Madin Darby Bovine Kidney (MDBK) cells and swine testicle (ST) cells were grown in high-glucose Dulbecco's Modified Eagle Medium (DMEM), which was added with 8% fetal bovine serum (FBS). The FBS was confirmed to be free of BVDV antibody with commercial BVDV antibody test kit (IDEXX Laboratories, Inc.). All cells and FBS were tested negative for the presence of Pestivirus antigen by RT-PCR (18). The cpBVDV C24V and CSFV Shimen strains were obtained from the China Institute of Veterinary Drug Control, and propagated in MDBK and ST cells, respectively.

Ginsenosides Rb1, Rb2, Rb3, Rc, Re, Rf, Rg1, Rh1, 20(S)-Rh2, pseudo ginsenoside Rh2 (PRh2), pseudo ginsenoside F11 (PF11), and NotoginsenosideR1 (NR1), screened in this study, were obtained from the Jilin University. The purity of these ginsenosides was above 98%, as estimated by high performance liquid chromatography. Stock solutions (40 mg/mL) of the compounds were dissolved in dimethyl sulfoxide (DMSO) or ethanol and were subsequently diluted in culture medium. As a negative control, DMSO was also added to all no-drug control samples.

Cytotoxicity assays were carried out as described previously (19). In short, MDBK cells, exponentially grown in the 96-well plates (approximately 5 × 10³ cells/well), were incubated in 5% CO₂ at 37°C for 24 h. Then, the medium was removed and serial dilutions of compounds were added. DMSO was used as a negative control. The cells were proliferated at 37°C for 3 days, and their overall metabolic activity was determined by the method of MTS/PMS (Promega, USA). The 50% cytotoxic concentration (CC₅₀) was calculated using a non-linear regression fitting of the data as the compound concentration necessary to reduce 50% cell viability compared to control non-treated cells.

The activity of compounds against BVDV was achieved by inhibiting the cpBVDV C24V strain induced cytopathogenicity in MDBK cells. Cells were seeded in 96-well plates and incubated overnight in growth medium in 5% CO₂ at 37°C. Then, the cells were infected with virus at 10³ TCID₅₀/mL, using a multiplicity of infection (MOI) of 0.05. After 2 h, the virus was removed and maintenance medium (DMEM supplemented with 2% FBS) with or without serial dilutions of compounds was added. Concentrations of compounds were from 6.25 to 200 μg/mL. DMSO was used as a control sample. The cells were incubated at 37°C for 3 days. Subsequently, the antiviral activity was evaluated through the viral titration and the cytopathic effect (CPE) inhibition, by MTS/PMS (Promega, USA). The 50% effective concentration (EC₅₀) was calculated using a non-linear regression fitting of the data as the compound concentration necessary to reduce 50% cytopathic effect on MDBK cells compared to DMSO treated control cells.

MDBK cells were seeded in 6-well plates and cultured overnight, then inoculated with cpBVDV strain C24V with 0.05 MOI and Rb2 or Rb3 (200 μg/mL) mixture. Uninfected cells, added with DMSO, were used as controls. The plates were incubated for 72 h at 37°C to be further submitted for BVDV RNA and E2 protein detection by real-time PCR and Western blot.

Total RNA was extracted, retrotranscribed, and quantified by the real-time PCR. Following manufacturer's instructions, TRIzol reagent was used to extract RNA (Invitrogen, China) and the M-MLV Reverse Transcriptase was used to conduct the reverse transcription reaction (Promega, USA). The specific primer of BVDV, as described previously, was used to conduct the real-time PCR (20). In brief, a 135-bp fragment from the 5′UTR of BVDV was amplified: 5′UTR forward, 5′-GGTAGCAACAGTGGTGAGTTC-3′, and 5′UTR reverse, 5′- CTCAGGTTAAGATGTGCTGTG-3′. The qPCR reactions were performed with the GoTaq qPCR master mix (Promega, USA) in the Bio-Rad iQ5 Real-Time PCR System (Bio-Rad Laboratories, USA). A 91-bp fragment of bovine β-actin mRNA in each sample was amplified and used as the endogenous control (BBA forward, 5′-CCCACACGGTGCCCATCTAT-3′, and BBA reverse, 5′-CCACGCTCCGTGAGGATCTTC-3′), in order to normalize the target amplification data (21). The 2–ΔΔCT method was used to perform the relative quantification of RNA of treated cells compared with that of untreated and infected cells. Each sample was tested in triplicate, including the positive and negative controls.

