Design of customized coronavirus receptors

Abstract

Although coronaviruses use diverse receptors, the characterization of coronaviruses with unknown receptors has been impeded by a lack of infection models^1,2. Here we introduce a strategy to engineer functional customized viral receptors (CVRs). The modular design relies on building artificial receptor scaffolds comprising various modules and generating specific virus-binding domains. We identify key factors for CVRs to functionally mimic native receptors by facilitating spike proteolytic cleavage, membrane fusion, pseudovirus entry and propagation for various coronaviruses. We delineate functional SARS-CoV-2 spike receptor-binding sites for CVR design and reveal the mechanism of cell entry promoted by the N-terminal domain-targeting S2L20-CVR. We generated CVR-expressing cells for 12 representative coronaviruses from 6 subgenera, most of which lack known receptors, and show that a pan-sarbecovirus CVR supports propagation of a propagation-competent HKU3 pseudovirus and of authentic RsHuB2019A³. Using an HKU5-specific CVR, we successfully rescued wild-type and ZsGreen-HiBiT-incorporated HKU5-1 (LMH03f) and isolated a HKU5 strain from bat samples. Our study demonstrates the potential of the CVR strategy for establishing native receptor-independent infection models, providing a tool for studying viruses that lack known susceptible target cells.

Extended Data Fig. 1 |. Development and optimization of a modular design strategy for CVR with a type I transmembrane topology.

a, Schematic illustration of the four miniprotein-based CVRs. b, Immunofluorescence analysis of ACE2/CVRs expression and authentic SARS-CoV-2 infection in HEK293T cells stably expressing the receptors. Upper: Receptor expression examined by C-terminal fused 3×FLAG tags. Lower: SARS-CoV-2 infection efficiency as indicated by intracellular N proteins at 24 hpi. Data representative of two independent authentic SARS-CoV-2 infection assays with similar results. c, Cartoon illustrating the framework of the CVRs for TM evaluation. d, Immunofluorescence analysis of the expression of the 31 CVRs in HEK293T cells by detecting the C-terminal fused 3×FLAG tags. e, SARS-CoV-2 PSV entry efficiency promoted by CVRs carrying different TMs. The detailed information on the TMs is summarized in Supplementary Table 2. Data are presented as mean ± s.d. (n = 3 independently infected cells), representative of two independent experiments with similar results. f, Schematic diagram showing the LCB1-based CVRs with indicated TM or TMC substitutions. g, h, SARS-CoV-2 PSV entry in HEK293T cells transiently expressing the indicated CVRs examined by RLU(g) or GFP (h). Data are presented as mean ± s.d. (n = 3 independently infected cells), Statistical analysis was performed using One-way ANOVA analysis followed by Dunnett’s test. Data represented were performed in at least two independent experiments with similar results. i, Immunofluorescence analysis of the subcellular distribution of LCB1-Mxra8 TMC-based CVRs with or without EPM transiently expressed in HEK293T cells. The white dashed boxes highlight the cell surface distribution with a higher magnification. j, SARS-CoV-2 PSV entry efficiency in HEK293T cells transiently expressing CVRs with or without EPM. Data are mean ± s.d. (n = 3 independently infected cells) and analyzed with unpaired two-tailed Student’s t-tests, representative of two independent experiments with similar results. Scale bars: 100 μm in b, d, h, and i. *P<0.05, **P<0.01, ****P<0.0001.

Extended Data Fig. 2 |. Exploring factors that contribute to the receptor function of CVRs with different topologies or modules.

