key: cord-0767679-m0rqg9lx
authors: Gan, Hanjie Jonathan; Harikishore, Amaravadhi; Lee, Jihye; Jeon, Sangeun; Rajan, Sreekanth; Chen, Ming Wei; Neo, Jun Long; Kim, Seungtaek; Yoon, Ho Sup
title: Antiviral activity against Middle East Respiratory Syndrome Coronavirus by Montelukast, an anti-asthma drug
date: 2020-12-10
journal: Antiviral Res
DOI: 10.1016/j.antiviral.2020.104996
sha: 4caff4355e853202a1032b31307556b3140ca43b
doc_id: 767679
cord_uid: m0rqg9lx

Middle East respiratory syndrome (MERS) is a respiratory disease caused by a coronavirus (MERS-CoV). Since its emergence in 2012, nosocomial amplifications have led to its high epidemic potential and mortality rate of 34.5 %. To date, there is an unmet need for vaccines and specific therapeutics for this disease. Available treatments are either supportive medications in use for other diseases or those lacking specificity requiring higher doses. The viral infection mode is initiated by the attachment of the viral Spike glycoprotein to the human Dipeptidyl Peptidase IV (DPP4). Our attempts to screen antivirals against MERS led us to identify montelukast sodium hydrate (MSH), an FDA-approved anti-asthma drug, as an agent attenuating MERS-CoV infection. We showed that MSH directly binds to MERS-CoV-Receptor-Binding Domain (RBD) and inhibits its molecular interaction with DPP4 in a dose-dependent manner. Our cell-based inhibition assays using MERS pseudovirions demonstrated that viral infection was significantly inhibited by MSH and was further validated using infectious MERS-CoV culture. Thus, we propose MSH as a potential candidate for therapeutic developments against MERS-CoV infections.

Middle East respiratory syndrome (MERS) emerged in 2012, extending its spread from Saudi Arabia, Qatar, and Jordan to 27 different countries (Sharif-Yakan and Kanj, 2014; WHO, 2020; Zaki et al., 2012) . Subsequently, it was found to be a coronavirus (CoV) infection that has undergone nosocomial amplifications contributing to its spread (Majumder et al., 2017) . To date, there are a total of 2,494 lab-confirmed cases and 858 deaths resulting from MERS, leading to a high mortality rate of 34.5%. There are currently no available vaccines or therapeutics specific to MERS. Treatment is mostly supportive, and broad-spectrum therapeutics were previously used for severe cases (CDC, 2015; WHO, 2018) . However, the lack of specificity and the need for high dosages to exert any therapeutic effect resulted in adverse side effects such as ribavirininduced anemia and carcinogenicity (Dyall et al., 2014) . Therefore, due to this unmet need for more effective treatments and its high epidemic potential, MERS-CoV is placed on the list of priority diseases and pathogens for critical research and development by the World Health Organization (WHO) (WHO, 2018) .

Until today, vaccine development against MERS led to the identification of potential candidates such as DNA vaccines and antibodies that are in clinical trials and preclinical phases (Cho et al., 2018; Perlman and Vijay, 2016; Zhang et al., 2014; Zhou et al., 2019) . Although vaccine development is critical combating infectious diseases, concurrent research on therapeutic development is still required to address the need for targeted treatment options. As such, our study prioritizes the development of inhibitors against the viral-host interface to suppress viral progression. For this, the primary focus was on the 204-residue long viral receptor-binding domain (RBD) on the S1 glycoprotein of MERS-CoV, which interacts with its host cell receptor:

Dipeptidyl peptidase IV (DPP4) .

J o u r n a l P r e -p r o o f 4 Within the S1 glycoprotein, RBD was found to interact with DPP4 by mutation studies (Du et al., 2013; Wang et al., 2013; Zhang et al., 2014) . More importantly, the MERS-CoV RBD-DPP4 complex's crystal structure revealed the precise molecular basis of the MERS-CoV RBD-DPP4 interface. A unique feature of the interaction is binding the MERS-CoV RBD to DPP4 at blades 4 and 5 of the β-propeller domain exposed to the surface, instead of its canonical catalytic site . Consistent with other studies, it provided a probable explanation for DPP4 inhibitors' ineffectiveness in blocking this viral attachment (Raj et al., 2013) . Anyhow, the MERS-CoV RBD-DPP4 complex structure directed the targeting of this interface region as the probable inhibition site of viral attachment and thereby its entry.

