Huh7 cells (ATCC, USA) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 5% (v/v) fetal calf serum, non-essential amino acids, 100 U/mL penicillin, and 100 μg/mL streptomycin. BSR-T7/5 (generous gifts from Klaus Conzelmann, Germany), BHK-1 (ATCC, USA), and Vero-E6 cells (ATCC, USA) were cultured in DMEM supplemented with 10% (v/v) fetal calf serum, non-essential amino acids, 1 mM sodium pyruvate, 2 mM L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin.

The SOSV minigenome segment was cloned into a Pol-I promoter-based plasmid (S1 Fig., Genbank #MG880223), and transfected into Huh7 cells in conjunction with Pol-II expression plasmids (pCAGGS) supplying SOSV polymerase (L), nucleoprotein (NP), and phosphoprotein (P) proteins in trans. The minigenome assay was performed in a 96-well plate on cells seeded at 2 × 10⁴ cells per well, with the following amounts of plasmids (150 ng total) transfected per well: 75 ng minigenome, 12.5 ng L, 50 ng NP, and 12.5 ng P. Minigenome activity was based on ZsGreen1 (ZsG) fluorescence levels determined at 72 h post transfection (hpt). To rescue recombinant virus, a T7 promoter-based plasmid containing full-length anti-genomic sense SOSV (S2 Fig.; Genbank #MG880224) was constructed. To rescue the reporter SOSV expressing ZsG, a modified genome was constructed in which the M open reading frame (ORF) was replaced with one encoding ZsG-P2A-M, leaving the M gene start and gene end intact (S2 Fig.; Genbank #MG880225). Virus rescue was performed in 6-well plates seeded with BSR-T7 cells (approximately 70% confluent) transfected with genome plasmid in conjunction with the SOSV-L, -NP, and -P expression plasmids (2 μg genome, 1 μg SOSV-L, 0.5 μg SOSV-NP, and 0.25 μg SOSV-P). At 96 hpt, cell culture supernatants were harvested, clarified by low-speed centrifugation, and used to infect BHK-21 cells. At 96 h post infection (hpi), cell culture supernatants were harvested, clarified, and titered by tissue culture infective dose 50 (TCID₅₀) assays in Vero-E6 cells based on crystal violet visualization of cytopathic effects [12].

All compounds were obtained from either Selleckchem (Houston, TX, USA) or Sigma-Aldrich (Saint Louis, MO, USA) with the exception of T-705 (BOC Sciences, Shirley, NY, USA) and 09167 (Vitas-M Laboratory, Champaign, IL, USA). For both screening assays, Huh7 cells were seeded in a 96-well plate at 2 × 10⁴ cells per well 16–20 h prior to treatment with DMSO-diluted compounds (final DMSO concentration 0.5%). For the minigenome screening assay, cells were transfected with the required plasmids using TransIT- LT-1 (Mirus Bio, Madison, WI, USA) 1 h post treatment. Total ZsG fluorescence was determined 72 hpt using a microplate reader (H1MD, BioTek Synergy, Winooski, VT, USA), with levels recorded at a height of 6 mm with 75 gain/sensitivity. ZsG levels were normalized to levels in mock-treated cells. For the reporter and wild-type screening assays, cells were infected 1 h post treatment with either rSOSV/ZsG or rSOSV at MOI 0.2. At 48 hpi, supernatants were collected (rSOSV-infected cells) for titer determination, or total ZsG fluorescence in the infected cell monolayer determined (rSOSV/ZsG-infected cells) using a H1MD plate reader (height 6 mm; 100 gain/sensitivity). For both assays, cell viability was determined by measuring ATP content in parallel with compound-treated, mock-transfected/infected cells using CellTiter-Glo 2.0 (Promega, Madison, WI, USA). All experiments were performed in quadruplicate and repeated at least 3 times.

ZsG fluorescence in infected cells was imaged directly using an EVOS digital inverted microscope (Thermo-Fisher, Waltham, MA, USA). For immunofluorescence, cells were fixed in 10% formalin for 30 min, permeabilized in PBS containing 0.1% Triton-X100 (v/v) for 10 min, and blocked with PBS 5% BSA (v/v) for 30 min. Primary antibody was rabbit α-SOSV NP (Genscript, Waltham, MA, USA), visualized with an AlexaFluor 488-labeled secondary antibody (Thermo-Fisher, #A11034). Quantification of total rSOSV protein production in Huh7 cells was performed on infected cells labeled as described above, and further stained with Nuc-Blue and Cell-Mask Red (Thermo-Fisher). Images were collected on the Operetta high-content imaging system (PerkinElmer, Waltham, MA, USA) using a 20× objective, with 16 fields per well analyzed using the Harmony software package (PerkinElmer). Total 488 fluorescence per field was determined, averaged for the well, and then normalized to values from control wells containing rSOSV-infected, mock-treated cells.

