Methylene Blue has a potent antiviral activity against SARS-CoV-2 and H1N1 influenza virus in the absence of UV-activation in vitro

Methylene blue is an FDA (Food and Drug Administration) and EMA (European Medicines Agency) approved drug with an excellent safety profile. It displays broad-spectrum virucidal activity in the presence of UV light and has been shown to be effective in inactivating various viruses in blood products prior to transfusions. In addition, its use has been validated for methemoglobinemia and malaria treatment. In this study, we first evaluated the virucidal activity of methylene blue against influenza virus H1N1 upon different incubation times and in the presence or absence of light activation, and then against SARS-CoV-2. We further assessed the therapeutic activity of methylene blue by administering it to cells previously infected with SARS-CoV-2. Finally, we examined the effect of co-administration of the drug together with immune serum. Our findings reveal that methylene blue displays virucidal preventive or therapeutic activity against influenza virus H1N1 and SARS-CoV-2 at low micromolar concentrations and in the absence of UV-activation. We also confirm that MB antiviral activity is based on several mechanisms of action as the extent of genomic RNA degradation is higher in presence of light and after long exposure. Our work supports the interest of testing methylene blue in clinical studies to confirm a preventive and/or therapeutic efficacy against both influenza virus H1N1 and SARS-CoV-2 infections.


Introduction
Viral pandemics cause significant morbidity and mortality.Influenza A viruses and coronaviruses are among the major human threats due to the presence of an important animal reservoir and the virus ability to cross the species barrier via mutation and recombination.The ongoing SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2) pandemic unveiled the inability to develop specific antivirals and vaccines in a short time frame.In addition, co-infection with influenza contributes to severity of SARS-CoV-2 pneumonia (1) and antiviral drugs for Influenza are far from ideal, due to resistance and administration early after symptoms to be effective.This scenario highlights the urgent clinical need for broad-spectrum antiviral drugs ready to use during pandemics and epidemics.Methylene Blue (MB) is an FDA and EMA approved drug with an excellent safety profile.Due to its antimicrobial, anti-inflammatory and antitoxic effects, photoactivated MB is used for a wide range of applications including treatment of methemoglobinemia or malaria (2).Photoactivated MB is also widely used to obtain blood product preparations free of viruses such as HIV (3), Ebola or Middle East Respiratory Syndrome coronavirus (MERS) (4), or SARS-CoV-2 virus (5).The antiviral activity appears to rely on multiple mechanisms, and is more potent for enveloped viruses (6).Among other actions, MB is known to corrupt DNA or RNA integrity.This is due to a redox reaction in which the molecule accepts electrons on its aromatic thiazine ring, thus being reduced to leuko-methylene blue (MBH2) which in turn transfers electrons to other molecules such as nucleic acids.Moreover, MB in combination with oxygen and a source of energy results in the production of singlet oxygen, a highly reactive reaction partner which induces guanine oxidation (8-oxo-7,8-dihydroguanine (8-oxoGua) lesions) damaging DNA or RNA.Other mechanisms include but are not limited to a) modified carbonyl moieties on proteins, b) single-strand breaks (ssb) in the RNA genome and c) RNA-protein crosslinks, all lesions correlating well with a broad-spectrum virucidal activity (7).In addition to this nucleic acid damaging activity, MB was shown to inactivate viruses such as HIV by targeting also the viral envelope and core proteins (8).MB presents few side effects at doses below 5mg/kg.Long term, high dose administrations such as for malaria treatment, may result in temporary fully reversible blueish coloration of urine, sclera and skin.Sensitivity to thiazine dyes and glucose 6 phosphate dehydrogenase (G-6-PD) deficiency are other contraindications (9).Using a classical virus neutralization assay for both H1N1 influenza virus and SARS-CoV-2, we present in vitro data demonstrating a potent antiviral activity of MB with and without UV-activation.We report a strong antiviral activity without light (experiments performed in a closed stainless-steel box) at 2, 4 and 20 hours of incubation with the virus.Interestingly, degradation of genomic RNA was only observed in presence of light and upon long exposure, confirming that the antiviral effect relies on multiple mechanisms of action and can be efficient also in absence of RNA degradation.Additional experiments carried on with SARS-CoV-2 further highlight that MB is effective also when added on cells already infected.Finally, we demonstrate an

