Inhibitory effects of an aqueous extract from Cortex Phellodendri on the growth and replication of broad-spectrum of viruses in vitro and in vivo

Background Cortex Phellodendri (C. Phellodendri), the dried trunk bark of Phellodendron amurense Ruprecht, has been known as a traditional herbal medicine, showing several bioactivities. However, antiviral activity of C. Phellodendri aqueous extract (CP) not reported in detail, particularly aiming the prophylactic effectiveness. Methods In vitro CP antiviral activity evaluated against Influenza A virus (PR8), Vesicular Stomatitis Virus (VSV), Newcastle Disease Virus (NDV), Herpes Simplex Virus (HSV), Coxsackie Virus (H3-GFP) and Enterovirus-71 (EV-71) infection on immune (RAW264.7) and epithelial (HEK293T/HeLa) cells. Such antiviral effects were explained by the induction of antiviral state which was determined by phosphorylation of signal molecules, secretion of IFNs and cytokines, and cellular antiviral mRNA expression. Furthermore, Compounds present in the aqueous fractions confirmed by HPLC analysis and evaluated their anti-viral activities. Additionally, in vivo protective effect of CP against divergent influenza A subtypes was determined in a BALB/c mouse infection model. Results An effective dose of CP significantly reduced the virus replication both in immune and epithelial cells. Mechanically, CP induced mRNA expression of anti-viral genes and cytokine secretion in both RAW264.7 and HEK293T cells. Furthermore, the main compound identified was berberine, and shows promising antiviral properties similar to CP. Finally, BALB/c mice treated with CP displayed higher protection levels against lethal doses of highly pathogenic influenza A subtypes (H1N1, H5N2, H7N3 and H9N2). Conclusion CP including berberine play an immunomodulatory role with broad spectrum antiviral activity, due to induction of antiviral state via type I IFN stimulation mechanism. Consequently, C. Phellodendri could be a potential source for promising natural antivirals or to design other antiviral agents for animal and humans. Electronic supplementary material The online version of this article (doi:10.1186/s12906-016-1206-x) contains supplementary material, which is available to authorized users.


Background
Several viruses cause malignant diseases worldwide, resulting in significant mortality and economic losses. For instance, Influenza A viruses are responsible for seasonal epidemics and have caused three pandemics in the 20th century (1918, 1957, and 1968) [1]. In addition, viral diseases cause serious threats to livestock sector and animal welfare together with productivity losses, uncertain food security and negative impacts on human health [2].
Although various prophylactic or therapeutic drugs and vaccines have been developed to prevent and treat viral diseases, the emergence of novel mutants or resistant virus strains reduces their efficacy and lead to the public health problems. Especially, with a view to provide effective methods for preventing viral diseases, researches have been attempted to identify novel materials with antiviral activities from natural or synthetic resources.
Natural products are one of relevant sources for antiviral substances. Therefore, many screening efforts have been made to identify antiviral substances from natural sources with high efficacy, low toxicity and minor side effects. Hence, treatments with herbal extracts having antiviral activities may assist to recover many animals through effective feed conversions and weight gains to result better economic returns. Consequently, as an alternative to conventional chemical agents, a large number of phytochemicals which containing extracts or pure substances from medical plants have been recognized as antivirals [3].
To identify potential antiviral agents, 200 natural oriental herbal medicines were screened and among them, Cortex Phellodendri (C. Phellodendri) found to display a wide array of antiviral activities. C. Phellodendri (Huangbai in Chinese), the dried bark of Phellodendron amurense Ruprecht (Family: Rutaceae), is one of the 50 fundamental herbs of traditional Chinese medicine and has been used against osteoarthritis, weight loss, obesity, diarrhea, diabetes, pneumonia and eye infections for a long period of time [4]. This herb is widely found in China and the Korean peninsula, and a wide range of primary scientific articles are already available on the activities of extracts of Phellodendron bark, although the underline mechanisms in the therapeutic process remain unclear. Further, the antiviral activity of C. Phellodendri has not been scientifically described.
In this study, the antiviral activities of C. Phellodendri aqueous extracts (CP) against wide array of viruses in vitro and in vivo were evaluated. Additionally, the immune-modulatory potential of C. Phellodendri which regulates the antiviral immune response was confirmed.

