key: cord-253438-k8iqv1jb
authors: Li, Yujun; Wang, Haimin; Tang, Xiaojuan; Fang, Shisong; Ma, Danting; Du, Chengzhi; Wang, Yifei; Pan, Hong; Yao, Weitong; Zhang, Renli; Zou, Xuan; Zheng, Jie; Xu, Liangde; Farzan, Michael; Zhong, Guocai
title: SARS-CoV-2 and Three Related Coronaviruses Utilize Multiple ACE2 Orthologs and Are Potently Blocked by an Improved ACE2-Ig
date: 2020-10-27
journal: J Virol
DOI: 10.1128/jvi.01283-20
sha: 
doc_id: 253438
cord_uid: k8iqv1jb

The ongoing coronavirus disease 2019 (COVID-19) pandemic has caused >20 million infections and >750,000 deaths. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the etiological agent of COVID-19, has been found closely related to the bat coronavirus strain RaTG13 (Bat-CoV RaTG13) and a recently identified pangolin coronavirus (Pangolin-CoV-2020). Here, we first investigated the ability of SARS-CoV-2 and three related coronaviruses to utilize animal orthologs of angiotensin-converting enzyme 2 (ACE2) for cell entry. We found that ACE2 orthologs of a wide range of domestic and wild mammals, including camels, cattle, horses, goats, sheep, cats, rabbits, and pangolins, were able to support cell entry of SARS-CoV-2, suggesting that these species might be able to harbor and spread this virus. In addition, the pangolin and bat coronaviruses, Pangolin-CoV-2020 and Bat-CoV RaTG13, were also found able to utilize human ACE2 and a number of animal-ACE2 orthologs for cell entry, indicating risks of spillover of these viruses into humans in the future. We then developed potently anticoronavirus ACE2-Ig proteins that are broadly effective against the four distinct coronaviruses. In particular, through truncating ACE2 at its residue 740 but not 615, introducing a D30E mutation, and adopting an antibody-like tetrameric-ACE2 configuration, we generated an ACE2-Ig variant that neutralizes SARS-CoV-2 at picomolar range. These data demonstrate that the improved ACE2-Ig variants developed in this study could potentially be developed to protect from SARS-CoV-2 and some other SARS-like viruses that might spillover into humans in the future. IMPORTANCE The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the etiological agent of the currently uncontrolled coronavirus disease 2019 (COVID-19) pandemic. It is important to study the host range of SARS-CoV-2, because some domestic species might harbor the virus and transmit it back to humans. In addition, insight into the ability of SARS-CoV-2 and SARS-like viruses to utilize animal orthologs of the SARS-CoV-2 receptor ACE2 might provide structural insight into improving ACE2-based viral entry inhibitors. In this study, we found that ACE2 orthologs of a wide range of domestic and wild animals can support cell entry of SARS-CoV-2 and three related coronaviruses, providing insights into identifying animal hosts of these viruses. We also developed recombinant ACE2-Ig proteins that are able to potently block these viral infections, providing a promising approach to developing antiviral proteins broadly effective against these distinct coronaviruses.

