Gramicidin S and melittin: potential anti-viral therapeutic peptides to treat SARS-CoV-2 infection

The COVID19 pandemic has led to multipronged approaches for treatment of the disease. Since de novo discovery of drugs is time consuming, repurposing of molecules is now considered as one of the alternative strategies to treat COVID19. Antibacterial peptides are being recognized as attractive candidates for repurposing to treat viral infections. In this study, we describe the anti-SARS-CoV-2 activity of the well-studied antibacterial peptides gramicidin S and melittin obtained from Bacillus brevis and bee venom respectively. The EC50 values for gramicidin S and melittin were 1.571 µg and 0.656 µg respectively based on in vitro antiviral assay. Significant decrease in the viral load as compared to the untreated group with no/very less cytotoxicity was observed. Both the peptides treated to the SARS-CoV-2 infected Vero cells showed viral clearance from 12 h onwards with a maximal viral clearance after 24 h post infection. Proteomics analysis indicated that more than 250 proteins were differentially regulated in the gramicidin S and melittin treated SARS-CoV-2 infected Vero cells against control SARS-CoV-2 infected Vero cells after 24 and 48 h post infection. The identified proteins were found to be associated in the metabolic and mRNA processing of the Vero cells post-treatment and infection. Both these peptides could be attractive candidates for repurposing to treat SARS-CoV-2 infection.

The pandemic caused by SARS-CoV-2 has led to intense research not only on the biology of the virus but also therapeutic interventions with a multi-pronged approach 1 . Vaccines have been developed at "warp" speed. The overall efficacies of different vaccines, though variable are excellent 2 and have played a major role in controlling the disease 3 . However, vaccines are not available universally and there have been cases of infection with SARS-CoV-2 even in vaccinated individuals, though not severe 4 . Also, the effect of vaccines would wane over time. The vaccines may also be less effective against newly emerging strains such as omicron which has a very large number of substitutions in the viral genome as compared to the earlier strains (https:// www. cdc. gov/ coron avirus/ 2019-ncov/ scien ce/ scien ce-briefs/ scien tific-brief-omicr on-varia nt. html). There is clearly a need for development of therapeutic agents in addition to vaccines. In the area of anti-infective agents against SARS-CoV-2, efforts have been taken to generate therapeutic antibodies that would neutralize the virus and prevent its interaction with cellular receptors to gain entry into cells 5 . Considering the time scales in developing a drug de novo, there have been several attempts to re-purpose drugs to treat COVID19 1,6,7 . However, repurposed drugs have had very limited success in treating SARS-CoV-2 infection including remdesivir 8 . There is no drug to-date that can be used specifically to treat COVID19 disease. Although two drugs from Pfizer (Paxlovid) and Merck (Molnupiravir) appear to show promise, their effectiveness is still to be established unequivocally 9 .
Infection in the case of SARS-CoV-2 is initiated by binding of the spike protein to ACE2 followed by a series of steps leading to fusion and internalization of the virus and propagation 1, 10 . If the binding of the spike protein to ACE2 is prevented, then the virus will no longer be able to enter cells and propagate. SARS-CoV-2 is an enveloped virus where the RNA is encapsulated within a lipid vesicular structure with the spike protein decorating on the external side giving the "corona" appearance 1, 10 . Disruption of the lipid structure would lead to the disintegration of the virus. Naturally occurring membrane-active peptides have potent antimicrobial activity which stems from their ability to disrupt bacterial membranes 11,12 . We have explored the antiviral activity of two extensively studied peptides, gramicidin S having the sequence: [cyclo-(Val-Orn-Leu-D-Phe-Pro)2] 13 and the bee venom peptide, melittin having the sequence: GIGAVLKVLTTGLPALISWIKRKRQQ-amide 14 . We reasoned that if the peptides could destabilize the viral membrane, the virus would disintegrate and would thus be rendered inactive. The peptides could also conceivably bind to the spike protein and prevents its interaction with ACE2 or inhibit fusion.