Plant materials and total aqueous extract preparation
Crude plant material, the dried bark of Phellodendron amurense Ruprecht (specimen number: Phellodendron amurense Rupr., PSNA200005151048; Department of Pharmacology at Busan National University, Korea), was purchased from Jaecheon Oriental Herbal Market (Jaecheon, Korea) and verified by Professor Ki-Hwan Bae at the College of Pharmacy, Chungnam National University. Water-soluble herbal extract of C. Phellodendri was prepared by Vitabio Corporation (Daejeon, Korea).

Cytotoxicity assay (CC 50 ) of Cortex Phellodendri in vitro
The CC 50 assay was performed in 72-well tissue culture plates and the CC 50 was determined through trypan blue exclusion test as described previously [6]. Increasing concentrations (0.1-16 μg/ml) of the extract were added to 75-80 % confluent RAW264.7, HEK293T and HeLa cell monolayers. After 24 h, the cell viability of each treatment group was determined by trypan blue staining. Concentrations of the extract plotted against the cell viabilities, and CC 50 was calculated as the concentration of the extract resulting 50 % cell viability.
Virus titration from Cortex Phellodendri-treated cell supernatants or cells Viral titers were determined with cell supernatants (VSV-GFP and H3-GFP) or cells (PR8-GFP and HSV-GFP) in Vero cells using a standard plaque assay. EV-71 titer measured by the 50 % tissue culture infectious doses (TCID 50 ) using HeLa cells with some modifications. Briefly, 75-80 % confluent HeLa cells grown in 96-well micro-titer plates were infected with 10-fold serial diluted supernatants (50 μl/well) collected from the infected cells. At 2 hpi, DMEM (10 % FBS) containing L-1-tosylamido-2-phenylethyl chloromethyl ketone (TPCK) trypsin (Thermo Fisher Scientific, Rockford, USA) was added to the infected wells and incubated for another 2 days. CPE were observed daily under 200X magnification, and the titers were determined by TCID 50 .
Oral administration of Cortex Phellodendri and virus challenge in BALB/c mice Five-week-old 52 BALB/c mice (16 ± 1 g) were used to conduct in vivo experiments (Table 1) For the survival rate analysis, 40 mice were grouped in to 4 main groups (n = 10), each containing 2 subgroups, control (n = 5) and CP treated group (n = 5). For the lung virus titration another 12 mice were grouped into 2 sub-groups, control (n = 6) and CP treated group (n = 6), and lung tissue was extracted at 3 (n = 3) and 5 dpi (n = 3) from each control and treated group. CP were orally administered to the mice (0.8 μg/g body weight) in a total volume of 200 μl at 1, 3 and 5 days before infection. PBS (200 μl) was orally administered to control groups mice. Treatment and challenge experiments were conducted in an approved BSL-2 + laboratory facility. Mice were intranasally infected with 5 times 50 % minimum lethal dose (MLD 50 ) of H1N1, H5N2, H7N3 or H9N2 in 20 μl of PBS per mouse. Body weight and survival were monitored up to 13 dpi. Mice exhibiting more than 25 % body weight loss were considered as an experimental end point and were humanely killed. Animal study was conducted under appropriate conditions with the approval of the Institutional Animal Care and Use Committee of Bioleaders Corporation, Daejeon, South Korea, Protocol number: BSL-ABLS-13-008.

Determination of lung viral titers
Lung tissues from euthanized mice were collected aseptically, and lung viral titer was calculated by the Reed and Muench method and expressed as Log 10 TCID 50 /Lung. In the modified hemagglutination assay (HA), instead of turkey red blood cells (RBC), 0.5 % chicken RBC was used and wells containing HA were scored as positive [10].
Determination of the effects of Cortex Phellodendri on Type I IFN-related protein phosphorylation by Immunoblot analysis

Level of mRNA induction by Cortex Phellodendri in vitro
RAW264.7 or HEK293T cells were treated with DMEM supplemented with media alone (negative control), DMEM with 1000 units/ml rmIFN-β or rhIFN-β (positive control), or DMEM with 1.0 μg/ml CP and the cells were harvested at 0, 3, 6, 12, and 24 hpt. The total RNA isolation, cDNA synthesis, and visualization of band intensity after PCR were as previously described. Additionally, different cDNA levels from HEK293T cells were measured by quantitative RT-PCR (qRT-PCR) using a QuantiTech SYBR Green PCR kit (Qiagen, Seoul, Korea) on a Mygenie96 thermal block (Bioneer, Daejeon, Korea). The PCR primers are listed in Tables 3 and 4.

