Abstract
Treatment options for human cytomegalovirus (CMV) remain limited and are associated with significant adverse effects and the selection of resistant CMV strains in transplant recipients and congenitally infected infants. Although most approved drugs target and inhibit the CMV DNA polymerase, additional agents with distinct mechanisms of action are needed for the treatment and prevention of CMV. In a large high throughput screen using our CMV-luciferase reporter Towne, we identified several unique inhibitors of CMV replication. Here, we synthesize and test in vitro 13 analogs of the original NCGC2955 hit (1). Analogs with no activity against the CMV-luciferase at 10 µM and 30 µM (2-6, 10-14) were removed from further analysis. Three analogs (7-9) inhibited CMV replication in infected human foreskin fibroblasts. The EC(50) of (1) was 1.7 ± 0.6 µM and 1.99 ± 0.15 µM, based on luciferase and plaque assay, respectively. Compounds 7, 8, and 9 showed similar activities: the EC(50) values of 7 were 0.21 ± 0.06 µM (luciferase) and 0.55 ± 0.06 (plaque), of 8: 0.28 ± 0.06 µM and 0.42 ± 0.07, and of 9: 0.30 ± 0.05 µM (luciferase) and 0.35 ± 0.07 (plaque). The CC(50) for 7, 8, and 9 in non-infected human foreskin fibroblasts was > 500µM, yielding a selectivity index of >1500. Compounds 1, 7, and 8 were also tested in CMV-infected primary human hepatocytes and showed a dose-response against CMV by luciferase activity and viral protein expression. None of the active compounds inhibited herpes simplex virus 1 or 2. Compounds 7 and 8 inhibited mouse CMV replication in vitro. Both inhibited CMV at late stages of replication; 7 reduced virus yield at all late time points, although not to the same degree as letermovir. Finally, the activity of analog 8 was additive with newly identified CMV inhibitors (MLS8969, NFU1827, MSL8554, and MSL8091) and with ganciclovir. Further structural activity development should provide promising anti-CMV agents for use in clinical studies.