Human cytomegalovirus resistance to deoxyribosylindole nucleosides maps to a transversion mutation in the terminase subunit-encoding gene UL89

Abstract

Human cytomegalovirus (HCMV) infection can cause severe illnesses, including encephalopathy and mental retardation, in immunocompromised and immunologically immature patients. Current pharmacotherapies for treating systemic HCMV infections include ganciclovir, cidofovir, and foscarnet. However, long-term administration of these agents can result in serious adverse effects (myelosuppression and/or nephrotoxicity) and the development of viral strains with reduced susceptibility to drugs. The deoxyribosylindole (indole) nucleosides demonstrate a 20-fold greater activity in vitro (the drug concentration at which 50% of the number of plaques was reduced with the presence of drug compared to the number in the absence of drug [EC50] = 0.34 μM) than ganciclovir (EC50 = 7.4 μM) without any observed increase in cytotoxicity. Based on structural similarity to the benzimidazole nucleosides, we hypothesize that the indole nucleosides target the HCMV terminase, an enzyme responsible for packaging viral DNA into capsids and cleaving the DNA into genome-length units. To test this hypothesis, an indole nucleoside-resistant HCMV strain was isolated, the open reading frames of the genes that encode the viral terminase were sequenced, and a G766C mutation in exon 1 of UL89 was identified; this mutation resulted in an E256Q change in the amino acid sequence of the corresponding protein. An HCMV wild-type strain, engineered with this mutation to confirm resistance, demonstrated an 18-fold decrease in susceptibility to the indole nucleosides (EC50 = 3.1 ± 0.7 μM) compared to that of wild-type virus (EC50 = 0.17 ± 0.04 μM). Interestingly, this mutation did not confer resistance to the benzimidazole nucleosides (EC50 for wild-type HCMV = 0.25 ± 0.04 μM, EC50 for HCMV pUL89 E256Q = 0.23 ± 0.04 μM). We conclude, therefore, that the G766C mutation that results in the E256Q substitution is unique for indole nucleoside resistance and distinct from previously discovered substitutions that confer both indole and benzimidazole nucleoside resistance (D344E and A355T).

FIG 1

Structures of 2-bromo-5,6-dichloro-1-(β-d-ribofuranosyl)benzimidazole (BDCRB) and 3-acetyl-2-bromo-5,6-dichloro-1-(2-deoxy-β-d-ribofuranosyl)indole (UMJD 1896, indole nucleoside 1896).

FIG 2

Initial characterization of an indole nucleoside 1896-resistant HCMV isolate (1896^r). HFFs infected with an indole nucleoside 1896-resistant HCMV isolate (open symbols) were subjected to increasing concentrations of either BDCRB (dashed lines, ■ and □) or indole nucleoside 1896 (solid lines, ▲ and Δ) and compared to HFFs infected with wild-type HCMV (Towne; closed symbols) subjected to the same concentrations of drug. The y axis represents the percentage of plaques compared to the amount for a no-drug control. The values represent the mean ± standard deviation from at least three experiments. *, P < 0.001, in which the replication of 1896^r was significantly different from that of wild-type virus (for indole nucleoside 1896 only; no statistically significant difference was found for BDCRB).

FIG 3

Decreased susceptibility of HCMV pUL89 E256Q to indole nucleoside 1896. HFFs infected with a wild-type HCMV strain (AD169) engineered with the UL89 G766C point mutation (open symbols) were subjected to increasing concentrations of either BDCRB (dashed lines, ■ and □) or indole nucleoside 1896 (solid lines, ▲ and Δ) and compared to HFFs infected with wild-type HCMV (AD169; closed symbols) subjected to the same concentrations of drug. The y axis represents the percentage of plaques compared to the amount for a no-drug control. The values represent the mean ± standard deviation from at least four experiments. *P < 0.0001, in which the replication of HCMV E256Q was significantly different from that of wild-type virus (for indole nucleoside 1896 only; no statistically significant difference was found for BDCRB).