HPLC analysis for chemical characterization of aqueous plant extract of Cortex Phellodendri
The liquid chromatography-mass spectrometry (LC-MS) analysis was performed on an Agilent 1200 Series highperformance liquid chromatography (HPLC) (Agilenttechnologies CO, USA) connected to a linear ion trap mass spectrometer 4000 QTRAP system (AB Sciex CO., Canada) equipped with an ESI TurboV source. Chromatography was carried out on a column ZORBAX Eclipse XDB-C18 (4.6 x 150 mm, I.D. Determination of antiviral characteristics of berberine PA-6 antiviral activity and cytokine inducing ability was further tested in immune (RAW264.7) cells using an effective dose (10.0 μg/ml). Inhibition of virus replication (PR8-GFP) in RAW264.7 cells, virus titration and induction of cytokine secretion were monitored according to the protocols described above sections.

Statistical analysis
Statistical analysis was performed using GraphPad Prism software version 6 for Windows (GraphPad Software, USA). Data are presented as the means ± standard deviations (S.D.) and are representative of at least three independent experiments. Differences between groups were analyzed by analysis of variance (ANOVA). Unpaired ttest was performed at each time points to compare the control and NDV infected extract-treated groups. Results for percent initial body weight were also compared by Student's t test. Comparison of survival was done by log-rank test using GraphPad Prism 6 version. P < 0.05 were regarded as significant.

Results
Inhibition of viral replication in Cortex Phellodendri-treated RAW264.7 cells In vitro antiviral effects of CP were evaluated by observing the amount of GFP expressed in the infected cell to analyze the effect on RNA and DNA viral replications. RAW264.7 cells treated with CP exhibited a marked reduction in GFP expression as compared to the untreated groups against PR8-GFP (Fig. 1a), VSV-GFP (Fig. 1b) and NDV-GFP (Fig. 1c). GFP expression was significantly reduced in extract-treated cells compared to the untreated groups (data not shown), which correlated with viral titers. CP treatment reduced the viral titers by 2.3and 2-fold against PR8-GFP and VSV-GFP, respectively ( Fig. 1a and b). Importantly, CP-treated cells had ≥ 70 % cell viability within 24 hpi for all tested viruses compared to untreated cells which had significantly higher cell death following virus infection (Fig. 1a-c). M-gene mRNA expression which represent the NDV virus replication, decreased time coarsely in extract treated cells compared to untreated groups ( Fig. 1c right panel).
These results indicate that the CP has potential to inhibit VSV-GFP, PR8-GFP and NDV-GFP replication in RAW264.7 cells.

Inhibition of viral replication in Cortex Phellodendri-treated HEK293T and HeLa cells
CP antiviral activity on non-immune epithelial cell lines such as HEK293T and HeLa cells was determined. Upon pre-treatment of the CP, HEK293T cells displayed marked reduction in GFP expression and viral titers by 6-and 2.6fold against VSV-GFP and HSV-GFP, respectively, at 24 hpi compared to untreated groups ( Fig. 2a and b). Moreover, CP antiviral activity in HeLa cells against H3-GFP and EV-71 was determined. CP pre-treatment exhibited markedly reduced H3-GFP expression and reduced viral titers by 4.5-fold at 24 hpi together with significant reduction in cell death in the extract treated group compared to untreated group (Fig. 2c). Furthermore, EV-71 induced CPE were markedly inhibited and upon pre-treatment with the extract, a low level of viral replication was observed (Fig. 2d). Collectively, it is also clear that the CP has ability to suppress DNA or RNA viral replication in epithelial cell lines.

EC 50 and CC 50 of Cortex Phellodendri extracts in vitro
As shown in Table 2 CP had CC 50 of 7.31 ± 0.12 μg/ml, 6.36 ± 0.06 μg/ml and 8.06 ± 0.04 μg/ml in Raw264.7, HEK293T and HeLa cells, respectively (Table 2). Importantly, CC 50 of CP were several magnitudes higher than the EC 50 of tested viruses in their respective cell lines. Selection indexes (SI) of CP for PR8, VSV-GFP and NDV-GFP in RAW264.7 cells were 8.2, 10.2 and 12.8, respectively; for VSV-GFP and HSV-GFP in HEK293T cells were 10.1 and 7.7, respectively; and for H3-GFP and EV-71 in HeLa cells were 10.7 and 20.9, respectively. These results suggest that CP is relatively non-toxic and could be expedient as a wide prophylactic and/or therapeutic potentials.

