Berberine Hampers Influenza A Replication through Inhibition of MAPK/ERK Pathway

Background: Berberine (BBR) is an isoquinoline alkaloid which exhibits a variety of biological and therapeutic properties, and has been reported by some to block replication of the influenza virus. However, contradictory results have also been presented, and the mechanistic explanation is lacking. Methods: A panel of cell lines (Madin–Darby canine kidney (MDCK), adenocarcinoma human alveolar basal epithelial cells (A549), lung epithelial type I (LET1)) and primary human airway epithelial cells (HAE) susceptible to influenza virus infection were infected with a seasonal influenza A virus in the presence or absence of BBR. Cytotoxicity towards cell lines was measured using XTT assay. The yield of the virus was analyzed using RT-qPCR. To study the molecular mechanism of BBR, confocal microscopy and Western blot analyses of cellular fractions were applied. Results and conclusions: Our results show cell-type-dependent anti-influenza properties of BBR in vitro which suggests that the compound acts on the cell and not the virus. Importantly, BBR hampers influenza replication in primary human airway epithelium 3D cultures that mimic the natural replication site of the virus. Studies show that the influenza A virus upregulates the mitogen-activated protein kinase/extracellular signal-related kinase (MAPK/ERK) pathway and hijacks this pathway for nucleolar export of the viral ribonucleoprotein. Our results suggest that BBR interferes with this process and hampers influenza A replication.


Introduction
The influenza viruses are among the most important human pathogens, with types A and B being the most clinically relevant for humans [1,2], and type A being the sole cause of pandemics. Influenza is a severe, acute, and highly contagious respiratory system disease. The symptoms include high fever, chills, general weakness, diffuse headaches, runny nose, muscle aches, sore throat, dry paroxysmal cough, and conjunctivitis. It is estimated that yearly seasonal influenza causes 3 to 5 million severe cases, and these are associated with about 300,000 to 650,000 deaths [3].
Influenza A viruses infect a wide variety of species [4], and currently, two strains are believed to circulate in humans (A(H1N1) and A(H3N2)), while a number of strains reside in birds and believe that this observed effect is relevant in the human body. Moreover, we show that BBR blocks the influenza-induced mitogen-activated protein kinase/extracellular signal-related kinase (MAPK/ERK) pathway, which is required for the transport of viral ribonucleoproteins to the cytoplasm.
Human airway (tracheobronchial) epithelial (HAE) cells were obtained from airway specimens resected from patients undergoing surgery at the Silesian Center for Heart Diseases. The study was approved by the Bioethical Committee of the Medical University of Silesia in Katowice, Poland (approval no: KNW/0022/KB1/17/10 dated 16 February 2010). Written consent was obtained from all patients. Primary cells were cultured on plastic to generate passage 1 cells and plated at a density of 3 × 10 5 cells/well on permeable Transwell inserts supports. HAE cultures were generated by the provision of an air-liquid interface for 6 to 8 weeks to form fully differentiated, polarized cultures that resemble pseudostratified mucociliary epithelium.
Human seasonal influenza virus type A H3N2, strain A/Hong Kong/4801/2014, was obtained from the European Virus Archive EVA project (EVAg, Marseille, France). Virus stocks were generated by infecting MDCK cells at 90% confluency for 48 h. After that time, cultures were aliquoted and stored at −80 • C. Mock samples were prepared in the same manner, using uninfected cells. Virus stocks were quantified by titration, as previously described [22].

Replication Inhibition Assay
MDCK, A549, or LET1 cells were seeded in 96-well plates (TPP, Trasadingen, Switzerland) at the density of 10 4 cells per well. After 24 h, cells were infected with influenza A virus at a dose of 400 50% tissue culture infectious dose (TCID 50 )/mL in the presence of BBR in a total volume of 100 µL in DMEM supplemented with penicillin/streptomycin and TPCK-treated trypsin 1 µg/mL (Sigma-Aldrich, Poznan, Poland). In the case of the HAE cultures, virus and BBR were applied to the apical surfaces of the HAE cultures. Control samples were inoculated in the same manner with the same volume of mock and/or DMSO. After 2 h of incubation at 37 • C, cells were rinsed thrice with PBS and fresh medium supplemented with BBR was added. The infection was carried out for 48 h (cell lines) or 72 h (HAE). Virus yield was measured using the RT-qPCR method described below.

