In-silico analysis of potential interaction of drugs and the SARS-CoV-2 spike protein


 The world is suffering its deadly pandemic in decades from Corona virus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Till the end of April, 2020, the death toll rose to more than 2 lacs worldwide (224 301, on May 1st), affecting 215 countries (source: WHO). Since the invention of a new drug or vaccine takes time, drug repurposing is one of the avenues. Keeping that in mind, we have tried to shed light on some of the drugs, using molecular docking/modeling studies, which could be of immense importance in the current scenario. Using bioinformatic approaches, we have tried to work on the viral spike (S) protein, majorly involved in pathogenicity through its receptor binding event and fusion to the host cells. More emphasis was given on the spike proteins (6vsb) and (6lxt). We predicted the drugs that target to the S proteins by in-silico docking analysis. We determined the functional part of the spike of SARS-CoV-2 and the drugs were predicted by analysis of the functional part using the PDBsum/DrugPort. Molecular Docking unveiled the binding of those drugs to the target spike. This in-silico study actually envisions the efficacy of the predicted drugs against the spikes of SARS-CoV-2 and the binding of the drugs might propose the combinatorial effect of the drugs to neutralize the spike effect in the human body.


Introduction
The outbreak of COVID-19 was rst noticed in China in December, 2019 and later on it has infected throughout the world, resulting in more than 200,000 deaths up to April end, 2020. The virus causing COVID-19 belongs to beta coronavirus of Coronaviridae family and has similarity with early Severe Acute Respiratory Syndrome (SARS) and Middle East Respiratory Syndrome (MERS).The new strain is called Severe Acute Respiratory Syndrome-Coronavirus-2 (SARS-CoV-2) 1 The symptoms of COIVD-19 includes different types of illness like fever, cough, respiratory obstruction, enteric, neurologic problems, skin disease, lack of olfaction sense etc. 2 . The entry of Coronavirus in the human body is mediated through the viral spike proteins in an amazing way of receptor recognition and binding of its S1 subunit to the host cells and fusion by its S2 subunit there 3,4 . It has been identi ed that SARS-CoV-2 transfused its infection via binding with a metallopeptidase, angiotensin converting enzyme-2 (ACE-2) of the host cell 5,6 . Lopinavir, chloroquine, chlorpromazine have been used in COVID-19 7 and success has been found in many cases. Here we approach via in silico targeting of the spike protein molecules deposited to Protein Data Bank. The proteins studied here, are possibly good drug targets as suggested by a few studies. Hence, we proceeded with the structures from PDBsum portal (http://www.ebi.ac.uk/pdbsum) 8 to focus on the drugs they bind, using various bioinformatic tools described in the methods and results. Our experimental analysis reveals many drugs, having a potential role for the treatment of COVID-19, explored from the molecular recognition, based on the interaction between macro-molecules and small molecules. So, the interaction between a protein molecule and a ligand molecule likewise protein-drug interaction will explain the various signi cant biological signaling cascade. And the physicochemical mechanism i.e. binding kinetics (Binding A nity) and thermodynamics (Global Energy, Free Energy) are the rational for the understanding of drug-protein interaction study. So, the feasibility of the drug to the target protein could be theoretically/computationally predicted 9 .

Methods
Structure retrieved: To study our investigation on COVID-19, structure of spike proteins, pre-fusion spike glycoprotein containing receptor binding site (6VSB) 10 and the post-fusion structure (6LXT) 11 were retrieved from Protein Data Bank (PDB)https://www.rcsb.org/. We retrieved PDB structure of 6VSB and FASTA sequence of both structures. The structures (3D structure) of drugs were retrieved from PubChem. Those 3D structures were converted to a PDB le by SMILES Translator (NCI/CADD group).
Visualization of structures: All the retrieved structures and molecular docking analysis were visualized by PyMOL (Schrodinger/pymol-open-source) and UCSF Chimera 12 .
Cluster analysis: To predict the functional part of the 6VSB, chain-C (6VSB is a trimer, we selected chain-c) the cluster was analyzed using the InterProSurf, which predicted the interacting sites on protein surfaces [13][14][15] .
Details of the molecular analysis by PROCHECK (PDBsum and DrugPort): To study more on the functional part, the sequence of the protein (6VSB) of that part and the similar sequence part of 6LXT with 6VSB was analyzed by PDBsum and followed by sequence search on DrugPort, which demonstrated the drug molecules and their target proteins by using PROCHECK (https://www.ebi.ac.uk/thorntonsrv/software/PROCHECK/),PDBsum, (http://www.ebi.ac.uk/pdbsum) and DrugPort server 8,16,17 .
Pocket analysis: Pockets on the surface of the protein was analyzed by CASTp 3.0 (Computed Atlas Surface Topography of the Protein) 18 .
Protein-ligand, Protein-Protein interaction: Entropy-Enthalpy calculation is an important way to analysis binding and here GalaxyWeb server was used for the Binding A nity calculation and for the analysis of protein-ligand molecular docking 19 . To determine the global energy of the molecular docking between the protein and ligand, PatchDock (based on shape complementarity principle 20