In this direction, we studied the MERS-CoV RBD-DPP4 complex interface and screened potential therapeutic drugs for MERS. The crystal structure of MERS-CoV RBD-DPP4 complex suggests that MERS-CoV RBD recognizes its cognate receptor DPP4 at a different site from the catalytic/sitagliptin binding site in DPP4. MERS-CoV RBD was found to interact with residues ranging from 281 to 296 of DPP4. Along this 15-amino acid stretch, a short helical region (residues 290-296) of DPP4 fits nicely into RBD's hydrophobic surface pocket. Thus, in this study, we targeted the protein-protein interactions harboring the RBD's hydrophobic pocket (Y540-Q544, W553-S557, N501-D510 residues) that anchored the short helix region at the binding site to disrupt receptor recognition of and thereby prevent viral entry into host cells.

Based on the PLP fitness scores and the molecular interactions at the MERS-CoV as mentioned above RBD residues, seven FDA-approved drugs were selected to evaluate in vitro binding and cell-based viral entry assays. Through in vitro fluorescence quenching titrations, Montelukast Sodium Hydrate (MSH) (NCBI, 2020; Song et al., 2018) was found to interact with MERS-CoV RBD in a dose-dependent manner. Subsequently, assays using MERS pseudovirions revealed J o u r n a l P r e -p r o o f 5 significant inhibition of viral entry by MSH. Further validation was then performed using an infectious virus culture (MERS-CoV/EMC/2012) (CDC, 2015; Park et al., 2017) .

J o u r n a l P r e -p r o o f 6

The crystal structure of MERS-CoV RBD-DPP4 (PDB: 4L72) was employed to virtually screen the FDA approved library dataset. The protein subunits of the complex structure were prepared by removing the heteroatoms, adjusting the charges, potentials, bond orders, and missing atoms with CHARMM force field and energy minimized with heavy atom constraint for 5000 steps of conjugate gradient algorithm in Biovia Discovery Studio 4.0 (Dassault Systèmes, 2016) . Thus, the minimized protein is used for further molecular docking simulation studies. 3D coordinates of ligands in the FDA approved library were obtained from the Zinc database (Sterling and Irwin, 2015) . The ligands were energy minimized using the smart minimizer algorithm, and this minimized library was taken into molecular docking studies. Molecular studies were carried out using GOLD, CCDC suite of programs (Jones et al., 1997) . As there were no known ligands characterized for this protein, default docking parameters were employed to generate docking poses for the ligands in FDA-approved library. All modeling and simulation studies were carried on a Linux Workstation.

pFastBac DUAL vectors containing the MERS-CoV RBD and DPP4 constructs were kindly gifted by Professor Xinquan Wang. Truncations were performed on the RBD construct to shorten the target sequences to the reported critical region for downstream analyses before cloning into a pFastBac-Sec-NH vector. After that, transposition, isolation of bacmids, and subsequent bacmid transfection for the generation of RBD and DPP4 baculoviruses were systematically conducted. The low-titer P 1 baculovirus stocks were amplified to high-titer P 3 baculovirus stocks before the expression of recombinant proteins. J o u r n a l P r e -p r o o f 7 For recombinant protein expression, baculoviruses were used to infect sf9 cells at a seeding density of 2.5 x 10 6 cells/mL. Cells were then incubated for 72 hours in a shaking incubator at 27 o C at 130 rpm, protected from light before harvesting. The culture was then subjected to centrifugation at 9,000x g for 15 minutes, and the supernatant containing the soluble secreted proteins were purified by Ni 2+ -NTA affinity chromatography. The elution fractions were collected, concentrated, and injected for secondary purifications via the HiLoad Superdex 75 16/60 and Superdex 200 Increase 10/300 GL columns.

Maltose-binding protein (MBP) tagged MERS-CoV RBD wild type (WT), and mutant constructs (V555A, R542D, and Y540A) were cloned into pHLmMBP-10 vector and subsequently transfected using Expifectamine 293 (Gibco) transfection reagent and expressed in ExPi293 cells. The culture was incubated at 37°C at 120 rpm for 120 hours before harvesting. The culture was then subjected to centrifugation at 3,000x g, and the supernatant containing the soluble secreted proteins were purified by Ni2+-NTA affinity chromatography. The elution fractions were collected, concentrated, and injected for secondary purifications via the HiLoad Superdex 200 16/60 column.

