In silico evaluation of the compounds of the ayurvedic drug, AYUSH-64, for the action against the SARS-CoV-2 main protease

Background Outbreak of Corona Virus Disease in late 2019 (COVID-19) has become a pandemic global Public health emergency. Since there is no approved anti-viral drug or vaccine declared for the disease and investigating existing drugs against the COVID-19. Objective AYUSH-64 is an Ayurvedic formulation, developed and patented by Central Council of Research in Ayurvedic Sciences, India, has been in clinical use as anti-malarial, anti-inflammatory, anti-pyretic drug for few decades. Thus, the present study was undertaken to evaluate AYUSH-64 compounds available in this drug against Severe Acute Respiratory Syndrome-Corona Virus (SARS-CoV-2) Main Protease (Mpro; PDB ID: 6LU7) via in silico techniques. Materials and methods Different molecular docking software's of Discovery studio and Auto Dock Vina were used for drugs from selected AYUSH-64 compounds against SARS-CoV-2. We also conducted 100 ns period of molecular dynamics simulations with Desmond and further MM/GBSA for the best complex of AYUSH-64 with Mpro of SARS-CoV-2. Results Among 36 compounds of four ingredients of AYUSH-64 screened, 35 observed to exhibits good binding energies than the published positive co–crystal compound of N3 pepetide. The best affinity and interactions of Akuammicine N-Oxide (from Alstonia scholaris) towards the Mpro with binding energy (AutoDock Vina) of −8.4 kcal/mol and Discovery studio of Libdock score of 147.92 kcal/mol. Further, molecular dynamics simulations with MM-GBSA were also performed for Mpro– Akuammicine N-Oxide docked complex to identify the stability, specific interaction between the enzyme and the ligand. Akuammicine N-Oxide is strongly formed h-bonds with crucial Mpro residues, Cys145, and His164. Conclusion The results provide lead that, the presence of Mpro– Akuammicine N-Oxide with highest Mpro binding energy along with other 34 chemical compounds having similar activity as part of AYUSH-64 make it a suitable candidate for repurposing to management of COVID-19 by further validating through experimental, clinical studies.


