Abstract
Viral emergence and reemergence underscore the importance of developing efficacious, broad-spectrum antivirals. Here, we report the discovery of tetrahydrobenzothiazole-based compound 1, a novel, broad-spectrum antiviral lead that was optimized from a hit compound derived from a cytopathic effect (CPE)-based antiviral screen using Venezuelan equine encephalitis virus. Compound 1 showed antiviral activity against a broad range of RNA viruses, including alphaviruses, flaviviruses, influenza virus, and ebolavirus. Mechanism-of-action studies with metabolomics and molecular approaches revealed that the compound inhibits host pyrimidine synthesis and establishes an antiviral state by inducing a variety of interferon-stimulated genes (ISGs). Notably, the induction of the ISGs by compound 1 was independent of the production of type 1 interferons. The antiviral activity of compound 1 was cell type dependent with a robust effect observed in human cell lines and no observed antiviral effect in mouse cell lines. Herein, we disclose tetrahydrobenzothiazole compound 1 as a novel lead for the development of a broad-spectrum, antiviral therapeutic and as a molecular probe to study the mechanism of the induction of ISGs that are independent of type 1 interferons.
INTRODUCTION
Despite the economic and health care burden posed by viral infections, current treatments for associated diseases are limited mostly to prophylactic vaccines. Only a small number of viral diseases (e.g., human immunodeficiency virus and hepatitis C virus [HCV]) can be treated with virus-specific therapeutics (e.g., sofosbuvir) (1, 2). These agents, so-called direct acting antivirals (DAAs), target viral gene products for their activities. In general, DAAs are prone to develop resistant mutants and have a narrow antiviral spectrum. Given the emergence of new viruses and the rapid spread of emerging viral diseases to previously unaffected geographic areas, there is an urgent need for the identification of agents that more efficiently target a broad range of viral diseases, which DAA approaches may not be able to deliver in time.
While broad-spectrum antivirals may overcome these limitations, the development of these agents has been hindered due to low efficacy or undesirable toxic effects, which are intrinsic characteristics of most broad-spectrum antivirals. For example, ribavirin has been studied since 1972 and tested against many RNA viruses; however, its useful antiviral spectrum is relatively narrow (3). Many RNA viruses, such as alphaviruses, are not susceptible to ribavirin, and patients may not benefit from the treatment due to its limited therapeutic window (4–6). T-705, an RNA-dependent RNA polymerase inhibitor, was also reported to have antiviral activity against a variety of RNA viruses. It is under development as a therapeutic candidate; however, its potency (50% inhibitory concentration [IC50]) falls in the few hundred micromolar range for most viruses, with the exception of influenza viruses (7).
Previously, we reported the discovery of new anti-Venezuelan equine encephalitis virus (VEEV) inhibitors from a high-throughput screening (HTS) campaign (8). VEEV is an RNA virus that causes encephalitis in humans and equids, and effective therapeutics for the disease have not yet been developed. We screened a library of 348,000 small-molecule compounds with a cell-based assay that measured the protection of cells from VEEV-induced cytopathic effect (strain TC-83) and discovered five active compounds (hits) with 50% effective concentrations (EC50s) that were better than 15 µM. One of these hits and the resulting optimized lead, ML336, turned out to be a DAA that inhibits viral RNA synthesis by targeting the amino terminal domains of viral nonstructural proteins 2 and 4 (8, 9).
In this current study, we investigated whether the HTS had identified a broad-spectrum antiviral inhibitor. Since the screen was based on a functional readout, i.e., reduction in virus-induced cell death, we hypothesized that the screen could identify a broad-spectrum antiviral compound as well. Indeed, we found that one of our hit compounds, CID 847035, did show antiviral effects in many cell-based antiviral assays, including the Marburg virus assay (http://pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?aid=540276), the Lassa virus assay (http://pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?aid=540256), and the respiratory syncytial virus (RSV) assay (http://pubchem.ncbi.nlm.nih.gov/assay/assay.cgi?aid=2391). Based on these observations, we undertook a study aimed at the development of novel, broad-spectrum antiviral inhibitors based on hit compound CID 847035 and the mechanism of action underlying their activity against multiple viruses.
Herein, we present the broad-spectrum antiviral behavior and mechanism of tetrahydrobenzothiazole compound 1 (Fig. 1), an analogue of our primary hit compound CID 847035. Using metabolomics and genomics approaches, we found that compound 1 inhibits pyrimidine biosynthesis and establishes an antiviral state by activating the genes involved in innate immunity, including those for retinoic acid-inducible gene I protein (RIG-I; encoded by DDX58), interferon-induced protein with tetratricopeptide repeats 1 (ISG56; encoded by IFIT1), and 2′-5′-oligoadenylate synthetase-like (OASL) protein (encoded by OASL). Interestingly, the antiviral status induced by compound 1 was independent of type 1 interferons (IFNs) or exogenous RNA. Importantly, we also determined that mouse cell lines are not capable of establishing antiviral status when treated with compound 1.
FIG 1.
Antiviral activity of compound 1 against VEEV. The structures of CID 847035 (A) and compound 1 (B). (C) Viral RNA (vRNA) analysis. Viral RNA was quantified using a quantitative real-time RT-PCR method with the total RNA from the cells. RNA amounts were compared to the DMSO-treated controls. (D) Titer reduction assay results for compound 1. Vero 76 cells grown in 12-well plates were infected with TC-83 at an MOI of 0.05 and then incubated in the presence of compound 1 at the denoted concentrations. Forty hours later, the supernatant was harvested and the titer of the progeny virus was determined. Each point represents the mean from three biological replicates. (E) Time of addition study. Test compound, compound 1, was added to the designated wells by replenishing the culture medium with fresh culture medium containing 5 μM compound at the time points denoted on the x axis. The cells were infected with virus at an MOI of 3, and the virus titers at 16 h postinfection from various time-of-addition points were depicted. Each data point is the mean from two independent replicates with triplication in titration.
MATERIALS AND METHODS
Cells and viruses.
Vero 76 (ATCC CRL-1587), BHK (ATCC CCL-10), HEp-2 (ATCC CCL-23), Neuro 2A (ATCC CCL-131), SH-SY5Y (ATCC CRL-2266), MRC-5 (ATCC CCL-171), and HEK 293T (ATCC CRL-3216) were obtained from ATCC and maintained in minimum essential medium with Earl's modification (MEM-E), containing 10% fetal bovine serum (FBS) and 1× GlutaMAX (Gibco 35050-061) at 37°C with 5% CO2. MDCK (product no. 84121903; Sigma-Aldrich), RD (ATCC CCL-136), and NIH 3T3 (ATCC CRL-1658) were maintained in Dulbecco's modified Eagle's medium (DMEM) with 10% FBS and 2 mM l-glutamine at 37°C with 5% CO2. TZM-bl cells were obtained through the NIH AIDS Research and Reference Reagent Program from John C. Kappes, Xiaoyun Wu, and Tranzyme, Inc. The TZM-bl indicator cell line, used for infectivity assays of HIV-1, is a genetically engineered HeLa cell clone that expresses CD4, CXCR4, CCR5, and Tat-responsive firefly luciferase and Escherichia coli β-galactosidase under the control of an HIV-1 long terminal repeat. TZM-bl cells were cultivated in DMEM (containing 4.5 g/liter glucose, l-glutamine, and sodium pyruvate) with 10% fetal calf serum, 50 IU/ml penicillin, and 50 μg/ml streptomycin at 37°C with 5% CO2.