CoV-2 WHU01, Pangolin-CoV-2020, Bat-CoV RaTG13, and SARS-CoV BJ01. Purified RBD proteins were then used to perform surface staining of 293T cells transfected with each of the 16 ACE2 orthologs or a vector plasmid control (Fig. 2) . All of the RBD proteins showed binding to a number of ACE2 orthologs. Unexpectedly, although the SARS-FIG 1 SARS-CoV-2 and ACE2 contact residues are conserved among four SARS-like viruses and 16 ACE2 orthologs, respectively. (A) Interactions between the SARS-CoV-2 receptor binding domain (RBD, red) and ACE2 (blue) involve a large number of contact residues (PDB accession no. 6M0J). RBD residues Ͻ5 Å from ACE2 atoms and ACE2 residues Ͻ5 Å from RBD atoms are shown. (B) The sequences of the SARS-CoV-2 WHU01, a pangolin coronavirus identified in Manis javanica (Pangolin-CoV-2020), a bat coronavirus identified in R. affinis (Bat-CoV RaTG13), and the SARS-CoV BJ01 are aligned, with residues different from the corresponding ones in SARS-CoV-2 highlighted in blue. The stars indicate RBD residues Ͻ5 Å from ACE2 atoms. The yellow lines indicate the RBM region. N-linked glycosylation motifs are indicated in green. (C) Sequences of ACE2 orthologs from the 16 indicated species are aligned, with only residues Ͻ5 Å from RBD atoms shown here. The numbering is based on human ACE2 protein, and the residues different from the corresponding ones in human ACE2 are highlighted in blue. CoV-2 and SARS-CoV RBDs differ significantly in the RBM region and ACE2-contact residues, both RBDs bound to 11 ACE2 orthologs with 10 identical ones. Moreover, the Pangolin-CoV-2020 RBD, which differs from SARS-CoV-2 RBD with only one amino acid within the RBM region, kept nine SARS-CoV-2 RBD-interacting ACE2 orthologs and gained three additional interacting ones, including that of rat, mouse, and chicken. The RBD of Bat-CoV RaTG13 then showed a binding profile significantly different and narrower than the other three RBDs. Note that human ACE2 and ACE2 orthologs of some domestic animals, including camels, cattle, horses, goats, sheep, cats, and rabbits, support efficient binding to all the four tested RBDs, suggesting that these ACE2 orthologs might be generally functional for supporting cell entry of the four tested viruses. A wide range of ACE2 orthologs can support entry of the four coronaviruses. To evaluate spike protein-mediated entry of these coronaviruses, we generated retrovirus-based luciferase reporter pseudoviral particles (pp) enveloped with one of six different spike proteins, including a wild-type SARS-CoV-2 spike (SARS-CoV-2 WHU01 FIG 2 A wide range of ACE2 orthologs support binding to RBD proteins of SARS-CoV-2 and three related coronaviruses. (A) 293T cells were transfected with adjusted amounts of the indicated ACE2-ortholog plasmids to have similar expression levels of the ACE2 ortholog proteins. Cells were then stained with an RBD-mouse IgG2 Fc fusion protein of SARS-CoV-2 WHU01, Pangolin-CoV-2020, Bat-CoV RaTG13, or SARS-CoV BJ01, followed by staining with an Alexa 488-goat anti-mouse IgG secondary antibody. RBD-ACE2 binding was detected using flow cytometry. (B) Percentages of cells positive for RBD binding in panel A are presented as a heatmap according to the indicated color code. (C) Expression levels of the indicated ACE2 orthologs were detected using Western blotting. The data shown are representative of two independent experiments performed by two different people with similar results. pp), a furin site deletion mutant of SARS-CoV-2 spike (SARS-CoV-2 Δfurin pp), a wild-type SARS-CoV spike (SARS-CoV BJ01 pp), a SARS-CoV spike carrying the Pangolin-CoV-2020 RBD (Pangolin-RBD/BJ01 pp), a wild-type Bat-CoV spike (Bat-CoV RaTG13 pp), and a Bat-CoV spike carrying the Pangolin-CoV-2020 RBD (Pangolin-RBD/RaTG13 pp). These reporter pseudoviruses were used to infect 293T cells expressing each of the 16 ACE2 orthologs. A vesicular stomatitis virus protein G (VSV-G)-pseudotyped reporter retrovirus whose entry is independent of ACE2 was used as a control virus. As expected, all the orthologs that supported RBD binding were also functional on supporting pseudovirus infection. Again, ACE2 orthologs of humans and most domestic mammals, including camels, cattle, horses, goats, sheep, cats, and rabbits, supported entry of all the tested pseudoviruses ( Fig. 3A to G). It is of note that, although furin-cleaved and uncleaved SARS-CoV-2 spike trimers have significant structural difference (23) , infection with SARS-CoV-2 Δfurin pseudovirus produced stronger reporter signals but an almost identical pattern of ACE2-ortholog usage as the wild-type pseudovirus ( Fig. 3A and B) . These data are consistent with the findings that both furin-cleaved and uncleaved SARS-CoV-2 spike trimers can adopt an ACE2-binding-competent conformation albeit at different frequency (17 to 50%), and furin cleavage reduces overall stability of the spike protein (23) (24) (25) . The ability of ACE2 orthologs to support SARS-CoV-2 infection was further confirmed by infection assays using SARS-CoV-2 live virus (Fig. 3H) . These data indicate that humans and these domestic animals might be generally susceptible to infections of the four distinct coronaviruses.

