Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Adenosine, an endogenous modulator of a wide range of biological functions, interacts with at least four cell surface receptor subtypes classified as A₁, A_2A, A_2B, and A₃. These receptor subtypes belong to the superfamily of G protein-coupled receptors and have been cloned in several animal species (Fredholm et al., 1994). Typically, G protein-coupled receptors show sequence homologies ranging from 85% to 95% among different species (Ralevic and Burnstock, 1998). On the contrary, the A₃ subtype exhibits a lower degree of homology that is only 74% between rats and humans or sheep and 85% between sheep and humans (Zhou et al., 1992; Linden, 1994). Moreover no considerable changes in binding affinity of several agonists and antagonists have been found between rat and human A₁ and A_2A receptors (Dionisotti et al., 1997), whereas the rat A₃ receptor differs from the human in the antagonist binding (Salvatore et al., 1993). Furthermore, A₃ adenosine receptors have large interspecies differences in peripheral distribution. The rat transcript has been detected in testis, lung, kidneys, heart, and brain (Hill et al., 1997; Jacobson et al., 1998b). The human A₃ receptor transcript is widespread, with the most abundant expression being found in the lung and liver. This suggests that numerous physiological effects of adenosine may be mediated by the A₃ adenosine receptor (Jacobson et al., 1995). Several studies indicate that adenosine A₃ receptors may play a basic role in different pathologies such as inflammation and neurodegeneration (Kohno et al., 1996), ischemic brain damage (Von Lubitz et al., 1994), asthma (Jacobson et al., 1998b), and cardiac ischemia (Liang and Jacobson, 1998). To obtain more information on the physiological role of A₃ receptors, new selective agonists and antagonists should be synthesized. The human cloned A₃ adenosine receptor was first characterized with N⁶-(4-amino-3-[¹²⁵I]iodobenzyl)adenosine (Salvatore et al., 1993). Subsequently, [¹²⁵I]AB-MECA has been widely used as a high affinity radioligand for A₃ receptors (Olah et al., 1994; Jacobson, 1998a) even if it shows a moderate A₃/A₁ selectivity (Klotz et al., 1998). The lack of potent and selective radiolabeled A₃ receptor antagonists has been the major obstacle in the characterization of structure, function, and regulation of this adenosine receptor subtype. Antagonists are generally considered more acceptable in the receptor classification than are the results obtained using agonists, the latter being complicated by the different receptor affinity states and cell-dependent effector coupling. In the last few years important progress has been made on the development of selective A₃ receptor antagonists, which have an interesting pharmacological profile (Jacobson et al., 1997; Jiang et al., 1997; Kim et al., 1998; Li et al., 1999). Recently, our group has identified a series of substituted pyrazolo triazolo pyrimidines as potent and selective antagonists to human A₃ adenosine receptors (Baraldi et al., 1999).

This paper describes for the first time the pharmacological characterization of the human A₃ adenosine receptors transfected in Chinese hamster ovary (CHO) cells by using the tritium-labeled form of the most representative compound of this series, [³H]5N-(4-methoxyphenylcarbamoyl)amino-8-propyl-2-(2-furyl)pyrazolo[4,3-e]-1,2,4-triazolo[1,5-c]pyrimidine ([³H]MRE 3008F20) (see Fig. 1). To verify the species specificity of MRE 3008F20, its affinity to rat adenosine receptors was also evaluated. Binding parameters obtained in hA₃CHO using [³H]MRE 3008F20 or [¹²⁵I]AB-MECA were compared. MRE 3008F20 displayed high selectivity for human A₃ (K_i = 0.85 ± 0.02 nM) when compared with its affinity to human A₁ (K_i = 1100 ± 100 nM), A_2A (K_i = 140 ± 15 nM), and A_2B (K_i = 2100 ± 300 nM) adenosine receptors. Moreover, the abilities of typical agonists to inhibit cAMP accumulation and the potency of a series of antagonists in blocking the IB-MECA-induced inhibition of adenylyl cyclase have been evaluated. Finally, with the aim of obtaining insights on the forces driving the coupling of the human A₃ adenosine receptor with a selective ligand, a thermodynamic analysis of [³H]MRE 3008F20 binding was performed and the enthalpic (H°) and entropic (S°) contributions to the standard free energy (G°) of the binding equilibrium were determined.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. [¹²⁵I]AB-MECA (specific activity 2000 Ci/mmol) was obtained from Amersham Laboratories, Milan, Italy. [³H]DPCPX (specific activity, 120 Ci/mmol) was obtained from NEN Research Products (Boston, MA). [³H]SCH 58261 (specific activity, 68 Ci/mmol) was obtained from the Schering-Plough Research Institute (Milan, Italy). NECA, R-PIA, S-PIA, CGS 21680, IB-MECA, AB-MECA, CGS 15943, DPCPX, and XAC were obtained from Research Biochemical International (Natick, MA). SCH 58261, MPC-NECA, MPC-MECA, MRE 3010F20, and MRE 3008F20 were synthesized by Prof. P.G. Baraldi (Department of Pharmaceutical Sciences, University of Ferrara, Italy). CHO cells transfected with the rat recombinant A₃ adenosine receptor were obtained from NEN Life Sciences Products (Boston, MA). HEK-293 cells transfected with the human recombinant A_2B adenosine receptor were obtained from Receptor Biology, Inc. (Beltsville, MD). CHO cells transfected with the human recombinant A₁, A_2A, and A₃ adenosine receptor were obtained from sources described earlier (Klotz et al., 1998). All other reagents were of analytical grade and obtained from commercial sources.