Results And Discussion
Cluster analysis of the SARS-CoV-2 protein We have worked on a SARS-CoV-2 spike protein (6VSB) from protein data bank 9 . The functional sites of 6VSB, chain-c were predicted by cluster analysis by using InterProSurf and here the amino acids were distributed based on their E1 vector and hydrophobicity, hydrophilicity and polar charged residues in the clusters [13][14][15] .

Identi cation of pockets by the analysis of surface topography of protein
Pockets on the surface of the proteins are the mouth-opening connection of the interior to the external bulk solution. Those sites are signi cant for the proteins-ligand interaction or might be the target sites for the therapeutic purpose. Here, we investigated the pockets of SARS-CoV-2 spike protein (6VSB, chain-c) protein by surface topography 18 .

Drug interaction to the target sites of the protein
We investigated the drug-target on the spike of SARS-CoV-2 protein by using the PROCHECK, PDBsum and DrugPort 8,16,17 by the protein sequence search which unveiled the drug molecules and their target proteins. We found the drugs targeted to the protein and performed molecular docking analysis by using the docking server [19][20][21][22][23][24][25][26][27] of those selected drugs to the spike protein of SARS-CoV-2. In each case, we demonstrated the Binding A nity by GalaxyWeb 12 ; Global energy by using PatchDock 20,21 and followed by Desmopressin is a synthetic analogue of Arginine Vasopressin. It helps in the water resorption by binding to the V2 receptors. Desmopressin can induce the eNOS activity. It might inhibit the viral or endotoxin induced septic shock 28,29 . So, Desmopressin might be one of the supportive drugs of choice in COVID-19, even if while the bleeding complicacy.

Docking with Telaprevir (PubChem CID: 3010818):
Telaprevir is an anti-viral used to treat hepatitis C virus (HCV) 30 . It is the inhibitor of HCV protease 31  Broad spectrum antibiotic Erythromycin binds to the bacterial ribosome (50S subunit) and inhibits bacterial protein synthesis by interfering with translocation of amino acids to the protein synthesismachinery during translation. It has been shown to impair the activity of S. pneumoniae infection in the respiratory tract 34 . In Fig-6, binding of Erythromycin with COVID was shown with Binding A nity: -17.877; Global Energy: -45.45 and Binding Energy: -5.80 (kcal/mol).
We also found the interaction of Gentamicin with S1 subunit ( gure not shown) with the Binding A nity: It is an anti-malarial, anti-viral and anti-neoplastic agent. Artesunate shows its activity through the breakage of the endoperoxide bridge in the presence of heme ion and generates ROS and free radicals, that helps to damage DNA and kill the malaria parasite 36 and is considered to be the best drug against many complicated malaria patients 37 . Artesunate is also effective against herpes virus, hepatitis virus, cytomegalovirus 38 . Fig-8  Rifabutin, an antibiotic is used in HIV infected patients for the controlling of Mycobacterium 45 .Here, the binding of Rifabutin to spike protein is shown in Fig-11(Binding A nity: -17.634) in (Fig-14); Global Energy: -54.04 and Binding Energy: -7.6 (kcal/mol).
Docking with Bedaquiline (PubChem CID: 5388906): Bedaquiline is also an antimycobacterial drug. It is very effective in the drug resistance pulmonary tuberculosis in HIV 46 , however, safety concerns remain associated with the usage of it 47 . Bedaquiline binds (Binding A nity: -17.420) with spike is depicted in Fig-12 FMN is a cofactor of many oxido-reductase reaction. It can also protect the lungs from toxic compounds 49 . The Binding A nity of FMN with spike protein is: -16.080 (Fig-14); Global Energy is -56.27 and Binding Energy is -5.8 (kcal/mol).
We have also analyzed the docking of NAD (PubChem CID: 5893), NAD binds spike with Binding A nity: -15.973; Global Energy: -22.43 and Binding Energy: -6.4 (kcal/mol). NAD is an electron carrier and redox reactor. It is also involved in the prevention of HIV and pneumonia 50 .

