Drug Susceptibility and Replicative Capacity of Multidrug-Resistant Recombinant Human Cytomegalovirus Harboring Mutations in UL56 and UL54 Genes

ABSTRACT

Letermovir is an investigational antiviral agent with a novel mechanism of action involving the viral terminase (pUL56). We evaluated the impact of the V236M mutation in the UL56 gene alone and in combination with the E756K mutation in the UL54 gene on drug susceptibility and viral replicative capacity of recombinant human cytomegalovirus. The double mutant exhibited at least borderline resistance to all antivirals tested (ganciclovir, foscarnet, cidofovir, brincidofovir, and letermovir) and replicated less efficiently than the wild-type virus in vitro.

TEXT

Morbidity and mortality associated with human cytomegalovirus (HCMV) infections in immunocompromised patients remain important issues despite existing prevention and treatment strategies (1). All currently approved antiviral agents (ganciclovir [GCV], foscarnet [FOS], and cidofovir [CDV]) target the viral DNA polymerase (2). The main concerns associated with these antivirals are related to their side effects (e.g., nephrotoxicity, myelosuppression) and the emergence of drug-resistant isolates. Moreover, because of their common viral target, cross-resistance to all drugs may occur, resulting in limited therapeutic options for clinicians (3). For instance, the E756K mutation in DNA polymerase (encoded by the UL54 gene) was reported to confer cross-resistance to GCV, FOS, and CDV and to reduce the viral replicative capacity (4). The orally bioavailable lipid ester prodrug of CDV (brincidofovir [BCV]) may not have the dose-limiting toxicity of the parent drug but does not constitute an interesting alternative for multidrug-resistant isolates from immunocompromised patients (5). Letermovir (LMV), a new antiviral agent under clinical development, has a novel mode of action in that it targets the viral terminase subunit pUL56 and interferes with DNA maturation and packaging (6). In this respect, mutations that map in the UL56 gene have been elicited in vitro and shown to confer low- to high-grade resistance to LMV (7). The V236M mutation was initially selected in vitro and was detected in an isolate obtained from a patient who was enrolled in a phase 2b dose-range-finding prophylactic study in hematopoietic stem cell transplant (HSCT) recipients who received suboptimal LMV dosing (8). In a phase 3 prophylaxis trial, LMV was shown to be highly effective in preventing HCMV infection in HSCT recipients (9). Such LMV-based prophylaxis may be of particular interest in this setting, to avoid the neutropenia associated with the use of GCV (10). In this study, we analyzed the impact of the V236M mutation in the UL56 gene alone and in combination with the E756K mutation in the UL54 gene on the drug resistance phenotype and growth kinetics of recombinant viruses in order to anticipate potential mutants emerging during sequential antiviral therapy.

Human lung fibroblasts (MRC-5, ATCC CCL-171) were maintained in minimal essential medium (MEM) (Gibco/Invitrogen) supplemented with 10% fetal bovine serum (FBS) (Wisent). We used a bacterial artificial chromosome (BAC) plasmid containing the wild-type HCMV genome derived from the laboratory strain AD169 (pHB5 [11]) in which we integrated the Gaussia luciferase (GLuc) reporter gene under the control of the CMV major immediate early promoter (pHB5-GLuc [12]). The different mutations were introduced alone or sequentially in the UL56 and UL54 gene(s) of pHB5-GLuc by en passant mutagenesis as previously described (12, 13). MRC-5 cells were transfected with wild-type (WT) and mutant pHB5-GLuc (bacmid) viruses by electroporation, and the reconstituted recombinant viruses were collected after 10 to 14 days. Introduction of the desired mutation(s) was confirmed by sequencing the entire UL56 and UL54 genes of mutated pHB5-GLuc and reconstituted viruses.