Synthesis of [³H]MRE 3008F20. The synthesis of [³H]MRE 3008F20 (specific activity, 67 Ci/mmol) (Fig. 1) was performed at Amersham International (Buckinghamshire, UK) from tritium gas through a method developed by Nycomed Amersham plc. The product was purified by reversed-phase high performance liquid chromatography using a water methanol/methanol triethylamine gradient.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Chemical structure of [3H]MRE 3008F20.

Human Cloned A₁, A_2A, A_2B, and A₃ Adenosine Receptor Binding Assay. The expression of the human A₁, A_2A, and A₃ receptors in CHO cells has been previously described (Klotz et al., 1998). The cells were grown adherently and maintained in Dulbecco's modified Eagle's medium with nutrient mixture F12 without nucleosides, containing 10% fetal calf serum, penicillin (100 U/ml), streptomycin (100 µg/ml), L-glutamine (2 mM), and Geneticin (G418, 0.2 mg/ml) at 37°C in 5% CO₂/95% air. Cells were split two or three times weekly at a ratio of between 1:5 and 1:20.

[³H]MRE 3008F20 Binding Assay. Kinetic studies of 2 nM [³H]MRE 3008F20 were performed by incubating membranes obtained by CHO cells transfected with the human A₃ receptors in a thermostatic bath at 4°C. For the measurement of the association rate, the reaction was terminated at different times (from 5 to 200 min) by rapid filtration under vacuum, followed by washing with 5 ml of ice-cold buffer four times. For the measurement of the dissociation rate, the samples were incubated at 4°C for 120 min, and then 1 µM MRE 3008F20 was added to the mixture. The reaction was terminated at different times from 5 to 150 min.

Cyclic AMP Accumulation Assay. Membrane preparation obtained by CHO cells transfected with the human A₃ receptors was suspended in 0.5 ml of incubation mixture (50 mM Tris-HCl, 10 mM MgCl₂, 1 mM EDTA, pH 7.4) containing 5 µM GTP, 0.5 mM 4-(3-butoxy-4-methoxybenzyl)-2-imidazolidinone (Ro-20-1724) as phosphodiesterase inhibitor, and 2.0 I.U./ml adenosine deaminase and preincubated for 10 min in a shaking bath at 37°C. Then the respective agonist and ATP (1 mM) were added to the mixture, and the incubation was continued for 10 min. Adenylyl cyclase was stimulated with 10 µM forskolin, which typically produced a 6- to 8-fold increase of activity over basal levels. Maximal inhibition of adenylyl cyclase by agonists amounted to about 60% of total stimulation. The potencies of antagonists were determined by antagonism of the inhibition of cAMP production induced by 100 nM IB-MECA. The reaction was terminated by transferring the tubes to a boiling water bath for 2 min. Then the tubes were cooled to room temperature, centrifuged at 2,000g for 10 min at 4°C, and the supernatants were used for the determination of cAMP by a competition protein binding assay carried out essentially according to the method of Varani et al. (1998b).