Introduction
The outbreak of corona virus disease (now known as COVID- 19) in late 2019 in Wuhan, China led to a global public health emergency that currently affects more than 200 countries worldwide. Coronavirus, crown-like spikes positive-sense ribonucleic acid (RNA) from Coronaviridae family. Over the past two decades, severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV) led to pandemic in 2002 and 2012 resulted in a 10% and 37% mortality rate respectively [1]. In December 2019, a new coronavirus of probable bat origin caused an outbreak of human pulmonary disease in Wuhan, China, that quickly spread throughout the world [2,3]. The causative agent initially named as 2019-novel coronavirus (2019-nCoV) by world health organisation (WHO). As per name recommendation from International Committee of Corona virus Study Group (ICCSG), the virus named as Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2), its RNA genome shares 82.30% and similar pathogenesis (host) of its identity with SARS coronavirus (SARS-CoV) [1,4]. Ultimately, the disease became known as corona virus disease 2019 (COVID-19) (WHO, 2020). On January 30, 2020, WHO officially declared the COVID-19 epidemic a public health emergency of international concern, and on March 11, 2020, the outbreak was confirmed a pandemic (WHO, 2020). The SARS-CoV and SARS-CoV-2 shares the similar pathogenesis (host) due to similarities in their genome [1].
Though several drug targets of coronaviruses have been identified, the main protease (M pro ), also known as 3-chymotrypsin-like protease (3CL pro ), has emerged as the best-described drug target [5]. The M pro processes the large polypeptide, translated from the viral RNA, specifically at 11 splicing regions mostly Leu to Gln (Ser, Ala and Gly). These significant splicing regions non-homologous to human. Inhibiting enzyme activity might results in blocking viral pathogenesis, making it an attractive drug target for SARS-CoV-2. Further, a mechanism-based peptide-like inhibitor (N3) has also been identified via computer-aided drug design, and then crystal structure of the M pro of SARS-CoV-2 in complex with this compound has been determined [6].
Given the lack of any validated, approved therapeutic intervention, such as a drug or vaccine, the COVID-19 pandemic continues to spread around the world. In such circumstances, the concept of drug repositioning might be a cost-effective, time-efficient, and less labour-intensive option for development of possible therapeutic and/or prophylactic lead candidate(s) from existing traditional and/or approved drugs [7]. In comparison to synthetic inhibitors, plant based-drugs are known to have low toxicity and are much safer to use. Natural products, such as traditional medicines and plant-derived compounds (phytochemicals) are considered rich sources of promising antiviral drugs [8].
Natural products were the sources of approximately 44% of the approved antiviral drugs from 1981 to 2006 [9]. Recent studies have suggested that drug poisoning and severe to moderate adverse effects associated with hydroxychloroquine have been reported in diabetic and hypersensitive patients, in whom COVID-19 tends to affect severely. Further, hydroxychloroquine is also known for inhibition of pro-inflammatory cytokines which leads to Acute Respiratory Distress Syndrome (ARDS) [10]. Due to the severe adverse effects shown by these synthetic compounds, natural, nonsynthetic, target specific drugs with minimal side effects are urgently required for prophylaxis and treatment of COVID-19 [11].
Plant compounds and natural phyto-molecules are most economical and ideal for exploration to develop drugs through the drug discovery process [12]. Recent studies via integrated computational approaches have highlighted the repurposing of approved and known drugs against SARS-CoV-2 drug targets, either individually or in combination, to combat the virulence of COVID-19. Several plant compounds have been reported, using in silico approaches, as potential inhibitors of the M pro of SARS-CoV-2. Mimicked drugs of bictegravir, doultegravir, paritaprevir, and raltegravir have been studied against 3CLpro and 2 0 -OMTase through in silico approaches [13,14]. Another study reported darunavir, remdesivir, and saquinavir, the natural compounds derivatives from coumarine and flavone which inhibits 3CLpro [13e15]. In another in silico study also shown that, Andrographis paniculata's compound Andrographolide, inhibits M pro of SARS-COV-2. Its good solubility, pharmacodynamic properties, target accuracy, and adherence to Lipinski's rule of five were also predicted using computational tools [11].
The potent plant compounds nelfinavir, lopinavir, kaempferol, quercetin, luteolin-7-glucoside, demethoxycurcumin, naringenin, apigenin-7-glucoside, oleuropein, curcumin, catechin, and epicatechin-gallate, have been shown to inhibit the M pro [12]. At the same time, several potential anti-SARS-CoV-2 lead compounds have been revealed in the medicinal plant library of traditional Chinese medicine compounds [16]. Another study identified oolonghomobisflavan-A from the tea plant as a potential inhibitor against M pro of SARS-CoV-2 [17]. In addition, a promising drug candidate from plant sources, bonducellpin D, has been reported to exhibit broad-spectrum inhibition potential against both the SARS-CoV M pro and the MERS-CoV M pro [18].
AYUSH-64 is a poly-herbal compound developed by the Central Council for Research in Ayurvedic Sciences, the main organization for research and development in Ayurveda in India under the Ministry of AYUSH, Government of India. It consists of four ingredients viz., Saptaparna (Alstonia scholaris R. Br.) bark aqueous extract 100 mg, Katuki (Picrorhiza kurroa Royle ex. Benth) root extract 100 mg, Kiratatikta (Swertia Chirata Pexbex. Karst) wholeplant extract 100 mg, and Kuberaksha (Caesalpinia crista L.) seed powder 200 mg [19].
To summarize the information provided in Table 1, it is understood that all four ingredients of AYUSH-64 are characterized by tikta rasa (bitter taste) and therefore amapachak (able to digest the Ama or undigested form of food) and hence act as Jvaraghna (antipyretic). The combined effects of these herbs are jwarahara (able to relieve fever), sannipata jwarahara (able to relieve intermittent fevers), krimihara (wormicidal), jantuhara (anthelmintic/antimicrobial/antiviral), and shothahara (anti-inflammatory), rendering this compound a potent combination against conditions such as Influenza Like Illnesses having the symptoms of cough, cold, headache and fever. Several compounds found in the drug have been reported to exhibit several pharmacological activities including antimalarial, anti-viral, anti-inflammatory, immunomodulatory etc. (Table 2).
Various pre-clinical studies were performed on AYUSH-64 to study its safety and efficacy aspects ( Table 3). The major studies, namely anti-malarial property in albino mice, found safe and nontoxic in a dose of 500 mg/kg of body weight for 12 weeks (Anonymous, 1987) and potential antiviral activity of Chirakin (marketed name of AYUSH-64 by Zandu) against Chikungunya virus were investigated at the department of Molecular Virology Laboratory, Rajiv Gandhi Centre for Biotechnology, Thiruvananthapuram, by protection and plaque reduction assay. The efficacy of the compound was expressed in terms of activity index and selectivity index in protection assay and Vero cells were infected with Chikungunya Virus (CHIKV) and treated with an optimum concentration of Chirakin (25 mg/mL). Chirakin treated cells and the reduction in virus yield was determined by plaque assay in Vero cells. Chirakin shows antiviral activity against Chikungunya virus in this study. Its activity was found better than ribavirin in protection assays, where as in plaque reduction assays, both perform almost equally with a 2 log 10 reduction in virus number (CK Katiyar. Technical note on Chirakin Tab eUnpublished report).
Attempts have also been made to clinically evaluate the compound for management of Microfilariasis [20,21] effectiveness against Plasmodium vivax in ring stage than in gamete stage and also the mixed infection cases of P. vivax [22] decline in infectivity rate of malaria, double-blind study for comparative efficacy in comparison with chloroquine/primaquine in P. vivax malaria and lead candidate in the management of influenza like illnesses [23] ( Table 4). Overall, the compound is considered to be safe and is prescribed widely by Ayurvedic experts for the effective management of malaria, joint pains, fever, and influenza like illnesses. In this in silico study, 36 major compounds were analysed which are four ingredients of AYUSH-64 drug might be potent inhibitors against M pro of SARS-CoV-2 observed via two different docking strategies of AutoDock Vina and Discovery studio molecular docking tools.

