Close relatives of MERS-CoV in bats use ACE2 as their functional receptors

Xiong, Qing; Cao, Lei; Ma, Chengbao; Tortorici, M. Alejandra; Liu, Chen; Si, Junyu; Liu, Peng; Gu, Mengxue; Walls, Alexandra C.; Wang, Chunli; Shi, Lulu; Tong, Fei; Huang, Meiling; Li, Jing; Zhao, Chufeng; Shen, Chao; Chen, Yu; Zhao, Huabin; Lan, Ke; Corti, Davide; Veesler, David; Wang, Xiangxi; Yan, Huan

doi:10.1038/s41586-022-05513-3

Download PDF

Article
Open access
Published: 07 December 2022

Close relatives of MERS-CoV in bats use ACE2 as their functional receptors

Nature volume 612, pages 748–757 (2022)Cite this article

22k Accesses
35 Citations
767 Altmetric
Metrics details

Subjects

Abstract

Middle East respiratory syndrome coronavirus (MERS-CoV) and several bat coronaviruses use dipeptidyl peptidase-4 (DPP4) as an entry receptor^1,2,3,4. However, the receptor for NeoCoV—the closest known MERS-CoV relative found in bats—remains unclear⁵. Here, using a pseudotype virus entry assay, we found that NeoCoV and its close relative, PDF-2180, can efficiently bind to and use specific bat angiotensin-converting enzyme 2 (ACE2) orthologues and, less favourably, human ACE2 as entry receptors through their receptor-binding domains (RBDs) on the spike (S) proteins. Cryo-electron microscopy analysis revealed an RBD–ACE2 binding interface involving protein–glycan interactions, distinct from those of other known ACE2-using coronaviruses. We identified residues 337–342 of human ACE2 as a molecular determinant restricting NeoCoV entry, whereas a NeoCoV S pseudotyped virus containing a T510F RBD mutation efficiently entered cells expressing human ACE2. Although polyclonal SARS-CoV-2 antibodies or MERS-CoV RBD-specific nanobodies did not cross-neutralize NeoCoV or PDF-2180, an ACE2-specific antibody and two broadly neutralizing betacoronavirus antibodies efficiently inhibited these two pseudotyped viruses. We describe MERS-CoV-related viruses that use ACE2 as an entry receptor, underscoring a promiscuity of receptor use and a potential zoonotic threat.

Broad host tropism of ACE2-using MERS-related coronaviruses and determinants restricting viral recognition

Article Open access 15 June 2023

Isolation of MERS-related coronavirus from lesser bamboo bats that uses DPP4 and infects human-DPP4-transgenic mice

Article Open access 11 January 2021

ACE2 receptor usage reveals variation in susceptibility to SARS-CoV and SARS-CoV-2 infection among bat species

Article 01 March 2021

Main

Coronaviruses are a large family of enveloped positive-strand RNA viruses classified into four genera: alpha-, beta-, gamma- and deltacoronaviruses. Alpha- and betacoronaviruses typically infect mammals such as bats and humans, whereas gamma- and deltacoronaviruses mainly infect birds and occasionally mammals^6,7. It is thought that the origins of most coronaviruses that infect humans can be traced back to their close relatives in bats—the most important animal reservoir of mammalian coronaviruses⁸. Cross-species transmission of coronaviruses has led to three major outbreaks in the past two decades caused by SARS-CoV, MERS-CoV and, most recently, SARS-CoV-2^9,10,11,12.

MERS-CoV belongs to lineage C of betacoronaviruses (merbecovirus subgenus) and has been reported to have a high case–fatality rate of approximately 35% (ref. ¹³). Merbecoviruses have also been found in several animal species, including camels, hedgehogs and bats. Although camels are confirmed intermediate hosts for MERS-CoV, bats, especially species in the family of Vespertilionidae, are widely considered to be the evolutionary source of MERS-CoV or its immediate ancestor¹⁴.

Specific coronavirus recognition of host receptors is usually mediated by the RBD (or domain B) within the spike protein S₁ subunit^15,16. Three out of four well-characterized human coronavirus receptors are transmembrane proteases, namely ACE2, DPP4 and aminopeptidase N (APN)^1,17,18. By contrast, carcinoembryonic antigen-related cell adhesion molecule 1 (CEACAM1) interacts with the murine hepatitis virus spike N-terminal domain (domain A)^19,20. Notably, the same receptor can be shared by distantly related coronaviruses with structurally distinct RBDs. For example, NL63 (an alphacoronavirus) uses ACE2 as an entry receptor, which is also widely used by sarbecoviruses (betacoronavirus lineage B)²¹. APN is a second example of cross-genera receptor use, as it mediates the entry of several alphacoronaviruses and a deltacoronavirus (PDCoV)⁷. However, DPP4 is known to be an entry receptor only for some merbecoviruses (betacoronavirus lineage C), such as MERS-CoV, HKU4, HKU25 and related strains^2,3,4,22.

Several merbecoviruses do not use DPP4 for entry, and their receptor use remains unclear, including the bat coronaviruses NeoCoV, PDF-2180, HKU5 and hedgehog coronaviruses EriCoV-HKU31^5,23,24,25. NeoCoV is the closest relative of MERS-CoV (85% nucleotide sequence identity at the whole-genome level) and was identified in Neoromicia capensis (Cape serotine, or Laephotis capensis, formerly classified in Neoromicia before phylogenetic analysis found that it belongs to Laephotis). NeoCoV was previously erroneously assumed to be sampled from Neoromicia zuluensis bats in South Africa^26,27. PDF-2180, which is closely related to NeoCoV phylogenetically (around 91% amino acid sequence identity in the S₁ subunits), infects Pipisrellus hesperidus bats, which are native to Southwest Uganda^23,28. The marked sequence divergence of the NeoCoV and PDF-2180 S₁ subunits relative to MERS-CoV (around 44–45% amino acid sequence identity) probably explains why these viruses do not use DPP4 as an entry receptor²³.

Here we report that NeoCoV and PDF-2180 use bat ACE2 as an entry receptor. We determined the cryo-electron microscopy (cryo-EM) structures of the PDF-2180 spike trimer as well as of the NeoCoV and PDF-2180 RBDs bound to Pipistrellus pipistrellus ACE2, revealing a binding mode that is distinct from that observed for SARS-CoV-2 and NL63. Although NeoCoV and PDF-2180 use human ACE2 (hACE2) suboptimally, adaptive mutations of this group of viruses might result in future spillover events.

Evidence of ACE2 use

To shed light on the evolutionary relationship among merbecoviruses, especially NeoCoV and PDF-2180, we conducted a phylogenetic analysis of the sequences of several human and animal coronaviruses. Maximum-likelihood phylogenetic reconstructions based on complete genome nucleotide sequences showed that NeoCoV and PDF-2180 form a sister clade to MERS-CoV (Fig. 1a). By contrast, a phylogenetic tree based on amino acid sequences of the S₁ subunit demonstrated that NeoCoV and PDF-2180 showed a distant evolutionary relationship with MERS-CoV but are closely related to hedgehog coronaviruses (EriCoVs) (Fig. 1b). A similarity plot analysis (Simplot) highlights that the S₁ subunits of NeoCoV and PDF-2180 are even more divergent from MERS-CoV than is the case for HKU4 (Fig. 1c). We first tested whether human DPP4 (hDPP4) could support the cell entry of several merbecoviruses using a pseudotype virus assay (Extended Data Fig. 1a). The result revealed that MERS-CoV S and HKU4 S pseudotyped viruses showed robust entry in BHK-21-hDPP4 and HEK293T-hDPP4 cells, although there is detectable MERS-CoV pseudovirus entry in other HEK293T cell lines, probably due to weak endogenous expression of hDPP4 (Fig. 1d and Extended Data Fig. 1b). Notably, we detected a significant increase in the entry of NeoCoV S and PDF-2180 S pseudotyped viruses in hACE2-expressing cells but not in hAPN-expressing cells, both of which were initially considered to be negative control cells (Fig. 1d and Extended Data Fig. 1b).

**Fig. 1: NeoCoV and PDF-2180 use ACE2 but not DPP4 for efficient entry.**

To investigate the possibility of more efficient use of bat ACE2 by these viruses, we screened an ACE2 library comprising 46 HEK293T stable cell lines individually expressing ACE2 orthologues spanning the Chiropteran phylogeny, as described in our previous study²⁹ (Extended Data Fig. 2a,b and Supplementary Table 4). NeoCoV S and PDF-2180 S, but not HKU4 S or HKU5 S pseudotyped viruses, showed efficient entry in cells expressing ACE2 from most Yangochiropteran bats, whereas no or very limited entry was observed using cells expressing ACE2 from Yinpterochiropteran bats or humans (Fig. 1e and Extended Data Fig. 3a–e). Consistent with previous reports, NeoCoV and PDF-2180 S-mediated pseudotyped virus entry could be substantially enhanced by exogenous trypsin treatment²⁸ (Extended Data Fig. 3f,g). As indicated by a dual split protein (DSP)-based fusion assay, ACE2 from Bat37 (Methods; Bat37ACE2) promoted more efficient cell–cell fusion than hACE2 in the presence of NeoCoV or PDF-2180 S, but not MERS-CoV S (Fig. 1f,g). The fact that human or hedgehog ACE2 did not support entry of EriCoV-HKU31 suggests that this virus may use a different receptor compared with NeoCoV and PDF-2180 (Extended Data Fig. 3h,i). Our results rule out bat DPP4 as a possible receptor for NeoCoV or PDF-2180, as none of the tested DPP4 orthologues could promote detectable entry of these two pseudotyped viruses (Fig. 1h and Extended Data Fig. 4a–d), in agreement with previous studies^23,28. Pseudotyped virus entry assays were also conducted using several other cell types from various species, including a bat cell line Tb 1 Lu, ectopically expressing ACE2 or DPP4 from Bat40 (Antrozous pallidus), and each test yielded similar results (Extended Data Fig. 5a–c).

Domain-B-mediated specific ACE2 binding

The inability of NeoCoV and PDF-2180 to use DPP4 is consistent with their highly divergent domain B (putative RBD) sequence compared with MERS-CoV and HKU4. We produced RBD–hFc (domain B fused to human IgG Fc domain) proteins to verify whether this domain was responsible for ACE2 receptor engagement. Live-cell binding assays using NeoCoV RBD–hFc showed species-specific recognition of various cell-surface-expressed bat ACE2 orthologues, in agreement with the results of the pseudotyped virus entry assays (Fig. 2a). Specific binding of NeoCoV RBD–hFc to several representative bat ACE2 orthologues was also confirmed by flow cytometry (Fig. 2b). Using biolayer interferometry (BLI), we found that the NeoCoV and PDF-2180 RBDs bound with the highest avidity to Pipistrellus pipistrellus (Bat37) ACE2 (apparent binding affinity (K_D,app) = 1.98 nM for NeoCoV and K_D,app = 1.29 nM for PDF-2180), whereas hACE2 binding was below the detection limit (Fig. 2c). An enzyme-linked immunosorbent assay (ELISA) also demonstrated the strong binding between NeoCoV and PDF-2180 RBDs and Bat37ACE2, but not hACE2 (Fig. 2d). As the ACE2 sequences of the hosts of NeoCoV and PDF-2180 are unknown, Bat37 represents the closest relative of the host of PDF-2180 (P. hesperidus) in our study. Although Bat37ACE2 is not the best receptor for supporting pseudotype entry, probably due to its moderate association rate (k_on) with the RBDs of the two viruses, the observed dissociation rate (k_off) is very slow, rendering it ideal for virus neutralization and cryo-EM analysis. The estimated binding affinities were further verified by competitive neutralization assays using soluble ACE2-ectodomain or RBD–hFc proteins. The entry of both NeoCoV and PDF-2180 S pseudotyped viruses was inhibited most efficiently by soluble Bat37ACE2 in a concentration-dependent manner (Fig. 2e,f). Moreover, the NeoCoV RBD–hFc and the PDF-2180 RBD–hFc neutralized NeoCoV S pseudotyped virus entry into cells expressing Bat37ACE2 (Fig. 2g). We further demonstrated the important role of the RBD in determining receptor use by assessing the entry specificity of pseudotyped viruses containing chimeric viral spike proteins. Bat ACE2 use was changed to hDPP4 use for a chimeric NeoCoV spike containing the MERS-CoV RBD (Fig. 2h). By contrast, a chimeric MERS-CoV spike containing the NeoCoV RBD was inefficient in pseudotyped assembly (Extended Data Fig. 5d–e). Taken together, these results demonstrate that the NeoCoV and PDF-2180 B domains are bona fide RBDs for species-specific interaction with ACE2.