Protection against influenza A virus infection by oral administration of Cortex Phellodendri in Balb/c mice
Minimum effective dose (0.8 μg/g body weight) was chosen based on our previous in vivo experimental experiences with various other herbal extracts (Data not shown), and groups of BALB/c mice were infected with lethal doses (5MLD 50 ) of H1N1, H5N2, H7N3 and H9N2.  Around 4 dpi, most of the control group mice exhibited severe clinical signs of respiratory disease. Additionally, mice showed obvious clinical signs, including decreased activity, huddling, hunched posture, and ruffled fur. Further, body weights were decreased progressively and the control groups succumbed to death by 9 dpi for all of the viruses tested. Strikingly, CP-treated mice showed ≤ 20 % body weight loss between 5 and 7 dpi and began to recover their lost weight by 8 dpi, returning to their normal state by 13 dpi (Fig. 3). All the groups that were orally inoculated with the herbal extract had higher survival rates (80 % survival for H1N1 and H5N2 and 60 % survival for H7N3 and H9N2) than control group. Survived mice in these groups did not show obvious clinical signs except slight weight losses. Lung tissues were collected from the H1N1-infected experimental group to examine the ability of CP to inhibit viral replication in the lung. In the control group, H1N1 replicated efficiently with viral titer of 4.19 and 5.34 log TCID 50 /Lung at 3 dpi and 5 dpi, respectively. In comparison, the viral burdens in CP treated groups were several folds lower than control group which had 2.3 and 1.9 log TCID 50 /Lung, at 3 dpi and 5 dpi, respectively (Fig. 3e). These findings suggest that CP can induce sufficient inhibition of viral replication in lung and promoted the survival of mice against lethal infections of diverse influenza A viruses.
Cortex Phellodendri extracts induced the secretion of IFN-β or pro-inflammatory cytokine and activation of Type I IFN signal molecules To determine the effect of CP on inhibition of the viral replication, the levels of Interferon-β (IFN-β) and proinflammatory cytokines were measured in the CP-treated cell supernatant of RAW264.7 and HEK293T cells (Fig. 4). CP (1.0 μg/ml) induced high levels of secreted TNF-α, IL-6, IL-12 and IFN-β, similar to the pattern observed in IFN-β treated cells in RAW264.7 cells (Fig. 4a). Though IL-6 and IFN-β secretion levels were lower than IFN-β treated cells, CP-treated HEK293T cells secreted detectable levels of IL-6 and IFN-β (Fig. 4b). Consequently, these data indicated that CP can mediate the antiviral responses in immune or epithelial cells by triggering the expression of IFNs and pro-inflammatory cytokines which may stimulate the antiviral state in cells, playing a crucial role in inhibiting viral replication. Phosphorylation levels of IFN related signal molecules and NF-kB activation related molecule (p65) were measured in CP-treated RAW264.7 cells to determine the type I IFN signal transduction. As shown in Fig. 4c, CP significantly up-regulated the IRF-3, STAT1, TBK1, p65, p38 and ERK phosphorylation, the main signaling molecules in the type I IFN and NF-kB pathways. Especially the phosphorylation of IRF3, the key indicator of IFN signal transduction was induced at 8 hpt, and this effect markedly increased with the time (Fig. 4c). Furthermore, increased STAT-1 phosphorylation indicates the transcriptional activation of interferon-stimulating genes (ISGs) which involved in controlling viral infection. Additionally, CPtreated RAW264.7 cells obviously activated NF-kB (p65), leading to strong secretion of pro-inflammatory cytokines. CP was tested with LPS (positive control) using a Limulus Amebocyte Lysate (LAL) assay. Only a trace amount of endotoxin was found in CP (Fig. 4d), thus confirmed that the induction of type I IFN and pro-inflammatory signaling was not due to LPS stimulation.  (Fig. 5a, b). Moreover, most of the antiviral or inflammation related gene transcription levels were up-regulated in HEK293T cells (Fig. 5c). Especially, IFN-β, Mx1, GBP-1, IL-6, IL-8 and TNF-α mRNA expression levels at 12 hpt were increased by 15-, 14-, 6-, 20-, 12-and 9-fold, respectively. Additionally, extracts induced the transcription of ISGs, such as ISG-15 and ISG-56 (23-and 35-fold), respectively at 12 hpt compared to non-treated cells. The observed patterns were similar to the pattern of the rhIFN-βtreated control (Fig. 5c). As a result, CP has the capacity to up-regulate IFN-β, ISGs and various antiviral gene transcriptional levels and this activation may have a direct relationship with the antiviral abilities of the extract, which were observed in RAW264.7, HEK293T and HeLa cells.

