Corrigendum: The citrus flavanone naringenin impairs dengue virus replication in human cells

Scientific Reports 7: Article number: 41864; published online: 03 February 2017; updated: 14 March 2017 In the original version of this Article, Claudia Nunes Duarte dos Santos and Juliano Bordignon were incorrectly affiliated with ‘Sección Virologia, Facultad de Ciencias, Universidad de La República, 11400, Montevideo, Uruguay’.

Using flow cytometry, virus titration and replicon assays, this study demonstrated the ability of naringenin to inhibit the replication of four DENV serotypes in Huh7.5 cells. Antiviral activity was evident even when naringenin was used to treat Huh7.5 cells 24 h after DENV infection. To extend our observations to a more relevant system, the antiviral activity of naringenin was also demonstrated in primary human monocytes (CD14 + ) after DENV-4 infection.

Naringenin inhibits infection with four different DENV serotypes in Huh7.5 cells. The cyto-
toxicity of naringenin is cell line-dependent. Different results have been observed in Vero and Hep2 cells 12,24 . Therefore, prior to the in vitro assessment of antiviral activity in Huh7.5 cells, cytotoxicity was evaluated using the neutral red assay 25 . Plotting cell viability against eight different concentrations of naringenin revealed concentration-dependent toxicity in Huh7.5 cells (Supplementary Figure 1). Also, a naringenin concentration of 311.3 μ M inhibited 50% of cell viability (CC 50 ). The non-toxic concentration (NTC) of naringenin was 250 μ M, which did not differ statistically from untreated control cells.
The anti-dengue activity of naringenin was evaluated using four DENV serotypes and recombinant IFN-α 2A as the positive control. Naringenin (250 μ M) was added to Huh7.5 cells during and after DENV infection to investigate the entire life cycle of the virus. After 72 h, a flow cytometry assay (FACS) revealed a reduction in the number of DENV-infected cells compared to non-treated controls (Fig. 1A,B). Additionally, the titration of DENV viable particles in cell culture supernatants using a foci-forming immunodetection assay confirmed . Flow cytometry data from three independent experiments representing the mean ± standard error (SEM) of Huh7.5 cells infected with the four DENV serotypes treated with naringenin or IFN-α 2A (B). The cell culture supernatants were titrated in C6/36 cells in a foci-forming immunodetection assay (C). Data represent the mean ± SEM from three independent experiments. One-way ANOVA and Dunnett's test for multiple comparisons (*p < 0.05 compared to DENV control).
Scientific RepoRts | 7:41864 | DOI: 10.1038/srep41864 the anti-DENV activity of naringenin (Fig. 1C). To further explore the anti-DENV effects of naringenin, seven other DENV strains belonging to the four DENV serotypes were tested. As shown in Fig. 2, the percentage of 4G2-positive Huh7.5 cells was reduced after treatment with either naringenin or IFN-α 2A for all strains tested (Fig. 2). Based on these results, the anti-DENV activity of naringenin is independent of the virus serotype or strain.
Furthermore, a concentration-response curve was performed to establish whether the activity in the antiviral assay was dependent on the amount of naringenin. According to our results, the anti-DENV activity of naringenin was concentration-dependent (Supplementary Figure 2 and Table 1). The 50% inhibitory concentration (IC 50 ) was calculated, as well as the selectivity index for each DENV serotype (SI = CC 50 /IC 50 ).

Naringenin inhibits DENV replication in Huh7.5 cells.
