Adenine Nucleotide-Induced Activation of Adenosine A_2B Receptors Expressed in Xenopus laevisOocytes: Involvement of a Rapid and Localized Adenosine Formation by Ectonucleotidases

Materials and Methods

Preparation of cRNA.

cRNAs for the human cystic fibrosis transmembrane conductance regulator (CFTR) and the β₂ adrenoceptor were prepared as described previously (Uezono et al., 1993, 1997). cDNA encoding the rat Ade A2B receptor was obtained by reverse transcriptase-polymerase chain reaction amplification of the rat brain cDNA (Ohkubo et al., 2001). PCR primers were designed to amplify the complete coding sequence of the rat Ade A_2B receptor (GenBank accession no.M91466). The nucleotide sequences and the locations in the cDNA are 5′-CACCTTAGCGGCTGTCCTGA-3′ (sense, 39–58) and 5′-GGGCCACATGCTTGAGAGGGTA-3′ (antisense, 1211–1190). PCR was carried out in 50 μl of a solution containing 25 U/ml Klen Taq DNA polymerase (CLONTECH, Palo Alto, CA). The conditions were 94°C for 1 min, followed by 30 cycles of 30 s at 94°C, 30 s at 58°C, and 2 min at 72°C, with final extension at 72°C for 5 min. A single PCR product having the expected length of 1173 base pairs was directly subcloned into pGEM-T vector (Promega, Madison, WI). The full-length rat Ade A_2B receptor cDNA was then isolated after digesting with NotI and SpeI and ligated into pcDNA 3.1+ (Invitrogen, Carlsbad, CA), which had been digested withNotI and XbaI. The resultant DNA construct was sequenced using an automated DNA sequencer (ABI Prism; Applied Biosystems, Tokyo, Japan). The Ade A_2Breceptor cDNA was linearized with StuI; cRNA was then transcribed using a T7 RNA polymerase kit (mMESSAGE mMACHINE T7 Kit; Ambion Inc., Austin, TX). After DNase treatment, cRNAs were purified by phenol/chloroform extraction, and precipitation was produced with ethanol. The cRNAs were then dissolved in RNase-free water.

Preparation and Injection of Oocytes.

X. laevisovaries were obtained from COPACETIC (Aomori, Japan). Ovaries were dissected into small pieces containing approximately 20 to 50 cells. They were washed with modified Barth's solution (MBS), consisting of 88 mM NaCl, 1 mM KCl, 2.4 mM NaHCO₃, 0.91 mM CaCl₂, 0.33 mM CaNO₃, 0.82 mM MgSO₄, 2.5 mM sodium pyruvate, 10 mM HEPES, pH 7.5, 100 μg/ml streptomycin, and 100 U/ml penicillin. They were then washed twice in Ca²⁺- and Mg²⁺-free MBS and incubated for 60 min with 1.5 mg/ml dispase II (Roche Molecular Biochemicals, Tokyo, Japan) changing the solution every 20 min. Dissociated oocytes were washed 4 to 5 times with Ca²⁺- and Mg²⁺-free MBS and then twice with normal MBS. Defolliculated healthy oocytes in stages V and VI were collected under a microscope and maintained at 18°C overnight. Healthy oocytes were selected again and injected with cRNAs (1 ng CFTR, 0.5 ng β₂ adrenoceptor, and/or 1 ng rat A_2B receptor) or sterile water (as a control) in a final volume of 50 nl using an automatic oocyte injector (Nanoject; Drummond Scientific Co., Broomall, PA). Oocytes were incubated at 18°C in MBS. The medium was changed every day, and the oocytes were used 3 to 10 days after injection.

Recordings.

Electrophysiological recordings were performed at room temperature with the two-electrode voltage-clamp method using a TEV-200 Voltage Clamp System (Dagan, Minneapolis, MN). An oocyte was placed in a 100-μl chamber containing ND96 solution composed of 96 mM NaCl, 2 mM KCl, 1.8 mM CaCl₂, 1 mM MgCl₂, and 5 mM HEPES, pH 7.6. Two microelectrodes with tip resistances of 0.2 to 1.0 MΩ filled with 3 M KCl were inserted. Membrane potential was then held at −60 mV. Oocytes were continuously superfused at a flow rate of 4 ml/min with ND96 solution. All test compounds were added to the ND96 superfusion solution. The duration of agonist application was 10 s. When the effects of inhibitors were examined, oocytes were perfused with the solution containing inhibitor for 1 min and then stimulated by agonists for 10 s in the presence of inhibitors. After stimulation by agonists, the inhibitor solution was applied for another 1 min. When the effects of ADA were examined, oocytes were continuously perfused with the solution containing ADA, and agonists were applied with a separate line for 10 s. The current-voltage relation was obtained by ramp pulses using a function generator (NF-121B; NF Corp, Yokohama, Japan). Currents were continuously recorded and analyzed using MacLab (ADInstruments Pty Ltd., Castle Hill, Australia).