Results
In order to define a non-toxic dose range on MB we calculated the 50% cytotoxic concentration (CC50) of MB in our in vitro infection systems, MDCK for H1N1 and Vero-E6 cells for SARS-CoV-2.We assessed MB toxicity in infection medium and upon different times of administration, according to the experimental conditions chosen to test the antiviral activity of the compound (Table1).
Table1 We first asked whether MB was able to exert antiviral activity independently of visible light.We thus incubated different doses of the compound spanning from 2 g/ml to 0.08 g/ml [1], with 1*10 7 plaque forming units (pfu) of human A/Netherlands/602/2009 (H1N1), for 20h or 2h at room temperature and in the presence or absence of visible light.We then measured the number of infectious particles in each condition.MB exerted virucidal activity in both the presence and absence of light and at both incubation time points, suggesting that the compound can inactivate irreversibly a relevant respiratory enveloped virus independently of visible light (Figure 1 a and b).As more than one mechanism of action can underlie the antiviral properties of MB [5], we asked whether the compound would damage H1N1 genome.We therefore measured genome integrity by RT-qPCR in each of the above-described conditions (Figure 1c and d).Surprisingly, the reduction of H1N1 infectiousness induced by MB did not always result in a reduction of viral genome integrity.Only upon 20h incubation under the light and at the highest doses, MB induced a significant decrease in the abundance of H1N1 genome, compared to the untreated control (Figure 1d).This finding suggests that RNA damage [5;6] is only partially responsible for MB antiviral activity under the light.Overall these preliminary results prompted us to evaluate the virucidal activity of MB against SARS-CoV-2.Hence, we incubated different doses of the drug, ranging between 50 g/ml and 0.08 g/ml, with 5*10 6 pfu of SARS-CoV-2 for 20 h in the dark and at room temperature.The results shown in Figure 2 evidence a complete inhibition of SARS-CoV-2 replication from 50 to 10 µg/ml and a 3.05 log reduction at 0.08 µg/ml, the lowest dose tested (Figure 2a).Of note Jin and colleagues (5) observed no virucidal activity of MB against SARS-CoV-2 when administered for 20' in the dark.We thus performed additional experiments incubating MB and the virus for 2h or 4h and evidenced a statistically significant effect at the 2h incubation time (Figure 2b).This effect was however lower compared to that observed at 20h incubation (5.3 log reduction after 2h treatment with 20 µg/ml and 3.6 log with 2 µg/ml).Subsequently we evaluated the possible combinatorial activity of MB and immunoglobulins present in the serum of a convalescent individual.Preliminary tests were conducted to identify the optimal serum dilution able to neutralize the virus and evidenced a 4.15 log decrease in presence of 1:16 dilution, a 2.3 log decrease in presence of 1:80 dilution, and absence of neutralization at lower serum dilutions (data not shown).In order to be able to evaluate any additive effect, the serum was diluted 1:80 for further experiments.5*10 6 pfu of SARS-CoV-2 were incubated for 20 hours in the dark with three different doses of MB in presence or absence of diluted convalescent serum.The results (Figure 2c) evidence a partial additive effect of MB and convalescent sera, with a complete inhibition of viral infectivity when combining MB (up to 0.4 µg/ml) and serum.To further verify the mechanism of action of MB in the dark against SARS-CoV-2, RT-qPCR was performed on the samples incubated for 2h and 4h.Despite the logarithmic decrease in viral infectivity (Figure 2b), no decrease of viral RNA was evidenced (Figure 2d).Finally, we assessed the efficacy of MB for therapeutic use.We thus tested whether MB was able to reduce the viral yield in cells previously infected with SARS-CoV-2.The treatment was administered at 4 or 24 hpi and the virus released by the infected cells was measured titrating the culture supernatants at 48 hpi.We observed, in both conditions, viral inhibition in a dose-response manner (Figure 3) with an EC50 of 0.11 g/ml at 4 hpi and of 0.13 g/ml at 24 hpi.

