Antiviral activity of T-705 against paramyxoviruses in vitro.
The inhibitory effect of T-705 on infection was evaluated for RSV-GFP, PIV-3, NDV, AMPV-C, MeV-Edm-GFP, and prototype strains of the 4 genotypes of HMPV (rHMPV NL/00/01-GFP [A1], HMPV NL/00/17 [A2], rHMPV NL/99/01-GFP [B1], and NL/94/01 [B2]). The addition of high concentrations of T-705 (up to 1,600 μM) 24 h prior to inoculation had no or little effect on cell viability as measured by cytotoxicity assay. Microscopic examination of the cells revealed that with the addition of increasing T-705 concentrations, the number of infected cells decreased for all viruses tested (
Fig. 1). For MeV-Edm-GFP, addition of 62.5 μM T-705 resulted in the absence of typical MeV-Edm-GFP-induced plaques. For most of the other viruses, a decrease in infection efficiency was observed starting at 125 μM T-705. Only for RSV-GFP were higher concentrations of the compound necessary to decrease infection of HEp-2 cells (
Fig. 1).
In the next experiment, as a preparation for the
in vivo experiments, we set out to mimic the effect of T-705 in preexposure and postexposure settings of virus infection
in vitro. Cells were treated at 3 time points: 24 h prior to, simultaneous with, and 24 h after inoculation. Influenza virus and HSV were used as sensitive and insensitive controls, respectively. Furuta et al. (
19) reported HSV to be insensitive to T-705 treatment, with EC
50 values of >625 μM. In our study, we observed some virus yield reduction when T-705 was administered 24 h prior to and simultaneously with infection, with EC
90 values of 230 and 540 μM, respectively (
Table 1). As these values are significantly higher than those obtained for the sensitive influenza virus (
P = 0.0286 for both treatment prior to and treatment simultaneous with inoculation; for the Mann-Whitney test,
P < 0.05) and treatment after infection with HSV had no inhibitory effect on virus release (EC
90 of >1,500 μM), our data confirm the insensitivity of HSV to T-705 treatment (
Fig. 2). Using plaque reduction assays, Furuta et al. reported EC
50 values for influenza virus ranging from 0.083 to 2.9 μM (
19). In the present study, we obtained EC
90 values of 1.5 μM and 2.0 μM for treatment prior to and simultaneous with infection, respectively. Taking into account that the employed assays and data were different, i.e., plaque reduction assays versus yield reduction assays and EC
50 versus EC
90, these values are in similar micromolar ranges, which confirms the activity of our batch of T-705 against influenza virus. However, influenza virus was significantly less sensitive to treatment after infection than to treatment prior to or simultaneous with infection (
P = 0.03 in both cases; for the Mann-Whitney test,
P < 0.05) (
Table 1;
Fig. 2).
Of all the paramyxoviruses tested, MeV-Edm-GFP was the most sensitive, with EC
90 values of 8.6, 9.7, and 13 μM for treatment prior to, simultaneous with, and after inoculation, respectively (
Table 1). For treatment prior to or simultaneous with infection, the values are significantly higher than those for influenza virus (
P = 0.03 in both cases; for the Mann-Whitney test,
P < 0.05), although the EC
90 values for MeV-Edm-GFP are still in the micromolar range. MeV-Edm-GFP and influenza virus were similarly sensitive to T-705 treatment 24 h after infection, with EC
90 values of 57 and 13 μM for influenza virus and MeV-Edm-GFP, respectively (
P = 0.2; for the Mann-Whitney test,
P < 0.05). Together, these data show that MeV-Edm-GFP is sensitive to treatment with T-705 over a micromolar range similar to that effective against influenza virus (
Fig. 2).
The four genotypes of HMPV were sensitive to treatment in the same micromolar range as that for MeV-Edm-GFP, with EC
90 values ranging from 11 to 26 μM, 12 to 34 μM, and 22 to 43 μM for treatment prior to, simultaneous with, and after infection, respectively (
Table 1). In general, serotype A strains were slightly less sensitive to T-705 than serotype B strains. Although no significant differences were observed between the four strains for treatment before infection (
P = 0.09; for ANOVA,
P < 0.05) or after infection (
P = 0.07; for ANOVA,
P < 0.05), significant differences between the four HMPVs were observed for treatment simultaneous with infection (
P = 0.003; for ANOVA,
P < 0.05) (
Fig. 2). This was mainly due to the significantly higher EC
90 value of NL/00/17 (34 μM) than those of the type B viruses (15 and 12 μM for NL/99/01 and NL/94/01, respectively) for this time of treatment (
P = 0.0286 in both cases; for the Mann-Whitney test,
P < 0.05). HMPV strains NL/00/01, NL/99/01, and NL/94/01 were as sensitive as MeV-Edm-GFP to treatment prior to inoculation (
P = 0.34, 0.48, and 0.52, respectively; for the Mann-Whitney test,
P < 0.05). Only NL/00/17 was less sensitive than MeV-Edm-GFP to this treatment (
P = 0.03; for the Mann-Whitney test,
P < 0.05). All four HMPVs, as well as MeV-Edm-GFP, were as sensitive to treatment after infection as influenza virus (
P = 0.14; for ANOVA,
P < 0.05) (
Fig. 2). These results obtained for four genotypes of HMPV demonstrate that this virus is sensitive to treatment over a micromolar range similar to that effective against influenza virus and MeV-Edm-GFP.