Mechanism of Action Assays
To determine whether BBR affects a particular stage of the replication cycle of the virus, a series of previously described mechanistic experiments [22,23] was carried out. All assays were conducted on sub-confluent LET1 cells.

Cell Viability
Cell viability was evaluated using the XTT Cell Viability Assay kit (Biological Industries, Cromwell, CT, USA) according to the manufacturer's instructions. Cells were incubated with BBR for 48 h at 37 • C in an atmosphere containing 5% CO 2 . After incubation, the medium was discarded and 100 µL of fresh medium was added to each well. Then, 50 µL of the activated 2,3-bis-(2-methoxy-4-nitro-5-sulphenyl)-(2H)-tetrazolium-5-carboxanilide (XTT) solution was added and samples were incubated for 2 h at 37 • C. The absorbance (λ = 450 nm) was measured using a Spectra MAX 250 spectrophotometer (Molecular Devices, San Jose, CA, USA). Data were presented as the ratio of signal from the examined sample and from the control sample (solvent-treated cells) × 100%.

Confocal Microscopy and Image Analysis
For confocal microscopy analysis, 0.1 × 10 6 A549 cells were seeded on coverslips in 12-well plates (TPP, Trasadingen, Switzerland). After 12 h, cells were washed with PBS and infected with influenza A virus in the presence of BBR.
Distribution of values was tested using the Shapiro-Wilk normality test, and equality of group variances was examined with Browne-Forsythe test. Statistical significance was calculated using Kruskal-Wallis test with Dunn's multiple comparisons test. p values <0.05 were considered significant and denoted with an asterisk.

Quantitative Real-Time PCR
Viral RNA was isolated using the Viral DNA/RNA Isolation Kit (A&A Biotechnology, Gdynia, Poland) according to the manufacturer's instructions. Then, reverse transcription was carried out with a high-capacity cDNA reverse transcription kit (Applied Biosystems, Waltham, MA, USA) according to the manufacturer's instructions using a Veriti thermal cycler (Applied Biosystems, Waltham, MA, USA). The whole cDNA was determined by quantitative real-time PCR (qPCR). The reaction was carried out in a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) (2 min at 50 • C, 10 min at 95 • C, followed by 40 cycles of 15 s at 95 • C and 1 min at 60 • C). The reaction mixture contained 1 × RT HS-PCR mix probe (A&A Biotechnology, Gdynia, Poland), a specific probe labeled with 6-carboxyfluorescein (FAM), a black hole quencher 1 (BHQ1) (sequence 5 -TCA GGC CCC CTC AAA GCC GA-BHQ1-3 , 100 nM), primers specifically targeting the influenza A matrix protein (450 nM each, sense primer 5 -AGA TGA GTC TTC TAA CCG AGG TCG-3 , antisense primer 5 -TGC AAA AAC ATC TTC AAG TCT CTG-3 ), and 2.5 µL of viral cDNA in 10 µL reaction mixture. For virus copy number quantification a standard DNA template of the known copy number was prepared and serially diluted, and the standard curve was prepared as described previously [22]. The lower limit of detection was < 10 3 copies/mL.

Statistics
The results are expressed as means ± standard deviations (SD). The half-maximal inhibitory concentration (IC 50 ) and 50% toxic concentration (TC 50 ) values were calculated using the Graph Pad Prism 8.0 function (GraphPad Software, San Diego, CA, USA). The selectivity index (SI) was defined as the TC 50 /IC 50 ratio. To determine the significance of the results obtained, Student's t-test was used and p values < 0.05 were considered significant. Western blot analyses were conducted using Image Lab 6.0 (Bio-Rad, Hercules, CA, USA).