Docking with Aprotinin (3GYM_2|Chains I, J|Pancreatic trypsin inhibitor|Bos Taurus (9913)
Aprotinin is known as BPTI i.e. bovine pancreatic trypsin inhibitor. It is a protein based drug used for the treatment of reducing blood loss during CABG or pericardititis 51,52 . Cleavage of spike protein by host protease is critical for its infection status. So, aprotinin could be therapeutically used to prevent spike protein infection 52,53 . Here we investigated the interaction between spike protein and aprotinin. From docking analysis, interaction without using template (Docking score: -267.79) could be observed between spike protein and aprotinin.
From the above stated drug-interaction studies, it could be predicted the possibility of e cacy of those drugs against COVID-19 patients; not only the above drugs and neutraceuticals but also few other drugs, like Sirolimus binds to the spike protein with a nity (table-1); it is an anti-neoplastic, antibiotic, immunosupressive drug and it could be the drug against spike but the drug has a potential pulmonary toxicity. In our analysis, we found some anticonvulsant medicines; Cannabidiol (table-1) is one of them, which can reduce of anxiety disorder. Sucralfate is used for the cure of damaged mucous has a potential Binding A nity (table-1) with spike protein. Sucralfate could be used a gel to protect from virus/bacteria. Ranolazine/Spiranolactone are the antihypertensive drugs both showed binding a nity (table-1) with spike proteins. Some anticancer antileukemic drugs have potential binding with spike of SARS-CoV-2; for example Vincristine is used for leukemia treatment and As 2 O 3 arsenic trioxide (ATO) was also found to be involved against spike of CoV; ATO with anti retroviral therapy (ART) is effective against the viral reservoir of HIV-1 in context of CD4+ T cells 54 . In case of tetraarsenic hexoxide (As 4 O 6 ) could be more effective in human papiloma virus associated cervical cancer 55 . It might be possible that ATO has the ability to induce apoptosis of viral protein by binding on the protein surface 56 (56). Though, use of arsenic could be potentially contraindictory, the dose of 0.15 mg/kg/day is well tolerated in the human body. Beside these, Por mer sodium, Benzyl penicillin, Tricalabendazole, Oseltamivir, Ribavirin and Tocilizumab antibody and other neutraceuticals like FAD, Ferrous Ascorbate, Glutathione, Zinc Chloride, Zinc acetate, Cu, Zn, Palmitic acid might be the crucial agent to combat against COVID-19 patients (shown in table-1).
It was found that that that older persons are more vulnerable to the COVID-19 and the persons who have diabetes, hypertension and cancer, are more susceptible to the complicacy of the disease and death. So, from the above analysis, Fostamatinib/Telaprevir/Hydroxychloroquine/Artesunate/Tenofovir could be used against spike protein and in combination with Cefpiramide/Erythromycin/Rifabutin/Bedaquiline as the protector of pulmonary injury due to the viral induced predisposing bacterial infection or Demopressin/Phylloquinone in case of renal dysfunction/bleeding disorder or Acarbose in hyperglycemia and Ranolazine/Spiranolactone in hypertension or FMN/FAD as neutraceutical and aprotinin as a protease inhibitor might be used to treat COVID-19 patients. It could be possible that various combinations of the those drugs would response better way.
Studies like these are important because, not all drugs work in the same e cacy in all the populations across the world. Disease association also varies in population speci c ways, many times. In case of COVID-19 too, data shows, that European or American population have been worstly affected in the world. (In case of infectious diseases, however, other factors like disease containment, health infrastructure etc. also play a crucial role, in addition to genetic makeup. Response to drugs also varied amongst them, eg. China did not nd Remdesivir very effective 57 , but U.S. Food and Drug Administration (FDA) authorized its emergency use, thinking it as potential drug to treat COVID-19 (https://www.fda.gov/media/137565). So, it is very important to keep eyes on many possibilities, where our study, could be an important one. It should also be mentioned here that potential drug interaction and contraindication should be followed up and it can be varied from patient to patient (personalized medicine), history of patients. It is a hypothetical insilico analysis by using the Bioinformatically found data which can be con rmed clinically by extended research work/clinical trials. Declarations Figure 1 Functional sites of chain-C of 6vsb protein: Fig-1A and 1B represented the distribution of amino acids based on their E1 vector and hydrophobicity, hydrophilicity and polar charged residues respectively. Red circles represent the functional sites; Fig-1C and 1D demonstrated the protein sequence of the red circle sites by entry into the PROCHECK, PDBsum server as described in method.                      This gure demonstrates the surface interaction between SARS-CoV-2 protein (6vsb) (yellow) and aprotinin (cyan) and the Docking score: -267.79