key: cord-0979247-hj7afmdl
authors: Lin, ChangDong; Li, Yue; Yuan, MengYa; Huang, MengWen; Liu, Cui; Du, Hui; Pan, XingChao; Wen, YaTing; Xu, Xinyi; Xu, Chenqi; Chen, JianFeng
title: Ceftazidime Is a Potential Drug to Inhibit SARS-CoV-2 Infection In Vitro by Blocking Spike Protein-ACE2 Interaction
date: 2020-09-15
journal: bioRxiv
DOI: 10.1101/2020.09.14.295956
sha: c6d6a13606ac6eaff782ee20ad0dfe9e7654d525
doc_id: 979247
cord_uid: hj7afmdl

Coronavirus Disease 2019 (COVID-19) spreads globally as a sever pandemic, which is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Cell entry of SARS-CoV-2 mainly depends on binding of the viral spike (S) proteins to angiotensin converting enzyme 2 (ACE2) on host cells. Therefore, repurposing of known drugs to inhibit S protein-ACE2 interaction could be a quick way to develop effective therapy for COVID-19. Using a high-throughput screening system to investigate the interaction between spike receptor binding domain (S-RBD) and ACE2 extracellular domain, we screened 3581 FDA-approved drugs and natural small molecules and identified ceftazidime as a potent compound to inhibit S-RBD–ACE2 interaction by binding to S-RBD. In addition to significantly inhibit S-RBD binding to HPAEpiC cells, ceftazidime efficiently prevented SARS-CoV-2 pseudovirus to infect ACE2-expressing 293T cells. The inhibitory concentration (IC50) was 113.2 μM, which is far below the blood concentration (over 300 μM) of ceftazidime in patients when clinically treated with recommended dose. Notably, ceftazidime is a drug clinically used for the treatment of pneumonia with minimal side effects compared with other antiviral drugs. Thus, ceftazidime has both anti-bacterial and anti-SARS-CoV-2 effects, which should be the first-line antibiotics used for the clinical treatment of COVID-19.

screened 3581 FDA-approved drugs and natural small molecules and identified 23 ceftazidime as a potent compound to inhibit S-RBD-ACE2 interaction by binding to S-24 RBD. In addition to significantly inhibit S-RBD binding to HPAEpiC cells, ceftazidime 25 efficiently prevented SARS-CoV-2 pseudovirus to infect ACE2-expressing 293T cells. 26

The inhibitory concentration (IC50) was 113.2 μM, which is far below the blood 27 concentration (over 300 μM) of ceftazidime in patients when clinically treated with 28 recommended dose. Notably, ceftazidime is a drug clinically used for the treatment of 29 pneumonia with minimal side effects compared with other antiviral drugs. Thus, 30 ceftazidime has both anti-bacterial and anti-SARS-CoV-2 effects, which should be the receptor recognition and membrane fusion. The N-terminal region of its S1 domain 49 contains receptor binding domain (RBD), which directly binds to angiotensin converting 50 enzyme 2 (ACE2) receptor on the plasma membrane of host cells and is responsible for 51 virus attachment 8,9 . Therefore, blocking the binding of spike protein to ACE2 is an 52 effective way to inhibit the infection of target cells by SARS-CoV-2. By now, several 53 studies have reported the development of monoclonal antibodies targeting spike protein 54 10,11 , however, the typical timeline for approval of a novel antibody for the management 55 viral infection is long. In addition, the side effects such as antibody-dependent 56 enhancement of viral infection need to be considered 11-13 , and the high cost of antibody 57 treatment will limit the clinical application. Therefore, repurposing of known small 58 molecule drugs to inhibit spike protein and ACE2 binding could significantly accelerate 59 the deployment of effective and affordable therapies for In this study, we expressed and purified Spike-RBD (S-RBD) and the extracellular 61 domain of ACE2 (ACE2-ECD) and then established an AlphaScreen-based high- 

In order to screen small molecules that block S protein-ACE2 binding, we firstly 75 established an AlphaScreen-based high-throughput system to detect the interaction 76 between S-RBD and ACE2-ECD (Fig. 1a ). S-RBD and ACE2-ECD were expressed in 77 293T cells and then purified. Biotinylated ACE2-ECD (ACE2-ECD-Biotin) binds to 78 streptavidin-coated Alpha donor beads and His-tagged S-RBD (S-RBD-His) binds to anti-79

