key: cord-0927508-qtnd15sh
authors: Khurana, Pankaj; Varshney, Rajeev; Gupta, Apoorv
title: A Network-Biology led Computational Drug repurposing Strategy to prioritize therapeutic options for COVID-19
date: 2022-05-11
journal: Heliyon
DOI: 10.1016/j.heliyon.2022.e09387
sha: ee672e2bca105f8f97a04f303ad91e57afa9bf79
doc_id: 927508
cord_uid: qtnd15sh

The alarming pandemic situation of novel Severe Acute Respiratory Syndrome Coronavirus 2 (nSARS-CoV-2) infection, high drug development cost and slow process of drug discovery have made repositioning of existing drugs for therapeutics a popular alternative. It involves the repurposing of existing safe compounds which results in low overall development costs and shorter development timeline. In the present study, a computational network-biology approach has been used for comparing three candidate drugs i.e. quercetin, N-acetyl cysteine (NAC), and 2-deoxy-glucose (2-DG) to be effectively repurposed against COVID-19. For this, the associations between these drugs and genes of Severe Acute Respiratory Syndrome (SARS) and the Middle East Respiratory Syndrome (MERS) diseases were retrieved and a directed drug-gene-gene-disease interaction network was constructed. Further, to quantify the associations between a target gene and a disease gene, the shortest paths from the target gene to the disease genes were identified. A vector DV was calculated to represent the extent to which a disease gene was influenced by these drugs. Quercetin was quantified as the best among the three drugs, suited for repurposing with DV of -70.19, followed by NAC with DV of -39.99 and 2-DG with DV of -13.71. The drugs were also assessed for their safety and efficacy balance (in terms of therapeutic index) using network properties. It was found that quercetin was a forerunner than other two drugs.

dynamics docking using energy-minimized states of the flavonoids was performed to assess 88 their use against SARS-Cov-2. Quercetin and isoquercetin were among the top 10 flavonoids 89

identified [23] . The S glycoprotein of the virus is cleaved by endosomal proteases cathepsin 90 B/L into S1 and S2 subunits and thereafter the S2 subunit-mediates membrane fusion [24] . 91 Quercetin which is a common and readily available supplement is a strong cathepsin inhibitor 92

with an IC50 in the low micromolar range [25] . Therefore, it is considered as an adjuvant to 93 prevent or to weaken the course of viral infections. The CoV genome encodes two proteases, 94

papain-like protease (PLpro) and 3-chymotrypsin-like protease (3CLpro), also known as the 95 main protease (Mpro), that are important for virus replication [26] . These proteases are among 96 the most promising drug targets. Interestingly, in 3CLpro all residues involved in the 97 dimerization, substrate binding, and catalysis are 100% conserved between SARS-CoV and 98 . Remarkably, quercetin has a strong binding affinity for 3CLpro [28] . 99 Quercetin thus may block viral replication via inhibition of the viral cysteine protease 100 3CLpro [29] . Some studies have also proposed quercetin could possibly replace HCQ and be 101 used as prophylactic applications, specifically for the more vulnerable individuals such as the 102 elderly, diabetics, immuno-compromised and those with coronary ailments [29] . A recent study 103 utilises the gene expression profiles of Vitamin D and Quercetin activities to asses them as 104 viable candidates for their potential utility as COVID-19 pandemic mitigation agents [30] . 105 Based on this study, a randomized interventional clinical trial entitled "Phase II Clinical Trial 106

of Estradiol to Reduce Severity of COVID19 Infection in COVID19+ and Presumptive 107 COVID19+ Patients" (https://clinicaltrials.gov/ct2/show/NCT04359329) and two 108

interventional randomized clinical trials evaluating effects of Vitamin D on prevention and 109 treatment of COVID-19 NCT04334005 and NCT04344041 have been listed on 110 ClinicalTrials.gov website. In a recent study, using supercomputer-based in silico drug-111

docking to the COVID-19 viral spike protein, quercetin was identified among top 5 scoring 112

ligands for viral S-protein-human ACE2 receptor interface [31] . Studies also show that 113

administration of quercetin has resulted in significantly decreased expression of the ACE2 gene 114 during differentiation of human intestinal cells [30] . Quercetin may thus inhibit the binding 115 and entry of the virus in the cell [18, 32] . Molecular docking evaluation of a marine drug 116 metabolite database show that that algae polyphenols and quercetin derivates can be 117 successfully used as treatment against SARS-CoV-2 [33, 34] . One of the world's most cited 118

