Cordycepin anti sarcovi 2

Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2 Cordycepin anti sarcovi 2

836 | wileyonlinelibrary.com/journal/cbdd 2020 John Wiley & Sons Ltd Chem Biol Drug Des 2021;97:836–853.

COVID-19 disease is caused by a positive-sense,

single-stranded RNA containing novel coronavirus, named severe

acute respiratory syndrome coronavirus 2 (SARS-CoV-2)

(previously provisionally known as 2019 novel coronavirus;

2019-nCoV) (Zu et  al.,  2020) The virus, which is now a pandemic (WHO, 2020), has infected at least 54.1 M people across the world, killing 1.31 and 34.8 M people have been recovered (till November, 15, 2020) The clinical symptoms associated with COVID-19 include high fever, mild cough, body aches, lack of smell and taste, self-limiting respiratory

R E S E A R C H A R T I C L E

Repurposing potential of FDA-approved and investigational

drugs for COVID-19 targeting SARS-CoV-2 spike and main

protease and validation by machine learning algorithm

Akalesh Kumar Verma1 | Rohit Aggarwal2

1 Cell and Biochemical Technology

Laboratory, Department of Zoology, Cotton

University, Guwahati, India

2 Cosmic Cordycep Farms, Badarpur Said

Tehsil Tigaon, Faridabad, Haryana, India

Correspondence

Akalesh K Verma, Cell and Biochemical

Technology Laboratory, Department of

Zoology, Cotton University, Guwahati

781001, India.

Email: akhilesh@cottonuniversity.ac.in

Abstract

The present study aimed to assess the repurposing potential of existing antiviral drug candidates (FDA-approved and investigational) against SARS-CoV-2 target proteins that facilitates viral entry and replication into the host body To evaluate molecu-lar affinities between antiviral drug candidates and SARS-CoV-2 associated target proteins such as spike protein (S) and main protease (Mpro), a molecular interac-tion simulainterac-tion was performed by docking software (MVD) and subsequently the applicability score was calculated by machine learning algorithm Furthermore, the STITCH algorithm was used to predict the pharmacology network involving multi-ple pathways of active drug candidate(s) Pharmacophore features of active drug(s) molecule was also determined to predict structure–activity relationship (SAR) The molecular interaction analysis showed that cordycepin has strong binding affinities with S protein (−180) and Mpro proteins (−205) which were relatively highest among other drug candidates used Interestingly, compounds with low IC50 showed high binding energy Furthermore, machine learning algorithm also revealed high applica-bility scores (0.42–0.47) of cordycepin It is worth mentioning that the pharmacology network depicted the involvement of cordycepin in different pathways associated with bacterial and viral diseases including tuberculosis, hepatitis B, influenza A, viral myocarditis, and herpes simplex infection The embedded pharmacophore features with cordycepin also suggested strong SAR Cordycepin's anti-SARS-CoV-2 activ-ity indicated 65% (E-gene) and 42% (N-gene) viral replication inhibition after 48h of treatment Since, cordycepin has both preclinical and clinical evidences on antiviral activity, in addition the present findings further validate and suggest repurposing potential of cordycepin against COVID-19

K E Y W O R D S

2019nCov, antiviral drugs, cordycepin, coronavirus, drug repurposing

tract illness to severe progressive pneumonia leading to

multi-organ failure leading to death (Shi et  al.,  2020; Liu

et al., 2020) The COVID-19 pandemic poses a big challenges

in the near future to global public health (Phelan et al., 2020),

appealing for the development of effective prophylactics and

therapies against the causative agent Due to lack of potent

antiviral drug(s) to treat the COVID-19 patients leads to a

clamoring to test existing antiviral drugs (alone or in

combi-nation) that are previously approved for the use of genetically

close human viruses, and the procedure specifically known as

drug repurposing (Nishimura & Hara, 2018) Drug

repurpos-ing or repositionrepurpos-ing of launched or even failed drugs against

new emerging viral diseases provides unique opportunities

in translational research It bears a substantially higher

prob-ability of success as compared to developing new

virus-spe-cific drugs and vaccines, in terms of cost, time, and clinical

availability (Ianevski et al., 2018)

