key: cord-1008464-ygn9oiir
authors: Chen, Yeh; Yang, Wen-Hao; Huang, Li-Min; Wang, Yu-Chuan; Yang, Chia-Shin; Liu, Yi-Liang; Hou, Mei-Hui; Tsai, Chia-Ling; Chou, Yi-Zhen; Huang, Bao-Yue; Hung, Chian-Fang; Hung, Yu-Lin; Chen, Jin-Shing; Chiang, Yu-Ping; Cho, Der-Yang; Jeng, Long-Bin; Tsai, Chang-Hai; Hung, Mien-Chie
title: Inhibition of Severe Acute Respiratory Syndrome Coronavirus 2 main protease by tafenoquine in vitro
date: 2020-08-15
journal: bioRxiv
DOI: 10.1101/2020.08.14.250258
sha: 81a21ed61594a7ae32c4f9e497ad321f32167cea
doc_id: 1008464
cord_uid: ygn9oiir

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causing the current pandemic, coronavirus disease 2019 (COVID-19), has taken a huge toll on human lives and the global economy. Therefore, effective treatments against this disease are urgently needed. Here, we established a fluorescence resonance energy transfer (FRET)-based high-throughput screening platform to screen compound libraries to identify drugs targeting the SARS-CoV-2 main protease (Mpro), in particular those which are FDA-approved, to be used immediately to treat patients with COVID-19. Mpro has been shown to be one of the most important drug targets among SARS-related coronaviruses as impairment of Mpro blocks processing of viral polyproteins which halts viral replication in host cells. Our findings indicate that the anti-malarial drug tafenoquine (TFQ) induces significant conformational change in SARS-CoV-2 Mpro and diminishes its protease activity. Specifically, TFQ reduces the α-helical content of Mpro, which converts it into an inactive form. Moreover, TFQ greatly inhibits SARS-CoV-2 infection in cell culture system. Hence, the current study provides a mechanistic insight into the mode of action of TFQ against SARS-CoV-2 Mpro. Moreover, the low clinical toxicity of TFQ and its strong antiviral activity against SARS-CoV-2 should warrant further testing in clinical trials.

The SARS-CoV-2 was identified in December 2019 as the cause of COVID-19 outbreak that originated in Wuhan, China (1) (2) (3) . It has since spread rapidly, infected more than twelve million people globally, and caused more than 548,822 deaths (4) . Currently, there are no scientifically proven drugs to control this outbreak. The SARS-CoV-2 genome shares about 83% identity with the SARS coronavirus that emerged in 2002 and contains approximately 30,000 nucleotides that are transcribed into 14 open reading frames (Orfs) (5) . Among them, Orf1a and Orf1ab are translated into two polyproteins which are then cleaved by the main protease (M pro ), yielding a number of protein products required for viral replication and transcription (6) (7) (8) (9) . Because there are no similar proteases in humans and that M pro is necessary for viral replication, M pro is considered as an ideal target for drug design. Drugs that are approved by the U.S. Food and Drug Administration (FDA) undergo rigorous evaluation for quality, safety and effectiveness, and thus identifying FDA-approved drugs that can inhibit M pro protease activity has the advantage to be used quickly for treatment in patients. To this end, we established a fluorescence resonance energy transfer (FRET)-based high-throughput drug screening platform to rapidly identify antiviral compounds from the FDA-approved drug library which can bind to M pro and inhibit its enzymatic activity.

Recombinant SARS-CoV-2 M pro (ORF1ab polyprotein residues 3264-3569, GenBank code: QHD43415.1) was expressed in Escherichia coli and purified to homogeneity ( fig. S1 ). To rapidly identify potential FDA-approved drugs targeting SARS-CoV-2 M pro , we established a fluorescence resonance energy transfer (FRET) assay by using a protein substrate consisting of the nsp4-5 N-terminal autocleavage site (TSAVLQ↓SGFRKM) of SARS-CoV-2 M pro inserted between mTurquoise2 and mVenus, an enhanced CFP-YFP pair with higher quantum yield and protein stability ( Fig. 1A and fig. S2 ) (10, 11) . The decrease in the FRET efficiency following cleavage of the protein substrate by SARS-CoV-2 M pro and the increase in time-dependent fluorescence emission of mTurquoise2 at 474 nm were used as a measurement of M pro activity.

