FACTORS DETERMINING THE ACTIVITY OF 2',3'-DIDEOXYNUCLEOSIDES IN SUPPRESSING HUMAN IMMUNODEFICIENCY VIRUS INVITRO

Abstract

Mitsuya and Broder [Proc. Natl. Acad. Sci. USA 83:1911-1915 (1986)] demonstrated that every purine (adenosine, guanosine, and inosine) and pyrimidine (cytidine and thymidine) nucleoside containing the 2'',3''-dideoxyribose configuration, when evaluated against human immunodeficiency virus (HIV) in vitro, significantly suppressed both the infectivity and the cytopathic effect of the virus, with 2'',3''-dideoxycytidine (ddCyd) being the most potent of the series (total antiviral protection at 0.5-1.0 .mu.M). We have compared three factors likely to be of significance in determining the pharmacological activity of these compounds, i.e., (i) their abilities to influence pool sizes of physiological deoxynucleoside-5''-triphosphates, (ii) their capacity to generate the corresponding 2'',3''-dideoxynucleoside-5''-triphosphates, and (iii) the effectiveness of these nucleoside-5''-triphosphates as inhibitors of HIV reverse transcriptase. In MOLT-4 cells (a human T cell line), ddCyd was the compound most efficiently converted to its 5''-triphosphate, whereas 2'',3''-dideoxyguanosine and 2'',3''-dideoxythymidine were the compounds least efficiently converted, generating levels of their corresponding 5''-triphosphates less than 0.1% of that seen with ddCyd when these nucleosides were compared on an equimolar basis (5 .mu.M). The 3''-azido analogue 2'',3''-dideoxythymidine fell intermediate between these two extremes. As inhibitors of HIV reverse transcriptase, however, all the 5''-triphosphates, with the exception of 2'',3''-dideoxyinosine-5''-triphosphate fell within a narrow range of activity (Ki, 0.10-0.26 .mu.M), affinities some 40-60-fold greater than those of the corresponding physiological 2''-deoxynucleoside-5''-triphosphates. Significant alterations in pool sizes of physiological 2''-deoxynucleoside-5''-triphosphates were not observed at pharmacologically effective drug levels. The relative ability of 2'',3''-dideoxynucleosides to generate 5''-triphosphates intercellularly thus correlates much more closely than do the other two factors examined, in capacity to block HIV replication. These studies support the conclusion that, for purposes of design of new compounds of this general class, factors influencing efficiency of nucleotide formation and degradation (e.g., membrane transport mechanisms, affinities for nucleoside kinases and for nucleotide kinases and phosphatases) may be of equal or even greater importance than differences in the relative abilities of the resultant 2'',3''-dideoxynucleoside-5''-triphosphates to inhibit the viral reverse transcriptase.