Heart and Circulatory Physiology


Presently, the physiological significance of myocardial adenosine A2a receptor stimulation is unclear. In this study, the influence of adenosine A2a receptor activation on A1 receptor-mediated antiadrenergic actions was studied using constant-flow perfused rat hearts and isolated rat ventricular myocytes. In isolated perfused hearts, the selective A2a receptor antagonists 8-(3-chlorostyryl)caffeine (CSC) and 4-(2-[7-amino-2-(2-furyl)[1,2,4]triazolo[2,3-a][1,3,5]triazin-5-ylamino]ethyl)phenol (ZM-241385) potentiated adenosine-mediated decreases in isoproterenol (Iso; 10−8 M)-elicited contractile responses (+dP/dt max) in a dose-dependent manner. The effect of ZM-241385 on adenosine-induced antiadrenergic actions was abolished by the selective A1 receptor antagonist 1,3-dipropyl-8-cyclopentylxanthine (10−7 M), but not the selective A3 receptor antagonist 3-ethyl-5-benzyl-2-methyl-4-phenylethynyl-6-phenyl-1,4-(±)-dihydropyridine-3,5-dicarboxylate (MRS-1191, 10−7 M). The A2a receptor agonist carboxyethylphenethyl-aminoethyl-carboxyamido-adenosine (CGS-21680) at 10−5 M attenuated the antiadrenergic effect of the selective A1 receptor agonist 2-chloro-N 6-cyclopentyladenosine (CCPA), whereas CSC did not influence the antiadrenergic action of this agonist. In isolated ventricular myocytes, CSC potentiated the inhibitory action of adenosine on Iso (2 × 10−7 M)-elicited increases in intracellular Ca2+concentration ([Ca2+]i) transients but did not influence Iso-induced changes in [Ca2+]itransients in the absence of exogenous adenosine. These results indicate that adenosine A2areceptor antagonists enhance A1-receptor-induced antiadrenergic responses and that A2a receptor agonists attenuate (albeit to a modest degree) the antiadrenergic actions of A1 receptor activation. In conclusion, the data in this study support the notion that an important physiological role of A2a receptors in the normal mammalian myocardium is to reduce A1 receptor-mediated antiadrenergic actions.

  • A3 receptor
  • perfused hearts
  • ventricular myocytes
  • isoproterenol
  • calcium

adenosine is a functionally important regulatory metabolite in the heart. Among the diverse cardiovascular actions of this nucleoside, an antiadrenergic effect mediated through an inhibitory A1 receptor subtype is readily discernible in isolated cell (6, 9) and whole heart preparations (4, 23). In contrast to its inhibitory effects, adenosine via a stimulatory A2 receptor subtype appears to increase myocardial contractility. Positive inotropic effects have been shown in ventricular muscle (1, 3, 17) and in isolated myocyte preparations (6, 13, 18, 26, 27).

Despite escalating evidence in favor of an A2 receptor-mediated positive inotropic action of adenosine, there is debate about the physiological significance of this effect. Adenosine A2 receptor agonists have been reported to be without effect on both mechanical performance in cardiac myocytes (24) and in ventricular muscle preparations (2) as well as on cAMP concentrations in cardiac tissue (25). In addition, in those studies where a positive inotropic influence of adenosine or its analogs was demonstrated in isolated myocyte preparations, either the A1 receptor had been uncoupled from its effector (18, 26, 27) or A2 receptor-induced effects were more readily apparent in the presence of an A1 receptor antagonist (6). These latter approaches employed to unmask the actions of A2 receptor-mediated effects cast doubt on an independent physiological role of this receptor in a cardiac preparation with intact A1receptors.

The unmasking of A2receptor-mediated inotropic actions in isolated cell preparations subsequent to an intervention that reduces A1 receptor or postreceptor cellular activity suggests that a significant interaction occurs between A1 and A2 receptor-induced effects. This, in turn, points toward an important modulating role of the A2 receptor on A1 receptor-mediated cardiac actions. Indeed, recent evidence showing an ability of an A2a receptor antagonist to enhance A1 receptor-induced antiadrenergic effects on mechanical performance in isolated ventricular myocytes, possibly through changes in Ca2+uptake (19), supports this hypothesis.

The aim of this study was to investigate whether A2a receptor-mediated effects have an important moderating influence on the antiadrenergic actions of adenosine in isolated perfused rat hearts and rat ventricular myocytes. In addition, this study evaluated whether these A2a receptor-induced effects are in part explained by alterations in intracellular Ca2+ concentration ([Ca2+]i). The results indicate that A2areceptor activation occurs simultaneously with A1 receptor activation and subsequently produces an inhibitory influence on adenosine-induced antiadrenergic effects in both whole heart and cellular preparations.


The animals used in this study were maintained and used in accordance with recommendations in the Guide for the Care and Use of Laboratory Animals [Department of Health and Human Services Publication No. (NIH) 85–25, Revised 1985] and the guidelines of the Institutional Animal Care and Use Committee of the University of Massachusetts Medical School (Worcester, MA).