Materials and methods
In the drug discovery process for COVID-19 crisis, the development of novel drugs with potential interactions with therapeutic targets is of primal importance. Conventionally, promising-lead or drug identification is achieved by traditional or experimental highthroughput screening (HTS), which is time-consuming and costeffective [24]. In contrast to the traditional drug discovery method (classical or forward pharmacology), the rational drug design is noted as efficient and economical [25]. In these contrasts, approach to identifying therapeutics is to repurpose approved compounds developed for other uses, by taking advantage of existing detailed information on human pharmacology and toxicology to enable rapid clinical trials and regulatory quick review. Furthermore, in silico drug design method is also known as reverse pharmacology, as the initial step is to identify promising drug targets (proteins/enzymes), after which they are used for screening of novel small-molecule inhibitors for the candidates [14,26e28]. In the present study of in silico drug design method, molecular docking approach was used for SARS-CoV-2 M pro and 36 AYUSH-64 compounds, to screen for the novel inhibitors by two different docking strategies/algorithms, LibDock protocol of BIOVIA Discovery Studio and AutoDock Vina.

S.No Name of the plant
Chemical Compound(s) & pharmacological activities.

1.
Saptaparna (Alstonia scholaris R. Br.) Methanol extracted from bark of Alstonia scholaris Found to be more promising. Nevertheless, the observed antiplasmodium activity of Alstonia scholaris was less pronounced as compared to Alstonia macrophylla [44]. Alstonia macrophylla, which is seen in India, generally adultrated for Alstonia sholaris [44]. Four new alkaloids, alstiphyllanines A-D (1 e4), were isolated from Alstonia macrophylla, and their structures were determined by MS and 2D NMR analyses. Alkaloids 1e4 showed moderate anti-plasmodial activity against Plasmodium falciparum and vasorelaxant activity against phenylephrine-induced contraction of isolated rat aorta. Other Studies: 1. Inhibited the carrageenan-induced inflammation in the rat paw oedema study model [45].
2. Stimulates the cell-mediated and humoral components of the immune system as well as phagocytosis in experimental animals [53,54]. 3. Improves the immune system by increasing the proliferation of lymphocytes and cytokine levels (IL-4 and IFN-gamma) in serum, in HA titre, DTH, PFC, phagocytic index and CD4/CD8 population [55]. 4. Another study demonstrates the antioxidant and free radical scavenging activity of the leaf extract of same plant [56]. 5. Inhibits the growth of Plasmodium falciparum Significantly [57e59].