VEEV TC-83 (lyophilized vaccine from USAMRIID) and V3526 were amplified in and titrated in BHK-21 cells. V3526-luc was rescued from the BHK cells and transfected a full-length viral RNA derived from pV3526-luc as described elsewhere (10). VEEV TrD (a gift from R. Tesh, World Reference Center for Emerging Viruses and Arboviruses) and herpes simplex virus 2 (HSV-2) (a gift from J. Steinbach, University of Louisville) were grown in Vero 76 cells that were maintained in DMEM with 10% FBS. The lymphocytic choriomeningitis virus ARM strain (LCMV-ARM) was a gift from I. Lukashevich (University of Louisville). Chikungunya virus (CHIKV) S-27 (BEIR NR-13220), Western equine encephalitis virus California (WEEV California; ATCC VR-70), Japanese encephalitis virus (JEV) SA14 (BEIR NR-2335), West Nile virus (WNV) NY-99 (a gift from R. Tesh, World Reference Center for Emerging Viruses and Arboviruses), and yellow fever virus 17D (YFV 17D; BEIR VR-1506) were grown in Vero 76 cells in a virus infection medium (MEM-E with 10% FBS, 1× GlutaMAX, and 25 mM HEPES, pH 7.3). Human enterovirus D71 (EV D71; strain MP4, BEIR NR-472) and encephalomyocarditis virus (EMCV) MM (BEIR NR-19846) were amplified in RD cells maintained in DMEM with 10% FBS.
To generate wild-type HIV-1, HEK 293T cells were plated at a density of 6 × 106 cells/100-mm culture dish 24 h prior to transfection. Cells were transfected with 10 μg of a wild-type HIV-1 proviral clone, pNL4.3, using linear polyethylenimine (25 kDa; Polysciences, Inc.) as described elsewhere (11). Culture supernatants were collected 48 h after transfection, and cellular debris was removed by filtration through a 0.45-μm filter. Viral p24 was quantitated using a standard p24 enzyme-linked immunosorbent assay (ELISA).
Dose-response studies.
EC50s and 50% cytotoxic concentrations (CC50s) were evaluated in a dose-response format starting from 50 μM with a 2-fold dilution, each in triplicate, in a 96-well format. For a CPE-based assay, cells were seeded in white well plates at a cell density of 12,000 cells per well in a volume of 45 μl and were incubated in an actively humidified incubator with 5.0% CO2 at 37°C and 95% humidity for 18 h. Test compounds diluted in 30 μl of cell culture medium were added to each well. After a 2-h incubation at 37°C with 5% CO2, 600 PFU of virus (or cell culture medium for the cytotoxicity assay) was added to the wells in a volume of 15 μl and then incubated for 2 days in an actively humidified incubator with 5.0% CO2 at 37°C and 95% humidity. Cell viability was measured with 90 μl per well of CellTiter-Glo reagent (Promega). Vero 76 cells were used for alphaviruses, and RD cells were used for human enterovirus 71 (EV-71) and EMCV assays. For a luciferase-tagged virus assay (i.e., V3526-luc and pVSV-luc), an optimized amount of virus and cell numbers were used for each cell line tested. For HEK 293T, Neuro 2A, and SH-SY5Y cells, 24,000 cells and 2,400 50% tissue culture infective dose (TCID50) units of virus per well were used. For Vero 76 and BHK cells, 12,000 cells and 1,200 TCID50 units of virus per well were used. For NIH 3T3, 24,000 cells and 20,000 TCID50 units of virus per well were used. After an 18-h incubation with virus, the plates were developed with Bright-Glo reagent (Promega), and luciferase activity was measured as a readout for the virus replication.
For measurement of inhibition of ebolavirus (EBOV) infection, a recombinant ebolavirus with a green fluorescent protein (GFP) gene inserted into the genome (Ebola-eGFP virus, a kind gift from H. Feldmann) was used (12). The virus stock was generated by infecting Vero E6 cells with Ebola-eGFP virus, followed by pelleting the culture supernatant through a 20% sucrose cushion. HeLa cells were pretreated for 2 h with 2-fold dilutions of 10 to 0.005 μM compound and were incubated with virus for 24 h in the presence of the compound. Fixed cells were imaged by a fluorescence microscope. Total and infected cells were counted by CellProfiler image analysis software (Broad Institute, MIT, Boston, MA) that detected nuclei stained with 4′,6-diamidino-2-phenylindole (DAPI) and virus-encoded GFP expression (13). This work was performed in a biosafety level 4 (BSL4) laboratory at the Texas Biomedical Research Institute.
Titer reduction assay.
To measure virus titer reduction, 12-well plates with 100,000 Vero 76 cells grown overnight were pretreated with compound diluted in virus infection medium at 37°C with 5% CO2 for 8 h unless denoted. For virus adsorption, cell plates were incubated on ice for 15 min, and then the cell culture supernatant was removed. Virus diluted at a multiplicity of infection (MOI) of 0.05 (or 3 for a time-of-addition study) in 250 μl of virus infection medium was added to the cell, and the virus was allowed to adsorb to the cells on ice for 1 h. The unadsorbed virus was washed with 1 ml of phosphate-buffered saline (PBS) once, and the wells were replenished with virus infection medium with 5 μM ML416 or dimethyl sulfoxide (DMSO) (0.25%, vol/vol). The progeny virus was harvested after 26 h for VEEV, CHIKV, WEEV, HSV-2, or influenza virus, after 48 h for JEV, YFV 17D, or WNV, and after 72 h for RSV or LCMV. The progeny viruses in the supernatants were enumerated by using either virus infection center assay or TCID50 assay. The virus infection center assay was done with Vero 76 cells grown confluent in 24-well plates. The cells were infected with 167 μl of the serially diluted virus samples for 1 h at 37°C with 5% CO2. Wells were washed with PBS and replenished with virus infection medium with 0.75% methylcellulose. Three or four days after virus infection, virus infection centers were visualized with crystal violet staining (0.2% crystal violet, 4% paraformaldehyde, and 10% ethanol). For influenza virus and HSV-2, a TCID50 assay was used for titration.
HIV-1 viral infectivity assay.
TZM-bl indicator cells were plated at a density of 10,000 cells/well in a 96-well culture plate 24 h prior to HIV-1 infection and were incubated at 37°C (5% CO2). On the day of infection, the culture medium was removed, and the cells were inoculated in triplicate with 100 μl of 2-fold serial dilutions of viral supernatants in culture medium containing 20 μg/ml DEAE-dextran and incubated at 37°C (5% CO2). At the time of infection, DMSO as a control or 10 μM ML416 was added to test the effect of the compound on HIV-1 infection. After a 24-h incubation, the culture medium was removed from each well and replaced with 100 μl of britelite plus luciferase assay substrate (PerkinElmer). Following 5 min of incubation at room temperature, 70 μl of each cell lysate was transferred to a 96-well OptiPlate-96 (PerkinElmer), and luminescence was measured in a Victor X2 multilabel reader (PerkinElmer). Relative infectivity was calculated by plotting the luciferase activity of the viral particle with treated DMSO as 100%.
Microarray.
HEp-2 cells were treated for 18 h with either 5 µM CID 70698683 or DMSO for the control in cell culture medium. One hundred nanograms of total RNA was amplified and labeled by following the Affymetrix (Santa Clara, CA) standard protocol for their 3′ IVT Plus labeling kit, followed by hybridization to Affymetrix PrimeView human gene expression arrays. The arrays were processed by following the manufacturer recommended wash and stain protocol on an Affymetrix FS 450 fluidics station and scanned on an Affymetrix GeneChip 7G scanner using Command Console 4.0. The resulting .cel files were imported into Partek Genomics Suite 6.6, and transcripts were normalized and summarized using robust multiarray averaging (RMA) as the normalization and background correction method. A one-way analysis of variance (ANOVA) was set up to compare the treatment of 5 µM CID 70698683 to the control. False discovery rate (FDR) was chosen as the multiple test correction for the resulting P values.
Real-time PCR.