DISCUSSION
Here we show broad-spectrum virucidal activity of MB in absence of light activation, against both H1N1 influenza virus and SARS-CoV-2 after incubation times of 2, 4 or 20h.In a previous study (5), no antiviral activity was observed after shorter incubations in absence of light.The activity of the compound in absence of UV activation is thus suggestive of a different mechanism of action to that reported in presence of UV light or to a slower kinetics.This hypothesis is further sustained by the lack of RNA decrease for both H1N1 and SARS-CoV-2 in presence of MB in the dark (Figure 1d   and 2d).We also show here that the antiviral effect of MB is maintained in presence of convalescent serum suggesting that the compound could be effective when administered in vivo in patients infected with SARS-CoV-2 and that its efficacy could be further increased by immune serum.A limitation of our study with SARS-CoV-2 is that the experiments are performed in Vero-E6 cell line where other drugs, such as hydroxychloroquine (10), were reported to exert antiviral activity but then they did not show efficacy in clinical trials (11).However, in those cases the effect was intracellular while MB acts also extracellularly and directly on the virus, therefore, we do not expect to observe such a discrepancy when going into clinical trials.MB produces a vasoconstriction in distributive (hypovolemic) shock by inhibition of nitric oxide synthase and guanylate cyclase.This is a concomitant and unexpected clinically beneficial effect, because end stage viral infections present often the clinical status of a distributive shock(12), (13).A recent French publication of a cohort of 2500 end stage cancer patients treated with MB during the first wave of CoviD-19 mentions a possible protective role of MB against respiratory viruses as in this cohort there was no reported cases of Influenza or SARS-CoV-2 infections (14).Another report describes inhibition by MB in vitro of the SARS-CoV-2 Spike -ACE2 protein-protein interaction.(15).Finally, a recent study compared the in vitro median Effective Concentration (EC50) of 3 drugs against SARS-COV-2, when administered 4h post infection.MB was the most efficient antiviral drug with an EC50 of 0.30 ± 0.03 µM, followed by Hydroxychloroquine (1.5 ± 0.3 µM) and Azithromycin (20.1 ± 4.5 µM) (16).In view of the wide variety of singlet oxygen driven chemical reactions corroding nucleic acids and proteins, one is surprised by the well documented absence of severe side effects of MB, even at doses up to 5 mg/kg.This may be linked to the fact, that the body's immune system uses MB for its own purposes.It is known that antibodies have, close to their binding site, a catalytic site capable of producing singlet oxygen in the presence of water (17).Riboflavin, also known as Vitamin B2 is another efficient singlet oxygen producing compound (18).It seems reasonable to assume that mammals have developed so far unknown control mechanisms to keep singlet oxygen collateral damages at bay.Altogether, our results support the interest of testing MB in clinical studies to confirm a preventive or therapeutic efficacy, alone or combined with immune sera, against two major public health concerns: Influenza virus and SARS-CoV-2 whose co-infection is a significant risk factor for prolonged hospitalization (1).Further experiments will be required to fully understand the different antiviral mechanisms of action of MB against these two enveloped viruses.Based on previously published data, MB might inactivate viruses limiting viral replication and spread beyond the upper respiratory tract.In addition, its anti-inflammatory activity (2) may lower the side effects linked to the host response.

Compound and convalescent serum
Methylene blue (Methylthonium chloride solution) was purchased from ProVepharm.A convalescent serum tested positive for IgG against SARS-CoV-2 by ELISA was separated through centrifugation, aliquoted and stored at -80°C.Written informed consent was obtained from the subject involved and the research protocol was approved by the local ethics committee (commission cantonale d'éthique de la recherche CCER de Genève).