Results
Antiviral activity of peptides. The cytolytic activity of gramicidin S and melittin was determined using MTT assay (Fig. 1). Results showed 75-80% cell survival at all the concentrations tested (up to 5 µg) indicating the safe use of these peptides (Fig. 1). The SARS-CoV-2 viral particles enumerated by the RT-qPCR showed that treatment of gramicidin S and melittin effectively reduced viral load in vitro Log EC 50 value of gramicidin S (0.1963) corresponds to 1.571 µg and Log EC 50 value (2.826) of melittin corresponds to 0.656 µg (Fig. 2) The antiviral activity of gramicidin S (3.0 µg) and melittin (1.5 µg) at 12 and 24 h was examined along with remdesivir (1 µM) as assay control. The data shown in Fig. 3 indicates that the peptides show antiviral activity at 12 h and is more pronounced at 24 h. The gramicidin S and melittin showed 99% and 95% viral reduction respectively at 12 h compared with remdesivir (20%). At 24 h remdesivir showed 90% viral reduction whereas both gramicidin S and melittin showed 99% viral reduction. The SARS-CoV-2 antiviral activity of gramicidin S and melittin was compared with remdesivir by confocal microscopy (Fig. 4). Panel B (green fluorescence) indicates infection of cells, more prominent at 24 h. Panels D and E correspond to virus incubated with gramicidin S and melittin before incubating with cells. The considerable decrease in green fluorescence indicates that both the peptides have good anti-viral activity as with remdesivir shown in Panel E. Although, antiviral activity is observed at 12 h, it is more pronounced at 24 h.  high-throughput proteomic analysis. iTRAQ based quantitative proteomics analysis identified 7 SARS-CoV-2 proteins as up regulated in the control infected Vero Cells. Nsp9, ORF1ab, ORF10 and nucleocapsid phosphoprotein were found to be up-regulated in the Vero cells after 24 h of infection, whereas the same proteins were found to be down regulated in the cells upon gramicidin S and melittin treatment (Table 1). Similarly, at 48 h post infection (hpi), ORF1ab, S protein, nucleocapsid phosphoprotein, helicase and RNA-dependent RNA polymerase were found to be up-regulated in the infected Vero cells which were found to be down regulated in the gramicidin S and melittin treated Vero cells (Table 1). Based on proteomics analysis, a total of 254 proteins were found to be differentially regulated and associated in infected and peptide treated Vero cells (Table 2 and Fig. 5a). It was found that majority of up and down regulated proteins were reversed with gramicidin S and melittin treatment at 24 and 48 hpi. RS28, K22E, K2C1, RL17 are few of the major down-regulated proteins which were found to be reversing their expression post peptide treatment. NPM, ACLY, CALX and F184B were found to be up-regulated in Vero cells after 24hpi, whereas peptide treatment showed reversal of the protein expression (Table 2). Heat map analysis showed that gramicidin S and melittin-treated cells at their respective time point post infection are rooted together against out rooting with  Interaction of gramicidin S and melittin with RBD domain of spike protein: In silico analysis. Both gramicidin S and melittin are membrane-active peptides and exert their antimicrobial activity by interacting with membrane components 13,14 . Both the peptides may exert their antiviral activity by targeting multiple regions of the virus. The ability of the peptides to bind to the receptor binding domain (RBD) of the SARS-CoV-2 spike protein was examined by molecular docking using web version of the program ZDOCK 27 . The structures shown in Fig. 6 indicate that both gramicidin S and melittin can bind to the RBD binding domain. Both the structures were in the top ten predictions in the ZDOCK output. Panel A shows the crystal structure of RBD-ACE2 complex. The interface between RBD and ACE are shown in violet color and stick representation respectively. The models of gramicidin S and melittin are shown in panels B and C respectively. The LigPlot 28 of the interacting amino acids are shown in panels D for gramicidin S and E for melittin. The residues highlighted in yellow are involved in RBD binding to ACE2 10 . The modeling study indicates that both the peptides can bind to RBD although their sequences are considerably different from the RBD binding region of ACE2. Both the peptides did not dock to the N-terminal region of ACE2 that binds to RBD region of the spike protein in the top 10 predictions (data not shown).