Total RNAs from cells in a 12-well plate were isolated with RNAzol RT (Molecular Research Center, Inc.) reagent as per the manufacturer's recommendations and were dissolved in 50 µl of the RNA storage solution (Life Technologies). One-microgram RNA samples were subjected to a cDNA synthesis with Maxima H Minus reverse transcriptase (Life Technologies), random hexamers, and oligo(dT) by following the manufacturer's recommendations. For quantitation of gene expression, we used a real-time PCR with the 2−ΔΔCT method in a total of 20 μl per well with 2 µl of 2-fold-diluted cDNA mixture in a multiplex mode in conjunction with TaqMan chemistry. Information on the primers and probes is in Table S2 in the supplemental material. The copy number of the viral genome was quantitated as described elsewhere (8). To quantitate the relative viral and human gene RNA copy numbers, 18S rRNA (catalog no. 4319413E; Life Technologies) and human GAPDH (glyceraldehyde-3-phosphate dehydrogenase) (catalog no. 4326317E; Life Technologies), respectively, were used as the endogenous controls. Three biological replicates, each with two technical replicates, were used for the quantitation.
Interferon assays.
Interferon-α and interferon-β in the cell culture supernatant were detected by using LumiKine Xpress hIFN-α and LumiKine hIFN-β kits (InvivoGen) as per the manufacturer's recommendations. One-day-old HEK 293T cells that were cultured in a 12-well plate were treated with 5 μM compound 1 or DMSO in virus infection medium for 18 h, and the cell culture supernatants were harvested and cleared by centrifugation at 3,000 × g for 10 min. For each treatment, six replicates were used per group, and three technical replicates were used for the controls, HEK 293-expressed human IFN-α2, and CHO-expressed human IFN-β.
The HEK-Blue IFN-α/β cell reporter assay (InvivoGen) was performed by following the manufacturer's recommendations. One-day-old HEK-Blue IFN-α/β cells plated in a 96-well plate were treated with 2-fold dilutions of 50 to 0.4 μM compound 1 or DMSO for 24 h. The expression of secreted alkaline phosphatase (SEAP) under the control of interferon-stimulated response element 9 (ISRE9), which is activated by type 1 interferons, was measured by determining the absorbance at 620 nm after 20 min of incubation with 100 μl of SEAP substrate. Three replicates were used for each data point.
Metabolomics analysis.
Metabolites were extracted from cell samples using a solvent mixture of acetonitrile, water, and chloroform (2:1.5:1 by volume). The metabolite extract from each sample was then derivatized using N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA). All derivatized samples were further analyzed on a LECO Pegasus 4D two-dimensional gas chromatography (GC) with time-of-flight mass spectrometer instrument (St. Joseph, MI). The instrument data were first processed using LECO's instrument control software ChromaTOF for peak picking and tentative metabolite identification. MetPP software was used for retention index matching, peak merging, peak list alignment, normalization, and statistical significance testing (14). The abundance test was performed using the pairwise two-tailed t test with sample permutation to standardize the abundance variation of each metabolite between sample groups. The presence/absence test was performed using Fisher's exact test.
RESULTS
Discovery of tetrahydrobenzothiazole compound 1.
Our lead, compound 1, resulted from a rigorous medicinal chemistry optimization effort focused on a hit compound (CID 847035) that was prioritized from our high-throughput screen that measured the protection of Vero 76 cells from a VEEV-induced CPE (Fig. 1A and B) (8). Medicinal chemistry optimization of this hit compound, which will be described separately in due course, was executed in parallel to the development of a structurally and mechanistically distinct anti-VEEV probe, ML336. Unlike ML336, compound 1 showed a potent antiviral effect toward multiple alphaviruses with CPE EC50s of 0.35 μM, 0.29 μM, and 0.96 μM for VEEV TC-83, CHIKV S27, and WEEV, respectively (Table 1).
TABLE 1.
Antiviral activity of compound 1a
| Virus family | Virus | EC50 (µM) | Log titer reduction at 5 µMb |
|---|---|---|---|
| Togaviridae | VEEV TC-83 | 0.35 | −2.54 |
| VEEV TrD | 0.46 | −4.77 | |
| VEEV V3526-luc | 0.17c | NTd | |
| CHIKV S17 | 0.29 | −4.024 | |
| WEEV | 0.96 | NT | |
| Rhabdoviridae | pVSV-luc | 0.19c | NT |
| Filoviridae | Ebola virus-GFP | 0.26 | NT |
| Picornaviridae | EV-71, MP4 | 11.1 | NT |
| EMCV MM | >25 | NT | |
| Orthomyxoviridae | Influenza virus A (H1N1) | NT | −3.03 |
| Arenaviridae | LCMV-ARM | NT | −3.466 |
| Paramyxoviridae | RSV Long | NT | −2.27 |
| Retroviridae | HIV-1 | 94% inhibition at 10 µM | NT |
| Flaviviridae | JEV | NT | −1.3 |
| YFV 17D | NT | −2.32 | |
| WNV | NT | −2.05 | |
| Herpesviridae | HSV-2 | NT | 0.01 |
For titration reduction assays, the progeny virus titers were compared to the controls (DMSO treated). For dose-response assays, cells were pretreated for 2 h prior to infection, and the luciferase activity was measured 16 h later.
Negative values indicate a decrease in progeny virus titer compared to the mock-treated controls. The data represent the means of at least three replicates.
These values are associated with a luciferase-tagged virus assay.
NT, not tested.
To verify the antiviral activity of compound 1 against multiple alphaviruses, we employed various assays with different readouts. In a luciferase-based anti-VEEV assay (V3526-luc) (Table 1), compound 1 showed promising antiviral activity with an EC50 of 0.17 μM. In a quantitative real-time PCR (qRT-PCR) assay with the VEEV TC-83 strain (Fig. 1C), compound 1 (20 µM) decreased viral RNA copy numbers by more than 1,000-fold compared to the control. Finally, we employed a titer reduction assay with a variety of alphaviruses to confirm the antiviral activity of compound 1. For VEEV TC-83, the progeny virus production decreased by more than 15,000-, 2,600-, or 340-fold at 20, 10, or 5 µM, respectively (Fig. 1D). VEEV TrD and CHIKV S17 strains were more sensitive to compound 1. At a concentration of 5 µM, the compound decreased the level of viral replication by more than 10,000-fold (4 log) for the two viruses. From these experiments, we concluded that tetrahydrobenzothiazole compound 1 potently inhibits multiple alphaviruses.
Time-of-addition assay.
To understand at which stage of virus replication compound 1 exerted its antiviral activity, we employed a time-of-addition assay with Vero 76 cells synchronously infected with TC-83 at a multiplicity of infection (MOI) of 3 (Fig. 1E). Compound 1 was added to cells at various time points with regard to the virus replication cycle and maintained until the progeny virus was harvested. The experiment showed that pretreatment was necessary in order to observe full antiviral activity. Treatment at time zero, right after virus adsorption to cells on ice, resulted in only a 65% reduction in progeny virus titer; however, pretreatment for 6 h resulted in an ∼400-fold reduction in virus titer. Antiviral efficacy of compound 1 was dependent on the length of pretreatment time between 4 and 0 h prior to infection. This result clearly shows that compound 1 requires the induction of a cellular response for its activity.
Cytotoxicity of tetrahydrobenzothiazole compound 1.
The cytotoxicity of compound 1 was evaluated, as a cytotoxic compound may misleadingly exhibit broad-spectrum antiviral activity. Vero 76 cells were plated in 96-well plates and incubated for 48, 72, and 96 h in the presence of compound 1 at various concentrations. Compound 1 did not show apparent cytotoxicity up to 12.5 µM (see Fig. S1A in the supplemental material). The CC50 values were 74.1, 31.0, and 34.6 µM after 2, 3, and 4 days of exposure, respectively, resulting in a selective index 50 (SI50) greater than 100 (CC50 at day 3/EC50 of CHIKV = 106.9). This indicated that the agent is not toxic to cells at the effective concentrations and that the inhibition of virus replication is not due to a nonspecific cytotoxicity. Cells treated with compound 1, however, looked larger and more stretched than those in the control group under visual observation. The compound was further tested in viability, cytotoxicity, and/or apoptosis assays with HEK 293T cells for 4 days. No significant cytotoxicity or apoptosis was induced with up to 25 µM of compound 1 (see Fig. S1B in the supplemental material).