After demonstrating the antiviral effects of naringenin, we determined the stage of the DENV life cycle during which this compound exerts its effects. Initially, we tested the virucidal effects of naringenin and observed its inability to destroy DENV particles (Supplementary Figure 3). Next, we used the time-of-drug addition approach to observe whether the compound interfered with the early or late phases of the DENV life cycle 26 . In most cases, naringenin treatment before   Table 1. Naringenin half maximal inhibitory concentration (IC 50 ) and selectivity index (SI) for anti-DENV activity in Huh7.5 cells. SI = CC50/IC50. and during infection did not reduce the percentage of Huh7.5 DENV-infected cells, although decreases were observed in the percentages of cells infected with DENV-2/ICC265. Overall, naringenin treatment was most effective either during and after infection or only after infection with the four DENV serotypes. These results were similar to those observed with IFN-α 2A treatment after infection (Fig. 3) and suggest the ability of naringenin to impair DENV replication and/or virus maturation. To determine if naringenin affects viral replication, DENV-1 and -3 replicons (RepDV1 and RepDV3) were employed. These subgenomic RNA systems contain DENV non-structural viral proteins (NS) that enable RNA replication and translation without viable DENV particle assembly. These features make DENV replicons good tools to study virus replication for the purpose of antiviral drug development. The DENV replicon assay is used to identify antiviral compounds that specifically impair DENV replication and/or translation 27,28 . According to our results, Huh7.5 cells were successfully transfected, and naringenin reduced DENV replication with an efficiency similar to IFN-α 2A and ribavirin treatments ( Fig. 4A-C). Furthermore, DENV-RNA transfection of Huh7.5 cells did not affect cell viability (data not shown).
To determine the efficiency of naringenin after Huh7.5 cells were infected with DENV, treatment was administered at different time points after challenge with the DENV-1 serotype (FGA/89). Naringenin efficiently reduced the percentage of Huh7.5 DENV-1 infected cells even when added 24 h after the infection (Fig. 5). Also, the naringenin treatment 6 h after the DENV-1 infection reduced the DENV-titer at the cell culture supernatant (Fig. 5). Treatment with IFN-α 2A was no longer effective when added 4 h after infection, confirming the potential anti-DENV activity of naringenin (Fig. 5).

Naringenin decreases DENV infection in human monocytes.
Once naringenin anti-DENV activity was confirmed in a human cell line (Huh7.5 cells), the ability of naringenin to impair DENV replication in primary human monocytes from healthy donors was investigated. First, it was established the non-toxic concentration of naringenin for human PBMCs using annexin V/7-AAD assay and monocyte quantification (CD14 + )
Flavonoids are amply present in fruits and vegetables. Importantly, flavonoids present low toxicity in different cell lines 24,31 . Here, the CC 50 of naringenin was determined to be 311.3 μ M (84.75 μ g/mL) in Huh7.5 cells, similar to that reported by Zandi et al. 12 in Vero cells (CC 50 of 304.85 μ M; 83 μ g/mL) 12 . We also established the NTC of naringenin in Huh7.5 cells (250 μ M; 68.06 μ g/mL). In a previous study by Khachatoorian et al. 32 , naringenin was not toxic to Huh7.5 cells at concentrations between 25 and 125 μ M 32 . However, for primary human monocytes, a lower concentration of naringenin had to be used (62.5 μ M; 16.94 μ g/mL). Additionally, Zanello et al. 33 studying anti-dengue virus activity of quinic acid derivatives, had demonstrated that primary human PBMC seems to be more sensible than cell lines, confirming our observation 33 .
The flavonoid fisetin inhibits DENV-2 infection in Vero cells, whereas naringenin and rutin do not 12 . Additionally, naringenin has virucidal effects on DENV-2 12 . Surprisingly, using flow cytometry and virus titration, we demonstrated the ability of naringenin to efficiently inhibit infection by eleven different DENV strains representing the four different virus serotypes in the Huh7.5 cell line. In contrast to results reported by Zandi et al. 12 , our RT-PCR assay results did not reveal any virucidal effects of naringenin on any of the four DENV serotypes 26,34 . Zandi et al. 12 only used the DENV-2 strain (New Guinea C), while we employed an RT-PCR assay and recent clinical isolates of the four different DENV serotypes to investigate the virucidal effects of naringenin. These differences may explain the contradictory results.