Thermodynamic Analysis. For a generic binding equilibrium, L + r = LR (where L = ligand and r = receptor), the affinity association constant K_A = 1/K_D is directly related to the standard free energy G° (G° = RT ln K_A), which can be separated in its enthalpic and entropic contributions according to the Gibbs equation: G°= H°  TS°. The standard free energy was calculated as G° = RT ln K_A at 298.15 K, the standard enthalpy, H°, from the van't Hoff plot ln K_A versus (1/T) (the slope of which is H°/R) and the standard entropy as S° = (H° G°)/T with T = 298.15 K and r = 8.314 JK¹mol¹.

Rat A₁, A_2A, and A₃ Adenosine Receptor Binding Assay. Binding assays to rat brain and striatum A₁ and A_2A receptors were performed according to Borea et al. (1994) and Borea et al. (1995), respectively.

Data Analysis. All binding studies (kinetics, saturation, competition) were analyzed with the program LIGAND (Munson and Rodbard, 1980). EC₅₀ and IC₅₀ values in the cAMP assay were calculated with the nonlinear least-squares curve fitting program Prism (GraphPAD, San Diego, CA).

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

The kinetic behavior of [³H]MRE 3008F20 binding was studied at 4°C in CHO cells expressing the cloned human A₃ adenosine receptors. Figure 2A shows that [³H]MRE 3008F20 binding reached equilibrium after approximately 40 min and was stable for at least 5 h. [³H]MRE 3008F20 binding was rapidly reversed by the addition of 1 µM MRE 3008F20 as shown in Fig. 2B. Association and dissociation curves were fitted to a one-component model significantly better than to a two-component model (P < .05). The rate constants were: k_obs = 0.195 ± 0.020 min¹ and k₁ = 0.042 ± 0.002 min¹. From the k₊₁ (k₊₁ = 0.076 ± 0.002 min¹ nM¹) and k₁ values, the apparent equilibrium dissociation constant (K_D) was estimated to be 0.55 nM. Figure 3A shows a saturation curve of [³H]MRE 3008F20 to the adenosine A₃ receptor with a K_D value of 0.80 ± 0.06 nM and a B_max value of 300 ± 33 fmol/mg protein (n = 3). The Scatchard plot was essentially linear, and computer analysis of the data (Munson and Rodbard, 1980) failed to show a significantly better fit to a two-site as compared with a one-site binding model, indicating that only one class of high affinity binding sites is present under our experimental conditions. Figure 3B shows a saturation curve of [¹²⁵I]AB-MECA to hA₃CHO cells with a K_D value of 0.85 ± 0.05 nM and a B_max value of 110 ± 12 fmol/mg protein. Table 1 shows the affinity, expressed as K_i values, and selectivity of MRE 3008F20 to human and rat adenosine receptors. MRE 3008F20 bound with high affinity (K_i = 0.85 ± 0.02 nM) to human A₃ adenosine receptors, with low affinity (K_i = 140 ± 15 nM) to human A_2A adenosine receptors, and with affinity values in the micromolar range to human A₁ (K_i = 1.1 µM) and A_2B (K_i = 2.1 µM) receptors. Moreover, MRE 3008F20 displayed a weak affinity to rat A_2A receptors (K_i = 2.0 µM) and no affinity to A₁ and A₃ receptors. All these data clearly indicate that MRE 3008F20 is endowed with affinity and selectivity for the human A₃ adenosine receptors. Table 2 shows the comparison of affinities, expressed as K_i, K_H, and K_L values, of selected adenosine receptor agonists and antagonists to human A₃-cloned receptors expressed in CHO cells using [¹²⁵I]AB-MECA or [³H]MRE 3008F20, and the percentage of receptors in the high affinity state (R_H) is also shown. The resultant order of potency in [¹²⁵I]AB-MECA displacement assays for adenosine receptor agonists was as follows: IB-MECA > MPC-NECA > AB-MECA > NECA > R-PIA > MPC-MECA > S-PIA > CGS 21680 (Fig. 4A). The order of potency of the receptor antagonists was as follows: MRE 3008F20 > MRE 3010F20 > XAC > CGS 15943 > DPCPX. SCH 58261 showed a K_i value > 10 µM (Fig. 4B). The order of potency in [³H]MRE 3008F20 displacement assays for adenosine receptor agonists was as follows: IB-MECA > MPC-NECA > AB-MECA > NECA > R-PIA > MPC-MECA > S-PIA > CGS 21680 (Fig. 5A). Displacement of [³H]MRE 3008F20 binding was stereoselective, with R-PIA (K_H = 66 nM; K_L = 2800 nM) being approximately 6- to 7-fold more active than its stereoisomer, S-PIA (K_H = 482 nM; K_L = 16000 nM). The order of potency of the receptor antagonists was: MRE 3008F20 > MRE 3010F20 > XAC > CGS 15943 > DPCPX. SCH 58261 showed a K_i value > 10 µM (Fig. 5B). To assess adenosine A₃ receptor-G protein interactions, inhibition experiments were performed also in the presence of 100 µM GTP. Its inclusion did not significantly affect antagonist binding. In contrast, agonists were 30- to 50-fold less active in the presence of 100 µM GTP, although the same order of potency was observed as that seen in the absence of GTP. The resultant Hill coefficients of agonists were significantly different from unity, whereas only one affinity state could be detected with a slope factor near unity in the presence of GTP. The Hill coefficients of antagonists were not significantly different from unity both in the absence or in the presence of GTP. Computer analysis of binding curves revealed that, in the presence of GTP, a one-component binding model adequately described the agonist binding