scientists, Michel Chrétien, has received 1 million dollars donation to begin a clinical trial of 119 quercetin against COVID-19. Quercetin has already been approved as safe for human 120 consumption by U.S.-based Food and Drug Administration, which means that if the results are 121 encouraging, the treatment will be readily available. 122 123 2-DG and nSARS-CoV-2 124 2-Deoxy-D-Glucose (2-DG) is a dual decoy of glucose and mannose. It has one of the hydroxyl 125 groups found in the chemical structure of natural glucose replaced by hydrogen. This renders 126 2-DG ineffective to be converted into energy or support proper glycosylation. It is thus an 127

inhibitor of hexokinase during glycolysis. The present study tries to explore the effectiveness of Quercetin or NAC or 2-DG as a 210

repurposed drug using shortest path in the drug-gene-gene-disease network to identify the 211 extent these drugs affect the SARS/MERS-associated genes. This extent has been quantified 212 and represented in a vector called DV that helps to prioritise these repurposed drugs. 213 214

Collection of Disease-genes and construction of Disease-gene network. 217 Figure 1 provides a graphical view of the network-based approach to compare Quercetin, NAC, 218 and 2-DG effectiveness against COVID-19 as a repurposed drug. The methodology has been 219 J o u r n a l P r e -p r o o f accepted from previous study of Taekeon Lee and Youngmi Yoon et al.[69] . First a directed 220 disease-gene network was constructed. COVID-19 infection is caused by a recently discovered 221 nSARS-CoV-2 that began spreading in December 2019, so there were very few validated-gene-222 targets available and these were not sufficient to construct a comprehensive disease-gene 223

network. The nSARS-CoV-2 shares an ancestral origin with betacoronavirus like SARS-CoV, 224

MERS-CoV, 229E, HKU1, NL63, and OC43. But SARS-CoV and MERS-CoV are the only 225 two betacoronavirus that cause severe respiratory illnesses in humans. Also, nSARS-CoV-2 226

shares 82% of its genome with human SARS-CoV and around 43% identity with MERS-CoV 227 as second best [70] . Therefore, SARS and MERS were used to identify the disease associated 228

genes. The disease (SARS and MERS)-gene associations were fetched from DisGeNet [71] . 229

DisGeNet is the largest publicly available repository of genes targets associated with human 230 diseases that have been collected from curated repositories, GWAS catalogs, animal models, 231 and the scientific literature. 232

The genes associated with these diseases were submitted to Graphite tool to construct a directed 233

disease (positive/negative/neutral) to every interaction between the genes. This relationship was 237 "positive" when activation or expression was recorded, was "negative" when inhibition or 238 repression was recorded, and "neutral" when "binding" was recorded [77] . A directed disease-239 gene network was thus constructed for further analysis. For each drug (Quercetin, NAC, and 2-DG) a comprehensive non-redundant list of associated 250 drug-genes was curated by extensive, manual literature-survey. To increase reliability, the 251 search was constrained to docking and experimentally validated in-vivo and in-vitro studies. 252

These curated drug-gene interactions were assigned as positive/negative/neutral as per drug 253 agonist/antagonist/binding effect on its gene target. These drug-gene targets were mapped to 254 the disease-gene-gene network to construct a directed drug-gene-gene-disease network. Thus 255 three such networks corresponding to three drugs namely quercetin, NAC and 2DG were 256 J o u r n a l P r e -p r o o f constructed. The drug-gene-gene-disease directed networks were visualized in 257

Cytoscape.3.6.0. 258 259

Shortest paths and DV Calculation from a drug-gene to a disease-gene. 260

To calculate the associations between a drug and a disease, the shortest paths from drug-genes 261

to disease-genes were identified. For this, the methodology was adopted from Taekeon Lee &  262 Youngmi Yoon et al which has been validated on 7083 associations between 1163 drugs and 263 298 diseases. [69] . They presume that neutral relationships maintain general actions in the body 264 and drugs affect diseases by regulating disease-related genes either by activation or inhibition 265

Pesca 3.8.0 plugin of Cytoscape was used to identify all the shortest paths between the drug-266

gene to the disease-gene in the three networks [78] . 267

The extent to which a drug affects the disease-associated genes can be determined using vector 268

DV [69] . To calculate DV, the associations between a drug-gene and a disease-gene in the 269 shortest path were calculated using types and weights of the shortest paths. The type of path 270

