key: cord-0957668-41xb34e3
authors: Toor, Himanshu G.; Banerjee, Devjani I.; Lipsa Rath, Soumya; Darji, Siddhi A.
title: Computational drug re-purposing targeting the spike glycoprotein of SARS-CoV-2 as an effective strategy to neutralize COVID-19
date: 2020-11-06
journal: Eur J Pharmacol
DOI: 10.1016/j.ejphar.2020.173720
sha: de7981aafdf94a4584baa66b5497fc8174da475d
doc_id: 957668
cord_uid: 41xb34e3

COVID-19 has intensified into a global pandemic with over a million deaths worldwide. Experimental research analyses have been implemented and executed with the sole rationale to counteract SARS-CoV-2, which has initiated potent therapeutic strategy development in coherence with computational biology validation focusing on the characterized viral drug targets signified by proteomic and genomic data. Spike glycoprotein is one of such potential drug target that promotes viral attachment to the host cellular membrane by binding to its receptor ACE-2 via its Receptor-Binding Domain (RBD). Multiple Sequence alignment and relative phylogenetic analysis revealed significant sequential disparities of SARS-CoV-2 as compared to previously encountered SARS-CoV and MERS-CoV strains. We implemented a drug re-purposing approach wherein the inhibitory efficacy of a cluster of thirty known drug candidates comprising of antivirals, antibiotics and phytochemicals (selection contingent on their present developmental status in underway clinical trials) was elucidated by subjecting them to molecular docking analyses against the spike protein RBD model (developed using homology modelling and validated using SAVES server 5.0) and the composite trimeric structures of spike glycoprotein of SARS-CoV-2. Our results indicated that Camostat, Favipiravir, Tenofovir, Raltegravir and Stavudine showed significant interactions with spike RBD of SARS-CoV-2. Proficient bioavailability coupled with no predicted in silico toxicity rendered them as prospective alternatives for designing and development of novel combinatorial therapy formulations for improving existing treatment regimes to combat COVID-19.

cell. It is a homotrimeric protein (each chain containing 1273 amino acids) which 134 recognizes and binds to its receptor Angiotensin Converting Enzyme-2 (ACE-2) with a 135 greater affinity attributed to the mutations at L455, F486, Q493, S494, N501 and Y505 136 residues (Yan et al., 2020) . Microbiological digital resources data insinuate a 137 comprehensive schematic of the spike glycoprotein ( Fig. 1-2) . (Table 1) . 170 The crystallization propensity of the SARS-CoV-2 spike glycoprotein sequence (Table 2) . Overall, 325 the spike glycoprotein RBD model was deemed non-crystallizable in its native form.

The MobiDB web tool indicated 0% disorderness in the spike glycoprotein sequence (Table S1 ; Fig. S1 in supplementary data). 

The spike glycoprotein sequence from UniProtKB was submitted to MODELLER 371 v9.24. It generated two models whose characteristics are listed in table 3. GA341

(range: 0-1, value near 1 being preferred), z-DOPE (lowest score preferred), MPQS 373 (ModPipe Quality Score; highest value preferred) and sequence identity (highest value 374 preferred) scores were employed for model quality assessment.

The results indicated that model 2 was the most reliable one as compared to model 1.

The MPQS score of model 1 was also predicted to be unreliable as compared to that of (Table 4) .

Sequential analysis of the spike RBD model and the template (6YLA chain E) revealed 419 that our RBD model lacked the first three residues (GLU327, THR328 and GLY329) 420 incorporated by 6YLA chain E. In addition, the residues 1, 2, 3, 4…. in the spike RBD (Table 5) . 459 Earlier studies expounded that the spike RBD comprised of significant loops (a-f) 460 which were then re-categorized into loop 1, 2 and 3 sub-segments. It was observed that (Table 6) .

The criteria for selection of the potential drug candidates as potential inhibitors of interactions involving loop residues, etc. Based on the above factors, we selected ten 481 drugs out of the thirty studied for further analysis whose detailed significant molecular 482 interactions are entailed (Table 7) . All selected drug candidates are illustrated in Fig. 9 .