SARS-CoV-2 genome comprised of 6 open reading

frames and code for several structural and non-structural

proteins (Walls et  al.,  2020) The structural protein

in-cludes spike protein (S), membrane protein (M), envelope

protein (E), and the nucleocapsid protein (Ncp) (Kahn &

McIntosh, 2005; Walls et al., 2020) that are tightly bound

to surface of the mature virion Spike protein (S) made

up of two distinct functional subunits (S1 and S2)

respon-sible for binding (S1 subunit) and fusion (S2 subunit) with

the host cell membranes (Kirchdoerfer et al., 2016; Walls

et  al.,  2020) The distal S1 subunit accommodates

recep-tor-binding domain and stabilized the prefusion state of the

membrane-anchored S2 subunit that contains the fusion

ma-chinery For all CoVs, S is further cleaved and processed

by host proteases, TMPRSS2 (Walls et al., 2020) at the S2′

site and in a recent study it has been reported that a

ser-ine protease inhibitor, which act on TMPRSS2 significantly

inhibit novel coronavirus entry (Wan et al., 2020) Thus, it

is evident that coronavirus entry into susceptible cells is a

complex and multistep process that requires the concerted

action of receptor-binding and proteolytic processing of the

spike protein to promote successful virus-cell fusion (Walls

et al., 2017)

Furthermore, the maturation of SARS-CoV-2 also

re-quired a series of highly complex proteolytic events

me-diated by main protease (CoV Mpro; also known as 3Cl

protease or 3CLpro) on the polyproteins to control viral gene

expression (Zhang et al., 2020) Mpro (~306 amino acid) is a

cysteine protease with a chymotrypsin-like two-domain fold

at the N terminus consists of three functional domains

(I-III) The structural analysis of Mpro revealed that two Mpro

molecules form an active homodimer A Cys-His catalytic

dyad is positioned in a cleft located between domains I

and II, and the Mpro N-terminal residues mostly 1 to 7 (or

N finger) are involved in the proteolytic activity, whereas,

the C-terminal domain III is reported to be required for di-merization (Xue et al., 2008; Zhang et al., 2020) Most mat-uration cleavage events within the precursor polyprotein are mediated by the CoV main protease to forestall the spread of disease by restraining the cleavage of the viral polyprotein (Xue et al., 2008)

Till now, no clinically proven vaccines or antiviral drug (s) are available for the prevention and treatment of COVID-19 pandemic Due to the gravity of the situation and worldwide rapid spread of SARS-CoV-2, an urgent and complemen-tary efforts are necessary to find new preventive methods The combination of α-interferon and the anti-HIV drugs Lopinavir/Ritonavir (Kaletra®) has been tested at a different levels of infection, but the curative effect is limited due to severe side effect in the host (Cao et al., 2020) A broad-spec-trum antiviral drug, remdesivir, (by Gilead Sciences, Inc.)

is also in the race of trial for the treatment of COVID-19, but lacking satisfactory data to prove its efficacy (Wang

et al., 2020) It is also worth mentioning, the Indian Council

of Medical Research (ICMR), under the Ministry of Health and Family Welfare (MHFW), has recommended the use of hydroxychloroquine (400 mg twice on day 1, then 400 mg once a week thereafter) as chemoprophylaxis for asymptom-atic healthcare workers directly involved in treating

COVID-19 patients with suspected or confirmed COVID-COVID-19, and for asymptomatic household contacts of confirmed cases (Rathi

et al., 2020)