Antiviral spectrum of tetrahydrobenzothiazole compound 1.
As compound 1 showed similar antiviral effects on all of the alphaviruses that we tested, we questioned whether the compound could inhibit a broader spectrum of viruses. To address this question, we tested compound 1 against various viruses, including vesicular stomatitis virus (pVSV-luc), respiratory syncytial virus (RSV), ebolavirus (EBOV), HIV-1, Japanese encephalitis virus (JEV), herpes simplex virus 2 (HSV-2), yellow fever virus 17D (YFV 17D), West Nile virus (WNV), influenza virus (IFNV), EV-71, and encephalomyocarditis virus (EMCV).
As shown in Table 1, compound 1 showed an antiviral effect against a broad-spectrum of viruses, with different sensitivity to each virus. Alphaviruses, LCMV, and IFNV were most sensitive. Treatment with compound 1 (5 μM) resulted in a more than 4-log reduction in the progeny virus titers of the alphaviruses tested and a more than 3-log reduction for IFNV and LCMV. RSV, WNV, and YFV 17D were sensitive as well with an ∼2-log titer reduction; however, JEV was less sensitive than the other viruses tested (∼1.3-log reduction). HIV-1 infection was sensitive as well, with 94% reduction in viral infectivity using the TZM-bl reporter system that measures luciferase activity with 10 µM concentration of compound 1. Interestingly, the replication of HSV-2 was not inhibited by compound 1, resulting in no changes in progeny virus titer.
The compound's antiviral activity was also assessed using a dose-response format (EC50 determination). The replication of pVSV-luc or EBOV-GFP was sensitive to the compound treatment with EC50s of 0.19 µM and 0.30 µM, respectively. The agent, however, did not inhibit the replication of the picornaviruses efficiently (i.e., EC50s of EV D71 and EMCV MM were 11.1 µM and >25 µM, respectively). From these experiments, we found that compound 1 has a broad-spectrum antiviral effect with different sensitivities depending on virus. This result implies that the antiviral mechanism of compound 1 has specificity to certain types of viruses.
Pyrimidine synthesis inhibition by tetrahydrobenzothiazole compound 1.
To test whether our lead compound could interfere with cellular metabolism for its antiviral activity, we measured the difference in cellular metabolites using a metabolomics approach. HEK 293T cells were treated either with 5 µM compound 1 or with DMSO for 18 h, and the cellular metabolites were analyzed and compared using GC mass spectrophotometry. The treatment with compound 1 changed the abundance of cellular metabolites significantly (see Table S1 in the supplemental material). For example, the amounts of l-glutamic acid and fumaric acid differed by 0.09- and 11-fold when the group treated with compound 1 was compared to mock-treated cells. For some metabolites, the difference was more evident (Table 2). For example, dihydroorotic acid (DHO) and orotic acid were found only in the groups treated with compound 1. DHO and orotic acid are metabolites that are related to de novo pyrimidine synthesis, suggesting that compound 1 inhibits pyrimidine synthesis after the synthesis of orotic acid from DHO. Consistent with this result, free uridine was not detected in the groups treated with compound 1.
TABLE 2.
Difference in cellular metabolites with the treatment of compound 1
| Group | Metabolite(s) | P value |
|---|---|---|
| Metabolites detected only in cells treated with compound 1 (not detected in mock-treated cells) | dl-Dihydroorotic acid | 4.11 × 10−4 |
| Orotic acid | 2.26 × 10−3 | |
| l-Serine | 1.52 × 10−2 | |
| Citrulline | 4.11 × 10−5 | |
| Alloxanoic acida | 4.11 × 10−5 | |
| 2-Ketoisovaleric acida | 4.11 × 10−5 | |
| Metabolites detected only in mock-treated cells (not detected in cells treated with compound 1) | Uridine | 4.11 × 10−5 |
| d-(-)-Erythrofuranose,(isomer)a | 4.16 × 10−4 | |
| (3R,4R)-tetrahydrofuran-2,3,4-triola | 4.11 × 10−5 | |
| Dihydromuconic acid; trans-3-hexenedioic acida | 1.52 × 10−2 | |
| Uridine phosphatea | 4.11 × 10−4 | |
| Methanesulfinic acida | 4.11 × 10−4 | |
| N-Acetyl-l-glutamic acida | 4.11 × 10−5 | |
| Putrescine; 1,4-diaminobutanea | 4.11 × 10−5 | |
| Pidolic acida | 3.30 × 10−3 | |
| 2-Hydroxyglutaric acid | 2.32 × 10−3 |
These compounds have not been confirmed using standard compounds.
To validate this finding, we tested to see if the antiviral effect of compound 1 could be overcome by the addition of exogenous pyrimidines. We measured virus replication in HEK 293T cells that were treated with mycophenolic acid (MPA) or compound 1 in the presence of various exogenous nucleosides and then compared the virus replication to that in the controls (Fig. 2). In groups treated with MPA, the antiviral effect of MPA was completely reversed when guanosine was supplemented in the culture, from 47% to 105% and 17% to 62% for pVSV-luc and V3526-luc, respectively. This result was consistent with our expectation that MPA inhibits inosine-5′-monophosphate dehydrogenase (IMPDH) as its antiviral mechanism, which is a key enzyme to synthesize guanosine de novo. Similarly, the antiviral effect of compound 1 greatly decreased when the cells were supplemented with the pyrimidines, cytidine or uridine. For example, while 5 µM compound 1 decreased the replication of V3526-luc to 1.9% compared to that of the control, the addition of cytidine or uridine restored the viral replication to 83.4% or 77.5% compared to that of the control, respectively.
FIG 2.
Reversion of antiviral effect of compound 1 by exogenous nucleosides. One-day-old HEK 293T cells were treated with DMSO, MPA (1 µM), or ML416 (1 µM) in the presence of the denoted supplements for 2 h, and then cells were infected with V3526-luc (A) or pVSV-luc (B). After 18 h of incubation, the luciferase activity from the infected cells was measured. The values represent the means and their standard deviations of 4 replicate samples as a percentage of the values for DMSO control wells.
Two pyrimidine synthesis intermediates, DHO and orotic acid, were also tested to confirm the metabolomics results with the accumulation of the molecules in the group treated with compound 1. The addition of DHO did not affect the antiviral activity of compound 1 at all, suggesting that compound 1 inhibits a downstream step from DHO. Orotic acid showed a moderate reversion effect to compound 1 (21.9% to 55.0% and 1.9% to 7.1% for pVSV-luc and V3526-luc, respectively).
Induction of innate immune genes by tetrahydrobenzothiazole compound 1 without virus infection or type 1 interferons.
It has been reported that the inhibition of pyrimidine biosynthesis may amplify the cellular response to single-stranded RNA (ssRNA) via the type 1 IFN system (15); however, we questioned this as Vero 76, the primary cell line used for antiviral activity testing of compound 1 and its derivatives, is known to be deficient in type 1 IFN production (16).
