key: cord-1017851-f03san07
authors: Zhao, Xuesen; Chen, Danying; Szabla, Robert; Zheng, Mei; Li, Guoli; Du, Pengcheng; Zheng, Shuangli; Li, Xinglin; Song, Chuan; Li, Rui; Guo, Ju-Tao; Junop, Murray; Zeng, Hui; Lin, Hanxin
title: Broad and differential animal ACE2 receptor usage by SARS-CoV-2
date: 2020-07-15
journal: bioRxiv
DOI: 10.1101/2020.04.19.048710
sha: 2fcbc36554e419b9e3416eedb92b3600cb1e2b3a
doc_id: 1017851
cord_uid: f03san07

The COVID-19 pandemic has caused an unprecedented global public health and economy crisis. The origin and emergence of its causal agent, SARS-CoV-2, in the human population remains mysterious, although bat and pangolin were proposed to be the natural reservoirs. Strikingly, comparing to the SARS-CoV-2-like CoVs identified in bats and pangolins, SARS-CoV-2 harbors a polybasic furin cleavage site in its spike (S) glycoprotein. SARS-CoV-2 uses human ACE2 as its receptor to infect cells. Receptor recognition by the S protein is the major determinant of host range, tissue tropism, and pathogenesis of coronaviruses. In an effort to search for the potential intermediate or amplifying animal hosts of SARS-CoV-2, we examined receptor activity of ACE2 from 14 mammal species and found that ACE2 from multiple species can support the infectious entry of lentiviral particles pseudotyped with the wild-type or furin cleavage site deficient S protein of SARS-CoV-2. ACE2 of human/rhesus monkey and rat/mouse exhibited the highest and lowest receptor activity, respectively. Among the remaining species, ACE2 from rabbit and pangolin strongly bound to the S1 subunit of SARS-CoV-2 S protein and efficiently supported the pseudotyped virus infection. These findings have important implications for understanding potential natural reservoirs, zoonotic transmission, human-to-animal transmission, and use of animal models. Importance SARS-CoV-2 uses human ACE2 as primary receptor for host cell entry. Viral entry mediated by the interaction of ACE2 with spike protein largely determines host range and is the major constraint to interspecies transmission. We examined the receptor activity of 14 ACE2 orthologues and found that wild type and mutant SARS-CoV-2 lacking the furin cleavage site in S protein could utilize ACE2 from a broad range of animal species to enter host cells. These results have important implications in the natural hosts, interspecies transmission, animal models and molecular basis of receptor binding for SARS-CoV-2.