(Tk) was calculated by only "positive" and "negative" relationships between genes comprising 271 the path. Thus, 1 and −1 are assigned to each positive and negative edge in the path respectively. 272

The "neutral" path was assigned as 0 to nullify the neutral relationship. The path type (Tk) was 273 defined by multiplying the values corresponding to edges on the path. Path weight (Wk), was 274 calculated as the product of the reciprocals of the out-degree of the corresponding node on a 275

path. To consider the extent to which a disease-gene was affected by a drug-gene, V was 276 calculated. V is the sum of all products of weight (W) and path type (T) on the given shortest 277

path (Equation 1). 278

To express the effects of quercetin, NAC and 2-DG on the disease, a vector (DV) for each 281 drug-gene-gene-disease path was calculated. DV is denoted as the product of (Vk) and the 282 corresponding type of drug-target interaction (Tdk) (Equation 2). 283 284 J o u r n a l P r e -p r o o f 297 298

1 Degree

The number of edges linked to a node 2 Scaled Connectivity The degree of a studied node relative to the most connected node within the same module 3

Number of Selfloops The number of edges starting and ending at the same node 4

Number of Triangles The number of triangles that include the studied node as a vertex 5 Z Score A connectivity index based on degree distribution of a network. 6

Clustering Coefficient The number of the connected pairs between all neighbors of node 7

Neighborhood Connectivity

The average connectivity of all neighbors 8 Topological Coefficient The extent to which a node in network shares interaction partners with other nodes 9

Interconnectivity A connectivity index indicating the quality of the studied nodes being connected together 10

Bridging Coefficient The extent of the studied node lying between any other densely connected nodes in the network 11

Degree Centrality The number of links incident upon a studied node 12

Avg Shortest Path Length

The average length of shortest paths between the studied node and all other ones 13

Distance Sum The sum of all shortest paths starting from the studied node 14 Eccentricity

The maximum non-infinite shortest path length between the studied node and all other nodes in the network 15 Eccentric The absolute difference between nodes' eccentricities and network's average eccentricity 16

Deviation

The variation between sum of node distances and network unipolarity 17

Distance Deviation The absolute difference between nodes' distance sum and network's average distance 18

Radiality The level of reachability of a studied node via various shortest paths within the entire network 19

Closeness Centrality (avg)

The average number of steps required to reach the studied node from any node in a network 20

Closeness Centrality (sum)

The reciprocal of the sum of the shortest paths between the studied node and all other nodes in the network J o u r n a l P r e -p r o o f 21 Eccentricity Centrality

The largest geodesic distance between the node and any other node 22

Harmonic Closeness Centrality

The sum of the reciprocals of the average shortest path lengths of each node in network 23

Residual Closeness Centrality

The closeness measured by removing the studied node 24

Load Centrality The fraction of all the shortest paths that pass through the studied node 25

Betweenness Centrality The number of times the studied node serving as a linking bridge along shortest path between any two nodes 26

Normalized Betweenness

The fraction of network shortest paths that a given protein lies on 27

Bridging Centrality The product of the bridging coefficient and betweenness centrality 28

CurrentFlow 

Further 52, 18, and 40 experimentally validated drug-gene interactions of quercetin, NAC, and 320 2-DG respectively were manually curated by extensive literature survey (Supplementary Table  321 S2). Also, with each drug its interaction characteristics with the target gene are considered i.e. 322 positive (+1) for agonist, negative (-1) for antagonist and neutral (0) for just binding interaction. 323

Further, these genes were mapped to the disease-gene-gene network. This way 33, 11, and 31 324 genes for quercetin, NAC, and 2-DG were retained in the network. Finally, three networks i.e. To calculate the association between a drug i.e. quercetin, 2-DG, NAC, and diseases (SARS 354 and MERS), the shortest paths between drug-genes and disease-genes were identified in all the 355 three networks as explained in detail in the methodology section. 356

In quercetin-gene-gene-disease network, 528 shortest paths were identified ( Figure 5 ). 1% of 357 the shortest path had common genes interacting with both the drug and the disease directly. 358

These include ESR1, EGFR, ACE, CASP3, BCL2L1 (highlighted as yellow nodes in Figure  359 5). 7% of the shortest path had no connecting gene between drug-gene and disease-gene pairs. 360