Most significant interactions were observed in case of Stavudine, Doxycycline and 484 Favipiravir (Fig. 10a, 10b and 10c ). Our results also indicated that Tenofovir, Eugenol, (Table S2 ; Fig. 6 supplementary data).

In Fig. 11 we can see the residues at the interface of RBD of the spike protein and the 495 ACE-2 receptor as listed earlier (Table 5) . To understand whether the binding of 496 inhibitor causes any structural changes in the RBD-ACE2 complex, we superimposed 497 the structure of the RBD with and without the inhibitors. Favipiravir was found to 498 interact directly with residues Y505, R403 and L441 of the RBD, where R403 is 499 considered to be a possible mutation (Fig. 12a, Fig. 6, Table 7) . Surprisingly, upon 500 superimposition of RBD and the RBD-Favipiravir complex we found deviation in side 501 chain conformation of R403 (Fig. 12a) . Apart from that few changes were seen in V401 . 1-2) .

The present study focused on the drug re-purposing based inhibition of SARS-CoV-2 574 attachment wherein the spike glycoprotein of SARS-CoV-2 undergoes conformational 575 reorganization to distinguish its receptor ACE-2 and initiates a cascade of molecular 576 progressions resulting in the integration with the cell membrane and liberation of its 577 genome inside the host cell for replication. 578 We developed our validated model of the spike glycoprotein RBD based on the 579 UniProtKB sequence P0DTC2 utilizing 6YLA chain E as the template (Fig. 5a) . 

We implemented a computational biology-based approach for drug re-purposing to 588 screen drug candidates as potential therapeutic inhibitors via molecular docking 589 analyses impacting the spike glycoprotein mechanism of action (Tables 6-7) . Table S6 ). (Tables S6-S7) . Table S7 ).

Raltegravir is an anti-retroviral drug belonging to the class of integrase inhibitors, is different than that of direct interaction with spike protein (Table 6 ; Fig. 12b ).

Chloroquine inhibits the viral attachment to its receptor along with the subsequent J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f J o u r n a l P r e -p r o o f formed a single hydrogen bond with residues ASN259, ASN136 and GLU132 of ACE-2 (bond lengths: 2.9, 2.2 and 2.7 respectively).

TYR449 depicted 2 hydrogen bonds with residues ASN259 (bond length 2.8) and THR258 (bond length 2.7) respectively. The yellow- The protein-protein complex of RBD of Spike protein (magenta) and the ACE-2 receptor (yellow) are shown as cartoon. A large number of polar residues from the Spike protein interact with the ACE2 receptor by hydrogen bonding and hydrophobic interactions ( Table 5 ).

The primary residues from the Spike protein which interact with the ACE2 are shown as sticks and colored by CPK. 

The Human Coronavirus Disease 722 COVID-19: Its Origin, Characteristics, and Insights into Potential Drugs and Its 723

In silico prediction and docking of 725 tertiary structure of LuxI, an inducer synthase of Vibrio fischeri

The 728 proximal origin of SARS-CoV-2

Sequence-specific determination of protein and peptide 731 concentrations by absorbance at 205 nm

UniProt: the Universal Protein knowledgebase

DeepLoc: prediction of protein subcellular localization using deep learning

Synthesis, in vitro antifungal 741 evaluation and in silico study of 3-azolyl-4-chromanone phenylhydrazones. DARU 742

Sequence and structure-based prediction of 749 eukaryotic protein phosphorylation sites

A Review of Raltegravir and its Use in HIV-1 Infection

The FDA-754 approved drug Ivermectin inhibits the replication of SARS-CoV-2 in vitro

Boosting the arsenal against COVID-19 through computational drug 757 repurposing

The spike glycoprotein of the new coronavirus 2019-nCoV contains a furin-like 761 cleavage site absent in CoV of the same clade

iLOGP: A Simple, Robust, and Efficient 764 Description of n-Octanol/Water Partition Coefficient for Drug Design Using the 765

SwissADME: a free web tool to evaluate 768 pharmacokinetics, druglikeness and medicinal chemistry friendliness of small 769 molecules

Intracellular metabolism and in vitro activity of Tenofovir against Hepatitis B 777 virus

New insights on the antiviral 780 effects of chloroquine against coronavirus: what to expect for COVID-19?