To determine the stage of the DENV life cycle during which naringenin exerts its antiviral effects, time-of-drug addition experiments were carried out 26 . Aside from the partial inhibition observed with DENV-2, naringenin did not affect binding and entry into Huh7.5 cells. Indeed, naringenin appeared to target other stages of the DENV life cycle, including replication and/or maturation. To determine if naringenin impairs DENV replication, we used the DENV-1 and DENV-3 replicon systems 27,28 . In accordance to quinic acid derivatives and 2-bromo-α -ergocriptine, naringenin appears to reduce DENV replication 33,35 . Recently, a stable BHK-21 cell line carrying the chikungunya replicon was employed to demonstrate the ability of naringenin to impair chikungunya virus (CHIKV) replication 31 . It seems that some flavonoids could impair the activity of viral proteins important for virus replication, like the serine protease activity (NS2B-NS3) of dengue and Zika virus 36,37 , and also the protease (NS2) of HCV 23 . Furthermore, naringenin inhibits intracellular HCV protein production and viral assembly, in agreement with our findings 32 . Flavonoids also impair influenza A replication and spread 38 . C2 group replacement in the chemical bond between the C2 and C3 rings may interfere with the cellular response to influenza A infection via MAPK signaling pathway modulation 38 .
Even when naringenin was added to Huh7.5 cell cultures several hours after infection (6 h and 24 h), the synthesis of viable virus particles and the percentage of infected cells were significantly reduced. Naringenin controlled DENV replication as efficiently as IFN-α 2A, a known antiviral cytokine 39 . 2-bromo-α -ergocriptine impairs DENV translation and/or RNA synthesis as late as 6 h after infection 35 . Treatment of DENV-2-infected cells with NITD-618, a benzomorphan core structure, reduces infection even when administered 12 h after infection 40,41 . The effects of naringenin 24 h after DENV infection in Huh7.5 cells demonstrated the potential of this flavanone as an anti-DENV compound.
In addition to the Huh7.5 cell line, the antiviral effects of naringenin were also tested in PBMCs isolated from healthy individuals and infected in vitro. Naringenin efficiently reduced infection in primary human monocytes, the primary cell targets of DENV 42 . Thus, the antiviral effects of naringenin were also observed in primary human monocytes, reinforcing the data obtained in the cell line. However, more studies are needed in order to better understand the mechanism of naringenin action at these cells.
In conclusion, data from multiple assays (flow cytometry, viral titration and replicon system) employed to assess infection with eleven different strains representing the four DENV serotypes in two cell types (the Huh7.5 cell line and primary human monocytes) support the ability of naringenin to target DENV replication, making naringenin a suitable candidate for the treatment of DENV infection. Our results provide novel insights for the development of specific anti-DENV drugs to treat infected patients.  DENV-2/ICC-266, DENV-3/97 (GenBank EF629367), and DENV-3/98 (GenBank EF629368). Additionally, a clinical isolate from a non-fatal case of DENV-4 with hemorrhagic manifestations was employed (DENV-4/ LRV13/422; GenBank: KU513441) 43 . Virus stocks were propagated in C6/36 cells and titrated in a foci-forming immunodetection assay 44 .
Naringenin (≥ 95% purity) was purchased from Santa Cruz Biotechnology and prepared in a 100% solution of dimethyl sulfoxide (DMSO, Sigma-Aldrich, St. Louis, MO, USA). Cytotoxicity Assays. Huh7.5 cell viability was tested using a neutral red (NR; Sigma-Aldrich, Irvine, UK) uptake assay as previously described 25 . Serial dilutions of naringenin (50 mM stock) in DMEM-F12 medium and 1% DMSO were added to 2 × 10 4 Huh7.5 cells/well in a 96-well plate and incubated for 72 h. The supernatant was removed, and a solution containing neutral red (33 μ g/mL) was applied for 2 h. After incubation, the supernatant was removed, and an acidified ethanol solution was used to extract the dye and perform optical density (OD) quantification at an absorbance of 540 nm using a spectrophotometer. Data from three independent experiments were normalized using the following equation: cell viability (%) = (OD sample value − OD blank control)/(OD cell control − OD blank control) × 100. The NTC was defined as the highest concentration that did not differ statistically from untreated control cells. The concentration that inhibited viability in 50% of cells (CC 50 ) was obtained by performing nonlinear regression followed by the construction of a sigmoidal concentration-response curve (variable slope; GraphPad Prism; La Jolla, CA, USA). Additionally, dose response curves were obtained with a serial dilution starting at the NTC of naringenin. The concentration that inhibited 50% of virus infection (IC 50 ) was obtained using nonlinear regression, followed by the construction of a sigmoidal concentration-response curve (variable slope; GraphPad) and calculation of the selectivity index (SI = CC 50 /IC 50 ). All assays were performed in triplicate.