T-705 also demonstrated an inhibitory effect on virus release for APMV-C, NDV, PIV-3, and RSV (
Table 1). No significant differences were observed between these four viruses for the three treatments (
P = 0.35, 0.08, and 0.37 for treatment prior to, simultaneous with, and after infection, respectively; for ANOVA,
P < 0.05). However, the EC
90 values for these four viruses were all significantly higher than those for HMPV strains NL/00/01 (
P = 0.01, 0.002, and 0.02 for treatment prior to, simultaneous with, and after infection, respectively; for ANOVA,
P < 0.05), NL/99/01 (
P = 0.002, 0.0003, and 0.02, respectively), and NL/94/01 (
P = 0.0015, <0.0001, and 0.01, respectively) but not for HMPV strain NL/00/17 (
P = 0.07, 0.02, and 0.08, respectively).
Together, these data show that T-705 has an antiviral effect, in the micromolar range, against all the paramyxoviruses tested, with higher EC90 values for AMPV-C, NDV, PIV-3, and RSV than for HMPV and MeV-Edm-GFP.
In vivo antiviral activity of T-705 against HMPV.
To test the in vivo activity of T-705, an established animal model for HMPV infections was employed. Syrian golden hamsters were treated with 50, 100, 150, 200, or 400 mg/kg/day for 4 days, starting 24 h before nasal inoculation with 106 TCID50 HMPV strain NL/00/01 (type A1). During this experiment, the animals did not show any signs of illness or weight loss.
Real-time RT-PCR assays conducted on throat swabs collected on days 3 and 4 after inoculation and qRT-PCR assays conducted on nasal turbinate and lung samples collected on day 4 after inoculation revealed the presence of viral genomes in these samples at all given concentrations (
Fig. 3A and
B, top panels). However, the numbers of viral genomes in the throat swabs and nasal turbinates decreased significantly with administration of increasing concentrations of T-705: starting at day 3 for the dose of 100 mg/kg/day (
P = 0.026; for the Mann-Whitney test,
P < 0.05) and at day 4 for the dose of 200 mg/kg/day (
P = 0.01; for the Mann-Whitney test,
P < 0.05) for the throat swabs and at a dose of 200 mg/kg/day (
P = 0.026; for the Mann-Whitney,
P < 0.05) for the nasal turbinates. T-705 treatment did not result in a significant decrease in viral genome titers in the lungs (
P = 0.2; for ANOVA,
P < 0.05), although the titers did decrease with increasing concentrations of T-705 (
Fig. 3B). Virus titration of the throat swab samples revealed levels of infectious virus below the limit of detection in 3 (day 3) or 4 (day 4) of the 6 animals treated at a dose of 100 mg/kg/day. In general, infectious virus titers decreased significantly in the throat swabs starting at a dose of 100 mg/kg/day (
P = 0.0022; for the Mann-Whitney test,
P < 0.05) (
Fig. 3A, lower panels) and in the nasal turbinates and lungs at 200 mg/kg/day (
P = 0.015 and
P = 0.022 for nasal turbinates and lungs, respectively; for the Mann-Whitney test,
P < 0.05) (
Fig. 3B, lower panels). Most importantly, at a dose of 200 mg/kg/day, 3 of 6 animals displayed levels of infectious virus in the lungs below the limit of detection, and at a dose of 400 mg/kg/day, none of the animals had levels of infectious virus in the lungs above the limit of detection (
Fig. 3B, lower right panel).