An initial screening of about 2,000 compounds using this FRET assay showed that tafenoquine (TFQ) exhibited the most significant inhibitory effect against SARS-CoV-2 M pro (Fig. 1B) .

TFQ (brand name Krintafel/Kozenis in U.S./Australia, owned and developed by GlaxoSmithKline) is an 8-aminoquinoline anti-malarial drug that was approved by the U.S. FDA in July 2018 and the Australian Therapeutic Goods Administration (TGA) in September 2018 for the radical cure of Plasmodium vivax (12) (13) (14) , a parasite that causes malaria. In addition, TFQ (brand name Arakoda/Kodatef in U.S./Australia, owned by 60 Degrees Pharmaceuticals) was later approved by the FDA and the TGA for malaria prophylaxis (14, 15) . However, the molecular target of TFQ is still unknown. Recently, two 4-aminoquinoline derivatives, chloroquine (CQ) and hydroxychloroquine (HCQ), were shown to be effective in inhibiting SARS-CoV-2 infection in vitro (16, 17) . Many clinical trials using CQ or HCQ to treat patients with COVID-19 have also been reported, but some have found no benefit and possible harm in patients (18, 19) . CQ is thought to inhibit virus entry by modifying glycosylation of ACE2 receptor and spike protein or by interfering with the pH-dependent endocytic pathway (20, 21) .

To further characterize TFQ, we compared the inhibitory effects of TFQ and HCQ against SARS-CoV-2 M pro at various concentrations by FRET and differential scanning fluorimetry (DSF) (22) . As shown in Figure 1c , TFQ exhibited almost 90% inhibition against SARS-CoV-2 M pro at a concentration of 90 μ M whereas HCQ did not demonstrate any significant inhibitory effects. Using a protein thermal shift assay, we showed that TFQ caused a negative shift in the melting temperature (Tm) of SARS-CoV-2 M pro in a dose-dependent manner (Fig. 1 , D and E). In contrast, HCQ had no influence on the thermal stability of SARS-CoV-2 M pro (Fig. 1 , D and F).

DSF is a powerful tool in early drug discovery (22) with the basic principle that drugs which bind to the therapeutic protein target will stabilize it and cause a positive shift in its Tm.

However, small-molecule inhibitors have been shown to cause negative shifts in the Tm values of target proteins by disrupting their oligomeric interfaces, leading to thermal destabilization and subsequent loss of interaction between the protein subunits (23, 24) . Some examples include 6hydroxy-DL-dopa binding to RAD52 (25) and SPD304 binding to TNF-α (26) . In other cases, a ligand can bind more strongly to the non-native state than the native state of its target protein, such as that of Zn 2+ and porcine growth hormone (27) . To test whether TFQ binding disrupts the dimerization interface or binds to the non-native state of SARS-CoV-2 M pro , various biophysical methods were utilized to assess the conformational changes of SARS-CoV-2 M pro . Analytical ultracentrifugation studies revealed identical sedimentation coefficient at various concentrations of TFQ, suggesting the absence of dimer-to-monomer conversion of SARS-CoV-2 M pro in the presence TFQ ( Fig. 2A) . Interestingly, results from circular dichroism (CD) spectroscopy revealed an increase in the far-UV signals (molar ellipticity at 222 nm) with increasing concentrations of TFQ, indicating that the total helical content of SARS-CoV-2 M pro decreased upon TFQ binding (Fig. 2B ). The decreased α -helical content was accompanied by reduced M pro protease activity (Fig. 2 , B and C). Together, these data suggested that TFQ may cause a local conformational change within its binding site, disrupting nearby α -helices and subsequently reducing M pro 's protease activity (Fig. 2 , B and C). Moreover, because the sedimentation coefficient remained unchanged with TFQ, it is unlikely that TFQ caused unfolding of the overall structure of SARS-CoV-2 M pro ( Fig. 2A) . To confirm this assumption, the solubility and To further probe the conformational changes of SARS-CoV-2 M pro , we performed a limited proteolysis assay by trypsin digestion (28) . The cleavage pattern indicated a greater degree of protection of SARS-CoV-2 M pro from trypsin digestion at higher concentrations of TFQ (Fig. 2E ).

