Sarbecovirus RBD indels and specific residues dictating multi-species ACE2 adaptiveness

Abstract

Our comprehensive understanding of the multi-species ACE2 adaptiveness of sarbecoviruses remains elusive, particularly for those with various receptor binding motif (RBM) insertions/deletions (indels). Here, we analyzed RBM sequences from 268 sarbecoviruses categorized into four RBM indel types. We examined the ability of 20 representative sarbecovirus Spike glycoproteins (S) and derivatives in utilizing ACE2 from various bats and several other mammalian species. We reveal that sarbecoviruses with long RBMs (type-I) can achieve broad ACE2 tropism, whereas viruses with single deletions in Region 1 (type-II) or Region 2 (type-III) exhibit narrower ACE2 tropism. Sarbecoviruses with double region deletions (type-IV) completely lost ACE2 usage, which is restricted by clade-specific residues within and outside RBM. Lastly, we propose the evolution of sarbecovirus RBM indels and illustrate how loop lengths, disulfide, and residue determinants shape multi-species ACE2 adaptiveness. This study provides profound insights into the mechanisms governing ACE2 usage and spillover risks of sarbecoviruses.

To understand why sarbecoviruses display different abilities in using ACE2 from various species, the authors investigate the ACE2 orthologue tropism of these viruses and elucidate how they have evolved along three distinct paths through RBD indels and specific residues.

Introduction

The Severe Acute Respiratory Syndrome (SARS) outbreak and COVID-19 pandemic significantly raised global awareness of the zoonotic risks posed by sarbecoviruses^1,2. The Sarbecovirus subgenus, also known as lineage B β-coronaviruses, encompasses hundreds of SARS-related coronaviruses exhibiting varying RBM sequences^{3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13–14}. Most sarbecoviruses naturally infect Rhinolophus (horseshoe) bats, the primary natural reservoir^{15, 16–17}. Additionally, sarbecoviruses sharing high receptor binding domain (RBD) similarity to SARS-CoV-2 have been identified in pangolins, such as GX/P2V/2017 (GX_P2V), GD/1/2019, and GX_P1E^10,18. Sarbecoviruses exhibit extensive genetic diversity in RBM, likely arising from frequent recombination and the high selective pressure associated with inter-species host jumping in bats and pangolins, underscoring the risks of the emergence and outbreak of new human sarbecovirus^{19, 20, 21, 22–23}. However, many sarbecoviruses are known only as viral sequences and their ability to jump species and spillover to humans remains unclear.

Although ACE2 has been documented as a receptor for selected groups of setracovirus (e.g., NL63) and merbecoviruses (e.g., NeoCoV)^24,25, it has been primarily studied as the receptor for sarbecoviruses^1,9,24. Notably, not all sarbecoviruses have been confirmed to use ACE2 as their receptor, especially RBD clade 2 sarbecoviruses, which are proposed to utilize a distinct (yet unknown) receptor^3,4,15. Nevertheless, ACE2 usage has been demonstrated in most representative sarbecoviruses other than clade 2 sarbecoviruses^{3,6,7,16,22,26}. Structural analysis of ACE2 in complex with RBDs from various sarbecoviruses reveals a similar interaction mode, albeit with variations in specific residues involved in recognition^20,22,27,28. Specifically, the bridge-shaped RBM spanning amino acids (aa) 439-508 of SARS-CoV-2, formed by an extended loop connecting two β strands of the RBD core subdomain and stabilized with disulfide-bridging, interacts with ACE2 at two distinct patches^29,30. The interface on ACE2 mainly comprises the amino-terminal (N-terminal) α1 helix, along with limited interactions with the α2 helix and a loop connecting the β3 and β4 strands²⁹.

Given the importance of receptor recognition in determining host barriers, assessing multi-species ACE2 tropism for sarbecoviruses with distinct RBM features is crucial for understanding their zoonotic potential^31,32. Previous studies have provided substantial insight into distinct receptor preferences among bats and other mammalian species for SARS-CoV-1, SARS-CoV-2, GX_P2V, RaTG13, NeoCoV, and others^{20,22,25,33, 34, 35, 36, 37, 38, 39, 40, 41–42}. Varying entry-supporting abilities have also been observed in ACE2 orthologues from the same bat species but with different polymorphisms, particularly in residues involved in sarbecovirus binding^{21,43, 44, 45, 46, 47–48}.

Sarbecoviruses are commonly classified into several clades based on the RBD phylogeny and ACE2 usage^3,4,23. Despite sharing a similar RBD core subdomain, sarbecoviruses exhibit diversity in RBM sequences, particularly the presence of various indels in Region 1 (residues 443-450_SARS-CoV-2) or Region 2 (residues 470-491_SARS-CoV-2)^4,23. Clade 1 sarbecoviruses are all ACE2-using sarbecoviruses which can be further divided into subclades 1a, 1b, and 1c based on RBD phylogeny³. Most clade 1a (SARS-CoV-1 lineage) and 1b (SARS-CoV-2 lineage) sarbecoviruses have the longest RBM and lack RBM deletions of more than one amino acid³. Several sarbecoviruses with Region 1 or Region 2 single RBM deletions that are identified as clade 1b viruses, such as RshSTT182, RshSTT200⁴⁹, Rc-o319, and Rc-kw8⁵, which were recently found in Cambodia and Japan, respectively. Clade 1c sarbecoviruses include a subgroup of recently reported Asian sarbecoviruses carrying single Region 1 deletion, such as RmYN05, RaTG15, and RsYN04, which are also known as clade 4 sarbecoviruses in some studies^6,7,50. Here, we described these sarbecoviruses as 1c subclade considering their RBD phylogeny, geographical distribution, and ACE2 usage compared with 1a and 1b. Clade 2 sarbecoviruses are phylogenetically close to clade 1 and characterized by the two deletions (or indels) within RBM^{3,4,15,16,32,51, 52–53}. Clade 3 sarbecoviruses, such as BM48-31, Khosta-1/2, BtKY72, PDF-2386, and PRD-0038, discovered in Africa and Europe are considered closer to the sarbecovirus ancestors and all carry single deletions (indels) in RBM Region 1^{3,9,14,54, 55, 56–57}. Several clade 3 sarbecoviruses have been demonstrated as ACE2-using viruses, supporting this receptor usage as an ancestral trait of sarbecoviruses^3,28,51,58. Although proposed to have evolved from ACE2-using ancestors through the subsequent receptor switch^3,54, whether all clade 2 sarbecoviruses have lost ACE2 usage across all ACE2 orthologues remains an open question.

Our understanding of the key determinants affecting sarbecoviruses multi-species ACE2 adaptiveness and the factors restricting ACE2 usage remains incomplete. With an increasing number of sarbecoviruses identified with various single RBM indels, addressing the impact of these indels on multi-species ACE2 tropism becomes crucial. Moreover, sarbecoviruses with similar RBM deletion patterns exhibit marked differences in ACE2 tropism profile, emphasizing the role of critical RBD residues impacting multi-species ACE2 recognition beyond loop deletions^20,21,45,51.

In this study, we analyze the 268 sarbecovirus S sequences to delineate the overall RBM indel features and categorize them into four RBM indel types. Employing an ACE2 library consisting of 56 orthologues, we extensively evaluate cellular RBD binding and pseudovirus entry of 20 representative sarbecoviruses and various derivatives, encompassing RBM loop chimera and mutations. We elucidate multiple key determinants for ACE2 recognition and show how loop lengths, disulfide, and restrictive residues dictate ACE2 tropism and adaptiveness to ACE2 orthologues from various species. Our data lead to a more comprehensive understanding of the multi-species ACE2 adaptiveness across sarbecoviruses, as well as the coevolution of RBM indels and ACE2 adaptiveness.

Results

Four RBM indel types for sarbecoviruses

We retrieved 2318 Non-human β-coronavirus S glycoprotein sequences from the NCBI and GISAID databases, with 876 distinguished as sarbecovirus based on phylogenetic analysis. After reducing redundancy by excluding identical sequences and those highly similar to SARS-CoV-1 and SARS-CoV-2 (>99% identity), we obtained 265 sarbecovirus S sequences for further investigation, with 17 from pangolins, 1 Hipposideros bat and 247 Rhinolophus bats. We also included 3 representative human sarbecovirus sequences (SARS-CoV-1, SARS-CoV-2-WT, and SARS-CoV-2-Omicron BA.1) for analysis (Fig. 1a and Supplementary Data 1). Phylogenetic analysis based on RBD protein sequences revealed five sub-clades (RBD clades 1a, 1b, 1c, 2, and 3), with clade 2 accounting for the largest (Fig. 1b, Supplementary Fig. 1a). Multi-sequence alignment and Sequence Logo analysis highlighted three highly variable regions in RBMs, with Regions 1 and 2, but not Region 3, being the hot spots of loop indels (Fig. 1c, and Supplementary Data 2).

[See PDF for image]

Fig. 1

Phylogenetic and structural analysis of sarbecoviruses categorized into four indel types.

a Flow diagram illustrating the retrieval of 265 non-redundant S glycoprotein sequences from non-human sarbecoviruses and three additional human sarbecoviruses. b The RBD clade information of the 268 sarbecoviruses. c RBM sequence logo plot illustrating the three high variable regions. The SARS-CoV-2 residue numbering is shown. d Phylogenetic tree based on RBD amino acids sequence for the 268 sarbecoviruses (See also Supplementary Fig. 1a, b for the complete tree and selection strategy) and multi-sequence alignment of 35 selected sarbecoviruses displayed with four RBM indel types with relevant sequences were boxed with corresponding colors. The 20 sarbecoviruses for subsequent functional analysis are indicated in red. The two cysteines for disulfide bridging are highlighted in orange. e Summary of the deleted residue numbers in Region 1 and Region 2 compared to the SARS-CoV-2 RBM sequence. The counts of each deletion length among the 268 sarbecoviruses are indicated in parentheses. f The counts of S glycoprotein sequences of the four RBM indel types. g Distribution plot of RBD clades grouped by RBM indel types. h Analysis of the deleted residue numbers of Region 1 and Region 2 indels grouped by different RBM types. i Structural display of the two interaction patches in the SARS-CoV-2 RBD/hACE2 complex (6M0J). Selected residues involved in receptor recognition and the C480-C488 disulfide are indicated in the close-up views of the two interaction patches. j Superimposing of SARS-CoV-2 RBD (6M0J) with RBDs from the indicated sarbecoviruses belonging to each RBM indel type. The discrepancies in loop lengths are highlighted with red dashed ellipses. Disulfide bridges are indicated in orange.