**Fig. 2: The domain B (RBD) of NeoCoV and PDF-2180 S proteins are required for species-specific ACE2 binding.**

The structural basis for ACE2 binding

To determine the molecular basis of ACE2 recognition, we performed structural studies of Bat37ACE2 in a complex with the NeoCoV and PDF-2180 RBDs. 3D classification revealed that the NeoCoV RBD–Bat37ACE2 complex primarily adopts a dimeric state with two copies of ACE2 bound to two RBDs, whereas only a monomeric state was observed in the PDF-2180–Bat37ACE2 complex (Figs. 3a,b and Extended Data Fig. 6a,b). We determined the structures of these two complexes at a resolution of 3.8 Å, and C₂ symmetry expansion was performed for the NeoCoV RBD–Bat37ACE2 complex to further improve the resolution to 3.5 Å, enabling a reliable analysis of the interactions involved in binding (Fig. 3a,b, Extended Data Fig. 6a,b and Supplementary Tables 1 and 2). Despite existing in different oligomeric states, the structures revealed that both NeoCoV and PDF-2180 recognize Bat37ACE2 in a very similar manner (Figs. 3a,b and Extended Data Fig. 7a). Thus, we used the NeoCoV RBD–Bat37ACE2 complex structure for detailed analysis. Both RBDs comprise a receptor-binding motif (RBM) and a core subdomain, as described for other betacoronaviruses (Fig. 3c and Extended Data Fig. 7b). The RBM folds as a four-stranded anti-parallel β-sheet (β6–β9) and exposes its tip surface for ACE2 engagement (Fig. 3c). By contrast, the MERS-CoV RBD recognizes the side surface of the DPP4 β-propeller through its four-stranded β-sheet (Fig. 3c). The differences in receptor use for these two viruses can be explained by (1) the local configuration of the NeoCoV RBM, which shows a conformational shift of η3 and β8 disrupting the flat DPP4-binding surface, and (2) the longer MERS-CoV 6–7 and 8–9 loops, relative to NeoCoV, would impair binding to the shallow cavity of bat ACE2 (Fig. 3c and Extended Data Fig. 7b,c).

**Fig. 3: Cryo-EM structures of NeoCoV RBD and PDF-2180 RBD in a complex with Bat37ACE2.**

In the NeoCoV–Bat37ACE2 complex structure, relatively smaller surface areas (498 Å² in NeoCoV RBD and 439 Å² in Bat37ACE2) are buried by the two binding partners compared with their counterparts in the MERS-CoV–DPP4 complex (926 Å² in MERS-CoV RBD and 1,037 Å² in DPP4, using Protein Data Bank (PDB) 4KR0) or the SARS-CoV-2-RBD complex (864 Å² in SARS-CoV-2 RBD and 823 Å² in hACE2, using PDB 6M0J). The NeoCoV RBM inserts into an apical depression constructed by the α9 and α10 helices and a loop connecting α10 and β4 of Bat37ACE2 through a network of electrostatic and hydrophobic interactions involving the tip of the RBM four-stranded β-sheet (Fig. 3d and Supplementary Table 2). Polar interactions are predominantly mediated by residues Asn504, Asn506, Asn511, Lys512 and Arg550 from the NeoCoV RBM and residues Glu305, Thr334, Asp338 and Arg340 from Bat37ACE2 (Fig. 3d and Supplementary Table 2). Moreover, the methyl groups of the NeoCoV RBM Ala509 and Thr510 residue side chains are packed into a hydrophobic pocket formed by residues Phe308, Trp328 and Leu333 of Bat37ACE2. The PDF-2180 RBM contains a T510F substitution relative to NeoCoV, which further improves hydrophobic interactions with Bat37ACE2 and is consistent with the increased binding affinity observed for this NeoCoV point mutant (Figs. 3d and Supplementary Table 2). Moreover, the Bat37 ACE2 glycans at positions Asn54 and Asn329 sandwich the strands (β8–β9), forming extensive interactions with RBD residue Trp540 and to a lesser extent with Asn532, Leu539, Ala541, Gly545, Gly546 and Val547 from the NeoCoV RBD, underpinning virus–receptor associations (Fig. 3d and Supplementary Table 2).

The importance of key interface residues was verified by mutagenesis and assessment of receptor binding and pseudotyped virus entry. As expected, the mutations N511Y and R550N in the NeoCoV RBD, which abolish the polar contacts and introduce steric clashes, resulted in a significant decrease in RBD binding and viral entry (Fig. 3e–g). Similarly, the E305K mutation in Bat37ACE2, which eliminates a salt bridge, also impaired receptor function. Moreover, the Bat37ACE2 N54A mutation (abolishing the N-glycosylation) and NeoCoV W540G/N532G substitutions (eliminating interactions with the glycan) hindered binding. These results confirmed the importance of the protein–glycan interactions identified in viral–receptor recognition, as recently described for the human-infecting CCoV-HuPn-2018³⁰. By contrast, the mutation N329A, which abolishes the N-glycosylation at site Asn329, did not have a major effect on receptor function (Fig. 3h–j).

Determinants of human ACE2 recognition

As mentioned above, the NeoCoV and PDF-2180 RBDs cannot efficiently interact with hACE2. Here we first examined the molecular determinants that restrict hACE2 from supporting the entry of these two pseudotyped viruses. By comparing the binding interface of the other three hACE2-using coronaviruses, we found that the SARS-CoV, SARS-CoV-2 and NL63 RBDs share overlapping footprints on hACE2 that barely overlapped with the region engaged by NeoCoV (Fig. 4a). Indeed, only residues 329–330 are in contact with all four virus RBDs (Fig. 4b). On the basis of sequence alignment and structural analysis of hACE2 and Bat37ACE2, we predicted that inefficient hACE2 binding could be attributed to suboptimal residues at the binding interfaces with NeoCoV and PDF-2180, especially around the divergent residues around 337–342 (Fig. 4c). To test this hypothesis, we replaced these hACE2 residues with those from the Bat37ACE2 orthologue (Fig. 4c,d). The mutations led to a significant gain of binding affinity with RBDs relative to wild-type hACE2, and an approximately 15-fold and 30-fold increase in the entry efficiency of NeoCoV and PDF-2180 S pseudotyped viruses, respectively (Fig. 4e–g). These results demonstrate that this region is critical for receptor engagement and host range determination, with residue Asn338 having a crucial role in restricting human receptor use, consistent with the intensive interactions that it forms with the NeoCoV RBD (Fig. 3d).

**Fig. 4: Molecular determinants affecting hACE2 recognition by NeoCoV and PDF-2180.**

To assess the risk of NeoCoV and PDF-2180 adaptation to hACE2, we sought to identify the RBM amino acid changes that would enable more efficient engagement of hACE2 based on our cryo-EM structures assisted by the mCSM-PPI2 software (Fig. 4h). We predicted that increasing hydrophobicity around the residue Thr510 of NeoCoV might enhance interactions with hACE2 (Fig. 4i). The PDF-2180 spike equivalent residue is Phe511, which is consistent with its slightly higher affinity for hACE2 (Extended Data Fig. 8a). Indeed, the NeoCoV T510F RBD mutant had a substantially increased binding affinity for hACE2 (K_D,app = 16.9 nM), and the corresponding pseudotyped virus entered HEK293T-hACE2 cells markedly better than the wild-type pseudotyped virus (Fig. 4j–l and Extended Data Fig. 8b). However, the PDF-2180 S pseudotyped virus entered these cells with much lower efficiency than the NeoCoV T510F S pseudotyped virus, indicating that other RBM residues might restrict efficient interactions with hACE2. Indeed, a G to A mutation at site 510 (corresponding to A509 in NeoCoV), increasing the local hydrophobicity, marginally enhanced PDF-2180 RBD binding to hACE2 (Fig. 4j–l). Additional replacement of residues 537–543 of the PDF-2180 spike with those of the NeoCoV spike further increased PDF-2180 binding affinity for hACE2 (Extended Data Fig. 8c–g). Moreover, the NeoCoV T510F S pseudotyped virus entered human colon cell line Caco-2 with much higher efficiency than wild-type NeoCoV, and was inhibited by the ACE2-targeting antibody H11B11³¹ (Fig. 4m).

Spike architecture and antigenicity

To understand the antigenic landscape of these ACE2-using bat merbecoviruses and provide a blueprint for vaccine design, we determined a cryo-EM structure of the prefusion PDF-2180 spike trimer. 3D classification of the data revealed the presence of a single conformation corresponding to a closed trimer for which we determined a structure at a resolution of 2.5 Å (Fig. 5a,b, Extended Data Fig. 9a–d and Supplementary table 1). The PDF-2180 S₁ and S₂ subunits are similar to the MERS-CoV equivalent subunits with which they can be superimposed with a root‐mean‐square distance of 2.8 Å and 1.4 Å for 566 and 443 aligned residues, respectively, reflecting their degree of sequence conservation (Fig. 5c–g). Although both spikes contain a furin-cleavage site at the S₁–S₂ junction, processing during synthesis was detected for MERS-CoV but not for PDF-2180 S pseudotyped virions (Extended Data Fig. 10a). Moreover, the S₁–S₂ cleavage site is resolved in the PDF-2180 S ectodomain trimer, but not MERS-CoV S, reflecting the different accessibility of this motif and putatively contributing to the closed conformation of the RBDs³² (Fig. 5g). Differential S cleavage has previously been shown to modulate plasma membrane versus endosomal fusion and suggests that these two viruses might use different host cell entry pathways besides their distinct receptor usage^33,34,35.

**Fig. 5: The architecture and antigenicity of the PDF-2180 spike glycoprotein.**

Previous research indicated that the RBD-directed LCA60 MERS-CoV neutralizing antibody does not neutralize a MERS-CoV virus chimera containing the PDF-2180 spike, in agreement with the aforementioned structural differences between the RBMs of these two viruses^28,36,37. Furthermore, no inhibition of the PDF-2180 or NeoCoV S pseudotyped viruses was observed with SARS-CoV-2 antisera obtained from individuals who were vaccinated with inactivated SARS-CoV-2 vaccine or a panel of ten RBD-neutralizing MERS-CoV nanobodies^38,39 (Fig. 5h,i and Extended Data Fig. 10b–e). As the SARS-CoV-2 S₁ subunit, especially the RBD, accounts for most of the infection- or vaccine-elicited serum neutralizing activity^40,41,42 and several potent MERS-CoV neutralizing antibodies target the RBD, our findings suggest that immunity to either SARS-CoV-2 or MERS-CoV would not elicit appreciable titres of cross-neutralizing antibodies owing to marked divergence of their RBDs^28,36,43.

We previously described the B6 and S2P6 cross-reactive and broadly neutralizing betacoronavirus antibodies targeting a conserved stem helix epitope in the spike S₂ fusion machinery^44,45. As the stem helix is strictly conserved between MERS-CoV, PDF-2180 and NeoCoV S, we assessed the ability of these two monoclonal antibodies to block PDF-2180 and NeoCoV S-mediated entry. B6 and S2P6 inhibited entry of PDF-2180, NeoCoV and NeoCoV-T510F S pseudotyped viruses into HEK293T cells that were transiently transfected with Bat40ACE2 (A. pallidus) or hACE2 (Fig. 5j–o) in a dose-dependent manner. Moreover, the strict architectural conservation of the fusion peptide region between PDF-2180 and MERS-CoV (Fig. 5g) explains that recently described monoclonal antibodies targeting this motif neutralize PDF-2180 S pseudovirus as well⁴⁶. Collectively, these data underscore that the conservation of the fusion machinery between NeoCoV, PDF-2180 and MERS-CoV S is associated with the retention of inhibitory activity of several stem-helix- and fusion-peptide-directed broadly neutralizing betacoronavirus antibodies, some of which have been shown to reduce the viral burden in small-animal challenge models^45,47.

