Our studies have shown that the acyclic nucleotide analogues PMEA and HPMPC are able to penetrate into cells and are then activated to mono- and diphosphate derivatives. The latter correspond to triphosphate analogues and presumably serve an important role in the biological activity exerted by these antiviral agents. In support of this idea, the inhibitory effect of PMEApp on HIV reverse transcriptase has been demonstrated with both RNA and DNA template-primer systems. Further studies will be undertaken to determine the effect of HPMPCpp on viral DNA polymerases. Whereas the metabolism of PMEA in CEM cells gives rise to only PMEAp and PMEApp, additional metabolites were obtained in MRC-5 cells; the identity of these metabolites remains to be determined. In the case of HPMPC, a third metabolite was obtained in addition to HPMPCp and HPMPCpp, which has been tentatively assigned as a phosphate-choline adduct by analogy with activation of cytosine-based nucleoside derivatives. The metabolism of HPMPC was unchanged between uninfected and infected cells, indicating that viral enzymes are not necessary for the activation of HPMPC. The long intracellular half-lives of the HPMPC metabolites may have implications for the antiviral efficacy of this compound. The persistence of activated metabolites suggests that infrequent dosing may be possible due to a prolonged antiviral effect. Our results on the effectiveness of infrequent dosing schedules with HPMPC in the treatment of HSV 2 infections in mice support this hypothesis. It is also possible that HPMPCp-choline may serve as a reservoir for HPMPC and therefore for the presumed active metabolite HPMPCpp.