The Antiviral Effect of BBR Is Cell-Type Dependent
To investigate the antiviral activity of BBR in vitro, a series of cell lines permissive to the influenza A virus were infected in the presence of BBR (10-160 µM). First, the most common in vitro model was used, i.e., the cell line derived from Madin-Darby canine kidney cells (MDCK). We performed a standard virus replication assay by infecting the MDCK cells in the presence or absence of BBR. Concomitantly, cell viability was monitored using the XTT assay. A significant inhibitory effect of BBR in the MDCK cells was observed in our experimental setting in samples treated with BBR at 80 and 160 µM, while IC 50 was equal to 52 µM. We did not observe any significant decrease in cell viability (TC 50 = 1035 µM, SI = 20) ( Figure 1A, Table 1). These results indicate inhibitory properties similar to those obtained by Wu et al. (IC 50 = 0.025 µg/mL ≈ 67 µM). We conclude that BBR hampers replication of the influenza A virus in MDCK cells, but to a much lesser extent than in other in vitro models [19,28]. Next, the human pulmonary tumor cell line A549 was assayed, as inhibition of the influenza virus by BBR in this cell line has been consistently reported [19,21]. Many reports show a high toxicity of BBR to cancer cells, including the A549 cell line [29,30], therefore BBR has been proposed as an antineoplastic drug candidate. Considering that BBR cytotoxicity was not assayed in these cells, we decided to verify the antiviral BBR activity in A549 cells. Our results indicate a significant decrease in influenza A replication in the presence of BBR, but the compound was cytotoxic at the active concentrations (IC 50 = 17 µM, TC 50 = 107 µM, SI = 6) ( Figure 1B, Table 1).
Finally, we decided to test the immortalized murine lung epithelial type I cell line (LET1). This cell line is non-cancerous and maintains most of the natural signaling pathways [31]. First, we verified the cytotoxicity of BBR on these cells and it was acceptable (TC 50 = 521 µM). Next, we tested the effect of BBR on the replication of the influenza A virus in LET1 cells. RT-qPCR analysis revealed a vast decrease in viral replication at non-toxic concentrations (IC 50 = 4 µM, SI = 123) ( Figure 1C, Table 1).   To summarize, we have tested three different in vitro models, and we have obtained three different results. To sort out the actual activity of BBR, we decided to evaluate BBR in a more natural system, mimicking the human airway epithelium (HAE). Briefly, HAE was reconstituted on the interface of air and media using primary human cells. This model has previously been well-characterized and successfully used by us in other studies [22,32,33]. To the best of our knowledge, it is the most natural cellular model for testing of potential inhibitory compounds against a variety of respiratory viruses. 3D HAE cultures form well-differentiated, polarized cultures that resemble in vivo pseudostratified mucociliary epithelium lining the conductive airways of the human respiratory tract. We did not observe cytotoxicity at tested concentrations (i.e., up to 160 µM, 0.054 µg/mL), suggesting that BBR is not toxic towards non-cancerous cells, consistent with previously published data. However, BBR inhibited influenza A virus replication in HAE cultures only at higher concentrations compared to LET1 cells (IC 50 = 16 µM). This may be explained by the fact that BBR was not continuously present on the cultures, but only during daily washes of the apical surface ( Figure 2). cell line is non-cancerous and maintains most of the natural signaling pathways [31]. First, we verified the cytotoxicity of BBR on these cells and it was acceptable (TC50 = 521 μM). Next, we tested the effect of BBR on the replication of the influenza A virus in LET1 cells. RT-qPCR analysis revealed a vast decrease in viral replication at non-toxic concentrations (IC50 = 4 μM, SI = 123) ( Figure 1C, Table 1).
To summarize, we have tested three different in vitro models, and we have obtained three different results. To sort out the actual activity of BBR, we decided to evaluate BBR in a more natural system, mimicking the human airway epithelium (HAE). Briefly, HAE was reconstituted on the interface of air and media using primary human cells. This model has previously been wellcharacterized and successfully used by us in other studies [22,32,33]. To the best of our knowledge, it is the most natural cellular model for testing of potential inhibitory compounds against a variety of respiratory viruses. 3D HAE cultures form well-differentiated, polarized cultures that resemble in vivo pseudostratified mucociliary epithelium lining the conductive airways of the human respiratory tract. We did not observe cytotoxicity at tested concentrations (i.e., up to 160 μM, 0.054 μg/mL), suggesting that BBR is not toxic towards non-cancerous cells, consistent with previously published data. However, BBR inhibited influenza A virus replication in HAE cultures only at higher concentrations compared to LET1 cells (IC50 = 16 μM). This may be explained by the fact that BBR was not continuously present on the cultures, but only during daily washes of the apical surface ( Figure 2).
Taken together, BBR is expected to inhibit the influenza A virus during the natural infection, but care should be taken during the selection of the in vitro model.  Taken together, BBR is expected to inhibit the influenza A virus during the natural infection, but care should be taken during the selection of the in vitro model.