Lastly, 79% of these shortest-path contain 1 connecting gene (highlighted as green nodes in 361 Figure 5 ) between drug-gene and disease-gene, whereas 13% of these shortest-path contains 2 362 connecting genes (highlighted as pink nodes in Figure 5 ) between drug-gene and disease-gene 363 ( Figure 5 ). This shows that quercetin could be an effective drug and would be able to affect 364 many disease-associated genes. Thus quercetin has a more number of associations with SARS and MERS disease-genes than 412 that of NAC or 2-DG, as shown by the shortest paths between the drug-gene to disease-gene. 413

In quercetin, 87% of its total shortest paths have at least 1 connecting gene. While NAC and 2-414 DG have 30% and 82% of their total shortest paths have at least 1 connecting gene respectively. 415

The influence of NAC on disease-genes seem to be complicated as 70% of the shortest paths 416 have atleast 2 connecting genes between drug-gene and disease-gene. 417 418

Further, this effect was quantified and expressed as a vector DV. DV is a vector and denoted 419 the extent to which a disease gene was influenced by a drug. has also been found as a direct inhibitor of NFKB1 in the shortest path network that further 447 regulates expression of pro-inflammatory cytokines such as IL12RB1, IL6, TNF, IL1B. and it 448

was observed that treatment with drugs that inhibited NF-κB activation led to a reduction in 449 inflammation and lung pathology in SARS-CoV-infected cultured cells and also increases the 450 mice survival rate [91] . Literature reports have shown that during both SARS-CoV and nSARS-451

CoV-2 infection there is an overproduction of pro-inflammatory cytokines (TNF, IL-6, and IL-452 1β) that results in a cytokine storm [95] . When this high cytokine concentrations get persist 453 over time could lead to an increased risk of vascular hyperpermeability, multiorgan failure, and 454 mortality [95] . Therefore, most of the therapeutical strategies develop until now are directed 455 towards maintaining an adequate inflammatory response for pathogen clearance this includes 456 the use of interleukin-1 inhibitors drug-like anakinra for immune-based therapy of nSARS-457

CoV-2. Both quercetin and anakinra are antagonists of the human IL- It makes bronchial mucous less viscous and being a cysteine derivative, helps in breaking 471 disulfide bridges between macromolecules, which leads to a reduction in mucus viscosity [99] . 472 NAC is also an inducer of Glutathione synthetase (GSS), which is an important enzyme in 473 glutathione biosynthesis. Glutathione (GSH) in the body has an antioxidant effect and it 474 reduces the formation of proinflammatory cytokines, such as IL-9 and TNF-α, and also has 475 vasodilator properties by increasing cyclic GMP levels and by contributing to the regeneration 476 of endothelial-derived relaxing factor. The inhibition of glutamate receptors (GRIN1,  477 GRIN2A, GRIN2D, GRIN3A) also promotes glutathione synthesis. This interaction can be 478 easily visualized in NAC targeted shortest path network, where NAC promotes the GSS and 479

its downstream signaling and inhibits glutamate receptors (GRIN1, GRIN2A, GRIN2D, 480 GRIN3A). Because of these two properties NAC is already proposed as a potential treatment, 481 preventive, and/or adjuvant against nSARS- . This shows that NAC could be an 482 important drug for the treatment of nSARS-CoV-2 infection but DV calculation shows it 483

presumably would be less effective as an nSARS-CoV-2 therapeutic agent as compared to 484

quercetin. 485 2-DG is a mimicking agent of glucose and is an antagonist of glucokinase/hexokinase 486 (GCK/HCK), an enzyme that converts glucose to glucose-6 phosphate in the glycolysis cycle 487

[100]. It is also a rate-limiting step of the oxidative metabolism. might not be as effective as quercetin and NAC as a re-purposed drug for COVID-19. 500 501 502

Network properties and Biological System Features to assess safety and efficacy balance 503

Therapeutic Index is a measure of relative safety of a drug. It is calculated as ratio of the dose 504 that produces toxicity to the dose needed to produce the desired therapeutic response. Drugs 505 with TI≤3 are considered less safe and referred to as Narrow Therapeutic Index (NTI) drugs. 506

Drugs with TI >3 are considered better than NTI on safety standards and referred to as Not -507

Narrow Therapeutic Index (NNTI) drugs. However, determination of TI is very complicated 508 for many drugs and is also highly susceptible to the variations of drug responses. 509