human population remains mysterious, although bat and pangolin were proposed to be 27 the natural reservoirs. Strikingly, comparing to the SARS-CoV-2-like CoVs identified in 28 bats and pangolins, SARS-CoV-2 harbors a polybasic furin cleavage site in its spike (S) 29 glycoprotein. SARS-CoV-2 uses human ACE2 as its receptor to infect cells. Receptor 30 recognition by the S protein is the major determinant of host range, tissue tropism, and 31 pathogenesis of coronaviruses. In an effort to search for the potential intermediate or 32 amplifying animal hosts of SARS-CoV-2, we examined receptor activity of ACE2 from 33 14 mammal species and found that ACE2 from multiple species can support the 34 infectious entry of lentiviral particles pseudotyped with the wild-type or furin cleavage 35 site deficient S protein of SARS-CoV-2. ACE2 of human/rhesus monkey and rat/mouse 36 exhibited the highest and lowest receptor activity, respectively. Among the remaining 37 species, ACE2 from rabbit and pangolin strongly bound to the S1 subunit of 38 SARS-CoV-2 S protein and efficiently supported the pseudotyped virus infection. These RmYN02, and bat CoV RaTG13, respectively, suggesting that SARS-CoV-2 probably 74 has bat origins (2, 3, 5) . This finding is not surprising as bats are notorious for serving 75 as the natural reservoir for two other deadly human coronaviruses, SARS-CoV and 76 Middle East respiratory syndrome coronavirus (MERS-CoV), which previously caused 77 global outbreak, respectively (6, 7). 78 Although SARS-CoV-2 may have originated from bats, bat CoVs are unlikely to 79 jump directly to humans due to the general ecological separation. Other mammal 80 species may have been served as intermediate or amplifying hosts where the progenitor 81 virus acquires critical mutations for efficient zoonotic transmission to human. This has 82 5 been seen in the emergence of SARS-CoV and MERS-CoV where palm civet and 83 dromedary camel act as the respective intermediate host (7). The Huanan seafood and 84 wild animal market in Wuhan city would otherwise be a unique place to trace any 85 potential animal source; however, soon after the disease outbreak, the market was 86 closed and all the wild animals were cleared, making this task very challenging or even 87 impossible. As an alternative, wide screening of wild animals becomes imperative. 88 Several recent studies identified multiple SARS-COV-2-like CoVs (SL-CoVs) from 89 smuggled Malayan pangolins in China. These pangolin CoVs (PCoV) form two 90 phylogenetic lineages, PCoV-GX and PCoV-GD (8-11). In particular, lineage 91 PCoV-GD was found to carry a nearly identical receptor-binding motif (RBM) in the 92 spike (S) protein to that of SARS-CoV-2 ( Fig.1) The S protein of SARS-CoV-2 is a type I membrane glycoprotein, which can be cleaved 103 to S1 and S2 subunit during biogenesis at the polybasic furin cleavage site (RRAR) 104 ( Fig.1) (15) (16) (17) (18) . Previous studies have shown that furin cleavage is not essential for 105 6 coronavirus-cell membrane fusion, but enhances cell-to-cell fusion (19-23), expands 106 coronavirus cell tropism (24), increases the fitness of sequence variant within the 107 quasispecies population of bovine CoV (25). Recent studies indicated that the cleavage 108 at the S1/S2 boundary by furin in virus-producing cells is a critical prime step that 109 facilitates conformation change triggered by receptor binding during virus entry and 110 subsequent fusion-activating cleavage at the S2′site, which is located immediate 111 upstream of fusion peptide in S2 subunit (18, 24, 26) . Also, furin cleavage in HA was 112 found to convert avirulent avian influenza virus isolate to a highly pathogenic isolate 113 (27). Interestingly, this cleavage site is not present in the S protein of SARS-CoV, bat 114 SL-CoVs or pangolin SL-CoVs identified so far (5, 15). Besides furin-mediated 115 cleavage in virus-producing cells, SARS-CoV-2 S protein is also cleaved for fusion 116 activation by cell surface protease TMPRSS2 and lysosomal proteases, e.g. cathepsin L, 117 during virus entering target cells (15, 18) . 118 During cell entry, S1 binds to the cellular receptor, subsequently triggering a 119 cascade of events leading to S2-mediated membrane fusion between host cells and 120 coronavirus particles (28). S1 protein contains an independently folded domain called 121 the receptor binding domain (RBD), which harbors an RBM that is primarily involved 122 in contact with receptor ( Fig. 1) . Human ACE2 (hACE2) has been identified as the 123 cellular receptor for both 15, 17, 29) and . In addition 124 to hACE2, ACE2 from horseshoe bat (Rhinolophus alcyone) was found to support cell 125 entry of SARS-CoV-2 S-mediated VSV-based pseudotyped virus (15). By using 126 infectious virus it has also been shown that ACE2 from Chinese horseshoe bat 127 (Rhinolophus sinicus), civet and swine, but not mouse, could serve as functional 128 7 receptors (3). However, in this infection system, the entry step was coupled with other 129 steps during virus life cycle, i.e. viral genome replication, translation, virion assembly 130 and budding, and thus the receptor activity of these animal ACE2 orthologs were not 131 directly investigated. 