtypes (12). Activation of A3 receptors by native adenosine requires relatively high concentrations of the purine; the inhibitory constant (K i) value of adenosine at the rat A3 receptor is ∼1 μM vs. 10–30 nM at rat A1 and A2a receptors, respectively (12). At first, the nonselective adenosine-receptor agonistN 6-2-(4-aminophenyl)ethyladenosine (APNEA) had been used to activate A3 receptors, in combination with a xanthine such as 8-(p-sulfophenyl)theophylline to eliminate non-A3 receptor-mediated effects. Consequently, more specific A3 agonists like IB-MECA have been developed. IB-MECA has an affinity of 1.1 nM and is ∼50-fold selective over adenosine A1 and A2a receptors in vitro (8). The chlorinated derivative, Cl-IB-MECA, displays aK i value of 0.33 nM and is highly selective for A3vs. A1 and A2a receptors by 2,500- and 1400-fold, respectively (12).
In the present study, Cl-IB-MECA selectivity was confirmed in its cardioprotective effect by the coadministration of a selective adenosine A1 antagonist, KW-3902, to rule out spillover into the A1receptor. Furthermore, KW-3902 did not reverse the decreased release of CK after pretreatment with Cl-IB-MECA. Thus the cardioprotective effects of Cl-IB-MECA did not appear to involve A1-mediated effects resulting from a spillover effect mediated by the adenosine A1 receptor. This selectivity data are consistent with reports by Nonaka et al. (22), Nomura et al. (21), and Zhao et al. (41) showing selectivity of KW-3902 for adenosine A1 receptor. Coronary vasodilation was apparent when a higher concentration (1 μM) of Cl-IB-MECA was administered. However, because we were interested in subvasodilator effects of Cl-IB-MECA, we did not show these functional data or determine its involvement of adenosine A2-receptor activation.
The role of activation of the specific adenosine A3 receptor in mediating cardioprotection during ischemia per se (as opposed to effects during reperfusion) has not been clearly defined. Stambaugh et al (29), with chick embyro ventricular myocytes in ovo, have suggested that activation of adenosine A3receptors with IB-MECA and Cl-IB-MECA during hypoxia can attenuate myocyte injury. In agreement, the present study found that stimulation of the adenosine A3 receptor with Cl-IB-MECA in the cell-free isolated rat heart model preserved cardiodynamics and reduced CK enzyme release. Although Cl-IB-MECA has been observed to inhibit selected neutrophil functions relevant to neutrophil-mediated postischemic injury (14), the cardioprotection observed in the present study was independent of any putative antineutrophil effects.
Studies evaluating the mechanisms or affectors leading to A3-mediated cardioprotection remain scarce. However, several studies have investigated the involvement of the A3 receptor with the cardioprotection mimicking preconditioning (1, 3, 19). With the use of a series of cell culture techniques in isolated rabbit myocytes, Armstrong and Ganote (1) suggest that the protection afforded by preconditioning may be secondary to stimulation of A3 receptors by a combination of APNEA (an A1/A3agonist) and 8-cyclopentyl-1,3-dipropylxanthine (a selective A1-antagonist) to unmask the A3 effects. Liu et al. (19), in isolated rabbit hearts, and Auchampach et al. (3), in a rabbit chronic in vivo model, have suggested that stimulation of the adenosine A3 receptor with APNEA and 8-cyclopentyl-1,3-dipropylxanthine or IB-MECA, respectively, may mediate cardioprotection afforded by preconditioning. In the current study, we did not mimic chemical preconditioning, which incorporates a drug washout period. Instead, global myocardial ischemia was initiated immediately at the end of drug administration, without a washout interval so that the drug of interest remained in the myocardium during ischemia. In agreement with the aforementioned studies, the activation of adenosine A3 receptors before myocardial ischemia in our model was cardioprotective.
