Relative importance of mechanisms needs clarification.

1In the February issue of The FASEB Journal, Rayment and colleagues report on experiments addressing the mechanism by which ADP elicits relaxation of coronary smooth muscle (1). Based on measurements of protein expression, receptor binding assays, and classical receptor blocking, the authors show that the action of adenosine via A2A-receptors contributes significantly to the relaxatory response of ADP. Addressing the mode by which adenosine is recruited to mediate this effect, the authors have favored the view that ADP, via (a yet unknown) surface receptor interaction, triggers cellular adenosine release. We believe that the authors’ own data support a different conclusion. Of the different nucleotides applied (i.e., ATP, UTP, ADP), ADP exerted the most prominent relaxation over a concentration range from 10 to 5 10 M. Relaxation was also elicited by ADP S, a stable analogue of ADP which has been shown to act similarly to the native nucleotide. The involvement of adenosine in smooth muscle relaxation toward ADP is supported by experiments which interfere with the transport and metabolism of adenosine. Application of adenosine deaminase to the organ bath shifted the dose-response curve of ADP to the right. And application of blockers of equilibrative nucleoside transport (ENT), NBTI and dipyridamole, shifted the dose-response curve of ADP to the left. Thus, membrane nucleoside transport inhibition increased the effects of ADP at lower concentrations. The important question to be solved is the mode by which this adenosine is recruited to mediate the effects of ADP. The authors have favored the view that ADP triggers cellular release of adenosine which then acts on extracellular A2A-receptors—we question this interpretation because their Figure 5 shows that blockade of nucleoside membrane transport (ENT) augments the relaxatory effect of ADP application (1). If adenosine mediating this effect would have been produced inside the smooth muscle cells, then a smaller release would be expected and one would expect a shift of the dose-response curve to the right. Because ENT is equilibrative by definition, block of this transporter cannot impair cell uptake of adenosine and at the same time leave cellular release unaffected—a concept in agreement with previous experimental data on smooth muscle cell adenosine import and export (2). The argument that ENTs with different sensitivities toward NBTI could account for this effect can be ruled out using mass balance based mathematical modeling. Furthermore, assumption of concentrative nucleoside transport (CNT) in parallel to ENT does not change this conclusion. A possibility of asymmetrical transmembrane adenosine transport exists for sodiumdependent purine transport. However, to the best of our knowledge, functional significance of CNTs has never been documented for smooth muscle cells. Moreover, because the sodium concentration in extracellular fluid is far above the intracellular concentration, this transport would facilitate adenosine transport toward the cytosol. Because CNT is less sensitive toward inhibition by NBTI or dipyridamole, effective block of ENT would still permit sodium-dependent uptake of adenosine via CNT but simultaneously prevent cellular release. This result is opposite to that observed by Rayment and associates. Their result therefore clearly indicates that adenosine originated in the extracellular region and because its transport into the cytosol, and thereby its metabolism to AMP and inosine, was prevented, the relaxant effect of ADP via A2A-receptors was increased. Rayment et al. could not inhibit the ADP effect using different blockers of adenine nucleotide metabolism. Unfortunately, effectiveness of these blockers to inhibit adenosine production was not tested. In smooth muscle cells, extracellular adenosine is produced via ecto-5 nucleotidase (CD73), whereas alkaline phosphatase is unimportant (2). Because ecto-5 -nucleotidase was not considered in the Rayment et al. study, extracellular production of adenosine may have remained undetected. With respect to other ecto-ATPases it needs to be realized that parallel extracellular routes of ATP and ADP hydrolysis exist (3) and the effectiveness of ATPase block by ARL67156 may be weak (4). A further weakness of the concept of cellular adenosine release mediated by ADP is found in experiments reporting an increased release of radioactivity after [H]-adenine labeling. While adenine may be transported separately (5), unlabeled adenosine derived from extracellular ADP breakdown may have competed with radioactive inosine for equilibrative membrane nucleoside transport, potentially causing this effect. In conclusion, the provocative study by Rayment and colleagues adds an interesting perspective to the discus-