Isolated Perfused Heart Preparation

Sixteen-week-old male Sprague-Dawley rats (Harlan Sprague-Dawley, Indianapolis, IN), weighing 380–540 g, were anesthetized with an intraperitoneal injection of ketamine (75 mg/kg) and xylazine (15 mg/kg). The hearts were then excised and immediately rinsed in ice-cold physiological saline solution (PSS). Hearts were perfused via the aorta at a constant flow with 37°C PSS containing (in mM) 118 NaCl, 4.7 KCl, 2.5 CaCl2, 25 NaHCO3, 1.2 KH2PO4, 1.2 MgSO4, and 10 glucose, at a pH of 7.40 gassed with 95% O2-5% CO2. The coronary flow rate was determined volumetrically and adjusted to achieve a flow of 10 ml ⋅ min−1 ⋅ g heart wt−1, according to the approximate weight of the heart measured immediately before initiation of heart perfusion. The coronary perfusion pressure (CPP) was monitored from a sidearm of the aortic perfusion cannula with a Statham P23 transducer. Developed left ventricular (LV) pressure was determined using a water-filled, latex balloon-tipped cannula inserted via the left atrium into the LV cavity. The hearts were paced at 300 beats/min with the voltage 10% above threshold, via platinum wire electrodes attached to the left atrium and the apex of the heart. LV pressure and CPP were recorded using a polygraph (model RS 3400, Gould Instrument recorder). The maximum rates of LV pressure development (+dP/dt) and relaxation (−dP/dt) were obtained using a differentiator (model 13–4616–71, Gould Instrument Systems, Valley View, OH) with a high-frequency cutoff set at 300 Hz. Coronary flow per gram of ventricular heart weight was calculated at the end of each experiment.

Ventricular Myocyte Isolation

Isolated adult rat ventricular myocytes were prepared according to methods previously described (21) with several modifications. Sprague-Dawley rats were decapitated, and the hearts were excised and perfused at a constant pressure of 70 cmH2O for 10 min through the aorta with a filtered (0.45-μm membrane filter) perfusing solution (PS; pH 7.4, 37°C). The PS contained (in mM) 118 NaCl, 10 glucose, 25 NaHCO3, 4.69 KCl, 1.18 MgSO4, 1.18 KH2PO4, and 1.0 CaCl2. After equilibration, the hearts were constant-pressure perfused with PS containing no added Ca2+ until spontaneous contractions ceased. The hearts were then perfused for 4–10 min with PS containing 0.48 mg/ml collagenase, 0.187 mg/ml hyaluronidase, 1.67 mg/ml recrystallized BSA, and 48.4 μM Ca2+ at a rate of 3–4 ml/min for each heart. Ventricles were removed from the perfusion system, cut into pieces, and incubated with 5 ml of PS containing 0.72 mg/ml collagenase, 0.28 mg/ml hyaluronidase, 2.5 mg/ml BSA, and 50 μM Ca2+ in a reciprocating water bath (40 cycles/min) with continuous gassing (95% O2-5% CO2) for 7 min at 37°C. This procedure was repeated with fresh incubation solution three to five times. After the final incubation period, the solution was replaced with 10 ml of fresh incubation solution and shaken at 120 cycles/min for 12 min with gassing to dissociate the myocytes. The solution was filtered through a 250-μm nylon mesh into a 50-ml polypropylene tube to which 40 ml of PS containing 5 mg/ml BSA and 100 μM Ca2+ (wash solution) were gradually added.

The myocytes were allowed to settle for 15 min, the upper two-thirds were aspirated, and 30–50 ml of fresh wash solution were added. After the myocytes were allowed to settle for a further 15 min, the wash solution was aspirated, and the myocyte pellet was resuspended in 30–50 ml MEM containing 200 μM Ca2+. After 15 min of settling, the modified MEM solution was aspirated, and the myocyte pellet was resuspended in MEM containing 500 μM Ca2+ (∼10 ml). A 2-ml volume of the myocyte suspension was seeded onto each of five 60-mm culture dishes. These myocytes were incubated (37°C) and gassed (5% CO2 in room air) for 1–3 h before use.