Discussion

The lack of knowledge of the receptors used by many bat coronaviruses limits our understanding of their entry and interspecies transmission mechanisms. We show that close relatives of MERS-CoV, such as NeoCoV and PDF-2180, engage bat ACE2 for efficient cellular entry. However, HKU5 and EriCoV do not use bat DPP4 or hedgehog ACE2 for entry, highlighting the complexity of receptor use by coronaviruses.

Many sarbecoviruses, alphacoronavirus NL63 and a group of merbecoviruses reported in this study share ACE2 for cellular entry. Our structural analysis indicates that NeoCoV and PDF-2180 bind to an apical side surface of ACE2, which is largely different from the surface engaged by other ACE2-using coronaviruses (Fig. 4a). The different interaction modes of the three groups of ACE2-using coronaviruses along with the distinct structures of their RBMs or receptor-binding loops suggest a convergent acquisition of ACE2 use during evolution²¹. The evolutionary advantage of ACE2 use for different coronaviruses remains unclear and deserves further investigation.

Our results support the previous hypotheses that the origin of MERS-CoV might be a result of an intraspike recombination event between a NeoCoV-like virus and a DPP4-using virus²⁶. RNA recombination can occur during the co-infection of different coronaviruses, giving rise to a new virus with different receptor use and host tropism⁴⁸. It remains unclear whether the event took place in bats, camels or other hosts, and when the host switching would have happened. Although bat merbecoviruses are geographically widespread, the two known ACE2-using merbecoviruses were both discovered in Africa. Considering that both NeoCoV and PDF-2180 S glycoproteins mediate relatively inefficient entry into human cells, the acquisition of the hDPP4-binding domain or the S₁ subunit in an unknown host coinfected with a DPP4-using virus might have driven the emergence of MERS-CoV. Further studies will be necessary to determine the evolutionary trajectory of MERS-CoV.

The host range determinants on ACE2 are primary barriers for cross-species transmission of these viruses. Our results show that NeoCoV and PDF-2180 favour ACE2 from Yangochiropteran bats, especially vesper bats (Vespertilionidae), which their host belongs to, but not ACE2 orthologues from Yinpterochiropteran bats. To date, most merbecoviruses were found in species belonging to the Vespertilionidae group, a highly diverse and widely distributed family⁸. Although the two viruses could not use hACE2 efficiently, our study also reveals that single-residue substitution increasing local hydrophobicity around NeoCoV site 510 could enhance their affinity for hACE2 and enable them to enter human cells endogenously expressing ACE2. Considering the extensive mutations in the RBD regions of the SARS-CoV-2 variants, especially the heavily mutated omicron variant, these viruses may hold a latent potential to infect humans by further adaptation through mutations^49,50. However, zoonotic spillover is a complex process involving multiple factors other than receptor recognition, including proteolytic spike activation, replication, immune response and contact opportunity. To date, there is no evidence that NeoCoV or PDF-2180 can infect mammals other than bats. Our results also did not confirm that NeoCoV carrying T510F, which has not been found in nature, can infect humans in vivo.

Although investigations using authentic viruses will support further evaluation of their risk for humans, the biosafety concern of artificially altering viral tropism should be taken seriously as antibodies elicited by current COVID-19 vaccination and anti-MERS-CoV RBD antibodies are inadequate to neutralize PDF-2180 and NeoCoV. Our results showed that broadly neutralizing betacoronavirus antibodies targeting the conserved stem helix retained their activity in inhibiting the entry of NeoCoV and PDF-2180. Thus, the clinical advancement of these antibodies could therefore enable therapeutic deployment in case of zoonotic spillover of these viruses.

Overall, we identified the functional receptor of a potential MERS-CoV ancestor in bats, which unexpectedly turned out to be ACE2, thereby enabling in-depth research of these important viruses presenting a possible risk of zoonotic emergence. Our study adds to the knowledge about the complex receptor use of coronaviruses, highlighting the importance of surveillance and research on these viruses to prepare for potential future outbreaks.

Methods

Receptor and virus sequences

The acquisition of sequences of 46 bat ACE2 and hACE2 was described in our previous study²⁹. The five bat DPP4 and hDPP4 sequences were directly retrieved from GenBank (human DPP4, NM_001935.3; Bat37, P. pipistrellus, KC249974.1) or extracted from whole genome sequence assemblies of the bat species retrieved from GenBank (Bat25, Sturnira hondurensis, GCA_014824575.2; Bat29, Mormoops blainvillei, GCA_004026545.1; Bat36, Aeorestes cinereus, GCA_011751065.1; Bat40, A. pallidus, GCA_007922775.1). The whole genome sequences of different coronaviruses were retrieved from GenBank. Accession numbers are as follows: MERS-CoV (NC_019843.3), Camel MERS-CoV KFU-HKU 19Dam (KJ650296.1), HKU4 (NC_009019.1), HKU5 (NC_009020.1), ErinaceusCoV/HKU31 strain F6 (MK907286.1), NeoCoV (KC869678.4), PDF-2180 (NC_034440.1), ErinaceusCoV/2012-174 (NC_039207.1), BtVs-BetaCoV/SC2013 (KJ473821.1), BatCoV/H. savii/Italy (MG596802.1), BatCoV HKU25 (KX442564.1), BatCoV ZC45 (MG772933.1), SARS-CoV-2 (NC_045512.2), NL63 (JX504050.1) and 229E (MT797634.1). All receptor and viral gene sequences used in this study were commercially synthesized by Genewiz, GenScript or GeneArt. The sources, accession numbers and sequences of the receptors and viruses are summarized in Supplementary Table 4.

Collection of SARS-CoV-2 antisera

All of the vaccinated sera were collected from volunteers at about 21 days after the third dose of the WHO-approved inactivated SARS-CoV-2 vaccine (CoronaVac, Sinovac). The median age of volunteers was 37 years. A total of 44% of participants were male and 56% were female. All of the volunteers were recruited by Sinovac. None of the participants had a history of previous SARS-CoV-2 infection, and none reported serious adverse events after vaccination. All of the volunteers were provided informed written consent forms, and the whole study was conducted according to the requirements of Good Clinical Practice of China. The procedures about human participants were approved by the Ethics Committee (seal) of Beijing Youan Hospital, Capital Medical University with an approval number of LL-2021-042-K.

Bioinformatic and computational analyses

Protein sequence alignment was performed using the MUSCLE algorithm by MEGA-X software (v.10.1.8) or ClustalW (https://www.genome.jp/tools-bin/clustalw). For phylogenetic analysis, nucleotide or protein sequences of the viruses were first aligned using the ClustalW and the MUSCLE algorithm, respectively. Phylogenetic trees were subsequently generated using the maximal-likelihood method in MEGA-X (1,000 bootstraps). The model and the other parameters used for phylogenetic analysis were applied following the recommendations after finding the best DNA/protein models using the software. The nucleotide similarity of coronaviruses was analysed using SimPlot (v.3.5.1) with a sliding windows size of 1,000 nucleotides and a step size of 100 nucleotides using gap-stripped alignments and the Kimura (two-parameter) distance model. Molecular dynamics prediction of the effect of residue mutations on protein–protein interactions was conducted by mCSM-PPI2 (http://biosig.unimelb.edu.au/mcsm_ppi2/)⁵².

Plasmids

Human codon-optimized sequences of various receptors and their mutants were cloned into a lentiviral transfer vector (pLVX-EF1a-Puro, Genewiz) with a C-terminal 3×Flag tag (DYKDHD-G-DYKDHD-I-DYKDDDDK). The DNA sequences of human codon-optimized NeoCoV S (GenBank: AGY29650.2), PDF-2180 S (GenBank: YP_009361857.1), HKU4 S (GenBank: AWH65899.1), HKU5 S (GenBank: YP_001039962.1), HKU31 S (GenBank: QGA70692.1), SARS-CoV-2 S (GenBank: YP_009724390.1) and MERS-CoV S (GenBank: YP_009047204.1) were cloned into the pCAGGS vector or pcDNA3.1(−) vector with a deletion of the last 13–15 residues (or 18 amino-acids in the SARS-CoV-2 S construct) or replacement by an HA tag (YPYDVPDYA) for higher VSV pseudotyping efficiency⁵³. Plasmids expressing coronavirus RBD–IgG-hFc fusion proteins were generated by inserting the coding sequences of NeoCoV RBD (380–585), PDF-2180 RBD (381–586), HKU4 (382–593), HKU5 RBD (385–586), HKU31 RBD (366–575), SARS-CoV-2 RBD (331–524) and MERS-CoV RBD (377–588) into the pCAGGS vector with an N-terminal CD5 secretion leading sequence (MPMGSLQPLATLYLLGMLVASVL) and C-terminal hFc tag for easy purification and detection. Plasmids expressing soluble bat and human ACE2 proteins (corresponding to residues 18–740 in hACE2) were constructed by inserting the ectodomain-coding sequences (containing a collectrin-like dimerization domain) into the pCAGGS vector with an N-terminal CD5 leading sequence and a C-terminal twin-strep tag and 3×Flag tandem sequences (WSHPQFEKGGGSGGGSGGSAWSHPQFEK-GGGRS-DYKDHDGDYKDHDIDYKDDDDK). Virus S proteins or receptor mutants or chimeras were generated by overlapping PCR. For DSP-based cell–cell fusion assays, the dual reporter split proteins were expressed by pLVX-EF1a-Puro-based plasmids carrying the rLucN(1–155)-sfGFP1–7(1–157) and sfGFP8–11(158–231)-rLucC(156–311) coding sequences (summarized in Supplementary Table 4), which were constructed in the laboratory based on a previous study^54,55. Plasmids expressing codon-optimized anti-ACE2 antibodies (H11B11; GenBank MZ514137 and MZ514138)³¹, B6⁴⁴, S2P6⁴⁵ and S2H14⁴⁰ were constructed by inserting the heavy-chain and light-chain coding sequences into the pCAGGS vector with N-terminal CD5 leader sequences, respectively. For anti-MERS-CoV nanobody–hFc fusion proteins, nanobody coding sequences were synthesized and cloned into the pCAGGS vector with N-terminal CD5 leader sequences and C-terminal hFc tags^38,39 (Supplementary Table 4). MERS-CoV-specific nanobodies and H11B11 coding sequences were synthesized by Sangon Biotech or Tsingke Biotechnology. For S trimer cryo-EM analysis, PDF-2180 S full-length gene (GenBank: YP_009361857.1) was synthesized by GeneArt, codon-optimized for expression in mammalian cells, cloned into pcDNA3.1(−) between XbaI and BamHI in frame with a Kozak’s sequence to direct translation and with the endogenous signal peptide. PDF-2180 S ectodomain gene was derived from the S full-length construct and comprised residues 1 to 1,286 followed by a foldon trimerization domain and a C-terminal His tag to assist purification. The full-length gene for expression of A. pallidus ACE2 (QJF77789) used in pseudotyped virus assays was synthesized by GeneScript and cloned in pCDNA3(+). The full-length gene for expressing hACE2 was described in one of our previous studies⁴¹.