Induction of antiviral gene expression by Cortex Phellodendri
HPLC analysis of aqueous plant extract of Cortex Phellodendri, identification of berbrine and effects of berberine on virus replication To investigate the critical components in CP, quantitative analysis was carried out by a reversed-phase HPLC method for the various fractions present in the CP and 6 fractions (PA-1, PA-2, PA-3, PA-4, PA-5 and PA-6) were successfully purified. Among 6 fractions, PA-6 (berberine) was detected at a wavelength and retention time of 280 nm & 12.885 min respectively, and 47.7 μg/mL of PA-6 present in 16.8 mg/mL (w/v) of CP (Fig. 6). The typical full-scan ESI-MS analyses of berberine and PA-6 is described in Additional file 1: Figure S1. These ESI-MS data compared to those reported in the literature [11][12][13], and PA-6 was identified as berberine. The antiviral effect of identified fractions (PA-1 to PA-6) was tested in RAW264.7 cells. An effective dose (10.0 μg/ml) was selected based on our preliminary experiments on the efficacy of purified fractions (Data not shown) and viral replication was observed with PR8-GFP in response to pre-treatment of various fractions. Remarkably, PA-6 inhibited the GFP expression compared to untreated groups and the viral titers were reduced by 3-fold at 24 hpi (Fig. 6c). Additionally, to find out whether PA-6 is the main compound in CP to stimulate cytokine secretion, IFN-β, IL-6 and TNF-α levels were measured after treatment of PA-6. Markedly, levels of IFN-β and cytokines secretion increased in RAW264.7 cells (Fig. 6d). These results suggest that berberine which is a major constituent of CP has the potential to inhibit the virus replication by inducing the antiviral state in the cells.

