| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* School of Biomedical Sciences, University of Nottingham Medical School, Queen’s Medical Centre, Nottingham, UK; and
School of Pharmacy, University of Nottingham, University Park, Nottingham, UK
1 Correspondence: School of Biomedical Sciences, University of Nottingham Medical School, Queen’s Medical Centre, Nottingham NG7 2UH, UK. E-mail: steve.alexander{at}nottingham.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: RT-PCR • adenosine A2A receptor • purine release • adenine nucleotides
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4)
. ATP may be released from perivascular nerves, from which it has been suggested to act at members of the ligand-gated channel family, P2X receptors, to elicit depolarization of smooth muscle leading to vasoconstriction (1
2
3
4)
. Luminal ATP may be released by blood components, including platelets (5)
and erythrocytes (6)
, or it may be released by the endothelium under conditions of shear stress (7)
or receptor activation (8)
. Luminal ATP appears to interact predominantly with P2Y receptors (7-transmembrane receptors) on either the endothelium or smooth muscle. Activation of endothelial P2Y receptors appears to lead to generation of mediators to elicit vasodilatation, while P2Y receptors on the smooth muscle tend to evoke vasoconstriction (1
2
3
4)
.
To date, eight distinct mammalian P2Y receptors have been identified (4, 9) and expressed in heterologous systems. In heterologous expression systems, P2Y1, P2Y2, P2Y4, P2Y6, and P2Y11 receptors couple primarily via members of the Gq/11 family of G-proteins to activate phospholipase C, leading in turn to elevation of intracellular calcium levels (3, 4, 10). Unusually, the P2Y11 receptor can activate adenylyl cyclase activity via Gs family members as well as elevating intracellular calcium (11)
. P2Y12, P2Y13, and P2Y14 receptors, on the other hand, couple primarily to inhibition of adenylate cyclase, leading to a reduced accumulation of cyclic AMP (4, 9).
P2Y receptors can also be divided on the basis of their endogenous agonists into adenine nucleotide-preferring (P2Y1, P2Y11, P2Y12 and P2Y13) and uracil nucleotide-preferring (P2Y2, P2Y4, P2Y6 and P2Y14) receptors (4, 9). Among the adenine nucleotide group, the human P2Y11 is selectively activated by ATP and fails to respond to ADP (11)
, although it appears that the dog orthologue responds to both ADP and ATP (12)
. P2Y1, P2Y12, and P2Y13 receptors have been proposed to be physiologically activated by ADP, while ATP is a reduced efficacy agonist (13
14
15
16)
.
In a variety of species and blood vessels, ADP-evoked relaxations are mediated by endothelial P2Y1 receptors (1
2
3
4)
. It appears, however, that vasorelaxation by adenine nucleotides acting directly on the smooth muscle is common in coronary arteries of various species, including rabbit, guinea pig, lamb, and human (17
18
19
20)
. In our preliminary studies, we observed that ADP mediated a relaxation of the porcine coronary artery (PCA), which was independent of the presence of an intact endothelium. Little is known about the receptor subtype or mechanism mediating this response, although the pharmacological and distribution profiles suggest the involvement of P2Y1 receptors. Given the association of Gq/11-coupled receptors, including the P2Y1 receptor, with muscle contractions, we sought to investigate further the mechanism of ADP-induced smooth muscle relaxation of the PCA. In doing so, we obtained evidence for a novel mechanism of action of adenine nucleotides in the vasculature, in that they stimulate the release of adenosine, which goes on to act at cell-surface A2A receptors to cause relaxation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 22)
. Hearts from modern hybrid white pigs (either sex, age less than 6 mo, wt 
50 kg) were removed and rapidly transported in iced, Krebs’-Henseleit solution to the laboratory. The proximal portion (ca. 5 cm) of the left circumflex coronary artery was removed and then stored overnight at 4°C in gassed (95% O2, 5% CO2) Krebs’-Henseleit solution containing 2% Ficoll (type 70000). After allowing time for equilibration to room temperature, a fine dissection was conducted to prepare tissue rings. Except where indicated, rings were denuded of endothelium by gentle rolling of the segment on a tissue paper pad after insertion of the ends of fine-tipped forceps into the lumen. An absence of substance P-evoked relaxation confirmed that this methodology was successful.