BBR Inhibits the Influenza A Virus Replication at Late Stages of the Infection
As we confirmed the anti-influenza activity of BBR, our next step was to determine whether BBR affects a particular stage of the virus replication cycle. To investigate this, we made a series of mechanistic experiments as described in our earlier works [22,23]. Our assays I-IV did not show a significant decrease in the influenza A virus RNA copy number measured with RT-qPCR, suggesting that BBR did not interfere with the virus itself, virus entry, trafficking, and genome replication ( Figure 3). However, we observed a decrease in assay V, which suggests that BBR affects virus assembly, maturation, or egress ( Figure 4).
Viruses 2020, 12, 344 9 of 16 affects a particular stage of the virus replication cycle. To investigate this, we made a series of mechanistic experiments as described in our earlier works [22,23]. Our assays I-IV did not show a significant decrease in the influenza A virus RNA copy number measured with RT-qPCR, suggesting that BBR did not interfere with the virus itself, virus entry, trafficking, and genome replication ( Figure  3). However, we observed a decrease in assay V, which suggests that BBR affects virus assembly, maturation, or egress ( Figure 4).

BBR Inhibits the Influenza A Virus Replication through Downregulation of the MAPK/ERK Pathway
Data presented above and recently published works on other viruses suggest that BBR acts on cellular pathways that are essential for efficient virus replication.
To confirm whether BBR inhibits the influenza A virus through downregulation of the MAPK/ERK pathway, confocal imaging of infected A549 cells treated with two different

BBR Inhibits the Influenza A Virus Replication through Downregulation of the MAPK/ERK Pathway
Data presented above and recently published works on other viruses suggest that BBR acts on cellular pathways that are essential for efficient virus replication.
To confirm whether BBR inhibits the influenza A virus through downregulation of the MAPK/ERK pathway, confocal imaging of infected A549 cells treated with two different concentrations of BBR and with an active concentration of specific MAPK/ERK inhibitor U0126 (10 µM) was conducted. After 12 h, we observed characteristic accumulation of a viral RNP complex inside nuclei, in both cells treated with BBR and U0126 ( Figure 5A). This effect was confirmed by image analysis. Significant differences in NP signal accumulation were observed for both the U0126 and BBR cells compared to the untreated control ( Figure 5B). More than 90% of NP colocalized with nuclear staining in cells treated with U0126 or BBR compared to about 70% in control cells. This effect was clearly visible between 12 and 20 h p.i. After this time, large amounts of RNP were also visible in the cytoplasm (data not shown). This outcome may be due to the removal of BBR from the cell or its degradation. Western blot analyses indicate moderate inhibition of ERK1 activation by BBR at 20 µM concentration (45% inhibition) compared to the U0126 cell (76% inhibition) (Figure 6.). We observed, however, a slight increase in the amount of phospho-ERK in control-infected cells, as previously reported [34]. To confirm confocal imaging results, we performed a similar experiment and isolated cytoplasmic and nuclear fractions for Western blot analysis. The analysis showed higher quantities (~3 fold increase) of viral NP in nuclear fractions of cells treated with U0126 or 20 µM BBR (0.007 µg/mL) compared to untreated infection, confirming our confocal imaging observations ( Figure 7B). Lower BBR concentration, however, did not cause significant accumulation of influenza A virus NP (data not shown). In the case of cytoplasmatic fraction, we observed a moderate yet clear decrease of NP (about 40% inhibition) signal in treated cells which may suggest that BBR does not interfere with the assembly and release of virions. Taking these results together, obtained data suggest an inhibition on the influenza A virus by BBR through the downregulation of the MAPK/ERK1 pathway. Viruses 2020, 12, x FOR PEER REVIEW 11 of 16