For the five subclades, RBM sequences of 35 representative sarbecoviruses from each main phylogenetic branch were selected to demonstrate the diversity of RBM sequences. The selection criteria include RBD phylogeny, RBM sequence similarity, indel types, indel lengths, and preference for sequences with cryo-EM structure availability (Fig. 1d, and Supplementary Fig. 1a–d). To better investigate the impact of RBM indels on multi-species ACE2 adaptiveness, we categorized sarbecoviruses into four RBM indel types in addition to the clade-based classification. Specifically, RBM type-I describes most clade 1a and 1b sarbecoviruses without RBM deletions relative to SARS-CoV-1 or SARS-CoV-2 and are considered RBM prototypes, RBM type-II and type-III are viruses with single RBM deletions in Region 1 or Region 2, respectively, RBM type-IV viruses correspond to clade 2 sarbecoviruses with dual RBM deletions (Fig. 1d).

In analyses of RBM deletions of the 268 sequences, it was found that deletions of 1, 2, 3, 4, or 5 amino acids (aa) can occur in Region 1, and deletions of 1, 9, 13, or 14 aa can occur in Region 2 (Fig. 1e). Notably, the 5aa deletion in Region 1 is always accompanied by a 13 or 14 aa deletion in Region 2. Accordingly, four RBM types were defined based on the presence of deletions (>1aa) in Region 1 (type-II), Region 2 (type-III), or both (type-IV), while sequences with no deletion or only 1aa deletion were classified as type-I. This classification led to different subgroups of sarbecovirus (RBM types) compared to the clades based on RBD phylogeny (Fig. 1f–g). For example, all clade 1a sarbecoviruses (SARS-CoV-1 lineage) are RBM type-I, all clade 3 and clade 1c sarbecoviruses are RBM type-II, while clade 1b (SARS-CoV-2 lineage) sarbecoviruses can be divided into RBM type-I, II, and III (Fig. 1g). The deleted amino acid numbers in Region 1 and 2 of different sarbecoviruses are summarized based on different RBM types (Fig. 1h). After reducing redundancy by removing sarbecoviruses with high RBM sequence similarity, 20 representative sarbecoviruses were selected for subsequent functional profiling of multi-species ACE2 tropism, encompassing sarbecoviruses from all five subclades and covering all kinds of deleted numbers in either Region 1 or Region 2 (Supplementary Fig. 1b, c). Amino acid sequence identity analyses revealed that these sarbecoviruses share at least 57.07% RBD identity and 21.54% RBM identity in amino acid sequences, suggesting greater genetic variation in RBM than in RBD (Fig. 1d and Supplementary Fig. 1d).

From a structural perspective, the spatially proximate Region 1 and Region 3 loops form interaction patch 2, while the majority of residues in Region 2 loop contribute to interaction patch 1 (Fig. 1i). Interestingly, the conserved disulfide bridge for stabilizing loop in Region 2 is absent in RBM type-III and IV sarbecoviruses (Fig. 1d, i)⁵⁹. Superimposition of the solved or AlphaFold2-predicted RBDs with that of SARS-CoV-2 highlighted shortened extended loops due to specific deletions (Fig. 1j). Given that the two deletions are situated in critical RBM extensions for ACE2 interaction, their presence might impact multi-species ACE2 tropism.

We utilized a well-established RBD-hFc-based assay to assess the live cell receptor binding (Supplementary Fig. 2a–c). Due to the unavailability of the authentic sarbecovirus strains, we employed a dual reporter-based vesicular stomatitis virus (VSV) pseudotyping system assembled with sarbecovirus S glycoproteins to assess receptor functionality of various ACE2 orthologues (Supplementary Fig. 2d–f)⁴. The S glycoproteins from these sarbecoviruses incorporated into the pseudovirus particles were adjusted at comparable levels for entry assays (Supplementary Fig. 2g). The two different functional assays provide cross-validation and migrate the potential impact of other S components on viral entry efficiencies, such as NTD and S2.

Multi-species ACE2 usage profile

To illustrate a comprehensive ACE2 usage spectrum of each sarbecovirus, we examined 56 ACE2 orthologues from 51 bat and 5 representative non-bat mammalian species (Supplementary Fig. 3a, b). The bat species represent a broad genetic diversity spanning 11 bat families with global distribution and genetic diversity, including eight Rhinolophus bats geographically across Europe, Africa, and Asia (Fig. 2a, and Supplementary Fig. 3c)³³. Sequence analysis of these ACE2 orthologues exhibited significant variations in residues potentially involved in sarbecovirus interactions (Supplementary Fig. 3a, b). HEK293T cells stably expressing ACE2 orthologues were established and maintained with verified expression³³ (Supplementary Fig. 4). RBD binding assays based on the 20 selected sarbecoviruses with distinct RBM features were conducted to investigate their multi-species ACE2 binding spectra (Fig. 2a). According to the binding data, a subset of 17 sarbecoviruses were tested with the pseudovirus entry assays to evaluate their ability to utilize different ACE2 orthologs for entry. (Fig. 2b).

[See PDF for image]

Fig. 2

Multi-species ACE2 tropism of representative sarbecoviruses and the impact of critical RBM residues.

a, b Multi-species ACE2 usage spectra of sarbecoviruses of different indel types. RBD binding of 20 sarbecoviruses (a) and PSV entry of 17 sarbecoviruses (b) were assessed in HEK293T cells stably expressing the 56 ACE2 orthologues from bats and selected mammals. Rhinolophidae as the natural sarbecovirus reservoir is indicated in red. Asterisk: sarbecoviruses not tested in PSV entry experiments. c Efficiencies of selected sarbecovirus RBD-hFc recombinant proteins binding to HEK293T cells stably expressing the indicated ACE2 orthologs from their hosts or preferred species. d BLI analyses of binding kinetics of the soluble dimeric ACE2 ectodomains to immobilized viral RBD-hFc. e, f Structural demonstration (7DRV for RaTG13 and 7DDP for GX_P2V) (e) and S glycoprotein pseudovirus packaging efficiencies (f) of RaTG13 and GX_P2V swap mutants. For f, data are representative blots from two independent experiments. g, h Heat map displaying PSV entry efficiencies of RaTG13 (g) and GX_P2V (h) swap mutants (SARS-CoV-2 residue numbering) in HEK293T cells expressing the indicated ACE2 orthologues. PSV entry > 5% control ACE2 is considered an effective entry and the numbers of supportive orthologues are shown in parentheses. i Negative charged surface of the consensus ACE2 (based on 56 ACE2 orthologues) spatially proximate to residue 501 of RaTG13 and GX_P2V. The AlphaFold2 predicted structure based on consensus ACE2 sequence, with interaction predicted by HDOCK. RLU: Relative Luminescence Units; RFU: Relative Fluorescence Units. The amino acid usages of the selected surface-exposed residues are indicated. Heatmap plotted by the mean values (n = 2 independently infected cells) in a, b, g, and h, which are representative results out of two independent experiments. Source data are provided as a Source Data file.

These two assays displayed generally consistent ACE2 usage patterns, with a few exceptions. These apparent discrepancies in two different assays are commonly observed when testing other coronaviruses. Except for the RBD concentrations or viral titer/infectivities that may result in differences in assay sensitivity, other factors also contribute to the discrepancies. For example, the multi-step entry process is not always guaranteed by RBD binding alone, weak RBD binding can sometimes be sufficient for entry as the binding is a dynamic process and multivalent binding through spike trimmers in pseudoviral particles can increase binding avidity. Except for the type-IV RBM sarbecoviruses, the other sarbecoviruses displayed confirmed ACE2 usage, albeit with different ACE2 usage spectra. Several type-I RBM viruses, like SARS-CoV-1, SARS-CoV-2, and GX_P2V, efficiently use most orthologues, including human ACE2 (hACE2) (Fig. 2a). In contrast, RBM type-II or type-III sarbecoviruses generally showed narrower ACE2 tropism, and most are unable to use hACE2, representing a relatively low ACE2 adaptiveness to achieve broad ACE2 tropism compared to RBM type-I sarbecoviruses. Notably, GX_P1E, an RBM type-II pangolin sarbecovirus closely related to GX_P2V, with only 2aa (NY) deletion in RBM Region 1 displayed narrower ACE2 tropism, suggesting that 2aa reduction in Region 1 length is sufficient to affect the breadth of multi-species ACE2 tropism. Although PRD-0038 has been proposed as a sarbecovirus with broad ACE2 tropism among Rhinolophus bats²⁸, this virus, along with four other clade 3 sarbecoviruses (Khosta-2, BtKY72, PDF-2386, and BM48-31) with 2 to 4aa deletions in Region 1, display a narrow or moderate breadth of ACE2 tropism in our study (Fig. 2a). The RBD binding of the five sarbecoviruses to their optimal ACE2 orthologues, mostly from their hosts, was further analyzed through flow cytometry and Bio-layer interferometry (BLI) (Fig. 2c, d).

Notably, two close-related RBM type-I sarbecoviruses with identical RBM lengths, GX_P2V and RaTG13, displayed contrasting breadth of ACE2 tropism (Fig. 2a, b). We hypothesize that the narrow tropism of RaTG13 could be attributed to suboptimal residues restricting its binding with ACE2 orthologues. Supporting this hypothesis, the previously reported RaTG13 mutations, T372A^43,47,60 (removing a glycan at N370) and T403R/K^45,46 (enabling additional ACE2 interaction), significantly expanded the spectrum of ACE2 usage in our study (Supplementary Fig. 5a, b). Furthermore, analysis based on eight swap mutants exchanging RBM residues on 493, 498, 501, and 505 (SARS-CoV-2 numbering) reveals point mutation at position 501, in addition to position 372 and 403, can also markedly expand the breadth of ACE2 tropism of RaTG13, suggesting a relatively high muli-species ACE2 adaptiveness of this virus (Fig. 2e–h, Supplementary Fig. 5c)^3,43,61,62.

To investigate the mechanism of restrictive effect of D501, residue usage analysis was conducted on six ACE2 positions spatially close to position 501, which revealed an overall negatively charged surface among the 56 orthologues, thereby disfavoring D501 due to electrostatic repulsion (Fig. 2i). This hypothesis is further supported by similar phenotypes of RshSTT200, SARS-CoV-1, and SARS-CoV-2 carrying D/T mutations at this position (Supplementary Fig. 5d,e and Supplementary Fig. 6a–d). Since the N501Y mutation became dominant during the spread of SARS-CoV-2 in humans, we also compared the multi-species ACE2 usage spectra of SARS-CoV-1 and SARS-CoV-2 carrying N or Y at position 501_SARS-CoV-2⁶³. The Y mutation at this position slightly increased the number of acceptable ACE2 orthologues for SARS-CoV-2 while dramatically reducing acceptable ACE2 to only four orthologues for SARS-CoV-1 (Supplementary Fig. 6a–d). Structural analysis suggests Y487_SARS-CoV-1 may cause steric hindrance with local Y41_hACE2 and K353_hACE2, whereas the Y501_SARS-CoV-2 instead forms a π-π stacking interaction with Y41_hACE2 and cation-π interaction with K353. This indicates the virus-specific phenotypes can be attributed to the structural discrepancies of residues at equivalent positions (Supplementary Fig. 6e)⁶⁴. Similarly, the 501 T mutations in RmYN05 or Rc-o319 have no significant impact on their breadth of ACE2 tropism (Supplementary Fig. 5d, e). Moreover, the different efficiencies of SARS-CoV-2, SARS-CoV-2-N501Y, and SARS-CoV-2-Omicron BA.1 in using various ACE2 orthologues indicated the presence of residues, other than the 501 residues, affecting the ACE2 tropism of Omicron BA.1, a phenotype that also observed in authentic SARS-CoV-2 infection assays (Supplementary Fig. 7a, b). Fine mapping of the mutations in BA.1 underscores the critical contribution of residues at position 493 in affecting multi-species ACE2 tropism (Supplementary Fig. 7c–g).