The efficacy-safety balance of a drug may be inferred from the network properties and 510

biological system profile of the drug-genes [105] . So, an effort was made to assess the safety 511 efficacy balance (i.e. NTI or NNTI) of the three drugs (quercetin, NAC and 2-DG) in the drug-512

gene-gene-disease network. Several connectivity and adjacency-based network properties, 513

properties based on shortest path length, have been found to be significantly different ( Some key features such as average shortest path length show increase from the targets of NTI 526 drug to that of NNTI one [105] . Among the three, quercetin showed the highest average shortest 527 path length, followed by NAC and 2-DG (Table 2) . Also some others such as average closeness 528 centrality demonstrated a decrease from the targets of NTI drug to that of NNTI drugs [105] . 529

The lower value of average closeness centrality of drug-target has been shown to demonstrate 530 a less lethality risk [106] . The interconnectivity values were lower for lethal diseases like 531 cardiovascular and oncogenic diseases [107] . Quercetin had the smallest average closeness 532 centrality among the three (Table 2) . 533

Based on these parameters, Quercetin and NAC were found to be closer to the profile of NNTI 534 drugs. Properties which determine the connectivity of the target (i.e average shortest path 535 length, bridging-coefficient, interconnectivity) showed an increasing trend from the targets of 536 NTI to NNTI drugs [105] . These parameters for quercetin, NAC and 2-DG are depicted in 537 Figure 8 . Average shortest path length, bridging-coefficient were highest in quercetin; whereas 538 interconnectivity was highest in NAC. 539 540

Parameters which determine the centrality of the target in the network (i.e. average closeness 541 centrality, degree, radiality) showed a decreasing trend from the targets of NTI to NNTI drugs 542

[105]. These parameters for quercetin, NAC and 2-DG are depicted in Figure 8 . Average 543 closeness centrality, degree, radiality were lowest in quercetin. 544 545 J o u r n a l P r e -p r o o f Biological system properties (i.e. affiliated pathways, number of similarity proteins) also 546 showed decreasing trend from the targets of NTI to NNTI drugs [105] . The affiliated pathways 547 and similarity proteins were lowest in quercetin targets (Table 2) . 548 549

It has been proven that the TI-related mechanism could be a result of synergistically effects 550 among these eight features [105] . As per these parameters, among the three, quercetin followed 551 the trends of NNTI drug. Other studies have also shown that number of similarity proteins and 552 affiliated pathways as a good indicator of target drugability [108, 109] . 553

Thus the targets of NTI drugs were highly centralized and connected in network, and the 554 numbers of similarity proteins and target-affiliated pathways were higher than those of NNTI 555 drugs [105] . 556

A list of 1580 core-essential-genes was obtained [110] . These genes without any context-557 dependence have been important for survival. For each drug-gene, its top 50 direct interactors 558

were obtained from STRING. The percentage of these direct interactors which were part of 559 core-essential-genes were identified. It is presumed that the drugs which target the core-560 essential-genes would be NTI drugs. It was found that quercetin had the minimum number of 561 core-essential-gene interactors (Table 2) . 562

Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2): An Update

How to improve R&D productivity: the pharmaceutical industry's grand 628 challenge

In Silico Oncology Drug Repositioning and Polypharmacology

Individualized network-based drug repositioning infrastructure for precision 632 oncology in the panomics era

Drug Repurposing: New Treatments for Zika Virus 634 Infection?

A comprehensive map of molecular drug targets

Network-based approach to prediction and population-based validation of in 638 silico drug repurposing

Drug repurposing against COVID-19: focus on 640 anticancer agents

More than 80 clinical trials launch to test coronavirus treatments

Virtual Screening of Inhibitors Against Spike Glycoprotein of 644 SARS-CoV-2: A Drug Repurposing Approach

The origin, transmission and clinical therapies on coronavirus disease 2019 646 (COVID-19) outbreak -an update on the status

Pleiotropic Effects of Tocotrienols and Quercetin on Cellular Senescence: 651 Introducing the Perspective of Senolytic Effects of Phytochemicals

Prophylactic Efficacy of Quercetin 3-beta-O-d-Glucoside against Ebola Virus 654 Infection

Quercetin and Vitamin C: An Experimental, Synergistic 656 Therapy for the Prevention and Treatment of SARS-CoV-2 Related Disease (COVID-19)

Coronavirus disease 2019 (COVID-19): first indication of efficacy of 659

Novirin in SARS-CoV-2 infection

Flavonoid-mediated inhibition of SARS coronavirus 3C-like protease 662 expressed in Pichia pastoris

Natural product-derived phytochemicals as potential agents against 664 coronaviruses: A review

