In Silico Study of Entry Inhibitor from Moringa oleifera Bioactive Compounds against SARS-CoV-2 Infection

al. In Silico Study of Entry Inhibitor from Moringa oleifera Bioactive Compounds against SARS-CoV-2 Infection. Pharmacogn J. 2022;12(5): 565-574.


INTRODUCTION
Coronavirus (CoV) also known as COVID 19, has spread worldwide in December 2019 and became a pandemic in January 2022. 1,2 WHO declared in March 2020 that this pandemic transmission is a person to person. The case of COVID 19 has enlarged widely to 213 countries and it caused more than 270 million infections over 5 million cases and those numbers still rising. 3 WHO confirmed the symptoms of COVID 19 were fever, dry cough, respiratory disorders, and olfactory and taste disorders. [4][5][6][7] The human coronavirus (HCoV) was positivestranded RNA virus. There were 2 types of protein of HCoV structural and non-structural protein that have different characteristics. The structural protein has characteristics including envelope, matrix, nucleocapsid, and spike. Besides, the non-structural protein has RNA-dependent RNA polymerase (RdRp). 8 RdRp has an important role in the HCoV life cycle and also became the main target factor for COVID 19 therapeutics. According to the Genome report, SARS-CoV-2depends on structure also reviewed the Ramachandran Plot value according to favored, allowed, and outlier regions. 21 Bioactivity and drug likeness prediction Bioactivity of active compounds were predicted according to probability values (Pa) through PASS online site (http://way2drug. com/passonline/). To be an effective drug, potential active compounds must be able to reach the target in the body. There were several characteristics that a drug must possess in order to reach the target in the body to be selected as a drug potential. The characteristics reviewed include molecular mass, TPSA value, solubility in lipids, and others. There were several parameters to be reviewed in drug-likeness, named Lipinksi, Ghose, Veber, Egan, and Muegge Parameter. Prediction of drug-likeness could be done through the SWISS ADME (http://www. swissadme.ch) website. Active compounds that fulfill five parameters will be selected. 8,22,23 Ligand and protein preparation The minimization energy process of the ligand was prepared with PyRx software. Ligand preparation aimed to increase flexibility and change the sdf format to pdb. Ligand preparation also to minimize the binding affinity. The target protein in this paper was the M pro and RdRp. Sterilization of the target protein from water and contaminant ligands was carried out by Discovery Studio software to increase the optimization of binding energy. 20,24,25 Molecular docking and dynamic simulation Molecular docking with the PyRx software was performed to predict the interaction of protein inhibition on SARS-COV 2 by active compounds from M. oleifera. 26 Molecular docking of the M pro target was done by blind docking. On the other hand, molecular docking of the RdRp target was done by specific docking at its catalytic sites: Gly-616, Trp-617, Asp-618, Tyr-619, Leu-758, Ser-759, Asp-760, Asp-761, Ala-762, Lys-798, Tys-799, Trp-800, Glu-811, Phe-812, Cys-813, and Ser-814. 27 The validation of the docking results were carried out with a dynamic molecular test by using CABS-flex 2.0. At this stage, the Fluctuation Plot tab on CABS-flex 2.0 showed the residue fluctuation profile due to the RMSF value for protein target. 28

Docking visualization
The analysis of the docking results was reviewed based on the 2D and 3D forms. Visualization of 2D docking results was done by Discovery Studio software. Moreover, the 3D visualization was carried out with PyMOL. Types of interactions and chemical bonds formed were analyzed using the Discovery Studio software. 26,29
Drug-likeness analysis aimed to identify molecules considered to be drugs built upon their physicochemical properties. The properties approaches aimed to measure drug likeness consist of octanol-water partition coefficient (ALOGP), number of H-bond acceptors (HBAs), number of H-bond donors (HBDs), molecular weight (MW), molecular polar surface area (PSA), number of aromatic rings (AROMs), number of structural alerts (ALERTS) and number of rotatable bonds (ROTBs). 31 Based on these properties, there are several relevant drug likeness rules such those proposed by Lipinski, Ghose, Veber, Eggan and Muegge. These rules suggest the compound as a drug based on their physicochemical properties. The Lipophilicity (log P o/w ) is the partition coefficient between water and n-octanol. Water solubility is the value of a drug's ability for oral targeting. SwissADME provides the number of violations in every rule. 30 The drug likeness prediction of bioactive compounds of M. oleifera ( Table 2) and drug likeness parameter of bioactive compounds of M. oleifera ( Table 3).
The Binding activity ability and molecules interaction of the bioactive compounds in the M. oleifera and target protein M pro was a cysteine protease that moderates the maturation cleavage of polyprotein in virus replication and also plays a crucial role in the Mawaddani N, et            A recent report has shown that remdesivir is able to inhibit replication of SARS-CoV-2 in in vitro and in vivo experiment. 38,39 Remdesivir is an adenosine analogue that inhibits SARS-CoV-2 RdRp. [32][33][34][35][36][37][38][39][40] Therefore, remdesivir can be used as a positive control in this study. According to the result docking simulations. 41 Based on the research showed that remdesivir probably binds to M pro stronger than to RdRp. 32 42 While the interaction residues between RdRp and remdesivir are 56 residues, 10 of those residues were involved in a catalytic activity such as Ala-558, Asp-684, Asp-760, Asp-761, Cys-813, Gly-559, Ser-682, Ser-759, and Ser-814. 43 In line with this study result shows that interaction RdRp to anthraquinone (Glu-811 These results indicate that catalytic sites Asp-760, Asp-761, and Cys813 were found in the interaction of RdRp and remdesivir, apigenin, chrysin, dibutyl phthalate, pterygospermin, and quercetin.

Molecular dynamics simulation of M. oleifera's bioactive compounds with SARS-CoV-2 glycoprotein
Simulation parameters and a set of distance restraints used by CABSflex. Molecular Dynamics (MD) simulations carried out to generate the best convergence between either CABS-flex simulation and protein fluctuation simulation in aqueous solution. MD was carried out by 10 nanoseconds in length. In addition, MD was derived by different force fields for globular protein. 28 MD aims to support molecular docking results. The root-mean-square fluctuation (RMSF) was a measure of the displacement of the position of the protein atom relative to the reference structure. RMSF analyzes the portions of structure that are fluctuating from their mean structure. 44 Figure 3 showed the information of flexibility of M pro with its interaction with Apigenin and Quercetin, figure 4 showed the information of flexibility of RdRp with its interaction with Pterygospermin and Quercetin. Based on, 45 the RMFS average value of M pro and Remdevisir is below 0.4. Meanwhile, based on the result of our molecular dynamics, the average of M pro -Apigenin's RMFS value at the catalytic site is 1.24 nm and the average of M pro -Quercetin's RMFS value at the catalytic site is 1.04 nm. It was a