Collectively, these data indicated that the RBM type-I sarbecoviruses exhibited superior multi-species ACE2 adaptiveness among the tested 20 sarbecoviruses, although the tropism breadth can be modulated by critical residues within or near the interaction surface, such as residues at positions 403, 501, and 493 (SARS-CoV-2 numbering).

The impact of RBM indels on ACE2 tropism

To investigate the impact of loop lengths on multi-species ACE2 tropism, we generated chimeras with specific loop substitutions in Region 1 or Region 2. These include SARS-CoV-2 with single deletions in each region and other sarbecoviruses carrying partial or entire loop substitutions with SARS-CoV-2 equivalent sequences (Fig. 3a). The VSV packaging efficiency of all S chimeras was validated by Western blot (Fig. 3b).

[See PDF for image]

Fig. 3

The impact of Region 1 and Regon 2 indels on multi-species ACE2 tropism of sarbecoviruses.

a Schematic illustration of the Region 1 and Region 2 substitutions in sarbecoviruses of different indel types. Light pink background: insertions corresponding to SARS-CoV-2 counterparts; gray background: deletions corresponding to sequences from type-II or type-III sarbecoviruses. (*) RBM swaps based on the indel boundaries. b Western blot detecting the S glycoprotein packaging efficiency in PSV particles. c, d RBD binding (c) and PSV entry (d) efficiencies of indicated sarbecoviruses and their swap mutants in HEK293T cells stably expressing the 56 ACE2 orthologues. e–h Disulfide bond in Region 2 is critical for multi-species ACE2 usage of sarbecoviruses. Schematic illustration of the Region 2 disulfide-related mutants based on SARS-CoV-2, Rc-o319, and their Region 2 substitution mutants (e), S glycoprotein packaging efficiency (f), RBD binding efficiency (g), and PSV entry efficiency (h) of SARS-CoV-2 and Rc-o319 mutants in HEK293T cells stably expressing the 56 ACE2 orthologues (dots in g and h). Dashed lines: background cut-off of RBD binding and PSV entry assays. For b and f, data are representative of two independent experiments with similar results. Heatmap plotted by the mean values (duplicates of independently bound or infected cells) in c and d, which are representative results out of two independent experiments. Two-sided Chi-squared test was used for statistical analysis of significance for g and h (n = 56, each point represents a value based on an ACE2 orthologue from a specific species). *P < 0.05, ****P < 0.0001. RLU Relative Luminescence Units, RFU Relative Fluorescence Units. Source data are provided as a Source Data file.

SARS-CoV-2 with a KVNY deletion in Region 1 (ΔRegion1*, relative to RshSTT200) displayed reduced multi-species ACE2 tropism but still can use ACE2 from a subset of species, including humans (Fig. 3c, d). However, the 9aa deletion in Region 2 (ΔRegion2*, relative to Rc-o319) abolished its ability to use any tested ACE2 orthologues. BM48-31 and Rc-o319 with regions substituted by SARS-CoV-2 equivalents can use more ACE2 orthologues, yet remain unable to achieve a broad tropism as RBM type-I sarbecoviruses. For RshSTT200 and Rc-o319, the increase of multi-species ACE2 tropism was not achieved by filling the deletions unless the entire RBM region was substituted with SARS-CoV-2 counterparts. This indicates the importance of side chains or local conformations apart from loop lengths (Fig. 3c, d). Notably, the highly conserved RBM disulfide bridge is present in Rc-o319-R2 but absent in Rc-o319-R2*, the importance of which was confirmed by the loss of infectivity of SARS-CoV-2 C480S and Rc-o319-R2 C462S mutants in using all tested ACE2 orthologues^59,65. However, introducing a disulfide to Rc-o319-R2* via K454C mutation remains insufficient to restore its ability to use ACE2, suggesting the presence of incompatible residues for Region 2-ACE2 interaction (Fig. 3e–h).

Substituting both R1 and R2 regions, which also introduced the featured disulfide, in the three RBM type-IV (clade 2) sarbecoviruses (ZC45, RmYN02, HKU3) failed to achieve any detectable ACE2 usage signal in both binding and pseudovirus entry assays (Fig. 3c, d). This indicates the presence of determinants other than loop deletions that restrict ACE2 usage in RBM type-IV sarbecoviruses⁴.

Clade 2-specific residues restricting ACE2 usage

In earlier attempts to identify determinants restrict ACE2 usage for clade 2 sarbecoviruses, we unexpectedly found that HKU3 and ZC45 remained unable to bind any ACE2 even with the entire RBM (aa439-508) replaced with SARS-CoV-2 RBM, indicating the presence of determinants restricting ACE2 recognition outside the RBM (Supplementary Fig. 8a). When comparing RBD sequences from 172 RBM type-IV (clade 2) sarbecoviruses with the other 96 ACE2-using sarbecoviruses, 22 clade 2-specific residues situated within or outside the RBMs were identified (Fig. 4a). It has been proposed that two residues (D496 and P502) within the Region 3 of RBM type-IV sarbecoviruses may restrict potential ACE2 interaction based on structural modeling, while the impact of these two residues and other clade 2-specific residues outside RBM on ACE2 recognition remains to be investigated using cell-based functional assays⁶⁶.

[See PDF for image]

Fig. 4

Fine mapping of clade 2-specific residues outside the RBM restricting ACE2 recognition.

a RBD residue usage (SARS-CoV-2 residue numbering) of 268 sarbecoviruses grouped by ACE2 dependence. Red fonts and arrows: strict clade 2-specific residues. Orange fonts and arrows: limited clade 2 specificity. b RBD amino acid sequences alignment of SARS-CoV-2, SARS-CoV-1, and HKU3. Red: HKU3-specific; Orange: shared by HKU3 and SARS-CoV-1 only. The boundaries of three fragments (Frag.) for subsequent mapping are indicated. c–g Fine mapping of residues restricting ACE2 usage outside the RBM. Mapping strategy for narrowing down the range of determinants restricting ACE2 recognition (c). Orange and green circles: capability of using hACE2 for entry (>1% of SARS-CoV-2 entry) and binding (background subtracted RFU > 0.2), respectively. Gray: unable to use hACE2. Underlines: two critical residues limiting ACE2 binding. PSV entry (d) and RBD binding (e) of the HKU3 mutants carrying SARS-CoV-2 corresponding sequences in HEK293T-hACE2. Yellow highlighted the mutants critical for analyzing ACE2-restricting determinants. Data representative of three independent experiments for d and e. RBD binding (f) and PSV entry (g) of HKU3 mutants with restored ACE2 binding affinity. h Dose-dependent inhibition of amplification of recombinant pcVSV-CoV carrying HKU3 S + NNSVGD + RBM_SARS-CoV-2 spike with h11B11 in Caco2-hACE2 cells. i, j pcVSV-HKU3 serves as a negative control. PSV entry (i) and RBD binding (j) of SARS-CoV-2 mutants carrying clade 2-specific restricting residues in HEK293T-hACE2. k BLI analyses of binding kinetics of the soluble dimeric ACE2 ectodomains to immobilized viral RBD-hFc. Scale bars, 200 μm. Heatmap plotted by the mean values (n = 3 independently infected cells) in f and g. Data are mean ± s.d. for d and i (n = 3 biologically independent cells), one-way ANOVA analysis, followed by Dunnett’s test for statistical analysis. *P < 0.05, ****P < 0.0001; NS, not significant (P > 0.05). Data representative of at least two independent experiments for h, i, and j with similar results. Source data are provided as a Source Data file.

To identify the determinants restricting ACE2 recognition, we conducted chimeric HKU3 S mutants with corresponding RBD sequences substituted with SARS-CoV-2 equivalents. HKU3 was selected as it represents the clade 2 sarbecovirus showing the highest SARS-CoV-2 RBD amino acid sequence identity (63.92%) in our study (Supplementary Fig. 1d). Sequence alignment of SARS-CoV-1, SARS-CoV-2 and HKU3 RBD displayed 16 HKU3-specific residues upstream of the RBM region (Fig. 4b). The subsequent functional dissection initiated from large fragment swaps and then proceeded to fine mapping of single residue combinations (Fig. 4b, c). In addition to SARS-CoV-2 RBM replacement (HKU3-RBM_SARS2), further substituting fragment A (aa 385-417) enabled HKU3 to use hACE2 for entry but remained deficient in binding hACE2 efficiently. Extending by fragment B (aa354-417) and fragment C (aa349-417) underscore the critical contribution of S349_SARS-CoV-2 for efficient ACE2 binding (Fig. 4d, e). Fine mapping of fragment A highlighted the crucial role of six clade 2-specific residues at positions 388, 394, 399, 401, 404, 405 (SARS-CoV-2 numbering) that restricting hACE2 usage, with S₃₄₉ + V₄₀₁G₄₀₄D₄₀₅ (S + VGD) being the minimal combination (Fig. 4c–e and Supplementary Fig. 8b). Similar results were obtained when testing two other clade 2 (RBM type-IV) sarbecoviruses, ZC45 and RmYN02 (Supplementary Fig. 8b, d–h). The expanded multi-species ACE2 usage spectra of HKU3-RBM_SARS2 carrying S + NNSVGD or S + VGD mutations were demonstrated by RBD binding and pseudovirus entry assays (Fig. 4f, g). Furthermore, the ACE2 dependence of HKU3 S mutant (HKU3-RBM S + NNSVGD) was verified by the efficient amplification of a propagation-competent recombinant VSV genetically encoding the S mutant in Caco2-hACE2 cells, which could be dose-dependently neutralized by the hACE2-specific antibody h11B11 (Fig. 4h and Supplementary Fig. 8c)⁶⁷.

The restrictive effect of these clade 2-specific residues was further demonstrated by the loss of ACE2 usage of SARS-CoV-2 carrying mutations at equivalent positions. SARS-CoV-2 carrying the corresponding mutants within or outside the RBM region (S349N, V401L, V401L + G404S + D405S, G496D + G502P, and P507A + Y508T) all underwent a significantly reduced efficiency in using hACE2 (Fig. 4i, j). The loss of hACE2 binding affinity of SARS-CoV-2 RBD harboring S349N or V401L point mutations was further verified by BLI assays (Fig. 4k). The restrictive effect of residues at these two positions can also observed in RmYN02 (clades 2) and SARS-CoV-1 (clade 1a), but is less pronounced in BM48-31 (clade 3) with higher RBD sequence divergence (Supplementary Fig. 8d–j).