Responses in the porcine isolated coronary artery
Arterial rings were mounted onto wires in a tissue bath (10 ml) containing warm (37°C), oxygenated Krebs’-Henseleit solution and were connected via isometric force transducers (Grass FT03) to a PC running the computer program SPIKE 2.0 (CED, Cambridge, UK). Rings were put under tension (10 g) and allowed to equilibrate for 60 min before assessing viability with two challenges of 60 mM KCl. Another 60 min was allowed before addition of the thromboxane A2 analog U46619 (11
,9-epoxymethano-PGH2 up to 100 nM) to elicit a contraction of 60% of that to KCl. Concentration-relaxation curves were constructed in the presence of increasing cumulative concentrations of agonist. Unless otherwise stated, the effects of antagonists or modulators were examined by application prior to the addition of U46619, which ensured a minimal contact time of 20 min prior to agonist application.
RT-PCR and sequence analysis of the porcine P2Y1 and A2A receptor
Total cellular RNA was isolated from the PCA using the RNeasy Mini protocol (Qiagen, Crawley, UK) for heart, muscle, and skin tissue. RNA was incubated with 25 Kunitz units DNase for 30 min at 37°C to remove any contaminating DNA before further purification using the RNeasy cleanup protocol (Qiagen). RNA was subjected to reverse transcription, and only samples that showed no evidence of genomic DNA contamination were used in subsequent analysis. Polymerase chain reaction (PCR) was performed using primers specific for hypoxanthine phosphoribosyltransferase (hprt), p2ry1, and adora2a (Table 1
) using an Eppendorf gradient thermocycler. Primers for hprt were designed based on available sequences (Genbank accession number AF143818, Table 1
). At the time of conducting these experiments, sequence data were not available for either the P2Y1 or A2A receptor genes in pigs. For p2ry1, primers were initially designed based on sequence homology between bovine and human sequences (Genbank accession numbers X87628 and U42030); for adora2a, primers were based on the PCR product from mouse-specific primers (23)
. cDNA was amplified using a multiplex PCR kit (Qiagen) to generate PCR products for hprt, p2ry1, and adora2a in the same reaction. PCR was initiated using the following profile: 1 cycle of 95°C for 15 min; 28 cycles of 94°C for 90 s, 58°C for 90 s and 72°C for 90 s; 1 cycle of 72°C for 10 min. RT-PCR reaction products were fractionated through 3% agarose gels stained with Sybr green I, and visualized on the Chemi-doc system (Bio-Rad, Hemel Hempstead, UK).
|
Immunoblotting
Protein samples were prepared from endothelium-denuded tissue segments by homogenization in buffer (20 mM Tris, 1 mM EGTA, 320 mM sucrose, 0.1% Triton X-100, 1 mM sodium fluoride, 10 mM glycerol-2-phosphate, pH 7.6) containing a protease inhibitor cocktail. After electrophoresis on 10% SDS-PAGE gels (5 µg protein per lane), samples were transferred to nitrocellulose membranes. Blots were blocked for 1 h in TBS-Tween containing 5% milk before exposure to antibodies (rabbit anti-hP2Y1 1:1000 and rabbit anti-hA2A 1:500 in blocking solution) overnight at 4°C. After incubation with HRP-conjugated goat anti-rabbit secondary antibody (Ab) (1:2000 in blocking solution), blots were developed for chemiluminescent detection.
Radioisotope procedures
Binding to A2A receptors was investigated using pig caudate membranes and [3H]-ZM241385 as described previously (24)
.
To investigate the potential effects of adenine nucleotides on nucleobase/nucleoside/nucleotide release, endothelium-denuded tissue segments were labeled at 37°C with [3H]-adenine for 30 min at 740 kBq/mL in Krebs’-Henseleit Ringer. After removing excess [3H]-adenine, tissue sections were transferred to tissue chambers of a Brandel Suprafusion System (25)
and suprafused with warm, oxygenated Krebs’-Henseleit solution continuously for 60 min at 37°C at a flow rate of 0.5 ml/min. Subsequent suprafusate was collected in 4 min fractions. After baseline values had been obtained, samples were incubated with Krebs’-Henseleit solution containing U46619 (3 nM) for 6 fractions, followed by a period of exposure to U46619 and ADP (30 µM), before returning to suprafusion with U46619 alone. At the end of the experiment, tissue was removed from the chamber and remaining radioactivity was assessed by liquid scintillation counting.