Cell lines

HEK293T (CRL-3216), Vero E6 cells (CRL-1586), A549 (CCL-185), BHK-21 (CCL-10), Caco-2 (HTB-37) and the bat epithelial cell line Tb 1 Lu (CCL-88) were purchased from American Type Culture Collection (ATCC). The human hepatocellular carcinoma cell line Huh-7 (SCSP-526) were obtained from the Cell Bank of Type Culture Collection, Chinese Academy of Sciences. All these cells were cultured in Dulbecco’s modified Eagle medium, (DMEM, Monad) supplemented with 10% fetal bovine serum (FBS), 2.0 mM of l-glutamine, 110 mg l⁻¹ of sodium pyruvate and 4.5 g l⁻¹ of d-glucose. An I1-hybridoma (CRL-2700) cell line secreting a neutralizing mouse monoclonal antibody against the VSV glycoprotein (VSVG) was cultured in minimum essential medium with Earles’s balanced salts solution, 2.0 mM l-glutamine (Gibco) and 10% FBS. All cells were cultured at 37 °C in 5% CO₂ with regular passaging every 2–3 days. HEK Expi 293F cells (A14527, Thermo Fisher Scientific) expressing protein for cryo-EM analysis was cultured in OPM-293 CD03 serum-free medium (OPM, Shanghai OPM Biosciences).

Stable cell lines

Stable cell lines overexpressing different receptors were generated by lentivirus transduction and antibiotic selection. Specifically, the lentiviruses carrying the target genes were produced by co-transfection of lentiviral transfer (pLVX-EF1a-Puro,) and packaging plasmids pMD2G (Addgene, 12259) and psPAX2 (Addgene, 12260) into HEK293T cells using Lip2000 transfection reagent (Biosharp, BL623B). The lentivirus-containing supernatant was collected and pooled at 24 and 48 h after transfection. Cells were transduced by the lentivirus after 16 h in the presence of 8 μg ml⁻¹ polybrene. Stable cells were selected and maintained in the growth medium with puromycin (1 μg ml⁻¹). Cells selected for at least 10 days were considered to be stable cell lines and were used in various experiments.

Protein expression and purification

Antibodies, nanobody–hFc, soluble ACE2 ectodomain and RBD–hFc fusion proteins were expressed in HEK293T cells by transfecting the corresponding plasmids with GeneTwin reagent (Biomed, TG101-01) according to the manufacturer’s instructions. Then, 4 h after transfection, the culture medium of the transfected cells was replenished by SMM 293-TII Expression Medium (Sino Biological, M293TII). Protein-containing supernatants were collected every 2–3 days. Antibodies, nanobody–hFc and recombinant RBD–hFc proteins were captured by Pierce Protein A/G Plus Agarose (Thermo Scientific, 20424), washed with wash buffer (100 mM Tris/HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA), eluted with pH 3.0 glycine buffer (100 mM in H₂O) and then immediately balanced by 1/10 volume of UltraPure 1 M Tris-HCI, pH 8.0 (15568025, Thermo Fisher Scientific). The twin-strep-tag-labelled proteins were captured by Strep-Tactin XT 4Flow high-capacity resin (IBA, 2-5030-002), washed with washing buffer and eluted with buffer BXT (100 mM Tris/HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 50 mM biotin). Eluted proteins were concentrated and buffer-changed to PBS through ultrafiltration. Protein concentrations were determined using the Omni-Easy Instant BCA Protein Assay Kit (Epizyme, ZJ102). Purified proteins were aliquoted and stored at −80 °C. For cryo-EM analysis, NeoCoV RBD(380–588), PDF-2018 RBD(381–589) and Bat37ACE2(18–740) were synthesized and subcloned into the pCAGGS vector with a C-terminal twin-strep tag. In brief, these proteins were expressed by transient transfection of HEK Expi 293F cells (Gibco, Thermo Fisher Scientific, A14527) using polyethylenimine MAX (MW 40,000; Polysciences) or 293-free Transfection Reagent (Thermo Fisher Scientific). After 4 days, the supernatant was collected and cells were kept in culture for an additional 4 days, yielding two collections per transfection. The RBD and ACE2 protein samples were further purified by size-exclusion chromatography using a Superdex 75 10/300 Increase column (GE Healthcare) or a Superdex 200 10/300 Increase column (GE Healthcare) in 20 mM HEPES, 100 mM NaCl, pH 7.5. For the RBD–receptor complex (NeoCoV RBD–Bat37ACE2/PDF-2180 RBD–Bat37ACE2), NeoCoV RBD or PDF-2180 RBD was mixed with Bat37ACE2 at the ratio of 1.2:1, and incubated for 30 min on ice. The mixture was then processed for gel-filtration chromatography. The fractions containing the complex were collected and concentrated to 2 mg ml⁻¹.

For PDF-2180 S ectodomain expression, HEK293F cells were grown in suspension, with rotation at 130 rpm, in FreeStyle 293 Expression Medium (Life Technologies) at 37 °C in a humidified 8% CO₂ incubator. The wild-type PDF-2180 S ectodomain construct was transfected into 500 ml cultures with cells grown to a density of 1 million cells per ml using 293-free Transfection Reagent (Thermo Fisher Scientific). After 4 days, the supernatant was collected and cells were kept in culture for an additional 4 days, yielding two collections per transfection. The supernatants were clarified by centrifugation at 800g for 10 min, supplemented with 350 mM NaCl and 25 mM Tris-HCl pH 8.0, further centrifuged at 14,000g for 30 min and passed through a 1 ml His trap HP column (Cytiva) that had been equilibrated with binding buffer (25 mM Tris pH 7.4 and 350 mM NaCl). The PDF-2180 S ectodomain was eluted using a linear gradient of 500 mM imidazole. Purified protein was concentrated, buffer-exchanged into Tris-saline buffer (25 mM Tris pH 8, 150 mM NaCl) and quantified using absorption at 280 nm. Purified S glycoprotein was concentrated, and flash-frozen before negative staining and cryo-EM analysis.

Cryo-EM sample preparation and data collection

For cryo-EM sample preparation, the NeoCoV RBD–Bat37ACE2 or PDF-2018 RBD–Bat37ACE2 complexes were diluted to 0.5 mg ml⁻¹. Holy-carbon gold grids (Cflat R1.2/1.3 mesh 300) were freshly glow-discharged with a Solarus 950 plasma cleaner (Gatan) for 30 s. A 3 μl aliquot of the mixture complex was transferred onto the grids, blotted with filter paper at 16 °C and 100% humidity, and plunged into the ethane using the Vitrobot Mark IV (FEI). For these complexes, micrographs were collected at 300 kV using the Titan Krios microscope (Thermo Fisher Scientific), equipped with a K2 detector (Gatan), using SerialEM automated data collection software (v.3.8). Videos (32 frames, each 0.2 s, total dose 60 e⁻ Å⁻²) were recorded at a final pixel size of 0.82 Å with a defocus of between −1.2 and −2.0 μm. For PDF-2180 S trimer sample preparation, 3 μl of PDF-2180 S at 1 mg ml⁻¹ was applied to a 2/2 C-flat grid (Protochips) that had been glow discharged for 30 s at 20 mA. The grids were plunge-frozen into liquid ethane using an FEI Mark IV Vitrobot with a 6.5–7.5 s blot time at 100% humidity and 20 °C. Data were acquired using the FEI Titan Krios transmission electron microscope operated at 300 kV and equipped with a Gatan K2 Summit direct detector and Gatan Quantum GIF energy filter, operated in zero-loss mode with a slit width of 20 eV. Automated data collection was carried out using Leginon (v.3.1)⁵⁶ at a nominal magnification of ×130,000 with a super-resolution pixel size of 0.525 Å and a defocus range between −0.8 μm and −1.5 μm. The dose rate was adjusted to 8 counts per physical pixel per s, and each video was dose-fractionated in 50 frames of 200 ms.

Image processing

For the NeoCoV RBD–Bat37ACE2 complex, a total of 4,234 micrographs were recorded. For the PDF-2018 RBD–Bat37ACE2 complex, a total of 3,298 micrographs were recorded. Both datasets were similarly processed. Raw data were processed using MotionCor2 (v.1.3.0). The raw data were aligned and averaged into motion-corrected summed images, after which, defocus values for each micrograph were determined using Gctf. Next, particles were picked and extracted for two-dimensional alignment. Well-defined partial particles were selected for initial model reconstruction in Relion⁵⁷. The initial model was used as a reference for 3D classification. After refinement and post-processing, the overall resolution of the PDF-2018 RBD–Bat37ACE2 complex was up to 3.8 Å based on the gold-standard Fourier shell correlation (FSC; threshold = 0.143)⁵⁸. For the NeoCoV RBD–Bat37ACE2 complex, the C₂ symmetry was expanded before 3D refinement. Finally, the resolution of the NeoCoV RBD–Bat37ACE2 complex reached 3.5 Å. The quality of the local resolution was evaluated using ResMap (v.1.95)⁵⁹. For the PDF-2180 spike trimer, a total of 1,746 micrographs were collected. Video frame alignment, estimation of the microscope contrast-transfer function parameters, particle picking and extraction were performed using Warp (v.1.0.9)⁶⁰. Particles were extracted with a box size of 800 binned to 400 px². Reference-free 2D classification was performed using Relion (v.3.0) to select well-defined particles images before 3D classification without symmetry applied using a MERS-CoV cryo-EM map⁶¹ as an initial model in Relion. 3D refinements and CTF refinement (to refine per-particle defocus values) were performed in Relion (v.3.0)⁶². Particle images were processed using the Bayesian polishing procedure implemented in Relion (v.3.0)⁶³ before performing another round of 3D refinement and per-particle defocus refinement. Subsequently, 3D classification without alignment was performed using a mask focused on the N-terminal domain to improve its resolution, and the selected particles were processed for 3D refinement imposing C₃ symmetry using non-uniform refinement in cryoSPARC (v.3.3.1)⁶⁴, which yielded a final reconstruction of the PDF-2180 S at a resolution of 2.5 Å. Local-resolution estimation, filtering and sharpening were performed using CryoSPARC (v.3.3.1)⁶⁵. Reported resolutions are based on the gold-standard FSC of 0.143 criterion, and FSC curves were corrected for the effects of soft masking by high-resolution noise substitution^66,67.

Model building and refinement

The NeoCoV RBD–Bat37ACE2 complex structures were manually built into the refined maps using Coot (v.0.9.4)⁶⁸. The atomic models were further refined by positional and B-factor refinement in real space using Phenix (v.1.19)⁵⁹. For building the PDF-2018 RBD–Bat37ACE2 complex model, the refinement NeoCoV RBD–Bat37ACE2 complex structures were manually docked into the refined maps using UCSF Chimera (v.1.15) and further corrected manually by real-space refinement in Coot. The atomic models were further refined by using Phenix (v.1.19) and Rosetta (v.1.2.5)^69,70. For the PDF-2180 S model building, UCSF ChimeraX (v.1.1)⁷¹ and Coot were used to fit a MERS-CoV spike atomic model (PDB: 5W9J) into the PDF-2180 cryo-EM map. The model was subsequently manually rebuilt using Coot. N-linked glycans were hand built into the density where visible, and the models were rebuilt and refined using Rosetta (v.1.2.5)^69,70. All of the models were validated and analysed using MolProbity⁷² and Privateer⁷³. The figures were generated using ChimeraX. The datasets and refinement statistics are shown in Supplementary Table 1.

Immunofluorescence assay

Expression levels of receptors were evaluated by immunofluorescence assay detecting the C-terminal 3×Flag tags. Cells expressing receptors were seeded in the 96-well plate (poly-lysine-pretreated plates for HEK293 cell lines) at a cell density of about 1–5 × 10⁴ cells per well and cultured for 24 h. Cells were fixed with 100% methanol at room temperature for 10 min and then incubated with a mouse monoclonal antibody (M2) targeting the Flag-tag (Sigma-Aldrich, F1804) diluted in 1% BSA/PBS at 37 °C for 1 h. After one wash with PBS, cells were incubated with 2 μg ml⁻¹ of the Alexa Fluor 594-conjugated goat anti-mouse IgG (Thermo Fisher Scientific, A32742) diluted in 1% BSA/PBS at 37 °C for 1 h. The nuclei were stained with Hoechst 33342 (1:5,000 dilution in PBS). Images were captured with a fluorescence microscope (Mshot, MI52-N).