Thus, the existing data confirmed that spike protein (S) and main protease (Mpro) in SARS-CoV-2 play vital roles during viral entry, genome replication, and self multiplica-tion in the host body (Huang et al., 2020; Wang et al., 2020; Wrapp et  al.,  2020; Zhang et  al.,  2020) Therefore, in the present study, these protein targets have been utilized during molecular interactions simulation with FDA-approved and in-vestigational drugs using different computational tools The present paper highlights the repurposing potential of known antiviral drug candidates against SARS-CoV-2

Highlights

• Cordycepin showed strong chemical interactions with SARS-CoV-2 RBD domain

• Cordycepin served as a pivot molecule against target proteins (S and Mpro)

• The repurposing potential of cordycepin was vali-dated by ECFP6 and Bayesian algorithm

• Pharmacology network also revealed the involve-ment of cordycepin in different pathways

TABLE 1

IC 50 (µM)

IC 50

PubChem AID/bioassay record/ References

Viral RNA polymerase inhibitor

Influenza C virus

Hepatitis C virus

Viral RNA polymerase inhibitor

Inosine monophosphate dehydrogenase inhibitor

Encephalitis virus

Antimalarial agent; chemo/radio sensitizer

SARS coronavirus

BCX4430 (Galidesivir)

RNA-dependent RNA polymerase inhibitor

Antimetabolite; Nucleoside analog

CVB3 and EV71

Rapamycin (Sirolimus)

IL-2-dependent T-cell proliferation inhibitor

HCV genotype 1b

Hepatitis C virus

Hepatitis C virus

SARS coronavirus

An irreversible inhibitor of ornithine decarboxylase

Trypanosoma brucei

Sl Nos.

IC 50 (µM)

IC 50

PubChem AID/bioassay record/ References

Chikungunya virus

Inhibitor of translation initiation by targeting the RNA helicase

Chikungunya virus

Non-competitive ß-adrenergic inhibitor

Abl; Src; Kit; EphR inhibitor

Transglycosylation and transpeptidation inhibitor

Peptidoglycan polymerization inhibitor

Protein translation inhibitor

Herpes simplex virus type 1

HIV1 subtype A

HT-29 viability

Abl; Kit; PDGFRB inhibitor

Synthetic serine protease inhibitor

MERS corona virus

Inhibits assembly of clathrin-coated pits

Sl Nos.

IC 50 (µM)

IC 50

PubChem AID/bioassay record/ References

N-methyl-D-aspartate glutamate receptor antagonist

Cyclo-oxygenase (COX) enzyme or prostaglandin G/H synthase inhibitor

SARS coronavirus

Bacterial cell wall synthesis inhibitor

Inhibits assembly of clathrin-coated pits

Mitogen activated protein-kinase inhibitor

Dengue virus type 2

Investigational anticancer and Phase II (Leukemia)

Polyadenylation inhibitor

Herpes simplex virus type-1

2 | MATERIAL AND METHODS

2.1 | Selection and acquisition of chemical

compounds

The information of broad-spectrum antiviral agents (BSAAs,

i.e., compounds targeting viruses belonging to two or more viral

families) was collected from a freely accessible database (https://

drugv irus.info/) (Andersen et  al.,  2020) The selected drug

molecules were belonged to investigational or FDA-approved

category against SARS-CoV-2, HCoV-229E, HCoV-OC43,

MERS-CoV, and SARS-CoV as well as other human diseases

Forty-five potential drug candidates as shown in Table 1 were

identified from the PubMed-NCBI literature with strong

antivi-ral activity against genetically close human viruses The present

status of all the drug candidates against SARS-CoV-2 target

proteins is also shown in Table 2 The 3D structures of all the

compounds were downloaded in SDF/Mol file format

embed-ded with 3D properties from the PubChem compound database

(http://pubch em.ncbi.nlm.nih.gov/), DrugBank (https://www

drugb ank.ca/) and ZINC database (https://zinc.docki ng.org/)