Fura 2 Fluorescence Measurements of [Ca2+]i

Aliquots of ventricular myocytes were transferred to a 0.8-ml superfusion chamber that was mounted on the stage of a Nikon inverted microscope (Diaphot 300), and the cells were viewed using ×40 oil-immersion fluorescence objective lens (Nikon Fluor 40). The temperature was maintained at 37°C using a heated microscope stage plate (Fryer A-50 temperature controller). The myocytes were incubated with 6.25 μM fura 2-AM added to the modified PS. After 20 min, the cells were then superfused with PS-HEPES (PS modified by adding and changing the following: 1 mM glucose, 0.3 mM KCl, 0.25 mM CaCl2, and 10 mM HEPES, pH 7.4) for 15 min. Myocytes that were rod shaped, clearly striated, and mechanically quiescent, but responsive to electrical stimulation with a vigorous and reversible contraction, were selected for Ca2+ transient measurements. Electrical stimulation (model SD9, Grass Instruments, Quincy, MA) at 0.5 Hz was provided by platinum wire electrodes attached to the bottom of the chamber. After a myocyte was selected, all but the myocyte was blocked from the field of view, thereby minimizing background fluorescence. Because the autofluorescence of the myocytes was negligible (∼1% of the individual 340- and 380-nm signals), there was no need to subtract this from the recorded signals. Fluorescent dye internalized within the cell was excited by incident light, provided by a high-speed dual-wavelength scanning illuminator (xenon arc lamp, 75 W) capable of switching between excitation light of 340 and 380 nm at a speed of 650 ratios/s (Delta Scan, Photon Technology International, Brunswick, NJ). Light emitted by the cells was detected at 520 nm by a photomultiplier tube and recorded at 50 points/s using a computer-based data acquisition system (Felix, Fluorescence Analysis Software, Photon Technology International). In vivo calibrations were performed at the end of the experiment by exposure of the myocyte, in the presence of Triton X-100, to CaCl2 followed by EGTA in the absence of Ca2+, to determine maximum ratio (Rmax) and minimum ratio (Rmin). [Ca2+]i was calculated according to the following formula: [Ca2+]i =K dβ[(R − Rmin)/(Rmax − R)] (9, 14), using an estimated dissociation constant (K d) of 200 nM and β, the ratio of permeabilized myocyte 380-nm fluorescence in the presence of EGTA to the 380-nm fluorescence in the presence of the maximum concentration of CaCl2.

Protocol for Isolated Hearts

Agents were infused just proximal to the aortic cannula at rates <3% of the perfusion rate to achieve the desired concentration in the perfusate. For each protocol, the preparation was allowed to stabilize for at least 15 min before eliciting baseline adrenergic responses. Adrenergic responses were induced every 5 min by 20-s infusions of isoproterenol (Iso) to achieve an Iso concentration of 10−8 M in the PSS. The reproducibility of peak +dP/dt values to repeated infusions of Iso over a period of >1 h was established in pilot studies. At least two reproducible baseline Iso-mediated responses were determined before proceeding with one of the following protocols.

Group A.

To examine the influence of simultaneous A1 and A2 receptor activation on the antiadrenergic effects of adenosine, the effect of the selective A2a receptor antagonists, 8-(3-chlorostyryl)caffeine (CSC) (11) and 4-(2-[7-amino-2-(2-furyl)[1,2,4]triazolo[2,3-a][1,3,5]triazin-5-ylamino]ethyl)phenol (ZM-241385) (16, 20), on Iso-elicited contractile responses was assessed in the presence and absence of adenosine. First, experiments were conducted with a series of progressively increasing concentrations of adenosine A2 receptor antagonists in the presence of either 10−6 M adenosine or the adenosine vehicle (0.9% NaCl). Iso challenges were repeated at the end of each 5-min infusion of either CSC (ranging from 10−8 to 10−6 M) or ZM-241385 (ranging from 10−9 to 10−6 M). This protocol was employed to determine the concentration range of CSC or ZM-241385 over which adenosine-induced antiadrenergic effects may be influenced by adenosine A2a receptor activity.

Second, the inhibitory effect of adenosine on adrenergic responses was ascertained in the presence of a continuous infusion of 10−6 M CSC, 10−6 M ZM-241385, or the vehicle of these agents. Iso challenges were repeated at the end of each 5-min infusion of progressively increasing concentrations of adenosine ranging from 10−8to 10−4 M. An adenosine concentration inhibition curve was determined in the presence of a continuous infusion of either the vehicle of CSC or ZM-241385, (10−6 M CSC or 10−6 M ZM-241385). The above time periods of either adenosine or A2a antagonist infusion were established as producing either no effects or minimal effects on CPP. Longer periods of infusion resulted in alterations in CPP that would have obscured the interpretation of the changes in cardiac mechanical performance noted subsequent to A2a receptor antagonist administration. Finally, the effect of incremental concentrations of DMSO, the vehicle of CSC and ZM-241385, on Iso-mediated responses in the absence or presence of 10−6 M adenosine was also assessed.

Group B.

To exclude whether CSC, independent of A2a receptor inhibition, modulates A1 receptor-induced antiadrenergic effects, the A1 receptor selective agonist 2-chloro-N 6-cyclopentyladenosine (CCPA) was infused either with CSC (10−6 M) or with the vehicle for CSC. To determine the CCPA dose-response relationship, Iso challenges were repeated at the end of a 5-min infusion of each incremental concentration of CCPA ranging from 10−9 to 10−5 M.

Group C.