In contrast, no concentrations of HCQ tested reduced the cleavage of SARS-CoV-2 M pro by trypsin digestion (Fig. 2E) . Results from binding constant measurement by isothermal titration calorimetry (ITC) indicated TFQ bound to SARS-CoV-2 M pro with micromolar affinity (Kd = ~10 -5 M, Fig. 2F ). These findings further supported the notion that TFQ binding induces local conformational changes in M pro that trigger an active-to-inactive form transition, reduce its Tm and protease activity, and render it more resistant to trypsin digestion.

To elucidate the inhibitory mechanism of TFQ against SARS-CoV-2 M pro , molecular docking was performed using SwissDock (29) . The resulting complex showed that TFQ fits well in the that TFQ occupies the sites equivalent to P2, P1 and P1′ of the N3 inhibitor in the substratebinding site (Fig. 3E) . The S1 subsite of M pro is highly specific to Gln at the P1 position of peptide substrate. The pentan-1,4-diamine moiety of TFQ appears to meet this requirement by mimicking the Gln side chain to form two hydrogen bonds with the side chain of E166 and main chain of F140 (Fig. 3 , C and E). The hydrophobic quinoline core of TFQ is in close proximity to the S2 subsite, which prefers hydrophobic residues (Fig. 3E ). In conclusion, the docking studies showed that TFQ binds to SARS-CoV-2 M pro by mimicking its preferred peptide substrate.

Since TFQ treatment inactivates SARS-CoV-2 M pro , a key protease for viral replication in of TFQ was around 2.5 μ M (Fig. 4C ). Viral infection can lead to changes in cell morphology and death of host cells, also known as cytopathic effect (CPE) (32, 33) . Vero E6 cells are susceptible to SARS-CoV-2 infection, which induces CPE (34) . We observed a significant decrease in SARS-CoV2-induced CPE in Vero E6 cells treated with 5 μ M TFQ treatment compared with the DMSO treatment group (Fig. 4D) , indicating that TFQ mitigates cell damages caused by SARS-CoV-2. Therefore, the data in Figure 4b observed on day 3 showed that there is no significant difference of viral RNA between DMSO-treated and TFQ (2.5 μ M)-treated groups because the former lacked sufficient number of surviving host cells for virus production. Collectively, these data demonstrated that TFQ potently reduces SARS-CoV-2 production in the host cells.

Drug repurposing is an efficient way to accelerate the development of therapies for COVID-

Here, we identified TFQ as a potent drug that inhibits SARS-CoV-2 replication by targeting its M pro from an FDA-approved compound library. We first demonstrated that TFQ inhibits the enzymatic activity of SARS-CoV-2 M pro by using a FRET-based assay. Subsequent molecular docking study indicated that TFQ binds directly to the substrate-binding pocket of SARS-CoV-2 M pro as a competitive inhibitor. Moreover, binding of TFQ prevented M pro from trypsin degradation and induced a negative shift in its Tm, supporting the conversion of SARS-CoV-2 M pro from an active to inactive form in the presence of TFQ. Using CD spectroscopy, we showed that increasing TFQ concentrations reduced the α -helical content of M pro , suggesting possible unraveling of some α -helices. However, the results from trypsin digestion indicated that TFQ binding rendered M pro more resistant to trypsin digestion, which indicates the presence of a more ordered structure, preventing it from trypsin digestion. In addition, the results from analytical ultracentrifugation ( Fig. 2A ) also suggested the formation of an ordered structure of the TFQ-M pro complex. Thus, inhibition of M pro by TFQ seems to convert M pro from an active to inactive conformation. This mechanism differs from the typical mechanism of action of inhibitors that bind to the active site of the enzyme to block substrate binding. Therefore, we proposed a model shown in Fig. 5 , illustrating the possible inhibition mechanism of TFQ on SARS-CoV-2 M pro .