Discussion
Influenza A virus infections are among major public health threats. Despite many efforts, the list of approved drugs is short, and the emergence of resistance is rapid [8]. Consequently, there is a continual need for new, broad-spectrum, anti-influenza drugs, and many different sources for such drugs are considered. Natural products constitute a rich library of compounds, which may be used to identify new structures and new molecular targets. However, knowledge of the mechanism of action is essential. Berberine (BBR) exhibits a number of biological activities and, among others, has been described as a potent inhibitor of the influenza virus [12]. In this work, we undertook a comprehensive study of the antiviral properties of BBR in vitro and ex vivo. We also delineated the mechanism of action of the compound.
Some reports on the anti-influenza properties of BBR are conflicting. For MDCK cells, Cecil et al. showed that BBR did not affect virus replication, while Wu et al. reported substantial inhibition in this culture model [19,20]. In our study, we observed noticeable influenza A virus inhibitory properties at high concentrations, similar to those obtained by Wu et al. [20]. The main difference between the studies was the concentration range used, as the effective concentration in the MDCK cells is unusually high.
The results obtained with a second commonly used model based on A549 cells are also slightly confusing. While the activity of BBR has been reported in these cells [21], other reports show a strong cytotoxic activity of BBR towards cancerous cells, including A549 cells [10,30]. Indeed, while the effective concentration was rather low, the cytotoxic concentration was also low, and it was impossible to determine whether the observed effect is not an artifact.
To sort out whether the antiviral effect of BBR has the potential to be used for treatment, we tested two further models employing non-cancerous cells originating from respiratory cells. First, mouse immortalized lung epithelial type 1 (LET1) line was used. Second, fully differentiated 3D primary human airway epithelium (HAE) cultures were employed. In both cases, we observed significant anti-influenza properties of BBR at non-toxic concentrations. These results confirm the anti-influenza properties of BBR, the low toxicity of the compound for the primary cells, and show the potency of the compound to inhibit the infection in the tissue.
In the subsequent step, we aimed to understand the molecular mechanism of action of the BBR on the influenza A virus. Functional assays demonstrated the effect of BBR on virus replication, and not the receptor binding or virus entry. The cell-type-dependent differences in the compound antiviral activity suggested that the compound likely acts on the host cell, and not the virus itself. For this reason, we made an effort to link the known activities of BBR with pathways that are important for the replication of the influenza A virus. Two recently published reports show BBR-mediated inhibition of the Chikungunya virus (CHIKV) and the enterovirus 71, which was caused by interference with the virus-induced mitogen-activated protein kinase (MAPK) signaling pathways, including extracellular signal-related kinase (ERK) [35,36]. It is also known that the active form of ERK1 (p-ERK1) is required for the nuclear export of the influenza A virus RNP complex during the infection [37] and that this pathway is modulated during the infection. First, the ERK pathway is upregulated shortly after the infection, to rapidly decrease after the first hour [38]. Next, the pathway is upregulated again at the late stage of the infection, when RNP must be exported from the nucleus to assemble into progeny virions [37]. ERK inhibition with a specific inhibitor U0126 impairs RNP transport and results in retention of RNPs inside the nucleus, and consequent inhibition of viral replication [37]. We verified if BBR exhibited activity similar to that of U0126. Confocal microscopy revealed a similar RNP retention within the cell nuclei in both BBR-treated and MAPK/ERK inhibitor U0126-treated A549 cells ( Figure 5). Accumulation of the nucleoprotein was also observed when nuclear fractions of cells treated with BBR and U0126 were analyzed ( Figure 6). The literature does not link the specific MAPK/ERK inhibition and the activity of BBR in A549 cells [39,40], which suggests that the antiviral effect is specific.
In our studies, we used BBR diluted in DMSO. While this solvent is convenient for use in cell culture, regulatory agencies recommend it only for topical use [41,42]. Considering the route of administration of anti-influenza drugs (oral, inhalation, or intravenous) [43], other solvents should be considered, which may include conjugates or nanocages. Such new formulations may also increase the bioavailability of BBR [44].
Summarizing, our results demonstrate inhibition of the influenza A virus by BBR at non-toxic concentrations. The antiviral effect differs between cell lines, but we confirmed its relevance using the most advanced model of the human respiratory epithelium. A subsequent study revealed that the inhibition of the influenza A virus replication results from BBR-specific inhibition of the MAPK/ERK1 pathway, required for the effective production of progeny virions. It is worth noting that due to its mechanism of action, the emergence of resistant mutants is less likely than for commonly used inhibitors targeting viral proteins.