To examine whether A2a receptor inhibition modulates adenosine-induced antiadrenergic effects by potentiating A1 and/or A3 receptor actions, either the selective A1 receptor antagonist 1,3-dipropyl-8-cyclopentylxanthine (DPCPX; 10−7 M) or the selective A3 receptor antagonist 3-ethyl-5-benzyl-2-methyl-4-phenylethynyl-6-phenyl-1,4-(±)-3,5-dicarboxylate (MRS-1191, 10−7 M) (15) was infused with adenosine (10−6M) and ZM-241385 (10−6 M). Iso challenges were repeated before and at the end of each 5-min sequential infusion of adenosine alone, adenosine with ZM-241385, and then either adenosine with both ZM-241385 and DPCPX, or adenosine with both ZM-241385 and MRS-1191. Separate groups of rat hearts were used to examine the influence of either DPCPX or MRS-1191 on the effect of ZM-241385 on the antiadrenergic action of adenosine.

Group D.

To further evaluate whether modulation of adrenergic responses can be attributed, in part, to an interaction that occurs between A2a and A3 receptors, the effect of the selective A3 receptor agonist 4-aminobenzyl-5-N-methylcarboxyamidoadenosine (AB-MECA; 10−6 M) (12) on Iso-elicited contractile responses was assessed in both the presence and the absence of ZM-241385 (10−6 M).

Group E.

To examine whether A1receptor-induced antiadrenergic effects could be modified by the simultaneous activation of A2receptors, the A2-selective agonist carboxyethylphenethyl-aminoethyl-carboxyamido-adenosine (CGS-21680) was infused with or without CCPA. The antiadrenergic actions of CCPA were evaluated using Iso challenges repeated at the end of each 5-min infusion of incremental concentrations of CCPA ranging from 10−9 to 10−5 M. A CCPA concentration inhibition curve was determined in the presence of a continuous infusion of either the vehicle of CGS-21680, 10−6 M CGS-21680, or 10−5 M CGS-21680. The effect of the simultaneous infusion of the vehicle of CCPA and CGS-21680 on adrenergic responsiveness was also examined.

Protocol for Isolated Ventricular Myocytes

To evaluate whether adrenergic-mediated increases in myocyte [Ca2+]itransients are simultaneously modulated by A1 and A2 receptor activation, isolated myocytes were exposed to Iso as well as to Iso and adenosine with or without CSC. After myocyte [Ca2+]itransients had stabilized, [Ca2+]itransients were obtained with myocytes exposed to a continuous infusion of 2 × 10−7 M Iso in the presence of the vehicles of CSC and adenosine. The antiadrenergic effects of adenosine and the ability of CSC to modulate this response were then assessed by infusing either 10−6, 10−5, or 10−4 M adenosine with Iso and, subsequently, adenosine together with CSC (10−6 M) and Iso. Each adenosine concentration required separate myocyte preparations because the action of CSC on adenosine-induced antiadrenergic effects was difficult to reverse during the washout period. To determine whether CSC alters adrenergic responsiveness independent of adenosine, Iso responses were obtained in a separate group of cell preparations in the presence or absence of 10−6M CSC.

Statistical Analysis

The concentration of adenosine or CCPA that produced 50% of the maximal inhibitory response (IC50) in isolated perfused hearts was determined from nonlinear regression analysis using sigmoid curve fitting. If the concentration of a ligand required to produce a maximal response could not be definitely determined, an apparent IC50 was calculated.

Repeated measures ANOVA, followed by either a Dunnett’s or a Student-Newman-Keuls test, were used to evaluate the effects of adenosine receptor agonist (adenosine, CCPA, CGS-21680, or AB-MECA) and antagonist (CSC, ZM-241385, DPCPX, or MRS-1191) ligands on either adrenergic or antiadrenergic responsiveness. An unpaired Student’st-test was used to compare IC50 values between groups. AP value of <0.05 was taken to indicate a statistically significant difference. All data are expressed as means ± SE.


Isoproterenol and adenosine were dissolved in 0.1% sodium metabisulfite or H2O, respectively. CSC, ZM-241385, MRS-1191, AB-MECA, CCPA, and CGS-21680 stock solutions were all dissolved in DMSO and diluted with water, such that the highest concentration of each substance used was associated with a 0.05% DMSO concentration in the perfusate.

Isoproterenol, HEPES, EGTA, BSA, Triton X-100, and DMSO were purchased from Sigma Chemical (St. Louis, MO). Buffer salts were obtained from Fisher Scientific (Medford, MA). Adenosine was supplied by Boehringer Mannheim (Indianapolis, IN), and MEM was acquired from GIBCO Laboratories (Grand Island, NY). Fura 2-AM was purchased from Calbiochem (La Jolla, CA). Crude collagenase and hyaluronidase were obtained from Worthington Biochemical (Freehold, NJ). CSC, MRS-1191, AB-MECA, CCPA, and CGS-21680 were supplied by Research Biochemicals (Natick, MA), and ZM-241385 was supplied by Tocris Cookson (Ballwin, MO).