His-conjugated AlphaLISA acceptor beads. When S-RBD binds to ACE2-ECD, the two 80 beads come into close proximity. Upon illumination at 680 nm, the donor beads generate 81 singlet oxygen molecules that diffuse to acceptor beads and transfer energy to thioxene 82 derivatives in the acceptor beads resulting in light emission at 520-620 nm. The results 83

showed that the incubation of ACE2-ECD-Biotin with S-RBD-His produced very strong 84

AlphaScreen signal, and the signal decreased to the basal level in the absence of either of 85 the two proteins (Fig. 1b) . To confirm the specificity of this AlphaScreen system, we 86 replaced S-RBD-His with His-tagged extracellular domains of other membrane proteins, 87

including mucosal vascular addressin cell adhesion molecule 1 (MAdCAM-1) and vascular 88 cell adhesion molecule 1 (VCAM-1). Co-incubation of MAdCAM-1-His or VCAM-1-His 89 with ACE2-ECD-Biotin did not generate AlphaScreen signal, indicating that the system 90 detects S-RBD-ACE2 interaction specifically (Fig. 1c) . 91

Next, we used the AlphaScreen-based high-throughput system to screen small molecules 93 that block S-RBD-ACE2 interaction. A total of 3581 small molecule compounds with 94 known molecular structures from FDA Approved Drug Library, Spectrum Collection and 95

Targetmol-Natural Compound Library were assessed (Fig. 1d) . The assay was conducted 96 at a final compound concentration of 10 μM and the interaction between S-RBD-His (0.1 97 μM) and ACE2-ECD-Biotin (0.2 μM) was analyzed. After first round screening, 75 98 candidate compounds were identified, which showed inhibitory effect on S-RBD-ACE2 99 interaction (Fig. 1d ). All these compounds showed over 45% inhibition rate according to 100 the changes in AlphaScreen signal. To exclude the interference of the compounds to the 101

AlphaScreen system per se, we designed a negative selection system in which the 102 biotinylated S-RBD-His links streptavidin-coated Alpha donor bead and anti-His-103 conjugated AlphaLISA acceptor bead together to generate AlphaScreen signal directly (Fig.  104 1e). After the negative selection, 10 compounds, including bleomycin sulfate, levodopa, 105 norepinephrine, trientine hydrochloride, ceftazidime, chiniofon, hematein, theaflavin, 106 bleomycin and myricetin, from the 75 candidate compounds were validated to inhibit S-107 RBD-ACE2 interaction effectively. Among the 10 compounds, ceftazidime was the most 108 potent inhibitor which showed a relative inhibition rate of 80.7% (Fig. 1f) . Thus, 109 ceftazidime was selected for further investigation considering the best inhibitory effect on 110 S-RBD-ACE2 interaction, the anti-inflammatory effect and the minimal side effect of 111 this drug compared with the other 9 compounds 14-17 (Fig. 1g) . AlphaScreen-based high-throughput system and 75 candidates were identified from 3581 121 compounds in positive selection. The inhibition rate was calculated by the decrease of 122

AlphaScreen signal of each compound compared with that of DMSO vehicle control group. 123 e, Schematic diagram of negative selection using AlphaScreen system. Biotinylated S-124 RBD-His simultaneously links streptavidin-coated donor bead and anti-His-conjugated 125

acceptor bead together to generate AlphaScreen signal directly. f, Relative inhibition of 10 126 candidate compounds on S-RBD-ACE2 interaction using AlphaScreen system. The 127

relative inhibition rate was calculated by subtracting the inhibition rate in negative 128 selection from that in positive selection. g, Molecular structure of ceftazidime. Data 129

represent the mean ± SEM (n ≥ 2) in b, c and f. *** p < 0.001, ns: not significant (Student's 130 t test). 131 132

Ceftazidime specifically binds to S-RBD protein 133

To investigate whether S-RBD or ACE2 is the binding target protein of ceftazidime, we 134 applied a bio-layer interferometry experiment to examine the binding affinity between 135 ceftazidime and S-RBD or ACE2-ECD. Along with the elevated concentration of 136 ceftazidime, this compound showed increased binding to S-RBD protein with an KD value 137 of 260 ± 38 μM (Fig. 2a) . Notably, ceftazidime hardly dissociated from S-RBD, indicating 138 a strong and stable interaction between ceftazidime and S-RBD. By contrast, ceftazidime 139 and ACE2-ECD showed no specific binding signal (Fig. 2b) . Thus, ceftazidime binds to S-140 RBD specifically. 141 142

Octet RED96 instrument. The biotin-conjugated S-RBD or ACE2-ECD was captured by 145 streptavidin that was immobilized on a biosensor and tested for binding with gradient 146 concentrations of ceftazidime. 147 148