Coronavirus Diseases (COVID-19) current status and future 784 perspectives: A narrative review

JPred4: A protein secondary 787 structure prediction server

Favipiravir: Pharmacokinetics and concerns about clinical trials 790 for 2019-nCoV Infection

Advance of promising targets and agents against 793 COVID-19 in China

Molecular Docking 798 and Structure-Based Drug Design Strategies

Modeling of loops in protein structures

Protein Identification and Analysis Tools on the ExPASy 804

The Proteomics Protocols Handbook

The species severe acute respiratory syndrome-related coronavirus: 814 classifying 2019-nCoV and naming it SARS-CoV-2

Prediction of glycosylation across the human proteome and 817 the correlation to protein function

A Review of 819 COVID-19

Safety and efficacy review of Doxycycline

iGEMDOCK: a graphical environment of 827 enhancing GEMDOCK using pharmacological interactions and post-screening 828 analysis

Practical considerations for developing nucleoside 831 reverse transcriptase inhibitors

A candidate multi-epitope vaccine against SARS-CoV-2

Comparative In Silico studies on phytochemicals of ocimum as 837 natural inhibitors of Ebola VP-35 protein

Study of combining virtual 841 screening and antiviral treatments of the Sars-CoV-2 (Covid-19)

NetAcet: prediction of N-terminal 844 acetylation sites

The neighbor-joining method: A new method for 953 reconstructing phylogenetic trees

Comparative Protein Modelling by Satisfaction of Spatial 956

A systematic and reverse vaccinology approach 958 to design novel subunit vaccines against Dengue virus type-1 (DENV-1) and 959 human Papillomavirus-16 (HPV-16)

Database resources of the 965 National Center for Biotechnology Information

Ligand docking and binding site analysis with 968 PyMOL and Autodock/Vina

Drug repurposing strategies for COVID-19

In silico studies on therapeutic agents for COVID-973 19: Drug repurposing approach

Coronavirus pathogenesis and the emerging 1004 pathogen severe acute respiratory syndrome coronavirus. Microbiol

HawkDock: A 1007 web server to predict and analyze the protein-protein complex based on 1008 computational docking and MM/GBSA

Analysis of therapeutic targets for 1012

SARS-CoV-2 and discovery of potential drugs by computational methods

The SARS-CoV-2 outbreak: What we know

Structural basis for the 1018 recognition of SARS-CoV-2 by full-length human ACE2

Protein-ligand binding site recognition using 1021 complementary binding-specific substructure comparison and sequence profile 1022 alignment

BioLiP: a semi-manually curated database for 1025 biologically relevant ligand-protein interactions

A novel coronavirus from patients with pneumonia in china

Evolutionary Divergence and Convergence, in 1035 Proteins

A:ASN259:CA -B:TYR449:OH 2.46628* Hydrogen Bond Carbon Hydrogen Bond A:ASN259:CA H-Donor B:TYR449:OH H-Acceptor B:LYS444:CE -A:ASP274:OD1 3.0135* Hydrogen Bond Carbon Hydrogen Bond B:LYS444:CE H-Donor A:ASP274:OD1 H-Acceptor B:TYR449:CA -A:SER262:OG 3.60435* Hydrogen Bond Carbon Hydrogen Bond B:TYR449:CA H-Donor A:SER262:OG H-Acceptor A:LYS229:NZ -B

Electrostatic Pi-Cation A:LYS229:NZ Positive B:PHE490 Pi-Orbitals A:GLU132:OE1 -B:TYR505 4.89665

Electrostatic Pi-Anion A:GLU132:OE1 Negative B:TYR505 Pi-Orbitals B:TYR489 -A:PRO235 5.23117

Hydrophobic Pi-Alkyl B:TYR489 Pi-Orbitals A:PRO235 Alkyl

 Table 5 ACE2-spike RBD model molecular interactions analysis.

Bond length Bond category Bond type Donor Acceptor