ACE2-Ig variants that have soluble ACE2 domain truncated at residue 740 but not 615 potently block SARS-CoV-2 entry. Recombinant RBD and soluble ACE2 proteins have been shown to potently block SARS-CoV entry (26, 27) . To investigate whether similar approaches could also be applied to SARS-CoV-2, we first produced mouse IgG2a Fc fusion proteins of RBD (RBD-Ig) and soluble ACE2 (ACE2-Ig) variants (Fig. 4A) . Specifically, the RBD variants include wild-type RBDs of SARS-CoV, Pangolin-CoV-2020, and SARS-CoV-2, and four mutants of SARS-CoV-2 RBD that were expected to bind ACE2 better via additional possible aromatic-stacking (F486W, Y505W) or salt-bridge (K417R, G496D) interactions. The ectodomain of cell-surface ACE2 spanning its residues 18 to 740 contains an enzymatic domain (18-615) and a Collectrin-like domain (CLD). Previous crystal-structure studies showed that soluble ACE2 protein truncated at its residue 615, preceding the CLD domain, express well and forms stable complex with the RBD of SARS-CoV and SARS-CoV-2, respectively (21, 22) . Therefore, the ACE2-Ig variants include human ACE2 truncated at its residue 615 (615-wt) and 740 (740-wt), respectively, and mutants of the 615-and 740-version ACE2-Ig proteins that were expected to inactivate ACE2s' enzymatic activity (NN) or bind SARS-CoV-2 RBD better via additional possible hydrophobic (Y83W, H34Y, M82K) or salt-bridge (D30E) interactions. We then evaluated these proteins for their potency of blocking SARS-CoV-2 Δfurin pseudovirus infection ( Fig. 4B and C) . Among all the RBD-Ig variants, the Y505W mutant of SARS-CoV-2 RBD showed modestly improved potency over the wild type, and wild-type RBD of Pangolin-CoV-2020 showed the best neutralization activity among all the tested RBDs (Fig. 4B ). Among the tested ACE2-Ig variants, interestingly, all the 740-version variants showed significantly better potency than the 615-version variants (two-tailed two-sample t test, P Ͻ 0.001; Fig. 4C ). In addition, D30E mutants of both 615-and 740-version proteins showed improved potency over the corresponding wild types (two-sample t test, P Ͻ 0.01), and the 740-D30E variant outperformed all the RBD-Ig and ACE2-Ig variants.

The 740-D30E variant of ACE2-Ig is a broadly neutralizing immunoadhesin. We further tested the 740-wt and 740-D30E variants of ACE2-Ig for their neutralization activities against SARS-CoV-2, SARS-CoV, Pangolin-CoV-2020, and Bat-CoV RaTG13 pseudotypes (Fig. 5) . Interestingly, the D30E mutation improved the protein's neutralization activity against SARS-CoV-2 and Bat-CoV RaTG13 pseudoviruses (two-sample t test, P Ͻ 0.01; Fig. 5A and C) but not Pangolin-RBD/BJ01 or SARS-CoV pseudovirus (two-sample t test, P Ͼ 0.05; Fig. 5B and D). The D30 residue of human ACE2 was consistently found to form a salt-bridge interaction with the K417 residue of SARS-CoV-2 RBD in multiple released structures of SARS-CoV-2 RBD in complex with ACE2 (20, 21) (Fig. 5E ). The residue 417 of Bat-CoV RaTG13 is also a lysine, while the same residue is an arginine for Pangolin-CoV-2020 RBD and a valine for SARS-CoV RBD. Thus, the mechanism of the D30E-mediated improvement is likely that the mutation enhances the salt-bridge interaction between the residue 30 of the ACE2 and residue 417 of SARS-CoV-2 and Bat-CoV RaTG13 RBDs. These data suggest that the 740-D30E variant of ACE2-Ig is a broadly neutralizing immunoadhesin against SARS-CoV-2, SARS-CoV, Pangolin-CoV-2020, and Bat-CoV RaTG13.