Interestingly, unlike R403_SARS-CoV-2 which directly interacts with ACE2, the two clade 2-specific residues crucial for ACE2 binding, S349_SARS-CoV-2 and V401_SARS-CoV-2, are situated underneath the canonical RBM. Thus, S349N and V401L mutations (SARS-CoV-2 numbering) in HKU3 may slightly alter the RBM conformation due to their relatively larger side chains (Supplementary Fig. 8i). The resulting conformational shift may lead to the mismatch of critical residues for ACE2 interaction, thereby restricting ACE2 usage of clade 2 sarbecoviruses even when the entire RBM region is replaced by SARS-CoV-2 counterparts.

Coevolution of RBM indels and ACE2 adaptiveness

To trace the coevolution of sarbecoviruses RBM indels and ACE2 adaptation, we integrated biological functional data on multi-species ACE2 tropism with in silico analyses based on RBD clades, RBM types, and residue usages in the two indel-containing regions (Fig. 5a–c and Supplementary Fig. 9a, b). The breadth of multi-species ACE2 tropism and hACE2 fitness jointly affect the overall spillover risks of sarbecoviruses of different clades and RBM indel types (Fig. 5a, b). The scrutiny of the sequence features reveals an intriguing indel pattern in Region 1, characterized by one or two conserved and centrally located glycine (G), while Region 2 displays a complex indel rather than straightforward 9aa or 13/14aa deletions in RBM type-III and IV sarbecoviruses, respectively (Fig. 5c and Supplementary Fig. 9b). Notably, a potential evolutionary trace of Region 1 insertion was identified by the likely duplication of NY/NF sequences on the right side of the G (Fig. 5c).

[See PDF for image]

Fig. 5

The proposed coevolution of sarbecovirus RBM indels and their impact on multi-species ACE2 adaptiveness.

a The phylogenetic tree based on RBD amino acid sequences using maximum likelihood analysis (See also Supplementary Fig. 1a). The red lines mark the sarbecoviruses tested in this study. b The numbers of supportive ACE2 orthologues (data based on Fig. 2a with RLU > 2 × 10⁵) and hACE2 compatibility of the indicated sarbecoviruses. Coloring is based on RBD clades. c Region 1 sequence logoplot (SARS-CoV-2 numbering) of sarbecoviruses grouped by different indel types in each clade. The conserved D442, F/Y451 for defining the boundary of Region 1 are highlighted in black. The featured glycine (G) and H-bond-associated tyrosine (Y) are highlighted in red and orange, respectively. d The proposed evolutionary pathway of sarbecoviruses RBD clades and RBM indel types. e Details of Region 1 sequence changes along with the emergence of different clades during the evolution of sarbecoviruses in bats, pangolins, and humans. The Region 1 lengths in each group are indicated in blue. The conserved glycine (1 G) and the double G (2 G) in most clade1b sarbecoviruses are highlighted in red. f The RBD-ACE2 complex structures or models of sarbecoviruses with distinct Region 1 sequences. Region 1 loops of various lengths are highlighted in green without transparency, the featured G is marked in red. Close-up views of the interface between Region 1 loops and the indicated ACE2 orthologues are shown. ACE2 orthologs are displayed in light gray. Orange dashed lines: H-bonds. g PSV entry of SARS-CoV-2 R1 mutants in HEK293T expressing the indicated ACE2 orthologues. Data are presented by mean values (n = 3 independently infected cells). Dashed lines indicate the events with low confidence for d and e. Source data are provided as a Source Data file.

Combining these data, we proposed an evolutionary pathway delineating the emergence of various RBM types, highlighted by critical events driving the evolution (Fig. 5d). While the origin of the common ancestor of sarbecoviruses remains elusive, Africa/Europe sarbecoviruses (clade 3) maintained a relatively ancient state of RBM indel type-II³. The Asia sarbecoviruses underwent extensive evolution and developed into clade 1 and 2 sarbecoviruses with significant genetic diversity. These viruses evolved in three different directions with distinct ACE2 adaptiveness. Clade 1c sarbecoviruses maintained RBM type-II with limited genetic diversities based on currently known sequences. Clade 2 sarbecoviruses underwent R1 (-5aa) and R2 (-13/14aa) deletions (or indels) and switched to a non-ACE2 receptor, coupled with the emergence of clade 2-specific residues that further restricted ACE2 usage. In contrast, clade 1a and 1b viruses underwent Region 1 insertion (or indels with increased residue numbers), generating the longest (8aa) Region 1 for underpinning the interactions, thereby achieving the highest multi-species ACE2 adaptiveness. Some clade 1b viruses may have undergone further or different indels events in Region 1 (-2 to -4aa) and Region 2 (-9aa), resulting in RBM type-II (e.g. RshSTT200 and GX_P1E) and type-III (e.g. Rc-o319 and Rc-kw8), respectively, with moderate multi-species ACE2 adaptiveness. (Fig. 5d, e).

While Region 1 is shorter than Region 2, it exhibits more diverse sequence changes in ACE2-dependent sarbecoviruses, fine-tuning the species-specific ACE2 adaptiveness. By contrast, no further RBM indel change was observed among all 172 clade 2 sarbecoviruses. Intriguingly, despite high sequence variability in Region 1, only two out of the 268 sequences, BM48-31 and BB9904, lack a G in this region. In RBM type-I sarbecoviruses with 8aa length, most clade 1b viruses have two conserved G (2 G), whereas most 1a viruses have an SG or TG (Fig. 5c, e).

From a structural aspect, indels in Region 1 resulted in different loop lengths, with a G close to the turn of the loop conferring flexibility. The elongated Region 1 loop allows closer distance and potential H-bond formation with ACE2, reinforcing the patch 2 ACE2 interaction network along with the Region 3 loop, which explains the superior multi-species ACE2 adaptiveness of RBM type-I sarbecoviruses (Fig. 5f).

The importance of residue usage in Region 1 is supported by the data showing a dramatic decrease in multi-species ACE2 usage in SARS-CoV-2 carrying a G447Y mutation (Fig. 5g and Supplementary Fig. 10a–c). Additionally, we observed a reduced multi-species ACE2 adaptiveness of the “G-free” 4aa deletion mutant, SARS-CoV-2-ΔGGNY (Region 1: DSKVNY), compared to SARS-CoV-2-ΔKVNY (Region 1: DSGGNY) (as shown in Fig. 3), the former only recognize R.alc ACE2, similar to the phenotype of BM48-31 which also lacks G in region 1 (Fig. 5g and Fig. 2a, b). SARS-CoV-2-ΔGGNY may employ a similar ACE2 recognition mode as BM48-31, considering the importance of position 31 _R.alcACE2 for both viruses in a swap mutagenesis experiment based on R.alc ACE2 and its closely related orthologue, R.fer ACE2 (Supplementary Fig. 11a–d). This suggests that Region 1 deletion might be more tolerable for binding with multiple ACE2 orthologues if smaller side chains, like G, are maintained in this region.

Discussion

The persistent evolution of sarbecoviruses in Rhinolophus bats drives the emergence of sarbecovirus clades with varying RBM sequences. Frequent sequence changes, particularly indels within the RBM, pose challenges in predicting the potential of sarbecoviruses to cross species barriers and spillover to humans. To better investigate the influence of indels and other determinants on multi-species ACE2 usage, we propose a new RBM indel-based classification, categorizing all currently identified sarbecoviruses into four distinct RBM indel types.

Our functional data, combined with extensive sequence analyses, provide a comprehensive profile of the multi-species ACE2 tropism of sarbecoviruses belonging to specific clades and RBM indel types (Supplementary Fig. 12). Despite narrower ACE2 tropism, all tested sarbecoviruses carrying single RBM deletion in either Region 1 or Region 2 exhibited confirmed ACE2 usage, typically adapted to ACE2 from their hosts. Furthermore, we hypothesize an evolutionary history regarding the emergence of sarbecoviruses with distinct RBM indel types. As the number of sarbecovirus sequences from different clades increases, the intricate evolutionary history of sarbecoviruses remains to be updated or amended.

The driving force behind the emergence of different sarbecovirus Region 1 and Region 2 indels remains elusive. Virus recombination may play a crucial role, as the RBM or even Region 1 has been predicted as a breaking point for combinations between sarbecoviruses⁵⁴. Although various NTD-indels emerged in SARS-CoV-2 during the pandemic, no indels have been detected in RBM Region 1 or Region 2 in prevalent variants, suggesting a different evolution mechanism of RBM indels formation of various sarbecoviruses in bats compared to SARS-CoV-2 in humans⁶⁸.

Our results reveal a coevolution between sarbecovirus indels and multi-species ACE2 adaptiveness. Remarkably, the fine-tuning of RBM Region 1 through various indels and specific side chains promotes the emergence of sarbecoviruses with distinct multi-species ACE2 usage spectra. This could be attributed to the dispensable trait of Region 1 for interaction with a specific ACE2 orthologue (including hACE2), compensated by the Region 3 loop without indels. Additional interactions mediated by the extended Region 1 loop in RBM type-I sarbecoviruses might be crucial for achieving broader host tropism and facilitating host jumping. Interestingly, a conserved G within Region 1 suggests that greater flexibility or the absence of a large side chain in this site may confer certain evolutionary advantages. Comparatively, indels in Region 2 are less diverse than in Region 1 and generally have a more dramatic impact on ACE2 recognition, or even result in a receptor switch to a yet-unknown receptor. It is worth noting that limited multi-species ACE2 adaptiveness may restrict viral host jumping ability. However, this does not preclude the ability to use or adapt to hACE2, as is exemplified in Khosta-2 or other clade 1b or 3 sarbecovirus mutants described in this and prior studies^{3,28,45,57,58}.

Filling RBM deletions with SARS-CoV-2 counterparts does not guarantee a broader ACE2 usage spectrum and may sometimes result in reduced or lost ACE2 usage. This underscores the enhanced ACE2 adaptiveness achieved during adaptive evolution, with both length and residues being optimized in specific indels. Consequently, substituting the entire loop sometimes is necessary for achieving higher ACE2 compatibility. However, RBM type-IV sarbecoviruses, even after gap-filling or entire RBM substitution, remained unable to use any ACE2 orthologues, which led to the identification of clade-specific determinants outside the RBM that restrict ACE2 usage, likely due to adaptation to another receptor usage. Some critical RBD core residues are underneath the RBM, indirectly restricting ACE2 binding by affecting the RBM conformation. Future structural analysis could elucidate how these determinants affect receptor recognition.