Purified recombinant proteins were analyzed by SDS-PAGE on a 10 % SDS-gel. Bands of interest were extracted and sent for MALDI-TOF Mass Spectroscopy Protein ID analysis.

Binding of the recombinant MERS-CoV RBD sample with its host cell receptor, DPP4, was performed to assess its functionality. The purified RBD and DPP4 proteins were incubated at 4 o C for 2 hours before injecting into a Superdex 200 Increase 10/300 GL column for analysis.

Negative control runs were prepared beforehand with DPP4 only, and MERS-CoV RBD only samples for comparison (Suppl. Figure 2 ). J o u r n a l P r e -p r o o f 8

Purified MERS-CoV RBD samples were centrifuged at 16,000x g for 10 minutes to pellet down any aggregates. The protein sample was diluted in 25 mM Tris-HCl (pH 8.0), 30 mM NaCl and 0.01 % NaN 3 to a final concentration of 5 µM per well in a Corning® 96 well, flat bottom black plate. Ligands diluted in DMSO were added into each well at a protein to ligand ratio of 1:5.

DMSO was added in place of the ligands for the negative control. Subsequent screens were carried out for a protein to ligand ratio of 1:5, 1:10, 1:20, 1:30, and 1:40.

The plate was then loaded onto a Tecan Safire 2 ™ microplate reader for endpoint measurements.

The absorption wavelength was pre-set at 290 nm, and the emission range was set between 290 to 430 nm, respectively. Quenching data was first baseline corrected and fitted into a non-linear regression Gaussian model using the following equation: Y=Amplitude*exp(-0.5*((X-Mean)/SD)^2). Then, the amplitude of each sample was recorded and transformed to fit a saturation curve. The K D was then calculated using the following equation: Y=Bmax*X/(K D + X).

Purified MBP-tagged MERS-CoV RBD was diluted in 25 mM Tris-HCl (pH 8.0), 30 mM NaCl and 0.01 % NaN 3 to a final concentration of 0.2 mg/ml per well in a Framestar® 96, semi-skirted PCR plate (4titude, Surrey). 9 two-fold serial dilutions were conducted for MSH in DMSO, from a final concentration of 100 µM down to 195 nM. DMSO was added in place of MSH for the control set. SYPRO Orange dye (Life Technologies, CA) was then added into each well to a final concentration of 1x. The plate was then sealed with a MicroAmp® Optical Adhesive Film from Applied Biosystems (Life Technologies, CA) and loaded into the Applied Biosystems® 7500

Real-Time PCR System. The system was initiated with a temperature increase from 25 o C to 95 J o u r n a l P r e -p r o o f 9 o C, with a ramp in 0.5 o C per 5 seconds and plate-reads with respective temperature increments.

The absorption and emission wavelengths were measured at 490 and 575 nm, respectively. Data was processed and fitted to a Boltzmann sigmoidal model for melting temperature (T m ) analyses.

Analysis of the ThermoFluor data via the equation: Y=Bottom + (Top-Bottom)/(1+(IC 50 /X)^HillSlope), revealed a K D calculation of 8.41 ± 0.25 µM.

Samples were pre-incubated in 4 o C before the analytical gel filtration using a Superdex 200

Increase 10/300 GL column. For the two-sample conditions, (1) RBD-DPP4 was incubated for 3 hours, followed by incubation with MSH overnight, and (2) RBD was incubated with MSH overnight followed by incubation with DDP4 for 3 hours. The molar concentration of MSH 10fold excess compared to the protein samples. The samples were monitored via Ultra-violet (UV) absorption at 280 nm, and shifts in fractionation peaks were analyzed.

