Computational Study of Scorpion Venom ( Lychas Mucronatus ) Activity as Antimicrobial Peptides ( AMPs ) to the SARS-CoV-2 Main Protease for the Future Coronavirus Disease ( COVID-19 ) Inhibitors

The 2019 coronavirus pandemic disease (COVID-19) is still declared a global pandemic by the World Health Organization (WHO). Therefore, an effort that is considered effective in finding therapeutic agents is needed to prevent the spread of COVID-19 infection. One of the steps that can be chosen is by utilizing antimicrobial peptides (AMPs) from animal venom by targeting the specific receptor of SARS-CoV-2, namely the main protease (Mpro). Through this research, a computational approach will be conducted to predict antiviral activity, including protein-peptide docking using PatchDock algorithm, to identify, evaluate, and explore the affinity and molecular interactions of four types of antimicrobial peptides (AMPs), such as Mucroporin, Mucroporin-M1, Mucroporin-S1, and Mucroporin-S2 derived from scorpion venom (Lychas mucronatus) against main protease (Mpro) SARS-CoV-2. These results were then confirmed using protein-peptide interaction dynamics simulations for 50 ns using Gromacs 2016 to observe the molecular stability to the binding site of SARS-CoV-2 Mpro. Based on protein-peptide docking simulations, it was proven that the Mucroporin S-1 peptides have a good affinity against the active site area of SARS-CoV-2 Mpro, with an ACE score of −779.56 kJ/mol. Interestingly, Mucroporin-S1 was able to maintain the stability of its interactions based on the results of RMSD, RMSF, and MM/PBSA binding free energy calculations. The results of the computational approach predict that the Mucroporin-S1 peptide is expected to be useful for further research in the development of new antiviral-based AMPs for the COVID-19 infectious disease.

At the end of December 2019, several local health authorities reported groups of patients with pneumonia whose cause was unknown and were epidemiologically linked to the seafood market in Wuhan, Hubei Province, China . SARS-CoV-2 has been identified as a new pathogenic coronavirus which is predicted to be a major cause of these events Zhu et al., 2020). For this phenomenon, the World Health Organization (WHO) announced a global health emergency on 30 January 2020 . Pandemics increased rapidly with first emergency meeting estimated the COVID-19 mortality rate to be around 4% . Literature search related to this COVID-19 is needed to summarize the characteristics of SARS-CoV-2 especially for diagnosis, therapy, and prevention of this infectious disease (Stoecklin et al., 2020).
Main protease (Mpro) from coronavirus can be one of the potential targets in the discovery of new antiviral candidates. Mpro plays an important role in the proteolytic process to produce two polyproteins (pp1a and pp1ab) which are used for coronavirus replication and transcription (Ge et al., 2013;Lu et al., 2020). Previous studies revealed that SARS-CoV-2 has structural similarities with SARS-CoV including the Mpro section based on complete phylogenetic analysis of the genome (Chen et al., 2009;Letko, Marzi, Munster, 2020). Interestingly, aside from the fact that this enzyme only exists in viruses and not in humans, the high conservation of Mpro among related viruses and its importance in virus replication makes this enzyme an attractive target for potential antivirals (Walls et al., 2020;Zhang et al., 2020). In addition, because no human proteases with the known specific cleavage specificity, such inhibitors are unlikely to be toxic (Park et al., 2016;Walls et al., 2019).
Among several sources of natural products, animal venoms have revealed great potential for drug development. Even though there is a dangerous mechanism of action from these animal venoms, but it has components that are potential drugs to cure disease. It is widely reported in the literature that animal venoms are rich in antimicrobial substances and contain a variety of biologically active compounds with different chemical structures (Hmed, Serria, & Mounir, 2013). Antimicrobial peptides (AMPs) are a group of peptides with important functions in the response of innate immune hosts when attacked by pathogenic organisms, such as fungi, bacteria, and viruses that are considered the first line of defense of many organisms, including plants, insects, and vertebrates (Bahar, & Ren, 2013;Jenssen, Hamill, & Hancock, 2006).
To date, there is an increasing need to design effective antiviral candidates for SARS-CoV-2. Several AMPs that have activity against SARS-CoV, namely peptides Mucroporin, Mucroporin-M1, Mucroporin-S1, and Mucroporin-S2 which are derived from scorpion venom (Lychas mucronatus) (Chen et al., 2012;Hong et al., 2014;Li et al., 2011;Zhao et al., 2012). In this research, an observation will be made of the molecular interactions between the AMPs and SARS-CoV-2 Mpro. Computational studies can be used to identify, evaluate and explore the affinity of AMPs as potential inhibitors of SARS-COV-2 (Kumar, Maurya, Prasad, Bhatt, & Saxena, 2020). In particular, SARS-CoV-2 Mpro is considered a target because it is a major part of forming coronavirus characteristics. Therefore, through this research, it is expected to obtain the molecular structure of reference AMPs for the treatment of COVID-19 infection.