Discussion
During the past two decades, the world had witnessed infection by three highly pathogenic human corona viruses namely, SARS-Co-V, MERS, SARS-CoV-2 29,30 . They belong to the group of β-coronavirus and have the ability to cross animal-human barriers and cause serious illness in humans. The timely development of specific antivirals is of utmost importance.The development of vaccines at "warp speed" has led to decrease in mortality and serious illness caused by SARS-CoV-2 3 . However, it is still not well established that whether vaccines are equally effective against the several variants that are emerging, such as the omicron variant, or the time frame when immunity will be present. To-date, there is no specific drug for SARS-CoV-2. Hence, there is clearly an urgent need to develop therapeutic molecules that would effectively neutralize the virus rather than repurposing drugs used to treat   www.nature.com/scientificreports/ other viral infections. In case of HIV, drugs specific to the virus has been effective to treat the disease although there is still no effective vaccine. Likewise, malaria can be treated with specific drugs rather than vaccines. However, development of specific therapeutic antiviral drugs for clinical use in a short span is extremely challenging. Repurposing of drugs already known for their therapeutic effects have been extensively screened and tested for inhibition of SARS-CoV-2 1,6 . While many repurposed drugs have shown excellent anti SARS-CoV-2 activity in vitro, they have shown very little success when used clinically. We have shown that host defence peptide such as β-defensins may have a role in infection by SARS-CoV-2 as they are down regulated in COVID-19 patients 31 .
Bee venom has immunosuppressive activity and is generally used in contemporary medicine to treat Multiple Sclerosis, Parkinson's disease, and arthritis. Bee venom activates foxP3-expressing cells, CD25 and CD4 + T cells, and thus modulates the IgE antibody ratio, resulting in a variety of allergic reactions to antigens 32 . This immunosuppressive activity was observed in Wuhan beekeepers against COVID-19 32 . Melittin is the main component of bee venom, and it is active against both enveloped and non-enveloped viruses by activating the Toll-like receptors (TLRs) pathway, which reduces inflammatory cytokines like nuclear factor-kappa B (NF-kB), extracellular signal-regulated kinases (ERK1/2), and protein kinase Akt 32 . Melittin exhibited antiviral activity against several viruses in vitro 18 . Significant antiviral activity of the sitagliptin and melittin-nanoconjugates complex (IC50 value-8.439 µM) was observed against SARS-CoV-2 in vitro 19 . Binding of sitagliptin and melittin nano conjugates to the active site of SARS-CoV-2 3CLpro (a protease) was also observed through molecular docking 19 .
Gramicidin S has potent antibacterial and fungicidal activity 13 . Molecular docking revealed that gramicidin S has a binding affinity of 11.4 kcal/mol to the SARS-CoV-2 spike glycoprotein and SARS-CoV-2 papain like protease, implying that gramicidin S could be an effective drug against the SARS-CoV-2 virus 33 . SARS-CoV-2 is an enveloped virus, with the viral membrane essential for its integrity and function 1, 10 . We reasoned that membrane-active peptides would disrupt the viral membrane and render the virus ineffective. We have investigated the anti-SARS-CoV-2 activity of two well studied membrane-active antibacterial peptides gramicidin S and melittin.
Our present in vitro studies using gramicidin S and melittin showed EC 50 value of 1.571 µg for gramicidin S and 0.656 µg for melittin (Fig. 1). The results were comparable with remdesivir which we used as the assay control in our study (Fig. 5). The immunofluorescence studies are in agreement with our RT-qPCR data showing the decrease in the viral load in the gramicidin S and melittin treated experimental groups compared to peptide non treated group (Fig. 4). Molecular docking studies suggest that both the peptides have structural features that would favor binding to RBD of the spike protein (Fig. 6). ZDOCK used in the present study, has been used extensively to study protein-protein/peptide interactions 34 . However, our prediction that melittin and gramicidin S have structural features that favor interaction with RBD needs experimental validation. In fact, membraneactive antibacterial peptides have multiple targets in bacteria. It is conceivable that gramicidin S and melittin also have multiple targets on the virus thereby acting as an effective antiviral agent. Both gramicidin S and melittin have hemolytic activity at concentrations higher than the EC50 values reported in this study. Hemolytic activity can be attenuated or eliminated by engineering the peptide for specific antiviral activity. This approach has been successfully used in generating gramicidin S and melittin analogs with only antibacterial activity without any hemolytic activity 35,36 . In order to prevent off-target toxic effects of melittin, formulations with polymeric nanoparticles are being explored 37 . Formulations for use of peptides as therapeutics is a challenge but there are considerable efforts towards this goal 12 .