Cells and Virus
MDCK and (ATCC CCL34) cells and Vero C1008 (clone E6) (ATCC CRL-1586) cells (kindly provided by Prof Gary Kobinger from the University of Laval), were propagated in DMEM High Glucose + Glutamax supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptavidin (pen/strep).SARS-CoV-2/Switzerland/GE9586/2020 was isolated from a clinical specimen in the University Hospital of Geneva in Vero-E6.Cells were infected and supernatant was collected 3 days post infection, clarified, aliquoted and frozen at -80°C before titration by plaque assay in Vero-E6.Human H1N1 virus A/Netherlands/602/2009 was a gift from Prof Mirco Schmolke (University of Geneva) and was produced and titrated in MDCK cells by plaque assay.Toxicity assay Vero-E6 and MDCK cells (13,000 and 20,000 cells per well respectively,) were seeded in 96-well plate one day before the assay.For toxicity evaluation in Vero-E6 methylene blue was serially diluted 1:2 in DMEM supplemented with 5% FBS and added on cells for 1h, followed by a washout and the addition of DMEM supplemented with 5% FBS for additional 48h hours.Alternatively, the dose range of MB was added on the cells for 48h.The toxicity evaluation in MDCK cells was performed in serum free DMEM supplemented with 0.2 µg/ml of TPCK-Trypsin (Sigma), using the same MB doserange and times of treatment applied to Vero-E6.MTS reagent (Promega) was added on cells for 3h at 37°C according to manufacturer instructions, subsequently absorbance read at 570 nm.Percentages of viability were calculated by comparing the absorbance in treated wells and untreated.50% cytotoxic concentration (CC50) were calculated with Prism 8 (GraphPad, USA).Neutralization assay H1N1 (1.5 *10 7 pfu) or SARS-CoV-2 (5*10 6 pfu) were incubated at room temperature in the dark (in an opaque stainless-steel box) or under visible light, in a final volume of 100 µl, with different dilutions of MB for 2h or 20h.The infection mixtures were then serially diluted (1:10 factor) and added for 1h at 37°C on the respective host cell lines: MDCK cells (1,000,000 cells per well, seeded in 6-well plate 24h in advance), H1N1 neutralization assay; Vero-E6 cells (100,000 cells per well, seeded in 24-well plate 24h in advance) for SARS-CoV-2 neutralization assay.MDCK monolayers were then washed and overlaid with 0.8% agarose in medium supplemented with TPCK trypsin 1µg/ml.Two days after infection, cells were fixed with paraformaldehyde 4% and stained with crystal violet solution containing ethanol.Whereas, after infection Vero-E6 cells were washed and overlaid with 0.8% avicel rc581 in medium supplemented with 5% FBS.Two days after infection, cells were fixed with paraformaldehyde 4% and stained with crystal violet solution containing ethanol.Plaques were counted for both viruses, and the titer of the different conditions was calculated.Statistical analysis was done with Prism software (Prism 8, GraphPad).RT-qPCR analysis H1N1 1.5*10 7 pfu was incubated at room temperature in the dark (in an opaque stainless steel box) or under visible light, in a final volume of 100 µl, with different dilutions of MB for 2h or 20h.SARS-CoV-2 5*10 6 pfu was incubated in the dark with 2 or 20 g/ml of MB for 2 or 4h.Viral RNA was extracted with EZNA viral extraction kit (Omega Biotek) and quantified by using RT-qPCR with the QuantiTect kit (#204443;Qiagen, Hilden, Germany) in a StepOne ABI Thermocycler.

Viral yield reduction assay
Vero-E6 cells (100,000 cells per well) were seeded in 24-well plate.Cells were infected with SARS-CoV-2 (MOI, 0.005 pfu/cell) for 4 h at 37°C.The monolayers were then washed and overlaid with medium supplemented with 5% FBS containing serial dilutions of compound, alternatively the compound was added 24 hpi.Supernatants were harvested 48 hpi and titrated on Vero-E6 cells.

Statistical Analysis
Where possible, half-maximal antiviral effective concentration (EC50) values were calculated by regression analysis using the dose-response curves generated from the experimental data using GraphPad PRISM 8 (GraphPad Software, San Diego, CA, USA).The 50% cytotoxic concentration (CC50) was determined using logarithmic viability curves.One-way ANOVA, followed by Dunnet's multiple comparison test, was used to assess the statistical significance of the differences between treated and untreated samples.Significance was set at the 95% level.

Figure 1 .
Figure 1.Assessment of MB antiviral activity on human H1N1.MB was incubated with H1N1 virus A/Netherlands/602/2009 (1.5*10 7 pfu) for 2h or 20h in the darkness a) and c) or in the presence of light b) and d).At the end of the incubation, mixtures were serially diluted and added for 1h at 37°C on MDCK cells a) and b).Mixtures were then removed and the cells were overlaid with medium containing 0.8% agarose and TPCK trypsin 1µg/ml.At 48 hours post infection (hpi) the cells were fixed in order to count the plaques and determine the viral titer in each experimental condition.In parallel, viral RNA was quantified by RT-qPCR c) and d).Results are mean and SD of two independent experiments.2 µg/ml (6.2 µM), 0.4 µg/ml (1.25 µM), 0.08 µg/ml (0.25 µM).UT = untreated.***p<0.001****p<0.0001.

Figure 3 .
Figure 3. Viral yield reduction assay.Vero-E6 were infected with SARS-CoV-2 (MOI 0.005).4h or 24 hpi, medium containing or not serial dilutions of MB was added on cells.24h later, cell supernatants were collected and titrated on Vero-E6 cells.Results are mean and SD of two independent experiments performed in duplicate.*p<0.05,**p<0.01.