Sample analysis was performed using HPLC using an Agilent HP 1090 HPLC (Waldbron, Germany), fitted with a Kromasil 3.5 µm particle size column (100x2.1 mm, Phenomenex, Macclesfield, UK), eluted at 0.2 ml/min and controlled by Chemstation software. Fractions from six different arterial segments stimulated with U46619 in the absence and presence of ADP were separately pooled and subjected to solid-phase extraction (Strata X-C columns, Phenomenex) according to the standard protocol, concluding with a final elution vol of 6 ml methanol. After evaporation of the methanol under a stream of N2, samples were redissolved in 60 µl water for HPLC separation. Purine standards (200 µM) were used in each analytical run to confirm retention times. Elution of purine standards (retention times), as measured by UV absorbance (254 nm), were: 2.22 (hypoxanthine); 3.25 (adenine); 3.57 (inosine); and 7.76 min (adenosine). Mobile phase A was 20 mM ammonium acetate, pH 4; mobile phase B was methanol. The gradient used was 5% B for 1 min, increasing linearly to 50% B, over 6 min; 50% B was maintained for 4 min before re-equilibration to 5% B for 5 min. Sample effluate fractions were collected every 30 s and quantified by liquid scintillation counting.
Data analysis
Unless specified otherwise, data reported are means ± SEM of results from at least five separate preparations. Concentration-response data were initially compared for statistical significance by two-way ANOVA (Prism, GraphPad, San Diego, CA, USA) and then Student’s unpaired t test was applied to test for significant effects on pEC50 and Rmax values. Antagonist affinities were calculated as pKi values using the Gaddum transformation: pKi = log(CR-1) – log[B], where CR is the concentration-ratio of EC50 values in the absence and presence of the antagonist and [B] is the concentration of antagonist. Tissue [3H]-nucleotide/nucleoside release in individual fractions was calculated as a percentage of the total radioactivity of tissue and suprafusate. Evaluation of sequence data homology to known sequences was achieved using the ClustalW program (26)
.
Materials
Nucleotides, levamisole, and NBTI were obtained from Sigma-Aldrich (Poole, UK). MRS2179, ZM241385, and ARL67156 were obtained from Tocris-Cookson (Bristol, UK). Protease inhibitor cocktail set 1 (Calbiochem/Merck Biosciences, Nottingham, UK) was used in sample preparation for immunoblotting. Antibodies were obtained from Alomone Labs/Caltag-Medsystems (Buckingham, UK, anti-P2Y1 Ab), Calbiochem/Merck Biosciences (Nottingham, UK, anti-A2A Ab), and DAKO/Cytomation (Ely, UK, HRP-conjugated goat anti-rabbit Ab). Samples were prepared for RNA analysis using kits from Qiagen and cDNA generated using Expand RT enzyme (Roche, Lewes, UK). All primers were designed in-house and synthesized by Eurogentec (Southampton, UK). [3H]-Adenine was obtained from Amersham International (Little Chalfont, UK), while [3H]-ZM241385 and SCH58261 were kind gifts from Tocris-Cookson (Bristol, UK) and Dr. Silvio Dionisotti, Schering-Plough (Milan, Italy), respectively.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). Denudation of the endothelium failed to alter significantly relaxation parameters for either ADP or ATP (Fig. 1A
), although relaxations to uridine triphosphate (UTP) were abolished (data not shown). At a concentration of 300 µM, the relaxation evoked by ATP was reversed, presumably through activation of contractile P2X receptors. ADP was utilized in the majority of subsequent experiments, as it gave a larger relaxation that was not reversed at higher concentrations.