MERS-Spike pseudovirions were designed and produced using a 3 rd generation lentiviral pseudotyping system involving Envelope, Transfer, and Packaging vectors in HEK 293T cells (ATCC, VA). Prior to usage, viral supernatants were concentrated, and immunoblot analyses using an anti-MERS-S1 center antibody (Sino-Biological, CN) was conducted to detect the presence of MERS Spike proteins. Infection assays were used to test the functionality as well as the target specificity of these pseudovirions on Vero E6 cells (ATCC C1008, VA). Quantitative measurements were obtained via luciferase assay. Measurements were obtained using the Tecan Safire 2 ™ microplate reader. Luminescence data positively correlates to viral entry, and comparisons between untreated and treated wells were used for viral entry inhibition analysis.

Vero E6 cells were grown and maintained on a 10 cm 3 culture plate in Dulbecco's Modified Eagle Medium (DMEM) medium with high glucose and 4 mM L-glutamine, supplemented with 10 % (v/v) fetal bovine serum (FBS) and 1 % (v/v) 100x Penicillin-Streptomycin solution at 37°C in a 5% humidified CO 2 incubator. To determine the highest non-toxic dose of MSH for subsequent assays, a cell viability assay using WST-1 reagent (Roche, CH) was performed on Vero E6 cells with increasing MSH concentration. Vero E6 cells were incubated at 37 o C for 48 hours with MSH treatment to simulate inhibition assay conditions before adding WST-1 and measurements according to the Cell Proliferation Reagent WST-1 manufacturer's protocol.

MERS-Spike pseudovirions were pre-incubated with MSH for 30 minutes in room temperature before inoculation to Vero E6 cells. Inoculums were left incubated with the cells at 37 o C for 48 hours before harvesting for luciferase measurements. Luminescence measurements were subjected to baseline-correction with the uninfected wells, and comparisons between untreated and treated wells were analyzed. 

The viral inoculums and drug treatments were incubated with the cells for 24 hours. After this, the infected cells were fixed at 24 hours post-infection (hpi) with 4% PFA and permeabilized with 0.25% Triton X-100 (Sigma Aldrich, USA) in Phosphate Buffer Saline (PBS) for 30 min. 

For assays with three or more data sets compared to the control as a reference, One-way ANOVA statistical analyses were performed with Dunnett's Post-hoc test to compare all columns with the control column. In experiments with fewer than 3 data sets, an unpaired one-tailed t-test was conducted instead. * p < 0.05, ** p < 0.01, *** p < 0.001.

Based on our in silico screening, one of the best fitting ligand, Montelukast Sodium Hydrate was involved in π-π stacking interactions with W553 residue. The ligand-binding was further stabilized by strong π-π stacking interaction of 7-chloroquinoline moiety with Y540 and hydrophobic interactions with V555, sidechain CH 2 atoms of K502, R542, and S557 residues.

( Figure 1B, C) .

For binding studies, the recombinant RBD was expressed via a baculoviral expression system for the required post-translational glycosylations (Alfalah et al., 2002; Ikushima et al., 2000) , followed by purification (Suppl. Figure 1 , 2). Seven compounds identified from our structureguided in silico screening of FDA-approved drugs were assessed for potential binding using intrinsic fluorescence quenching. Preliminary screening of these ligands was conducted at a fixed protein-ligand molar ratio of 1:5, and the appropriate negative controls were implemented (Suppl. Figure 3A and B). Results suggested that Nalmefene (N), Cefaclor (Cef), and MSH could potentially interact with the MERS-CoV RBD sample (Suppl. Figure 3C ). Figure 2C ).

Additional mutation studies to ascertain our in silico modeling data was carried out. Three specific point mutations, Y540A, R542D, and V555A, were generated, followed by fluorescence quenching with MSH, and was compared with the wild type MERS-CoV RBD. The study revealed that V555A exhibited a notable reduction in quenching, suggesting that the predicted hydrophobic interactions between V555A and MSH could be important for binding between MERS-CoV RBD and MSH ( Figure 2D ).

and MSH exhibited a slight but distinct and consistent peak shift in fractionation volume from the RBD-DPP4 complex peak towards the DPP4-only peak regardless of mixing order, suggesting potential inhibition and complex dissociation effects of MSH in RBD-DPP4 complex formation ( Figure 3 ).

We then investigated the potential inhibitory effect of MSH using non-replicative pseudovirions.