Measurement of Nucleotide Hydrolysis.

Oocytes in a 48-well plate (5 cells/well) were washed twice with ND96 solution and preincubated for 2 min in the presence or absence of inhibitors [α,β-MeADP, xanthine-amine congener (XAC), or pyridoxalphosphate-6-azophenyl-2′, 4′-disulphonic acid (PPADS)] in 200 μl of ND96 solution. Oocytes were then mixed with an equal volume of solution containing ATP or β,γ-MeATP and incubated at room temperature (approximately 22°C) for 10 min. In preliminary experiments, nucleotide hydrolysis increased linearly for at least 30 min. The incubation was terminated by the addition of EDTA to a final concentration of 10 mM. The supernatants were collected and stored at −20°C. Adenine nucleotides and their metabolites were measured by reverse-phase HPLC on an analytical C18 column (100 × 4.6 mm; YMC Co. Ltd., Kyoto, Japan) using 50 mM NaH₂PO₄ (pH 5.5) as the solvent at a flow rate of 1 ml/min (Matsuoka et al., 1995). Absorbance at 258 nm was monitored online with a UV detector (Nihon Bunko, Tokyo, Japan).

For the measurement of [³H]Ade formation from [³H]AMP on the membrane surface, oocytes in a 0.5-ml microcentrifuge tube (5 cells/tube) were washed twice and were incubated with ND96 solution containing ADA (1 U/ml) for 2 min. Cells were then pretreated with ATP, β,γ-MeATP, AMP, or α,β-MeADP for 30 s, followed by the addition of [³H]AMP (5 μCi/tube, to a final concentration of 2.5 μM) and incubation for 1 min. To investigate whether [³H]Ade existed on the cell surface, the reaction was stopped by aspirating the medium, followed immediately by adding 100 μl of 2.5% perchloric acid. The acid-extract (90 μl) was mixed with 9 μl of 4.2 M KOH to neutralize and deposit potassium perchlorate and was stored at −20°C until analysis by the use of HPLC. The sample was mixed with an equal volume of 100 μM Ade as a standard marker and [³H]Ade in 150 μl of mixture was separated by HPLC. Ade was monitored by UV detection, and the fraction containing [³H]Ade was collected and measured by a liquid-scintillation counter.

Statistics.

All experiments were repeated at least three times, and similar results were obtained. Statistical analyses of the data were performed by the paired Student's t test for two data comparison and one-way analysis of variance with the Dunnett two-tailed test for multiple data comparison. P values of less than 0.05 were considered to be statistically significant.

Drugs.

ADA, ATP, ADP, AMP, adenosine-5′-O-(3-thio)triphosphate (ATPγS), α,β-MeADP, and UTP were obtained from Sigma Chemical Co. (St. Louis, MO). β,γ-MeATP was purchased from Nacalai Tesque Inc. (Kyoto, Japan). 2-Methylthio ATP (2MeS-ATP), XAC, and PPADS were obtained from Sigma/RBI (Natick, MA). [2,8-³H]AMP (20 Ci/mmol, 1 mCi/ml) was obtained from Moravek Biochemicals (Brea, CA). All other chemicals were reagent grade or the highest quality available.

Results

Functional Expression of A_2B Receptors in X. laevis Oocytes.