Pseudotype virus production and titration

Coronavirus S pseudotyped viruses (CoV-PSVs) were generated according to a previously described protocol based on a replication-deficient VSV-based rhabdoviral pseudotyping system (VSV-dG)⁷⁴. The VSV-G glycoprotein-deficient VSV co-expressing GFP and firefly luciferase (VSV-dG-GFP-fLuc) was rescued using a reverse genetics system in the laboratory and helper plasmids from Karafast. For CoV-PSV production, HEK293T or Vero E6 cells were transfected with the plasmids overexpressing the coronavirus spike proteins using the Lip2000 transfection reagent (Biosharp, BL623B). After 24–36 h, the transfected cells were transduced with VSV-dG-GFP-fLuc diluted in DMEM with 8 μg ml⁻¹ polybrene for 4 h at 37 °C with a 50% tissue culture infectious dose (TCID₅₀) of 1 × 10⁶ TCID₅₀ per ml. After 2–5 h infection, virus inoculum was removed, and the cells were washed twice with DMEM or PBS before addition of DMEM supplemented with anti-VSV-G-antibody-containing supernatant (from IL-mouse hybridoma) with 10–50-fold dilution to minimize background from the parental viruses. CoV-PSV-containing supernatants were collected at 24 h after the medium change, clarified at 4,000 rpm for 5 min to remove cellular debris, aliquoted and frozen at −80 °C. The TCID₅₀ of pseudotyped viruses was determined using a threefold-serial-dilution-based infection assay on HEK293T-bat40ACE2 cells for NeoCoV and PDF-2180 S pseudotypes, HEK293T-hDPP4 cells for MERS-CoV and HKU4 S pseudotypes, and BHK21-hACE2 cells for SARS-CoV-2 S pseudotypes. The TCID₅₀ was calculated according to the Reed–Muench method^75,76. Relative light unit values of ≥1,000 were considered to be positive. Viral titres (genome equivalents) of HKU5 and HKU31 without an ideal infection system were determined by quantitative PCR with reverse transcription using the HiScript II Q RT SuperMix (Vazyme, R223-01). RNA copies in the virus-containing supernatants were detected using primers for the VSV-N gene sequences (VSV-N-F, 5′-ACGGCGTACTTCCAGATGG-3′; VSV-N-R, 5′-CTCGGTTCAAGATCCAGGT-3′). For PDF-2180-, NeoCoV- and NeoCoV-T510F-neutralization assays, the S pseudotyped viruses were generated by transfecting the HEK293T cells using Lipofectamine 2000 (Life Technologies) according to the manufacturer’s instructions. After 5 h at 37 °C, DMEM supplemented with 20% FBS and 2% penicillin–streptomycin was added, and cells were incubated at 37 °C. The next day, cells were washed three times with DMEM and were transduced with VSVΔG-luc. After 2 h, virus inoculum was removed, and the cells were washed five times with DMEM before addition of DMEM supplemented with anti-VSV-G antibody (IL-mouse hybridoma supernatant diluted 1:25 (v/v)) to minimize parental background. After 18–24 h, the supernatants containing pseudotyped VSVs were collected, centrifuged at 2,000g for 5 min to remove cellular debris, filtered with a 0.45 µm membrane, concentrated ten times using a 30 kDa cut-off membrane (Amicon), aliquoted and frozen at −80 °C.

Pseudotyped virus entry assay

Cells obtained by plasmid transfection or lentiviral transduction were trypsinized and incubated with different pseudotyped viruses (1 × 10⁵ TCID₅₀ per 100 μl) in a 96-well plate (5 × 10⁴ per well) to allow attachment and entry simultaneously. For viruses without known susceptible cells, infections were performed using the same genome copies of a reference virus with a calculated TCID₅₀ titre (commonly 1 × 10⁵ TCID₅₀ per 100 μl). For TPCK-treated trypsin (Sigma-Aldrich, T8802) treatment, pseudotyped viruses in serum-free DMEM were incubated with 100 μg ml⁻¹ TPCK-treated trypsin for 10 min at 25 °C. The reactions were stopped by adding 100 μg ml⁻¹ soybean trypsin inhibitor (Sigma-Aldrich, T6414) in DMEM + 10% FBS. After 16–20 h, GFP images were acquired using a fluorescence microscope (Mshot, MI52-N), and intracellular luciferase activity was determined using the Bright-Glo Luciferase Assay Kit (Promega, E2620) and measured using the SpectraMax iD3 Multi-well Luminometer (Molecular Devices) or a GloMax 20/20 Luminometer (Promega).

Pseudotyped virus neutralization assays

For viral RBD or soluble ACE2 neutralization assays, serial dilutions of proteins were prepared in DMEM. Pseudotyped viruses (1 × 10⁵ TCID₅₀ per well) were mixed with 25 µl of each dilution and the mixtures were incubated for 30–45 min at 37 °C before addition to receptor-expressing cells seeded the day before at 2 × 10⁴ cells per well in a 96-well plate. After 16–20 h, the luciferase activity was measured in the same way as described for the pseudotype virus entry assay.

For pseudotyped virus neutralization with B6, SP6 and S2H14 IgGs, HEK293T cells were transfected with plasmids encoding for full-length hACE2 or Bat40 (A. pallidus) ACE2. In brief, HEK293T cells at 90% confluency were transfected using Lipofectamine 2000 (Life Technologies) and trypsinized at 5 h after transfection and seeded into 96-well plates at 50,000 cells per well overnight at 37 °C. For neutralizations, twofold serial dilutions of B6, S2P6 or S2H14 IgGs were prepared in DMEM. Then, 5 µl of the corresponding pseudotyped viruses was mixed with 20 µl of DMEM and 25 µl of each IgG dilution, and the mixtures were incubated for 45 min at 37 °C. Transfected HEK293T cells were washed three times with DMEM before adding 40 μl of the mixture containing virus and IgG. Then, 1 h later, 40 μl DMEM was added to the cells. After 17–20 h, 70 μl of One-Glo-EX substrate (Promega) was added to each well and incubated on a plate shaker in the dark. After 5–15 min incubation, the plates were read using the Biotek Neo2 plate reader. Measurements were taken in duplicate with biological replicates. Relative light units were plotted and normalized in Prism (GraphPad, v.8). Cells alone without pseudotyped virus was defined as 0% infection, and cells with virus only was defined as 100% infection.

Western blot

Cells washed with PBS were lysed with 2% TritonX-100/PBS containing 1 mM freshly prepared PMSF (Beyotime, ST506) on ice for 10 min. Cell lysates were clarified by centrifugation at 12,000g at 4 °C for 5 min, mixed with 1:5 (v/v) 5× SDS-loading buffer, and incubated at 95 °C for 5 min. To detect the HA tag or VSV-M on pseudotyped viruses, 1 ml of virus-containing supernatant was incubated with 8% PEG at 4 °C overnight, precipitated by centrifugation at 7,000g for 10 min at 4 °C and resuspended with 50 μl 1×SDS loading buffer. After SDS–PAGE and PVDF membrane transfer, the blots were blocked with 5% milk in PBS containing 0.1% Tween-20 (PBST) or TBST (20 mM Tris-HCl pH 8.0, 150 mM NaCl) supplemented with 0.05% Tween-20 at room temperature for 1 h. Primary antibodies against Flag (Sigma-Aldrich, F1804), HA (BioLegend, 901515), VSV-M [23H12] (Kerafast, EB0011) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (AntGene, ANT325) were added at a 1:10,000 dilution in PBST with 1% milk, or 1:250 dilution in TBST with 1% milk in the case of the stem-helix targeting monoclonal antibody B6, and incubated on a shaker at 4 °C. After three washes in PBST or TBST, the blots were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody AffiniPure goat anti-mouse IgG (H+L) (Jackson Immuno Research, 115-035-003), AffiniPure goat anti-rabbit IgG (H+L) (Jackson Immuno Research, 111-035-003, 1:10,000 dilution) or Alexa Fluor 680-conjugated goat anti-human secondary antibody (1:50,000 dilution, Jackson ImmunoResearch, 109-625-098) for 1 hour. The blots were subsequently washed three times before visualization using the LI-COR Odyssey CLx or the Omni-ECL Femto Light Chemiluminescence Kit (EpiZyme, SQ201) and a ChemiDoc MP Imaging System (Bio-Rad). Information about antibodies is provided in Supplementary Table 3. Uncropped and unprocessed full scans of gel source data are provided in Supplementary Figs. 1 and 2.

Coronavirus RBD–hFc live-cell binding assays

Recombinant coronavirus RBD–hFc proteins were diluted in DMEM at the indicated concentrations and then incubated with the cells for 1 h at 37 °C. After incubation, cells were washed once with DMEM and then incubated with 2 μg ml⁻¹ of Alexa Fluor 488-conjugated goat anti-human IgG (Thermo Fisher Scientific; A11013) diluted in Hanks’ balanced salt solution (HBSS) with 1% BSA for 1 h at 37 °C. Cells were washed twice with PBS and incubated with Hoechst 33342 (1:5,000 dilution in HBSS) for nucleus staining. Images were captured using a fluorescence microscope (MI52-N). For flow cytometry analysis, cells were detached by 5 mM of EDTA/PBS and analysed using the CytoFLEX Flow Cytometer (Beckman). The dead cells, as indicated by SSC/FSC, were excluded by gating. A total of 10,000 events in a gated live-cell population were analysed for all of the samples. The RBD–hFc-bound cells were gated as indicated by the fluorescence intensity compared with HEK293T control cells without receptor expression. The flow cytometry data were analysed using FlowJo (v.10). The gating strategy for flow cytometry analysis (Fig. 2b) is exemplified in Supplementary Fig. 3.

BLI binding assays

Protein binding affinities were determined using BLI assays performed on the Octet RED96 instrument (Molecular Devices). In brief, 20 μg ml⁻¹ RBD–hFc recombinant proteins were loaded onto protein A (ProA) biosensors (ForteBio, 18–5010) for 30 s. The loaded biosensors were then dipped into the kinetic buffer (PBST) for 90 s to wash out unbound RBD–hFc proteins. Subsequently, biosensors were dipped into the kinetic buffer containing soluble ACE2 ectodomain proteins with concentrations ranging from 0 to 500 nM for 120 s to record association kinetic and then dipped into kinetics buffer for 300 s to record dissociation kinetics. Kinetic buffer without ACE2 was used to define the background. The corresponding affinities were calculated with the Octet Data Analysis software (v.12.2.0.20) using curve-fitting kinetic analysis or steady-state analysis with global fitting. K_D,app values were reported because of the use of dimeric ACE2.

ELISA

To evaluate binding between viral RBD and ACE2 in vitro, 96-well immuno-plates were coated with ACE2 ectodomains at the indicated concentrations in BSA/PBS (100 μl per well) overnight at 4 °C. After three washes with PBST, wells were blocked by 3% skimmed milk/PBS at 37 °C for 2 h. Next, different concentrations of RBD–hFc proteins (1–9 μg ml⁻¹) diluted in 3% milk/PBST were added to the wells and incubated for 1 h at 37 °C. After extensive washes, wells were incubated with 1:2,000 diluted HRP-conjugated goat anti-human Fc antibody (Sigma-Aldrich, T8802) in 3% skimmed milk/PBS for 1 h at 37 °C. Finally, the substrate solution (Solarbio, PR1210) was added to the plates, and the absorbance at 450 nm after reaction termination was measured using the SpectraMax iD3 Multi-well Luminometer (Molecular Devices).