132 In an effort to search for potential animal hosts, we examined the receptor activity 133 of ACE2 from 14 mammal species, including human, rhesus monkey, Chinese To test if other animal ACE2 orthologs can also be used as receptor for 164 SARS-CoV-2, we cloned or synthesized ACE2 from rhesus monkey, Chinese horseshoe 165 bat (Rs bat), Mexican free-tailed bat (Tb bat), rat, mouse, palm civet, raccoon dog, ferret 166 badger, hog badger, canine, feline, rabbit, and pangolin. These animals were chosen as 167 being either the proposed natural hosts for SARS-CoV-2 (bat, pangolin) (3, 10), 168 intermediate hosts for SARS-CoV (civet, raccoon) (12), common animal model (rat, 169 mouse, monkey), or household pets (canine, feline, rabbit). These ACE2 molecules were 170 transiently expressed in 293T cells (Fig.3A) , which were then infected with 171 pseudotyped virus of SARS-CoV-2 (SARS-CoV-2pp). The luciferase activity was 172 measured and normalized to hACE2 (Fig. 3B) . The results showed that (1) ACE2 of 173 9 human and rhesus monkey were the most efficient receptors; (2) ACE2 of rat and mouse 174 barely supported virus entry (<10% of hACE2); (3) the receptor activities of the other 175 10 animal ACE2s were between human/monkey and rat/mouse. Among these, ACE2 of 176 canine, feline, rabbit and pangolin could support virus entry at levels >50% of hACE2. 177 To examine receptor binding ability, we performed immunoprecipitation (IP) 178 analysis by using both S1 and receptor binding domain (RBD) as probe. Among the 14 179 different ACE2s tested, ACE2 from human, monkey, feline, rabbit and pangolin 180 exhibited significant and consistent association with S1 and RBD ( (Fig.3B ). 199 We also tested the receptor usage of these 14 ACE2 by SARS-CoV (Fig.3D ). The To help understand the molecular basis of different ACE2 receptor activities, we 208 first examined the overall sequence variation between these ACE2s. For this purpose, 209 we constructed a phylogenetic tree based on the nucleotide sequences of ACE2s (Fig.4) . 210 Interestingly, the phylogenetic clustering of ACE2s is correlated with their abilities to 211 support SARS-CoV-2 entry. For example, ACE2s in subclade IIA (human, rhesus 212 monkey and rabbit) and IIB (rat and mouse) were the most efficient and poorest receptor, been experimentally demonstrated that introduction of K353H into hACE2 significantly 241 reduces binding to SARS-CoV S1; in contrast, introduction of H353K into rat ACE2 242 12 significantly increases binding to SARS-CoV S1 (37). Our homology models indicate 243 that other residue substitutions may also be contributing to the low viral entry activity in 244 mouse and rat ACE2. Substitutions Q24N, Q27S, M82N, Q325P and E329T in rat 245 ACE2, and L79T, M82S and E329A in mouse ACE2, are all predicted to disrupt 246 interactions with RBD residues (Fig. 6 and Table 1 ). 247 Both Bat ACE2s are also inefficient receptors for viral entry (Fig.3B ). Since the 248 profile of residues at the receptor/RBD interface is significantly different from rat and 249 mouse ACE2, we examined other bat-specific residue substitutions that may be 250 contributing to receptor dysfunction. There are 8 and 10 critical residue substitutions in 251 the Rs bat and Tb bat ACE2s, respectively (Fig.5 ). Among these, we examined the 252 substitutions at positions Y41, H34 and E329 as they are only seen in bat ACE2s. The 253 Y41H substitution in both bat ACE2s appears to be disrupting the same H-bond network 254 that was disrupted by K353H in rat and mouse ACE2. Although Y41 is not as centrally 255 located in the H-bond network as K353, it directly contacts N501 from the RBD, which is 256 the same residue that is stabilized by K353. A second interaction which appears to be 257 disrupted in only bat ACE2s occurs at position H34. In humans, H34 forms a H-bond 258 with Y453 from the RBD, which is broken through a H34T substitution in bat ACE2s. 259 Finally, the bat-unique substitution E329N appears to be disrupting H-bonds connecting 260 two ACE2 residues (E329, Q325) and two RBD residues (N439, Q506). In Tb bat ACE2, 261 all connections in the H-bond network are disrupted by the single E329N substitution, 262 however the H-bond network is predicted to be restored by an additional substitution, 263 Q325E in Rs bat. In addition, other residue substitutions, i.e. T27M and M82N in Rs bat, 264 and D30Q and L79H in Tb bat, are also disruptive (Fig. 6 ). PCoV-GD has only one non-critical amino acid substitution (Q483H) in the RBM when 299 compared to SARS-CoV-2 ( Fig.1) (10) . Therefore, PCoV-GD most likely can also use 300 hACE2 and other animal ACE2s as functional receptors. 