a, Schematic diagram showing CVRs carrying LCB1 or mNb1 displayed in either type I or type Ⅱ transmembrane topology. b, Evaluation of SARS-CoV-2 or MERS-CoV PSV entry efficiency supported by the indicated CVRs with different transmembrane topologies in HEK293T cells. Data are mean ± s.d. of biological triplicates examined over three independent infection assays. Unpaired two-tailed Student’s t-tests. c, Assessment of CVR expression, SARS-CoV-2 RBD-mFc binding, and PSV entry efficiency supported by the CVRs carrying varying copies of TR23 repeats transiently expressed in HEK293T cells. Data are representative of three independent experiments. Scale bars: 100 μm. d, Schematic representation of the CVRs carrying different numbers of immunoglobulin (Ig) domains (left) or an Fc mutant with abolished dimerization ability. e, Western blot analysis of CVRs expression in HEK293T cells under either reducing or non-reducing conditions, respectively. f, Assessment of SARS-CoV-2 PSV entry efficiency in HEK293T cells transiently expressing the indicated CVRs. Data are mean ± s.d. (n = 3 independently infected cells.) and analyzed by unpaired two-tailed Student’s t-tests. g, Schematic representation of the CVRs carrying different numbers of Ig-like domains (left) from mCEACAM1a. h, Western blot analysis of CVRs expression in HEK293T cells. i, SARS-CoV-2 PSV entry efficiency in HEK293T cells transiently expressing the indicated CVRs. Data are mean ± s.d. (n = 3 independently infected cells.). One-way ANOVA analysis followed by Dunnett’s test. j, Schematic representation of the CVRs carrying different SARS-CoV-2 RBD targeting miniproteins. k, l, Expression (k) and SARS-CoV-2 entry-supporting (l) ability of different CVRs in 293T cells. Data are mean ± s.d. (n = 3 independently infected cells). One-way ANOVA analysis followed by Dunnett’s test. m, Schematic diagram showing CVRs carrying different types of VBDs, the two representative VBDs for each type are indicated. n, Immunofluorescence analyzing the expression of the indicated CVRs transiently expressed in HEK293T cells by detecting the C-terminal fused 3×FLAG tags, and the SARS-CoV-2 RBD binding. Scale bars: 100 μm. o, PSV entry are supported by indicated CVRs transiently expressed in HEK293T cells. Data are mean ± s.d. (n = 3 independently infected cells). Data representative of two independent transfections, expression verification, and infection assays with similar results for d-f, g-i, j-l, and m-o, respectively. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; NS, Not significant (P>0.05).

Extended Data Fig. 3 |. Investigating receptor function of CVRs with VBDs connected in various ways.

a-c, Illustration (a), RBD binding efficiency (b), and PSV entry-supporting efficiency (c) of a SARS-CoV-2/MERS-CoV bi-specific CVR transiently expressed in HEK293T cells. Data are presented as mean ± s.d. (biological triplicates of infected cells), representative of three independent experiments with similar results. Unpaired two-tailed Student’s t-test. d-f, Illustration (d), expression (e), and PSV entry-supporting efficiencies (f) of CVRs carrying single or trimeric VBD. Data are mean ± s.d. (biological triplicates of infected cells), analyzed by unpaired two-tailed Student’s t-test. Representative of two independent experiments. g-i, Schematic illustration of bispecific adapter protein (g) and MERS-CoV PSV entry efficiency in BHK-21-hACE2 cells in the presence of indicated concentrations of adaptor proteins (h11B11-mNb1) throughout the infection. Entry efficiency is examined by GFP intensity (h) or RLU (i). Data are mean ± s.d. (biological triplicates of infected cells), representative of three independent infection assays. j-l, Schematic illustration of FcγR (CD32a) mediated antibody-dependent coronavirus entry (j). CD32a expression, antibody (CB6) binding (k), and SARS-CoV-2 PSV entry (l) into HEK293T-CD32a cells pretreated with indicated concentration (con.) of the CB6 antibodies. Data are mean ± s.d. (n=4 biologically independent cells), examined over two independent experiments. m-o, Entry of pre-attached PSV promoted by soluble neutralizing antibody (Nb27-hFc) or bi-specific neutralizing antibody with membrane-associating ability (Nb27-hFc-h11B11). Schematic illustration (m), Nb27-hFc promoted entry (n), and Nb27-hFc-h11B11 promoted PSV entry (o) in Caco2 cells with virus pre-attachment by 1500 rpm centrifugation at 4°C for 1 hour. Data are mean ± s.d. (biological triplicates of infected cells). Data are representative of three independent infection assays. Scale bars: 100 μm for b, e, and k, and 200 μm for h. One-way ANOVA analysis followed by Dunnett’s test for i, l, n, and o. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; NS, Not significant (P>0.05).