Cell–cell fusion assays

A cell–cell fusion assay based on DSPs was conducted on HEK293T cells stably expressing different receptors⁵⁴. Group A cells were transfected with spike and rLucN(1–155)-sfGFP1–7(1–157) expressing plasmids. Group B cells were transfected with spike and sfGFP8–11(158–231)-rLuc(156–311) expressing plasmids. Then, 12 h after transfection, cells from both groups were trypsinized and mixed into a 96-well plate at 8 × 10⁴ cells per well. Next, 24 h after transfection, cells were washed once with DMEM and then incubated with DMEM with or without 10–50 μg ml⁻¹ TPCK-treated trypsin for 10 min at room temperature. Then, 6 h later, nuclei were stained with Hoechst 33342 (1:5,000 dilution in HBSS) for 30 min at 37 °C. Fluorescent images were subsequently captured using a fluorescence microscope (MI52-N; Mshot). For measurements of live-cell luciferase activity, 20 μM of EnduRen live-cell substrate (Promega, E6481) was added to the cells in DMEM and incubated for at least 1 h before detection using the Varioskan LUX Multi-well Luminometer (Thermo Fisher Scientific).

Biosafety

The infection-related experiments were conducted in the State Key Laboratory of Virology, Wuhan University, strictly following the bio-safety regulations. Indeed, a VSV-based pseudotype virus system was used for all entry and neutralization assays, including new mutations in NeoCoV or PDF-2180 spikes that expand tropism for human cells. These pseudotyped viruses are non-replicating and non-pathogenic to humans and are commonly used by researchers to study the mechanisms of coronavirus entry and host range determination. The gain-of-function mutation, NeoCoV-T510F, was generated as we found that an Phe residue was present in PDF-2180 S at the equivalent position and its effect on binding affinity is consistent with our prediction models. We cautiously avoided extensively testing and showing the gain of function mutations on NeoCoV to expand its tropism on human cells.

Statistical analysis

Most infection and live-cell binding-related experiments were repeated between 2 and 5 times with around 3–4 biological repeats. In vitro protein-binding-related experiments were conducted with 2–3 technical repeats. Similar results were obtained in all of the experiments, and representative data are shown. Data are presented as mean ± s.d. or mean ± s.e.m. as specified in the figure legends. Most statistical analyses were conducted using GraphPad Prism (v.8) using unpaired two-tailed Student’s t-tests, unless otherwise specified. P < 0.05 was considered to be significant; *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.001.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The cryo-EM maps have been deposited at the Electron Microscopy Data Bank under the following accession numbers: EMD-32686 (NeoCoV RBD–Bat37ACE2 complex), EMD-32693 (PDF-2180 RBD–Bat37ACE2 complex) and EMDB-26378 (PDF-2180 S trimer). Atomic models corresponding to EMD-32686, EMD-32693 and EMDB-26378 have been deposited at the PDB and are available under the following accession numbers: 7WPO, 7WPZ and 7U6R, respectively. The accession numbers (NCBI Genbank or GISAID), protein sequences or species information of receptor, viral, antibody and reporter genes are provided in the Methods and Supplementary Table 4. All other data supporting the findings of this study are available with the Article and the Supplementary Information. Source data are provided with this paper.

References

Raj, V. S. et al. Dipeptidyl peptidase 4 is a functional receptor for the emerging human coronavirus-EMC. Nature 495, 251–254 (2013).
Article ADS CAS Google Scholar
Yang, Y. et al. Receptor usage and cell entry of bat coronavirus HKU4 provide insight into bat-to-human transmission of MERS coronavirus. Proc. Natl Acad. Sci. USA 111, 12516–12521 (2014).
Article ADS CAS Google Scholar
Lau, S. K. P. et al. Receptor usage of a novel bat lineage C betacoronavirus reveals evolution of Middle East respiratory syndrome-related coronavirus spike proteins for human dipeptidyl peptidase 4 binding. J. Infect. Dis. 218, 197–207 (2018).
Article CAS Google Scholar
Luo, C. M. et al. Discovery of novel bat coronaviruses in South China that use the same receptor as Middle East respiratory syndrome coronavirus. J. Virol. 92, e00116-18 (2018).
Article Google Scholar
Ithete, N. L. et al. Close relative of human Middle East respiratory syndrome coronavirus in bat, South Africa. Emerg. Infect. Dis. 19, 1697–1699 (2013).
Article Google Scholar
Cui, J., Li, F. & Shi, Z. L. Origin and evolution of pathogenic coronaviruses. Nat. Rev. Microbiol. 17, 181–192 (2019).
Article CAS Google Scholar
Li, W. et al. Broad receptor engagement of an emerging global coronavirus may potentiate its diverse cross-species transmissibility. Proc. Natl Acad. Sci. USA 115, E5135–E5143 (2018).
CAS Google Scholar
Latinne, A. et al. Origin and cross-species transmission of bat coronaviruses in China. Nat. Commun. 11, 4235 (2020).
Article ADS CAS Google Scholar
Ksiazek, T. G. et al. A novel coronavirus associated with severe acute respiratory syndrome. N. Engl. J. Med. 348, 1953–1966 (2003).
Article CAS Google Scholar
Zaki, A. M., van Boheemen, S., Bestebroer, T. M., Osterhaus, A. D. & Fouchier, R. A. Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia. N. Engl. J. Med. 367, 1814–1820 (2012).
Article CAS Google Scholar
Zhou, P. et al. A pneumonia outbreak associated with a new coronavirus of probable bat origin. Nature 579, 270–273 (2020).
Article ADS CAS Google Scholar
Chen, L. et al. RNA based mNGS approach identifies a novel human coronavirus from two individual pneumonia cases in 2019 Wuhan outbreak. Emerg. Microbes Infect. 9, 313–319 (2020).
Article CAS Google Scholar
Middle East Respiratory Syndrome Coronavirus (MERS-CoV) (WHO, 2012); http://www.who.int/emergencies/mers-cov/en/.
Mohd, H. A., Al-Tawfiq, J. A. & Memish, Z. A. Middle East respiratory syndrome coronavirus (MERS-CoV) origin and animal reservoir. Virol. J. 13, 87 (2016).
Article Google Scholar
Li, F. Receptor recognition mechanisms of coronaviruses: a decade of structural studies. J. Virol. 89, 1954–1964 (2015).
Article Google Scholar
Tortorici, M. A. & Veesler, D. Structural insights into coronavirus entry. Adv. Virus Res. 105, 93–116 (2019).
Article CAS Google Scholar
Li, W. et al. Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus. Nature 426, 450–454 (2003).
Article ADS CAS Google Scholar
Yeager, C. L. et al. Human aminopeptidase N is a receptor for human coronavirus 229E. Nature 357, 420–422 (1992).
Article ADS CAS Google Scholar
Williams, R. K., Jiang, G. S. & Holmes, K. V. Receptor for mouse hepatitis virus is a member of the carcinoembryonic antigen family of glycoproteins. Proc. Natl Acad. Sci. USA 88, 5533–5536 (1991).
Article ADS CAS Google Scholar
Walls, A. C. et al. Cryo-electron microscopy structure of a coronavirus spike glycoprotein trimer. Nature 531, 114–117 (2016).
Article ADS CAS Google Scholar
Hofmann, H. et al. Human coronavirus NL63 employs the severe acute respiratory syndrome coronavirus receptor for cellular entry. Proc. Natl Acad. Sci. USA 102, 7988–7993 (2005).
Article ADS CAS Google Scholar
Wang, Q. et al. Bat origins of MERS-CoV supported by bat coronavirus HKU4 usage of human receptor CD26. Cell Host Microbe 16, 328–337 (2014).
Article CAS Google Scholar
Anthony, S. J. et al. Further evidence for bats as the evolutionary source of Middle East respiratory syndrome coronavirus. mBio 8, e00373-17 (2017).
Article Google Scholar
Corman, V. M. et al. Characterization of a novel betacoronavirus related to middle East respiratory syndrome coronavirus in European hedgehogs. J. Virol. 88, 717–724 (2014).
Article Google Scholar
Lau, S. K. P. et al. Identification of a novel betacoronavirus (Merbecovirus) in Amur hedgehogs from China. Viruses 11, 980 (2019).
Article CAS Google Scholar
Corman, V. M. et al. Rooting the phylogenetic tree of Middle East respiratory syndrome coronavirus by characterization of a conspecific virus from an African bat. J. Virol. 88, 11297–11303 (2014).
Article Google Scholar
Geldenhuys, M. et al. A metagenomic viral discovery approach identifies potential zoonotic and novel mammalian viruses in Neoromicia bats within South Africa. PLoS ONE 13, e0194527 (2018).
Article Google Scholar
Menachery, V. D. et al. Trypsin treatment unlocks barrier for zoonotic bat coronavirus infection. J. Virol. 94, e01774-19 (2020).
Article Google Scholar
Yan, H. et al. ACE2 receptor usage reveals variation in susceptibility to SARS-CoV and SARS-CoV-2 infection among bat species. Nat. Ecol. Evol. 5, 600–608 (2021).
Article Google Scholar
Tortorici, M. A. et al. Structure, receptor recognition, and antigenicity of the human coronavirus CCoV-HuPn-2018 spike glycoprotein. Cell 185, 2279–2291 (2022).
Article CAS Google Scholar
Du, Y. et al. A broadly neutralizing humanized ACE2-targeting antibody against SARS-CoV-2 variants. Nat. Commun. 12, 5000 (2021).
Article ADS CAS Google Scholar
Raghuvamsi, P. V. et al. SARS-CoV-2 S protein:ACE2 interaction reveals novel allosteric targets. eLife 10, e63646 (2021).
Article CAS Google Scholar
Park, J. E. et al. Proteolytic processing of Middle East respiratory syndrome coronavirus spikes expands virus tropism. Proc. Natl Acad. Sci. USA 113, 12262–12267 (2016).
Article ADS CAS Google Scholar
Millet, J. K. & Whittaker, G. R. Host cell entry of Middle East respiratory syndrome coronavirus after two-step, furin-mediated activation of the spike protein. Proc. Natl Acad. Sci. USA 111, 15214–15219 (2014).
Article ADS CAS Google Scholar
Hoffmann, M. et al. SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor. Cell 181, 271–280 (2020).
Article CAS Google Scholar
Walls, A. C. et al. Unexpected receptor functional mimicry elucidates activation of coronavirus fusion. Cell 176, 1026–1039 (2019).
Article CAS Google Scholar
Corti, D. et al. Prophylactic and postexposure efficacy of a potent human monoclonal antibody against MERS coronavirus. Proc. Natl Acad. Sci. USA 112, 10473–10478 (2015).
Article ADS CAS Google Scholar
Wrapp, D. et al. Structural basis for potent neutralization of betacoronaviruses by single-domain camelid antibodies. Cell 181, 1004–1015 (2020).
Article CAS Google Scholar
Stalin Raj, V. et al. Chimeric camel/human heavy-chain antibodies protect against MERS-CoV infection. Sci. Adv. 4, eaas9667 (2018).
Article ADS CAS Google Scholar
Piccoli, L. et al. Mapping Neutralizing and immunodominant sites on the SARS-CoV-2 spike receptor-binding domain by structure-guided high-resolution serology. Cell 183, 1024–1042 (2020).
Article CAS Google Scholar
Walls, A. C. et al. Structure, function, and antigenicity of the SARS-CoV-2 spike glycoprotein. Cell 181, 281–292 (2020).
Article CAS Google Scholar
Bowen, J. E. et al. SARS-CoV-2 spike conformation determines plasma neutralizing activity elicited by a wide panel of human vaccines. Sci. Immunol. https://doi.org/10.1126/sciimmunol.adf1421 (2022).
Li, Y. et al. A humanized neutralizing antibody against MERS-CoV targeting the receptor-binding domain of the spike protein. Cell Res. 25, 1237–1249 (2015).
Article CAS Google Scholar
Sauer, M. M. et al. Structural basis for broad coronavirus neutralization. Nat. Struct. Mol. Biol. 28, 478–486 (2021).
Article ADS CAS Google Scholar
Pinto, D. et al. Broad betacoronavirus neutralization by a stem helix-specific human antibody. Science 373, 1109–1116 (2021).
Article ADS CAS Google Scholar
Low, J. S. et al. ACE2-binding exposes the SARS-CoV-2 fusion peptide to broadly neutralizing coronavirus antibodies. Science 377, 735–742 (2022).
Article ADS CAS Google Scholar
Zhou, P. et al. A human antibody reveals a conserved site on beta-coronavirus spike proteins and confers protection against SARS-CoV-2 infection. Sci. Transl. Med. 14, eabi9215 (2022).
Article CAS Google Scholar
Banner, L. R., Keck, J. G. & Lai, M. M. A clustering of RNA recombination sites adjacent to a hypervariable region of the peplomer gene of murine coronavirus. Virology 175, 548–555 (1990).
Article CAS Google Scholar
Cameroni, E. et al. Broadly neutralizing antibodies overcome SARS-CoV-2 Omicron antigenic shift. Nature 602, 664–670 (2022).
Article ADS CAS Google Scholar
McCallum, M. et al. Structural basis of SARS-CoV-2 Omicron immune evasion and receptor engagement. Science 375, 864–868 (2022).
Article ADS CAS Google Scholar
Ashkenazy, H. et al. ConSurf 2016: an improved methodology to estimate and visualize evolutionary conservation in macromolecules. Nucleic Acids Res. 44, W344–W350 (2016).
Article CAS Google Scholar
Rodrigues, C. H. M., Myung, Y., Pires, D. E. V. & Ascher, D. B. mCSM-PPI2: predicting the effects of mutations on protein-protein interactions. Nucleic Acids Res. 47, W338–W344 (2019).
Article CAS Google Scholar
Comparison of vesicular stomatitis virus pseudotyped with the S proteins from a porcine and a human coronavirus. J. Gen. Virol. 90, 1724–1729 (2009).
Article CAS Google Scholar
Ishikawa, H., Meng, F., Kondo, N., Iwamoto, A. & Matsuda, Z. Generation of a dual-functional split-reporter protein for monitoring membrane fusion using self-associating split GFP. Protein Eng. Des. Sel. 25, 813–820 (2012).
Article CAS Google Scholar
Wang, J., Kondo, N., Long, Y., Iwamoto, A. & Matsuda, Z. Monitoring of HIV-1 envelope-mediated membrane fusion using modified split green fluorescent proteins. J. Virol. Methods 161, 216–222 (2009).
Article CAS Google Scholar
Suloway, C. et al. Automated molecular microscopy: the new Leginon system. J. Struct. Biol. 151, 41–60 (2005).
Article CAS Google Scholar
Scheres, S. H. RELION: implementation of a Bayesian approach to cryo-EM structure determination. J. Struct. Biol. 180, 519–530 (2012).
Article CAS Google Scholar
Brown, A. et al. Tools for macromolecular model building and refinement into electron cryo-microscopy reconstructions. Acta Crystallogr. D 71, 136–153 (2015).
Article CAS Google Scholar
Afonine, P. V. et al. Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallogr. D 68, 352–367 (2012).
Article CAS Google Scholar
Tegunov, D. & Cramer, P. Real-time cryo-electron microscopy data preprocessing with Warp. Nat. Methods 16, 1146–1152 (2019).
Article CAS Google Scholar
Pallesen, J. et al. Immunogenicity and structures of a rationally designed prefusion MERS-CoV spike antigen. Proc. Natl Acad. Sci. USA 114, E7348–E7357 (2017).
Article CAS Google Scholar
Zivanov, J. et al. New tools for automated high-resolution cryo-EM structure determination in RELION-3. eLife 7, e42166 (2018).
Article Google Scholar
Zivanov, J., Nakane, T. & Scheres, S. H. W. A Bayesian approach to beam-induced motion correction in cryo-EM single-particle analysis. IUCrJ 6, 5–17 (2019).
Article CAS Google Scholar
Punjani, A., Zhang, H. & Fleet, D. J. Non-uniform refinement: adaptive regularization improves single-particle cryo-EM reconstruction. Nat. Methods 17, 1214–1221 (2020).
Article CAS Google Scholar
Punjani, A., Rubinstein, J. L., Fleet, D. J. & Brubaker, M. A. cryoSPARC: algorithms for rapid unsupervised cryo-EM structure determination. Nat. Methods 14, 290–296 (2017).
Article CAS Google Scholar
Rosenthal, P. B. & Henderson, R. Optimal determination of particle orientation, absolute hand, and contrast loss in single-particle electron cryomicroscopy. J. Mol. Biol. 333, 721–745 (2003).
Article CAS Google Scholar
Chen, S. et al. High-resolution noise substitution to measure overfitting and validate resolution in 3D structure determination by single particle electron cryomicroscopy. Ultramicroscopy 135, 24–35 (2013).
Article CAS Google Scholar
Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development of Coot. Acta Crystallogr. D 66, 486–501 (2010).
Article CAS Google Scholar
Frenz, B. et al. Automatically Fixing Errors in Glycoprotein Structures with Rosetta. Structure 27, 134–139 (2019).
Article CAS Google Scholar
Wang, R. Y. et al. Automated structure refinement of macromolecular assemblies from cryo-EM maps using Rosetta. eLife 5, e17219 (2016).
Article Google Scholar
Pettersen, E. F. et al. UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Sci. 30, 70–82 (2021).
Article CAS Google Scholar
Williams, C. J. et al. MolProbity: more and better reference data for improved all-atom structure validation. Protein Sci. 27, 293–315 (2018).
Article CAS Google Scholar
Agirre, J. et al. Privateer: software for the conformational validation of carbohydrate structures. Nat. Struct. Mol. Biol. 22, 833–834 (2015).
Article CAS Google Scholar
Whitt, M. A. Generation of VSV pseudotypes using recombinant DeltaG-VSV for studies on virus entry, identification of entry inhibitors, and immune responses to vaccines. J. Virol. Methods 169, 365–374 (2010).
Article CAS Google Scholar
Reed, L. J. & Muench, H. A simple method of estimating fifty per cent endpoints. Am. J. Epidemiol. 27, 493–497 (1938).
Article Google Scholar
Nie, J. et al. Quantification of SARS-CoV-2 neutralizing antibody by a pseudotyped virus-based assay. Nat. Protoc. 15, 3699–3715 (2020).
Article CAS Google Scholar