For the monitoring and quantitative analysis of PV entry, a luciferase reporter gene was incorporated into the PV ( Figure 4A ) (Grehan et al., 2015) . Before the inhibition assays, the 50% J o u r n a l P r e -p r o o f 15 cytotoxic concentration (CC 50 ) of MSH was found to be 35.1 µM (Suppl. Figure 4C ). Based on this value, three different MSH concentrations were systematically tested in our infection assays:

10 µM, 15 µM, and 20 µM. The toxicity screening for these selected MSH concentrations was carried out in parallel with the viral entry inhibition assay to ensure the minimal effect of cell death on inhibition assay results. Here, our results reflected an enhanced reduction in luciferase activity at 20 µM MSH, signifying maximum viral entry inhibition with minimal effect arising from cell death due to drug toxicity ( Figures 4B and C ).

An additional time-point assay was conducted to study the time-dependent inhibitory effect of MSH. Our data reflected a significant decrease in luminescence, directly correlating to a distinct viral entry inhibition, after 6 hours of treatment of 20 µM MSH ( Figure 4D ). This indicates that although the PV inoculum was incubated with the cells for a short duration of 6 hours before being replaced with fresh media for the subsequent 42 hours for luminescence detection, it is sufficient to observe a significant decrease in PV entry. This signifies that although the PV inoculum was incubated with the cells for a short duration of 6 hours before being replaced with fresh media for the subsequent 42 hours for luminescence detection, it is sufficient to observe a significant decrease in PV entry. Therefore, this finding suggests that MSH is involved in the early stages of viral infection, in line with our hypothesis that MSH targets the MERS-CoV spike glycoprotein to inhibit viral progression.

To validate our results from the PV assays, we performed further experiments with infectious MERS-CoV (MERS-CoV/EMC/2012) under two different conditions: Pre-incubation and Cotreatment ( Figure 5 ). In the pre-incubation experiment, the virus and MSH were mixed and incubated for one h before the cells' addition. On the other hand, both drug and MSH were added J o u r n a l P r e -p r o o f 16 to the cells simultaneously without prior incubation in the co-treatment experiment. The concentration of MSH was maintained until the infected cells were fixed for immunofluorescence imaging. As controls, chloroquine and lopinavir's antiviral efficacy were assessed in parallel (Chan et al., 2015) . The IC 50 of MSH with the infectious MERS-CoV was determined by the dose-response curve (DRC) analysis. It was approximately 3 µM in both preincubation and co-treatment experiments ( Figure 5 ). This IC 50 value was substantially lower than those with chloroquine and lopinavir, suggesting a more potent antiviral efficacy of MSH against infectious MERS-CoV.

Due to the lack of vaccines and drugs against MERS, there is a critical need for research and therapeutic discovery and development. Such studies are vital, as we cannot ignore the possibility of a mutant form of MERS emerging in the future, such as SARS-CoV2 from SARS-CoV. In this direction, structure-guided in silico screening of FDA-approved drug libraries was conducted, revealing seven potential hits. Intrinsic fluorescence quenching titrations performed using recombinant MERS-CoV RBD samples showed a significant dose-dependent quenching effect upon MSH addition. In addition to the substantial quenching of fluorescence signals, a redshift in the emission maxima was observed upon increasing MSH concentration. This red-shift correlates to a conformational change to an increasingly polar microenvironment (Akbar et al., 2016; Di Bartolo, 2013; Möller and Denicola, 2002) . Further studies conducted with DSF agreed with our previous quenching analyses, which revealed that MSH is a potential U-binder, resulting in destabilizing interactions with MERS-CoV RBD (Cimmperman et al., 2008) .

Structure elucidation attempts were performed using co-crystallization and soaking methods, but diffraction data and analysis revealed the absence of the inhibitor electron density, indicating the requirement for further optimization.

Cell-based PV infection assays demonstrated a significant inhibitory effect exerted on PV entry.

The current hypothesis is that MSH interacted with the RBD on the MERS-CoV Spike glycoproteins, which then inhibited RBD/DPP4 complex formation and, consequentially, viral entry inhibition. A limitation of PV studies is the absence of co-factors and proteins on the virion surface, which might help compensate or strengthen viral attachment (Jolly and Sattentau, 2013) .

As such, the infectious virus isolated from MERS-infected patients were used in subsequent inhibition assays on Vero cells.