An antibody-like ACE2-Ig variant further improves neutralization potency by ϳ10-fold. To obtain a more potent ACE2-Ig that may serve as an anti-SARS-CoV-2 drug candidate, we generated more variants and used human IgG1 domains to replace the mouse counterparts in the ACE2-Ig fusion proteins. Two new variants, which have an antibody like configuration and contain four or six soluble ACE2 domains in a single molecule, are named as ACE2-Ig-v3 and ACE2-Ig-v4, respectively (Fig. 6A ). We first tested these variants for their neutralization potency against SARS-CoV-2 pseudovirus. Both of the new variants showed pronounced improvements over ACE2-Ig-v1.1, the original 740-D30E dimer variant ( Fig. 6A to C). Specifically, ACE2-Ig-v3 and ACE2-Ig-v4 have estimated 50% inhibitory concentration (IC 50 ) values of 40 to 70 pM and IC 90 values of 250 to 730 pM, representing Ͼ10-fold improvement on IC 50 and Ͼ25-fold improvement on IC 90 compared to those of ACE2-Ig-v1.1 ( Fig. 6B and C) . Because the ACE2-Ig-v4 configuration significantly impairs protein yield during production (data not shown), we chose to proceed with ACE2-Ig-v3 to test its neutralization potency against SARS-CoV-2 live virus (Fig. 6D) . Consistent with the pseudovirus neutralization data, ACE2-Ig-v3 at 0.16 g/ml showed a more potent inhibition of SARS-CoV-2 live virus infection than ACE2-Ig-v1 at 0.8 g/ml. Moreover, ACE2-Ig-v3 at 0.8 g/ml (1.85 nM) already completely abolished viral nucleocapsid protein (NP) immunofluorescent signal. These data demonstrate that ACE2-Ig-v3 is a markedly improved ACE2-Ig variant as a potent entry inhibitor against SARS-CoV-2 virus.

In this study, we investigated the ability of SARS-CoV-2 to utilize animal orthologs of ACE2 for cell entry. We observed that ACE2 orthologs of a wide range of domestic animals, including camels, cattle, horses, goats, sheep, cats, and rabbits, efficiently supported binding and entry of this virus, suggesting that these domestic mammals might be susceptible to this viral infection ( Fig. 2B and Fig. 3G ). Consistent with this, during preparation of the manuscript, two studies independently reported laboratory ). SARS-CoV-2 and RaTG13 have a K417 residue at their spike proteins, while Pangolin-CoV has an R417 residue and SARS-CoV has a V417 residue at their spike proteins, respectively. Thus, a stabilized salt bridge interaction between E30 of the ACE2-Ig protein and K417 of the virus spike protein is likely responsible for the D30E mutation-mediated neutralization enhancement. The data shown are representative of two or three experiments independently performed by two different people with similar results, and data points in panels A to D represent the means Ϯ the SD of three or four biological replicates. and natural infection of cats by SARS-CoV-2 (28, 29) . Therefore, it is necessary to further investigate the susceptibility of these domestic animals to SARS-CoV-2. In addition to the species investigated in this study, farmed mink has been found susceptible to SARS-CoV-2 and able to transmit the virus back to humans (30, 31) . The currently available information therefore suggests that surveillance of livestock and farmed mammals in marketplaces for SARS-CoV-2 infection might be necessary.

Pangolins have been proposed as potential intermediate or natural hosts of SARS-CoV-2 (11) (12) (13) . It has been proposed that SARS-CoV-2 might originate from recombination of a Pangolin-CoV-like coronavirus and a RaTG13-like coronavirus (13, 32) . Thus, the intermediate host of SARS-CoV-2 should also be susceptible to both of the "parental viruses." Here, we found that pangolin (Manis javanica) ACE2 does not support binding or entry of Bat-CoV RaTG13, a coronavirus known to have the highest genome sequence identity (96.2%) to SARS-CoV-2 so far, suggesting a lower likelihood of Malayan pangolins (Manis javanica) being intermediate hosts for SARS-CoV-2. On the other hand, we also found that SARS-CoV-2, SARS-CoV, and Pangolin-CoV-2020 (11) can all efficiently utilize pangolin (Manis javanica) ACE2 for cellular binding and entry, supporting the hypothesis that pangolins might be natural hosts of SARS-CoV-like coronaviruses (11, 12) . It is also noteworthy that Pangolin-CoV-2020 can also efficiently utilize human ACE2, as well as a wide range of domestic-and wild-animal ACE2 orthologs for cell entry (Fig. 2B and 3G) , indicating that this virus has very broad host range and high risk of spillover into human population in the future.