Treatment efficacy analysis of traditional Chinese medicine for novel coronavirus 666 pneumonia (COVID-19): an empirical study from Wuhan

COVID-19 and chronological aging: senolytics and 669 other anti-aging drugs for the treatment or prevention of corona virus infection? Aging (Albany 670 NY)

Sanghamitra Pati Molecular docking-672 simulation edge assessment of potential and less-toxic 'anti-HIV-drug and phyto-flavonoid' 673 combination against COVID-19. Preprints

COVID-19 and Flavonoids: In Silico Molecular Dynamics Docking to the Active 675

Catalytic Site of SARS-CoV and SARS-CoV-2 Main Protease

Cysteine cathepsins: from structure, function and regulation to new frontiers

Quercetin, a flavonoid with anti-inflammatory activity, suppresses the 679 development of abdominal aortic aneurysms in mice

The SARS-coronavirus papain-like protease: 682 structure, function and inhibition by designed antiviral compounds

Prediction of the SARS-CoV

structure: virtual screening reveals velpatasvir, ledipasvir, and other drug 686 repurposing candidates

Silico Exploration of the Molecular Mechanism of Clinically 688

Oriented Drugs for Possibly Inhibiting SARS-CoV-2's Main Protease

Potential new treatment strategies for COVID-19: is there a role for 691 bromhexine as add-on therapy?

Tripartite Combination of Candidate Pandemic Mitigation Agents: Vitamin D

Estradiol Manifest Properties of Medicinal Agents for Targeted Mitigation of 694 the COVID-19 Pandemic Defined by Genomics-Guided Tracing of SARS-CoV-2 Targets in Human 695 Cells

Repurposing Therapeutics for COVID-19: Supercomputer-Based Docking to the 697

SARS-CoV-2 Viral Spike Protein and Viral Spike Protein-Human ACE2 Interface

Small molecules blocking the entry of severe acute respiratory syndrome 700 coronavirus into host cells

Natural products' role against

Putative Inhibitors of SARS-CoV-2 Main Protease from A Library of Marine 704

Natural Products: A Virtual Screening and Molecular Modeling Study

Rhinovirus induces an anabolic reprogramming in host cell metabolism 706 essential for viral replication

Improving cancer radiotherapy with 2-deoxy-D-glucose: phase I/II clinical 708 trials on human cerebral gliomas

Acute toxicity and cardio-respiratory effects of 2-deoxy-D-glucose: a 710 promising radio sensitiser

Clinical studies for improving radiotherapy with 2-deoxy-D-glucose: 712 present status and future prospects

Glucose antimetabolite 2-Deoxy-D-Glucose and its derivative as

Dengue virus induces and requires glycolysis for optimal replication

Triggering unfolded protein response by 2-Deoxy-D-glucose inhibits porcine 720 epidemic diarrhea virus propagation

Interference of nucleoside diphosphate 722 derivatives of 2-deoxy-D-glucose with the glycosylation of virus-specific glycoproteins in vivo

Selective down-regulation of human papillomavirus transcription by 2-725 deoxyglucose

Activation of the unfolded protein response by 2-deoxy-D-glucose inhibits 727

Kaposi's sarcoma-associated herpesvirus replication and gene expression

2-Deoxy-d-Glucose and Its Analogs: From Diagnostic to Therapeutic Agents

EXTH-07. Design and evaluation of WP1122, in inhibitor of 733 glycolysis with increased CNS uptake, Design and evaluation of WP1122, in inhibitor of 734 glycolysis with increased CNS uptake

Proteomics of SARS-CoV-2-infected host cells reveals therapy targets

N-Acetylcysteine: A potential therapeutic agent for SARS-CoV-2. Med 738 Hypotheses

N-acetylcysteine as a potential treatment for COVID-740 19

N-Acetylcysteine as an antioxidant and disulphide breaking agent: the reasons 742 why

COVID-19: The Potential Role of Copper and N-acetylcysteine (NAC) in a 744 Combination of Candidate Antiviral Treatments Against SARS-CoV-2

N-acetylcysteine decreases airway inflammation and 747 responsiveness in asthma by modulating claudin 18 expression

Coronavirus envelope protein: current knowledge

N-acetylcysteine decreases angiotensin II receptor binding in vascular 752 smooth muscle cells

Endogenous Deficiency of Glutathione as the Most Likely Cause of Serious 754 Manifestations and Death in COVID-19 Patients

Influence of nebulized unfractionated heparin and N-acetylcysteine in acute 756 lung injury after smoke inhalation injury

