Adenosine receptors in an enriched fraction of plasma membranes from canine ventricular myocardium

Differential effects of the adenosine A(1) receptor allosteric enhancer PD 81,723 on agonist binding to brain and adipocyte membranes

The benzoylthiophene analog, PD 81,723, has been shown to allosterically enhance agonist binding and functional activation of the mammalian adenosine (ADO) A(1) receptor subtype by putatively maintaining the receptor in a high affinity state. The present studies were conducted to evaluate the ability of PD 81,723 to enhance the binding of [3H]cyclohexyladenosine ([3H]CHA) to A(1) receptors of neural (cerebral cortex) and non-neural (adipocyte) origin in three different species; rat, guinea pig and dog. PD 81, 723 (0.3-100 microM) produced a concentration-dependent enhancement of [3H]CHA binding to rat brain A(1) receptors. These effects were also species-dependent with larger enhancements (150-200% of control) observed in guinea pig and dog brain membranes as compared to the rat (120% of control). In contrast, PD 81,723 did not produce any enhancement of [3H]CHA binding to A(1) receptors in adipocyte membranes from any of the species examined. Additional binding studies were conducted using pharmacological manipulations that have previously been shown to enhance the allosteric effects of PD 81,723. In the presence of 1 mM GTP, the allosteric effects of PD 81,723 (15 microM) were increased in rat, guinea pig and dog brain membranes, however, in adipocyte membranes from each species, no significant alteration in agonist binding was observed. Similarly, the A(1) receptor selective antagonist 8-cyclopentyl-1, 3-dipropylxanthine (added to effectively reduce the intrinsic antagonist properties of PD 81,723) was found to enhance the allosteric effects of PD 81,723 (15 microM) in brain, but produce no alteration of agonist binding in adipocyte membranes from each species. Examination of the dissociation kinetics of [3H]CHA binding from rat brain and adipocyte membranes revealed that PD 81,723 (15 microM) differentially slowed agonist dissociation from brain, but not adipocyte, membranes. Taken together, the present data support the hypothesis that in tissues that are sensitive to PD 81,723, this benzyolthiophene functions to maintain the A(1) receptor in a high-affinity state and that the relative proportions of high-affinity A(1) receptors present in specific tissues may contribute, at least in part, to the apparent differential effects of PD 81,723 on agonist binding. The tissue specific modulation of A(1) receptor function by PD 81,723 also illustrates the possibility that the locus of allosteric modulation by PD 81,723 may be manifest via a specific, but indirect and tissue-dependent, interaction with the A(1) receptor.

Species comparison of adenosine and beta-adrenoceptors in mammalian atrial and ventricular myocardium

The antagonist radioligand 1,3-[3H]dipropyl-8-cyclopentylxanthine ([3H]DPCPX) was used to characterize adenosine A1 receptors in membrane preparations from atrial and ventricular myocardium of rat, rabbit, guinea pig and pig. Kd values in crude membranes from guinea pig atria and ventricles (3.3 and 3.0 nM) were higher than those in the other species (ranges, 1.5-1.8 and 1.5-1.9 nM). Bmax values were greater in atria than in ventricles in all four species, and in atria and ventricles of guinea pig (76 and 34 fmol/mg), than in the other species (ranges, 15-17 and undetectable to 12 fmol/mg). In contrast, guinea pig Kd and Bmax values for beta-adrenoceptors, which were labelled with (-)3-[125I]iodocyanopindolol, fell within the range of values for the other three mammalian species. In semipurified membrane preparations from pig, [3H]DPCPX and the agonist radioligand [125I]-N6-4-aminobenzyladenosine appeared to label a similar population of receptors and gave comparable Kd values in atria (0.73 and 0.66 nM) and in ventricles (0.57 and 0.70 nM). In semipurified preparations from pig, the agonist R-(-)-N6-(2-phenylisopropyl)adenosine (R-PIA) displaced [3H]DPCPX in a manner consistent with the presence of both high- and low-affinity adenosine A1 receptors. The data from this study indicate that the density of adenosine A1 receptors in atria is greater than in ventricles, but similar Kd values suggest that the A1 receptor population is the same in the two cardiac tissues. Also, the data demonstrate that the [3H]DPCPX antagonist binding characteristics of guinea pig myocardium differ from those in rat, rabbit and pig.

A1-adenosine receptor inhibition of adenylate cyclase in failing and nonfailing human ventricular myocardium

BACKGROUND

Receptors that couple via the stimulatory G protein, Gs, to adenylate cyclase and to a positive inotropic response have been extensively investigated in falling human heart. In contrast, much less is known about receptors, such as the A1-adenosine receptor, that couple to adenylate cyclase via the inhibitory G protein, Gi, to give a negative inotropic response. Activation of such Gi-coupled receptors might worsen heart failure. Furthermore, alpha Gi is increased in failing human ventricular myocardium, which may enhance inhibitory receptor coupling to adenylate cyclase.

