The inhibitory action of 9-(1,3-dihydroxy-2-propoxymethyl)guanine on the replication of human cytomegalovirus was studied. Three laboratory strains (AD-169, Towne, and Davis) and three early passage (<10) clinical isolates were all inhibited in yield inhibition assays. In cultures infected with AD-169, virus yields could be inhibited if the drug was added as late as 3 days after the replication cycle had begun. The effects of the drug were' fully reversible during the first 4 days of the viral replication cycle. Viral infectivity and viral DNA synthesis were reduced more than viral protein synthesis. Synergistic antiviral effects were observed with betacysteine, and to a lesser extent, with beta-serine recombinant interferons, but only over a narrow range of dose combinations.The novel guanosine analog, 9-(1,3-dihydroxy-2-propoxymethyl)guanine (DHPG), is active in vitro against human cytomegalovirus (CMV) (1, 13, 24). Acycloguanosine and other nucleoside analogs have limited activity against human CMV (3,5,12,16,19,23 various concentrations was then added to the tissue cultures. When a multiplicity of infection (MOI) of 0.1 was used, the virus was allow'ed to replicate for 5 days; replication was continued for multiple cycles (10 to 12 days) if the MOI was <0.1. Infected cultures were harvested by scraping the cells into the supernatant, pooling duplicate wells, sonicating to disrupt the cells, and then freezing the cells at -70°C. For some experiments, the supernatant and intracellular virus were harvested separately. Samples were assayed for PFU of virus on HEL cells (21). Inhibition was determined by comparing the yields from drug-treated virus-infected cultures with those from untreated virus-infected cultures and was expressed as a percentage of control yields. For plaque reduction assays, the drug was incorporated into the semisolid overlay. For experiments in which the effects of IFNs and DHPG in combination were tested, cells were first treated overnight with IFN and then infected with CMV. Both drugs were included in the maintenance medium, either DMEM or DMEM with 1% agarose, during virus replication. To evaluate drug interactions, we calculated a combination index (CI) by the formula derived by Spector et al. (28): CI = ln(A) + ln(B) -ln(AB) -ln(VC), where A and B are the mean virus yields in the presence of antiviral agents A and B, respectively, AB is the virus yield in cultures treated with both agents, and VC is the virus yield in untreated controls. The CI ± standard error (SE) was calculated for each drug combination with data from independent experiments. When the CI was within 2 SEs of 0, the combined drug effects were additive. When the CI was greater than ±2 SE, the effects were synergistic; when the CI was less than ±2 SE, the effects were antagonistic. For plaque reduction assays, the drug interactions were assessed by the formula [(A)(B)]i [(AB)(VC)]. When the numerical result was >1, the effects were synergistic; when the numerical result was equal to 1, the effects were additive; and wh...