|
Relaxations to ADP in the porcine coronary artery were compared to other nucleotides. Cytidine triphosphate (CTP) (1 mM) evoked a modest relaxation (14±6%), while TTP (1 mM) and 2'-deoxyATP (30 µM) failed to evoke significant relaxations (data not shown). At concentrations up to 100 µM, ATP evoked a relaxation of slightly lower potency and efficacy to ADP (Fig. 1B
, Table 2
). Compared with ADP, relaxations were also more variable, presumably due to the contribution of an ATP-evoked contractile effect. The poorly metabolized analog of ADP, ADPßS, evoked a less-potent relaxation than ADP, but with similar maximal relaxations (Fig. 1B
, Table 2
).
|
These data pointed toward the potential contribution of a P2Y receptor in the relaxatory responses to adenine nucleotides. The relative potencies of ADP and ATP indicated the possible involvement of P2Y1 receptors. We set out, therefore, to assess whether P2Y1 receptors were expressed in the porcine coronary artery. Since sequence data for P2Y1 receptors at the time of conducting these experiments were limited to human, cow, mouse, rat, chicken, and turkey, we had to design primers for mRNA detection in the pig based on sequence similarities of the defined orthologues. PCR analysis and subsequent screening of the product allowed refinement of primers to apply in RT-PCR screening of porcine tissues (Table 1)
. The primary sequence of the porcine receptor (GenBank AY691680, 805 bp, 268 aa) showed greatest homology to the bovine receptor (94% identity), with a high level of homology with the human (92%), rat, and mouse (88%) receptors. Since it is established in the literature that the porcine coronary artery expresses A2A adenosine receptors which mediate relaxations to adenosine (27
28
29)
, we used this receptor as a positive control. Again, sequence data for the porcine orthologue were unavailable at the time of conducting these experiments, and so we used mouse-specific primers (30)
to identify a region of the porcine A2A cDNA from coronary artery that allowed the generation of pig-specific primers. This region of the porcine A2A receptor (GenBank AY686733, 768 bp, 256 aa) showed greatest homology to the human receptor (90%) with a high degree of homology to the guinea pig (88%) and rat receptor (83%).
A representative agarose gel of the multiplex RT-PCR products in the PCA with products at the expected sizes (hprt 210 bp; p2ry1 256 bp; adora2a 287 bp) is shown in Fig. 2
A. Products corresponding to A2A and P2Y1 receptors, as well as the housekeeping gene (HPRT), were present in all samples examined.
|
It is accepted that expression at the level of transcription of the gene is not always reflective of protein expression, so we next examined for immunoreactivity for P2Y1 and A2A receptors in preparations from the PCA (Fig. 2B
), having verified cross-reactivity of the Ab with porcine receptors using the pig caudate as a positive control (data not shown). Using the anti-P2Y1 Ab, we observed multiple bands of immunoreactivity in the PCA. All of these bands were eradicated by preincubation of Ab with blocking peptide, with the exception of the 163 kDa band. Indicated (by the arrow on the right-hand side) is a band of estimated MW 42 kDa (Fig. 2B
, lane 1), the size of the full-length P2Y1 receptor from human and cow, which was also observed in samples from the pig brain (data not shown). For A2A receptor-like immunoreactivity in the PCA, we observed a diffuse band at 48 kDa (Fig. 2B
, lane 2, indicated by the arrow on the right-hand side), consistent with the expected size of the full-length receptor from human and cow is 46 kDa. This band was also observed in samples from the pig caudate (data not shown).
We then sought to assess the involvement of P2Y1 receptors through pharmacological means. PPADS is a useful general P2 receptor antagonist, while MRS2179 is a more selective P2Y1 antagonist (31, 32). Thus, we examined their ability to modify ADP-evoked relaxations in the porcine-isolated coronary artery (Fig. 3
A). However, neither PPADS (10 µM) nor MRS2179 (1 µM) evoked a significant change in either the potency (pEC50 values+PPADS 5.54±0.10; +MRS2179 5.30±0.10) or efficacy (Rmax values+PPADS 94±4%; +MRS2179 98±5% relaxation) of the ADP response (Fig. 3A
).