We first examined the effects of Ade and ATP on the electrophysiological response of noninjected oocytes. Ade and ATP (both at 10 μM) had no effect on membrane current observed with a holding potential at −60 mV (Fig.1A). In oocytes coexpressing β₂ adrenoceptor and CFTR, epinephrine (Epi, 1 μM) induced a slowly developing Cl⁻ current, a response elicited by β₂ receptor-mediated increase in cyclic AMP, which activates CFTR by cyclic AMP-dependent phosphorylation (Bear et al., 1991). Exposure of these oocytes to Ade or ATP did not induce any current (Fig. 1 B), indicating that oocytes do not possess endogenous adenylyl cyclase-linked purinoceptors. In oocytes injected with the A_2B receptor cRNA alone, Ade and ATP failed to induce any membrane current (data not shown). When A_2B receptor cRNA was coinjected with β₂ adrenoceptor and CFTR cRNAs, the brief application of Ade (10 μM) for 10 s induced a membrane current (Fig. 1C). The Ade receptor antagonist XAC (1 μM) inhibited the currents induced by 10 μM Ade without affecting the Epi-induced current (Fig. 1C). In contrast, the β receptor antagonist propranolol (10 μM) inhibited the Epi-induced current, but it had no effect on the current induced by 10 μM Ade (Fig. 1C). These results suggest that the A_2B receptor is functionally expressed in cRNA-injected X. laevis oocytes.

Effects of Nucleotide Agonists.

In oocytes coexpressing the β₂ and A_2B receptors together with CFTR, exposure to ATP (10 μM) for 10 s induced a current (Fig. 2 A). This response to ATP was similar to those to Epi (1 μM) and Ade (10 μM) with respect to the time course of development and magnitude of the current (Fig. 2). β,γ-MeATP, an ATP analog that is metabolically more stable than ATP, also induced a current (Fig. 2). We also examined the effects of several nucleotide agonists, such as ADP, AMP, UTP, α,β-MeATP, ATPγS, and 2MeS-ATP at a concentration of 10 μM each. Among them, ADP, AMP, and ATPγS induced the current, whereas UTP, α,β-MeATP, and 2MeS-ATP essentially had no effect (Fig. 2). The current induced by nucleotide agonists exhibited characteristics of the Cl⁻ current through CFTR, as determined from the reversal potential and its shift by changing the external concentrations of Cl⁻ (data not shown; Uezono et al., 1993). In additional experiments, we used β,γ-MeATP as the nucleotide agonist.

Comparison of the Effects of Ade, AMP, ATP and β,γ-MeATP.

Fig. 3 A shows a representative chart recording of the effects of different concentrations of Ade and β,γ-MeATP on the membrane current in an oocyte coexpressing the CFTR and A_2B receptors. Brief applications (10 s) of increasing concentrations of Ade or β,γ-MeATP induced the CFTR current in a concentration-dependent manner. A similar concentration-dependent increase in the CFTR current was observed with AMP and ATP. Average values of current amplitude in five to six different oocytes from four different batches are shown in Fig. 3B. EC₅₀ values for Ade, β,γ-MeATP, ATP, and AMP were 4.3 ± 1.4 (n = 8), 7.8 ± 1.6 (n = 8), 7.4 ± 1.8 (n = 5), and 6.2 ± 1.2 μM (n = 5), respectively.

Effects of ADA and Purinoceptor Antagonists.

When oocytes were perfused with ADA (1 U/ml) containing medium, Ade (20 μM) applied by a separate line for 10 s failed to induce the CFTR current (Fig.4A). Under the same condition, there was still an effect of β,γ-MeATP, although it was smaller than in the absence of ADA. When the ADA concentration in the perfusing medium was increased up to 5 U/ml, β,γ-MeATP-induced current was inhibited (Fig. 4A). This inhibition by 5 U/ml ADA was not nonspecific effects, because such treatment did not affect the CFTR current induced byN-ethylcarboxamide adenosine, an ADA-resistant Ade analog (data not shown). Similar results showing a resistance to ADA (inhibited by 5 U/ml but not by 1 U/ml) were obtained with ATP and AMP (data not shown). We examined next the effects of purinoceptor antagonists on Ade- and β,γ-MeATP-induced currents. The Ade-induced current was inhibited by XAC, as shown in Fig. 1C. The P2 receptor antagonist PPADS (1 mM) had no effect on the response to 10 μM Ade (Fig. 4B). In contrast, the current induced by β,γ-MeATP (10 μM) was inhibited by both 1 mM PPADS and 1 μM XAC (Fig. 4). We previously showed that these antagonists at the same concentrations effectively inhibited a P1 antagonist-sensitive ATP response in C6Bu-1 rat glioma cells, in which XAC inhibited the response directly by blocking agonist-receptor interaction, whereas PPADS inhibited the response indirectly by preventing extracellular adenine nucleotide metabolism (Ohkubo et al., 2001).