301 We also tested the receptor usage by a SARS-CoV-2 mutant that lacks the furin 302 cleavage site at the S1/S2 boundary. Our result showed that the mutant virus behaved 303 similarly to the wt virus. Namely, the entry of mutant virus could also be supported by As described above, it seems that dogs are not as susceptible as cats to 343 SARS-CoV-2 (41, 48). Interestingly, this is in agreement with results from IP analysis 344 that showed cat ACE2 could bind to S1 or RBD more efficiently than dog ACE2 345 (Fig.3C ). Structural models further suggest that, at those critical RBD-binding residues, 346 dog and cat ACE2 share 4 substitutions (Q24L, D30E, D38E, and M82T), while dog 347 ACE2 has an additional substitution, H34Y (Fig.5) . Based on structural modeling, both 348 Q24L and M82T are predicted to be disruptive, while both D30E and D38E are 349 tolerable (Table 1) . H34Y in dog ACE2 is predicted to disrupt the hydrogen bond with 350 Y453 of RBD (Table 1) . These atomic interactions explain why dog ACE2 binds to S1 351 or RBD less efficiently compared to cat ACE2, and both are less efficient than human 352 ACE2. 353 In addition to cat and dog, rabbits are also often raised as household pets. Our 354 results indicate that rabbit ACE2 is an efficient receptor ( Fig. 3B and 3C) , suggesting 355 that rabbit may be more susceptible to SARS-CoV-2 infection than cat. (Fig.3D ), but not SARS-CoV-2. An alternative way to make a mouse-adapted 379 SARS-CoV-2 strain could be achieved by rational design of the S gene. Based on the 380 structural model, we know that receptor dysfunction of mouse ACE2 is due to 381 disruptive D30N, L79T, M82S, Y83F, E329A and K353H substitutions (Fig.5, Fig.6 and 382 Table 1 ). Therefore, by specifically introducing mutations into the RBM of S gene it 383 may be possible to restore or at least partly restore interactions with these ACE2 384 substitutions. Consequently, the engineered virus may be able to efficiently infect 385 wildtype mice. 386 To date, several animals (i.e. rhesus monkey, ferret, dog, cat, pig, chicken and duck) 387 have been examined as potential animal models for SARS-CoV-2 (40, 41). Although 388 the rhesus monkey, ferret and cat may seem to be the promising candidates, none of 389 them are perfect in terms of recapitulation of typical clinical features in COVID-19 390 patients. Therefore, multiple animal models may be needed. Our results indicate that 391 rabbit ACE2 is a more efficient receptors than other animal ACE2s for both 392 SARS-CoV-2 and SARS-CoV (Fig.3) . Therefore, it may be worthy assessing rabbit as a 393 useful animal model for further studies. The nucleotide sequence of SARS-CoV-2 S gene was retrieved from NCBI 430 database (isolate Wuhan-Hu-1, GenBank No. MN908947). According the method 431 described by Gregory J. Babcock et al (59) , the codon-optimized S gene was 432 synthesized, and cloned into pCAGGS vector. The SARS-CoV-2 S gene mutant without 433 the furin cleavage site at the S1/S2 boundary was generated by an overlapping 434 PCR-based method as previously described (60). The S1 subunit (aa 14-685) and RBD 435 (aa 331-524) were cloned into a soluble protein expression vector, pSecTag2/Hygro-Ig 436 vector, which contains human IgG Fc fragment and mouse Ig k-chain leader sequence 437 (61). The protein expressed is soluble and has a human IgG-Fc tag. 438 Western blot assay 439 As previously described, the expression of ACE2-C9, S1-Ig, and RBD-Ig fusion 440 proteins were examined by western-blot (61). Briefly, lysates or culture supernatants of 441 293T cells transfected with plasmid encoding ACE2 orthologs and S1-Ig or RBD-Ig 442 were collected, boiled for 10 min, and then resolved by 4~12% SDS-PAGE. A PVDF 443 membrane containing the proteins transferred from SDS-PAGE was blocked with 444 blocking buffer (5% nonfat dry milk in TBS) for 1h at room temperature and probed 445 with primary antibody overnight at 4 °C. The blot was washed three times with washing 446 buffer (0.05% tween-20 in TBS), followed by incubation with secondary antibody for 447 1h at room temperature. After three-time washes, the proteins bounded with antibodies 448 21 were imaged with the Li-Cor Odyssey system. (Li-Cor Biotechnology) . 449 Immunoprecipitation (IP) assay. 450 The association between Ig-fused S1 protein or RBD protein and ACE2 protein 451 with C9 tag was measured by IP according to a previously described method (60). 

