Vero (ATCC CCL 81), A549 (ATCC CL-185) and HepG2 (ATCC CRL-10741) cells were grown in Eagle’s minimum essential medium (MEM, GIBCO) containing 5% inactivated calf serum and 50 μg/mL gentamycin. Maintenance medium (MM), pH 7.5, consisted of MEM supplemented with 1.5% calf serum and gentamycin. C6/36 mosquito cells (ATCC CRL-1660) were grown in L-15 medium (Leibovitz, GIBCO) supplemented with 10% fetal bovine serum, 0.3% tryptose phosphate broth, 1% MEM non-essential amino acids and 0.02% glutamine.

Stocks of DENV-1 (strain Hawaii, GenBank KM204119.1), DENV-2 (strain NGC, GenBank MK506264.1), DENV-3 (strain H87, GenBank KU050695), DENV-4 (strain 8124, provided by the Hospital de Niños Dr. Ricardo Gutierrez, Buenos Aires, Argentina), ZIKV (strains DAK-AR-41524, GenBank KX198134.2) and PRVABC59, GenBank MK713748.1) and the clinical ZIKV isolate INEVH116141 (provided by the Instituto Nacional de Enfermedades Virales Humanas, Pergamino, Argentina) were obtained in C6/36 cells and further titrated in Vero cells by plaque formation.

Vero, A549 and HepG2 cells grown in 96 well microplates were treated with MM containing 0.5% dimethyl sulfoxide (DMSO) used as drug diluent (control) or different concentrations of BBR (Sigma-Aldrich, USA) for 24 h at 37°C and cell viability was then determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT, Sigma- Aldrich, USA) method. The 50% cytotoxic concentration (CC₅₀), the concentration of BBR that reduced cell viability by 50% respect to control cultures, was determined. Similar cytotoxicity assays were performed for the synthetic compounds U0126 (#9903 Cell Signaling, USA), an inhibitor of ERK1/2 activation, and SB202190 (#8158, Cell Signaling, USA), a compound that impairs p38 phosphorylation, to determine the range of non-cytotoxic concentrations for each inhibitor.

Vero, A549 or HepG2 cells were infected with DENV or ZIKV (MOI = 1). After 1 h of incubation at 37°C, cells were incubated with MM containing DMSO 0.5% (control) or different concentrations of BBR, U0126 (15 µM) (Cell Signaling, USA) or SB202190 (40 µM) (Cell Signaling, USA) for 24 h at 37°C. Then, cell supernatants were harvested and extracellular virus yield was determined by plaque assay in Vero cells. The 50% effective concentration (EC₅₀) of BBR, the concentration of compound that reduced viral yield by 50% respect to control cultures, was determined.

Other Vero cell cultures were infected with ZIKV (INEVH116141) or DENV-2 (MOI = 1) and after 1 h at 37°C, cells were treated with MM containing DMSO 0.5% (control), BBR (40 µM) or SB202190 (40 µM). After 24 h, cell supernatants were harvested to quantify extracellular viral titer whereas cells were covered with MM and disrupted by two cycles of freezing and thawing. After removing cell debris by low-speed centrifugation intracellular virus titer was determined by plaque assay in Vero cells.

Vero cells were infected with ZIKV (INEVH116141) or DENV-2 (MOI = 1) and DMSO 0.5% (control) or BBR (40 µM) was added at 1, 2, 4, 6 or 8 h post- infection (p.i.). In other ZIKV (INEVH116141) (MOI = 1) infected cultures the inhibitor SB202190 (40 µM) was added at 1 or 8 h p.i. In both cases, at 24 h p.i. extracellular virus yield was determined by plaque assay in Vero cells. In other experiments Vero cells infected with ZIKV (INEVH116141) or DENV-2 (MOI = 1) were treated with DMSO 0.5% (control) or BBR (40 µM) during different intervals (1–24 h p.i, 1–8 h p.i. or 8–24 h p.i.). At 24 h p.i. extracellular virus yield was determined by plaque assay in Vero cells.