Extended Data Fig. 4 |. Comparing receptor function of CVRs with native receptors, or alternative receptors/coreceptors in different cell types.

a-c, The ability of CVRs to promote cell-cell membrane fusion, and authentic SARS-CoV-2 infection is comparable to their native receptors. Spike and receptor-mediated cell-cell fusion was demonstrated by reconstituted GFP intensity (a) and relative light unit (RLU) of Renilla luciferase activity (b). Authentic SARS-CoV-2 infection was examined by immunostaining of intracellular N proteins at 24 hpi (c). Data are mean ± s.d. (biological triplicates) for b. Data are representative of two independent fusion assays or infection assays with similar results. d, Receptor specificity of different coronavirus PSVs in HEK293T stably expressing the native receptor or the indicated CVRs. e, SARS-CoV-2 and MERS-CoV PSV entry into various cell types expressing the indicated receptors. Data representative of two independent transfection and infection assays with similar results for d and e. f, SARS-CoV-2 PSV entry efficiencies in HEK293T cells expressing different receptors or entry factors. Data are presented as mean ± s.d. (n=3 independently infected cells.), representative of two independent transfection and infection experiments. **P<0.01, ***P<0.001, ****P<0.0001; NS, Not significant (P>0.05).

Extended Data Fig. 5 |. Relationship between the binding affinity, neutralizing activity, and CVR entry-promoting efficiency of 25 SARS-CoV-2 RBD-targeting nanobody-fused receptors.

a, b, Assessment of the entry-promoting ability of 25 nanobody-CVRs in HEK293T cells, indicated by GFP (a) and the RLU (b), respectively. Data are presented as mean ± s.d. (n = 3 independently infected cells), representative of two independent experiments. One-way ANOVA analysis followed by Dunnett’s test. Scale bars: 200 μm. c, Comparing RBD binding, neutralization, and PSV entry-promoting ability of different nanobody-fused proteins in HEK293T cells. RBD-mFc binding and PSV entry assays were conducted in HEK293T transiently expressing the 25 nanobody-CVRs. The SARS-CoV-2 PSV neutralization assay was performed in HEK293T-ACE2 in the presence of indicated nanobody-Fc recombinant proteins (10 μg/mL). Data are representative results of two independent experiments with similar results and plotted by the mean (n = 3 independently infected/bound cells). ***P<0.001, ****P<0.0001; NS, Not significant (P>0.05).

Extended Data Fig. 6 |. Expression and antigen-binding ability of CVRs targeting distinct SARS-CoV-2 neutralizing epitopes.

a, Western blot analysis of the expression levels of indicated scFv-CVRs transiently expressed in HEK293T cells. Data are representative of two independent experiments. b, Binding of SARS-CoV-2 S-trimer to HEK293T cells expressing the indicated CVRs. Data are representative of three assays using independent preparations of proteins. c, Flow cytometry analysis of the binding efficiency of scFv-mFc with HEK293T cells transiently expressing the SARS-CoV-2 Spike proteins and ZsGreen simultaneously. The ZsGreen positive cells were gated for subsequent analysis of mFc binding efficiency. Data representative from a single experiment with mean values (n = 3 biologically independent cells) indicated. d, Trypsin-mediated S₂’ cleavage of SARS2-CoV-2 PSV in the presence of soluble receptors or CB6-scFv-mFc. The concentrated SARS-CoV-2 PSV particles were incubated with 100 μg/mL of soluble receptors or CB6-scFv-mFc for 1 hour, followed by incubation with the indicated concentration of TPCK-treated trypsin for 30 minutes. Western blot analysis was conducted by detecting the S2P6 epitope in the S₂ subunit. Data are representative of three independent assays with similar results.

Extended Data Fig. 7 |. Molecular basis of NTD-mediated coronavirus entry.

a-c, Package efficiency of PSVs carrying indicated sarbecoviruses spike glycoproteins (a) and the indicated mutants (b, c). Western blot was conducted by detecting the conserved S2P6 epitope. VSV-M serves as a loading control. Blots representative of two independent transfection assays for pseudovirus production. d, Structures of SARS-CoV-2 BA.4/5 spike trimer without antibody binding (left), or in complex with S2L20 (right). Dashed boxes highlighted the N370-glycan spatially proximate to the S2L20. e, Heatmap showing the inhibitory efficacy of indicated SARS-CoV-2 neutralizing antibodies against PSV entry in HEK293T-hACE2 or HEK293T-S2L20, with BSA as a control. Data are representative results of two independent neutralization assays and plotted by the mean (n=3 independently infected cells). f, Structures of SARS-CoV-2 BA.2 spike trimers with (upper) or without (lower) the binding of NTD-targeting 4A8, along with the side-view (top) and top-view (bottom) cryoEM structures of SARS-CoV-2 Wuhan-Hu-1 spike trimmers the binding of NTD-targeting CV3-13 and DH1052. Orange: NTD; Green: CTD; Red: S2L20; Magenta: 4A8; Blue: CV3-13; Pink: DH1052. g, CryoEM data processing workflow and validation of the S2L20-bound SARS-CoV-2 S CryoEM structure. h, Negative stain microscopy of prefusion SARS-CoV-2 S-glycoprotein (without stabilizing proline substitutions) incubated without antibody, or with S2H14, S2X28, or S2L20 as indicated. S2H14 is known to promote the transition to the postfusion state and was used as a control. Scale bars: 10 nm. Data representative of images captured from two independent experiments with similar results. i, Top-view and side-view CryoEM structures depicting soluble mCEACAM1a (cyan) in complex with MHV spike trimer (gray). NTD and CTD of MHV S are indicated in orange and green, respectively.