3.
Kiratatikta (Swertia Chirata Pexbex. Karst) The drug shows promising effect due to itsantipyretic (Bhargava, 2009), and antimalarial activity [58]. 1. Anti-protozoal activity:A MeOH extract of Swertia chirata found to inhibit the catalytic activity of topoisomerase I of Leishmania donovani was subjected to fractionation to yield three secoiridoid glycosides: amarogentin (a), amaroswerin (b), and sweroside (c). Amarogentin is a potent inhibitor of type I DNA topoisomerase from Leishmania and exerts its effect by interaction with the enzyme, preventing binary complex formation. Other Studies: 2. Inhibits the expression of viral protein R (an attractive target for HIV disease) in HeLa cells harbouring the TREx plasmid encoding full-length Vpr (TREx-HeLa-Vpr cells) [60]. 3. Inhibited HSV-1, plaque formation at more than 70% level and viral dissemination [61]. 4. Suppressive effects on inflammatory mediators by blocking the expression of COX-2 and phosphorylation of Akt, IKK-b, MAPK and NF-kB, activation in LPS-stimulated macrophages [62]. 5. A study on experimental arthritis in rats suggested amelioration of oxidative and inflammatory stress, which implies the immunomodulatory effect of leavesof the above said plant. Post treatment with the leaves of the said plant, the animal subjects showed marked reduction in inflammation as well as arthritic changes [63].

Ligand preparation
Selected 36 compounds of AYUSH-64 were obtained from the PubChem database, as 3D coordinates of structure-data file (SDF) format, and each compound was converted to PDBQT file format with the help of PyMol and ADT tool, generating an input compound/ligand file for docking study in AutoDock Vina. In BIOVIA Discovery studio, 225 maximum conformations were generated for selected 36 compounds, and separate isomer conformations were also created within the threshold of 20.0 kcal/mol relative energy. The CHARMm force field (fast and accurate) was applied to minimize the compounds with 1000 steps of steepest descent (SD) algorithm. The root-mean-square deviation (RMSD) with <1.0 Å were considered as duplicates, and the higher specified RMSD value will reduce the number of output ligand poses and RMS gradient 0.001 [15].

ADME and toxicity prediction
To predict the drug-likeness properties of selected 36 AYUSH-64 compounds, in silico absorption, distribution, metabolism, excretion, and toxicity (ADME/T) prediction was carried out. We explored the ADME/T properties of the 36 AYUSH-64 selected compounds using the ADME/T Protocol in the Discovery Studio software package (Accelrys, San Diego, CA, USA). These examinations were exclusively founded on the substance structure of the particle. Some of the parameters that were calculated included, ADME 2 Dimenctional Fast Polar Surface Area (ADME 2D FPSA), Atom-based Log P98 (ALogP98), Blood Brain Barrier (BBB), Cytochrome P4502D6 (CYP2D6), and Hepatotoxicity (HEPATOX).

Molecular docking
The prepared 36 AYUSH-64 compounds (ligands) were docked on an individual basis to the M pro of SARS-CoV-2 with grid coordinates (grid center). The selected compounds were set flexible with conformation condition while interacting with the macromolecule of M pro under rigid conditions of the protein. The prepared configuration file was run through ADT command line with PDBQT file as input with grid size of X: 25 Â Y: 25 Â Z: 25 points, grid centre set at X: 11.74, Y: 13.63 and Z: 70.63 dimensions having 1000 Å grid radius. By using AutoDock Vina scoring algorithm to measure Gibbs Free Energy (ÀDG kcal/mol) for selected AYUESH-64 compound binding affinities [29].
The LibDock protocol module of BIOVIA Discovery Studio was used for molecular docking studies of the 6LU7 and 36 AYUSH-64 compounds, the binding site is created with the same dimensions of the AutoDock Vina i.e. 4 Å radius of N3 peptide binding site. The LibDock protocol is a high-throughput algorithm developed by Diller and Merz, in 2001 [29], consisting the polar and a polar feature as "Hotspots." Among the 36 compounds of AYUSH-64, Table 4 Clinical studies of AYUSH-64.