Isolated Perfused Hearts

A2 receptor blockade potentiates the antiadrenergic effects of adenosine.

The A2a receptor antagonists CSC and ZM-241385 were found to decrease Iso-elicited contractile responses (+dP/dt max) in the presence, but not in the absence, of adenosine. Incremental concentrations of both adenosine A2a receptor antagonists enhanced the antiadrenergic effects of 10−6 M adenosine in a dose-dependent manner (Fig. 1). Adenosine (10−6 M) suppressed the contractile response to Iso (+dP/dt max) by 11 ± 4% in the absence of CSC and by 34 ± 7% in the presence of 10−6 M CSC infusion (P < 0.05). Similarly, 10−6 M adenosine attenuated the contractile response to Iso by 5 ± 2% in the absence of ZM-241385 to 28 ± 2% in the presence of 10−6 M ZM-241385 (P < 0.05). The infusion of either CSC or ZM-241385 together with the vehicle of adenosine did not alter adrenergic responses to Iso despite significant increases in the baseline contractile state (CSC increased +dP/dt max from 1,531 ± 77 to 1,922 ± 243 mmHg/s; and ZM-241385 increased +dP/dt max from 1,500 ± 103 to 2,328 ± 217 mmHg/s). In addition, DMSO, the vehicle of CSC and ZM-241385, did not alter adrenergic responses in either the presence or absence of adenosine (data not shown). CSC at higher concentrations resulted in an increase in CPP in the group receiving adenosine (CPP with 10−6 M adenosine alone = 71 ± 4 mmHg; CPP with adenosine and CSC = 89 ± 17 mmHg;P < 0.05). However, no significant changes in CPP occurred in response to ZM-241385 (data not shown).

Fig. 1.

Effect of incremental concentrations of adenosine A2a receptor antagonists 8-(3-chlorostyryl)caffeine (CSC) (top) or ZM-241385 (bottom) on contractile response (+dP/dt max) to a 20-s infusion of 10−8 M isoproterenol in presence of 10−6 M adenosine (•) or vehicle (○) in isolated rat hearts.Top: vehicle,n = 4; adenosine,n = 8.Bottom: vehicle,n = 3; adenosine,n = 8. * P < 0.05 adenosine vs. vehicle. † P < 0.05 vs. baseline. # P < 0.01 vs. 0 M CSC or ZM-241385 in presence of adenosine.

Both A2a receptor antagonists produced a leftward shift in the concentration-inhibition curve of adenosine (Fig. 2). Reducing the influence of A2 receptors with CSC and ZM-241385 unmasked the antiadrenergic effect of adenosine at a concentration (10−7 M) that under control conditions did not result in a significant attenuation of Iso-elicited responses (Fig. 2). The IC50 value for adenosine-induced antiadrenergic effects was decreased 8-fold after CSC (Fig. 2,top,inset) and 10-fold after ZM-241385 (Fig. 2, bottom,inset) administration. The differences in the IC50 values after either CSC or ZM-241385 compared with those exposed to the vehicle of CSC and ZM-241385 (Fig. 2) could not be accounted for by differences between the groups in coronary flow (in ml ⋅ min−1 ⋅ g heart wt−1; vehicle of CSC = 9.35 ± 0.30, CSC = 9.37 ± 0.41; vehicle of ZM-241385 = 10.35 ± 0.45, ZM-241385 = 11.0 ± 0.28) or CPP.

Fig. 2.

Effect of adenosine A2a receptor antagonists CSC (top) or ZM-241385 (bottom) on +dP/dt max to a 20-s infusion of 10−8 M isoproterenol during incremental concentrations of adenosine.Inset of each panel illustrates effect of CSC or ZM-241385 on IC50 value calculated from +dP/dt max data.Top: vehicle (VEH) + adenosine,n = 9; CSC + adenosine,n = 6.Bottom: vehicle + adenosine,n = 9; ZM-241385 + adenosine,n = 7. * P < 0.05, CSC or ZM-241385 vs. vehicle. † P < 0.05 vs. 0 M adenosine.

Finally, whereas CSC potentiated the antiadrenergic influence of adenosine, CSC did not alter the antiadrenergic effect of the selective A1 receptor agonist CCPA (data not shown). The IC50 values for CCPA remained unchanged after CSC administration (with CSC, 1.88 ± 0.54 × 10−7 M; without CSC, 1.9 ± 0.69 × 10−7 M). Thus A2 receptor blockade did not potentiate the antiadrenergic effect of an agonist that selectively activates A1 and not A2 receptors.

A1 receptor, but not A3 receptor, antagonists abolished the potentiating effect of ZM-241385 on the antiadrenergic actions of adenosine.