Extended Data Fig. 8 |. Generation and characterization of VBDs used for CVRs customized for various coronaviruses.

a, Workflow demonstrating the customization of nanobody-based CVRs for specific coronaviruses. b, Coronavirus CTD or S₁ binding in HEK293T cells transiently expressing the corresponding CVRs. Dashed lines indicate thresholds for positive ratio calculation. Data are presented as mean ± s.d. (n=3 biologically independent cells), representative of two independent experiments. c, The pan-sarbecovirus entry-promoting ability of CVR-Nb27 was evaluated by six different sarbecoviruses in 293T cells. Data are presented as mean ± s.d. (biological quadruples of infected cells), representative of two independent infection assays. Unpaired two-tailed Student’s t-tests. d, BLI analyses of binding kinetics of immobilized nanobody-Fc fusion proteins with the soluble RBD and S₁-hFc (for 229E) of the indicated coronaviruses. e, a summary of binding kinetics of nanobodies bound to the immobilized virus antigens. f, g, Expression (f) and entry-supporting efficiency (g) of the CVRs with or without EPM transiently expressed in the HEK293T cells. EPM: endocytosis prevention motif. Data are mean ± s.d. (n = 3 independently infected cells). Unpaired two-tailed Student’s t-tests. Experiments were performed twice with similar results, and representative data were shown. h-k, HKU1 specific 2D1-CVR exhibited a comparable receptor function to TMPRSS2. The expression (h) and HKU1-PSV entry promoting efficiency (h, i) of the two receptors were examined. The amplification of propagation-competent rVSV-HKU1-GFP in Caco2 cells expressing 2D1-CVR and TMPRSS2 was demonstrated by GFP (j) and RNA accumulation (k). Scale bars: 100 μm for f, h, and j. Data are mean ± s.d. (biologically triplicates of infected cells) analyzed by unpaired two-tailed Student’s t-tests for i. Data presented are RNA copies of two independently infected cells with each point representing the mean of technical duplicates (RT-qPCR) for k. Experiments presented were independently performed twice with similar results for h-j and single time for k. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; NS, Not significant (P>0.05).

Extended Data Fig. 9 |. Neutralization and inhibition assays based on CVR-expressing HEK293T cells.

a, Comparison of neutralization profiles of sera collected from COVID-19 convalescents (left) or vaccinated individuals (right) based on HEK293T cells expressing ACE2 or two different CVRs. Serum dilution: 1:200. Heatmap plotted by the mean values (biologically triplicates of infected cells), which are representative results out of two independent experiments. b, c, Neutralization assays of several broadly neutralizing antibodies against PSV entry of representative coronaviruses in HEK293T stably expressing the indicated CVRs. Neutralization curves (b) and a summary of IC₅₀ (c) against each virus are shown. The RBD-targeting REGN 19033 (REGN) was employed as a control. /: no inhibition detected. Data are mean ± s.d. (biologically triplicates of infected cells). d, The IC₅₀ of selected entry inhibitors against SARS-CoV-2 PSV entry was determined in both HEK293T-ACE2 or HEK293T-LCB1-CVR cells. Data are mean ± s.d. (biologically triplicates of infected cells). e, Inhibitory efficacy of inhibitors against PSV entry of SARS-CoV-2-D614G, HKU1, HKU3 and HKU5 in HEK293T cells stably expressing the indicated CVRs. Data are mean ± s.d. (biologically triplicates of infected cells), representative of two infection inhibition assays with similar results. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; NS, Not significant (P>0.05).