Treatment cells were washed using the icy PBS, and total protein samples were extracted using the RIPA Lysis and Extraction Buffer (Thermo Fisher Scientific, USA). According to the instructions of the manufacturer, after being centrifuged at 13,000 rpm for 20 min, concentration of proteins in supernatants was detected using a BCA Protein Assay Kit. Proteins were equally separated on 12–15% SDA-PAGE and then moved onto the PVDF membrane (Merck Millipore, USA). After being blocked in 5% skimmed milk for 2 h, the membrane was incubated with the anti-BVDV E2 antibodies (VMRD, USA) at 4°C overnight and washed, and then incubated with the HRP-conjugated goat anti-mouse IgG (Sigma, USA) at room temperature for 1 h. Signals were developed with an ECL Detection Kit (Thermo Fisher Scientific, USA).

To expound the stage of viral replication at which ginsenoside Rb2 and Rb3 exerts their activity, three different assays were performed as described previously (22, 23) with some modifications. MDBK cells were treated with Rb2 and Rb3 before virus infection (pre-treatment group), after infection (post-treatment group), or virucidal group (pre-treatment and post-treatment group) (Figure 6A). Rb2 and Rb3 were always used at the non-cytotoxic concentration (200 μg/mL) and the cpBVDV strain C24V titration was always used at 0.05 MOI.

The virucidal assay was conducted with equal amounts of BVDV and ginsenoside. The mixtures were incubated in the microcentrifuge tubes at 37°C for 2 h, and then they were added into the MDBK cell monolayers. The pre-treatment and post-treatment groups were conducted in the procedures that MDBK cell monolayers were incubated with ginsenoside or BVDV virus at 37°C for 2 h. After removal of the liquid, the BVDV or ginsenoside was added. Virus and DMSO were also included on each plate as controls. The plates were incubated at 37°C for 72 h, followed by fixation and staining with BVDV specific antibody labeled by FITC (VMRD, USA).

MDBK cells were seeded in 12-well plates and cultured overnight, then infected with cpBVDV strain C24V at a MOI of 0.05, then Rb2 (200 μg/mL) was added at different time points after infection and the cells were further incubated to complete 24 h of infection. All samples were processed for viral RNA analysis.

Using the FA-based virus inhibition assay, the antiviral efficacy of Rb2 against CSFV was investigated.

CSFV Shimen strain with 0.05 MOI was separated and added into the ST cell monolayers and allowed to inoculation for 2 h at 37°C. After inoculation, cells were washed with PBS and serial dilutions of ginsenoside Rb2 was added. Concentrations of compounds were from 25 to 200 μg/mL. Uninfected cell and virus treated with DMSO were used as controls. The plates were incubated at the temperature of 37°C for 72 h, cells were fixed with icy stationary liquid (acetone:methanol = 3:1) at 4°C for 20 min. The cells were further incubated with mouse anti-CSFV E2 mAb WH303 at 37°C for 1 h, washed three times with PBS and incubated with FITC-conjugated rabbit anti-mouse IgG (Sigma, USA) at 37°C for 30 min. After three washes, immunofluorescence was observed using a fluorescence microscope.

The BVDV and CSFV IRES were cloned into the firefly luciferase plasmids pGL4.20 (Promega, USA), respectively. The specific primer for CSFV was described previously (24). The BVDV primer sequences were: F: 5′CCGAGCTCGTATA CGAGAATTT GCCTAGGAC-3′; R: 5′ CCC AAGC TTGGCATAAACAGGTTCTTCCACCC3′. Similar 470 bp fragments of the BVDV were amplified and subcloned into the pGL4.20 using restriction enzymes with HindIII and SacI. All reconstructed plasmids were sequenced at Invitrogen Co., Ltd.