Proteomics studies indicate that metabolic change caused by SARS-CoV-2 pathogenesis result in long-term metabolic disorders in COVID-19 patients, and this varies according to pathogen severity. Carbon sources, specifically glycolysis and glutamino lysis pathways, have been found to play critical roles in SARS-CoV-2 viral replication and production 38 . Non-oxidative pentose phosphate pathways (PPP) are also involved in viral replication; Transketolase is a key mediator enzyme of PPP involved in ribonucleotide production. Benfo-oxythiamine, a TKT inhibitor, acted against SARS-CoV-2 infection and inhibited the viral replication 39 . Our findings from proteomic pathway analysis showed that several proteins are strongly associated with carbon metabolism and non-oxidative PPP. Specifically, LDHA, LDHB, ALDOA, TALDO, PGK1, and PGAM2 proteins were found to be up regulated in early viral replication, i.e. at 24 h viral induced cell control, and vice versa in gramicidintreated cells. TKT was found to be up regulated after 48 h of viral induced cell control, but it was substantially down regulated by melittin treatment. It suggests that gramicidin and melittin may function as viral inhibitors by suppressing intercellular metabolic regulators.
Another prominent pathway, mRNA processing, was identified during the local network pathway analysis. According to a recent study on SARS-CoV-2 RNA host protein interaction, the majority of virus which induced host RNA binding proteins prevents the virus induced cell death. Several mRNA binding proteins, including Heterogeneous nuclear ribonucleoproteins (HNRNPs), dead box RNA helicases (DDX), and NONO, were activated during the innate immune response to SARS-CoV-2 infection 40 . In our study, NONO, DDX5, RBMX, and HNRPM proteins were found to be up regulated with more than 1 log fold change in melittin-treated SARS-CoV-2 infected cells. This finding suggests that melittin may have antiviral activity during the early stages of viral infection by activating host RNA binding proteins.
The antimicrobial activity of gramicidin S and melittin have been well characterized. Our results strongly argue for development of peptides, gramicidin S and melittin, as potent therapeutic candidates to treat SARS-CoV-2 and possibly other influenza like viruses which are also enveloped viruses, for which there are no effective vaccines. Localized delivery at the site of infection in the nasopharyngeal region by appropriate formulations would avoid cytotoxicity due to systemic delivery.
In conclusion, our study indicates the potential of antibacterial peptides such as gramicidin S and melittin for development as therapeutic molecules to treat COVID19. These peptides have broad-spectrum antibacterial activity and resistance does not develop against them. It is likely that variants of SARS-CoV-2 which may escape immune surveillance, may be susceptible to membrane-active peptides such as gramicidin S and melittin. www.nature.com/scientificreports/ With tremendous advances in the formulation of drugs to minimize side-effects, it should be able to administer gramicidin S and melittin by appropriate formulations to avoid any non-specific cytolytic effects.

Materials and methods
Peptides. Gramicidin S and melittin were procured from commercial sources (Gramicidin S: 368108, Calbiochem CA, USA Melittin: M4171 from Sigma Chemical Co, USA). They were characterized by HPLC and mass spectrometry and found to be > 95% pure.
Cell viability using MTT assay. The Vero cells were plated in 96 well culture plate and incubated at 37 °C with 5% CO 2 . After attaining 90-95% cell confluency, different concentrations of gramicidin S and melittin (0.5, 0.7, 3, 5 µg for both) were added to the cells to check the effect of the peptides on the cells for 24 h. After 24 h, 100 µl (50 µg) of MTT substrate was added to the cells and the plate was continued to incubate for 3 h at 37 °C with 5% CO 2 . Later the formazan crystals formed were dissolved in 100 µl of DMSO and the absorbance was measured at 570 nm in Multimode Micro plate reader (Synergy HIM).