ZsG fluorescence values for compound-treated, rSOSV/ZsG-infected cells were normalized to those in mock-treated (DMSO only) cells, and used to fit a 4-parameter equation to semilog plots of the concentration-response data. From this, the compound concentrations inhibiting 50% of the ZsG expression (50% effective concentration, EC₅₀) were interpolated. Cell viability was similarly calculated in compound-treated, mock-infected cells to determine the 50% cell cytotoxicity concentration (CC₅₀) of each compound. The selectivity index (SI) was calculated by dividing the CC₅₀ by the EC₅₀. Data analysis was performed using GraphPad Prism v7. Suitability for high-throughput screening was determined using the Z prime (Z′) score, a measure of statistical effect size, with values between 0.5 and 1.0 considered acceptable [13].

SOSV Minigenome plasmid; pSOSV MG; Genbank # MG880223

Rescue plasmid for rSOSV; pSOSV FL; Genbank #MG880224

Rescue plasmid for rSOSV/ZsG; pSOSV/ZsG FL, Genbank #MG880225

We constructed a SOSV minigenome segment containing the parental SOSV 3′ and 5′ leader and trailer sequences, along with the gene start sequence for NP and the gene end sequence for L. All internal coding sequences and intergenic regions were replaced with the coding sequence for ZsGreen1 (ZsG), a green fluorescent protein (Fig 1A). Protein expression plasmids under control of the Pol-II promoter (pCAGGS) allowed in trans expression of SOSV NP, L, and P, the minimum requirement to initiate paramyxovirus replication and transcription [1]. The minigenome segment was cloned into a Pol-I expression plasmid in the negative orientation, meaning it has to be replicated first by the NP + L + P replication complex before transcription of ZsG mRNA can occur.

Next, the SOSV minigenome screen was optimized to achieve maximum reporter protein expression with the greatest signal-to-noise ratio in a 96-well plate format using Huh7 cells. Huh7 cells have previously proven useful for evaluating nucleoside analog activity against other viruses, and are highly permissive to SOSV infection [14–17]. Alongside a constant amount of minigenome plasmid, differing ratios of pCAGGS-L, -NP, and -P were transfected, and ZsG fluorescence was determined at both 48 and 72 hpt (Fig 1B). We determined the optimal ratio of plasmids for the assay to be 6:1:4:1 of genome, pCAGGS-L, -NP, and–P respectively, with ZsG fluorescence read at 72 hpt. To determine whether the optimized minigenome assay was suitable for screening antiviral compounds, we used the broad-spectrum antiviral ribavirin. Using 50 μM ribavirin as the positive control and DMSO as the vehicle-only control, the Z′ score of the assay was 0.58. Dose-dependent reduction of ZsG fluorescence was also observed in the SOSV minigenome assay using a serial 2-fold dilution of ribavirin, giving an EC₅₀ of 5.91 ± 0.85 μM with no cytotoxicity (CC₅₀ > 50 μM; Fig 1C). These data indicated that the assay was robust, and suitable for screening anti-SOSV compounds at BSL-2.

As minigenome assays can be used as surrogates for measuring viral replication and transcription, we evaluated a panel of 40 compounds predicted to inhibit these processes. These were a collection of nucleoside analogs supplemented with other select compounds reported to inhibit viral replication. Compounds were initially screened for activity at concentrations of 5000, 500, and 50 nM, with cell viability determined concurrently by assessing ATP content (S1 Table). Compounds inhibiting the minigenome by >50% whilst retaining >80% cell viability were then chosen for further analysis; these were 2′-deoxy-2′-fluorocytidine (2′-dFC), 5-fluorouridine, mycophenolic acid (MPA), and 6-azauridine. Dose-response analysis of these compounds demonstrated that both 2′-dFC and 6-azauridine inhibited minigenome activity in the absence of cell toxicity, with EC₅₀ values of 1.31 ± 0.15 μM and 5.17 ± 1.01 μM, respectively (Fig 2). While 5-fluorouridine was not cytotoxic at the concentrations evaluated, it did not completely inhibit minigenome signal even at the highest concentrations evaluated, and was therefore dropped at this point. MPA inhibited minigenome activity at sub-micromolar concentrations, but was cytotoxic at higher concentrations. However, with an SI of 7, it was incorporated into the secondary virus screen to determine if a larger therapeutic window became apparent in the context of authentic viral replication.