|
Having established the presence of A2A adenosine receptors in this tissue using molecular means, we confirmed the affinity of the A2A-selective antagonists SCH58261 and ZM241385 (29, 33) as antagonists of relaxations evoked by adenosine and NECA (a nonselective adenosine receptor agonist). Apparent pKi values (calculated using the Gaddum transformation) for ZM241385 as an antagonist of adenosine and NECA relaxations were 8.8 ± 0.4 and 8.8 ± 0.3, respectively. A similar exercise generated estimates of SCH58261 affinity against NECA- and CGS21680-evoked relaxations of 8.2 and 8.2, respectively, slightly below literature estimates for A2A affinity in this tissue (29)
. Somewhat unexpectedly, these two antagonists could also induce significant rightward shifts in the potency of ADP and ATP as relaxants in the PCA. Estimation of the antagonist affinity of ZM241385 as an antagonist of ADP and ATP responses allowed calculation of pKi values of 9.2 ± 0.4 and 9.0 ± 0.3, respectively. SCH58261 (10 nM) also induced a significant rightward shift in the potency of ADP, with a calculated pKi value of 8.9 (Fig. 3B
).
Since the affinities of ZM241385 and SCH58261 were entirely consistent with an action of ADP at A2A adenosine receptors, we investigated whether the active adenine nucleotides could occupy A2A receptors directly. [3H]-ZM241385 (0.34 nM) binding to A2A adenosine receptors was assessed in particulate preparations from the pig caudate nucleus (34)
at 37°C after a 5 min incubation, in order to minimize the potential effects of metabolic enzymes. At 1 mM, adenosine evoked a complete displacement of [3H]-ZM241385 binding (102±1%, n=3). In contrast, ADP and ADPßS (both at 1 mM) failed to alter significantly the binding of [3H]-ZM241385 (–2±2 and –1±3%, respectively).
Since these observations precluded a direct activation of A2A receptors as the mechanism of ADP-evoked relaxations in the PCA or the possibility of contaminating adenosine in the ADP samples, we investigated whether degradation of ADP to adenosine was a potential alternative explanation. Although responses to ADP were rapid and thus seemed to militate against metabolism as an explanation, metabolism of adenosine by inclusion of adenosine deaminase (1 U/ml, Fig. 3B
) in the organ bath significantly reduced the potency of ADP (pEC50 value+ADA 4.75±0.06) without altering significantly the efficacy (98±4% relaxation). To investigate the potential metabolic routes involved in the generation of adenosine from ADP, we used an inhibitor of ecto-ATPase/AMPase, ARL67156 (35, 36), and an inhibitor of alkaline phosphatase, levamisole, which can also hydrolyze nucleotides in intact preparations (37)
. The presence of ARL67156 (100 µM, n=4) or levamisole (300 µM, n=3) failed to alter either the potency (pEC50 values+ARL67156 5.19±0.12; +levamisole 5.41±0.07) or efficacy (Rmax values+ARL67156 100±7%; +levamisole 97±4%) of ADP. In parallel experiments, 100 µM ARL67156 significantly enhanced contractile responses to UTP in the porcine coronary artery (Rayment et al., manuscript submitted). 5'-(4-fluorosulfonyl)benzoyladenosine (10 µM), an inhibitor of apyrase (38, 39), also failed to alter the ADP-evoked relaxation (pEC50 value 5.33±0.12 and Rmax 93±6%).
Given that multiple routes for nucleotide hydrolysis exist in intact tissues (40)
, the metabolism of ADP was also investigated by using other nucleotides to compete with hydrolysis of ADP by metabolic enzymes. These experiments examined relaxatory responses to a single concentration of ADP (30 µM) after preincubation with TTP (1 mM), CTP (1 mM), or 2'-deoxyATP (30 µM). However, none of these agents reduced significantly ADP-evoked relaxations (control 68±8% of U46619 induced tone; +CTP 71±9; +TTP 83±8; +deoxyATP 68±11%).