The selective A1 receptor antagonist DPCPX at 10−7 M abolished the potentiating effect of the A2a receptor antagonist ZM-241385 on adenosine-induced antiadrenergic actions (Fig.3, top). However, the selective A3 receptor antagonist MRS-1191 at 10−7M failed to attenuate the action of ZM-241385 on adenosine-induced antiadrenergic actions (Fig. 3,bottom). In addition, the selective A3 receptor agonist AB-MECA at 10−6 M was unable to mediate antiadrenergic actions in either the presence or absence of the A2a receptor antagonist ZM-241385 at 10−6 M (data not shown). Infusion of the vehicles for the various agents did not alter Iso-elicited contractile responses.

Fig. 3.

Effect of adenosine (Ado) A1 and A3 receptor antagonists 1,3-dipropyl-8-cyclopentylxanthine (DPCPX) (top, n = 11) and MRS-1191 (bottom, n = 8), respectively, on modulating influence of ZM-241385 on Ado (10−6M)-induced antiadrenergic actions. Iso, isoproterenol. * P < 0.05 vs. Iso alone.

Interaction between A2 and A1 receptors.

The selective A2 receptor agonist CGS-21680 at 10−5 M (Fig.4), but not at 10−6 M (data not shown), reduced the antiadrenergic actions of the higher concentrations of the selective A1 receptor agonist CCPA. However, CGS-21680 (10−5 or 10−6 M) did not change the apparent IC50 (M) for CCPA (CCPA + CGS-21680 at 10−5 M, 0.24 ± 0.094 × 10−7 M; CCPA + vehicle, 1.33 ± 0.5 × 10−7 M). CGS-21680 alone produced no effect on the contractile response to a 20-s infusion of Iso. The vehicle of CGS-21680 and that of CCPA, when infused simultaneously, produced no change in adrenergic responsiveness (data not shown).

Fig. 4.

Effect of adenosine A2a receptor agonist CGS-21680 on +dP/dt max to a 20-s infusion of 10−8 M isoproterenol during incremental concentrations of selective A1receptor agonist 2-chloro-N 6-cyclopentyladenosine (CCPA). Vehicle, n = 13; CGS-21680 at 10−5 M,n = 10. * P < 0.01 CGS-21680 vs. vehicle.

[Ca2+]iTransients in Isolated Ventricular Myocytes

A2-receptor antagonist enhances the antiadrenergic effect of adenosine.

The A2a receptor antagonist CSC potentiated the antiadrenergic influence of adenosine, as determined from Iso-elicited delta changes (Δ) in amplitude or %Δ amplitude of the [Ca2+]itransients (Figs. 5 and6) and the systolic [Ca2+]i(Table 1). Infusion of 2 × 10−7 M Iso resulted in a 90 ± 22% increase in [Ca2+]itransients. When 10−5 M adenosine was administered with Iso, the [Ca2+]i transients decreased significantly by 32 ± 5% in the absence of CSC and by 58 ± 8% in the presence of 10−6 M CSC. CSC alone produced no effect on the Iso-induced increment in the amplitude of the myocyte [Ca2+]itransients (Fig. 5, Table 1). Neither adenosine nor CSC produced significant effects on diastolic [Ca2+]i(data not shown).

Fig. 5.

Effect of Ado A2a receptor antagonist CSC (10−6 M) on either antiadrenergic action of 10−5 M Ado (left) or adrenergic responses in absence of Ado (right) in isolated rat ventricular myocytes, as determined from intracellular Ca2+ concentration ([Ca2+]i) transient measurements. Delta (Δ) amplitude of [Ca2+]i transient is change in response to 2 × 10−7 M Iso. Bottom left, n = 7; bottom right, n = 5. * P < 0.001 vs. Δ[Ca2+]i transient in presence of Iso alone. † P < 0.01 vs. Δ[Ca2+]itransient in presence of Iso plus Ado.

Fig. 6.

Effect of adenosine A2a receptor antagonist CSC (10−6 M) on antiadrenergic action of 10−6 to 10−4 M adenosine in isolated rat ventricular myocytes, as determined from [Ca2+]itransient measurements. Percent delta (Δ) amplitude of [Ca2+]itransient is response to 2 × 10−7 M isoproterenol obtained in presence of either adenosine or adenosine plus CSC expressed as a percent of Δ[Ca2+]i transient obtained with isoproterenol alone. Adenosine at 10−6 M,n = 4; 10−5 M adenosine,n = 7; 10−4 M adenosine,n = 4. * P < 0.001 adenosine + CSC vs. adenosine alone.

View this table:
Table 1.

Effect of the A2a receptor antagonist CSC, with and without Ado, on peak Iso-induced changes in systolic [Ca2+]i in isolated rat ventricular myocytes


Adenosine A2a receptors have been identified to coexist with A1receptors on the myocyte cell membranes (6, 18). However, the functional role of myocardial A2areceptors is as yet not fully understood. The principal finding of this study suggests that an important role of A2a receptor activation is to counteract the antiadrenergic actions of adenosine. This effect appears to be attributable to an interaction between A2a and A1, and not between A2a and A3, receptors. The findings of this study further indicate that a potential mechanism by which A2a receptor activation inhibits the antiadrenergic action of adenosine involves a reduction in β-adrenoceptor-mediated increases in cardiac myocyte [Ca2+]i.