To test whether compound 1 induced cellular immune response without type 1 IFN, we examined the changes in host gene expression after treatment with CID 70698683 without addition of ssRNA. CID70698683 is an N-phenylthiophene derivative of compound 1 that has similar antiviral effects on multiple alphaviruses with CPE EC50s of 0.60 μM, 0.66 μM, and 0.93 μM for VEEV TC-83, CHIKV S27, and WEEV, respectively. Human HEp-2 cells were treated with 5 μM CID 70698683 or DMSO for 18 h, and then the cellular mRNAs were subjected to a DNA microarray assay. Ninety-two genes were upregulated, and 145 genes were downregulated by a more than 2-fold difference (GEO accession no. GSE72167). Among these changes, certain sets of genes that are involved in the interferon pathways were clearly upregulated (Fig. 3), including RIG-I (4.63-fold) and OASL (3.76-fold increase). We found that some interferon-stimulated genes (ISGs) (e.g., GBP2, ISG20, IFI44, IRF9, or IFIT1) were also upregulated significantly (3.5, 2.85, 2.83, 2.4, or 2.3, respectively). While the ISGs were upregulated, the expressions of interferons were not (approximately 0.9-fold to 1.1-fold changes); rather, the expression of IFN-ε decreased by half. These findings are consistent with our hypothesis that induction of those ISGs occurs without functional IFNs.
FIG 3.
Induction of innate immune genes by CID 70698683. An asterisk indicates information from reference 25. Values in red indicate a fold change by the treatment of 5 µM CID 70698683.
We further validated this finding with a real-time PCR and ELISA. First, we sought to understand whether compound 1 simply amplified cellular innate immune responses after virus infection or if the compound was able to establish an antiviral state without an external interferon inducer, such as a virus infection. To test this, we measured the induction of innate immune response genes by compound 1 in mock- or pVSV-luc-infected HEK 293T cells.
As shown in Fig. 4, the induction of innate immune genes by compound 1 was independent of virus infection. First, we confirmed that treatment with compound 1 induced only certain genes in the interferon pathway. For example, the expressions of MYD88 and OAS1 genes did not show any changes by compound 1; however, RIG-I, IFIT1, IRF7, and OAS2 genes were upregulated more than 10-fold (solid bars in blue). More interestingly, the induction of the genes by compound 1 was not amplified by virus infection. Treatment of mock- or virus-infected cells (solid versus spotty blue bars) with compound 1 resulted in the same level of gene expression for the tested genes. We also found that the induction of the ISGs by compound 1 was much stronger than by pVSV-luc infection. pVSV-luc increased the expression of innate immune genes, such as IFIH1, IFIT1, OAS2, OASL, and IFNB1, by approximately 2- to 4-fold, indicating that HEK 293T cells are responsive to the infection of the virus. However, the effect was much weaker compared to that of compound 1. Interestingly, compound 1 repressed the expression of IFIH1 and IFNB1; whereas, these genes were induced rather than repressed in the virus-infected groups (orange spotty bars).
FIG 4.
Compound 1 induces innate immune response genes independently of virus infection. One-day-old HEK 293T cells were treated with 5 μM compound 1 or DMSO (control) for 8 h and were then infected with mock or pseudotype VSV-luc (pVSV-luc) at an MOI of 3. Cells were further incubated for 16 h in the presence of compound 1 or DMSO, and then the host gene expression level was measured using quantitative real-time RT-PCR.
The lack of induction of type 1 IFNs by compound 1 was confirmed at the cytokine level with ELISA and a cell-based reporter assay. HEK 293T cells were treated with 5 µM compound 1 for 18 h, and then the IFN α/β in the cell culture supernatant was detected in an IFN ELISA (Fig. 5A) or in a reporter assay (HEK-Blue IFN-α/β cells). HEK-Blue IFN-α/β cells directly respond to IFN-α/β and express the secreted alkaline phosphatase (SEAP) via the activation of IRF9. In the two assays, no measurable quantity of IFNs was detected compared to the controls, indicating the lack of IFN-α/β induction by treatment with compound 1. We also treated HEK-Blue IFN-α/β cells directly to test whether compound 1 could directly activate the type 1 IFN receptor (IFNAR) and IRF9 pathway or produce IFN in an autocrine manner (Fig. 5B). Treatment with compound 1 did not induce SEAP activity compared to the control, indicating no direct activation through IFNAR or any other receptor that can use IRF9.
FIG 5.
The treatment of compound 1 does not induce the production of type 1 IFNs. (A) HEK 293T cells were treated with compound 1 or DMSO for 18 h, and the cell culture medium was subjected to an IFN-α/β ELISA to detect interferons. A total of 125 pg/ml of IFN-α and -β was used as controls. (B) HEK-Blue IFN-α/β reporter cells were incubated with 5 µM compound 1 or DMSO for 18 h, and then the expressed SEAP, which is controlled by IRF9, was measured. OD620 nm, optical density at 260 nm.
In summary, these data indicate that compound 1 induces expression of certain sets of innate immunity genes, creating an antiviral state in cells in the absence of type 1 IFN
Cell line specificity of tetrahydrobenzothiazole compound 1.
Since the supporting data strongly suggested that compound 1 inhibits pyrimidine synthesis and induces the antiviral state of the host cells, we questioned whether the antiviral effect and mechanism of compound 1 is cell type dependent. To address this question, we measured the antiviral activity of compound 1 in several cell lines by determining an EC50 for each. The cell lines include HEK 293T (human embryonic kidney), SY-SH5S (human bone marrow derived neuroblast), Vero 76 (African green monkey kidney fibroblast), BHK C21 (hamster kidney fibroblast), Neuro 2A (mouse neuroblast), and NIH 3T3 (mouse fibroblast).
The antiviral activities of compound 1 in various cell lines are summarized in Table 3 as a function of EC50. As a control for the experiment, we used monensin, which is known to inhibit virus replication by hampering the acidification of endosome during virus entry. The antiviral activities of monensin were very close to each other in all cell lines tested in this experiment. The EC50s were within a range of 0.02 and 0.25 µM for V3526-luc and between 0.17 and 0.56 µM for pVSV-luc in all cell lines, with two exceptions in Vero 76 cells (2.87 and 3.87 μM). This result shows that endocytosis is a critical pathway for viruses in the cell lines and that monensin worked equally in the cell lines.
TABLE 3.
Antiviral activity of compound 1 in various cell linesa
| Cell line | EC50 (µM) of: |
|||
|---|---|---|---|---|
| Monensin |
Compound 1 |
|||
| V3526-luc | pVSV-Gp | V3526-luc | pVSV-Gp | |
| HEK 293T | 0.08, 0.03 | 0.19, 0.74 | 0.15 | 0.17 |
| SY-SH5S | 0.02 | 0.17, 0.48 | 0.49 | 3.82 |
| Vero 76 | 3.87 | 0.56, 2.57 | 0.17, 0.25 | 0.20, 0.68 |
| BHK | 0.09, 0.24 | 0.50, 0.19 | 1.12, 0.92 | 8.26, 7.16 |
| Neuro 2A | 0.25, 0.19 | 0.22, 0.23, 0.29 | >50 | >50 |
| NIH 3T3 | 0.02 | 0.24, 0.67 | >50 | >50 |
Each number represents the result of an independent determination of the EC50 from a dose-response study with concentrations starting at 25 μM with a 5-fold dilution, each in triplicate, in a 96-well format. For assays with HEK 293T, Neuro 2A, and SH-SY5Y cells, 24,000 cells and 2,400 TCID50 units of virus per well were used. For Vero 76 and BHK cells, 12,000 cells and 1,200 TCID50 units of virus per well were used. For NIH3T3, 24,000 cells and 20,000 TCID50 units of virus per well were used. EC50s were calculated using XLfit (IDBS) formula 205, a four-parameter Levenberg-Marquardt algorithm with maximum and minimum limits set at 100 and 0, respectively.
In contrast to monensin, the antiviral activities of compound 1 were cell line and/or species dependent. The EC50s of compound 1 in HEK 293T or Vero 76 cells were between 0.15 µM and 0.68 µM using either V3526-luc or pVSV-luc, thus implying a strong antiviral activity in these cell lines. Contrary to this, no antiviral activity of compound 1 was detected in the mouse cell lines (Neuro 2A and NIH 3T3) that we tested (EC50s of >50 µM for the two viruses). Interestingly, compound 1 still showed a moderate effect in BHK cells, a hamster cell line. These data clearly indicate that the antiviral activity of compound 1 is cell line dependent and much less effective in mouse cells.