Ceftazidime inhibits S-RBD binding to human pulmonary alveolar epithelial cells 149

Lung is the main organ infected by SARS-CoV-2, which cause severe acute respiratory 150 syndrome (SARS) 18,19 . Therefore, we examined the inhibitory effect of ceftazidime on the 151 binding of S-RBD protein to human pulmonary alveolar epithelial cells (HPAEpiC) which 152 express ACE2. Addition of 100 μM ceftazidime into the soluble S-RBD binding assay 153 system led to a significantly decrease in the S-RBD binding signal (Fig. 3a) , demonstrating 154 the efficient inhibition on S-RBD binding to HPAEpiC cells by ceftazidime. Further 155 analysis showed an IC50 of 39.90 ± 1.11 μM (Fig. 3b) . in vitro and the IC50 was 113.24 ± 1.23 μM (Fig. 4) . Of note, ceftazidime showed no 176 detectable cytotoxicity at a high concentration of 400 μM, indicating its safety for clinical 177 usage (Fig. 4) Fig. 1) . These results in 214 combination with a preliminary Structure Activity Relationship (SAR) analysis suggested 215 that the unique moieties in ceftazidime, including 2-aminothiazole, oxime protected with a 216 terminal-exposed isobutyric acid and the positive charged pyridine, might be involved in 217 mediating the binding to S-RBD and eventually blocked the protein interaction between S-218 RBD and ACE2. Furthermore, our data showed that ceftazidime hardly dissociated from 219 the S-RBD proteins (Fig. 2a) , which could be due to the covalent binding of ceftazidime to 220 S-RBD. 

Extended Data Fig. 1 | Effect of ceftazidime and the derivatives of cephalosporins on 223

The 260 coronavirus pandemic in five powerful charts

Novel human coronavirus (SARS-CoV-2): A lesson from 263 animal coronaviruses

Zoonotic origins of human coronaviruses

Assessing the application of a pseudovirus system for emerging 268 SARS-CoV-2 and re-emerging avian influenza virus H5 subtypes in vaccine 269 development

Inhibition of SARS-CoV-2 (previously 2019-nCoV) infection by a 271 highly potent pan-coronavirus fusion inhibitor targeting its spike protein that 272 harbors a high capacity to mediate membrane fusion

Cryo-EM structure of the 2019-nCoV spike in the prefusion 275 conformation

Cell entry mechanisms of SARS-CoV-2

Antigenic 279 modules in the N-terminal S1 region of the transmissible gastroenteritis virus spike 280 protein

Structural and Functional Basis of SARS-CoV-2 Entry by Using 282

Human monoclonal antibodies block the binding of SARS-CoV-2 284 spike protein to angiotensin converting enzyme 2 receptor

Potent Neutralizing Antibodies against SARS-CoV-2 Identified by 287

Single-Cell Sequencing of Convalescent Patients' B Cells

Implications of antibody-dependent enhancement of infection 290 for SARS-CoV-2 countermeasures

Viral-Induced Enhanced Disease 293 Illness

Endotoxin neutralization and anti-inflammatory effects of 295 tobramycin and ceftazidime in porcine endotoxin shock

Anti-inflammatory effects of the antibiotics ceftazidime and 298 tobramycin in porcine endotoxin shock: are they really anti-inflammatory?

Short-term side effects and patient-reported outcomes of 301 bleomycin sclerotherapy in vascular malformations

Side effects induced by the acute levodopa challenge in Parkinson's 304 Disease and atypical parkinsonisms

New understanding of the damage of SARS-CoV-2 infection 307 outside the respiratory system

SARS-CoV-2 and viral sepsis: observations and hypotheses

Characterization of spike glycoprotein of SARS-CoV-2 on virus entry 312 and its immune cross-reactivity with SARS-CoV

SARS-CoV-2 pseudoviruses were produced as previously described 20 . The pseudoviruses 341 were diluted in complete DMEM mixed with an equal volume (50 μl) of diluted DMSO or 342 ceftazidime, and then incubated at 37 °C for 1 h. The mixture was transferred to 293T 343 expressing human ACE2 stable cell line cells. The cells were incubated at 37 °C for 48 h, 344 followed by lysed with Bio-Lite Luciferase Assay Buffer and tested for luciferase activity 345 (Vazyme). The percent neutralization was calculated by comparing the luciferase value of 346 ceftazidime treatment group to that of DMSO control. 347

The BLI experiment was performed using an Octet Red 96 instrument (ForteBio, Inc.).