The molecular arrangement and geometry of all the compounds

were fully optimized using the semiempirical quantum

chem-istry method (PM3) (Gogoi et al., 2019) At last, the fully

op-timized 3D structure of all the drug molecules was exported

in Mol2 format and used for molecular interaction simulation

using different computational tools

2.2 | Selection of target protein

The SARS-CoV-2 associated target proteins, namely spike

protein receptor-binding domain (ID: 6VW1) (Shang

et al., 2020) and main protease (6LU7) (Jin et al., 2020), were

used in the present study The crystal structures of both the

target proteins were obtained from the RCSB Protein Data

Bank (http://www.rcsb.org/pdb/home/home.do) These

tar-get proteins have been reported for their pivotal roles during

SARS-CoV-2 infection, replication, survival, and

multiplica-tion in the host body (Liu et al., 2020b; Xi et al., 2020)

2.3 | Molecular interaction simulation

The intermolecular interactions between all the

FDA-approved and investigational drug candidates (Table 1) with

aforesaid target proteins were studied by using Molegro

Virtual Docker (MVD 2010.4.0 software for windows-7,

trial) (Kusumaningrum et  al.,  2014; Puspaningtyas,  2014)

The active site selection and protein preparations were

car-ried out by the inbuilt program of the software (Thomsen &

Christensen,  2006) Moreover, postdocking generated

pro-tein–ligand complex along with chemical interactions was

further analyzed and visualized by Discovery Studio (Sakkiah

et al., 2010) (https://www.3dsbi ovia.com/produ cts/colla borat ive-scien ce/biovi a-disco very-studi o/) and Chimera software (https://www.cgl.ucsf.edu/chime ra/) (Goddard et  al.,  2007; Thomsen & Christensen, 2006)

2.4 | Machine learning models for drug repurposing

Bayesian machine learning models (from FDA-approved drug screens) from Assay Central software (https://assay centr al.github.io/#) were used (Ekins et al., 2019) for iden-tifying the possibility of compounds that may work against SARS-CoV-2 A total of 45 drug molecules from previously reported screens for antiviral activities against different human viruses were used to validate their potential repurpos-ing ability usrepurpos-ing Bayesian machine learnrepurpos-ing models Each model in Assay Central used different metrics for evaluating predictive performance such as recall, precision, specific-ity, F1-Score, receiver operating characteristic (ROC) curve, Cohen's Kappa (CK), and the Matthews correlation coef-ficient (MCC) These models utilize extended-connectivity fingerprints of maximum diameter 6 (ECFP6) descriptors generated from the library

2.5 | Pharmacology network of potent compound(s)

Interactions between proteins and bioactive compounds or drugs are an integral part of biological processes in living or-ganisms In the present study, the pharmacology interaction network of the most active drug candidate (s) (in context with molecular interactions) was determined by STITCH (Search Tool for Interacting Chemicals) algorithm The interactions between drugs and receptors include direct (physical) and indirect (functional) associations and generated by compu-tational prediction from knowledge transfer between organ-isms, and from interactions aggregated from other databases (primary) Interactions in STITCH are derived from different sources such as genomic context predictions, (conserved) co-expression, automated text mining, and previous knowledge

in databases (Szklarczyk et al., 2016)

2.6 | Pharmacophore modeling

Pharmacophore features of the most active drug candidate (s)

in terms of higher affinity for target proteins were determined using the Ligandscout software which also reveals structural activity relationship (SAR) (Wolber & Langer, 2005) with a specific biological target (s) The fully optimized 3D structure

TABLE 2 Status of all the compounds used in the present study against SARS-CoV-2 target proteins for molecular interaction simulation

1 Favipiravir

2 Nitazoxanide

3 Remdesivir

4 Mycophenolic acid

5 Chloroquine

6 Niclosamide

(Galdecivir)

8 Gemcitabine

9 Rapamycin

(Sirolimus)