DISCUSSION
This report discloses the discovery and mechanistic characterization of tetrahydrobenzothiazole compound 1, a broad-spectrum antiviral agent that inhibits pyrimidine biosynthesis and induces antiviral responses. This structural series was identified from anti-VEEV HTS with a mechanism-independent readout, offering the possibility that multiple antiviral agents with different mechanisms of action could be identified from the same screen. In fact, our pursuit of hits from the VEEV HTS led us to develop a structurally distinct inhibitor, ML336, which targets a novel domain of VEEV nsP2 and highlights the domain's importance for viral replication and as a unique antiviral target. In this study, we revisited our CPE assay hit compounds with the intent of finding a mechanism of action orthogonal to that of ML336. This effort ultimately led to the development of compound 1. There have been many HTS campaigns that employed the same strategy with different viruses (PubChem assay identification no. 1635, 2310, 2440, 540278, 463114, 540249, 588723, 651637, 1074, 51831, and 504781), with each resulting in the discovery of novel hit compounds. As evidenced with our study, those hits may include compounds targeting diverse antiviral mechanisms as well.
As Lucas-Hourani et al. (15) indicated, several antiviral HTS campaigns independently discovered pyrimidine synthesis inhibitors targeting dihydroorotate dehydrogenase (DHODH) as an antiviral hit (17–19). This fact may indicate that the inhibition of de novo pyrimidine biosynthesis can elicit antiviral activity without hampering cell proliferation. In fact, our experiment addressing the effect of compound 1 on growing cells showed no or minimal impact to cell growth or cytotoxicity at a concentration that is 10-fold higher than the EC50 (see Fig. S1 in the supplemental material). We believe that the cellular pyrimidine level that initiates the activation of the ISGs is higher than the concentrations affecting cell proliferation or viability. Unlike the DHODH inhibitors discovered by others, however, compound 1 seems to inhibit an enzyme downstream from DHODH. Our metabolomics analysis clearly showed that treatment with compound 1 caused the accumulation of orotate as well as DHO (Table 2), which indicates an inhibition at a postorotate step (see Fig. S2 in the supplemental material). The accumulation of orotidine-monophosphate was not monitored; therefore, we hypothesize that compound 1 may inhibit orotate phosphoribosyltransferase (OPRTase) activity, which is executed by a dual functional enzyme, UMP synthetase (Umps) (20). Nevertheless, the addition of orotate lessened the antiviral effect of compound 1 (Fig. 2B), which could indicate DHODH as the inhibitory step. We believe that the inhibition of OPRTase by compound 1 may not be potent enough; therefore, high concentrations of exogenous orotate (substrate) may facilitate the reaction even in the presence of compound 1. Alternatively, exogenous orotate may affect the gene expression of ISGs through an unknown pathway. A follow-up study is needed to understand the detailed mechanism of compound 1 in the context of a target enzyme.
Our assay matrix showed that compound 1 has broad antiviral activity against most of the viruses tested, but the effect was virus species dependent (Table 1). While RNA viruses seemed more sensitive to compound 1 than DNA viruses (e.g., HSV-2), not all RNA viruses exhibited equal responsiveness. For example, enteroviruses were resistant to treatment with compound 1. While most of the RNA viruses that we tested are recognized by RIG-I in the infected cells, the infections of HSV-2 and enteroviruses are known to be recognized by melanoma differentiation-associated gene 5 (MDA5, encoded by the IFIH1 gene) and TLR/MYD88, respectively (21, 22). Consistent with this finding, MDA5 and MYD88 were not induced by treatment with compound 1 (Fig. 4). A detailed study will be required to elucidate the exact mechanism; however, these results imply that it may be associated with the RIG-I pathway.
We also found that JEV was less susceptible to compound 1 than other tested flaviviruses. Assuming that the viral RNA syntheses were similar to each other (based on titers of the viruses), this finding may indicate a potential difference in the sensitivity to the ISGs that are activated by pyrimidine synthesis inhibition. It is known that the antiviral effect of pyrimidine synthesis inhibitors is through functional ISGs driven by the IRF1 transcription factor (15). Therefore, the difference in sensitivity may indicate the resistance to ISGs by each virus. Interestingly, some dengue virus mutants that are resistant to brequinar, a DHODH inhibitor, have been reported (23). In that case, the authors proposed a resistance mechanism based on the enhancement of polymerase activity or viral assembly; however, our study with compound 1 as the probe suggests that those mutants may exert resistance by evading the ISGs expressed by pyrimidine deprivation. We have attempted to isolate compound 1-resistant mutant VEEV strain TC-83 by serial passages in the presence of compound 1 three times; however, we have not detected the emergence of resistant mutants (see Fig. S3 in the supplemental material). A future study for isolation of resistant mutants with more passages would be informative to determine the resistance mechanisms of alphaviruses to this class of compounds, which may include an enhancement of polymerase activity or evasion of ISGs.
One of the most interesting findings of our study is the induction of ISGs by compound 1, including OASL, in the absence of sensitizing the cells with virus, exogenous ssRNA, or interferons (Fig. 4). OASL is an antiviral gene that is known to be expressed under the control of IRF-3 by interferon and ISRE upon virus infection. However, our data showed that OASL could be induced more than 100-fold without exogenous RNA or induction of IRF-3. In our experiment, however, IRF-3 was downregulated after the treatment. Further, the infection of compound 1-pretreated cells with pVSV-luc did not change the cellular innate immune response. This may be explained by the strong antiviral activity of compound 1, which may prevent viral RNA accumulation to a level to trigger the RIG-I/MAVS pathway. Alternatively, the result may imply that the induction of ISGs by compound 1 is independent of viral infection. Nonetheless, our study may explain the mechanism of the amplification of cellular innate immune response by pyrimidine synthesis inhibitors reported by others (15, 24). Our results clearly show that the amplification of the response is due to the IFN-independent induction of ISGs (see below for further discussion), which will prime or amplify the antiviral state of the cells upon viral infection. Our findings confirm previous reports and further highlight the antiviral mechanism of nucleoside synthesis inhibitors in detail.
In fact, our study showed that the expressions of ISGs are not associated with type 1 IFNs. In our gene expression assays, no type 1 IFN genes were induced by compound 1; rather, they were downregulated. Expression of interferon may be transient, so the lack of interferon mRNA may not rule out a potential expression at an early time point in the process. To avoid this pitfall, we validated our findings using an ELISA-based assay in addition to a reporter-based assay to detect the interferons produced by compound 1; however, no IFN-α/β activities were detected. Our argument is also supported by the antiviral activity of compound 1 in Vero 76 cells, which allow a variety of virus replication due to the lack of functional type 1 IFN production systems (16). Considering that the antiviral activity of compound 1 in Vero 76 cells was similar to that in other human cell lines (Table 3), this observation strongly supports the idea that the presence of interferon is not essential for the antiviral activity of compound 1 or other pyrimidine synthesis inhibitors and that the expressed ISGs are not involved in the production of interferons. Our results differ from the current view of the antiviral mechanism of many ISGs such as OASL, which is believed to induce type I interferons by assembling a RIG-I complex in response to viral infection (25). Therefore, our results clearly highlight new insights on the novel antiviral mechanisms of ISGs.