The 293T cells were grown in 96-well plates until 80% confluence, reconstructed plasmids and Renilla luciferases plasmids pGL4.74 (Promega, USA) were co-transfected with the X-treme GENETM HP DNA transfection reagent (Roche, Philadelphia, PA), using standard operation, respectively. The parental vector pGL4.20 and pGL4.74 co-transfected group served as mock controls. The cells were allowed to inoculate for 4 h at 37°C. After inoculation, cells were washed with PBS, added with the ginsenoside Rb2 (200 μg/mL) and incubated at 37°C for 48 h. The luciferase activity was determined using the dual luciferase reporter assay system (Promega, USA).

Student's t-test was conducted to examine the statistical differences among ginsenoside treatment groups, virus control, DMSO control, and mock groups. Expression of results was in the form of mean ± standard deviation, representing three independent experiments. Statistical analyses were conducted with GraphPad Prism 6 (GraphPad Software, San Diego, CA). In the case of a value of p < 0.05, it was regarded statistically significant.

Most compounds had no effect on the cell viability at 200 μg/mL, while ginsenoside 20(S)-Rh2 and pseudo ginsenoside Rh2 reduced the viability of MDBK cells significantly (Figure 1). Therefore, ginsenoside 20(S)-Rh2 and pseudo ginsenoside Rh2 were not included in further experiments.

Concerning the antiviral tests undertaken on the 10 compounds, ginsenoside Rb2 and Rb3 significantly inhibited BVDV induced CPE. CPE inhibition was 94.5 and 91.2% at the 200 μg/mL concentration. All other compounds demonstrated a limited or no inhibitory effect on CPE (Figure 2). To further determine the antiviral effect of ginsenoside Rb2 and Rb3, and calculate the EC₅₀, different concentrations of Rb2 and Rb3 (6.25–200 μg/mL) were added into the BVDV infected MDBK cells. The viral titer and CPE inhibition were inversely proportional to the ginsenoside Rb2 and Rb3 concentrations (Figures 3, 4). The EC₅₀ of Rb2 was 57.76 μg/ mL, that of Rb3 was 60.25 μg/ mL (Table 1).

BVDV replication was significantly inhibited by Rb2 or Rb3, as shown by reduction of viral RNA levels when compared to DMSO (Figure 5A). Similar results were obtained on BVDV E2 protein with Western blot (Figure 5B). Compared with the virus control and DMSO treated group, there is almost nothing or slight bands in the Rb2 and Rb3 treated groups.

The different assays performed to elucidate the mechanism of action of the ginsenoside Rb2 and Rb3, applied at different periods, showed that Rb2 and Rb3 completely suppressed BVDV infection with virucidal assay and post-treatment of cells. Pre-treatment of cells assay also showed effects on BVDV infection, but the inhibitory effect was weaker than other assays (Figures 6B,C).

In order to give a better perspective about the time of action of Rb2, time of drug addition experiments were carried out. A concentration of Rb2 of 200 μg/mL was added at different points of time after infection, and then viral replication was evaluated as the yield of viral RNA synthesis. Viral RNA was markedly reduced when Rb2 was added during the first 8 h, after which a gradual reduction in drug inhibition was observed. The antiviral effect of Rb2 can be exerted within in 14 h p.i. (Figure 7).

To further understand the ginsenoside Rb2 antiviral activity for other pestiviruses, CSFV Shimen was investigated. Using the FA-based virus inhibition assay, ginsenoside Rb2 demonstrated to have a similar inhibition ability on CSFV to that observed on the BVDV reference strain CV24V (Figure 8).

Direct antiviral activity of ginsenoside Rb2 at the level of Pestivirus IRES mediated translation in 293T cell culture was investigated. The translation efficiency in 293T cells was analyzed by a reporter RNA encoding Flufly under the Pestivirus IRES control, in the presence or absence of the Rb2. As shown in Figure 9, the luciferase activity (Flufly/Renilla) in the presence of Rb2 was significantly reduced than that of the control in absence of Rb2. But the luciferase activity of vector was not changed in the presence or absence of Rb2. These results suggest that ginsenoside Rb2 can inhibit the translation process mediated by the Pestivirus IRES.