Table l. RELATIVE EFFECTIVENESS OF NUCLEOSIDES AS VASODEPRESSORS AND INHIBITORS OF PLATELET AGGREGATIONNucleosideAdenosine 2-Methoxyadenosine 2-Methylthioadenosine 2-TrinuoromethyladenosineInhibition of Vasodepressor potency platelet B-att Guinea-pig f aggregation J1 0-3 0-1 0-121 0-40-18 0-121 0-02 0-1 0-01Data are molar potency ratios.2-Methoxyadenosine6 and 2-trifluoromethyladenosine7 were synthesized as described before; our synthesis of 2-methylthioadenosine will be reported elsewhere. t Means of five or more separate observations. Doses were used which produced 25 per cent depression of blood pressure in animals anaesthetized with 'Nembutal'.t ADP-mediated platelet aggregation. Means of ten or more experiments using sheep platelet-rich plasma at 37 C. The inhibitors were pre-incubated in platelet-rich plasma for 2 min before ADP was added. 2 - Methylthioadenosine-5' - phosphate (2 - methylthio -AMP) was synthesized from the 2/,3'-O-isopropylidene derivative of 2-methylthioadenosine by the phosphoryla-tion procedure described by Tener2, and was isolated in good yield as white needles, melting point 192-195 C (decomposition), which were analytically pure (analysis calculated for C11H18N6O7PS-HaO: C, 32-12; H, 4-41; N, 17-03; P, 7-53 per cent. We found: C, 31-77; H, 4-42; N, 16-94; P, 7-45 per cent).Platelet aggregation studies were carried out on samples of platelet-rich plasma prepared from sheep, dog and human citrated blood using the turbidimetric technique reported before3. 2-Methylthio-AMP in micromolar concentrations inhibited the ADP-mediated aggregation of platelets in platelet-rich plasma from these three species and in Table 2 the potencies of 2-methylthio-AMP are compared with the potencies of AMP, adenosine and 2-methylthioadenosine. There was considerable species variation in the response of platelet aggregation to inhibition by 2-methylthio-AMP; sheep platelets were the most sensitive and human platelets the least sensitive. 2-Methylthioadenosine was twenty times less potent than 2-methylthio-AMP as an inhibitor of sheep and dog platelet aggregation, and 1-5 times less potent than 2-methylthio-AMP as an inhibitor of human platelet aggregation. When 2-methylthio-AMP was pre-incubated in sheep, dog and human platelet-rich plasma for up to 30 min before the addition of ADP the potency of the analogue was not affected, in contrast with the results of pre-incubating AMP and 2-chloro-AMP in sheep platelet-rich plasma4. With increasing periods of incubation the potency of AMP decreased while that of 2-chloro-AMP increased. This was presumably a consequence of the hydrolysis of AMP to adenosine and of 2-chloro-AMP to 2-chloroadenosirie, by plasma phosphatase or S'-nucleotid-ase and concomitant deamination of adenosine to inosine. 2-Methylthio-AMP can be presumed to be a substrate for plasma 5'-nucleotidase, for it has been shown to be a substrate for 5'-nucleotidase purified from rat heart, and it has a Fmax similar to AMP and a Km rather higher than that of AMP (unpublished results of M. Edwards and M. H. M.). 2-Methylthioadenosine was less potent than 2-methylthio-AMP as an inhibitor of platelet aggregation and this suggests that 2-methylthio-AMP acts directly on the adenosine-ADP platelet receptor3'5, rather than acting after hydrolysis to 2-methylthioadenosine. Using human platelet-rich plasma the log dose-response curves for ADP-mediated platelet aggregation in the absence of inhibitors and in the presence of 2-methylthio-AMP, AMP and adenosine were parallel, indicating a direct competitive mode of action for 2-methylthio-AMP at the adenosine-ADP receptor (Fig. 1). Table 2. INHIBITORS OF PLATELET AGGREGATIONConcentration (^M) required for 50 per cent inhibition2-Methylthio-AMP AdenosineAMP 2-MethylthioadenosineSheep 0-51-2 10-011-0 Platelets Dog1-3 2-4 18t 20 fHuman 8-61-0 10-5 12-8The inhibitors were pre-incubated in platelet-rich plasma at 37 C for 2 min before 0-6 ^uM ADP was added. The values represent the means of results obtained with three or more batches of platelet-rich plasma except for two cases when only one batch of plasma was used |-2-Methylthio-AMP (750 [Jt.g/kg) was administered to dogs anaesthetized with 'Xembutal', and blood was sampled up to 3 h later. The response of platelets to ADP decreased, indicating that 2-methylthio-AMP in vivo reduces the aggregatability of platelets, and that the effect of the analogue on the platelets lasts for up to 3 h. The aggregates which formed in this case disaggregated more rapidly than aggregates formed in platelet-rich plasma prepared from blood samples before the administration of 2-methylthio-AMP. Such accelerated disaggregation of ADP-induced platelet aggregates was not observed in platelet-rich plasma prepared from animals subjected to the same experimental treatment, but without receiving 2-methylthio-AMP. These observations suggest that 2-methylthio-AMP induced changes in the platelets themselves. Sheep, human and dog platelets aggregated by ADP in vitro in the presence of 2-methylthio-AMP also disaggregated more rapidly than control aggregates. 2-Methylthio-AMP, given intravenously in doses of 2 mg/kg, had no effect on the blood pressure of anaesthetized guinea-pigs and rats, while 680 (Jig/kg and 1,000 [xg/kg of 2-methylthioadenosine caused a 25 per cent decrease in blood pressure of guinea-pig and rat, respectively. A dose of 750 (xg/kg of 2-methylthio-AMP had no effect on the blood pressure, heart rate or respiration rate of the dog. The analogue administered intravenously to mice, in a dose of 75 mg/kg, was found to be non-toxic, and produced no observable overt effects.Inhibitory properties have been separated clearly from vasodilatory properties in the case of 2-methylthio-AMP platelet aggregation. The analogue seems to show great promise as a drug for use in clinical situations where platelet stickiness must be reduced, for example, in postoperative situations, for heart implants and in rejection crises after organ transplants when vascular blockades are caused by platelet aggregates. Fig. 1. Log dose-response curves for the aggregation of platelets in human platelet-rich plasma in response to ADP alone (o), and in the presence of 5 juM AMP (A), 1-2 (M. adenosine (O), and 15 ^M 2-methylthio-AMP (A). The initial rates of aggregation were measured as described earlier. Points shown are results obtained with two different batches of plasma. We thank Professor R. H. Thorp for his interest and encouragement, and Dr J. Fildes of the Australian National University for carrying out the micro-analysis of 2-methylthio-AMP.