curves (data not shown). The Spearman's rank correlation coefficient between affinity values of [¹²⁵I]AB-MECA and [³H]MRE 3008F20 binding to human A₃ adenosine receptor by selected receptor agonists and antagonists was 0.97 (P < .01). Table 2 also shows the inhibition of cAMP accumulation by agonists (EC₅₀) and the capability of the antagonists (IC₅₀) to block the effect of 100 nM IB-MECA on adenylyl cyclase in CHO cells. Figure 6A shows the log-dose-response curve for typical adenosine receptor agonists. All adenosine analogs were able to inhibit cAMP accumulation displaying an order of potency identical with that observed in binding affinities to the adenosine A₃ receptor. IB-MECA appeared to be the most potent compound (EC₅₀ = 14 nM) followed by MPC-NECA and AB-MECA (EC₅₀ in the range of 120-160 nM); R-PIA was more potent than its stereoisomer S-PIA (EC₅₀ = 550 nM and 2.1 µM, respectively). Figure 6B shows the capability of the antagonists to block the effect of 100 nM IB-MECA cAMP production in CHO cells by adenosine receptor antagonists. The most potent adenosine receptor antagonists were MRE 3008F20 and MRE 3010F20 (IC₅₀ = 4.5 and 5.3 nM, respectively). The linear correlation coefficient between affinity values of [¹²⁵I]AB-MECA and [³H]MRE 3008F20 binding to the human A₃ adenosine receptor by selected receptor agonists and antagonists was 0.99 (P < .01) (Fig. 7A). The comparison of K_i and EC₅₀/IC₅₀ values indicates that a high correlation exists between data obtained from binding and functional assays (r = 0.99; P < .01) (Fig. 7B). K_D and B_max derived from the saturation experiments of [³H]MRE 3008F20 binding to A₃ adenosine receptors performed at the six temperatures selected were found within the following range: K_D = 0.8 to 3.3 nM and B_max = 285 to 313 fmol/mg protein. Although the dissociation constant (K_D) changed with temperature, B_max values obtained from [³H]MRE 3008F20 saturation experiments appeared to be largely independent of it. Figure 8 shows the van't Hoff plot ln K_A versus 1/T of the [³H]MRE 3008F20 binding to the A₃ adenosine receptor and the final equilibrium thermodynamic parameters (expressed as mean values ± S.E. of three independent determinations) were: G° = 48.77 ± 0.12 kJ mol¹; H° = 33.11 ± 3.13 kJ mol¹; S° = 52.25 ± 5.53 J mol¹ K¹. The linearity of the plot was statistically significant (Cp°, equilibrium heat capacity change, approximately 0) and its slope (H°/R) was positive, a property that has been found to be typical for antagonist binding to adenosine receptors (Lohse et al., 1984; Borea et al., 1996).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   A, kinetics of 2 nM [3H]MRE 3008F20 binding to human A3 adenosine receptors with association curves representative of a single experiment. Inset, first-order plots of [3H]MRE 3008F20 binding. Be represents the amount of [3H]MRE 3008F20 bound at equilibrium, and B represents the amount of [3H]MRE 3008F20 bound at each time. Association rate constant was: K+1 = 0.076 ± 0.002 min1 nM1. B, kinetics of [3H]MRE 3008F20 binding to human A3 adenosine receptors with dissociation curves representative of a single experiment. Inset, first-order plots of [3H]MRE 3008F20 binding. Dissociation rate constant was: K1 = 0.04 ± 0.002 min1.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   A, saturation of [3H]MRE 3008F20 binding to human A3 adenosine receptors. The KD value was 0.80 ± 0.06 nM and the Bmax value was 300 ± 33 fmol/mg protein. B, saturation of [125I]AB-MECA binding to human A3 adenosine receptors. The KD value was 0.85 ± 0.05 nM and the Bmax value was 110 ± 12 fmol/mg protein. Experiments were performed as described in Experimental Procedures. Values are the means and vertical lines are the S.E. of the mean of four separate experiments performed in triplicate. In the inset the Scatchard plot of the same data is shown.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   A, competition curves of specific [125I]AB-MECA binding to human A3 adenosine receptors by adenosine agonists. B, competition curves of specific [125I]AB-MECA binding to human A3 adenosine receptors by adenosine antagonists. Curves are representative of a single experiment from a series of four independent experiments. Competition experiments were performed as described in Experimental Procedures.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   A, competition curves of specific [3H]MRE 3008F20 binding to human A3 adenosine receptors by adenosine agonists. B, competition curves of specific [3H]MRE 3008F20 binding to human A3 adenosine receptors by adenosine antagonists. Curves are representative of a single experiment from a series of four independent experiments. Competition experiments were performed as described in Experimental Procedures.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   A, inhibition curves of cAMP accumulation to human A3 adenosine receptors by adenosine agonists. B, inhibition curves representing the capability of the antagonists to block the effect of 100 nM IB-MECA on adenylyl cyclase to human A3 adenosine. Curves are representative of a single experiment from a series of four independent experiments.