Role of Metabolically Generated Ade in Nucleotide-Induced A_2B Receptor Activation.

To examine whether β,γ-MeATP activates A_2B receptors directly or indirectly through metabolically generated Ade, we determined the effects of α,β-MeADP, a potent inhibitor of ecto-5′-nucleotidase (Bruns, 1980), on the β,γ-MeATP-induced current. α,β-MeADP alone at the concentration tested (250 μM) had no effect on membrane current. Pretreatment of oocytes for 1 min with α,β-MeADP significantly inhibited the β,γ-MeATP-induced current without affecting the Ade-induced current (Fig.5). Similar results were obtained with the ATP- and AMP-induced response (data not shown).

We next examined whether oocytes could convert ATP or β,γ-MeATP into Ade. After a 10-min incubation, ATP (100 μM) was decreased by 57.7 ± 5.2% and converted into ADP, AMP, and Ade [24.5 ± 2.6, 9.2 ± 0.7, and 15.2 ± 2.2% (n = 5), respectively]. β,γ-MeATP (100 μM) was very resistant to hydrolysis, being decreased by 4.7 ± 1.6% and converted mainly to Ade 1.1 ± 0.2% (n = 5). ADP and AMP production from β,γ-MeATP was less than 0.5%. These results demonstrate that oocytes possess an ectonucleotidase cascade that can produce Ade from ATP and β,γ-MeATP. We next examined the effects of α,β-MeADP, PPADS, and XAC on Ade production from ATP and β,γ-MeATP (Fig. 6, A and B). Although the rate of hydrolysis of ATP in oocytes was largely different from that of β,γ-MeATP, Ade production from these nucleotides were affected similarly by inhibitors. α,β-MeADP (250 μM) and PPADS (1 mM) significantly inhibited Ade production from ATP and β,γ-MeATP, whereas XAC (1 μM) had no effect. These results indicate that the inhibitory effects of α,β-MeADP and PPADS on the CFTR current induced by ATP and β,γ-MeATP are caused by the inhibition of Ade production and that XAC acts as an A_2B receptor antagonist.

Local Ade Formation from Adenine Nucleotides on Oocyte Membrane Surface.

The results above suggest that ATP and β,γ-MeATP both activate the A_2B receptors after conversion into Ade. However, the Ade production from β,γ-MeATP was quite low compared with that from ATP, whereas these agonists induced nearly identical time- and dose-dependent responses, which were also similar to the Ade-induced response. A possible explanation for these observations is that the ectonucleotidase cascade led to a rapid and localized Ade formation from ATP and β,γ-MeATP to a similar extent on the membrane surface, where the Ade concentration is much higher than that measured in the bulk bath volume. To explore this possibility, we first examined [³H]Ade formation from [³H]AMP on the oocyte membrane surface. When oocytes were incubated with [³H]AMP (5 μCi/tube, to a final concentration of 2.5 μM) in the presence of ADA (1 U/ml), levels of membrane-associated [³H]Ade, which was detected in an acid-cell extract prepared after aspirating the incubation medium, was rapidly increased within 1 min (Fig.7). This increase in [³H]Ade was inhibited by α,β-MeADP (250 μM), suggesting that the Ade formation is mediated by ecto-5′-nucleotidase and that the generated Ade is retained in a membrane surface microenvironment even in the presence of ADA. If the conversions of ATP and β,γ-MeATP into Ade occur in the same environment, an excess amount of ATP and β,γ-MeATP should decrease the conversion of [³H]AMP into [³H]Ade through a competition with unlabeled AMP derived from ATP and β,γ-MeATP. As shown in Fig. 7, preincubation of oocytes with 250 μM ATP or β,γ-MeATP for 30 s significantly decreased [³H]Ade production to a similar extent, which was equivalent to the effects of 250 μM AMP. These results suggest that at the membrane surface, ATP, β,γ-MeATP, and AMP are equally converted into Ade.