A recent report examining rat hippocampal slices suggested that adenine nucleotides could elicit "heteroexchange", whereby adenosine (and other adenine nucleotides) were released through the ENT1 adenosine transporter (41)
. After prelabeling with [3H]-adenine, segments of coronary artery were suprafused with Krebs’-Henseleit medium and then U46619 in the absence and subsequent presence of ADP (Fig. 4
A). The addition of U46619 alone to the tissue did not affect release of radioactivity (mean basal release 0.090±0.005% accumulated [3H]-adenine metabolites; +U46619 0.097±0.003%), but the further addition of ADP caused an increase in the amount of [3H]-adenine metabolites (0.32±0.04%, P<0.0001). After ADP was removed, overflow of radioactivity was significantly reduced (0.17±0.01%, P<0.005).
|
To determine the nature of the released radioactivity, we used HPLC separation (Fig. 4B
). Comparing elution profiles of suprafusate with standards indicated an absence of adenine nucleotides, with the principal agents eluted being the nucleobases and nucleosides hypoxanthine, adenine, and inosine (the latter two agents were indistinguishable using this system), with a modest component of the nucleoside adenosine eluting later (Fig. 4B
). These peaks were not observed in tissues treated with U46619 alone.
To investigate the role of nucleoside transporters in the ADP-evoked release of adenosine and the subsequent relaxation, we examined responses in the presence of the ENT1 inhibitors NBTI and dipyridamole (9)
. Both agents (at 100 nM) significantly enhanced the potency (pEC50 values+NBTI 5.90±0.06; +dipyridamole 6.24±0.23; P<0.005) as well as the efficacy (Rmax values+NBTI 106±3; +dipyridamole 103±5%; P<0.05) of ADP-evoked relaxations (Fig. 5
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Porcine P2Y1 and A2A receptors
To assess the contribution of P2Y1 receptors to adenine nucleotide responses in the PCA, we found it necessary to sequence, at least in part, the porcine orthologue of this receptor. The nucleotide sequence we obtained (GenBank AY691680) exhibits 88% or greater homology with human, rat, mouse, and bovine orthologues. Similarly, the partial sequence we obtained for the porcine A2A receptor (GenBank AY686733) exhibits a high degree of similarity with A2A receptors from man (90%) and rodents (>83%).
Experiments assessing both mRNA and immunoreactivity for P2Y1 receptors indicated expression in the PCA. Since tissue taken for these analyses was endothelial denuded, these data suggest the expression of P2Y1 receptors in the smooth muscle of this tissue. The agonist rank order of potency for relaxation was also consistent with a contribution of P2Y1 receptors. However, P2Y1 receptors are thought to act primarily via activation of Gq/11 G-proteins to elevate intracellular calcium ions, which would be expected to lead to smooth muscle contraction, rather than relaxation. Furthermore, the lack of effect of the P2Y1-selective antagonist MRS2179 (31, 32) and the nonselective antagonist PPADS (3, 4) suggested lack of involvement of P2Y1 receptors and P2 receptors, respectively.
Given the presence of a relaxatory A2A adenosine receptor in the PCA, which was confirmed by RT-PCR, immunoblot (Fig. 2)
and quantitative pharmacological assessments using both selective agonist (CGS21680) and antagonists (SCH58261 and ZM241385), we assessed next the potential for the involvement of A2A receptors in the relaxatory response to ADP.
Involvement of the A2A receptor in ADP-mediated relaxation
The almost identical affinity estimates obtained for SCH58261 and ZM241385 as antagonists of A2A adenosine receptors and ADP-evoked relaxation indicate that the ADP response is indeed mediated via A2A receptors. We ruled out a direct activation of A2A receptors by ADP and ADPßS as well as the possible contamination of these nucleotides by adenosine, since [3H]-ZM241385 binding to pig caudate membranes indicated complete occupancy by millimolar adenosine, without a significant effect of the same concentration of the nucleotides.
We showed the involvement of extracellular adenosine in the relaxatory effect of ADP, since adenosine deaminase was an effective inhibitor of the ADP response. Using a variety of enzyme inhibitors, however, we could not support a role for extracellular hydrolysis of adenine nucleotides in the generation of adenosine. A further alternative, therefore, was that ADP evoked release of adenosine from the smooth muscle, which has previously been reported to occur in the rat hippocampal slice preparation, with a suggested involvement of ENT1 nucleoside transporters (41)
.
