Chemotherapy of AIDS (acquired immunodeficiency syndrome) is one of the most challenging scientific projects upon which much attention is currently focused. Although AZT (3'-azido-3'-deoxythymidine) is available as the sole compound formally approved for clinical use, its serious side effects, suppression of bone marrow cell growth, combined with the appearance of AZT-resistant HIV (human immunodeficiency virus) variants, give an incentive to search for other promising AIDS drug candidates having a higher selectivity against HIV. We have recently synthesized some 6-substituted acyclouridines on the basis of our own strategy using LDA (lithium diisopropylamide) which has been proved to be a general method for the modification of the base moiety. 1-[(2-Hydroxyethoxy)methyl]-6-(phenylthio)thymine (1, HEPT) was found to be a specific anti-HIV-1 agent. In contrast to 2',3'-dideoxyribonucleoside analogues, HEPT has no inhibitory activity against other retroviruses such as SIVMc (simian immunodeficiency virus), SRV (simian AIDS-related virus), MSV (murine Moloney sarcoma virus), and even against HIV-2. The '5''-triphosphate of HEPT however does not inhibit HIV-1 reverse transcriptase at concentrations much higher than the EC50 of HEPT for HIV-1 replication in MT-4 cells. Consequently, the presence of an hydroxyl group in HEPT may not be necessary for its anti-HIV-1 activity. In addition HEPT does not compete with [3H-Me] thymidine for phosphorylation by thymidine kinase derived from MT-4 cells. Our recent studies have suggested that modification of the base moiety of HEPT by substituting at C-2 or C-5 might enhance activity. In the present communication, we describe the synthesis and anti-HIV-1 activity of HEPT analogues, the acyclo portions of which were altered so they could not be phosphorylated. Furthermore, we show a dramatic improvement in the activity by replacing the 5-substituent with an ethyl group. Initially, preparation of O-alkyl derivatives of 1 was carried out. Lithiation of 1-[(2-methoxyethoxy)methyl]thymine with LDA and subsequent reaction with (PhS)2 gave the O-methyl derivative 2. The O-benzyl derivative 3 was prepared by selective alkylation of 1. The anti-HIV-1 activity and cytotoxicity of these compounds are listed in Table I. Although the presence of the bulky O-benzyl group results in an almost complete loss of activity, the O-methyl derivative 2 appeared to retain the activity accompanied by a slight increase in cytotoxicity. As a result of these experiments, we then synthesized some '5''-deoxy analogues of HEPT from the thymine derivatives (4-7). The LDA lithiation reaction was found to work even when the starting material included a halogeno or azido functionality. From the EC50 values of the resulting 6-phenylthio analogues (8-11) in Table I, it can be concluded that the hydroxyl function of HEPT does not contribute to its anti-HIV-1 activity. Another possible conclusion to be drawn from the data is that the value of the EC50 may correlate with the size of the substituent in the acyclo portion. Thus, this hypothesis was examined by the introduction of an aralkyl side chain. Compound 12 synthesized in this context was found to be much more active than 8 which corresponds to the genuine deoxy analogue of HEPT. However, due to the increased toxicity, the selectivity index (SI) of 12 remained much the same as that of 8. In the course of our studies, aimed at increasing the anti-HIV-1 activity of HEPT, we recently found that replacement of the 5-methyl group in HEPT with an ethyl group enhanced activity to a greater extent, as illustrated by the EC50, CC50, and SI of 13. This combined with the present results led us to synthesize 14 and 15 by the method described. Their activity shown in Table I, together with that of HEPT, clearly indicates that the initial activity of HEPT has been dramatically improved at this stage. In particular, the EC50 value of 15 is comparable to AZT. It should be emphasized that, in terms of CC50, both 5-ethyl analogues are much less cytotoxic than is AZT. When the activity was examined with some AZT-resistant HIV-1 strains, both compounds were equally effective. Their anti-HIV-1 activities were further confirmed by monitoring viral antigen expression in CEM cells as shown in Figure 1. HIV-1 antigen expression was almost completely suppressed at concentrations of 40-800 nM. It should be mentioned that the analogues synthesized in this study are uniformly inactive against HIV-2 following the original specificity of HEPT. Our recent observation using reverse transcriptase (RT) indicates that 14 and 15 were potent inhibitors of HIV-1 RT, irrespective of the source of enzymes. However, reflecting their lack of activity against HIV-2 in cell cultures, these compounds did not prove inhibitory to HIV-2 RT. These results suggest their mode of action against HIV-1 at RT is clearly distinct from that of AZT. Another point to be emphasized is the effect of the compounds on bone marrow cell proliferation. In our preliminary experiments, AZT suppressed approximately 50% of the colony formation of murine bone marrow progenitor cells at concentration of 1 μM, whereas no such inhibition was observed with 14 and 15 even at 10 μM. In conclusion, the present study demonstrates the anti-HIV-1 activity originally found in HEPT can be retained, or improved in certain cases, by removing the hydroxyl group. The 5-ethyl-'5''-deoxy analogues (14 and 15) obtained by further modification at the base moiety are much less toxic than AZT and, in particular, 15 is almost as active as AZT. Considering their effectiveness against AZT-resistant strains of HIV-1 and also their lower toxicity to bone marrow cells, we believe these compounds may constitute highly promising candidates for the chemotherapy of AIDS.