Discussion

In this study, we succeeded in having functional A_2B receptors expressed in X. laevisoocytes by injecting them with cRNA from the rat A_2B receptor. Although oocytes injected with the A_2B receptor cRNA alone did not have an electrophysiological response to Ade, oocytes coexpressing the A_2B receptor and CFTR carried out A_2B receptor-mediated cyclic AMP formation through an activation of CFTR current. Cyclic AMP formation through CFTR current has been demonstrated in oocytes coexpressing CFTR with several Gs-coupled receptors, such as adrenergic β₂ (Uezono et al., 1993), vasoactive intestinal peptide, and pituitary adenylyl cyclase-activating peptide receptors (Uezono et al., 1997). In oocytes coexpressing A_2B and β₂ receptors together with CFTR, the Ade receptor antagonist XAC selectively inhibited Ade-induced CFTR current without affecting the Epi-induced current. In contrast, the β receptor antagonist propranolol inhibited the CFTR current induced by Epi, but not that induced by Ade. Therefore, we concluded that the Ade-induced response is mediated by A_2B receptors.

Using oocytes expressing the A_2B receptors, we found that ATP and β,γ-MeATP stimulated the CFTR current. The following results indicate that currents induced by ATP and β,γ-MeATP are CFTR currents mediated by an activation of A_2B receptors. First, the adenine nucleotides did not induce a current in oocytes injected with water or CFTR cRNA alone, indicating that the responses are dependent on the A_2B receptor expression. Second, the reversal potentials of the current at different Cl⁻concentrations were consistent with the CFTR Cl⁻current. Finally, the P1 antagonist XAC inhibited the adenine nucleotide-induced currents. β,γ-MeATP is suggested to stimulate a certain class of Ade receptors directly (Hourani et al., 1991). However, the activation of A_2B receptors in oocytes by β,γ-MeATP clearly required its conversion into Ade, because an ecto-5′-nucleotidase inhibitor, α,β-MeADP, blocked the β,γ-MeATP-induced current without affecting the Ade-induced current. These observations indicate that α,β-MeADP does not have an antagonist effect on the A_2B receptor and that the inhibitory effect on β,γ-MeATP-induced current is caused by a blockade of Ade formation. Correspondingly, the conversion of ATP and β,γ-MeATP into adenosine was inhibited by α,β-MeADP in oocytes. Furthermore, PPADS also inhibited Ade production from ATP and β,γ-MeATP in oocytes. Although PPADS is widely used as a P2 antagonist, it also inhibits the extracellular adenine nucleotide metabolism (Grobben et al., 1999). Our data in oocytes suggest that PPADS inhibits the adenine nucleotide-induced CFTR current as an ectonucleotidase inhibitor rather than as a P2 antagonist.

The involvement of adenosine in the effects of ATP and β,γ-MeATP was also supported by experiments with ADA. The CFTR current induced by ATP and β,γ-MeATP were inhibited by exogenously added ADA. However, this inhibition required high concentration of ADA (5 U/ml), because ATP- and β,γ-MeATP-induced responses were little affected by ADA at 1 U/ml, a concentration which abolished Ade-induced effects. Similar results were reported in the rat hippocampal preparation, in which 5 to 10 U/ml ADA is necessary for the removal of the membrane-associated Ade (Cunha et al., 1996). Studies from other laboratories demonstrated the P1 antagonist-sensitive ATP responses in the rat hippocampus (Dunwiddie et al., 1997; Cunha et al., 1998).

β,γ-MeATP has been considered to be a metabolically stable P2 agonist (Hourani et al., 1991). However, the present results suggest that β,γ-MeATP is immediately converted into Ade on the oocyte membrane surface, thereby causing the CFTR current, which was very similar to the ATP- and Ade-induced responses concerning the time course, the maximal current amplitude, and the effective concentration range of the agonists. This is somewhat surprising because only 1% of an added dose of β,γ-MeATP was converted into Ade in the bulk incubation medium within 10 min. However, a lack of correlation between the Ade formation and the induced response was reported in the inhibition of neurotransmitter release by ATP and β,γ-MeATP in the rat hippocampal slice (Cunha et al., 1998). We also reported that β,γ-MeATP induced a nearly identical response to ATP or AMP in a manner that was dependent on Ade formation in C6Bu-1 and NG108–15 cells, despite their marked difference in the rates of hydrolysis in the bulk medium (Ohkubo et al., 2000a, 2001). In those cells, we showed previously that [³H]Ade converted from [³H]ATP or [³H]AMP was differently distributed in the cell surface and bulk incubation medium, especially in the presence of ADA (Ohkubo et al., 2000a, 2001). Using a similar protocol, we demonstrated a rapid increase in [³H]Ade level within 1 min on the oocyte membrane surface during incubation with [³H]AMP in the presence of ADA. This [³H]Ade formation was inhibited by α,β-MeADP, suggesting that the Ade formation is mediated by ecto-5′-nucleotidase. If this [³H]Ade accumulation by [³H]AMP hydrolysis reflected the rapid activation of A_2B receptors by AMP, the hydrolysis of ATP or β,γ-MeATP would occur in the same environment on the membrane. Indeed, we observed that a brief treatment of oocytes with an excess amount of ATP or β,γ-MeATP markedly reduced the following [³H]Ade accumulation by [³H]AMP hydrolysis. These results suggest that ATP and β,γ-MeATP are rapidly and efficiently hydrolyzed into Ade on the membrane surface, leading to the A_2Breceptor activation similar to a direct effect of Ade.