Mechanisms of release
We investigated the release of radioactivity from [3H]-adenine-prelabeled PCA segments and observed overflow of tritium only in the presence of ADP (Fig. 4)
. In the presence of U46619 alone, no change was found in radioisotope release compared with basal. Separation of the release radioactivity in pooled samples allowed identification of hypoxanthine, adenine/inosine, and adenosine as the nucleobases and nucleosides released. The proportion of adenosine was relatively low compared with hypoxanthine and adenine/inosine, which indicates a major influence of metabolism within the extracellular space of the arterial segment to degrade released adenosine. Given that the major peak of radioactivity eluted at a time where adenine and inosine would elute suggests that either the major entity released is an unmodified precursor or inosine derived from the degradation of adenosine via ecto-adenosine deaminase activity (42)
.
We attempted to make the conditions of release as similar as possible to those in the contractile experiments, but it is possible that inclusion of a transport or metabolic inhibitor would have increased the yield of evoked adenosine release. Indeed, when we investigated the potential influence of ENT1 transport inhibitors on relaxatory responses to ADP, we observed significant enhancements of the relaxatory response. From this, we infer that adenosine is released via a mechanism other than through equilibrative (bidirectional) ENT1 transporters. One possibility would be ENT2 transporters, for which there are no selective inhibitors identified.
The observed enhancement of relaxations to ADP in the presence of ENT1 inhibitors contrasts with the suggested mechanism of action of adenine nucleotides in the rat hippocampal slice, where ENT1 inhibitors prevented "heteroexchange" (41)
, although it is possible that other members of the nucleoside transporter families may be involved in the relaxatory responses to ADP in vascular tissue. An alternative explanation of our results involves a novel PPADS-insensitive P2 receptor at which ADP and ATP are slightly more potent than ADPßS, which stimulates the release of adenosine in the coronary artery. This pharmacological profile does not readily fit into the current taxonomy of P2 receptors.
Interestingly, although examples in the literature of "vasorelaxant smooth muscle P2Y receptors" are rare, they appear to be expressed commonly in coronary arteries of a variety of species, including rabbit (17, 18), guinea-pig (18)
, lamb (19)
, and human (20)
. It is interesting to speculate that the evoked release of adenosine by nucleotides, as we identify here, is a common property in the coronary arteries of all these species.
It is possible that this mechanism explains some of the findings in the literature, which describe common responses evoked by both adenosine and adenine nucleotides (43
44
45
46
47)
. In these reports, P1 (adenosine) receptor antagonists blocked responses to adenosine analogues and adenine nucleotides with similar affinities, leading to the hypothesis of a P3 receptor distinct from P2 and P1 (adenosine) receptors. One prediction based on the nucleotide-evoked release of adenosine observed in the current study is that the apparent profile of ‘P3' receptor antagonists will vary dependent on which adenosine receptor is involved in the particular tissue/response. Thus, in tissues where A1 receptors predominate, A1-selective antagonists would be more potent than A2A-selective antagonists, while in A2A-expressing tissues, the reverse would be true. This phenomenon may help to explain why such a receptor has not been identified in deorphanization programs.
Concluding remarks
In this study, we have identified a novel mechanism of vasoregulation, whereby adenine nucleotides stimulate adenosine release from the vascular smooth muscle of the pig isolated coronary artery to act in a paracrine fashion on A2A receptors in this tissue to cause relaxation.
|
ACKNOWLEDGMENTS |
|---|
Received for publication August 2, 2006. Accepted for publication September 6, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1-adrenoceptor stimulation. Br. J. Pharmacol. 113,1203-1208[Medline]
-methyleneATP- and CGS21680-induced cyclic AMP generation in NG108–15 cells. Br. J. Pharmacol. 129,291-298[CrossRef][Medline]Related Articles
This article has been cited by other articles:
|
|
 |
|
  P. Dieterich, C. von Klitzing, and A. Deussen Relative importance of mechanisms needs clarification FASEB J, July 1, 2007; 21(9): 1951 - 1952. [Full Text] [PDF]
|
|
|
|
|
|
S. J. Rayment, V. Ralevic, and S. P. H. Alexander Response to: "Relative importance of mechanisms needs clarification" FASEB J, July 1, 2007; 21(9): 1953 - 1953. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