functional receptor for the SARS coronavirus

GILT restricts the cellular entry mediated by the envelope glycoproteins of 605

Ebola virus and Lassa fever virus. Emerging microbes & infections

Identification of Residues Controlling Restriction versus Enhancing Activities of 608 IFITM Proteins on Entry of Human Coronaviruses

A Multibasic Cleavage Site in the Spike 610

Protein of SARS-CoV-2 Is Essential for Infection of Human Lung Cells

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

SARS-CoV-2 by full-length human ACE2

Structure of SARS coronavirus spike 618 receptor-binding domain complexed with receptor

Receptor and viral 621 determinants of SARS-coronavirus adaptation to human ACE2

Structure, Function, and Evolution of Coronavirus Spike Proteins

Coronaviruses post-SARS: update on replication and pathogenesis. 625

Respiratory disease in rhesus 629 macaques inoculated with SARS-CoV-2

Susceptibility of ferrets, cats, dogs, and other domesticated animals to SARS-coronavirus 2. 633

SARS-CoV-2 neutralizing serum antibodies in cats: 637 a serological investigation

Virology: SARS virus infection of cats and ferrets

An animal model of SARS 644 produced by infection of Macaca mulatta with SARS coronavirus

Macaque model for severe acute respiratory syndrome

Animal models and vaccines for SARS-CoV infection

Furin cleavage of the SARS coronavirus spike 652 glycoprotein enhances cell-cell fusion but does not affect virion entry

Canine SARS-CoV-2 infection

Animal-to-human SARS-associated coronavirus transmission? 658 Emerging infectious diseases

Animal models for highly pathogenic emerging 660 viruses

The pathogenicity of SARS-CoV-2 in hACE2 transgenic mice. 666 Nature

Lethal 669 infection of K18-hACE2 mice infected with severe acute respiratory syndrome coronavirus

Clinical features of patients infected with 2019 674 novel coronavirus in Wuhan

Labs rush to study coronavirus in transgenic animals -some are in short 676 supply

A new mouse-adapted strain of SARS-CoV as a lethal model for evaluating antiviral 679 agents in vitro and in vivo

A mouse-adapted

SARS-coronavirus causes disease and mortality in BALB/c mice

Identification of 684 three interferon-inducible cellular enzymes that inhibit the replication of hepatitis C virus

Efficient replication of severe acute respiratory syndrome coronavirus in mouse 688 cells is limited by murine angiotensin-converting enzyme 2

Amino acids 270 to 510 of 690 the severe acute respiratory syndrome coronavirus spike protein are required for interaction with 691 receptor