Discussion
In the present study, the broad spectrum antiviral activity of medicinal herb C. Phellodendri both in vitro and in vivo was evaluated to suggest the rationale for its immunemodulatory and biological properties. C. Phellodendri has been reported as a well-known source for the variety of therapeutics with rare side effects after a dose [14], and in the tested cell lines, CP also did not show any significant cytotoxic effect. Moreover, the cell CC 50 of CP was several magnitudes higher than the EC 50 of tested viruses and the SI of the herb for various viruses indicate the higher safety margin of the extract for therapeutic and/or prophylactic purposes (Table 1).
In the present study, the CP inhibited the replication of PR8-GFP (Fig. 1a), VSV-GFP (Figs. 1b and 2a), NDV-GFP (Fig. 1c), HSV-GFP (Fig. 2b), H3-GFP (Fig. 2c) and EV-71 (Fig. 2d) in immune (RAW264.7) and epithelial (HEK293T and HeLa) cells. Interestingly, oral administration of CP not  only increased the survival rate of mice subjected to lethal challenges with different influenza A virus subtypes including H1N1, H5N2, H7N3 and H9N2 but also led to rapid weight recovery (Fig. 3). Although CP-inoculated mice initially displayed little weight reduction, majority did not lose more than 25 % of their body weight. However, all of the mice in the control groups displayed more than 25 % losses within 9 dpi. Collectively, it suggest that CP is strong enough to inhibit the viral replication and promoted the survival of mice against lethal infections of diverse influenza A viruses. Upon viral infection, host cells initially recognize the virus infection and rapidly evoke the antiviral innate immune responses including secretion of type I IFN and pro-inflammatory cytokines [15]. Secreted IFNs and cytokines induce an antiviral state which is important to protect the host cells against invading viruses, as induction of antiviral state in early time of virus infection is critical to control the spread and pathogenesis of viruses [16]. In this study, CP induces an antiviral state via the induction of type I IFN and pro-inflammatory cytokines in vitro. To elucidate the delineate features of CP in antiviral signaling, the effect of CP on the phosphorylation of IRF-3, p65, TBK1, STAT1, ERK and p38, the important regulatory molecules which present in the type I IFN [17][18][19][20] and NF-kB [21,22] signaling pathways were studied and it shown to be up-regulated. Additionally, CP up-regulated the transcriptional levels of IFN-β, ISGs a b d c Fig. 6 Chemical characterization of Cortex Phellodendri by HPLC analysis and effects of PA-6 on virus replication. a Chemical compounds in CP was analyzed by the reversed-phase HPLC. b RAW264.7 cells treated with media alone, 1.0 μg/ml CP, 10.0 μg/ml PA-6, or 1000 unit/ml recombinant mouse IFN-β at 12 h prior to infection with PR8-GFP (MOI = 1.0). GFP expression images were obtained at 12 and 24 hpi (200X magnification). c PR8-GFP titer was determined by standard plaque assay. d Secreted cytokines of murine IFN-β, IL-6 and TNF-α measured by ELISA. Tests were performed in triplicate. The data shows representative means ± SD. (*P < 0.05 indicates a significant difference between groups) and various antiviral genes. These findings confirm that CP can induce the innate immune responses via type I IFN and NF-kB signaling for controlling viral replication.
In this study, murine macrophages (Raw264.7) were used to evaluate the antiviral effect of CP and also used epithelial cells such as HEK293T or HeLa cells which are very susceptible for viral replication. Recently, it is acknowledged that HEK293T [23] and HeLa [24] cells have less prominent pattern recognition receptors (PRRs), chiefly Toll-like receptors (TLRs). Thus it could postulate that CP may contain active components which can bind or penetrate the cell membrane and those active components may stimulate the cell surface PRRs, cytoplasmic PRRs or both and ultimately, can induce the antiviral immune responses in immune cell or epithelial cells. A bacterial endotoxin, LPS (or lipid A-associated protein) is a known immunomodulator and is often a contaminant in biological preparations. Therefore, it could speculate that macrophage-stimulating properties of CP might be due to contamination from bacterial endotoxin [25]. In this study, CP was tested with LPS using a LAL assay and was found only a trace amount of endotoxin (Fig. 4d), and antiviral function of CP was not due to LPS stimulation.
The constituents of CP are numerous and diverse, including aporphine alkaloids, protoberberine alkaloids, flavonoids, berberine, oxyberberine, palmatine, isovanillin, ferulic acid, limonin and various other molecules [4]. HPLC analysis was conducted to identify different active compounds present in the total aqueous fraction and found six major peaks (Fig. 6a), and six fractions (PA-1 to PA-6) which were successfully purified. Among six fractions, PA-6 was identified to have significant antiviral properties similar to a CP and later confirmed as berberine (Fig. 6c). Berberine has already been studied for its anticancer properties [26], antibacterial activity [27], antifungal activity [28] and antiviral activity [29]. Especially, it was reported that berberine inhibit influenza viral replication by reducing the viral neuraminidase activity, and reducing inflammatory responses which induced by influenza virus through inhibition of nitric oxide (NO) and repression of TNF-α and monocyte chemoattractant protein 1 (MCP-1) gene transcription [30]. Moreover, HB-13, a compound derived from berberine, also inhibits HSV-1 and HSV-2 activity [31,32]. However, in the present study, the antiviral effects of berberine was demonstrated with more specific mechanism via type I IFN stimulation. Consequently, antiviral or immunomodulatory effects of CP could be due to the cumulative effect of berberine or various other unknown active compounds present in the extract.
Modern pharmacological researches have further confirmed that alkaloids of C. Phellodendri could be used for treating many diseases. Among them, relationship between mechanisms of antiviral effects and active compounds including berberine should be required further study. Furthermore, the question of how to address the need for both individualizing (the basis of TCM) and standardizing (the basis of modern pharmacology) the treatments with C. Phellodendri must be settled. Once these issues are resolved, the prospects will be exists for widespread use of C. Phellodendri as a safe, effective, and affordable form of healthcare.

Availability of data and materials
The datasets supporting the conclusions of this article are presented in this main paper and supplementary materials.
Authors' contributions JHK and PW designed and performed all virus infection experiments, analyzed the data; MSK, CN, BHL, MBU, THK, JEY executed cell biological experiments and animal experiments; JYM and CP analyzed the data. HK and JSL designed the overall study and wrote the paper. All authors read and approved the final manuscript.

Competing interests
None of the authors have any financial or personal relationships with other people or organizations that could inappropriately influence or bias this study.

Consent for publication
This information is not relevant.