Although replicating virus could not be detected in the lungs of animals treated with a dose of 400 mg/kg/day, viral genome titers declined only minimally. These findings of decreased infectious virus titers in samples from treated animals and only limited decreases in viral genome titers resemble the findings reported for influenza virus. In the influenza virus study, the infectious virus load in treated samples decreased disproportionately to the RNA copy number. It was demonstrated that T-705 induced hypermutation of the viral genome, which explained the decrease in viral titers with equal titers of viral genome copies in samples treated with T-705 (
24). To investigate this possibility for the mechanism of action of T-705 against HMPV, samples (nasal turbinates and lungs) from untreated animals and high-dose-treated animals were subjected to next-generation sequencing (
Table 2). To this end, the region from nt 6748 to 12340 of the viral genome (the L open reading frame), for two animals each (untreated animals H013 and H014 and treated animals H044 and H045), was subjected to 454 deep sequencing using 3 overlapping PCR fragments. Overall, this region was covered by 6,892 to 17,447 reads (
Fig. 4A), and analysis of single nucleotide polymorphisms (SNPs) revealed the presence of 22 SNPs in the viral genomes isolated from treated animals that were also present in the viral genomes retrieved from untreated hamsters (see Table S1 in the supplemental material). This analysis also revealed the presence of approximately 25 reads with T-to-C hypermutation, but these were detected only for the sample isolated from the nasal turbinate of one treated hamster (animal H043) (
Fig. 4A, blue circles around nt 8000). These hypermutated reads were absent for the lung sample from the same hamster and the samples from the other treated hamster (H044). These hypermutated reads most likely represent (hypermutated) defective interfering virus particles (DIs) that occur during HMPV infection, as previously reported (
44), and based on their absence in the other (treated) samples, they were not a result of T-705 treatment. The T-C hypermutation detected in these reads was included in the calculation of the number of SNPs for
Fig. 4C, which explains the wide error bars.
Additional SNPs were detected that were present in untreated samples and absent in treated samples or the other way around. In the nasal turbinate samples from untreated animals, 9 (H013) and 10 (H014) SNPs were detected that were absent in samples from treated animals, and in the lung samples, 4 (H013) and 5 (H014) SNPs were detected that were absent in samples from treated animals; all of these were present in <10% of the reads (see Table S1 in the supplemental material). In the nasal turbinate samples from treated animals, 4 SNPs (in both H043 and H044) were detected that were absent in the untreated samples, and in the lung samples, 16 (H043) and 4 (H044) SNPs were detected that were absent in the untreated samples. Overall, the numbers of SNPs detected in this part of the genome did not differ significantly between the treated and untreated samples, even when the 25 hypermutated reads detected in the nasal turbinate sample from animal H043 were included (
Fig. 4C). To obtain more evidence and to avoid the detection of DIs, the nucleoprotein gene of the genomes isolated from the same animals, and including a third animal per group (H015 or H041), was also subjected to deep sequencing. To this end, one PCR fragment covering the complete ORF was analyzed, and deep sequencing resulted in coverage of this region by 22,655 to 49,335 reads (
Fig. 4B). Similar to the results obtained for the polymerase protein ORF, SNPs detected in the viral genomes isolated from treated animals were also present in those obtained from untreated animals and were present in fewer than 10% of the reads (see Table S2 in the supplemental material). There were no significant differences between the numbers of SNPs detected in the nucleoprotein ORF for the treated and untreated samples (
Fig. 4C). Thus, next-generation sequencing of approximately 50% of the viral genome in samples obtained from treated and untreated animals did not reveal significant differences in SNPs detected in treated and untreated samples (
Fig. 4C), which indicates that T-705 did not induce specific mutations in the analyzed parts of the viral genome.
To investigate whether T-705 has a direct effect on the activity of the polymerase proteins of HMPV, as described for those of influenza virus and chikungunya virus (
18,
47), the inhibitory effect of T-705 on polymerase complex activity was evaluated by use of a minigenome system. Addition of DMSO alone (data not shown) or increasing concentrations of T-705 had no inhibitory effect on the expression of the constitutively expressed renilla luciferase or on the expression of a T7 promoter-driven firefly luciferase expression plasmid (
Fig. 5A). In contrast, addition of increasing concentrations of T-705 to the minigenome system for HMPV, expressing firefly luciferase, demonstrated an inhibitory effect of T-705 on the HMPV polymerase activity (
Fig. 5). Addition of 80 μM T-705 reduced the activity by 50% compared to that in untreated samples, while addition of 800 μM T-705 significantly reduced the luciferase activity, to below 10% (
P = 0.002; for the
t test,
P < 0.05). These data demonstrate that for HMPV, the mechanism of action of T-705 is at least partially directed against the polymerase activity and is not the induction of lethal mutagenesis of viral genomes.