A main limitation of this study is our analyses are based on the publicly available sarbecovirus sequences which may not recapitulate all authentic sarbecoviruses in nature. There could be additional RBM indel types yet to be discovered and evaluated. Although we tried to include most S glycoprotein sequences with distinct features, there might still be exceptions. For example, certain RBM type-II sarbecoviruses not tested might achieve a broad ACE2 tropism through compensatory interactions not contributed by RBM region 1. Moreover, we acknowledge that it is insufficient to predict tropism patterns solely based on RBD clades and RBM indel types, as the pattern can be affected by single restriction residues. Considering the current data of restrictive clade 2-specific RBD core residues are based on RBM type IV chimeras with SARS-CoV-2 RBM, other ways of acquiring ACE2 binding may exist for these viruses harboring different RBM sequences. It should also be noted that although ACE2 compatibility is a primary barrier for sarbecoviruses to cross species, efficient ACE2 recognition alone does not guarantee susceptibility at the animal level. Other factors, such as host protease, immune response, and viral replication efficiency, also affect host tropism, which can be verified by authentic viruses and in vivo studies in the future^50,69,70.

In conclusion, our RBM indel-type classification offers a more precise way to describe sarbecoviruses when combined with RBD phylogenetic information. Our functional ACE2 usage data elucidate the mechanism governing multi-species ACE2 usage and adaptiveness, shaped by multiple factors such as the presence and features of RBM loop deletions, RBM disulfide bridges, critical RBM residues for direct interaction, and restrictive residues within and outside the RBM. These findings establish a solid scientific foundation for risk assessment and viral surveillance to mitigate potential future zoonoses caused by these viruses.

Methods

Cell culture

BHK-21 (ATCC, CCL-10), Caco2 (ATCC, HTB-37), HEK293T (ATCC, CRL-1586) cells and their derivatives were maintained in Dulbecco’s modified eagle medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS), 2.0 mM L-Glutamine, 110 mg/L sodium pyruvate, and 4.5 g/L D-glucose. I1-Hybridoma (CRL-2700), secreting a monoclonal antibody targeting VSV glycoprotein (VSV-G), was maintained in Minimum Essential Medium with Earle’s salts and 2.0 mM L-Glutamine (MEM; Gibco). All cells were cultured at 37°C in 5% CO₂ with regular passage every 2-3 days.

Gene sequences

Sarbecovirus S glycoprotein sequences are retrieved from NCBI Virus and GISAID databases. The keywords used for sequence search include “Betacoronavirus”, “Sequence length between 1000-1400” “protein” and “NOT Homo sapiens” for NCBI and “bat”, “pangolin”, “civet”, coronaviruses for GISAID. A comprehensive collection of 2,318 Betacoronavirus S glycoprotein sequences was obtained. After extracting 876 sarbecovirus sequences through phylogenetic analysis using Geneious, the dataset was refined to 265 unique sequences for further analysis by excluding redundant entries. The ACE2 orthologues sequences were summarized by previous reports³³. Several additional ACE2 orthologues tested in this study include Rhinolophine Malayanus (Provided by Professor Huanbin Zhao, Wuhan University, China), Rhinolophus shameli (GenBank: MZ851782), Rhinolophus cornutus (GenBank: BCG67443.1), Rhinolophus sinicus isolate Rs-3357(GenBank: KC881004.1), Rhinolophus affinis (GenBank: QMQ39227.1), Manis javanica (Pangolin)(GenBank: XP_017505752.1), Mouse (GenBank: NP_001123985), Camelus (GenBank: XP_006194263), Civet (GenBank: Q56NL1), and Rhinolophus alcyone (GenBank: ALJ94035.1). Human Aminopeptidase N precursor (APN) (GenBank: NP_001141.2) was included as a negative control. The sources and accession numbers of the receptors and the 268 sarbecoviruses were summarized in Supplementary Data 1.

Bioinformatic analysis

Amino acid or nucleotide sequences from viruses or ACE2 orthologues were aligned using Mafft v7.450⁷¹. Phylogenetic trees were generated with IQ-Tree (version 2.0.6)⁷² using a Maximum Likelihood model with 1000 bootstrap replicates. Tree annotations were performed using iTOL (https://itol.embl.de/). Sequence identities were analyzed by Geneious Prime (https://www.geneious.com/) after being aligned by Mafft. The residue usage frequency (Sequence Logo analysis) was generated by the Geneious Prime.

Plasmids

The coding sequences of various coronavirus S glycoproteins and their derivatives were human codon optimized and cloned into the pCAGGS vector with C-terminal 18-amino acids replaced with an HA tag (YPYDVPDYA) for improving VSV pseudotyping efficiency and enabling detection^73,74. The plasmids for expressing ACE2 orthologues are constructed by inserting human codon-optimized ACE2 sequences into a lentiviral transfer vector (pLVX-IRES-puro) with a C-terminal 3×FLAG-tag (DYKDHD-G-DYKDHD-I-DYKDDDDK) for detection. The plasmids expressing the recombinant coronaviruses RBD human IgG Fc (RBD-hFc) fusion proteins were constructed by inserting RBD sequences corresponding to SARS-CoV-2 (aa331−524) containing an N-terminal CD5 secretion signal peptide (MPMGSLQPLATLYLLGMLVASVL) and C-terminal hFc-twin-strep tandem tags for purification and detection.

ACE2 stable cell lines

ACE2 expressing stable cell lines were established by lentiviral transduction^33,75. Briefly, lentivirus carrying the ACE2 genes was generated by co-transfecting pLVX-IRES-puro-ACE2 orthologues, pMD2G (plasmid no. 12259; Addgene), and psPAX2 (plasmid no. 12260; Addgene) into HEK293T cells. HEK293T or Caco2 cells were subsequently transduced with the lentiviruses, and the stable cells expressing ACE2 orthologues were selected in the presence of puromycin (1 μg/mL). The expression levels of ACE2 orthologues were assessed using an immunofluorescence assay³³. Briefly, HEK293T cells were fixed with 4% paraformaldehyde for 10 min at room temperature, permeabilized with 0.2% Triton X-100/PBS for 10 min, and blocked with 1% BSA for 30 min at 37 °C. Subsequently, the cells were incubated with M2 antibody (anti-FLAG-tag, catalogue no. F1804A-5MG; Sigma) at 4 °C for 1 hour. After three washes with PBS, the cells were treated with 2 μg/mL Alexa Fluor 594-conjugated goat anti-mouse IgG (catalogue no. A11032; Thermo Fisher Scientific). Nucleus was stained blue with Hoechst 33342 (1:5,000 dilution in PBS). Images were captured with a fluorescence microscope (MI52-N; Mshot). Relative fluorescence unit of Alexa Fluor 596 and Hoechst 33342 was quantified by Thermo Varioskan LUX. The expression of most ACE2 orthologues was also verified by Western Blot analysis in our previous reports³³.

Recombinant protein expression and purification

Recombinant RBD-hFc fusion proteins or ACE2 ectodomains (amino acid sequences 18-740 correspond to Human ACE2) fused with FLAG-strep-tag proteins were generated through transient transfection of HEK293T cells using Lipofectamine 2000. The transfected cells were cultured in SMM 293-TIS Expression Medium (Serum-free, without L-Glutamine) (Sino Biological). The supernatant, containing the recombinant proteins, was collected at 2, 4, and 6 days post-transfection, and the expression was confirmed through Western Blot analysis using the Goat Anti-Human IgG-Fc secondary Antibody (HRP) (SinoBiological Inc, SSA002) for RBD or the M2 antibody for ACE2. Protein purification was performed using Protein A/G Plus Agarose (Thermo Fisher Scientific) for RBD and Strep-Tactin®XT 4Flow® high capacity resin (IBA) for ACE2 ectodomains. The protein concentration was quantified using the BCA protein determination kit (EpiZyme) and SDS-PAGE with Coomassie blue staining was employed for analysis.

Live cell RBD binding assay

HEK293T cells stably expressing ACE2 were seeded in poly-D-lysine-treated 96-well plates. After 12 hours, with cells were incubated with RBD-hFc protein (2 μg/mL) in growth medium for 0.5 hours at 4 °C. Subsequently cells were washed with Hanks’ Balanced Salt Solution (with Ca²⁺ & Mg^2 + )(HBSS) twice, and then treated with Alexa Fluor 488-conjugated goat anti-human IgG (catalogue no. A11013; Thermo Fisher Scientific) at a concentration of 2 μg/mL in DMEM with 2% FBS for 30 minutes (min) at 4 °C. Hoechst 33342 (1:5,000 dilution in PBS) was utilized for nuclear staining. Following fixation with methanol at 25 °C for 5 minutes and washed once by HBSS, images were captured by fluorescence microscopy (MI52-N; Mshot), and the fluorescence intensity was analyzed using Thermo Varioskan LUX Alexa. RFUs of each sample were normalized by subjecting background signals in control cells (expressing hAPN) before analysis. The RFUs showing negative values after subtraction were presented as zero.

Flow cytometry

HEK293T cells stably expressing ACE2 orthologues (R.aff, R.sha, R.alc, R.mal, and R.cor) were cultured in 6-well plates for 12 hours. Cells were detached by 5 mM EDTA and washed twice by PBS, and then incubated with indicated proteins (RaTG13 RBD, RshSTT200 RBD, BM48-31 WT RBD, BM48-31 A480Y RBD, RmYN05 RBD, Rc-o319 RBD with hFc tags) at a concentration of 20 μg/mL for 30 min at 4 °C. Following three PBS washes, cells were stained with 488-conjugated goat anti-human IgG (1:1000, Alexa Fluor) for 30 min. Subsequently, flow cytometry analysis was performed using a CytoFLEX analyzer, collecting 10,000 events per sample. In a separate assay demonstrating the sensitivity of live cell binding assay, HEK293T cells expressing hACE2 were plated 12 hours before incubation with two-fold serial diluted SARS-CoV−2 RBD-hFc (from 20 μg/mL) for 30 min. After three PBS washes, cells were stained with 488-conjugated goat anti-human IgG (1:1000, Alexa Fluor) and subjected to Flow cytometry analysis. For the pseudoviruses entry assays, GFP expressing VSV pseudotypes was 10-fold serial diluted from 1 × 10⁶ TCID₅₀/mL. After 12 hours post-infection incubation, cells were washed with PBS and trypsinized for analysis. FlowJo V10 software was employed for data analysis. Gating strategy is illustrated in Supplementary Fig. 13.

Biolayer interferometry (BLI)

The Octet RED96 system (ForteBio, Menlo Park, CA) was employed to determine the apparent affinity (K_D, app, due to the potential dimerization of ACE2) between the RBD and ACE2. The buffer for analysis was phosphate buffer saline with 0.05% Tween20 (PBST). The RBD (10 μg/mL) was captured on ProA biosensors, followed by binding of ACE2 ectodomains at 2-fold serial dilutions ranging from 500 nM for 120 s, followed by dissociated in the PBST for additional 300 s. Analysis was conducted with curve-fitting kinetic with global fitting with a 1:1 binding mode using ForteBio Octet analysis software v12.2.0.20 (ForteBio, Menlo Park, CA). Mean K_D, app values were derived by averaging all binding curves that conformed to the theoretical fit with an R² value ≥ 0.95.