11 Cyclosporine

12 Emetine

13 Ribavirin

14 Luteolin

15 Tilorone (Amixin)

16 Glycyrrhizin

17 Eflornithine

18 Monensin

19 Arbidol (Umifenovir)

20 Silvestrol

21 Amiodarone

22 Dasatinib

23 Lopinavir

24 Nelfinavir

25 Oritavancin

26 Hydroxychloroquine

27 Ritonavir

28 Dalbavancin

29 Teicoplanin

30 Homoharringtonine

31 Alisporivir

32 Cepharanthine

33 Hexachlorophene

34 Imatinib

35 Nafamostat

36 Chlorpromazine

37 Camostat

38 Memantine

39 Indomethacin

40 Saracatinib

41 Telavancin

42 Promethazine

(Continues)

files of drug candidate (s) (Mol2 format) were loaded into the

working space of Ligandscout software, and key

pharmaco-phore features were identified The unique pharmacopharmaco-phore

features considered during analysis were H–bond acceptor,

H–bond donor, hydrophobic, aromatic, halogen bond donor,

and positively and negatively ionizable groups (Wolber &

Kosara, 2006)

2.7 | Cytotoxicity assay

The in vitro antiviral SARS-CoV-2 testing service was

availed by Translational Health Science and Technology

Institute (THSTI), NCR Biotech Science Cluster,

Faridabad-01, Haryana, India Since the antiviral activity

was tested in the Vero E6 cells, the test substance(s), that is,

cordycepin, should not be cytotoxic to host cells at the test

concentration/s; therefore, cytotoxicity test was performed

before anti-SARS-CoV-2 experiment The assay is done in

a 96-well plate (Thermo Scientific Nunc Edge 2.0) format

in 3 wells for each sample 1x10e4 VeroE6 cells were seeded

per well and incubated at 37°C overnight for the monolayer

formation Next day, cells were incubated with the test

sub-stance (cordycepin) at the different concentration (1, 5, 10,

20, and 50 μM) The cell without test substance was used

as negative control, and Remdesivir was used as positive

reference drug After 24 and 48 hr, cells were stained with

Hoechst 33342 and Sytox orange dye Images were taken at

10X, 16 images per well, which covers 90% of well area using

ImageXpress Microconfocal (Molecular Devices, LLC, San

Jose, CA-95134 USA) Hoechst 33342 nucleic acid stain is

a popular cell-permeant nuclear counter stain that emits blue

fluorescence when bound to dsDNA It stains all the live and

dead cells Sytox orange dye stains nucleic acids in cells with

compromised membranes This stain is an indicator of cell

death Finally, percentage of cell viability was determined in

cordycepin treatment group as compared to untreated control

2.8 | Anti-SARS-CoV-2 testing

Briefly, the assay was performed in a 96-well plate (Thermo

Scientific Nunc Edge 2.0) format in 3 wells for each sample

(Caly et al., 2020) 1 × 10e4 cells were plated per well and in-cubated at 37°C overnight for the monolayer formation Cells were incubated with the culture medium with cordycepin

at the potent non-cytotoxic concentration (10  μM) deter-mined as mentioned above Soon after (within 5 min), virus was added to each well at a defined multiplicity of infec-tion (MOI; 0.1 for 2 hr) Control cells were incubated with culture medium with corresponding concentration of vehi-cle Then, the plate was incubated at 37°C and culture su-pernatant was harvested at 24 and 48  hr later Viral RNA was extracted using the QIAamp 96 Virus QIAcube HT Kit (Qiagen, Hilden, Germany) from 100  μl culture superna-tant Reverse transcription was carried out using the BioLine SensiFAST cDNA kit (Bioline, London, UK), total reaction mixture (20 μl), containing 10 μl of RNA extract, 4 μl of 5× TransAmp buffer, 1 μl of Reverse Transcriptase, and 5 μl of nuclease free water The reactions were incubated at 25°C for 10 min, 42°C for 15 min, and 85°C for 5 min qRT-PCR (Applied Biosystems, Foster City, CA, USA) was performed using cycling conditions of 95°C for 2 min, 95°C for 5 s, and 60°C for 24 s, and Ct values for N and E gene sequence were determined The obtained data were used for calculating the