EXPERIMENTAL SECTION Material
The crystal structure macromolecules used in this study were the main protease (Mpro) of SARS-CoV-2. The target macromolecules were obtained from the Protein Data Bank website (http://www.rcsb.org/pdb) with the PDB ID 6LU7. The antimicrobial peptide (AMPs) molecules used in this study was the sequencing of Mucroporin, Mucroporin-M1, Mucroporin-S1, and Mucroporin-S2 peptides derived from scorpion venom (Lychas mucronatus).
The main protease (Mpro) macromolecules used in this study were Mpro from the novel coronavirus 2019 (COVID-19 or SARS-CoV-2) downloaded from Protein Data Bank (http://www.rcsb.org/pdb) with PDB ID 6LU7 . SARS-CoV-2 Mpro macromolecular preparation was accomplished by removing water molecules and natural ligands, adding polar hydrogen atoms, and calculating the Kollman charge using MGLTools 1.5.6 with AutoDock 4.2.
The results of peptide molecular modeling were used as input for protein-peptide docking simulations.

Identification of the Peptide-Binding Sites in SARS-CoV-2 Mpro
The SARS-CoV-2 Mpro macromolecules that have been prepared were then identified, evaluated, and explored binding site areas that are most responsible for antimicrobial activity using BIOVIA Discovery Studio 2020 (BIOVIA, 2020). All amino acid residues that found around the natural ligand within a spherical radius distance set were prepared for the location of peptide-protein binding in protein-peptide docking simulations.

Validation of the Protein-Peptide Docking Methods
Before a protein-peptide docking simulation was performed between the Mucroporin peptide molecules and the SARS-CoV-2 Mpro macromolecules, the method must first be validated to determine the parameters to be used in the molecular docking simulations. Method validation of molecular docking parameters was done using the re-docking method. In this process, the value of the Root Mean Square Deviation (RMSD) is limited to a maximum radius of 2 Å, and the protein-peptide method was chosen for the complex type (Bikadi & Hazai, 2009;Hevener et al., 2009).

Protein-Peptide Docking Simulations
Protein-peptide docking simulations were performed using PatchDock (Aruleba et al., 2018;Fakih, 2020;Sathya & Rajeswari, 2016). All AMPs molecules for these simulations were modeled and added hydrogen atoms from the de novo methods. Complex types were selected as protein-peptide with default clustering RMSD 2 Å. The representation of the Connolly dot surface of the molecule into different components including convex, concave, and flat patch was generated through the PatchDock algorithm. PatchDock was optimized, refined, overhauled, and reselected the side chain interface from the top 10 candidate solutions. It also changes the orientation of the molecule relative by limiting flexibility in the side chains of the interacting surface and allowing the movements of small rigid-body. Observation of the protein-peptide docking simulation results was accomplished using BIOVIA Discovery Studio 2020 (BIOVIA, 2020).

Protein-Peptide Interaction Dynamics Simulations
Molecular dynamics simulations were conducted on all protein-peptide complexes resulting from molecular docking studies. Simulations were performed using Gromacs 2016 (Pronk et al., 2013;Abraham et al., 2015;Pall et al., 2015) and simulation results were analyzed with VMD 1.9.2 (Humphrey, Dalke, & Schulten, 1996) and BIOVIA Discovery Studio 2020 (BIOVIA, 2020). AMBER99SB-ILDN and AMBER general forces fields (GAFF) were used to parameterize the SARS-CoV-2 Mpro macromolecules and Mucroporin peptide molecules (Aliev et al., 2014). The longdistance electrostatic forces are determined using the Ewald Particle Mesh method (Darden, York, & Pedersen, 1993;Essmann et al., 1995). Neutralization of the system was done by adding sodium (Na+) and chloride (Cl−) ions. The cubic TIP3P water model was used to solve complex systems. The simulation step includes minimization, heating up to 310 K, temperature equilibrium (NVT), pressure equilibrium (NPT), and the production process with a timestep of 2 fs for 50 ns. System stability was identified and evaluated by analysis of energy, temperature, pressure, Root Mean Square Deviation (RMSD), and Root Mean Square Fluctuation (RMSF) of the SARS-CoV-2 Mpro macromolecules. Analysis of the interaction stability of Mucroporin peptide molecules and SARS-CoV-2 Mpro macromolecules was carried out by calculating the residual values of RMSD and RMSF at the binding site during the simulation.