J o u r n a l P r e -p r o o f 18 MSH cytotoxicity screens were conducted before cell-based assays with infectious MERS-CoV to ascertain any reduction in the viral progression due to MSH's inhibitory effect. Moreover, MSH was identified as a potent inhibitor of viral entry as it exhibited a more robust response observed with the pseudovirion experiments. Calculations from our infectious MERS-CoV assays revealed an IC 50 of ~ 3 µM, comparatively more potent to LPV, which is currently used in clinical trials as part of the MIRACLE trial for MERS-CoV (Chan et al., 2015) . Although the calculated IC 50 falls in the µM range compared to other therapeutic compounds (nM to pM ranges), MSH still exhibited potent viral entry inhibition (Makau et al., 2017) .

A future study to examine the effect of MSH on MERS-CoV infection in DPP4 transgenic mice models would be an ideal way forward Li et al., 2017) . Besides, a structure-based drug design approach could be adopted, wherein the structure elucidation of MSH with RBD followed by the design of MSH analogs and characterization will enable us to improve MSH potency.

In summary, in this study, we showed that the anti-asthmatic drug -MSH, a leukotriene D4 receptor antagonist, could bind MERS-CoV RBD and inhibit the viral progression, suggesting that MSH could be considered as a potential therapeutic lead for MERS-CoV infections. 

Tryptophan fluorescence quenching as a binding assay to monitor protein conformation changes in the membrane of intact mitochondria

Intestinal Dipeptidyl Peptidase IV Is Efficiently Sorted to the Apical Membrane through the Concerted Action of N-andO-Glycans as Well as Association with Lipid Microdomains

Middle East Respiratory Syndrome Coronavirus Outbreak in the Republic of Korea

Treatment with lopinavir/ritonavir or interferon-β1b improves outcome of MERS-CoV infection in a nonhuman primate model of common marmoset

Development of Middle East respiratory syndrome coronavirus vaccines-advances and challenges

A quantitative model of thermal stabilization and destabilization of proteins by ligands

Wild-type and innate immune-deficient mice are not susceptible to the Middle East respiratory syndrome coronavirus

Introduction to Fluorescence Spectroscopy with Applications to Biological Systems

Identification of a receptor-binding domain in the S protein of the novel human coronavirus Middle East respiratory syndrome coronavirus as an essential target for vaccine development

Repurposing of clinically developed drugs for treatment of Middle East respiratory syndrome coronavirus infection

An optimised method for the production of MERS-CoV spike expressing viral pseudotypes

Internalization of CD26 by mannose 6-phosphate/insulin-like growth factor II receptor contributes to T cell activation

Attachment factors

Development and validation of a genetic algorithm for flexible docking

Mouse-adapted MERS coronavirus causes lethal lung disease in human DPP4 knockin mice

Nosocomial amplification of MERS-coronavirus in South Korea

Identification of small molecule inhibitors for influenza a virus using in silico and in vitro approaches

Protein tryptophan accessibility studied by fluorescence quenching

PubChem Compound Database; CID=71311637

Hospital outbreaks of middle east respiratory syndrome

Middle East respiratory syndrome vaccines

Dipeptidyl peptidase 4 is a functional receptor for the emerging human coronavirus-EMC

Emergence of MERS-CoV in the Middle East: origins, transmission, treatment, and perspectives

Asymptomatic Middle East Respiratory Syndrome coronavirus infection using a serologic survey in Korea

ZINC 15-ligand discovery for everyone

BIOVIA, BIOVIA Discovery Studio

Structure of MERS-CoV spike receptor-binding domain complexed with human receptor DPP4

Annual review of diseases prioritized under the Research and Development Blueprint

WHO, 2020. MERS Situation Update

Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia

Current advancements and potential strategies in the development of MERS-CoV vaccines

Advances in MERS-CoV vaccines and therapeutics based on the receptor-binding domain

Montelukast (MSH) directly binds to MERS-CoV-Receptor-Binding Domain (RBD)

MSH inhibits MERS-Spike pseudovirion (PV) entry

MSH attenuates live MERS-CoV infection in host cells

We would like to express our thanks to various laboratory members and collaborators who assisted in this study. In particular, we would like to extend our gratitude to Dina Darwis and ShiHui, Wong for their advice and suggestions on the baculoviral expression system. We also thank Professor Xinquan Wang (Tsinghua University, China) for kindly providing the plasmid vectors