The present results support a hypothesis proposed in our previous study with C6Bu-1 cells, stating that cyclic AMP formation induced by ATP and β,γ-MeATP are mediated by locally generated Ade on the membrane surface and subsequent activation of A_2Breceptors (Ohkubo et al., 2001). In C6Bu-1 cells, the effects of ATP and β,γ-MeATP were not affected by ADA treatment, but were inhibited by XAC, PPADS, and α,β-MeADP. ADP, AMP, and ATPγS mimicked those responses, whereas typical P2 receptor agonists, such as α,β-MeATP, 2MeS-ATP, UTP, and UDP, had no effect. In this study, we found that such characteristics were all reproduced in oocytes expressing the A_2B receptor. Therefore, we suggest that the ATP-induced response having the characteristics described above might occur in native tissues and cells through a combination of ectonucleotidase-dependent Ade formation and subsequent activation of Ade receptors. Although we demonstrated adenine nucleotide responses with the A_2B receptor, such responses would also occur with other Ade receptor subtypes. Indeed, ATP-induced inhibition of neurotransmitter release in the rat hippocampus (Dunwiddie et al., 1997; Cunha et al., 1998) and neuromuscular junctions (Sebastião et al., 1999) are likely to involve the A₁ receptor subtype. Furthermore, we demonstrated recently that ATP-induced cyclic AMP accumulation in a neuroblastoma cell line is mediated by the A_2Areceptor subtype, in which Ade formation was catalyzed by ectoalkaline phosphatase instead of ecto-5′-nucleotidase (Ohkubo et al., 2000b).

Recently, several ectoenzymes involved in adenine nucleotide metabolism have been identified and characterized (Zimmermann and Braun, 1999). These include the ectonucleotide triphosphatase family, which hydrolyzes either nucleoside 5′-triphosphates or both nucleoside 5′-tri- and diphosphates; the ectophosphodiesterase/pyrophosphatase family, which hydrolyzes nucleoside 5′-triphosphates directly to nucleoside 5′-monophosphates; and the enzymes catalyzing the conversion of nucleoside 5′-monophosphates to the corresponding nucleosides, including ecto-5′-nucleotidase and alkaline phosphatase. Ectonucleotidase activity in X. laevis oocytes has been examined extensively by Ziganshine et al. (1995, 1996a,b). They demonstrated that although the enzyme activity in oocytes is localized mainly in the follicle cell layer, defolliculated oocytes also possess moderate enzyme activity (Ziganshin et al., 1995, 1996b). However, the enzymes involved in nucleotide hydrolysis in oocytes have not yet been identified. Previously, we showed that C6Bu-1 cells express ectophosphodiesterase/pyrophosphatase1 and ecto-5′-nucleotidase, and these two enzymes seem to be sufficient for producing Ade from β,γ-MeATP (Ohkubo et al., 2001). In addition, PPADS was demonstrated to inhibit ectophosphodiesterase/pyrophosphatase activity in C6 glioma cells (Grobben et al., 1999). Therefore, it seems likely that oocytes express at least two different enzymes: ecto-5′-nucleotidase and the PPADS-sensitive ectophosphodiesterase/pyrophosphatase1-like enzyme.

There are many reports that methylxanthine derivatives block the P2 receptor agonist-induced response in several tissue preparations. Different purinoceptors are generally expressed in the same tissue and even in the same cells, making it difficult to elucidate the mechanism of action of ATP. In this study, using the expression of A_2B receptors in X. laevis oocytes, we showed that ATP and several adenine nucleotides can stimulate A_2B receptors after their conversion to Ade. These findings suggest that the close association of P1 receptors to ectonucleotidases may constitute a functional receptor for ATP and that some important physiological responses to ATP may occur through this mechanism.