View larger version (19K): [in this window] [in a new window]   Fig. 7.   Comparison between affinity values of [3H]MRE 3008F20 and [125I]AB-MECA binding to human A3 adenosine receptors (r = 0.99; P < .01). B, comparison between affinity values of [3H]MRE 3008F20 binding and EC50 or IC50 obtained in cAMP assays to human A3 adenosine receptors (r = 0.99; P < .01).

View larger version (8K): [in this window] [in a new window]   Fig. 8.   Van't Hoff plot, ln KA versus 1/T, showing the effect of temperature on the equilibrium binding association constant, KA = 1/KD of [3H]MRE 3008F20.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

In this study we characterized the human A₃ adenosine receptor on CHO cells utilizing [³H]MRE 3008F20. The results of the paper show that this highly potent and selective antagonist represents a key advance toward the characterization of A₃ adenosine receptors. In saturation experiments [³H]MRE 3008F20 labeled a single class of recognition sites with affinity (K_D) of 0.80 ± 0.06 nM and receptor density (B_max) of 300 ± 33 fmol/mg protein (Fig. 3A). Moreover, saturation experiments in hA₃CHO indicate that [¹²⁵I]AB-MECA interacts with only one recognition site with affinity in the nanomolar range, K_D = 0.85 ± 0.05 nM, and with a binding capacity of 110 ± 12 fmol/mg protein (Fig. 3B). One possible explanation for this difference is that the agonist [¹²⁵I]AB-MECA recognizes only receptors in the high affinity state for agonists, whereas [³H]MRE 3008F20 as an antagonist can bind to both agonist low and high affinity states of the receptor with equal affinity. In inhibition experiments MRE 3008F20 displayed a subnanomolar affinity for A₃ adenosine receptors (K_i = 0.85 ± 0.02 nM) and very high selectivity versus A₁ (K_i = 1100 ± 100 nM), A_2A (K_i = 140 ± 15 nM), and A_2B (K_i = 2100 ± 300 nM) adenosine receptors, proving to be the most potent and selective antagonist for the characterization of the human A₃ receptor subtype (Table 1). Moreover, competition experiments of MRE 3008F20 indicated that the behavior of this antagonist examined in humans is completely different from that observed in rat A₃-cloned receptors (Table 1). The high interspecies rat-human differences in affinity constants is common to other xanthinic and nonxanthinic antagonists (Salvatore et al., 1993; Ji et al., 1994). In competition binding studies, various adenosine receptor agonists and antagonists bound human A₃ adenosine receptors in CHO cells with a rank order of potency and affinity range similar to that observed for [¹²⁵I]AB-MECA. Computer analysis of competition curves with [³H]MRE 3008F20 as a radioligand show that agonists interact with two recognition binding sites, whereas adenosine antagonists interact with only one (Fig. 5; Table 2). At human A₃ adenosine receptors, agonist competition isotherms were biphasic and fitted better to a two-site model (Fig. 5A), allowing for the characterization of the high and low affinity components, presumably reflecting binding to G protein-coupled and uncoupled states of the receptor, respectively. The addition of GTP made the curve become steeper and the slope factor approach unity. The K_i of agonists in the presence of GTP was in general reasonable close to the affinity of the low affinity state in the absence of GTP. On the contrary, competition binding curves with antagonists ligands were monophasic for