Animal models are essential for preclinical evaluation of efficacy and potential toxicity of candidate prophylactic vaccines or therapeutics for COVID-19, as well as for studying the transmission, pathogenesis, and immunology of this disease. In this study, rabbit ACE2 was found to efficiently support binding and entry of all the four coronaviruses ( Fig. 2B and 3G ). It is therefore worth exploring whether rabbit, a commonly used laboratory species, could serve as a common model animal for studying COVID-19, SARS, and diseases caused by other SARS-related coronaviruses. Adaptation of viruses to infect mice is another way of developing small animal models of COVID-19. We found in this study that Pangolin-CoV-2020, whose RBM only differs from that of SARS-CoV-2 with one amino acid, could efficiently utilize mouse ACE2 for binding and cell entry (Fig. 1B, 2B, and 3G) . These data suggest that mice might be susceptible to this viral infection, and thus this pangolin coronavirus could be used as a surrogate to SARS-CoV-2 for in vivo studies in wild-type mice.

Effective vaccines or targeted therapeutics against SARS-CoV-2 infections are not yet available (15) . ACE2-Ig that has soluble ACE2 truncated at its residue 615 had been proposed as a candidate therapeutic for SARS-CoV-2 infection in the beginning of the COVID-19 pandemic (33) . Two recently posted experimental studies by Lui et al. (34) and Case et al. (35) have also investigated the use of this 615-version ACE2-Ig to block SARS-CoV-2 infection and showed IC 50 values of ϳ1 and 29 g/ml, respectively. Consistent with these studies, we got in our study an estimated IC 50 value of ϳ4 g/ml for this variant (615-wt in Fig. 4C ). It is noteworthy that our studies here have identified three key improvements over the 615-version ACE2-Ig. First, the 740-wt ACE2-Ig variant showed a Ͼ20-fold potency improvement over the 615-wt variant, and all the 740version variants showed markedly enhanced neutralization potency over the 615version variants (Fig. 4C) . The reason for this improvement is not very clear yet, but it is possible that the CLD domain, included in the 740 version, stabilizes an orientation of the ACE2 enzymatic domain favorable to S-protein binding, or that the CLD has an independent anti-SARS-CoV-2 activity. Second, likely because SARS-CoV-2 has adapted in animals (e.g., pangolins) whose ACE2 orthologs have a glutamic acid at position 30, the D30E mutation further improves the 740-version ACE2-Ig's neutralization potency against SARS-CoV-2. Moreover, the D30E mutation also enables the protein to neutralize all the four distinct SARS-like coronaviruses (Fig. 5) . Third, by utilizing an antibodylike structure to build an ACE2 tetramer, we generated a variant ACE2-Ig-v3 that has at least 10-fold additional improvement on neutralization potency against SARS-CoV-2 pseudotype as well as live virus (Fig. 6 ). Through these changes, we have therefore successfully improved the originally very modest immunoadhesin inhibitor (IC 50 Ϸ 18 nM) to be a very potent entry inhibitor (IC 50 Ϸ 60 pM) against SARS-CoV-2.

Recently, a clinical-grade recombinant soluble ACE2 protein has been shown to block SARS-CoV-2 infection in engineered human organoid (36) . A phase 2 clinical trial to investigate this protein, injected twice daily, as a treatment for COVID-19 patients has already started recruiting patients in multiple countries (https://clinicaltrials.gov/ct2/ show/NCT04335136). Considering that fusion with IgG Fc improves soluble ACE2 protein's half-life in mice from Ͻ2 h to over a week (37, 38) and that we have markedly improved the neutralization potency of ACE2-Ig, we therefore optimistically expect that the improved ACE2-Ig variants described in this study could potentially be developed to provide effective protections from SARS-CoV-2 and other SARS-like viruses that might spillover into humans in the future, as well as SARS-CoV-2 variants that emerge over the course of the current pandemic. Cells. 293T cells and Vero cells were kindly provided by Stem Cell Bank, Chinese Academy of Sciences, confirmed mycoplasma-free by the provider, and maintained in Dulbecco modified Eagle medium (DMEM; Life Technologies) at 37°C in a 5% CO 2 -humidified incubator. Growth medium was supplemented with 2 mM GlutaMAX-I (Gibco, catalog no. 35050061), 100 M nonessential amino acids (Gibco, catalog no. 11140050), 100 U/ml penicillin and 100 g/ml streptomycin (Gibco, catalog no. 15140122), and 10% FBS (Gibco, catalog no. 10099141C). 293T-based stable cells expressing human ACE2 were maintained under the same culture condition as 293T, except that 3 g/ml of puromycin was added to the growth medium. 293F cells for recombinant protein production were generously provided by Yu J. Cao (School of Chemical Biology and Biotechnology, Peking University Shenzhen Graduate School) and maintained in SMM 293-TII serum-free medium (Sino Biological, catalog no. M293TII) at 37°C, 8% CO 2 , in a shaker incubator at 125 rpm.