Time-of-Drug Addition Assay. Initially, Huh7.5 cells at a density of 2 × 10 4 cells/well in 96-well plates were (i) treated with naringenin for 1 h prior to DENV infection; (ii) treated with naringenin during DENV infection; (iii) treated with naringenin after DENV infection; or (iv) treated with naringenin during and after DENV infection. An MOI of 10 was used for DENV infection, and naringenin was tested at a concentration of 250 μ M. The percentage of infected cells was determined using FACS as previously described.
Naringenin Impairs DENV Replication in Huh7.5 Cells. After confirming the anti-DENV effects of naringenin after viral entry into Huh7.5 cells, we determined if this flavanone impairs DENV replication. Two subgenomic replicons derived from DV1-BR/90 (RepDV1) and BR DEN3 290-02 (RepDV3) were used in these experiments 27,28 . DENV replicon RNAs were obtained and used to transfect Huh7.5 cells (2 μ g of RNA/2 × 10 6 cells) in a Nucleofector 2B device (Lonza; Cologne, Germany) according to the manufacturer's instructions 33 . After 1 h of transfection, the Huh7.5 cells were treated with naringenin (250 μ M). The plates were further incubated for 72 h. After incubation, the cells were recovered and subjected to FACS analysis, as described previously. As a negative control, Huh7.5 cells were transfected without RNA and were not treated. Huh7 Antiviral Effects in Primary Human Cells. PBMCs obtained from healthy donors who provided signed informed consents were purified using Ficoll-histopaque and a classical protocol 46 . After purification, PBMCs were plated onto 24-well plates at a density of 1 × 10 6 cells/well and treated with 200 μ M of naringenin for 5 days and then the exposure of annexin V/7-AAD was evaluated. Also, we quantified the number of human monocytes (CD14 + ) after treatment with naringenin (250, 125 and 62.5 μ M) for five days. After it was established that the NTC of naringenin for human monocytes was 62.5 μ M, infection with dengue was performed. Human PBMCs were plated onto 24-well plates at a density of 1 × 10 6 cells/well and infected with DENV-4 TVP/360 (MOI: 10) for 2 h. The inoculum was removed, and cells were treated with naringenin (62.5 μ M), IFN-α 2A (200 IU/mL) or RPMI media. After incubation for 5 days at 37 °C and 5% CO 2 , cells were stained for DENV E protein (mAb 4G2-FITC conjugated) and with a mouse anti-human CD14 PE-Cy7 antibody 47 (BD Biosciences, USA). CD14 expression was used to gate the monocytes. The percentage of DENV E-positive cells (4G2 + ) in this population was calculated.
Data Analysis. Statistical analysis consisted of one-way ANOVA followed by Dunnett's test for multiple comparisons with a significance of p < 0.05. Statistical analyses were performed using Prism software (GraphPad version 5.0; La Jolla, CA, USA). Primary human monocyte infection was analyzed using two-way ANOVA and the Bonferroni correction for multiple comparisons with a significance level of p < 0.05. Flow cytometry data were analyzed using FlowJo version X software (Tree Star Inc.; Ashland, OR, USA).