S.No Clinical studies of AYUSH-64 1
Clinically effective in cases of microfilariasis [20,21]. 2 81% curative effect on Plasmodium vivax, drug is more effective in ring stage than in Gamete stage. In the cases of mixed infection of P. vivax and P. falciparum the curative effect was found to be 75% after longer therapy [69]. 3 Good effect in the management of microfilaraemia [70]. 4 In a study group of 4500 participants, AYUSH-64 was observed to be safe and nontoxic with good anti-malarial activity and also a major decline in the infectivity rate was noted by administering AYUSH-64 along with anti-mosquito measures, in various types of fevers [70]. 5 A double-blind study demonstrated comparative efficacy of AYUSH-64 to the chloroquine/primaquine as the standard modern control in sixty cases of P. vivax malaria [70]. 6 In a prospective, open-label, nonrandomized, single group, single-centre pilot study with pre-test and post-test design one-week intervention of 'AYUSH-64' in a dose of 3 g/day effectively helped to recover from Influenza Like Illnesses, and also the quicker return to normal life with reduced frequency of usage of acetaminophen/antihistaminic. No adverse effects were found during the study [71] 7 A pilot study done on 112 participants produced 97.14% result with prophylactic treatment in P.F. Malaria [72]. which were passed the applied filters while ligand preparation process, subjected to dock with the prepared 6LU7 crystal structure. The prepared compounds were docked to analyses and compare interacting amino acids. The post-docking analyses of two different docking strategies were visualized for H-bond, hydrophobic molecular interactions and bond distances with 4 Å radii in the Discovery Studio v2020 in the M pro of SARS-CoV-2. The top ranked molecular docked complex selected for molecular dynamics simulations using Desmond v2019-4 for 100 nano seconds (ns).

Molecular dynamics simulations
The first ranked docked molecular complex, AYUSH-64 with M pro selected for carrying out the dynamic's simulations using Desmond v2019-4 protocol. Where initially the selected molecular docked complex atoms calibrated by the "Optimized Potentials for Liquid Simulation (OPLS-2005)" force field. Further, geometric chemical structure corrections in molecular docked complex are performed using "Protein Preparation Wizard (macro model)". To put the molecular docked complex in solvent environment, Simple Point charge-HOH model was used with the cubic box dimensions 10 Å Â 10 Å Â 10 Å, and for maintaining the acid-base balance for the molecular docked complex solvent, electrical ions NaCL were added with 0.15 M [15,30e32]. To model the maximum possible atomic bond energy (covalent and non-covalent) flexibility of molecular docked complex, hybrid algorithm "Steepest Descent and the limited-memory Broyden-Fletcher-Goldfarb-Shanno" is used. To create the initial minimized condition model of dynamic particles, pressure, and temperature of the molecular docked complex, "Berendsen thermostat and barostat NPT ensemble" method is used for 5 ns.
To assert constant temperature at 300 K and pressure of 1 atm throughout the molecular dynamics' simulation process, "Nose-eHoover thermostat algorithm [33]" and "Martyna-Tobias-Klein Barostat algorithm [34]" is used respectively. To affirm coulombic long-range atomic interactions in the molecular docked complex throughout the molecular dynamic simulations process "Smooth Particle Mesh Ewald" method is used. To maintain the accuracy and tolerance of long-range interactions with the smaller value of 1e-9 is set for accurate computational implemented by SHAKE algorithm. The final production run was carried out for 100 ns, and the trajectory sampling was done at an interval of 1.0 ps [15,30,31]. Subsequently, in the MD simulations study, the binding free energy (DG bind ) for all simulated protein-ligand complexes were estimated by Desmond and Prime-molecular mechanics/generalized born surface area (MM-GBSA) method using thermal_mmgbsa.py script embedded in Schr€ odinger suite.