Plasmids. DNA fragments encoding spike proteins of SARS-CoV-2 WHU01 (GenBank accession no. MN988668.1), SARS-CoV BJ01 (GenBank AY278488.2), Pangolin-CoV (National Genomics Data Center GWHABKW00000000; https://bigd.big.ac.cn/search/?dbIdϭgwh&qϭGWHABKW00000000&pageϭ1) (11) , and Bat-CoV RaTG13 (GenBank MN996532) (8), were synthesized by the Beijing Genomic Institute (BGI, China) and Sangon Biotech (Shanghai, China) and then cloned into pcDNA3.1(ϩ) plasmid or pCAGGS plasmid between EcoRI and XhoI restriction sites. Plasmids encoding recombinant RBD and soluble ACE2 variants were generated by cloning each of the gene fragments into a pCAGGS-based mouse-IgG2a or human IgG1 Fc fusion protein expression plasmid between NotI and BspEI sites. The retroviral reporter plasmids encoding a Gaussia luciferase or a green fluorescent protein (GFP) reporter gene were constructed by cloning the reporter genes into pQCXIP plasmid (Clontech), respectively. DNA fragments encoding C-terminally S-tagged ACE2 orthologs were synthesized in pUC57 backbone plasmid by Sangon Biotech (Shanghai, China). These fragments were then cloned into pQCXIP plasmid (Clontech) between SbfI and NotI restriction sites.

IgG Fc fusion protein production and purification. 293F cells at the density of 6 ϫ 10 5 cells/ml were seeded into 100 ml of SMM 293-TII serum-free medium (Sino Biological, catalog no. M293TII) 1 day before transfection. The cells were then transfected with 100 g of plasmid in complex with 250 g of PEI MAX 4000 (Polysciences, Inc., catalog no. 24765-1). Cell culture supernatants were collected at 48 to 72 h posttransfection. Recombinant Fc fusion proteins are purified using protein A-Sepharose CL-4B (GE Healthcare, catalog no. 17-0780-01), eluted with 0.1 M citric acid at pH 4.5, and neutralized with 1 M Tris-HCl at pH 9.0. Buffers were then exchanged to phosphate-buffered saline (PBS), and proteins were concentrated by 30-kDa cutoff Amicon Ultra-15 centrifugal filter units (Millipore, catalog no. UFC903096).

Flow cytometry for detecting interactions of RBD-Ig proteins with cell surface ACE2 orthologs. 293T cells were seeded at 20% density in 48-well plates at 12 to 15 h before transfection. Cells in each well were then transfected with 0.5 l of Lipofectamine 2000 (Life Technologies, catalog no. 11668019) in complex with 200 ng of plasmid encoding one of the 16 ACE2 orthologs or a D30E mutant of the human ACE2. Culture medium was changed at 6 h after transfection. Cells were then detached with 5 mM EDTA (Life Technologies, catalog no. 15575020) at 36 h posttransfection. The cells were then stained with 5 g/ml RBD-Ig proteins at 37°C for 10 min, washed three times, and then stained with 2 g/ml Alexa488-conjugated goat anti-mouse IgG secondary antibody (Invitrogen, catalog no. A-11001) at room temperature for 20 min. After another three washes, cells were analyzed by Attune NxT flow cytometer (Thermo Fisher), and signals of 10,000 FSC/SSC-gated cells were collected for each sample.