human A₃ adenosine receptors (Fig. 5B). Another aim of this study was to investigate the thermodynamic behavior of [³H]MRE 3008F20 binding and determine the enthalpic (H°) and entropic (S°) contributions to the standard free energy (G°) of the binding equilibrium. The linearity of the van't Hoff plot for [³H]MRE 3008F20 binding in the CHO cells indicates that Cp° values for the drug interaction are nearly zero, which means that H° and S° values are not significantly affected by temperature variations at least over the temperature range investigated (Borea et al., 1995).

The linearity of van't Hoff plots in a limited range of temperatures (4-30°) appears to be a common feature of practically all membrane receptor ligands so far studied from a thermodynamic point of view (Gilli et al., 1994). Thermodynamic data obtained from the van't Hoff plot indicate that [³H]MRE 3008F20 binding to human A₃ adenosine receptors is enthalpy- and entropy-driven (H° = 33.11 ± 3.13 kJ mol¹, S° = 52.25 ± 5.53 J mol¹ K¹). This binding behavior has previously been found to be typical of adenosine A₁ and A_2A receptor antagonists (Borea et al., 1996). To evaluate the regulation of adenylyl cyclase activity and to test whether the binding parameters correlated with the functional response, we determined the IC₅₀ values obtained for inhibition of cAMP production by receptor agonists and antagonists, respectively. In the cAMP assay, the compounds studied exhibited a rank order of potency similar to that observed in binding experiments. Agonists were generally less potent as inhibitors of cAMP in CHO cells expressing human A₃ receptors than predicted from binding data. This could be related to the fact that, apart from the fact that the conditions of these assays differ, recombinant A₃ receptors could not be well coupled to inhibition of cAMP accumulation in CHO cells. In addition, it has been observed for A₁ adenosine receptors that function correlates with the low affinity rather than the high affinity state (Lohse et al., 1986; Wilken et al., 1990). However, the new adenosine receptor antagonists MRE 3008F20 and MRE 3010F20 are potent in binding (K_i = 0.85 and 0.95 nM, respectively) and in functional assays (IC₅₀ = 4.5 and 5.3 nM, respectively). The high statistically significant Spearman's rank correlation coefficient between receptor affinity values of [³H]MRE 3008F20 binding and receptor affinity of [¹²⁵I]AB-MECA to human A₃ adenosine receptors expressed in CHO cells (Table 2) confirmed that both ligands can label the human A₃ adenosine receptor subtype. Likewise the Spearman's rank correlation coefficient between IC₅₀ values in the cAMP assay and receptor affinity values showed a highly significant positive correlation.

In conclusion, all these data indicate, for the first time, that MRE 3008F20 is one of the most selective high affinity human A₃ adenosine receptor antagonists ever reported. Furthermore, we demonstrated that its tritiated radiolabeled form, [³H]MRE 3008F20, is a ligand with subnanomolar affinity for the human A₃ adenosine receptor, exhibiting an appropriate A₃ pharmacological profile in human A₃ CHO cells. Due to its high selectivity for A₃ receptors, this new radioligand could be used to detect A₃ receptors in a variety of mammalian tissues and can be considered a new pharmacological tool to better elucidate the physiopathological role of A₃ adenosine receptors.