Identification of residues in the receptor-binding domain (RBD) of the spike protein of human 694

coronavirus NL63 that are critical for the RBD-ACE2 receptor interaction

Identification of residues in the receptor-binding domain (RBD) of the spike protein of human 698

coronavirus NL63 that are critical for the RBD-ACE2 receptor interaction

CLUSTAL W: improving the sensitivity of 701 progressive multiple sequence alignment through sequence weighting, position-specific gap 702 penalties and weight matrix choice

The Phyre2 web portal for 704 protein modeling, prediction and analysis

which is not present in SARS-CoV and other animal SARS-CoV-2-like

while the S2 subunit 730 containing a FP mediates membrane fusion. In SARS-CoV-2, the S1 contains NTD and 731 an independently folded domain known as RBD, which harbors a region called receptor 732 binding motif (RBM), that are primarily in contact with receptor. The most critical 733 hACE2-binding residues in the RBM of several SARS-CoV-2-related CoVs are 734 highlighted in yellow and referred from the crystal structure of RBD-hACE2 complex 735 (Shang et al, Nature)

The GenBank No. for these CoVs is: 738 SARS-CoV-2 (isolate Wuhan-Hu-1, MN908947)

Human ACE2 served as receptor for SARS-CoV-2. (A) ACE2 supported

At 48 h post transfection, the cells were infected by SARS-CoV-2 S protein 746 pseudotyped virus (SARS-CoV-2pp)

incubated with indicated concentration of hACE2 antibody or control antibody 750 (anti-IDE) for 1 h, and then infected by pseudotyped virus of SARS-CoV-2, Influenza 751 virus A (IAVpp) or human coronavirus (HCoV) OC43 (HCoV-OC43pp) in the presence 752 of indicated concentration of hACE2 antibody or control antibody (anti-IDE) for 753 another 3 h, then the virus and antibodies were removed

Multiple ACE2 orthologues served as receptors for SARS-CoV-2

Transient expression of ACE2 orthologs in 293T cells. The cell lysates were detected by 765 western blot assay, using an anti-C9 monoclonal antibody. (B) HIV-Luc-based 766 pseudotyped virus entry. 293T cells were transfected with ACE2s orthologs

Error bars reveal the standard deviation 770 of the means from four biological repeats. (C) IP assay. The upper panel showed the 771 input of ACE2 protein with C9 tag, S1 and RBD with IgG tag. The lower panel showed 772 the ACE2 pulled down by S1-Ig or RBD-Ig fusion protein. (D) SARS-CoV 773 spike-mediated entry. 293T cells were transfected with ACE2s orthologs. At 48 h post 774 transfection, the cells were infected by the pseudotyped virus of SARS-CoV. At 48 h 775 post infection

Phylogenetic clustering of ACE2s correlates with their receptor activities. 783 Upper panel: phylogram tree of 14 ACE2s. The tree was constructed based on 784 nucleotide sequences using the Neighbor-joining method implemented in program

The percentage of replicate trees in which the associated taxa clustered 786 together in the bootstrap test (1000 replicates) are shown next to the branches. The tree 787 was rooted by ACE2 of platypus (Ornithorhynchus anatinus). The taxonomic orders 788 where these animals are classified are shown on the right-hand side of the tree

Fig. 5. Critical RBD-binding residues in ACE2 orthologs

PDB: 6VW1) in the bound conformation was extracted 797 from the SARS-CoV-2 RBD/ACE2 complex and used as a template for homology 798 modeling (16). Low panel: Critical RBD-binding residues in ACE2 orthologs

ACE2s and both bats, respectively. The rest of residue substitutions are highlighted in

PDB: 6VW1) in the bound conformation was extracted from the 807 SARS-CoV-2 RBD/ACE2 complex and used as a template for homology modeling 808 (16). ACE2 Homology models were generated using the one-to-one threading algorithm 809 of Phyre2 (63)

The effects of residues substitution were predicted by homologous-based modeling 824 analyses based on the crystal structure of SARS-CoV-2 RBD/hACE2 complex (16)