Susceptibility testing of WT and mutant recombinant viruses to antiviral agents was performed in MRC-5 cells by the GLuc reporter-based assay (12). Briefly, cells at 75% confluence were infected with recombinant HCMV strains at a multiplicity of infection (MOI) of 0.001. After 90 minutes, serial 2-fold dilutions of each antiviral compound (GCV, FOS, CDV, BCV, and LMV) in MEM plus 2% FBS were added to triplicate wells. The supernatant was then collected on day 6 postinfection to measure the GLuc activity by use of the coelenterazine substrate (12). The concentration of each antiviral compound that reduced the GLuc activity by 50% (EC₅₀) was determined. Drug resistance was defined as EC₅₀ ≥3.0-fold compared with that of the WT virus, whereas a ratio between 1.8 and 3.0 was considered a borderline level of resistance.

The replicative capacity of WT and mutant recombinant viruses was determined by the GLuc reporter-based assay and a conventional plaque assay (12). MRC-5 cells at 75% confluence were infected with the different recombinant viruses at an MOI of 0.001. The GLuc activity was assayed in cell culture supernatant collected daily for 8 days. For the plaque assay, cell culture supernatant was collected daily in two separate wells. The extracellular virus yield was evaluated by infecting human foreskin fibroblasts with serial 2-fold dilutions of each supernatant sample in triplicate. Infected cells were incubated for 10 days in MEM plus 2% FBS containing 0.4% SeaPlaque agarose (Lonza, Rockland, ME), fixed, and stained, and the PFU in each well were counted.

Table 1 shows that the EC₅₀s of GCV, FOS, and CDV against the WT recombinant virus were 1.0 ± 0.4, 53.3 ± 9.6, and 0.8 ± 0.1 μM, respectively; whereas LMV and BCV were the most active compounds, with EC₅₀s of 3.2 ± 0.7 and 0.5 ± 0.1 nM, respectively. The recombinant viruses harboring the V236M mutation in the UL56 gene demonstrated a high level of resistance to LMV (mean, 27.0-fold increase) and were susceptible to all other antiviral agents. As reported in the literature, the recombinant virus harboring mutation E756K in the UL54 gene was resistant to GCV and FOS (7.1- and 3.5-fold increases, respectively) and had borderline levels of resistance to CDV and BCV (1.8- and 2.8-fold increases, respectively) (4, 12, 14, 15). This strain was susceptible to LMV (1.4-fold increase). The recombinant viruses with V236M and E756K mutations in UL56 and UL54 genes, respectively, were resistant to GCV and LMV (mean, 3.4- and 34.0-fold increases, respectively) and had borderline levels of resistance to FOS, CDV, and BCV (mean, 2.5-, 1.9-, and 2.0-fold increases, respectively).

TABLE 1.

Susceptibilities of wild-type and mutant recombinant viruses to antiviral agents by the Gaussia luciferase assay

Mutation (gene) and clone no.a	EC50 (mean ± SD [fold change]) ofb:
	Ganciclovir (μM)	Foscarnet (μM)	Cidofovir (μM)	Brincidofovir (nM)	Letermovir (nM)
Wild type (n = 5)	1.0 ± 0.4 (1.0)	53.3 ± 9.6 (1.0)	0.8 ± 0.1 (1.0)c	0.5 ± 0.1 (1.0)d	3.2 ± 0.7 (1.0)
V236M (UL56) 1 (n = 4)	1.4 ± 0.5 (1.4)	32.0 ± 4.2 (0.6)e	0.7 ± 0.2 (0.9)e	0.4 ± 0.1 (0.9)	100.0 ± 20.0 (31.0)
V236M (UL56) 2 (n = 1)	0.8 (0.8)	32.1 (0.6)	0.7 (0.9)	0.7 (1.4)	70.0 (22.0)
E756K (UL54) 1 (n = 3)	7.6 ± 0.9 (7.1)	188.0 ± 94.0 (3.5)	1.3 ± 0.2 (1.8)	1.3 ± 0.1 (2.8)	4.7 ± 0.4 (1.4)
V236M (UL56)/E756K (UL54) 1 (n = 3)	4.1 ± 0.9 (3.8)	148.2 ± 5.2 (2.8)	1.4 ± 0.3 (1.8)	0.9 ± 0.1 (2.1)d	110.0 ± 20.0 (34.0)
V236M (UL56)/E756K (UL54) 2 (n = 1)	3.2 (3.0)	117.1 (2.2)	1.5 (1.9)	1.0 (2.0)	110.0 (34.0)

^an, number of replicate experiments.