Extended Data Fig. 10 |. Characterization of CVR-promoted amplification of replication-competent pcVSV-CoVs or authentic coronaviruses.

a, Genetic organizations and workflow for generating replicable pcVSV-HKU3 or pcVSV-HKU5. b, Successful rescue (P0) and amplification (P1) of pcVSV-HKU3 assisted by VSV-G. Representative images of an experiment that was conducted for a single time. c, d, Trypsin-enhanced cell-cell fusion (c) and VSV-G-independence (d) of pcVSV-HKU3 infection in Caco2-Nb27 (MOI: 0.001). Data are representative of four independent infection assays with similar results. e, f, Accumulation of pcVSV-HKU3 (e) or pcVSV-HKU5 (f) RNA in the supernatant at indicated time points. g, pcVSV-HKU5 mediated cell-cell fusion in Caco2 or Caco2-1B4 cells at MOI=0.1. TPCK-Try+: 20 μg/mL TPCK-treated Trypsin in DMEM+2% FBS. Representative of two independent experiments. h, TCID₅₀ determination assay for RsHuB2019A in Caco2-Nb27 cells by the Red-Muench method. Caco2-Nb27 cells were inoculated with a 10-fold serial dilution of RsHuB2019A containing supernatant (Passage 6). The TCID₅₀ was determined using immunofluorescence to detect the presence of N protein expression of the inoculated cells at 4 dpi. i, Trypsin-dependent amplification of RsHuB2019A in Huh-7 cells. The RsHuB2019A genomic RNA copies in the supernatant collected at indicated time points of infected Huh-7 cells were quantified by RT-qPCR using RdRp-specific primers. Inoculation was conducted at an MOI of 0.0001, with or without trypsin treatment. Try: 100 μg/mL Trypsin in DMEM. j, Genetic organizations of the HKU5 (HKU5-1 LMH03f) ΔORF5-ZsGreen-HiBit (HKU5-ZGH). k-m, Supernatant RNA copies of HKU5-ZGH (k) in Caco2-1B4 cells, ZsGreen-HitBit signal (l), and increase in ZsGreen intensity (P0) (m). Representative of two independent infection assays are shown. Data are mean ± s.d. (n = 3 independently infected cells) for l. n, Isolation of HKU5 (strain PaGD2014/15) from bat samples by Caco2-1B4 cells and its trypsin-dependent propagation. o, Vero E6 and Caco2 cells were infected with HKU5-1 (LMH03f) with or without trypsin treatment. Data are representative of three independent experiments. p, Sequencing results show L76R and K519T mutations in HKU5-1 spikes after ten passages in Caco2-1B4 cells. q, HKU5-1 after ten passages carrying mutations remains unable to infect Caco2 cells without exogenous trypsin treatment. Caco2-1B4 with CVR expression was included as a positive control. r, Efficacy of indicated antiviral reagents against HKU5-1 infection in Caco2-1B4 cells assessed by intracellular N proteins at 48 hpi. s, Efficient inhibition of HKU5 entry by EK1 and EK1C4 peptides in Caco2-1B4. t, Inhibitory effect of selected anti-viral reagents against authentic HKU5-ZGH infection in Caco2-1B4. Inhibitors were coincubated with either the cells or the viruses for 1h and present in the culture medium during infection. The HiBit-based luciferase activity was determined at 48 hpi to assess the inhibitory effect of selected anti-viral reagents against the infection of authentic HKU5-ZGH in Caco2-1B4. Data are representative of two independent infection assays for r-t. u, Overview of the protease cleavage sites of selected coronaviruses. The residue responsible for reduced endosomal cysteine protease activity (ECP) is marked in red, numbering based on SARS-CoV-2. The HKU5 infection efficiencies in n, o, q, r, and s were assessed using rabbit polyclonal antibodies targeting the HKU5 N protein (Cy3) at 48 hpi. Data presented are RNA copies of two independently infected cells with each point representing the mean of technical duplicates (RT-qPCR) for e, f, i, and k, representative of two independent experiments. Scale bars: 125 μm for all images.