The activation of KATP channels has been well documented as a physiological effector of adenosine-receptor stimulation. In 1990, Kirsch et al. (15) reported that cardiac adenosine receptors activate the KATP in rat ventricular myocytes via adenosine A1 receptor-coupled activation of a Gi protein. Furthermore, KATP channels have been proposed to be involved in A1receptor-mediated protective effects in ischemic-reperfused myocardium (5, 11, 28). Glibenclamide, a KATP-channel antagonist, abolishes the cardioprotection afforded by stimulation of the adenosine A1 receptor in many animal species (2, 26, 27, 32, 33, 39). Because the adenosine A3 receptor is similar to the adenosine A1 receptor in homology and receptor transduction through an inhibition in adenylate cyclase (43) and cAMP and a stimulation of protein kinase C translocation (1), it is plausible that activation of the adenosine A3 receptor may mediate cardioprotection via activation of KATP channels. In the present study, we have shown that the cardioprotection mediated by the specific activation of the adenosine A3receptor with Cl-IB-MECA was reversed by the addition of the KATP-channel antagonist glibenclamide. Because the current study utilized pacing of the isolated perfused rat heart, conclusions regarding the arrhythmogenicity of stimulation of the KATP channels by A3-receptor activation cannot be made. In our model, the responses to the administration of glibenclamide alone did not significantly differ from those in control hearts. This observation suggests that A3-mediated cardioprotection is, in part, mediated via the activation of KATP channels. From the present study, we cannot delineate which KATP isoform is affected by the activation of the adenosine A3receptor because glibenclamide blocks both the mitochondrial and sarcolemmal isoforms of the KATPchannel.
Study limitations. The limitations in our study are inherent to the crystalloid perfused Langendorff model. Because our model is crystalloid perfused, adequate oxygen delivery to meet the demands of a heart contracting isovolumetrically against a filled balloon was a problem. However, we demonstrated a vasodilator reserve in our preparation with vasodilatory response to adenosine that increased perfusate flow by 160%. Because our perfusate lacked formed elements, the effects of the activation of the adenosine A3 receptor on other cells including neutrophils cannot be assessed. Furthermore, it has previously been demonstrated that KATP-channel activation may effect polymorphonuclear neutrophil function (24, 42). It was the intent of this study to investigate cardioprotective mechanisms independent of neutrophils. Other systemic inflammatory mediators (i.e., complement fragments) that may have direct effects on vascular endothelium or myocytes are absent. In the present in vitro study, we did not quantify the release of histamine from mast cells. Stimulation of the adenosine A3 receptor by some agonists has been shown to produce hypotension as a consequence of the release of histamine from mast cells in rats and mice (7, 10, 35). If this response occurs in humans, therapies targeting the stimulation of A3 receptors may be severely limited by side effects. We conclude that pretreatment with a specific adenosine A3-receptor agonist, via activation of KATP channels, may be a clinically relevant chemical preconditioning stimulus that attenuates postischemic cardiodynamic dysfunction and CK release without the vasodilator liability of nonspecific adenosine agonists.
We are grateful to Gail Nechtman for formatting the manuscript for publication and to Dr. Bakthavachalam for the generous gift of Cl-IB-MECA. KW-3902 was a gift from Kyowa Hakko Kogyo through the efforts of Dr. Akira Karasawa. Furthermore, we thank the Carlyle Fraser Heart Center of Crawford W. Long Hospital-Emory University School of Medicine for their continued support of the research effort.
Address for reprint requests and other correspondence: J. Vinten-Johansen, Cardiothoracic Research Laboratory, Crawford Long Hospital, 550 Peachtree St., NE, Atlanta, GA 30365–2225 (E-mail:).
Funded in part by the Thoracic Surgery Research Foundation Fellowship Award (V. Thourani). Cl-IB-MECA was synthesized by Research Biochemicals (Natick, MA) as part of the Chemical Synthesis Program of the National Institute of Mental Health.
Presented at the Seventieth Scientific Sessions, American Heart Association, November 9–12, 1997, Orlando, Florida.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 1999 the American Physiological Society