Prime MM-GBSA
The MM-GBSA is a suite of protocols that combines OPLS-2005 molecular mechanics energies (EMM), a Surface Generalized Born (SGB) solvation model for polar solvation (GSGB), and a nonpolar solvation term (GNP), composed of the nonpolar solvent accessible surface area and vdW interactions. The binding energy calculations are much accurate than the docking score [35]. The binding energy is calculated by the following equations [36,37].
Where E complex , E protein , and E ligand are the minimized energies of the receptorelead complex, receptor, and leads respectively DG solv ¼ G solv ðcomplexÞ À G solv ðproteinÞ À G solv ðligandÞ (3) Where G solv(complex) , G solv(protein) , and G solv(ligand) are the solvation free energies of the complex, protein, and inhibitor, respectively Where G SA (complex) , G SA (protein) , and G SA (ligand) are the surface area energies for the complex, protein and inhibitor, respectively.
The simulations were carried out using the GBSA continuum model in Prime [38]. The Prime uses a Surface Generalized Born (SGB) model employing a Gaussian surface instead of a vdW surface for better representation of the solvent-accessible surface area [35,39]. The binding free energy can be computed for all the trajectories or a set of frames within generated frames from MD simulations. Considering all frames, perfect (or imperfect) poses might have a crucial contribution in the binding free energy calculation, hence, each and every frame of the entire MD simulated trajectories were used, that is highly rigorous and more accurate approach to validate the docking and dynamics approaches in the study for calculating the binding free energy. Therefore, in the present study, all MD trajectory frames from start to end (~1000 frames) were selected to estimate the DG bind for proposed and native MD simulated complex.

Validation of docking protocol
Preferably, the distinguishing proof of the ideal docking and scoring blend will diminish the number of false positives and false negatives while ensuring optimal hit rates. It has accounted for enormous strategies for validating docking approach and scoring functions. Utilized strategy through pose selection whereby docking programs are exploited to re-dock into the target's active/ binding site with a known conformation and orientation, ordinarily from known coecrystal structure. The approaches that can restore poses under a pre-selected RMSD esteem from the known compliance/orientation (1.5 or 2 Å relying upon ligand size) are considered having performed effectively (Fig. 1). Pose selection is then trailed by scoring and positioning to contemplate which of the accessible scoring capacities most precisely positions the stances regarding their RMSD esteems. In the present study, the different docking protocols of Discovery studio and AutoDock protocols are validated through, re-docked with known coecrystal structure of 6LU7 with N3 peptide (Fig. 1). The re-docked results of RMSD revealed that, the AutoDock vina determined with 1.25 _ A and Discovery studio has 1.02 _ A. Also, 2D atomic coordination of the redocked complexes was supplanted experimentally derived structure. Thus, these docking protocols was viewed good enough for replicating the docking results similar to coecrystal structure and consequently can be applied for further molecular docking analysis.

Molecular interactions of M pro with AYUSH-64
The SARS-CoV-2, M pro plays a significant role in RNA translation and is essential for viral replication. The three-dimensional structure of M pro (6LU7) is a monomer structure, consisting three major structural domains: domain I (residues 8e101), domain II (residues 102e184), which contribute an antiparallel b-barrel structure, and domain III (residues 201e303), which contains a five-fold anti-parallel a-helix cluster and is attached to long loop (residues 185 to 200) of domain II [6]. The amino acid of cysteineeHistidine dyadplays significant role in the maturation cleavage, which is located in the cleft between domains I and II the precursor M pro [6]. Moreover, M pro was non-homologous to human and it is an ideal anti-viral drug target for SARS-CoV-2 [15]. The coecrystal ligand of N3 peptide is located in the cleft region between domain I and II. The SARS-CoV-2 M pro shares highly (~99%) similarity and identity (100%) with binding site residues of SARS-CoV M pro [15].
In the present study, the selected 36 compounds of AYUSH-64 were substituted (shows a similar inhibitor-binding mode) within the cleft between domain I and II, found to be having good docking score than the native substrate of N3 peptide (Supplementary Table 1 and Fig. 1A). ADME/T properties prediction results are illustrated in Supplementary Table 2 for drug-likenes. Two docking strategies, Akuammicine N-Oxide compound has best docking score, binding affinity and good ADME/T properties than the other selected 35 AYUSH-64 compounds and reference substrate of N3 peptide (Supplementary Table 3 and Fig. 2). The side chain residues, His41 (2), Met49 (2), Tyr54, Asn 142 (2), Gly 143, Ser144, Cys145 (2), His163, Met165, and His172 formed 14 H-bonds with Akuammicine N-Oxide within 4 Å region of M pro (Fig. 3). Along the side, Cys44, Pro52, Phe140, Leu141, Asn142, Arg188, Asp187, Gln189 were formed vdW interactions to the Akuammicine N-Oxide within 4 Å region in both docking strategies (Supplementary Table 2 and Fig. 3B). To summarize, Akuammicine N-Oxide compound reveal a similar inhibitory mechanism to occupy the substrate-binding pocket and block the enzyme activity of SARS-CoV-2 M pro . Further, the docked complex of M pro e Akuammicine N-Oxide and native M pro e N3 peptide complex was subjected to perform molecular dynamics simulations to check the stability at 100ns employing Desmond 2019-4.