Download references

Acknowledgements

We thank X. Huang, X. Li and L. Chen for cryo-EM data collection at the Center for Biological Imaging (CBI) in the Institution of Biophysics, CAS; Y. Chen, Z. Yang, B. Zhou, X. Ou and Z. Zhou (Sartorius) for technical support on BLI and SPR; Q. Wang for the gift of the hedgehog ACE2 plasmid; Y. Fu for comments about manuscript writing; and the Beijing Taikang Yicai Foundation for their support of this work. H.Y. was supported by China NSFC projects (32188101, 32270164, and 32070160), Fundamental Research Funds for the Central Universities (2042020kf0024 and 2042022kf1188) and a Special Fund for COVID-19 Research of Wuhan University. X.W. was supported by the Strategic Priority Research Program (XDB29010000 and XDB37030000), CAS (YSBR-010), the National Key Research and Development Program (2020YFA0707500 and 2018YFA0900801), the Beijing Natural Science Foundation (Z210014), the Ten Thousand Talent Program and the NSFS Innovative Research Group (no. 81921005). The study funded in part by the National Institute of Allergy and Infectious Diseases (DP1AI158186 and 75N93022C00036 to D.V.), a Pew Biomedical Scholars Award (to D.V.), an Investigators in the Pathogenesis of Infectious Disease Awards from the Burroughs Wellcome Fund (to D.V.), Fast Grants (to D.V.), the University of Washington Arnold and Mabel Beckman cryo-EM center, the National Institute of Health grant S10OD032290 (to D.V.). D.V. is an Investigator of the Howard Hughes Medical Institute.

Author information

These authors contributed equally: Qing Xiong, Lei Cao, Chengbao Ma, M. Alejandra Tortorici

Authors and Affiliations

State Key Laboratory of Virology, Institute for Vaccine Research and Modern Virology Research Center, College of Life Sciences, TaiKang Center for Life and Medical Sciences, Wuhan University, Wuhan, China
Qing Xiong, Chengbao Ma, Chen Liu, Junyu Si, Peng Liu, Mengxue Gu, Chunli Wang, Lulu Shi, Fei Tong, Meiling Huang, Jing Li, Chufeng Zhao, Chao Shen, Yu Chen, Ke Lan & Huan Yan
CAS Key Laboratory of Infection and Immunity, National Laboratory of Macromolecules, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China
Lei Cao & Xiangxi Wang
Department of Biochemistry, University of Washington, Seattle, WA, USA
M. Alejandra Tortorici, Alexandra C. Walls & David Veesler
Howard Hughes Medical Institute, Seattle, WA, USA
Alexandra C. Walls & David Veesler
Department of Ecology, Tibetan Centre for Ecology and Conservation at WHU-TU, Hubei Key Laboratory of Cell Homeostasis, College of Life Sciences, Wuhan University, Wuhan, China
Huabin Zhao
Humabs BioMed SA, subsidiary of Vir Biotechnology, Bellinzona, Switzerland
Davide Corti
University of Chinese Academy of Sciences, Beijing, China
Xiangxi Wang

Authors

Qing Xiong
View author publications
You can also search for this author in PubMed Google Scholar
Lei Cao
View author publications
You can also search for this author in PubMed Google Scholar
Chengbao Ma
View author publications
You can also search for this author in PubMed Google Scholar
M. Alejandra Tortorici
View author publications
You can also search for this author in PubMed Google Scholar
Chen Liu
View author publications
You can also search for this author in PubMed Google Scholar
Junyu Si
View author publications
You can also search for this author in PubMed Google Scholar
Peng Liu
View author publications
You can also search for this author in PubMed Google Scholar
Mengxue Gu
View author publications
You can also search for this author in PubMed Google Scholar
Alexandra C. Walls
View author publications
You can also search for this author in PubMed Google Scholar
Chunli Wang
View author publications
You can also search for this author in PubMed Google Scholar
Lulu Shi
View author publications
You can also search for this author in PubMed Google Scholar
Fei Tong
View author publications
You can also search for this author in PubMed Google Scholar
Meiling Huang
View author publications
You can also search for this author in PubMed Google Scholar
Jing Li
View author publications
You can also search for this author in PubMed Google Scholar
Chufeng Zhao
View author publications
You can also search for this author in PubMed Google Scholar
Chao Shen
View author publications
You can also search for this author in PubMed Google Scholar
Yu Chen
View author publications
You can also search for this author in PubMed Google Scholar
Huabin Zhao
View author publications
You can also search for this author in PubMed Google Scholar
Ke Lan
View author publications
You can also search for this author in PubMed Google Scholar
Davide Corti
View author publications
You can also search for this author in PubMed Google Scholar
David Veesler
View author publications
You can also search for this author in PubMed Google Scholar
Xiangxi Wang
View author publications
You can also search for this author in PubMed Google Scholar
Huan Yan
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

H.Y., X.W. and D.V. conceived and designed the study. Q.X., L.C., C.M., M.A.T., C.L., J.S., P.L. and F.T. performed the experiments. Q.X., L.C., C.M., M.A.T., C.L., C.Z., D.V., H.Y. and X.W. analysed the data. M.A.T., H.Y., X.W., Q.X., L.C., C.M., C.L. and D.V. interpreted the results. D.C. provided unique reagents. H.Y. and X.W. wrote the initial drafts of the manuscript. H.Y., D.V., X.W., L.C., Q.X. and C.M. revised the manuscript. M.A.T., C.L., P.L., A.C.W., M.G., C.W., L.S., F.T., M.H., J.L., C.S., Y.C., H.Z. and K.L. commented on the manuscript.