Vero cells infected with ZIKV (INEVH116141) or DENV-2 (MOI = 1) were treated with DMSO 0.5% (control) or BBR (40 µM) during different intervals: 1–8 h p.i. or 8–24 h p.i. Other set of Vero cell cultures were infected with ZIKV (INEVH116141) (MOI = 1) and after 1 h of incubation at 37°C cells were treated with DMSO 0.5% (control), BBR (40 µM), U0126 (15 µM) or SB202190 (40 µM) for 24 h. In all cases cells were lysed and Western blot assay was performed as previously described (Quintana et al., 2016). The detection of viral E glycoprotein was performed using a mouse monoclonal antibody anti-flavivirus group antigen antibody, clone D1-4G2-4-15 (MAB10216, Millipore, USA). The relative protein expression level was normalized against glyceraldehyde 3-phosphate dehydrogenase (GAPDH) using a mouse anti-GAPDH (ab-9484, Abcam, UK) as primary antibody. Peroxidase conjugated anti-mouse immunoglobulins (Sigma-Aldrich, USA) were employed as secondary antibody. The analysis of ERK1/2 and p38 activation was done as previously described (Brunetti et al., 2018; Quintana et al., 2020; Brunetti et al., 2022). To examine ERK1/2 activation the primary antibodies employed were rabbit anti-phospho-ERK1/2 (p-ERK) (#9101, Cell Signaling, USA) and rabbit anti-ERK1/2 (ERK) (sc-514302, Santa Cruz Biotechnology, USA), whereas p38 activation was determined using rabbit anti-phospho-p38 (p-p38) (# 9211, Cell Signaling, USA) and rabbit anti-p38 (#8690, Cell Signaling, USA) as primary antibodies. In both cases, peroxidase-conjugated anti-rabbit IgGs (Promega, USA) were employed as secondary antibody. Protein bands were detected by chemiluminescence and quantified using ImageJ software.

Aliquots of ZIKV (INEVH116141) or DENV-2 suspensions were incubated for 2 h at 37°C with MM containing DMSO 0.5% (control) or different concentrations of BBR. Then, samples were diluted in MM, in order to achieve a low BBR concentration (0.1 µM) that is below the active antiviral concentrations, and remaining infectivity was determined by plaque assay in Vero cells.

Data represent the mean values obtained from three independent experiments. Duplicate samples were performed in each experiment. Statistical significance of the differences between control and drug-treated cultures was determined either by 2-tailed paired Student’s t-test or one-way ANOVA followed by Dunnett’s multiple comparison tests (Prism 7, GraphPad Software, Inc.). A p-value <0.05 was considered to be statistically significant; *p < 0.05, **p < 0.01 and ***p < 0.001.

In order to assess the antiviral activity of BBR against ZIKV and DENV we first tested the effect of the compound on the viability of Vero cell monolayers. The incubation of cell cultures with BBR, within the concentration range 6.5–200 µM for 24 h did not exert cytotoxic effects (Figure 1A). By contrast, BBR treatment caused a dose dependent inhibition of virus yield in cells infected with the ZIKV clinical isolate INEVH116141 (Figure 1B) or the ZIKV strains PRVABC59 (Figure 1C) and DAK-AR-41524 (Figure 1D). The lowest value of EC₅₀ was obtained against PRVABC59 strain, whereas similar susceptibility to BBR was observed for ZIKV (INEVH116141) and ZIKV (DAK-AR-41524) (Table 1). More than 99.9% of virus titer inhibition was achieved after treatment with 40 µM BBR for all the ZIKV strains assayed. On the other hand, BBR exhibited effectiveness against all DENV serotypes (Figures 2A–D), with EC₅₀ values between 2.4 and 11.6 µM (Table 1). Treatment with the maximal concentration of BBR tested (40 µM) reduced virus yield by 98.0%, 99.1%, 96.4% and 92.4% for DENV-1, DENV-2, DENV-3 and DENV-4, respectively. We next examined the cytotoxic and antiviral effects of BBR in human cell lines. The inhibitory effect of BBR against ZIKV (INEVH116141) and DENV-2 was evident in A549 and HepG2 cells at non-cytotoxic concentrations of the compound and with similar EC₅₀ values for each virus in both cell types (Figure 3; Table 1). At the maximal concentration of BBR assayed, the antiviral effect was stronger in HepG2 than in A549 cells, with 99.3% and 99.9% inhibition of virus production for DENV-2 and ZIKV, respectively (Figures 3D, F).