Based on the results shown in this study, the binding of low concentration (10 to 90 μ M) of TFQ to M pro decreases its protease activity by reducing the α -helical structure content. The conformational change may destabilize SARS-CoV-2 M pro by exposing some hydrophobic residues to solvent, resulting decreased thermal stability. However, at concentration above 120 μ M TFQ, the exposed hydrophobic region of SARS-CoV-2 M pro may exceed a threshold, leading to protein aggregation and precipitation. Therefore, TFQ may inhibit the function of M pro through a two-step progressive process to reduce SARS-CoV-2 production (Fig. 5 ).

It is interesting to note that the N3 inhibitor blocked SARS-CoV-2 at a concentration of 10 μ M in cell-based assay (30) whereas TFQ exhibited strong antiviral effect at a concentration of 5 μ M in SARS-CoV-2 infected Vero E6 cells (Fig. 4) . In contrast to TFQ which can be immediately evaluated in patients with COVID-19 in clinical trials, there is currently no safety, oral bioavailability, or pharmacokinetics study of N3 inhibitor in patients.

TFQ is approved for prophylaxis and treatment of malaria in the U.S. and Australia (15, 35) . As a preventive measure, a dose of 200 mg TFQ is recommended for three days prior to traveling and 200 mg per week until one week after return. For radical cure, a single dose of 300 mg TFQ is recommended (https://wwwnc.cdc.gov/travel/news-announcements/tafenoquine-malariaprophylaxis-and-treatment). Those above dose recommendations for malaria prophylaxis suggested that TFQ at higher doses may be tolerated by the human body. Contrary to the long half-life (one month or longer) and possible severe side effects, such as bulls-eye maculopathy, dry eye, nausea, diarrhea, anemia, liver failure, and muscle paralysis, of CQ and HCQ, the halflife of TFQ is relatively shorter (~14 days) and the side effects are less severe (36, 37) . Together with our data showing that 5 μ M TFQ strongly inhibited SARS-CoV-2 infection in vitro (Fig. 4) , especially when applied as TFQ pre-treatment (full-time treatment) to mimic the prophylactic use against viral infection, the repurposing of TFQ for the prevention and treatment of COVID-19 is worth looking into for clinical evaluations.

Although FDA-approved drugs that target M pro have been identified using virtual docking methods (38, 39) , they have not been evaluated for their effects on M pro protease activity by functional assays. It is worth mentioning that among those docking-positive candidates, none of them that we tested (at least 10; Table 1 ) showed strong inhibitory effects against the M pro protease activity. Hence, the current study not only identifies the first anti-SARS-CoV-2 M pro drug that has been tested functionally to inhibit M pro protease activity and evaluated for safety (FDA-approved drug) but also provides a mechanistic insight into the mode of action of TFQ against SARS-CoV-2 M pro . The low clinical toxicity and mechanism-driven antiviral activity of TFQ against SARS-CoV-2 should warrant further testing in clinical trials.

The 

Three small-molecule compound libraries, including the FDA-approved Drug Library, Clinical

Compound Library, and Anti-COVID-19 Compound Library (MedChemExpress), were used to screen for drugs against SARS-CoV-2 M pro . 

Vero E6 (7 × 10 4 ) cells were seeded in 24-well plates and subjected to two modes of drug treatment, one in which cells were pre-treated with drugs for an hour prior to viral infection, and the other without drug pre-treatment. Cells were then infected with virus for one hour in the absence of drugs. After infection, cells were washed with PBS, and cultured with drugcontaining medium until the end of the experiment. The virus-containing supernatants were harvested at one to three days post-infection and subjected to qRT-PCR to determine the viral titers. The viral cytopathic effect (CPE) was observed under microscope and imaged at 3-day post infection.

The viral RNA in supernatant was extracted using the QIAamp Viral RNA Mini Kit (QIAGEN).

The extracted RNA was reverse transcribed using SuperScriptTM III reverse transcriptase 

Data of bar or curve graphs display as percentage or number compared to control groups with a standard deviation of two or three independent experiments. Microsoft Excel was used for statistical analyses. The two-tailed independent Student's t-test was used to compare continuous variables between two groups. All experiments were carried out at least twice. The statistical significance level of all tests is set to 0.05.

Supplementary Figures   Fig. S1 . The purification of SARS-CoV-2 M pro . 