A2a Receptor Blockade Enhances the Antiadrenergic Actions of Adenosine

The inhibitory effects of exogenous adenosine on Iso-elicited contractile responses were markedly potentiated after A2a receptor blockade in intact rat hearts and in isolated rat cardiac myocytes. A xanthine with selective A2a receptor antagonist properties, CSC (11), augmented the antiadrenergic actions of adenosine in the isolated perfused heart, but not that of the selective A1 receptor agonist CCPA (11). CSC further enhanced the adenosine-mediated decreases in Iso-elicited increments in isolated cardiac myocyte [Ca2+]i. In addition, a nonxanthine with selective A2a receptor antagonist properties, ZM-241385 (16, 20), similarly potentiated the antiadrenergic actions of adenosine. This action of ZM-241385 was abolished by DPCPX, an A1 receptor blocker (11). Neither CSC nor ZM-241385 had a significant effect on Iso-elicited responses in the absence of adenosine. This suggests that these selective A2a receptor antagonists enhanced the antiadrenergic actions of adenosine via A2a receptor blockade rather than by a nonspecific antiadrenergic effect. In general, the data on the actions of A2a receptor blockade are in agreement with prior work performed in myocyte shortening experiments with isolated avian cardiac myocytes by Liang and Morley (19). These investigators demonstrated an enhanced antiadrenergic action of an adenosine analog following A2a receptor blockade (19).

An important difference between the present results and the data of Laing and Morley (19) is the degree to which A2a receptor blockade accentuated A1 receptor-mediated antiadrenergic actions. Although the aforementioned investigators did not provide results that allowed for a quantitative interpretation of their data, the representative results provided suggest that CSC produced only a modest effect on A1 receptor-mediated antiadrenergic actions. In this study, A2 receptor blockade was associated with a profound increase in A1 receptor-mediated antiadrenergic effects. Possible explanations for the quantitative differences noted in this study and that of Liang and Morley (19) include, first, that the influence of CSC on adenosine-induced antiadrenergic actions was quantitatively characterized largely in intact hearts, rather than in isolated cardiac myocytes. Second, there are potential differences in the binding characteristics of the adenosine receptor agonist for the A2a receptor used in this study compared with that used by Liang and Morley (19). These investigators used an adenosine agonistR-phenylisopropyladenosine, with a lower potency at the A2a receptor (11) in comparison with adenosine, the agonist used in this study. Third, in the current study, a protocol was employed that induced marked adrenergic responses in perfused hearts (∼225% increase in +dP/dt max) in comparison with those produced by Liang and Morley (19) in isolated myocytes (∼37% increase in myocyte shortening). The former approach may have improved the ability to detect small but significant differences in adrenergic responses at relatively low concentrations of adenosine. Fourth, the A1-A2areceptor interaction in this study was examined over a broad ligand concentration range, and the most profound effect of both CSC and ZM-241385 was noted over an intermediate range of adenosine concentrations. This approach was not adopted by Liang and Morley (19), and it is possible that the investigators may not have used an ideal concentration of the adenosine analog to show the same degree of effect of CSC on A1 receptor-induced antiadrenergic actions as observed in this study. Finally, species differences might explain the quantitative differences in the results.

Taken together, the data on A2areceptor blockade suggest that if cardiac A1 and A2 receptors are simultaneously activated to modulate adrenergic responses, adenosine-mediated A2a receptor activation induces effects that oppose the A1receptor-mediated antiadrenergic actions. Furthermore, the fact that myocardial A2a receptor blockade was able to potentiate the action of intermediate to low concentrations of adenosine on adrenergic effects argues for its physiological importance. Infusion of either one of the A2a receptor antagonists with 10−7 M adenosine, a concentration that under control conditions did not produce a significant attenuation of Iso-elicited responses, resulted in a significant depression of adrenergic responses. The latter finding is consistent with the hypothesis that ventricular A2areceptors mediate functional effects at adenosine concentrations that are considered to be physiological (7, 10, 17). The other myocardial A2 receptor subtype, the A2b receptor, is thought to mediate physiological effects only at high adenosine concentrations (18).