Effect of N-acetylcysteine on gas exchange after 758 methacholine challenge and isoprenaline inhalation in the dog

Acetylcysteine on Respiratory Secretions in Mechanically Ventilated Patients

Aerosolized administration of N-acetylcysteine 764 attenuates lung fibrosis induced by bleomycin in mice

Therapeutic blockade of inflammation in severe COVID-19 infection with 767 intravenous N-acetylcysteine

Use of hydroxychloroquine and N-acetylcysteine for 769 treatment of a COVID-19 patient

A Network-Biology Informed Computational Drug Repositioning Strategy 771 to Target Disease Risk Trajectories and Comorbidities of Peripheral Artery Disease

A survey of current trends in computational drug repositioning

Lethality and centrality in protein networks

A network view of disease and compound 778 screening

Systematic evaluation of drug-disease relationships to identify 780 leads for novel drug uses

Network-based drug repositioning: A novel strategy for discovering potential 782 antidepressants and their mode of action

SAveRUNNER: A network-based algorithm for drug repurposing 785 and its application to COVID-19

Drug repositioning using drug-disease vectors based on an integrated 787 network

Systematic Comparison of Two Animal-to-Human Transmitted Human 789 Coronaviruses: SARS-CoV-2 and SARS-CoV. Viruses

DisGeNET: a comprehensive platform integrating information on human 791 disease-associated genes and variants

graphite -a Bioconductor package to convert pathway topology to gene 793 network

Reactome graph database: Efficient access to complex pathway data

PID: the Pathway Interaction Database

KEGG: new perspectives on genomes, pathways, diseases and drugs

Prediction of drugs having opposite effects on disease genes in a directed 802 network

Finding the shortest path with PesCa: a tool for network reconstruction

A protein network descriptor server and its use in studying protein, disease, 806 metabolic and drug targeted networks. Brief Bioinform

The Comparative Toxicogenomics Database: update 2019

Basic local alignment search tool

High-Resolution CRISPR Screens Reveal Fitness Genes and Genotype-Specific 811

STRING: a database of predicted functional associations between 813 proteins

Network-based drug repurposing for novel coronavirus 2019-nCoV/SARS-CoV-815 2

Estrogen receptors regulate innate immune cells and signaling pathways

The stimulation of cell proliferation by quercetin is mediated by the 819 estrogen receptor

Interactions of quercetin with receptor tyrosine kinases 821 associated with human lung carcinoma

Good ACE, bad ACE do battle in lung injury, SARS

In silico analysis and molecular docking studies of potential 825 angiotensin-converting enzyme inhibitor using quercetin glycosides

A SARS-CoV protein, ORF-6, induces caspase-3 mediated, ER stress and JNK-828 dependent apoptosis

SILICO DOCKING OF QUERCETIN 830 COMPOUND AGAINST THE HELA CELL LINE PROTEINS

Inhibition of NF-kappaB-mediated inflammation in severe acute 833 respiratory syndrome coronavirus-infected mice increases survival

Diagnosis, Prevention, and Treatment of Thromboembolic Complications 836 in COVID-19: Report of the National Institute for Public Health of the Netherlands

Clinical and immunological features of severe and moderate coronavirus 839 disease 2019

COVID-19 cytokine storm: the interplay between inflammation and 841 coagulation

Molecular Docking Analysis: Interaction Studies of Natural Compounds to 843

Anti-inflammatory Targets. Quantitative Structure-activity Relationship

Quercetin and Quercitrin Attenuates the Inflammatory Response and Oxidative 845

Stress in LPS-Induced RAW264.7 Cells: In Vitro Assessment and a Theoretical Model. Biomed 846 Res Int

Flavonoids differentially modulate liver X receptors activity-Structure-848 function relationship analysis

Effects of drugs on mucus glycoproteins and water in bronchial 850 secretion

Synergism of ursolic acid derivative US597 with 2-deoxy-D-glucose to 852 preferentially induce tumor cell death by dual-targeting of apoptosis and glycolysis

Participation of hindbrain AMP-activated protein kinase in 855 glucoprivic feeding

Phosphorylation of Janus kinase 1 (JAK1) by AMP-activated protein 857 kinase (AMPK) links energy sensing to anti-inflammatory signaling

The Complex Role of STAT3 in Viral Infections

Tyrosine dephosphorylation of STAT3 in SARS coronavirus-infected Vero E6 862 cells