A Novel Coronavirus from Patients with Pneumonia in China

A new coronavirus associated with human respiratory disease in China

An interactive web-based dashboard to track COVID-19 in real time

Genomic characterization of the 2019 novel human-pathogenic coronavirus isolated from a patient with atypical pneumonia after visiting Wuhan

An Overview of Severe Acute Respiratory Syndrome-Coronavirus

Protease Inhibitors: Peptidomimetics and Small Molecule Chemotherapy

Coronavirus main proteinase (3CLpro) structure: basis for design of anti-SARS drugs

Conservation of substrate specificities among coronavirus main proteases

A SARS-CoV-2 protein interaction map reveals targets for drug repurposing

A guide to choosing fluorescent proteins

Structure-guided evolution of cyan fluorescent proteins towards a quantum yield of 93%

Killing the hypnozoite--drug discovery approaches to prevent relapse in Plasmodium vivax

Tafenoquine and its potential in the treatment and relapse prevention of Plasmodium vivax malaria: the evidence to date

Clinical utility of tafenoquine in the prevention of relapse of Plasmodium vivax malaria: a review on the mode of action and emerging trial data

Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro

Hydroxychloroquine, a less toxic derivative of chloroquine, is effective in inhibiting SARS-CoV-2 infection in vitro

Breakthrough: Chloroquine phosphate has shown apparent efficacy in treatment of COVID-19 associated pneumonia in clinical studies

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

Hydroxychloroquine in patients with COVID-19: A Systematic Review and meta-analysis

High-density miniaturized thermal shift assays as a general strategy for drug discovery

Disrupting the Constitutive, Homodimeric Protein-Protein Interface in CK2beta Using a Biophysical Fragment-Based Approach

Broad Spectrum Inhibitor of Influenza A and B Viruses Targeting the Viral Nucleoprotein

Small-Molecule Disruption of RAD52 Rings as a Mechanism for Precision Medicine in BRCA-Deficient Cancers

Small-molecule inhibition of TNF-alpha

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

Limited proteolysis as a tool to probe the tertiary conformation of dysferlin and structural consequences of patient missense variant L344P

SwissDock, a protein-small molecule docking web service based on EADock DSS

Structure of M(pro) from SARS-CoV-2 and discovery of its inhibitors

Virus Isolation from the First Patient with SARS-CoV-2 in Korea

Coronaviridae and SARS-associated coronavirus strain HSR1

SARS coronavirus induces apoptosis in Vero E6 cells

Severe Acute Respiratory Syndrome Coronavirus 2 from Patient with Coronavirus Disease

Guidance for Using Tafenoquine for Prevention and Antirelapse Therapy for Malaria -United States

Pharmacokinetics of hydroxychloroquine and its clinical implications in chemoprophylaxis against malaria caused by Plasmodium vivax

Tafenoquine: A 2018 Novel FDA-Approved Prodrug for the Radical Cure of Plasmodium vivax Malaria and Prophylaxis of Malaria

Molecular docking and dynamics simulation of FDA approved drugs with the main protease from 2019 novel coronavirus

Screening of Clinically Approved and Investigation Drugs as Potential Inhibitors of SARS-CoV-2 Main Protease and Spike Receptor-Binding Domain Bound with ACE2 COVID19 Target Proteins: A Virtual Drug Repurposing Study

We would like to thank Ms. Yi-Jen Liau at National Taiwan University for technical assistance of SARS-CoV-2 preparation and SCI Pharmtech Inc. for providing HCQ. 

The materials used in this study should be requested form M.C.H. and P1′ (C-terminal to the cleavage site) and the corresponding binding subsites located on the substrate-binding pocket are named as S1 (for binding to P1), S2 (binding to P2) and S1′ (for binding to P1′). . Binding of TFQ alters nearby α -helices and to some extent reduces the proteolytic activity of SARS-CoV-2 M pro . At concentration above 120 μ M TFQ, the exposed hydrophobic region of SARS-CoV-2 M pro exceed a threshold, causing protein aggregation.

Lopinavir 32280433Ritonavir 32280433Atazanavir 32280433Indinavir 32280433Nelfinavir 32296570Darunavir 32280433Simeprevir 32280433Saquinavir 32280433Colistin 32296570 Table 1 . List of docking-positive FDA-approved compounds against SARS-CoV-2 M pro .