Several 5-iodotubercidin analogues in the pyrazolo[3,4-d]pyrimidine ring system were synthesized as potential inhibitors of adenosine kinase by a direct Lewis acid-catalyzed glycosylation procedure using both the preformed carbohydrate and the heterocyclic base as starting materials. The 5'-hydroxyl, -chloro, -azido, -deoxy, -amino, and -fluoro derivatives were prepared and evaluated in three systems for biological activity relative to adenosine, the true substrate, and 5-iodotubercidin, a known inhibitor. First, each compound was studied kinetically for inhibition of purified human placental adenosine kinase activity. The order of potency was: iodotubercidin > hydroxyl > amino > or = deoxy > fluoro > chloro >> azido. The Ki values for the 5'-hydroxyl and 5'-amino compounds, the two most potent inhibitors, were 80 and 150 nM, respectively. The inhibition appeared to be essentially competitive in nature, although a noncompetitive component of significance for the more potent inhibitors cannot be ruled out. Second, a bioassay was conducted in which the toxicity of 6-methylmercaptopurine riboside toward human CEM lymphoblasts was reversed by varying concentrations of the compounds. The order of effectiveness of the compounds in this system, representing a functional inhibition of adenosine kinase in cultured cells, was about the same as that with the purified enzyme, except that the 5'-chloro and 5'-fluoro compounds were ineffective. Third, the 5'-hydroxyl derivative was evaluated in vivo in a rat pleurisy inflammation model and displayed biological activity at a dose of 30 mg/kg given orally. Finally, the in vitro toxicity of each compound was assessed in CEM lymphoblasts. Results indicated that the two most potent inhibitors in the pyrazolo[3,4-d]pyrimidine ring system, the 5'-hydroxyl (7) and the 5'-amino (20), were 15-fold and 75-fold, respectively, less growth inhibitory than 5-iodotubercidin.

A number of 3,4-disubstituted pyrazolo[3,4-d]pyrimidine ribonucleosides were synthesized and tested for their biological activity. Glycosylation of persilylated as well as nonsilylated 3-bromoallopurinol with 1-O-acetyl-2,3,5-tri-O-benzoyl-D-ribofuranose (4) provided the key intermediate 3-bromo-1-(2,3,5-tri-O-benzoyl-beta-D-ribofuranosyl)-pyrazolo[3,4-d] pyrimidin-4(5H)-one (5a). Similar glycosylations of 3-cyanoallopurinol and 3-(methylthio)allopurinol furnished the corresponding protected N-1 glycosyl derivatives (5b and 5c). Debenzoylation of these nucleosides (5a-c) gave the corresponding 3-bromo-, 3-cyano-, and 3-(methylthio)allopurinol nucleosides (6a-c). The site of glycosylation and anomeric configuration of 6a and 6c were assigned on the basis of spectral studies as well as conversion to allopurinol ribonucleoside, whereas the structural assignment of 6b was made by single-crystal X-ray analysis. Conventional functional group transformation of 5a and 5b provided a number of novel 3-substituted allopurinol nucleosides, which included 10a and 18a-d. Glycosylation of 4-amino-3-bromopyrazolo[3,4-d]pyrimidine (14) with 4 and subsequent debenzoylation gave 3-bromo-4-aminopyrazolo[3,4-d]pyrimidine ribonucleoside (13a) from which 3,4-diamino-1-beta-D-ribofuranosylpyrazolo[3,4-d]pyrimidine (13b) was obtained by amination. Thiation of 5b, followed by deblocking, gave 3-cyanothiopurinol ribonucleoside (20). All of these compounds were tested in vitro against certain viruses, tumor cells, and the parasite Leishmania tropica. Among the 3-substituted allopurinol nucleosides, 18b and 18c showed significant activity against Para 3 virus and were found to be potent inhibitors of growth of L1210 and P388 leukemia. Compound 20 exhibited the most significant broad-spectrum in vitro antiviral and antitumor activity. 3-Bromoallopurinol ribonucleoside (6a) was found to be more active than allopurinol ribonucleoside against Leishmania tropica within human macrophages in vitro.