The host cell-type-dependent antiviral activity was particularly interesting. The species-dependent activity may be a universal characteristic of any host-targeting compound, including compound 1. Study of host-dependent activity is important for choosing an animal model to test these compounds. Compound 1 showed the strongest antiviral effect in human cells but no antiviral activity in the mouse cell lines that we tested. We also found that brequinar does not show antiviral effect in mouse cell lines (26), and Grandin et al. mentioned that GAC50, a potent novel DHODH inhibitor, showed antiviral activity only in primate cells (27). Interestingly, our findings are consistent with the lack of antiviral activity of pyrimidine synthesis inhibitors against RNA viruses in mice as reported by others. Some pyrimidine synthesis inhibitors have been evaluated for their in vivo antiviral activities in mouse-based animal models (except for the cotton rat model for RSV) (17–19, 28). Those tests have failed to show antiviral efficacy even with a significant decrease in the pyrimidine concentration in animal sera. Due to these results, pyrimidine synthesis inhibitors have not been well accepted as an antiviral strategy. Our findings, however, conversely suggest that such failure may be due to the use of nonresponsive animal models for the inhibitors. Leflunomide, a pyrimidine synthesis inhibitor, has been reported to have an antiviral effect in humans, suggesting pyrimidine synthesis inhibition as a potential broad-spectrum antiviral target for humans (29–31). Therefore, we believe that a worthy goal includes the development of an animal model for this class of inhibitors and the evaluation of them for antiviral activities. A recent study using rhesus macaques, however, failed to show the anti-RSV effect of a pyrimidine synthesis inhibitor when it was administered postinfection. It has been proposed that the level of circulating pyrimidines in the serum produced through the salvage pathway may be sufficient to reverse the antiviral effect or that the interference with cell-mediated immunity by the inhibitors may be proviral (27). In our study, a pretreatment was necessary to maximize the antiviral effect in the IFN-deficient cell line. It is not clear, however, whether a pyrimidine synthesis inhibitor would successfully mount an antiviral state and amplify the IFN-driven host innate immune response even at a postinfection stage in IFN-competent cells. More in vivo antiviral studies to test pyrimidine synthesis inhibitors in various animal models would be informative to evaluate the potential of pyrimidine biosynthesis inhibitors as antiviral therapeutics.
As the antiviral activity of compound 1 is based on pyrimidine synthesis inhibition and the subsequent expression of certain ISGs, it is difficult to determine whether the cell type specificity is related to pyrimidine synthesis inhibition or to inducing the ISGs. Differences in the sequences of the protein targeted by compound 1, presumably Umps, may account for this phenomenon. The sequence identity between human and mouse Umps is high (88.8%); however, this cannot rule out the potential difference in the binding residues in Umps. A future study is necessary to identify the molecular target. An alternative hypothesis is that the mouse lacks the mechanism to induce the ISGs upon the pyrimidine synthesis inhibition. This is based on the observation that pyrimidine synthesis inhibitors with various chemical structures have failed to demonstrate antiviral activity in mice as mentioned above. Currently, the molecular mechanism of ISG induction by pyrimidine depression is completely unknown. Lead compound 1 may shed light on the antiviral mechanism involving innate immune responses through the deprivation of pyrimidine.
In summary, tetrahydrobenzothiazole compound 1 is a molecular probe that enabled the study of innate cellular immune responses involved with viral infection. More specifically, compound 1 permits mechanistic scrutiny which demonstrates that antiviral ISGs, induced by the inhibition of pyrimidine synthesis, lead to broad-spectrum antiviral activity independent of type 1 IFNs.
Supplementary Material
ACKNOWLEDGMENTS
We thank Igor Lukashevich for LCMV, Jill Steinbach for HSV-2, and Robert Tesh at the World Reference Center for Emerging Viruses and Arboviruses for VEEV TrD and WNV. We also thank BEI resources for CHIKV, JEV, YFV 17D, EV D71, and EMCV MM.
Microarray experiments were performed with the assistance of the University of Louisville Genomics Facility, which is supported by the NIH, the J. G. Brown Foundation, and user fees.
Funding Statement
Chemistry support was provided by the NIH under grant U54HG005031 (KU SCC). Microarray experiments were performed with the assistance of the University of Louisville Genomics Facility, which is supported by NIH grant P20GM103436 (KY IDeA Networks of Biomedical Research Excellence), NIH grant P30GM106396 (University of Louisville J. G. Brown Cancer Center Phase III CoBRE), the J. G. Brown Foundation, and user fees. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Footnotes
Supplemental material for this article may be found at http://dx.doi.org/10.1128/AAC.00282-16.
REFERENCES
- 1.Stedman C. 2014. Sofosbuvir, a NS5B polymerase inhibitor in the treatment of hepatitis C: a review of its clinical potential. Therap Adv Gastroenterol 7:131–140. doi: 10.1177/1756283X13515825. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zeng QL, Zhang JY, Zhang Z, Wang LF, Wang FS. 2013. Sofosbuvir and ABT-450: terminator of hepatitis C virus? World J Gastroenterol 19:3199–3206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Sidwell RW, Huffman JH, Khare GP, Allen LB, Witkowski JT, Robins RK. 1972. Broad-spectrum antiviral activity of Virazole: 1-beta-d-ribofuranosyl-1,2,4-triazole-3-carboxamide. Science 177:705–706. doi: 10.1126/science.177.4050.705. [DOI] [PubMed] [Google Scholar]
- 4.Kumar R, Tripathi P, Baranwal M, Singh S, Tripathi S, Banerjee G. 2009. Randomized, controlled trial of oral ribavirin for Japanese encephalitis in children in Uttar Pradesh, India. Clin Infect Dis 48:400–406. doi: 10.1086/596309. [DOI] [PubMed] [Google Scholar]
- 5.Mertz GJ, Miedzinski L, Goade D, Pavia AT, Hjelle B, Hansbarger CO, Levy H, Koster FT, Baum K, Lindemulder A, Wang W, Riser L, Fernandez H, Whitley RJ, Collaborative Antiviral Study Group. 2004. Placebo-controlled, double-blind trial of intravenous ribavirin for the treatment of hantavirus cardiopulmonary syndrome in North America. Clin Infect Dis 39:1307–1313. [DOI] [PubMed] [Google Scholar]
- 6.Briolant S, Garin D, Scaramozzino N, Jouan A, Crance JM. 2004. In vitro inhibition of Chikungunya and Semliki Forest viruses replication by antiviral compounds: synergistic effect of interferon-alpha and ribavirin combination. Antiviral Res 61:111–117. doi: 10.1016/j.antiviral.2003.09.005. [DOI] [PubMed] [Google Scholar]