Western Blot to detect S-tagged ACE2 (ACE2-S-tag) or C9-tagged spike (Spike-C9-tag) expression in 293T cells. 293T cells were seeded at 20% density in 6-well plates at 12 to 15 h before transfection. Cells in each well were then transfected with 2 g of plasmid in complex with 5 l of Lipofectamine 2000 (Life Technologies, catalog no. 11668019). At 36 h after transfection, the cells were lysed, and 10 g of total protein was used for Western blotting. ACE2-S-tag expression was detected by using 6.2, a mouse anti-S-tag monoclonal antibody (Invitrogen, catalog no. MA1-981), and a horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG Fc secondary antibody (Invitrogen, catalog no. 31437). Beta-actin was used as an internal control. Spike-C9-tag expression was then detected by using 1D4, a mouse anti-C9-tag monoclonal antibody (Invitrogen, catalog no. MA1-722), and the HRP-conjugated goat anti-mouse IgG Fc secondary antibody (Invitrogen, catalog no. 31437).

MLV retroviral vector-based coronavirus-spike and VSV-G pseudotypes were produced using a previously described protocol (23) with some modifications. 293T cells were seeded at 30% density in 150 mm dish at 12 to 15 h before transfection. Cells were then transfected with 67.5 g of PEI MAX 4000 (Polysciences, Inc., catalog no. 24765-1) in complex with 11.25 g of plasmid encoding a coronavirus spike protein or VSV-G, 11.25 g of plasmid encoding murine leukemia virus) Gag and Pol proteins, and 11.25 g of a pQCXIP-based GFP or luciferase reporter plasmid. Eight hours after transfection, cell culture medium was refreshed and changed to growth medium containing 2% fetal bovine serum (FBS; Gibco, catalog no. 10099141C) and 25 mM HEPES (Gibco, catalog no. 15630080). Cell culture supernatants were collected at 36 to 48 h posttransfection, spun down at 3,000 ϫ g for 10 min, and filtered through 0.45-m filter units to remove cell debris. Coronavirus spike-pseudotyped viruses were then concentrated 10 times at 2,000 ϫ g using 100-kDa cutoff Amicon Ultra-15 centrifugal filter units (Millipore, catalog no. UFC910024).

Pseudovirus infection of 293T cells expressing ACE2 orthologs. 293T cells were seeded at 20% density in polylysine precoated 48-well plates 12 to 15 h before transfection. Cells in each well were then transfected with 0. Coronavirus pseudovirus neutralization assay. Coronavirus spike protein-pseudotyped luciferase reporter viruses were prediluted in DMEM (2% FBS, heat inactivated) containing titrated amounts of RBD-Ig or ACE2-Ig variant proteins. An Fc fusion protein of an anti-influenza hemagglutinin (HA) antibody, F10-scFv (25) , was used as a control protein here. Virus-inhibitor mixtures were then added to ACE2-expressing 293T or HeLa cells in polylysine (Sigma, catalog no. P4832-50ml) precoated 96-well plates and incubated overnight at 37°C. The cells were then washed with serum-free medium and incubated in 150 l of DMEM (2% FBS) at 37°C. Cell culture supernatants were collected for Gaussia luciferase assay at 48 h postinfection.

SARS-CoV-2 live virus infection of ACE2 ortholog-expressing 293T cells. 293T cells expressing one of the 16 ACE2 orthologs were inoculated with SARS-CoV-2 live virus at 800 50% tissue culture infective dose(s) (TCID 50 ) and incubated for 1 h at 37°C. The cells were then washed with serum-free medium and incubated in 150 l of DMEM (2% FBS) at 37°C for an additional 24 h. The cells were then fixed with 4% paraformaldehyde in PBS, permeabilized with 0.5% Triton X-100, and sequentially stained with 1:200-diluted rabbit anti-SARS-CoV-2 nucleocapsid polyclonal antibody (Sino Biological, catalog no. 40588-T62) at 37°C for 30 min, 4 g/ml of Alexa Fluor 568 goat anti-rabbit IgG (Invitrogen, catalog no. A-11011) at 37°C for 20 min, and 0.5 g/ml of DAPI (4=,6=-diamidino-2-phenylindole; Sigma-Aldrich, catalog no. D9542-5mg) at room temperature for 10 min. Stained cells were then examined under fluorescence microscope (IX73 microscope; Olympus).

SARS-CoV-2 live virus neutralization by ACE2-Ig variants. SARS-CoV-2 live virus at 800 TCID 50 were prediluted in DMEM (2% FBS, heat inactivated) containing titrated amounts of ACE2-Ig variant proteins and incubated at 37°C for 1 h. Virus-inhibitor mixtures were then added to ACE2-expressing 293T or HeLa cells in polylysine-precoated 48-well plates and incubated for 1 h at 37°C. The cells were then washed with serum-free medium and incubated in 400 l of DMEM (2% FBS) at 37°C for 24 to 40 h. Cells were then fixed for immunofluorescence staining of viral nucleocapsid proteins as described above.