Interactive Cardiac Effects of A2 and A1 Receptor Agonists

The effects of an A2 receptor agonist, CGS-21680, on the antiadrenergic actions of an A1 receptor agonist, CCPA, were evaluated in isolated perfused hearts to further explore the extent of the interaction between A2a and A1 receptors. Although quantitatively a modest effect, CGS-21680 attenuated the antiadrenergic actions of the higher concentrations of CCPA, whereas CGS-21680 alone had no effect on Iso-elicited responses. However, the A2 agonist did not alter the apparent IC50 value for CCPA-mediated antiadrenergic effects. The lack of effect of CGS-21680 on the apparent IC50 value for CCPA-mediated antiadrenergic effects may reflect an inability to use higher concentrations of CCPA. The dose-response relationship for CCPA alone illustrated in Fig. 4 is not a sigmoid relationship. To achieve a sigmoid fit to the latter curve, a higher concentration of CCPA was needed. However, higher concentrations of CCPA would have resulted in CCPA vehicle (DMSO)-mediated adverse effects on the heart. Finally, adenosine may mediate antiadrenergic effects via activation of both A1 and A3 receptors, whereas CCPA is expected to predominantly transduce antiadrenergic effects via the A1 receptor. Thus it may be argued that unmasking the antiadrenergic effects of A1 and A3 receptor activation as a consequence of A2 receptor blockade may result in greater augmentation of the antiadrenergic effects of adenosine than the inhibitory effects of the stimulatory action of the A2 receptor agonist on antiadrenergic properties of the selective A1 receptor agonist. There is reason to believe that this scenario is unlikely, since the potentiating effect of action of ZM-241385 on the antiadrenergic actions of adenosine was completely reversed by DPCPX, a selective A1 receptor blocker (11), but not by MRS-1191, a selective A3 receptor antagonist (15). Furthermore, infusion of AB-MECA, a selective A3 receptor agonist (12), did not result in altered adrenergic responses in either the presence or absence of an A2a receptor antagonist.

CGS-21680 did not augment Iso-induced adrenergic responses. This differs from results obtained in isolated avian cardiac myocytes (19). A possible reason for the discrepancy between these two lines of evidence is that the Iso concentration used in the present study produced a marked increase in contractility in comparison with that produced by the investigators who showed an additive effect of the action of CGS-21680 on adrenergic responses (19). Hence, in the present study, the capacity to augment the already profound adrenergic-mediated inotropic responses may have been limited. In support of a potentiating effect of A2 receptor activation on β-adrenoceptor-mediated contractile responses, CGS-15943, an A2 antagonist, attenuates Iso-elicited contractile responses in perfused hearts (22), presumably by inhibiting the actions of endogenous adenosine. In the present study, a possible endogenous adenosine-mediated effect on A2 receptors was not examined.

Despite some reservations about the expected effects of CGS-21680 on the antiadrenergic effects of CCPA, the effects of this A2 agonist at higher doses of CCPA support the hypothesis that a significant interaction occurs between A1 and A2a receptors in modifying adrenergic responses. Although CGS-21680 binds to A2b receptors, its main action is on A2a receptors (18). Therefore, although a possible influence of CGS-21680 on A2b receptors cannot be excluded, it is likely that the latter pharmacological agent attenuated A1 receptor-induced antiadrenergic effects through A2a receptor activation.

Isolated Myocyte [Ca2+]iTransients

One mechanism by which A2areceptor activation might attenuate A1 receptor-induced antiadrenergic effects was explored. It has been suggested that A2a receptors are coupled to Gs, which mediate contractile responses through cAMP-dependent and -independent pathways (6, 19), both of which may activate L-type Ca2+ channels and subsequently increase [Ca2+]i(19). In support of a hypothesis that alterations in [Ca2+]i, in part, mediate the potentiated antiadrenergic action of an A1 receptor agonist by an A2a receptor antagonist, CSC enhanced adenosine-mediated decreases in Iso-elicited increments in [Ca2+]i transients.

In conclusion, the present results indicate that A2a receptor activation or inactivation modulates the antiadrenergic influence of A1 receptor-mediated effects. Both xanthine and nonxanthine A2areceptor antagonists potentiate the antiadrenergic actions of adenosine through an effect that is blocked by A1, but not A3 receptor selective antagonists, whereas an A2 receptor agonist attenuates (albeit to a modest degree) the antiadrenergic effects of an A1 receptor agonist. In addition, the present results indicate that the augmented antiadrenergic action of adenosine on β-adrenoceptor-mediated contractile responses following A2a receptor blockade is in part mediated through a decrease in [Ca2+]itransients. These results suggest that although the predominant impact of adenosine on adrenergic responses is an A1 receptor-mediated antiadrenergic action, a profound modulating influence of proadrenergic A2a receptor-induced effects also occurs. The importance of the modulating influence of A2a receptor-mediated actions in the ischemic, hypoxic, or failing myocardium (5), where interstitial levels of adenosine are elevated (8), requires further investigation.


This study was supported, in part, by National Institutes of Health Grants HL-22828 and AG-11491.


  • Address for reprint requests: T. E. Meyer, Div. of Cardiology, Dept. of Medicine, Univ. of Massachusetts Medical Center, 55 Lake Ave. North, Worcester, MA 01655-0127.

  • 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.


View Abstract