- 7.Mendenhall M, Russell A, Juelich T, Messina EL, Smee DF, Freiberg AN, Holbrook MR, Furuta Y, de la Torre JC, Nunberg JH, Gowen BB. 2011. T-705 (favipiravir) inhibition of arenavirus replication in cell culture. Antimicrob Agents Chemother 55:782–787. doi: 10.1128/AAC.01219-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Chung DH, Jonsson CB, Tower NA, Chu YK, Sahin E, Golden JE, Noah JW, Schroeder CE, Sotsky JB, Sosa MI, Cramer DE, McKellip SN, Rasmussen L, White EL, Schmaljohn CS, Julander JG, Smith JM, Filone CM, Connor JH, Sakurai Y, Davey RA. 2014. Discovery of a novel compound with anti-Venezuelan equine encephalitis virus activity that targets the nonstructural protein 2. PLoS Pathog 10:e1004213. doi: 10.1371/journal.ppat.1004213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Schroeder CE, Yao T, Sotsky J, Smith RA, Roy S, Chu YK, Guo H, Tower NA, Noah JW, McKellip S, Sosa M, Rasmussen L, Smith LH, White EL, Aube J, Jonsson CB, Chung D, Golden JE. 2014. Development of (E)-2-((1,4-dimethylpiperazin-2-ylidene)amino)-5-nitro-N-phenylbenzamide, ML336: novel 2-amidinophenylbenzamides as potent inhibitors of Venezuelan equine encephalitis virus. J Med Chem 57:8608–8621. doi: 10.1021/jm501203v. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Bernard KA, Klimstra WB, Johnston RE. 2000. Mutations in the E2 glycoprotein of Venezuelan equine encephalitis virus confer heparan sulfate interaction, low morbidity, and rapid clearance from blood of mice. Virology 276:93–103. doi: 10.1006/viro.2000.0546. [DOI] [PubMed] [Google Scholar]
- 11.Durocher Y, Perret S, Kamen A. 2002. High-level and high-throughput recombinant protein production by transient transfection of suspension-growing human 293-EBNA1 cells. Nucleic Acids Res 30:E9. doi: 10.1093/nar/30.2.e9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Towner JS, Paragas J, Dover JE, Gupta M, Goldsmith CS, Huggins JW, Nichol ST. 2005. Generation of eGFP expressing recombinant Zaire ebolavirus for analysis of early pathogenesis events and high-throughput antiviral drug screening. Virology 332:20–27. doi: 10.1016/j.virol.2004.10.048. [DOI] [PubMed] [Google Scholar]
- 13.Leung DW, Borek D, Luthra P, Binning JM, Anantpadma M, Liu G, Harvey IB, Su Z, Endlich-Frazier A, Pan J, Shabman RS, Chiu W, Davey RA, Otwinowski Z, Basler CF, Amarasinghe GK. 2015. An intrinsically disordered peptide from Ebola virus VP35 controls viral RNA synthesis by modulating nucleoprotein-RNA interactions. Cell Rep 11:376–389. doi: 10.1016/j.celrep.2015.03.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Wei X, Shi X, Koo I, Kim S, Schmidt RH, Arteel GE, Watson WH, McClain C, Zhang X. 2013. MetPP: a computational platform for comprehensive two-dimensional gas chromatography time-of-flight mass spectrometry-based metabolomics. Bioinformatics 29:1786–1792. doi: 10.1093/bioinformatics/btt275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lucas-Hourani M, Dauzonne D, Jorda P, Cousin G, Lupan A, Helynck O, Caignard G, Janvier G, Andre-Leroux G, Khiar S, Escriou N, Despres P, Jacob Y, Munier-Lehmann H, Tangy F, Vidalain PO. 2013. Inhibition of pyrimidine biosynthesis pathway suppresses viral growth through innate immunity. PLoS Pathog 9:e1003678. doi: 10.1371/journal.ppat.1003678. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Diaz MO, Ziemin S, Le Beau MM, Pitha P, Smith SD, Chilcote RR, Rowley JD. 1988. Homozygous deletion of the alpha- and beta 1-interferon genes in human leukemia and derived cell lines. Proc Natl Acad Sci U S A 85:5259–5263. doi: 10.1073/pnas.85.14.5259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hoffmann HH, Kunz A, Simon VA, Palese P, Shaw ML. 2011. Broad-spectrum antiviral that interferes with de novo pyrimidine biosynthesis. Proc Natl Acad Sci U S A 108:5777–5782. doi: 10.1073/pnas.1101143108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wang QY, Bushell S, Qing M, Xu HY, Bonavia A, Nunes S, Zhou J, Poh MK, Florez de Sessions P, Niyomrattanakit P, Dong H, Hoffmaster K, Goh A, Nilar S, Schul W, Jones S, Kramer L, Compton T, Shi PY. 2011. Inhibition of dengue virus through suppression of host pyrimidine biosynthesis. J Virol 85:6548–6556. doi: 10.1128/JVI.02510-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Smee DF, Hurst BL, Day CW. 2012. D282, a non-nucleoside inhibitor of influenza virus infection that interferes with de novo pyrimidine biosynthesis. Antivir Chem Chemother 22:263–272. doi: 10.3851/IMP2105. [DOI] [PubMed] [Google Scholar]
- 20.Yablonski MJ, Pasek DA, Han BD, Jones ME, Traut TW. 1996. Intrinsic activity and stability of bifunctional human UMP synthase and its two separate catalytic domains, orotate phosphoribosyltransferase and orotidine-5′-phosphate decarboxylase. J Biol Chem 271:10704–10708. doi: 10.1074/jbc.271.18.10704. [DOI] [PubMed] [Google Scholar]
- 21.Kuo RL, Kao LT, Lin SJ, Wang RY, Shih SR. 2013. MDA5 plays a crucial role in enterovirus 71 RNA-mediated IRF3 activation. PLoS One 8(5):e63431. doi: 10.1371/journal.pone.0063431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rasmussen SB, Jensen SB, Nielsen C, Quartin E, Kato H, Chen ZJ, Silverman RH, Akira S, Paludan SR. 2009. Herpes simplex virus infection is sensed by both Toll-like receptors and retinoic acid-inducible gene-like receptors, which synergize to induce type I interferon production. J Gen Virol 90:74–78. doi: 10.1099/vir.0.005389-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Qing M, Zou G, Wang QY, Xu HY, Dong H, Yuan Z, Shi PY. 2010. Characterization of dengue virus resistance to brequinar in cell culture. Antimicrob Agents Chemother 54:3686–3695. doi: 10.1128/AAC.00561-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Yeo KL, Chen YL, Xu HY, Dong H, Wang QY, Yokokawa F, Shi PY. 2015. Synergistic suppression of dengue virus replication using a combination of nucleoside analogs and nucleoside synthesis inhibitors. Antimicrob Agents Chemother 59:2086–2093. doi: 10.1128/AAC.04779-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Zhu JZ, Zhang YG, Ghosh A, Cuevas RA, Forero A, Dhar J, Ibsen MS, Schmid-Burgk JL, Schmidt T, Ganapathiraju MK, Fujita T, Hartmann R, Barik S, Hornung V, Coyne CB, Sarkar SN. 2014. Antiviral activity of human OASL protein is mediated by enhancing signaling of the RIG-I RNA sensor. Immunity 40:936–948. doi: 10.1016/j.immuni.2014.05.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Chung D. 2016. The establishment of an antiviral state by pyrimidine synthesis inhibitor is cell type-specific. J Antimicrob Agents 1:101. [PMC free article] [PubMed] [Google Scholar]
- 27.Grandin C, Hourani ML, Janin YL, Dauzonne D, Munier-Lehmann H, Paturet A, Taborik F, Vabret A, Contamin H, Tangy F, Vidalain PO. 2016. Respiratory syncytial virus infection in macaques is not suppressed by intranasal sprays of pyrimidine biosynthesis inhibitors. Antiviral Res 125:58–62. doi: 10.1016/j.antiviral.2015.11.006. [DOI] [PubMed] [Google Scholar]
- 28.Bonavia A, Franti M, Pusateri Keaney E, Kuhen K, Seepersaud M, Radetich B, Shao J, Honda A, Dewhurst J, Balabanis K, Monroe J, Wolff K, Osborne C, Lanieri L, Hoffmaster K, Amin J, Markovits J, Broome M, Skuba E, Cornella-Taracido I, Joberty G, Bouwmeester T, Hamann L, Tallarico JA, Tommasi R, Compton T, Bushell SM. 2011. Identification of broad-spectrum antiviral compounds and assessment of the druggability of their target for efficacy against respiratory syncytial virus (RSV). Proc Natl Acad Sci U S A 108:6739–6744. doi: 10.1073/pnas.1017142108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Josephson MA, Javaid B, Kadambi PV, Meehan SM, Williams JW. 2006. Leflunomide in solid organ transplantation and polyoma virus infection. Adv Exp Med Biol 577:255–265. doi: 10.1007/0-387-32957-9_18. [DOI] [PubMed] [Google Scholar]
- 30.Dunn MC, Knight DA, Waldman WJ. 2011. Inhibition of respiratory syncytial virus in vitro and in vivo by the immunosuppressive agent leflunomide. Antivir Ther 16:309–317. doi: 10.3851/IMP1763. [DOI] [PubMed] [Google Scholar]
- 31.Jung YH, Moon KC, Ha JW, Kim SJ, Ha IS, Cheong HI, Kang HG. 2013. Leflunomide therapy for BK virus allograft nephropathy after pediatric kidney transplantation. Pediatr Transplant 17:E50–E54. doi: 10.1111/petr.12029. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.