Data collection and analysis. All the experiments were repeated two to four times. All of the infection assays for Fig. 3 were independently performed by two different people, and all the data are reproducible in different hands. The key neutralization assays of Fig. 4 to 6 were independently performed by two or three different people, and all of the data are reproducible in different hands. Attune NxT Software (Thermo Fisher) was used to collect and analyze flow cytometry data. Image Lab Software (Bio-Rad) was used to collect SDS-PAGE and Western blot image data. Cell Sens Software (Olympus) was used to collect fluorescence microscopy data. ICE Software (Berthold Technologies) was used to collect luciferase assay data. GraphPad Prism 6.0 software was used for figure preparation and statistical analyses.

Statistical analysis. Data expressed as mean values Ϯ the standard deviations (SD). Statistical analyses were performed using two-sided two-sample Student t test using GraphPad Prism 6.0 software when applicable. Differences were considered significant at P Ͻ 0.01. Data availability. The study did not generate unique data sets or code. Our research resources, including methods, plasmids, and protocols, are available upon reasonable request to qualified academic

The complete genome sequence of severe acute respiratory syndrome coronavirus strain HKU-39849 (HK-39)

Genomic characterization of the severe acute respiratory syndrome coronavirus of Amoy Gardens outbreak in Hong Kong

Epidemiology and cause of severe acute respiratory syndrome

People's Republic of China

Fatal swine acute diarrhoea syndrome caused by an HKU2-related coronavirus of bat origin

Isolation and characterization of viruses related to the SARS coronavirus from animals in southern China

Origin and evolution of pathogenic coronaviruses

Broad cross-species infection of cultured cells by bat HKU2-related swine acute diarrhea syndrome coronavirus and identification of its replication in murine dendritic cells in vivo highlight its potential for diverse interspecies transmission

A pneumonia outbreak associated with a new coronavirus of probable bat origin

China Novel Coronavirus Investigator Research Team. 2020. A novel coronavirus from patients with pneumonia in China

Genomic characterization and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding

Are pangolins the intermediate host of the 2019 novel coronavirus (SARS-CoV-2)?

Identifying SARS-CoV-2 related coronaviruses in Malayan pangolins

Isolation of SARS-CoV-2-related coronavirus from Malayan pangolins

Therapeutic options for the 2019 novel coronavirus (2019-nCoV)

The hallmarks of COVID-19 disease

Receptor recognition and cross-species infections of SARS coronavirus

Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus

Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronaviruses

SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor

Structural basis for the recognition of the SARS-CoV-2 by full-length human ACE2

Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor

Structure of SARS coronavirus spike receptor-binding domain complexed with receptor

SARS-CoV-2 and bat RaTG13 spike glycoprotein structures inform on virus evolution and furin-cleavage effects

Structure, function, and antigenicity of the SARS-CoV-2 spike glycoprotein

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

A 193-amino acid fragment of the SARS coronavirus S protein efficiently binds angiotensin-converting enzyme 2

Retroviruses pseudotyped with the severe acute respiratory syndrome coronavirus spike protein efficiently infect cells expressing angiotensin-converting enzyme 2

Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS-coronavirus 2

SARS-CoV-2 neutralizing serum antibodies in cats: a serological investigation

Coronavirus rips through Dutch mink farms, triggering culls

SARS-CoV-2 infection in farmed minks, the Netherlands

Evidence of recombination in coronaviruses implicating pangolin origins of nCoV-2019

Therapeutic strategies in an outbreak scenario to treat the novel coronavirus originating in Wuhan, China

Trimeric SARS-CoV-2 Spike interacts with dimeric ACE2 with limited intra-Spike avidity

Neutralizing antibody and soluble ACE2 inhibition of a replication-competent VSV-SARS-CoV-2 and a clinical isolate of SARS-CoV-2

Inhibition of SARS-CoV-2 infections in engineered human tissues using clinical-grade soluble human ACE2

Fusion proteins for half-life extension of biologics as a strategy to make biobetters

Novel ACE2-Fc chimeric fusion provides long-lasting hypertension control and organ protection in mouse models of systemic renin angiotensin system activation