METHODS AND RESULTS

A1-Adenosine receptor inhibition of adenylate cyclase was examined in crude particulate preparations derived from 12 nonfailing and 12 failing human left ventricles. Experimental conditions were designed for maximal inhibitory responses. Dose-response curves were performed with the selective A1-adenosine receptor agonist R-phenylisopropyl-adenosine (R-PIA). No differences in nonfailing versus failing heart were observed for basal adenylate cyclase activity (49.0 +/- 4.1 versus 45.7 +/- 2.6 pmol cyclic AMP/min/mg), maximal R-PIA-mediated inhibition (31.1 +/- 2.6 versus 30.2 +/- 1.6 pmol cyclic AMP/min/mg), ED50 (R-PIA x 10(-7) 1.28 +/- 0.10 versus 1.36 +/- 0.08), or slope (1.06 +/- 0.06 versus 1.03 +/- 0.10), respectively. Furthermore, fluoride, forskolin, and manganese adenylate cyclase activation were not different in failing heart, which is consistent with no change in the catalytic unit of adenylate cyclase. The inhibitory G protein alpha Gi, as quantitated by pertussis toxin-catalyzed ADP-ribosylation, was increased in failing heart (105.7 +/- 5.8 versus 132.7 +/- 3.4 optical density units, p less than 0.003). Basal adenylate cyclase activity was reduced in failing heart (7.8 +/- 0.8 versus 4.5 +/- 0.4 pmol cyclic AMP/min/mg, p less than 0.005) with assay conditions designed to assess G protein effects.

CONCLUSIONS

The A1-adenosine receptor pathway exerts a major inhibitory effect on human myocardial adenylate cyclase activity. Although alpha Gi was increased in failing heart, A1-adenosine receptor inhibition of adenylate cyclase was not altered in preparations of failing versus nonfailing human ventricular myocardium.

Endothelium-mediated coronary dilatation by adenosine does not depend on endothelial adenylate cyclase activation: studies in isolated guinea pig hearts

Adenosine, applied to the coronary system of guinea pigs at up to 10(-6) M, elicits dilatation solely via an endothelium-mediated process. We investigated the role of coronary A2 receptors in this dilation, since the coronary endothelium possesses adenosine A2-receptors with a stimulatory action on the adenylate cyclase. In situ, A2 receptor stimulation can be assessed by prelabeling the coronary endothelial adenine nucleotide pool with 3H-adenosine and subsequently determining the rate of release of radiolabeled cAMP induced by A2 agonists. Thus, perfusion of isolated hearts with 5'-N-ethylcarboxamidoadenosine (NECA) dose-dependently increased coronary flow and the release of 3H-cAMP from the endothelium. In the presence of 50 microM 2',5'-dideoxyadenosine (ddA), a P-site agonist which inhibits the catalytic activity of adenylate cyclase, coronary flow increases induced by both adenosine and NECA were unaffected. In contrast, ddA reduced the release of labeled cAMP in response to NECA by about 60%. In cultured endothelial cells, ddA likewise inhibited cAMP accumulation due to NECA by about 70%. Moreover, ddA antagonized the adenylate cyclase mediated flow response due to the PGI2 analogue, iloprost, as well as the positive chronotropic and inotropic actions of isoproterenol. The dissociation elicited by ddA between the coronary flow response and the release of cAMP strongly indicates that the endothelial A2 receptors which are linked to adenylate cyclase are not causally involved in endothelium-dependent coronary dilatation induced by adenosine.

Evidence for adenylate cyclase-coupled A1-adenosine receptors on ventricular cardiomyocytes from adult rat and dog heart

Isolated metabolically stable cardiomyocytes from adult rats and mongrel dogs were used to characterize the mechanism underlying the antiadrenergic effect of adenosine. In a system not affected by cellular heterogeneity, isoproterenol (3 x 10(-9) M - 10(-5) M) in the presence of adenosine deaminase (5U/ml) dose dependently increased cellular cAMP (5-80 pmol/mg). The effect of isoproterenol (0.1 microM) was inhibited by various adenosine derivatives, the rank order of potency being in the rat: (-)-N6-(R-phenyl-isopropyl)-adenosine (R-PIA) greater than 5'-N-ethylcarboxamidoadenosine (NECA) greater than S-PIA, and in the dog NECA greater than R-PIA greater than S-PIA. The cAMP increase induced by forskolin (1 microM) was attenuated in the rat by R-PIA. 8-phenyltheophylline (3 microM) antagonized the effect of R-PIA on isoproterenol-stimulated cAMP formation. Basal cAMP content was not influenced by R-PIA or NECA. Omission of adenosine deaminase from the incubation medium attenuated the isoproterenol-induced cAMP increase in the rat by about 30%. Our findings provide evidence for the presence of adenylate cyclase-coupled A1-adenosine receptors on cardiomyocytes which may mediate the antiadrenergic effect of adenosine in the heart.