Determining the Balance Between Drug Efficacy and Safety by the Network

Biological System Profile of Its Therapeutic Target

A novel paradigm for potential drug-targets discovery: quantifying 866 relationships of enzymes and cascade interactions of neighboring biological processes to 867 identify drug-targets

What are next generation innovative therapeutic targets? Clues from 871 genetic, structural, physicochemical, and systems profiles of successful targets

Therapeutic target database update 2018: enriched resource for facilitating 874 bench-to-clinic research of targeted therapeutics

High-Resolution CRISPR Screens Reveal Fitness Genes and Genotype-877

In this study, we compare three drug candidates i.e. quercetin, NAC, and 2-DG to be 575repositioned for nSARS-CoV-2 by exploring associations between drugs and disease genes 576 derived from a directed drug-gene-gene-disease network using the shortest path. 577In the case of quercetin, 1% of the shortest path had common genes interacting with both the 578 drug and the disease directly; 7% of the shortest path had no connecting gene between drug-579 gene and disease-gene pairs; 79% of these shortest-path contain 1 connecting gene between 580 drug-gene and disease-gene and 13% of these shortest-path contains 2 connecting genes 581 between drug-gene and disease-gene. In the case of NAC, there were no common genes 582interacting with both the drug and the disease directly; 8% of the shortest paths have no 583 connecting genes; whereas, 22% and 79% of the shortest path have 1 and 2 connecting genes 584 between drug-gene and disease-gene pair respectively. In the case of 2DG, there were no 585 common genes interacting with both the drug and the disease directly; 14% of the shortest paths 586in the 2-DG affected network have no connecting genes between drug-gene and disease-gene 587pair. While 68% and 18% of the shortest path contains either 2 connecting genes or 1 588connecting gene respectively. The shortest path calculation shows that both NAC and 2-DG 589 follow a complex route to effect the SARS and MERS genes as compared to quercetin. Further, 590the extent of a disease-gene influenced by a drug in the directed drug-gene-gene-disease 591 network was quantified through a vector called DV; it showed that quercetin has a greater 592 therapeutic influence on the SARS and MERS associated genes, followed by NAC and 2-DG. 593The DV values for quercetin, NAC and 2-DG are -70.19 , -39.99 and -13.71 respectively. Thus 594 from the present analysis, quercetin potentially appears to be a better drug that can be 595 repurposed against SARS and MERS than NAC and 2-DG. The shortest path calculation of 596 quercetin revealed five genes ESR1, EGFR, ACE, CASP3, BCL2L1 that are also directly 597 associated with SARS and MERS. As nSARS-CoV-2 shares an ancestral origin and genome 598 with betacoronavirus like SARS-CoV, MERS-CoV, therefore, quercetin could potentially be a 599 more potent drug candidate that could be effective against nSARS-CoV-2 infection. 600Detailed biological implications of the three drugs provided enough evidences for their 601 repurposing ability as a therapeutic intervention against nSARS-CoV-2. Further the efficacy-602 safety balance of these drugs were assessed by connectivity and centrality of the drug-targets 603 in the drug-gene-disease network. The network properties and biological system features were 604 used to assess safety and efficacy balance and hence classify the drugs as NTIs or NNTIs. The 605 network properties that have been found to be significantly different (p-value<0.05) were used 606 for evaluation. Based on these parameters, Quercetin and NAC were found to be closer to the 607 profile of NNTI drugs. Quercetin was found to follow the pattern of a better and safer drug. 608The limitation of the study includes the following. Drug repositioning can identify novel 609indications for existing drugs and is a robust and quick alternative to traditional methods. 610However, the likelihoods put-forth in the present study need to be validated experimentally. 611The method relies heavily on the previous knowledgebase that has been established in terms in 612 gene-gene, gene-disease and gene-drug interactions. The method would be more robust as more 613and more experimentally-validated interactions are incorporated in the drug-gene-gene-disease 614network. The method may be improvised further if the kinetics of each of these interactions 615 could be added as edge-weights to the network. However, such kinetics data is not available 616 for majority of the interactions. 617 J o u r n a l P r e -p r o o f Thus, in the present study, extent to which a SARS-CoV-2 disease genes were influenced by 618 quercetin, NAC and 2DG was computed and quercetin was quantified as the best among the 619 three drugs, suited for repurposing. Further using biological implications and network 620properties, the drugs were assessed for their safety and efficacy balance (in terms of therapeutic 621 index). It was found that quercetin was a forerunner than other two drugs. 622 623