Molecular dynamics simulations
For understanding the conformational deviations/fluctuations and binding free energy of docked and native complexes, 100 ns period of molecular dynamics simulations was performed for each trajectory. The M pro e Akuammicine N-Oxide and native N3 peptide complex depicted the total energy average of À9.2271.69 and À9.5674.02 kcal/mol, potential energy of À114792.08 and À121645.95 kcal/mol, was relatively stable with the average temperature 298.67 K, pressure -0.822 bar, and volumẽ 366704.2 Å 3 throughout the simulations run (Fig. 4).  The average RMSD value observed for M pro Ca atoms is 1.4 _ A and for proposed Akuammicine N-Oxideis 1.17 _ A, which was found stable throughout the simulation period (Fig. 5A). In the native complex of M pro e N3 pepetide, average RMSD values 9.5 _ A and 2.05 _ A were observed for the Ca atoms and maximum RMSD of M pro Ca atoms is 14.12 _ A, that constitutessignificantly not stable during MD simulations (Fig. 5B). To measure the atomic fluctuations in M pro residues, RMSF values were analysed for each residue Ca atoms bound with side-chain of docked and native complex of M pro of SARS-CoV-2. The average RMSF observed for each Ca atoms: 1.4 _ A and side chain: 2.5 _ A in the overall simulation run docked with Akuammicine N-Oxide (green lines) (Fig. 6A). In the native N3 peptide (green lines) with M pro complex observed in the 1000 frames of MD simulations where the results revealed with average RMSF values for Ca and side-chain have 3.77 _ A and 4.52 _ A respectively (Fig. 6B). Notably, post dynamics simulation of the M pro with Akuammicine N-Oxide and native N3 peptide, formed significant H-bonds, hydrophobic and water associated bridges correlating with coecrystal structure (Fig. 7). The Akuammicine N-Oxide formed seven H-bonds in M pro with Ser46 (~7%), Asn142 (~7%), His164 (~27%), Glu166 (~3%), Val186 (~30%), Arg188 (~4%) and Gln189 (~22%); three hydrophobic interactions with Leu27 (~30%), Met49 (~31%) and Cys145 (~39%); and more than 10% of water associated bridges with Asn142 (~18%), Glu166 (~39%) and Gln189 (~12%) within 4 _ A region of binding site (Fig. 7A). The coecrystal N3 peptide molecular interaction was substituted with the above furnished residue interactions of Akuammicine N-Oxide with M pro of SARS-COV-2.
From the comparative MM-GBSA plots between docked M pro-eAkuammicine N-Oxide and native M pro eN3 peptide complexes, in total of 1000 frames obtained from 100 ns MD simulations were considered for binding free energy calculations (DG bind ) ( Supplementary Fig. 1). The docked complex illustrated the stable average binding free energy of À44.57 kcal/mol ( Supplementary  Fig. 1A) than fluctuated native complex with average binding free energy of À38.47 kcal/mol ( Supplementary Fig. 1B). The stable DG bind values obtained from docked complex is self explanatory that, the proposed M pro eAkuammicine N-Oxide exhibited significantly stable and higher binding free energy in comparison with the native complex during 100 ns of MD simulations. While M pro of secondary structure elements (SSE) contributing to the overall protein stability was examined and noticed that, the protein was maintained an average of~42% SSE, mostly composed of~18.08% helices and~24.14% of strand during 100ns simulation period. Remarkably, the Akuammicine N-Oxide also showed the substantial ligand properties of Radius of Gyration (Rg:~4.0), Molecular Surface Area (MolSA:~363.37), Solvent Accessible Surface Area (SASA:~209.8) and Polar Surface Area (PSA:~138.47) ( Supplementary Fig. 2). Significantly, one intra-molecular interaction was observed during the simulation time of 100 ns period. Moreover, every rotatable bond of the Akuammicine N-Oxide formed H-bonds with M pro was stable throughout the simulation trajectory ( Supplementary Fig. 3). To summarize from all the observed findings of molecular docking and dynamics simulations, M pro e Akuammicine N-Oxide complex was highly stable and shown significant molecular interactions when compared with native M pro e N3 pepetide complex of SARS-COV-2.