Pseudovirus production and entry assays

Pseudovirus incorporating coronaviruses S glycoproteins were produced using a vesicular stomatitis virus (VSV)-based system with slight modifications to a well-established protocol^73,76,77. In general, HEK293T cells were transfected with plasmids expressing S proteins through Lipofectamine 2000 (Biosharp, China). After 24 hours, the transfected cells were infected with VSV-ΔG-fLuc-GFP (1×10⁶ TCID50/mL) diluted in DMEM followed by a two-hour incubation on a shaker at 37 °C, the cells were replenished with DMEM containing anti-VSV-G monoclonal antibody (I1, 1 μg/mL). After 24 hours, the pseudovirus-containing supernatant was harvested, centrifuged at 13,523 × g for 5 min at 4 °C, and stored at −80 °C. For the viral entry assay, the HEK293T cell lines expressing various ACE2 orthologues were inoculated with pseudotyped viruses in DMEM with 10% FBS. In general, 3 × 10⁴ trypsinized cells were incubated with pseudovirus (1.5 × 10⁵ TCID50/100 μL) in a 96-well plate to allow cell attachment and pseudovirus entry. At 16-20 hpi (hours post infection), images of the infected cells were captured by a fluorescence microscope (MI52-N; Mshot). Intracellular luciferase activity was determined using a Bright-Glo Luciferase Assay Kit (Promega Corporation, E2620) and measured with a Thermo Varioskan LUX, SpectraMax iD3 Multi-well Luminometer (Molecular Devices) or a GloMax 20/20 Luminometer (Promega Corporation).

pcVSV-CoV amplification assay

Plasmids for rescuing propagation-competent (pc) VSV-CoV genetically encoding HKU3 and HKU3 S + NNSVGD + RBM spike glycoproteins were generated by replacing the fLuc coding sequences in the vector pVSV-ΔG-fLuc-GFP with coronavirus spike coding sequences⁷⁵. Reverse genetics experiments were conducted to rescue recombinant pcVSV-CoV according to a previously described protocol⁷⁷. BHK-21 cells of 80% confluence in a 6-well plate were infected with recombinant vaccinia virus expressing T7 RNA polymerase (VVT7, a gift from Mingzhou Chen’s lab, Wuhan University) at a multiplicity of infection (MOI) of 5 for 45 minutes at 37 °C. Subsequently, the vvT7 was removed by PBS wash, and the cells were transfected with the pVSV-dG-GFP-S plasmids and helper plasmids in a ratio of 5:3:5:8:1 (pVSV-dG-GFP-S: pBS-N: pBS-P: pBS-G: pBS-L). Supernatant containing pcVSV-CoV (P0) was collected at 48 hours post-transfection and filtered through a 0.22-μm filter to remove vvT7. Subsequently, Caco2 cells transfected with plasmids expressing VSV-G for 24 hours were further infected with P0 supernatant for VSV-G efficient virus amplification assisted by VSV-G, which generated the passage 1 (P1) supernatant. P1 viruses were further amplified in Caco2 cells stably expressing either hACE2 or an HKU3-customized viral receptor, in the presence of anti-VSV-G antibody (I1-Hybridoma supernatant), resulting in passage 2 (P2) viruses carrying HKU3 S + NNSVGD + RBM_SARS-CoV-2 or HKU3 S glycoproteins, respectively⁷⁸. For a typical virus amplification assay, 3 × 10⁴ trypsinized cells were incubated with pcVSV-CoV (1 × 10⁴ TCID50/100 μL) in a 96-well plate. For the hACE2 antibody h11B11 neutralization assay, cells were incubated with indicated concentrations of antibody for 1 hour at 4 °C, washed by PBS, and then incubated with viruses. GFP images were captured after 24 hours post-infection.

Authentic virus infection

The SARS-CoV-2 WT strain (IVCAS 6.7512) was provided by the National Virus Resource, Wuhan Institute of Virology, Chinese Academy of Sciences. The BA.1 strain (YJ20220223) was provided by Hubei Provincial Center for Disease Control and Prevention. SARS-CoV-2 authentic viruses-related experiments were conducted in ABSL-3 facility at Wuhan University with the approval from the Biosafety Committee of ABSL-3 lab. HEK293T cells expressing ACE2 orthologues were seeded in poly-lysine-treated 96-well plates (1.25 × 10⁵ cells/well). At 12 hours post seeding, SARS-CoV-2 strains (WT and Omicron BA.1) were introduced to different stable cells and incubated for 1-2 hours. Following a medium change to DMEM with 2% FBS, cells were cultured for 24 hours, fixed with methanol, and treated with anti-SARS-CoV-2 Nucleocapsid (N) antibody (catalogue no. 40143-MM05; Sino Biological) at 1:1000 for one hour at 37 °C. After PBS wash, cells were treated with a secondary antibody (Alexa Fluor 594) and Hoechst 33342 (1:10,000 dilution in PBS) for nuclei staining. Images were captured using a fluorescence microscope (MI52-N, Mshot, China).

Structural analysis

Protein structures were predicted by AlphaFold2 and HDOCK^{79, 80–81}. Briefly, AlphaFold2, implemented in ColabFold, was utilized with default settings for predicting the protein structures of various sarbecovirus RBDs and ACE2 orthologues. The top-ranked model was used for all subsequent analyses. The docking of the ACE2 ectodomain in complex with RBD was accomplished using HDOCK (v.1.1). Structural representations and analyses were carried out within ChimeraX (v.1.8). The hydrogen bonds and clashes between the displayed amino acids were analyzed using the H-bonds and clashes command. The following cryo-EM complex structures in the PDB database were also used for structural analysis in this study: human ACE2/SARS-CoV-2-RBD (Protein Data Bank 6M0J), human ACE2/SARS-CoV-1-RBD (3SCI), human ACE2/RaTG13-RBD (7DRV), human ACE2/GX_P2V-RBD (7DDP), human ACE2/SARS-CoV-2 alpha variant-RBD (7EKF), human ACE2/RshSTT200-RBD (7XBH), Rhinolophus alcyone ACE2/PRD-0038-RBD (8U0T), and RsYN04 RBD/antibody S43 (8J5J).

Western Blot

To examine the intracellular sarbecoviruses S glycoprotein expression levels, HEK293T cells were transfected with plasmids encoding the viral S glycoproteins fused with a C-terminal HA-tag. After 24 hours, cells were washed with PBS, lysed on ice for 10 min in 2% TritonX-100/PBS containing 1 mM PMSF (Beyotime, ST506). The cell lysates were clarified by centrifugation at 13,523 × g for 5 mins at 4 °C. The supernatants were mixed with 1:5 (v/v) 5× SDS-loading buffer and incubated at 95 °C for 5 min. For evaluating the S glycoprotein levels in pseudovirus (PSV) particles in the cultured medium, PSV was concentrated with a 30% sucrose cushion (30% sucrose, 15 mM Tris–HCl, 100 mM NaCl, 0.5 mM EDTA) at 20,000 x g for 1.5 hours at 4 °C. The concentrated PSV was then resuspended in 1×SDS loading buffer and incubated at 95 °C for 30 min. Following SDS-PAGE and PVDF membrane transfer, the blots were blocked with 10% milk in TBST containing 0.1% TBS (20 mM Tris-HCl pH 8.0, 150 mM NaCl) supplemented with 0.05% Tween-20 at room temperature for 1 hour. Primary antibodies targeting HA (MBL, MBL-M180-3), β-tubulin (Immunoway, YM3030), or VSV-M (Kerafast, EB0011) were applied at a 1:10,000 dilution in TBST with 1% milk. After three washes with TBST, blots were incubated with the secondary antibody Peroxidase AffiniPure Goat Anti-Mouse IgG (H + L) (Jackson Immuno Research, 115-035-003). Blots were further washed three times before chemiluminescence detection (SQ201, Yamei Biotech) using the ChemiDoc MP Imaging System (Bio-Rad). Uncropped scans of all blots in the Figures are supplied in the Source Data file, and the uncropped scans of those presented in the Supplementary Figs. were included at the end of the Supplementary Information file (Supplementary Fig. 14).

Geographical distribution of bat species

The global distribution data of bat species were obtained from the IUCN Red List of Threatened Species 2020, the base layer of the map (version 5.1.1) was sourced from Natural Earth, available at (https://www.naturalearthdata.com/downloads/110mcultural-vectors/). GeoScene Pro 21 was utilized to visualize and analyze the bat distribution data.

Statistical and reproducibility

Most experiments were independently conducted 2−4 times with similar results, each with 2-3 biologically independent replicates. Representative results were shown. Heat maps were plotted based on combined data of mean values (at least n = 2 biologically independent wells of cells) of RLU or RFU that were obtained in one or several experiments, with background (control cells expressing APN) signals subtracted. Most data are presented as means ± standard deviation (s.d.) as indicated in the figure legends. All statistical analyses were conducted using Prism 9 software (GraphPad). Two-tailed unpaired (Student’s) t-test was performed if only two conditions were compared. One-way ANOVA analysis, followed by Dunnett’s test, was employed for multiple comparisons. The association between the entry/binding efficiency and the presence of RBM disulfide was assessed using the two-sided chi-squared test. P < 0.05 was considered significant. *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.001; NS, not significant (P > 0.05).

Reporting summary

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

Acknowledgements

We are grateful to the funding support from National Key R&D Program of China (2023YFC2605500 to H.Y.), National Natural Science Foundation of China (NSFC) projects (82322041, 32270164, 32070160 to H.Y. and 323B2006 to C.B.M), Fundamental Research Funds for the Central Universities (2042023kf0191, 2042022kf1188, and 2042024kf1028 to H.Y.), Natural Science Foundation of Hubei Province (2023AFA015 to H.Y.), Hubei Provincial Research Center of Basic Discipline of Biology (to H.Y.) and TaiKang Center for Life and Medical Sciences (to H.Y.). We thank Huabin Zhao (Wuhan University) for his help in providing the coding sequences of many bat ACE2 orthologues. We thank Ming Guo (Wuhan University) for his help in conducting SARS-CoV-2 authentic viruses-related experiments in ABSL-3. We thank Qiang Ding (Tsinghua University) for his kind offer of several plasmids expressing mammalian ACE2 orthologues. We also want to express our gratitude to the core facilities and ABSL-3 laboratory of the Key Laboratory of Virology, Wuhan University.

Author contributions

Conceptualization, H.Y., and J.Y.S.; methodology, J.Y.S., Y.M.C, Y.H.S, M.X.G, C.L.W., C.B.M, P. L., Q.X., L.L.S., M.L.H., X.Yu., X.Yang., Z.X.M., Y.C.S.; data analysis, J.Y.S., Y.M.C., Y.H.S., M.X.G.; writing—original draft, H.Y., J.Y.S., Y.M.C.; writing—review & editing, H.Y., J.Y.S., Z.L.S.; supervision and funding acquisition, H.Y.