The single-stranded, negative-sense RNA genome of SOSV encodes 6 genes to express 7 proteins, with 1 protein generated by V/P gene mRNA editing. The intergenic regions of SOSV contain transcriptional promoter and terminator sequences. We cloned the full-length SOSV genome into an expression plasmid under control of the T7 polymerase promoter. When transfected into cells and supplied with SOSV P, NP, L, and T7 polymerase in trans, this system successfully generated infectious recombinant SOSV (rSOSV). After optimizing the rSOSV rescue system, we designed another recombinant SOSV genome capable of expressing a fluorescent reporter protein in infected cells. Previous approaches to develop recombinant paramyxoviruses expressing reporter genes have included the reporter as either a separate open reading frame (ORF), or incorporated it within a parental ORF [18–22]. We employed the latter approach, and generated a SOSV genome in which the ZsG ORF was inserted upstream of the M ORF, fused to the viral protein via the self-cleaving P2A motif taken from porcine teschovirus 1 [23]. The authentic SOSV M gene start and gene end sequences therefore transcribe a single mRNA encoding the ORF for both M and ZsG, which, upon translation, results in the expression of 2 separate proteins via the P2A ribosomal skipping motif. Incorporation of this recombinant SOSV genome into the previously described SOSV rescue system (Fig 3A) resulted in the successful rescue of the reporter SOSV (rSOSV/ZsG).

Infecting Vero-E6 cells with rSOSV/ZsG at MOI 0.1 resulted in extensive syncytia formation in the cell monolayer similar to infection with rSOSV, with strong ZsG fluorescence detectable in these infected cells (Fig 3B). To compare growth kinetics of the recombinant viruses, Vero-E6 cells were infected with rSOSV, rSOSV/ZsG, or wild-type SOSV at MOI 0.1, with titers determined by TCID₅₀ at 24, 48, 72, 96, and 120 hpi. All 3 viruses grew to similar titers over the 120 h time course, with no attenuation apparent in either of the recombinant viruses compared to the wild-type (Fig 3C). This indicated that incorporation of the ZsG ORF into the SOSV genome did not negatively impact the virus phenotype. To assess whether rSOSV/ZsG could be used to confirm antiviral activity of a compound by measuring the reduction in ZsG fluorescence in treated cells, we optimized a screening assay in Huh7 cells using 50 μM ribavirin as the positive control; the resultant Z′ score was 0.74, indicating a robust assay. A concentration-response curve using a 2-fold serial dilution of ribavirin demonstrated dose-dependent reduction in ZsG fluorescence in rSOSV/ZsG-infected cells, giving an EC₅₀ of 6.10 ± 0.09 μM with no cytotoxicity (CC₅₀ > 50 μM; Fig 3D).

To confirm the inhibitory effect of 2′-dFC, 6-azauridine, and MPA in the context of authentic viral replication, we performed counter-screening with rSOSV/ZsG. Alongside these 3 compounds, we included 09167, an antagonist of the innate immune system that was shown to inhibit NiV and other paramyxoviruses in a similar screening assay [22,24]. Huh7 cells were treated with compound 1 h prior to infection with rSOSV/ZsG at an MOI of 0.2, with ZsG fluorescence read 48 hpi. All 3 compounds reduced ZsG expression in a dose-dependent manner, with EC₅₀ values similar to those observed during the minigenome screen (Fig 4A). 2′-dFC (EC₅₀ = 1.48 ± 0.22 μM) and 6-azauridine (EC₅₀ = 7.99 ± 1.32 μM) potently inhibited ZsG fluorescence without cytotoxicity at the highest compound concentrations. While MPA had an EC₅₀ value of 0.33 ± 0.04 μM, similar to that obtained in the minigenome screen, it again had a low SI of 5 due to its cytotoxicity at higher concentrations (Table 1). The 0.67 ± 0.05 μM EC₅₀ value of 09167 was consistent with the sub-micromolar concentrations used to inhibit other paramyxoviruses, indicating that SOSV was similarly sensitive to this compound [24].

To ensure the validity of using the ZsG-expressing virus to accurately identify specific inhibitors, we investigated whether these compounds also inhibited the replication of rSOSV. Huh7 cells were treated with 2′-dFC, 6-azauridine, MPA, or 09167, and infected with rSOSV (MOI = 0.2) 1 h post treatment. At 48 hpi, cells were fixed and total SOSV protein production was visualized using rSOSV-specific and fluorescently labelled antibodies. Total fluorescence due to rSOSV protein staining per well (average of 16 analyzed fields per well) was normalized to rSOSV-infected, mock-treated cells, allowing determination of the relative reduction in virus replication due to the inhibitory effects of each compound. All compounds tested reduced rSOSV protein production in a dose-dependent manner, indicating inhibition of viral replication and growth (Fig 4B). Direct visualization of the compound-treated rSOSV infected cell monolayers clearly demonstrated the reduction in viral protein production, and also demonstrated the ability of rSOSV to form extensive syncytia in Huh7 cells at sub-optimal concentrations of compound (Fig 5). These results indicate that inhibition caused by the compounds is indeed due to a SOSV-specific mechanism, rather than one unique to the ZsG-expressing recombinant.