Discussion
M pro is one of the best-characterized pan drug target for the coronavirus's family. Along with M pro an essential enzyme for the endurance of virus, processing the polyproteins which are   Fig. 2). The top scored docking pose compound, Akuammicine N-Oxide formed fourteen H-bond with M pro side chain residues along number bonds mentioned within brackets, His41 (2), Met49 (2), Tyr54, Asn142 (2), Gly143, Ser144, Cys145 (2), His163, Met165, and His172 of M pro significant binding site residues (Fig. 3). The Akuammicine N-Oxide formed vdW interactions with the binding site residues of Cys44, Pro52, Phe140, Leu141, Asn142, Arg188, Asp187, and Gln189 within 4 Å region in both docking strategies (Supplementary Table 2 and Fig. 3B). Thus, the molecular docking approaches utile to identify the novel hits/ leads against several pathogenesis. Docking aims to reproduce a conformational selection specific to target conformations that are recognized by the hit/lead to form molecular-interactions in the form of thermodynamical favored complex. The MD simulations were successfully used to achieve the major objective of substantially stabilize the M pro quaternary structure which involves an increase of the inter-molecular interactions by strengthening the ionic contacts and structural flexibility of the M pro binding pocket to allow flexible exploration of its conformational space, thus extracting representative frame from docked complexes.
To investigate the top scored docking pose complex (M pro e Akuammicine N-Oxide) binding conformational stability with SARS-CoV-2 M pro , we performed 100 ns MD simulations using Desmond. The MD simulations results revealed, Akuammicine N-Oxide formed seven stable H-bonds with Ser46, Asn142, His164, Glu166, Val186, Arg188 and Gln189; three hydrophobic interactions with Leu27, Met49 and Cys145; and water associated bridges with Asn142, Glu166 and Gln189 within 4 _ A region of binding site (Fig. 7A). The coecrystal N3 peptide molecular interaction was substituted with Akuammicine N-Oxide in M pro of SARS-COV-2. Furthermore, advanced MD simulations of MM-GBSA analysis (total frames obtained from 100 ns run) also confirmed the proposed M pro eAkuammicine N-Oxide complex has consistent binding free energy (DG bind ) score when compared to native complex. This study from sequence levels to advanced structural dynamics confirmed, Akuammicine N-Oxide of AYUSH-64 have potential inhibitor against SARSR-CoV-2. Superimposition of experimentally derived X-ray crystal structure of the M pro of MERS-CoV, SARS-CoV, and SARS-CoV-2, shares structural similarity and high conservativeness of substrate binding site. These observances are enlightened; Akuammicine N-Oxide was a potential pan-coronaviral inhibitor against corona family. Although these results were obtained from in silico approaches, the formation of these interactions provides insight to design and repurposing of novel drugs that might enlighten scientific community and researchers for further immune-modulatory investigations in human. Further, it is pertinent to note that AYUSH-64 is a compound herbal preparation, which is already tested to some extent for its safety and efficacy for treating malaria, joint disorders and influenza like illness. The currents in silico study provide a valid basis for its repurposing in the management of COVID-19.

Conclusion
This study revealed that, several compounds from an Ayurvedic drug are acting as potent inhibitors of SARS-CoV-2 M pro enzyme. Assessing the individual compounds through two different molecular docking strategies and extensive molecular dynamics simulations, explains that the Akuammicine N-Oxide is tightly bound (H-bond) with significant residues Cys145 and His164 of M pro , also found stable during pre-and post-dynamics simulations. The results indicate that AYUSH-64, an approved and safe drug for managing joint pains, fever, influenza like illnesses make it a good candidate for repurposing against COVID-19. Therefore, AYUSH-64 might be validated through further experimental, clinical studies that entailed to confirm the anti-protease properties against COVID-19.

Source(s) of Funding
None.

Conflicts of Interest
None.