Extended Data Fig. 11 |. Schematic diagram of the modular design strategy for CVRs and the potential applications of this technique.

a, Workflow outlining the process of creating artificial receptor scaffolds (ARS) and viral binding domains (VBDs) to construct functional customized viral receptors (CVRs) for establishing in vitro and in vivo infection models. b, The crucial role of CVRs in bridging infection models and virus strains, along with the potential applications of this technique in both basic research and applied science.

Fig. 1 |. Modular design of customized viral receptors (CVRs) for efficient coronavirus entry.

a-e, Dissecting the importance of ACE2 sequences for its viral receptor function. a, Schematic representation illustrates the LCB1-ACE2 chimera with stepwise truncated ACE2 sequences. Protein expression levels (b) and SARS-CoV-2 RBD binding efficiency (c) in HEK293T cells transiently expressing the specified chimeras. SARS-CoV-2 pseudovirus (PSV) entry in HEK293T cells expressing each chimera was demonstrated by GFP (d) or RLU (e), respectively. Scale bars: 200 μm. f-j, Functionality of chimeric receptors with remaining ACE2 sequences substituted by domains from other proteins. Schematic representation (f) delineates CVRs carrying exogenous spacer, transmembrane and cytosolic domain (TMC), and endocytosis-prevention motif (EPM) sequences. The CVR expression (g), SARS-CoV-2 RBD-mFc binding (h), and PSV entry (i, j) efficiencies in HEK293T transiently expressing the indicated receptors were shown. Scale bars: 200 μm. k-m, The impact of spacer length on CVR receptor function. Schematic representation (k) illustrates CVRs with various TR23 tandem repeats, displaying predicted spacer length. CVR expression (l) and SARS-CoV-2 PSV entry efficiency (m) were evaluated in HEK293T cells transiently expressing the indicated CVRs. n, Schematic illustration of the modular design strategy for CVRs. RLU: relative light units. Data representative of at least two independent transfections (Western blot for expression) and functional assays (RBD binding and PSV entry) with similar results for a-e, f-j, and k-m, respectively. Data are presented as mean ± s.d., biological triplicate of bound cells (dashed lines denote thresholds for positive binding) for c and h, and biological triplicates of infected cells for e, j, and m. One-way ANOVA analysis followed by Dunnett’s test for e and j; unpaired two-tailed Student’s t-tests for m. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; NS, not significant (P>0.05).

Fig. 2 |. The impact of binding epitopes on receptor function and its molecular basis.

a, Structural display of SARS-CoV-2 neutralizing epitopes in NTD, CTD, and S₂ subunit, respectively. Various epitope types of 22 neutralizing antibodies are indicated based on a SARS-CoV-2 spike protein structure (6XR8). FP: fusion peptide. SH: stem helix. b, Schematic representation of 44 single-chain variable fragments (scFv)-CVRs with N-terminal light chain (LH) or heavy chain (HL), respectively. c, Heat map displaying SARS-CoV-2 PSV entry efficiency in HEK293T cells transiently expressing the indicated scFv-CVRs. Data are plotted by mean (n = 3 independently infected cells), representative of two independent experiments with similar results. SH: stem helix; FP: fusion peptide. d, Demonstration of CVR expression, antigen binding, PSV entry, and spike-mediated cell-cell fusion in HEK293T cells expressing indicated scFv-CVRs. Scale bars: 100 μm. Data presented were performed in two independent assays with similar results. e, Cartoon elucidates the functional receptor-mediated RBD conformational change and the subsequent exposure of 76E1 binding epitope and proteolytic cleavage at the S₂’ cleavage sites. f, Flow cytometry analysis of 76E1 epitope exposure of SARS-CoV-2 S in the presence of indicated soluble scFv-mFc recombinant proteins. Data are presented as mean ± s.d. (n=3 biologically independent cells), representative of two independent experiments. g, Dose-dependent exposure of 76E1 epitope upon soluble hACE2 (sACE2) or S2L20 scFv-mFc coincubation, which was not detected in BG10-19 scFv-mFc. Data are representative of two independent experiments with similar results. h, Trypsin-mediated cleavage of S₂’ site in SARS-CoV-2 pseudovirus particles in the presence of 100 μg/mL of indicated scFv-mFc. TPCK-try: 10 μg/mL. Blots representative of at least four independent cleavage assays with similar results. Dashed lines denote thresholds for positive ratio calculation for f and g.