MM/PBSA Binding Free Energy Calculations
The calculation of binding free energy as a result of molecular dynamics simulations using the Molecular Mechanics Poisson-Boltzmann Surface Area (MM/PBSA) method was demonstrated using the g_mmpbsa package (Baker, Sept, Joseph, Holst, & McCammon, 2001;Kumari, & Kumar, 2014) which is integrated in the Gromacs 2016. Polar desolvation energy was calculated using the Poisson-Boltzmann equation using a grid size of 0.5 Å. The dielectric constant of the solvent was set to 80 to represent water as solvent (Hou et al., 2010;Spackova et al., 2003). The nonpolar contribution was determined by calculating the surface area accessible to the solvent with a solvent radius of 1.4 Å. The binding free energy of a protein-peptide complex was determined based on 50 snapshots taken from the beginning to the end of the molecular dynamics simulations trajectories of the complex molecule.

RESULTS AND DISCUSIONS Protein-Peptide Docking Simulations
All modeled antimicrobial peptides (AMPs) derived from scorpion venom (Lychas mucronatus) were then simulated by the protein-peptide binding method to the main protease (Mpro) of SARS-CoV-2 that acts as a target macromolecule (Figure 1). In addition, prediction of the properties and physico-chemical characteristics of the four types of AMP molecules was also performed. Table 1 shows that the Mucroporin-S1 and Mucroporin-S2 peptide molecules are composed of polar amino acid residues, while the Mucroporin and Mucroporin-M1 peptide molecules are constructed with non-polar amino acid residues. Therefore, it can be predicted that each peptide molecule will occupy part of the SARS-CoV-2 Mpro based on its physicochemical properties.
The docking simulation results show that Mucroporin, Mucroporin-M1, Mucroporin-S1, and Mucroporin-S2 have a good affinity compared to natural ligands with the active site of SARS-CoV-2 Mpro, with ACE scores of −300.83 kJ/mol, −377.19 kJ/mol, −779.56 kJ/mol, −648.52 kJ/mol, and −149.94 kJ/mol, respectively (Table 2). This phenomenon shows a promising sign that the four types of AMPs have strong interactions and bonds at the active site area of the target macromolecule. Thereafter, all protein-peptide docking complexes were selected for further studies using the protein-peptide interaction dynamics simulations.

Mucroporin
Mucroporin-M1 Mucroporin-S1 Mucroporin-S2  Note: PCP_PC = positively charged residues, PCP_NE = neutral charged residues, PCP_PO = polar residues, PCP_NP = non-polar residues, PCP_AL = residues having aliphatic side chain, PCP_AR = aromatic residues, PCP_BS = basic residues, PCP_NE_pH = neutral residues based on pH, PCP_HB = hydrophobic residues, PCP_HL = hydrophilic residues, PCP_NT = neutral residues, PCP_HX = hydroxylic residues, PCP_TN = tiny residues, PCP_SM = small residues, PCP_LR = large residues. Mucroporin-S1, and Mucroporin-S2 peptides exhibit molecular interactions that were identical to SARS-CoV-2 Mpro, which were in polar patches in the active binding sites of target macromolecules (Figure 2). Similar to the results of the ACE score, the peptide Mucroporin-S1 has many interactions with SARS-CoV-2 Mpro including eight hydrogen bonds (with Gly23, Thr24, Thr25, Cys44, Thr45, Ser46, Asn142, and His164), one hydrophobic interaction (with Cys145), and one electrostatic interaction (with His41). Whereas the only 5 interactions that can be formed by the peptide Mucroporin-S2 include two hydrogen bonds (with Phe 140 and His163) and three hydrophobic interactions (with Leu27, Met49, and Cys145). This interaction can be formed because of hydrogen bonds, especially AMPs which act as donor hydrogen bonds and protein amino acid residues as hydrogen bond acceptors. Most hydrogen bonds between proteinpeptide are quite strong, with average bond lengths in the range of 3 Å. In addition to hydrogen bonds, the interaction between the three AMPs and SARS-CoV-2 Mpro was also dominated by hydrophobic interactions and electrostatic interactions. It can be predicted that hydrogen bonds, hydrophobic interactions, and electrostatic interactions that contribute to macromolecular proteins play an important role in stabilizing protein-peptide complexes.
In contrast to other AMPs, Mucroporin attaches to the non-polar patches of the surface of SARS-CoV-2 Mpro. Moreover, the Mucroporin peptide had ten unfavorable interactions with amino acid residues at the active sites of SARS-CoV-2, including Arg4 and Lys5. Then, although occupying the same area as the peptides Mucroporin-S1 and Mucroporin-S2, many unfavorable interactions were formed with Gly23, Thr24, Ser46, Gly143, Cys145, Asn142, Gln189, and Glu166 causing the peptide Mucroporin-M1 to have a low effect on SARS-CoV-2 Mpro. This phenomenon shows that different conformations of the peptides Mucroporin and Mucroporin-M1 complexes can be caused by this interaction and predicted to affect their biological activity as antimicrobials.