Next we decided to investigate the mode of antiviral action of BBR on ZIKV and DENV infections. Taking into account that BBR was similarly effective against the four DENV serotypes, DENV-2 was chosen as a representative serotype to further analyze the antiviral activity of the alkaloid against DENV infection. In order to assess whether BBR affects early or late stages of viral multiplication cycle we investigated the effect of the time of addition of BBR on extracellular ZIKV and DENV-2 production in Vero cells. For these experiments we used the 40 µM concentration of BBR since this compound concentration exerted a significant inhibitory effect on viral yield (Figures 1B, 2B). Figure 4A shows that similar high levels of viral yield inhibition were achieved when BBR was added up to 8 h p.i. Additionally, treatment with the compound only during the first 8 h of infection did not significantly affect ZIKV or DENV-2 infectious titers measured at 24 h p.i. indicating that the compound would not affect early events of viral multiplication cycle. By contrast, the presence of BBR from 8 to 24 h p.i. markedly reduced ZIKV and DENV-2 yields (Figure 4B). To determine whether BBR impairs viral protein production we analyzed the expression of E glycoprotein by Western blot. After treatment of ZIKV or DENV-2 infected Vero cells with BBR, either from 1 to 8 h p.i. or from 8 to 24 h p.i. non-significant changes in the level of E glycoprotein expression was detected indicating that the inhibitory effect of the compound would be exerted after viral protein synthesis (Figures 4C, D). Therefore, BBR would impair either the formation of progeny infectious viral particles or the release of infectious virions to the extracellular medium. To explore these possibilities we investigated the effect of BBR on intracellular and extracellular infectivity. For this purpose, ZIKV or DENV-2 infected cells were treated with BBR for 24 h and after harvesting culture supernatants to quantify released virions, cells were treated with low pH to inactivate viral particles adsorbed to cell membrane and then cell monolayers were disrupted to determine the number of intracellular infectious viral particles. Treatment with BBR caused 89.8% (for ZIKV) and 89.2% (for DENV-2) of inhibition of intracellular viral titer indicating that the alkaloid impairs the formation of mature viral particles. On the other hand, a higher level of inhibition was detected in extracellular titers in comparison with intracellular ones. Extracellular viral titer was reduced by 99.1% and 98.9% in ZIKV and DENV-2 infected cells, respectively (Figures 5A, B). These results may suggest that the release of mature virions might be also affected. As it was mentioned in section 1, it has been described that BBR exhibits a direct inactivating effect on ZIKV and DENV-2 particles (Robinson et al., 2018; Batista et al., 2019; Ratanakomol et al., 2021) and this property may contribute to the strong reduction of extracellular viral titers. Therefore, we examined whether BBR displays virucidal action against ZIKV (INEVH116141) and DENV-2 under our experimental conditions. As it is shown in Figure 5C, the incubation of ZIKV or DENV-2 particles with different concentrations of BBR for 2 h at 37°C caused the inactivation of viral particles in a dose dependent manner corroborating that besides its inhibitory action, BBR exhibits inactivating properties mainly at high compound concentrations. Maximal virucidal action, 95.7% for ZIKV and 67.3% for DENV-2, was observed at 160 µM of BBR (Figure 5C), whereas at 40 µM of BBR, the compound concentration used in the mechanistic studies, the reduction of viral particle infectivity was 60.7% and 54.6% for ZIKV and DENV-2, respectively.

Finally we examined the effect of BBR on the activation state of ERK1/2 and p38 MAPK pathways. In these assays we also employed two well-known MAPK inhibitors: the compound U0126, which is a potent non-ATP competitive inhibitor of MEK, the upstream kinase that phosphorylates ERK1/2 and the compound SB202190 that inhibits the phosphorylation of p38 (Davies et al., 2000). In a previous report we demonstrated that DENV-2 was able to induce p38 activation in Vero cells and we also confirmed that treatment with SB202190 inhibited viral multiplication and p38 phosphorylation (Quintana et al., 2020). Therefore, we decided to analyze the effect of BBR and the inhibitors of ERK1/2 and p38 kinases on MAPK activation and viral production in the context of ZIKV infection. To this end, Vero cells infected or not with ZIKV (INEVH116141) were treated with BBR (40 µM), U0126 (15 µM) or SB202190 (40 µM). ZIKV infection promoted the activation of both ERK1/2 and p38 signaling pathways (Figures 6A, B). While BBR treatment did not alter basal levels of ERK1/2 phosphorylation in uninfected cultures the alkaloid produced a significant inhibition of p-ERK1/2 in ZIKV infected cells. On the contrary, the inhibitor U0126 caused a significant diminishment in p-ERK1/2 level both in ZIKV infected and uninfected cells (Figure 6A). On the other hand, the presence of BBR or SB202190 reduced p-p38 levels both in uninfected and ZIKV infected cells (Figure 6B) suggesting that BBR might affect p38 activation in a virus independent manner. In addition, the three assessed compounds markedly reduced viral yield after treatment for 24 h. Whereas BBR and SB202190 inhibited extracellular viral production by 99.7% and 99.8%, respectively, treatment with U0126 reduced viral yield by 94.0% (Figure 6C). In line with the results obtained with BBR (Figure 5A), besides the strong inhibition on extracellular virus production, treatment with SB202190 from 1 to 24 h p.i. also exerted a significant diminishment (97.8%) of intracellular virus titer with respect to untreated infected cells (Figure 6D). Furthermore, as it was observed in BBR treated cultures (Figure 4A) late addition of SB202190 at 8 h p.i. still caused a high reduction (98.5%) of viral yield respect to the control (Figure 6E).