Fig. 3 |. NTD-mediated sarbecovirus entry promoted by S2L20-CVR.

a, CVR expression, SARS-CoV-2 NTD/CTD-mFc binding, cell-cell fusion, PSV entry, and SARS-CoV-2(ΔN-GFP) infection in HEK293T stably expressing indicated CVRs. Data representative of two independent transfection and functional assays. b, c, PSV entry of SARS-CoV-2 VOCs (b) or various sarbecoviruses (c) in HEK293T stably expressing hACE2 or S2L20-CVR. d, Sarbecoviruses NTD-mFc binding efficiencies in HEK293T-S2L20-CVR. Data representative of three independent assays. e, S expression levels and corresponding sACE2 or S2L20-mFc binding efficiencies. Data representative of two independent binding assays. f, g, Impact of T372A mutation on S2L20-CVR-supported PSV entry of RaTG13 (f) or BANNAL-20-52 (g). R.aff: R.affinis allele 9479. h, SARS-CoV-2 Wuhan-Hu-1 Hexapro S ectodomain trimer bound to the S2L20 Fab. S protomers: pink, cyan, and gold; N-linked glycans: dark blue; S2L20 Fab variable heavy and light chains: dark and light green, respectively. i, (Top) Superimposition (based on the S₂ subunit) of the structures of S2L20-bound SARS-CoV-2 S with closed RBDs (white, 7N8H) and of S with closed RBDs (teal, PDB 7K43). (Bottom) Superimposition (based on the S₂ subunit) of the structures of S2L20-bound SARS-CoV-2 S with closed RBDs (white, 7N8H) and of S2L20-bound SARS-CoV-2 S with open RBDs (this study). Insets: zoomed-in views of the interface between the NTD, RBD, and S2L20. S2L20 is shown as a semi-transparent surface. The S2M11 Fab is not shown for clarity in both panels. (j) Proposed mechanism of S2L20-mediated RBD opening and stabilization. C: closed; O: open. *: NTD repositioning. For b, c, f, and g, data are mean ± s.d. (n = 3 independently infected cells) out of two independent experiments. Ratios indicate PSV entry supported by S2L20-CVR compared to that supported by ACE2. Unpaired two-tailed Student’s t-tests. Scale bars: 100 μm. *P<0.05, ***P<0.001, ****P<0.0001.

Fig. 4 |. CVRs supported entry and amplification of various coronaviruses.

a, Phylogenetic tree based on coronavirus S amino acid sequences. Underline: without known receptors. b, PSV entry of twelve coronaviruses in HEK293T transiently expressing indicated CVRs. Data are mean ± s.d. (biological triplicates), representative of three independent infection assays. c, Cell-cell fusion promoted by indicated coronavirus S in Caco2 cells with or without stable expression of corresponding CVRs with trypsin treatment. Data representative of three independent experiments. d, 229E-S₁-mFc binding, cell-cell fusion, and authentic 229E infection in HEK293T stably expressing APN or 4H5-CVR. e-f, MHV-A59 antigen binding, cell-cell fusion, and authentic MHV infection in HEK293T stably expressing mCEACAM1a, NTD-targeting (e), or CTD-targeting (f) CVRs. g, MHV-A59 RNA copies in supernatant from infected cells expressing mCEACAM1a or 1B3-CVR. h, Cartoon illustrating different trypsin dependence of RsHuB2019A propagation in Huh-7 and Caco2-Nb27 cells. i, CVR expression, N protein, and CPE in cells inoculated with RsHuB2019A at indicated MOI (no trypsin). j, Accumulation of viral RNA in supernatant of indicated cells infected with RsHuB2019A. Try: 100 μg/mL Trypsin in DMEM+2%FBS. k, Transmission electron microscopy analysis of HKU5-1 virions. Representative images of a single experiment are shown. l, CVR expression, N protein, and CPE in indicated cells inoculated with HKU5-1 at indicated MOI. m, Accumulation of HKU5 RNA in supernatant of cells inoculated with HKU5-1 at indicated MOI. Infection was examined by S immunofluorescence for d, e, and f and N protein immunofluorescence for i and l at 24 hours post-infection (hpi). For g, j, and m, data presented are RNA copies of two independently infected cells with each point representing the mean of technical duplicates (qRT-PCR). For d-g, i-j, and l-m, data presented are from at least two independent experiments with similar results for each virus, respectively. Scale bars for c-f, 100 μm; for k, 60 nm; for i,l, 125 μm.