^bEC₅₀, concentration of antiviral drug that reduces the Gaussia luciferase activity by 50%. Fold change represents the ratio of EC₅₀ of a mutant recombinant virus to that of wild-type virus. A fold change of ≥3.0 corresponds to antiviral resistance, whereas a fold change of between 1.8 and 3.0 is considered a borderline level of resistance.

^cn = 8.

^dn = 4.

^en = 3.

It was previously reported that a series of mutations in the UL56 gene did not alter the growth kinetics of recombinant viruses when assessed by plaque assay (6, 7, 16, 17) or by secreted alkaline phosphatase reporter-based assay (18). Figure 1A shows that the V236M mutation, which is not located in a conserved region of the UL56 gene, significantly reduced the viral replicative capacity, assessed by the GLuc assay, by 0.32 log (P < 0.01), 0.15 log (P < 0.05), 0.17 log (P < 0.001), and 0.14 log (P < 0.001) on days 5, 6, 7, and 8 postinfection, respectively, compared with that of the WT virus. As previously reported, recombinant virus harboring the E756K mutation in the UL54 gene had decreased viral growth kinetics by 0.24 log (P < 0.05), 0.18 log (P < 0.01), and 0.15 log (P < 0.001) on days 5, 6 and 8 postinfection, respectively, compared with that of the WT virus (4). Combination of the V236M and E756K mutations did not influence the growth of the recombinant virus compared with each mutant alone (reduction, 0.15 log [P < 0.01], 0.17 log [P < 0.001], and 0.11 log [P < 0.001] compared with that of the WT virus on days 6, 7, and 8 postinfection, respectively). These results were confirmed with a conventional plaque assay, as shown in Fig. 1B.

FIG 1.

Replicative capacity of wild-type and recombinant viruses harboring the V236M mutation in the UL56 gene and the E756K mutation in the UL54 gene alone or in combination as determined by the Gaussia luciferase (GLuc) reporter-based assay (A) and by a conventional plaque assay (B). MRC-5 cells were infected with the different recombinant viruses at an MOI of 0.001, and the GLuc activity was measured in cell culture supernatants sampled daily for 8 days. Supernatants were also collected daily and diluted to infect new human foreskin fibroblasts to determine the number of viral plaques. Results represent the mean ± SD of sextuplicate determinations and are representative of two independent experiments (GLuc assay) and one experiment (plaque assay). RLU, relative light unit. *, P < 0.05; **, P < 0.01; ***, P < 0.001, compared with wild-type virus by one-way analysis of variance with Dunnett's multiple-comparison test (GraphPad Prism version 5.00).

According to the different mechanisms of action of antiviral drugs, the recombinant virus harboring the V236M mutation conferring resistance to LMV was susceptible to GCV, FOS, CDV, and BCV. In a similar manner, LMV was effective against the E756K recombinant virus that was resistant to GCV, FOS, CDV, and BCV, suggesting that this drug has potential for the treatment of patients with HCMV disease that does not respond to all currently available antiviral agents. LMV can induce low- to high-grade levels of resistance with partially reduced viral fitness. This suggests the possibility of mixed or evolving virus subpopulations that may have a selective advantage in the presence of LMV. Thus, close monitoring of viral isolates obtained from patients treated with this antiviral should be performed (18). In conclusion, our mutant recombinant HCMV mimics an isolate that may be sequentially selected by LMV prophylaxis followed by treatment with currently approved drugs.