Peer review

Peer review information

Nature Communications thanks Frank Kirchhoff and the other anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its supplementary information files. Bat distribution shapefiles are available at https://www.iucnredlist.org/search?query=bat&searchType=species. The structural models of the ACE2 protein and RBD protein were downloaded from Protein Data Bank with PDB ID: human ACE2/SARS-CoV-2-RBD (6M0J), human ACE2/SARS-CoV-1-RBD (3SCI), human ACE2/RaTG13-RBD (7DRV), human ACE2/GX_P2V-RBD (7DDP), human ACE2/SARS-CoV-2 alpha variant-RBD (7EKF), human ACE2/RshSTT200-RBD (7XBH), Rhinolophus alcyone ACE2/PRD-0038-RBD (8U0T), and RsYN04 RBD/antibody S43 (8J5J). The accession numbers of 268 sarbecovirus spike proteins and 56 ACE2 orthologues used in this study are available in the Supplementary Data 1. The data generated in this study are provided in the Supplementary Information. Uncropped scans of all blots in the Supplementary Figs. were included at the end of the Supplementary Information file (Supplementary Fig. 14).  are provided with this paper.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s41467-024-53029-3.

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

References

1. Ksiazek, TG et al. A novel coronavirus associated with severe acute respiratory syndrome. N. Engl. J. Med.; 2003; 348, pp. 1953-1966.1:CAS:528:DC%2BD3sXjslajtbk%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/12690092]

2. Zhou, P et al. A pneumonia outbreak associated with a new coronavirus of probable bat origin. Nature; 2020; 579, pp. 270-273.2020Natur.579.270Z1:CAS:528:DC%2BB3cXksFKlsLg%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32015507][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7095418]

3. Starr, TN et al. ACE2 binding is an ancestral and evolvable trait of sarbecoviruses. Nature; 2022; 603, pp. 913-918.2022Natur.603.913S1:CAS:528:DC%2BB38XnvFSit7g%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35114688][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8967715]

4. Letko, M; Marzi, A; Munster, V. Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronaviruses. Nat. Microbiol; 2020; 5, pp. 562-569.1:CAS:528:DC%2BB3cXjvFyitL0%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32094589][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7095430]

5. Murakami, S et al. Isolation of Bat Sarbecoviruses, Japan. Emerg. Infect. Dis.; 2022; 28, pp. 2500-2503.1:CAS:528:DC%2BB3sXhvV2nsbc%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36417954][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9707576]

6. Guo, H et al. Identification of a novel lineage bat SARS-related coronaviruses that use bat ACE2 receptor. Emerg. Microbes Infect.; 2021; 10, pp. 1507-1514.1:CAS:528:DC%2BB3MXhslarsbnI [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34263709][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8344244]

7. Zhou, H et al. Identification of novel bat coronaviruses sheds light on the evolutionary origins of SARS-CoV-2 and related viruses. Cell; 2021; 184, pp. 4380-4391.e14.1:CAS:528:DC%2BB3MXhtlKrtbjK [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34147139][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8188299]

8. Wacharapluesadee, S et al. Evidence for SARS-CoV-2 related coronaviruses circulating in bats and pangolins in Southeast Asia. Nat. Commun.; 2021; 12, 2021NatCo.12.972W1:CAS:528:DC%2BB3MXjvFOgtbc%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33563978][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7873279]972.

9. Alkhovsky, S. et al. SARS-like Coronaviruses in Horseshoe Bats (Rhinolophus spp.) in Russia, 2020. Viruses14, 113 (2022).

10. Lam, TT-Y et al. Identifying SARS-CoV-2-related coronaviruses in Malayan pangolins. Nature; 2020; 583, pp. 282-285.2020Natur.583.282L1:CAS:528:DC%2BB3cXhtFCnu77L [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32218527]

11. Meta Djomsi, D. et al. Coronaviruses Are Abundant and Genetically Diverse in West and Central African Bats, including Viruses Closely Related to Human Coronaviruses. Viruses15, 337 (2023).

12. Murakami, S et al. Detection and Characterization of Bat Sarbecovirus Phylogenetically Related to SARS-CoV-2, Japan. Emerg. Infect. Dis.; 2020; 26, pp. 3025-3029.1:CAS:528:DC%2BB3MXkt1Cgsro%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33219796][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7706965]

13. Temmam, S et al. Bat coronaviruses related to SARS-CoV-2 and infectious for human cells. Nature; 2022; 604, pp. 330-336.2022Natur.604.330T1:CAS:528:DC%2BB38XnvFShsLg%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35172323]

14. Drexler, JF et al. Genomic characterization of severe acute respiratory syndrome-related coronavirus in European bats and classification of coronaviruses based on partial RNA-dependent RNA polymerase gene sequences. J. Virol.; 2010; 84, pp. 11336-11349.1:CAS:528:DC%2BC3cXhsVeisb%2FL [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20686038][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2953168]

15. Hu, B et al. Discovery of a rich gene pool of bat SARS-related coronaviruses provides new insights into the origin of SARS coronavirus. PLoS Pathog; 2017; 13, [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29190287][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5708621]e1006698.

16. Ge, X-Y et al. Isolation and characterization of a bat SARS-like coronavirus that uses the ACE2 receptor. Nature; 2013; 503, pp. 535-538.2013Natur.503.535G1:CAS:528:DC%2BC3sXhslSnsLrF [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24172901][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5389864]

17. Li, W et al. Bats are natural reservoirs of SARS-like coronaviruses. Science; 2005; 310, pp. 676-679.2005Sci..310.676L1:CAS:528:DC%2BD2MXhtFChsLjO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16195424]

18. Xiao, K et al. Isolation of SARS-CoV-2-related coronavirus from Malayan pangolins. Nature; 2020; 583, pp. 286-289.2020Natur.583.286X1:CAS:528:DC%2BB3cXhtlShtL%2FP [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32380510]

19. Guo, Z et al. SARS-CoV-2-related pangolin coronavirus exhibits similar infection characteristics to SARS-CoV-2 and direct contact transmissibility in hamsters. iScience; 2022; 25, 2022iSci..25j4350G1:CAS:528:DC%2BB38XhsVWgur7I [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35529312][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9065673]104350.

20. Niu, S et al. Molecular basis of cross-species ACE2 interactions with SARS-CoV-2-like viruses of pangolin origin. EMBO J; 2021; 40, 1:CAS:528:DC%2BB3MXht1KrtLjJ [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34018203][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8209949]e107786.

21. Li, P et al. The Rhinolophus affinis bat ACE2 and multiple animal orthologs are functional receptors for bat coronavirus RaTG13 and SARS-CoV-2. Sci Bull (Beijing); 2021; 66, pp. 1215-1227.2021SciBu.66.1215L1:CAS:528:DC%2BB3MXot1ClsLY%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33495713]

22. Liu, K et al. Binding and molecular basis of the bat coronavirus RaTG13 virus to ACE2 in humans and other species. Cell; 2021; 184, pp. 3438-3451.e10.1:CAS:528:DC%2BB3MXhtlKrt7bE [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34139177][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8142884]

23. Boni, MF et al. Evolutionary origins of the SARS-CoV-2 sarbecovirus lineage responsible for the COVID-19 pandemic. Nat. Microbiol; 2020; 5, pp. 1408-1417.1:CAS:528:DC%2BB3cXhsVyjt7vP [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32724171]

24. Hofmann, H et al. Human coronavirus NL63 employs the severe acute respiratory syndrome coronavirus receptor for cellular entry. Proc. Natl. Acad. Sci. USA.; 2005; 102, pp. 7988-7993.2005PNAS.102.7988H1:CAS:528:DC%2BD2MXkvF2isr8%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15897467][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1142358]

25. Xiong, Q et al. Close relatives of MERS-CoV in bats use ACE2 as their functional receptors. Nature; 2022; 612, pp. 748-757.2022Natur.612.748X1:CAS:528:DC%2BB38XjtVKht7zJ [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36477529][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9734910]

26. Tan, CCS et al. Genomic screening of 16 UK native bat species through conservationist networks uncovers coronaviruses with zoonotic potential. Nat. Commun.; 2023; 14, 2023NatCo.14.3322T1:CAS:528:DC%2BB3sXhtlGntrvE [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37369644][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10300128]3322.

27. Hu, Y et al. Host range and structural analysis of bat-origin RshSTT182/200 coronavirus binding to human ACE2 and its animal orthologs. EMBO J; 2023; 42, 1:CAS:528:DC%2BB3sXmtV2lug%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36519268][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9877840]e111737.

28. Lee, J et al. Broad receptor tropism and immunogenicity of a clade 3 sarbecovirus. Cell Host Microbe; 2023; 31, pp. 1961-1973.e11.1:CAS:528:DC%2BB3sXisVSntLrL [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37989312][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10913562]

29. Lan, J et al. Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor. Nature; 2020; 581, pp. 215-220.2020Natur.581.215L1:CAS:528:DC%2BB3cXoslOqtL8%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32225176]

30. Wu, K; Peng, G; Wilken, M; Geraghty, RJ; Li, F. Mechanisms of host receptor adaptation by severe acute respiratory syndrome coronavirus. J. Biol. Chem.; 2012; 287, pp. 8904-8911.1:CAS:528:DC%2BC38XjvVehtbs%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22291007][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3308800]

31. Lu, G; Wang, Q; Gao, GF. Bat-to-human: spike features determining ‘host jump’ of coronaviruses SARS-CoV, MERS-CoV, and beyond. Trends Microbiol; 2015; 23, pp. 468-478.1:CAS:528:DC%2BC2MXhtV2iu73E [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26206723][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7125587]

32. Wu, Z et al. A comprehensive survey of bat sarbecoviruses across China in relation to the origins of SARS-CoV and SARS-CoV-2. Natl Sci. Rev; 2023; 10, 1:CAS:528:DC%2BB2cXjsVaisbo%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37425654]nwac213.

33. 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; 2021; 5, pp. 600-608. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33649547]

34. Liu, Y. et al. Functional and genetic analysis of viral receptor ACE2 orthologs reveals a broad potential host range of SARS-CoV-2. Proc. Natl. Acad. Sci. USA. 118, e2025373118 (2021).

35. Zhang, Y et al. Cross-species tropism and antigenic landscapes of circulating SARS-CoV-2 variants. Cell Rep; 2022; 38, 110558.1:CAS:528:DC%2BB38XmvVyrsbc%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35303476][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8898742]

36. Latinne, A et al. Origin and cross-species transmission of bat coronaviruses in China. Nat. Commun.; 2020; 11, 2020NatCo.11.4235L1:CAS:528:DC%2BB3cXhs12rtbnO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32843626][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7447761]4235.

37. Wu, L et al. Broad host range of SARS-CoV-2 and the molecular basis for SARS-CoV-2 binding to cat ACE2. Cell Discov; 2020; 6, 68.1:CAS:528:DC%2BB3cXhvFKnt7fF [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33020722][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7526519]

38. Li, Y. et al. SARS-CoV-2 and Three Related Coronaviruses Utilize Multiple ACE2 Orthologs and Are Potently Blocked by an Improved ACE2-Ig. J. Virol. 94, e01283–20 (2020).