Corresponding authors

Correspondence to David Veesler, Xiangxi Wang or Huan Yan.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature thanks the anonymous reviewers for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 Pseudotyped virus entry efficiency of six merbecoviruses in HEK293T cells stably expressing the indicated human ACE2, DPP4, or APN.

a, The expression levels of coronaviruses S proteins in transiently transfected HEK293T cells were analysed by Western blot detecting the C-terminal fused HA tags. GAPDH was used as a loading control. Experiments were performed in triplicates. b, Pseudotyped S virus entry efficiency of six merbecoviruses in HEK293T cells stably expressing hACE2, hDPP4, or hAPN. The dashed lines indicate the baseline of background signals (mean values of vector-only groups with MERS-CoV excluded). Blot representative of experiments of three independent transfections for generating each pseudoviruses. Data are presented as mean ± SEM (n = 4 biologically independent infected cells), representative of three independent infection assays with similar results. Two-tailed unpaired Student’s t-test; ****P < 0.001. RLU: relative light units.

Source data

Extended Data Fig. 2 Bat ACE2 orthologs tested in this study and their ability to support the entry of NeoCoV and PDF-2180.

a, Phylogenetic tree and assessment of the ability of a panel of ACE2 orthologs from 46 bat species to support NeoCoV and PDF-2180 S-mediated entry (relative to Bat40 ACE2). ACE2 orthologs deficient in supporting the entry of both viruses are highlighted in red. The GenBank accession numbers and protein sequences were summarized in Supplementary Table 4. b, Immunofluorescence assay detecting the 46 bat ACE2 orthologs C-terminal 3×Flag tag in the HEK293T stable cells. Mock indicates cells were transduced with lentiviruses produced by empty vector only. Data representative of three independent immunostaining assays with similar results.

Extended Data Fig. 3 Entry of indicated pseudoviruses in HEK293T cells stably expressing different bat ACE2 orthologs.

a-e, Entry efficiency of NeoCoV (a) PDF-2180 (b-c), HKU4 (d), and HKU5 (e) S pseudotyped viruses as indicated by GFP (a-b) or luciferase (c-e) intensity. Data representative of two independent infection assays. Data are presented as mean ± SD (biological triplicates for a-c and biological duplicates for d-e). f-g, TPCK-treated trypsin treatment (100 μg/ml) significantly enhanced NeoCoV (f) and PDF-2180 (g) S pseudotyped virus entry in HEK293T cells expressing different ACE2 orthologs. Data are presented as mean ± SD (biological triplicates of infected cells), representative of two independent experiments. Two-tailed unpaired Student’s t-test; ****P < 0.001. h, The expression level of ACE2 was evaluated by immunofluorescence detecting the C-terminal fused 3×Flag tag. Experiments performed twice with similar results and representative data are shown. i, Entry of SARS-CoV-2 and HKU31 S pseudotyped viruses into cells expressing hACE2 or hedgehog hgACE2. Representative data are presented as mean ± SD (biological triplicates of infected cells) examined over two independent experiments.

Source data

Extended Data Fig. 4 ACE2 and DPP4 receptor usage of different merbecoviruses.

a, Western blot analysis of the expression levels of ACE2 and DPP4 orthologs in HEK293T cells. Blot representative of immunoblotting based on biological independent duplicates of transfected cells. b, Evaluation of bat ACE2 expression level by immunofluorescence assay detecting the C-terminal 3×Flag tag. Data representative of two independent immunofluorescence assays based on two different transfection experiments. c-d, Pseudotyped virus entry (c) and RBD binding (d) of different CoVs on HEK293T cells expressing different ACE2 or DPP4 orthologs. The experiment was repeated independently twice with similar results, and representative data are shown.

Extended Data Fig. 5 ACE2 and Domain B dependent entry of NeoCoV, PDF-2180 S pseudotyped viruses in different cell types.

a-c, BHK-21, HEK293T, Vero E6, A549, Huh-7, and Tb 1 Lu cells were transfected with either Bat40ACE2 or Bat40DPP4. Expression, pseudotyped virus entry (GFP) (a), and luciferase intensity (b-c) were detected at 16 h post-infection. Data are presented as mean ± SD (biological triplicates). Data representative of two independent transfection and infection assays. d, Expression levels of the indicated chimeric viral S glycoproteins (anti-HA) in HEK293T whole cell lysates. GAPDH was used as a loading control. e, Chimeric viral S glycoproteins (anti-HA) and VSV-M level in pseudotyped viruses precipitated from virus-containing supernatant. Blots representative of two independent transfection assays for pseudovirus production.

Source data

Extended Data Fig. 6 Cryo-EM data processing and validation of the NeoCoV RBD-Bat37ACE2 complex and PDF-2180 RBD-Bat37ACE2 complex cryo-EM datasets.

Electron micrograph and flowchart for cryo-EM data processing, resolution estimation of the EM maps, density maps, and atomics models of NeoCoV RBD-Bat37ACE2 complex (a) and PDF-2180 RBD-Bat37ACE2 complex (b).

Extended Data Fig. 7 Structural and sequence analysis of different merbecoviruses.

a, Superimposition of overall structures of NeoCoV RBD-Bat37ACE2 (red) and PDF-2180 RBD-Bat37ACE2 complexes (blue). b, Structures of RBDs from different merbecoviruses. MERS-CoV (PDB:4KQZ), HKU4 (PDB:4QZV), HKU5 (PDB:5XGR). c, RBD sequence alignments were generated with ClustalW and rendered with ESPript. Identical residues are highlighted with a red background and similar residues are coloured red in blue boxes.

Extended Data Fig. 8 Characterization of NeoCoV and PDF-2180 RBM mutations enhancing human ACE2 binding and pseudovirus entry.

a, Analysis of binding of various concentrations of the NeoCoV and PDF-2180 RBD-hFc to HEK293T cells stably expressing hACE2. The SARS-CoV-2 RBD-hFc was used as a positive control. Mock indicates no protein added during protein incubation. Data representative of two live cell binding assays using independent preparations of RBD-hFc proteins. b, Analysis of binding kinetics of the interaction between NeoCoV RBD WT or T510F with hACE2 using BLI. Reported K_D values correspond to avidities due to the utilization of dimeric ACE2 constructs. Representative of two independent experiments. Unfitted curves can be found in Supplementary Figure 4b. c, Identification of PDF-2180 S mutations enhancing hACE2 binding. Sequence alignment of the NeoCoV and PDF-2180 RBMs and definition of the PDF-2180 mutants generated. d, Binding of PDF-2180 RBD-hFc mutants to hACE2- or Bat37ACE2-expressing HEK293T cells. Data presented were performed in two independent assays with similar results. e, Western blot analysis of the expression levels of PDF-2180 S protein mutants in HEK293T cells. f, Western blot analysis of the packaging efficiency of PDF-2180 S mutants in VSV pseudotyped viruses. g, Entry efficiency of PDF-2180 S mutant pseudotyped viruses into hACE2- or Bat37ACE2- expressing HEK293T cells. Experiments presented were independently performed twice with similar results for d-g. Representative data of g are presented as mean ± SD (n = 3 biological triplicates). Two-tailed unpaired Student’s t-test; *P < 0.05,**P < 0.01; ***P < 0.005, and ****P < 0.001

Source data

Extended Data Fig. 9 Cryo-EM data processing and validation of the PDF-2180 S cryo-EM dataset.

a, Representative electron micrograph (top) and class averages (bottom) of PDF-2180 S embedded in vitreous ice. Scale bar of the micrograph: 1000 Å. Scale bar of the class averages: 100 Å. b, Flowchart for cryo-EM data processing of PDF-2180 S trimer. c, Gold-standard Fourier shell correlation curve for the PDF-2180 S trimer. The 0.143 cut-off is indicated by a horizontal dashed grey line. d, Local resolution map for the PDF-2180 spike trimer.

Extended Data Fig. 10 S₁/S₂ cleavage and antibody neutralization efficiency of PDF-2180, NeoCoV, NeoCoV-T510F mutant, or MERS-CoV S pseudotyped viruses.

a, Western blot of VSV pseudotyped particles harbouring PDF-2180, NeoCoV, NeoCoV-T510F mutant or MERS-CoV S glycoproteins (detected using the B6 antibody) used for the B6 and S2P6 neutralization assays. Experiments were performed in duplicates, and representative data were shown. Mw: molecular weight ladder. Full-length S and S₂ subunit bands are indicated on the right-hand side of the blot. b-c, Inhibitory activity of 10 MERS-CoV RBD specific nanobodies at 10 μg/ml (b) and 10 SARS-CoV-2 specific anti-sera or two unvaccinated control sera at 50-fold dilution (c). Data are presented as mean ± SD (biological triplicates of two independent experiments). d-e, Dose-dependent inhibition of the entry of the indicated pseudotyped viruses by representative MERS-CoV RBD specific nanobodies (d) and SARS-CoV-2 specific anti-sera (e). Data are presented as mean (n = 2 biological duplicates), representative of two independent experiments.

Source data

Supplementary information

Supplementary Information

Supplementary Figs. 1–4 and Supplementary Tables 1–3.

Reporting Summary

Supplementary Table 4

Source data

Source Data Fig. 1

Source Data Fig. 2

Source Data Fig. 3

Source Data Fig. 4

Source Data Fig. 5

Source Data Extended Data Fig. 1

Source Data Extended Data Fig. 3

Source Data Extended Data Fig. 5

Source Data Extended Data Fig. 8

Source Data Extended Data Fig. 10

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Xiong, Q., Cao, L., Ma, C. et al. Close relatives of MERS-CoV in bats use ACE2 as their functional receptors. Nature 612, 748–757 (2022). https://doi.org/10.1038/s41586-022-05513-3

Download citation

Received: 24 January 2022
Accepted: 02 November 2022
Published: 07 December 2022
Issue Date: 22 December 2022
DOI: https://doi.org/10.1038/s41586-022-05513-3

This article is cited by

Recombination analysis on the receptor switching event of MERS-CoV and its close relatives: implications for the emergence of MERS-CoV
- Jarel Elgin Tolentino
- Spyros Lytras
- Kei Sato
Virology Journal (2024)
Isolation and characterization of a pangolin-borne HKU4-related coronavirus that potentially infects human-DPP4-transgenic mice
- Luo-Yuan Xia
- Zhen-Fei Wang
- Wu-Chun Cao
Nature Communications (2024)
Potential recombination between SARS-CoV-2 and MERS-CoV: calls for the development of Pan-CoV vaccines
- Qiao Wang
- Lujia Sun
- Shibo Jiang
Signal Transduction and Targeted Therapy (2023)
Targetable elements in SARS-CoV-2 S2 subunit for the design of pan-coronavirus fusion inhibitors and vaccines
- Liyan Guo
- Sheng Lin
- Guangwen Lu
Signal Transduction and Targeted Therapy (2023)
Broad host tropism of ACE2-using MERS-related coronaviruses and determinants restricting viral recognition
- Chengbao Ma
- Chen Liu
- Huan Yan
Cell Discovery (2023)

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Subjects

Abstract

Similar content being viewed by others

Main

Evidence of ACE2 use

Domain-B-mediated specific ACE2 binding

The structural basis for ACE2 binding

Determinants of human ACE2 recognition

Spike architecture and antigenicity

Discussion

Methods

Receptor and virus sequences

Collection of SARS-CoV-2 antisera

Bioinformatic and computational analyses

Plasmids

Cell lines

Stable cell lines

Protein expression and purification

Cryo-EM sample preparation and data collection

Image processing

Model building and refinement

Immunofluorescence assay

Pseudotype virus production and titration

Pseudotyped virus entry assay

Pseudotyped virus neutralization assays

Western blot

Coronavirus RBD–hFc live-cell binding assays

BLI binding assays

ELISA

Cell–cell fusion assays

Biosafety

Statistical analysis

Reporting summary

Data availability

References

Acknowledgements

Author information

Authors and Affiliations

Contributions

Corresponding authors

Ethics declarations

Competing interests

Peer review

Peer review information

Additional information

Extended data figures and tables

Supplementary information

Source data

Rights and permissions

About this article

Cite this article

Share this article

This article is cited by

Comments

Search

Quick links