Protein-Peptide Interaction Dynamics Simulations
After the protein-peptide complex is formed in the previous stage, further identification, evaluation, and exploration are carried out using molecular dynamics methods. The purpose of this molecular dynamics simulation is to observe the stability of each proteinpeptide complex during a 50 ns simulation. The best complex stability and strong peptide bonds to SARS-CoV-2 Mpro are predicted to be able to inhibit the entry of coronavirus into cells and host tissues due to the inability to continue infection signaling. It is also important to explore visualizations, Root Mean Square Deviation (RMSD) graphs, and Root Mean Square Fluctuation (RMSF) graphs during molecular dynamics simulations.
Based on the snapshots taken at the beginning of the simulation (0 ns), 25 ns, and the end of the simulation (50 ns) it can be observed that the conformational changes of each peptide molecule generally occur when the simulation has reached 10 ns (Figure 3). The peptide molecule's position changes during the simulation, but significant changes are shown by the Mucroporin peptide molecule. The molecular structure of the peptide has transformed into a loop that was originally alpha-helical. Different phenomena occur in the peptide molecules Mucroporin-M1, Mucroporin S-1, and Mucroporin-S2 whose conformation does not change significantly.
The dynamics of molecular interactions between peptide molecules and their target macromolecules are studied based on simulations in solvents explicitly using Gromacs 2016. Strong affinity tends to reduce the movement of the atoms which are bound and generally stabilizes the binding site areas of SARS-CoV-2 Mpro. This phenomenon was analyzed by calculating the RMSD from the SARS-CoV-2 Mpro macromolecular binding site during the 50 ns simulation to ensure the stability and rationality of the selected conformation. The RMSD graph in Figure 4 shows that the three systems of Mucroporin-M1, Mucroporin-S1, and Mucroporin-S2 did not have significant differences. However, similar to the visualization of snapshots during simulation, the Mucroporin peptide molecule is unable to stabilize the SARS-CoV-2 Mpro macromolecule. This is evidenced by the average RMSD value above 4 Å so that it shows a fluctuating graph after the simulation reaches 10 ns.
The RMSD data are also supported by identifying the RMSF values of amino acid residues found at the SARS-CoV-2 Mpro macromolecular binding site ( Figure  5). However, the peptide molecules Mucroporin-M1, Mucroporin-S1, and Mucroporin-M2 were only able to bind strongly with amino acid residues at the active site of the SARS-CoV-2 Mpro macromolecule. These residues include Gly23,Thr24,Thr25,Thr26,His41,Cys44,Glu47,Met49,Asn119,Asn142,Gly143,Cys145,Met165,Glu166,Pro168,Gln189,and Ala191. Meanwhile, the Mucroporin peptide molecule moved away from the active site of SARS-CoV-2 Mpro during molecular dynamics simulations, namely interacting with the amino acid residues Ala285, Leu286, Asn277, Gly278, Arg279, Ser284, Ala285, and Leu286 (Table 3). High flexibility occurs around the amino acid residues found at the end of the peptide molecule because these residues play an important role in binding to the target.

MM/PBSA Binding Free Energy Calculations
Apart from identifying the visualization for each trajectory, the Root Mean Square Deviation (RMSD) graphs, and the Root Mean Square Fluctuation (RMSF) graphs, then the binding free energy calculation results from the molecular dynamics simulations from the beginning to the end of the simulations were also performed. Based on the results of the MM/PBSA calculations in Table 4, it can be observed that the Mucroporin-M1 and Mucroporin-S1 peptide molecules have better binding free energy compared to the Mucroporin and Mucroporin-S2 peptide molecules against the SARS-CoV-2 Mpro macromolecules.
The peptide molecules Mucroporin-M1 and Mucroporin-S1 have binding free energy values of −694.37 kJ/mol and −227.41 kJ/mol during molecular dynamics simulations. Meanwhile, the peptide molecules Mucroporin and Mucroporin-S2 have binding free energy values of −141.18 kJ/mol and −123.69 kJ/mol. Especially, the energy that contributed the most during the simulation was van der Waals interactions and electrostatic interactions. This phenomenon because the MM/PBSA approach allows observation of the influence of the contribution of van der Waals and electrostatics as well as changes in protein-peptide affinity that are influenced by the solvation process of complex systems.

CONCLUSION
Mucroporin-S1 was able to bind stably against binding site areas of SARS-CoV-2 Mpro. Interestingly, Mucroporin-S1 has the strongest affinity and interactions with the target macromolecular active site, with the ACE score of −779.56 kJ/mol. These AMPs were also able to stabilize the active site areas of the SARS-CoV-2 macromolecule during molecular dynamics simulations for 50 ns. Therefore, the results of this study indicate that Mucroporin-S1 has the potential to be further developed as SARS-CoV-2 Mpro inhibitor candidates in the treatment of COVID-19 infectious diseases.