39. Li, S et al. Cross-species recognition and molecular basis of SARS-CoV-2 and SARS-CoV binding to ACE2s of marine animals. Natl Sci. Rev; 2022; 9, 1:CAS:528:DC%2BB3sXmslGnur0%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36187898][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9517163]nwac122.

40. Mou, H et al. Mutations derived from horseshoe bat ACE2 orthologs enhance ACE2-Fc neutralization of SARS-CoV-2. PLoS Pathog; 2021; 17, 1:CAS:528:DC%2BB3MXps1Wns7g%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33836016][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8059821]e1009501.

41. Conceicao, C et al. The SARS-CoV-2 Spike protein has a broad tropism for mammalian ACE2 proteins. PLoS Biol; 2020; 18, 1:CAS:528:DC%2BB3MXpsVSg [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33347434][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7751883]e3001016.

42. Chen, J. et al. The binding and structural basis of fox ACE2 to RBDs from different sarbecoviruses. Virol. Sin. https://doi.org/10.1016/j.virs.2024.06.004 (2024).

43. Li, P et al. Effect of polymorphism in Rhinolophus affinis ACE2 on entry of SARS-CoV-2 related bat coronaviruses. PLoS Pathog; 2023; 19, 1:CAS:528:DC%2BB3sXjtFSjtrc%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36689489][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9904459]e1011116.

44. Guo, H. et al. Evolutionary Arms Race between Virus and Host Drives Genetic Diversity in Bat Severe Acute Respiratory Syndrome-Related Coronavirus Spike Genes. J. Virol. 94, e00902–20 (2020).

45. Zech, F et al. Spike residue 403 affects binding of coronavirus spikes to human ACE2. Nat. Commun.; 2021; 12, 2021NatCo.12.6855Z1:CAS:528:DC%2BB3MXisFyht77F [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34824253][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8617078]6855.

46. Wang, Q et al. Determinants of species-specific utilization of ACE2 by human and animal coronaviruses. Commun Biol; 2023; 6, 1051.1:CAS:528:DC%2BB3sXitFOru77L [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37848611][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10582019]

47. Kang, L et al. A selective sweep in the Spike gene has driven SARS-CoV-2 human adaptation. Cell; 2021; 184, pp. 4392-4400.e4.1:CAS:528:DC%2BB3MXhsF2gtrnO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34289344][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8260498]

48. Ou, X et al. Host susceptibility and structural and immunological insight of S proteins of two SARS-CoV-2 closely related bat coronaviruses. Cell Discov; 2023; 9, 78.1:CAS:528:DC%2BB3sXhsFOls73N [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37507385][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10382498]

49. Delaune, D et al. A novel SARS-CoV-2 related coronavirus in bats from Cambodia. Nat. Commun.; 2021; 12, 2021NatCo.12.6563D1:CAS:528:DC%2BB3MXisVGktbbE [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34753934][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8578604]6563.

50. Guo, H et al. ACE2-Independent Bat Sarbecovirus Entry and Replication in Human and Bat Cells. MBio; 2022; 13, [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36409074]e0256622.

51. Roelle, SM; Shukla, N; Pham, AT; Bruchez, AM; Matreyek, KA. Expanded ACE2 dependencies of diverse SARS-like coronavirus receptor binding domains. PLoS Biol; 2022; 20, e3001738.1:CAS:528:DC%2BB38XitVOrs73N [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35895696][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9359572]

52. Hu, D et al. Genomic characterization and infectivity of a novel SARS-like coronavirus in Chinese bats. Emerg. Microbes Infect.; 2018; 7, [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30209269][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6135831]154.

53. Ren, W et al. Difference in receptor usage between severe acute respiratory syndrome (SARS) coronavirus and SARS-like coronavirus of bat origin. J. Virol.; 2008; 82, pp. 1899-1907.1:CAS:528:DC%2BD1cXhvFartLg%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/18077725]

54. Wells, HL et al. The evolutionary history of ACE2 usage within the coronavirus subgenus Sarbecovirus. Virus Evol; 2021; 7, 1:STN:280:DC%2BB3sfgvFyiug%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33754082][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7928622]veab007.

55. Crook, JM et al. Metagenomic identification of a new sarbecovirus from horseshoe bats in Europe. Sci. Rep.; 2021; 11, 2021NatSR.1114723C1:CAS:528:DC%2BB3MXhs1ClsbrI [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34282196][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8289822]14723.

56. Tao, Y. & Tong, S. Complete Genome Sequence of a Severe Acute Respiratory Syndrome-Related Coronavirus from Kenyan Bats. Microbiol Resour Announc8, e00548–19 (2019).

57. Su, C. et al. Structural characteristics of BtKY72 RBD bound to bat ACE2 reveal multiple key residues affecting ACE2 usage of sarbecoviruses. MBio15, e0140424 (2024).

58. Seifert, SN et al. An ACE2-dependent Sarbecovirus in Russian bats is resistant to SARS-CoV-2 vaccines. PLoS Pathog; 2022; 18, 1:CAS:528:DC%2BB38XisFemur7O [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36136995][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9498966]e1010828.

59. Grishin, AM et al. Disulfide Bonds Play a Critical Role in the Structure and Function of the Receptor-binding Domain of the SARS-CoV-2 Spike Antigen. J. Mol. Biol.; 2022; 434, 167357.1:CAS:528:DC%2BB3MXisFWhsLbP [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34780781]

60. Zhang, S et al. Loss of Spike N370 glycosylation as an important evolutionary event for the enhanced infectivity of SARS-CoV-2. Cell Res; 2022; 32, pp. 315-318.1:CAS:528:DC%2BB38XlsVChsrY%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35017654][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8752327]

61. Wang, Q et al. Key mutations on spike protein altering ACE2 receptor utilization and potentially expanding host range of emerging SARS-CoV-2 variants. J. Med. Virol.; 2023; 95, 1:CAS:528:DC%2BB38XitlOqtrjP [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36056469]e28116.

62. Zhao, Z et al. Structural basis for receptor binding and broader interspecies receptor recognition of currently circulating Omicron sub-variants. Nat. Commun.; 2023; 14, 2023NatCo.14.4405Z1:CAS:528:DC%2BB3sXhsFejtbjI [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37479708][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10362042]4405.

63. Starr, TN et al. Shifting mutational constraints in the SARS-CoV-2 receptor-binding domain during viral evolution. Science; 2022; 377, pp. 420-424.2022Sci..377.420S1:CAS:528:DC%2BB38XhvV2rtrrI [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35762884]

64. Han, P et al. Molecular insights into receptor binding of recent emerging SARS-CoV-2 variants. Nat. Commun.; 2021; 12, 2021NatCo.12.6103H1:CAS:528:DC%2BB3MXit1yhtrjO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34671049][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8528823]6103.

65. Hati, S; Bhattacharyya, S. Impact of Thiol-Disulfide Balance on the Binding of Covid-19 Spike Protein with Angiotensin-Converting Enzyme 2 Receptor. ACS Omega; 2020; 5, pp. 16292-16298.1:CAS:528:DC%2BB3cXht1WisL7I [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32656452][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7346263]

66. Gao, B; Zhu, S. Mutation-driven parallel evolution in emergence of ACE2-utilizing sarbecoviruses. Front. Microbiol.; 2023; 14, 1118025. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/36910184][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9996049]

67. Du, Y et al. A broadly neutralizing humanized ACE2-targeting antibody against SARS-CoV-2 variants. Nat. Commun.; 2021; 12, 2021NatCo.12.5000D1:CAS:528:DC%2BB3MXhvVehsrbO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34404805][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8371079]5000.

68. Mykytyn, AZ; Fouchier, RA; Haagmans, BL. Antigenic evolution of SARS coronavirus 2. Curr. Opin. Virol.; 2023; 62, 101349.1:CAS:528:DC%2BB3sXhslamsr%2FO [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37647851]

69. Menachery, V. D. et al. Trypsin Treatment Unlocks Barrier for Zoonotic Bat Coronavirus Infection. J. Virol. 94, e01774–19 (2020).

70. Richard, M et al. Factors determining human-to-human transmissibility of zoonotic pathogens via contact. Curr. Opin. Virol.; 2017; 22, pp. 7-12. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27907884]

71. Nakamura, T; Yamada, KD; Tomii, K; Katoh, K. Parallelization of MAFFT for large-scale multiple sequence alignments. Bioinformatics; 2018; 34, pp. 2490-2492.1:CAS:528:DC%2BC1MXhtVWrtb7P [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29506019][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6041967]

72. Nguyen, L-T; Schmidt, HA; von Haeseler, A; Minh, BQ. IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies. Mol. Biol. Evol.; 2015; 32, pp. 268-274.1:CAS:528:DC%2BC2MXivFGltrs%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25371430]

73. Schwegmann-Weßels, C et al. Comparison of vesicular stomatitis virus pseudotyped with the S proteins from a porcine and a human coronavirus. J. Gen. Virol.; 2009; 90, pp. 1724-1729. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19264610]

74. Fukushi, S et al. Vesicular stomatitis virus pseudotyped with severe acute respiratory syndrome coronavirus spike protein. J. Gen. Virol.; 2005; 86, pp. 2269-2274.1:CAS:528:DC%2BD2MXpvVCgtL0%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16033974]

75. Ma, C et al. Broad host tropism of ACE2-using MERS-related coronaviruses and determinants restricting viral recognition. Cell Discov; 2023; 9, 57. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37321999][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10272122]

76. Nie, J et al. Quantification of SARS-CoV-2 neutralizing antibody by a pseudotyped virus-based assay. Nat. Protoc.; 2020; 15, pp. 3699-3715.1:CAS:528:DC%2BB3cXhvFWlsLnM [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32978602]

77. Whitt, MA. Generation of VSV pseudotypes using recombinant ΔG-VSV for studies on virus entry, identification of entry inhibitors, and immune responses to vaccines. J. Virol. Methods; 2010; 169, pp. 365-374.1:CAS:528:DC%2BC3cXht1WisrjM [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20709108][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2956192]

78. Liu, P. et al. Engineering customized viral receptors for various coronaviruses. bioRxiv 2024.03.03.583237 (2024) https://doi.org/10.1101/2024.03.03.583237.

79. Jumper, J et al. Highly accurate protein structure prediction with AlphaFold. Nature; 2021; 596, pp. 583-589.2021Natur.596.583J1:CAS:528:DC%2BB3MXhvVaktrrL [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/34265844][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8371605]

80. Yan, Y; Tao, H; He, J; Huang, S-Y. The HDOCK server for integrated protein-protein docking. Nat. Protoc.; 2020; 15, pp. 1829-1852.1:CAS:528:DC%2BB3cXmsleisLY%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32269383]

81. Yan, Y; Wen, Z; Wang, X; Huang, S-Y. Addressing recent docking challenges: A hybrid strategy to integrate template-based and free protein-protein docking. Proteins; 2017; 85, pp. 497-512.1:CAS:528:DC%2BC2sXhtlGqsb0%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28026062]