% virus inhibition, if any Ramdesivir was used as a positive control (Caly et al., 2020)

2.9 | Statistical analysis

The data of molecular docking score of five different poses

were expressed as mean ± SD The data were analyzed by

using one-way ANOVA followed by Tukey's range test

con-sidering *p ≤ 0.05 as statically significant values.

On the highest point of all questions associated with COVID-19 is; what are the most effective therapeutic op-tions to cure COVID-19? To answer this, the one among many approaches may be the use of pre-existing antivi-ral drugs alone or in combination to combat COVID-19 Similar was the case when FDA-approved hydroxychlo-roquine was implemented to modulate cellular response

43 Trametinib

44 Mefloquine

45 Cordycepin

Note: Dark blue: Cell culture/co-cultures, yellow: Primary cells/organoids, green: animal model, pink: Phase II, orange: Phase III, purple: phase IV, red: approved,

black: investigational.

TABLE 2 (Continued)

by suppressing the inflammatory response, thus improving

organ functions in COVID-19 patients (Zhou et al., 2020)

To date, no treatment specific to SARS-CoV-2 has been

reported to tackle the pandemic situation In this context,

drug repurposing, also known as repositioning or

reprofil-ing, is a unique and alternative strategy for generating

ad-ditional value to the existing investigational or approved

drugs by targeting disease other than that for which it

was originally reported (Huang et al., 2020; Nishimura &

Hara, 2018) This has several advantages over new drug

discovery since chemical synthesis, preclinical (animal

model), and clinical information (phase 0, I, and IIa)

in-cluding safety data, doses, and pharmacokinetics results

are already available for the molecules that may assist the

rapid drug development process Therefore, the present

study was undertaken with a strong commitment to

iden-tify potential therapeutic agent(s) against SARS-CoV-2

from the available antiviral drug candidates by using

com-putational approach The antiviral potential (Table  1) of

all the drug candidates used in the present study has been

validated in cell cultures/co-cultures, primary cells, animal

model, and phase II-IV and also approved against

SARS-CoV-2, HCoV-229E, HCoV-OC43, MERS-CoV, and

SARS-CoV (Table 2)

3.1 | Molecular interaction simulation

Molecular interaction simulation is the most commonly used method particularly for lead identification in com-puter-aided drug designing (CADD) The binding energy revealed the affinities between ligands and their corre-sponding target receptor molecule Lower binding energy (negative) indicates a higher affinity of the ligand for the re-ceptor (Kusumaningrum et al., 2014; Puspaningtyas, 2014) The heavily glycosylated large transmembrane spike gly-coprotein (type I) of SARS-CoV-2 accounts for its notable feature and plays an important role during viral attachment, fusion, and entry into the host body (Wrapp et al., 2020), and therefore, it is suggested that the inhibition of spike protein may be associated with inhibition of viral multipli-cation The transmembrane spike glycoprotein exists in a heterotrimeric form with three separate polypeptide chains: chain A, B, and C, forming each monomer The spike gly-coprotein has two functional domains, named as S1 and S2, both of which are responsible for successfully entry of coronavirus into the host cells (Wrapp et al., 2020) The molecular interaction study revealed that cordycepin has

a strong binding affinity followed by nitazoxanide, rapa-mycin, monensin, silvestrol, amiodarone, cepharanthine,

FIGURE 1 Comparative docking scores of different drug candidates are shown with SARS-CoV-2 spike protein Data are mean ± SD of 5

different poses (n = 5), One-way ANOVA, *p ≤ 0.05 [Colour figure can be viewed at wileyonlinelibrary.com]