Structure-guided multivalent nanobodies block SARS-CoV-2 infection and suppress mutational escape

Abstract

The pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to spread, with devastating consequences. For passive immunization efforts, nanobodies have size and cost advantages over conventional antibodies. In this study, we generated four neutralizing nanobodies that target the receptor binding domain of the SARS-CoV-2 spike protein. We used x-ray crystallography and cryo-electron microscopy to define two distinct binding epitopes. On the basis of these structures, we engineered multivalent nanobodies with more than 100 times the neutralizing activity of monovalent nanobodies. Biparatopic nanobody fusions suppressed the emergence of escape mutants. Several nanobody constructs neutralized through receptor binding competition, whereas other monovalent and biparatopic nanobodies triggered aberrant activation of the spike fusion machinery. These premature conformational changes in the spike protein forestalled productive fusion and rendered the virions noninfectious.

Bivalent nanobodies neutralize by inducing postfusion conformation of the SARS-CoV-2 spike.

On virions, SARS-CoV-2 spike trimers are mostly in an inactive configuration with all RBDs in the down conformation (left). Binding of bivalent nanobody VE stabilizes the spike in an active conformation with all RBDs up (middle), triggering premature induction of the postfusion conformation, which irreversibly inactivates the spike protein (right).

Fig. 1. Camelid nanobodies against two epitopes on the SARS-CoV-2 spike RBD neutralize infection.

(A) Average distance tree of nanobody candidates identified by phage display, calculated by percentage identity (66). (B) Binding of 100 nM HA-tagged VHHs to immobilized SARS-CoV-2 spike RBD or a control protein (MBP) was quantified by ELISA with horseradish peroxidase (HRP)–coupled anti-HA antibody. Unrelated VHH SN was used as a negative control. O.D., optical density. (C) SARS-CoV-2 spike–pseudotyped VSV ΔG eGFP was incubated with the indicated concentrations of HA- or LPETG-tagged (VHH LaM-4, VHH 72) VHHs or ACE2-Fc at 37°C for 1 hour, followed by infection of Vero E6 cells for 8 hours. eGFP-positive cells were measured by flow cytometry to quantify infection. Normalized values from three independent experiments ± SEM and IC₅₀ values are plotted. (D) SARS-CoV-2 was incubated with the indicated concentrations of HA- or LPETG-tagged VHHs as in (C), followed by plaque assay on Vero E6 cells. Plaques were enumerated 3 days after infection; normalized values of three independent experiments ± SEM and IC₅₀ values are plotted. (E and F) Biotinylated SARS-CoV-2 spike RBD was immobilized on SPR spectroscopy chips. (E) Indicated HA-tagged VHHs were injected for 90 s, followed by dissociation for 180 s. Dissociation constants (K_D) were determined on the basis of fits, applying a 1:1 interaction model. (F) Epitope binning was performed by first injecting a single VHH for 120 s, followed by injection of a 1:1 mixture of the first nanobody in combination with VHH E, U, V, or W for 80 s.

Fig. 2. X-ray crystallography defines the binding sites of neutralizing VHHs on the SARS-CoV-2 RBD.

(A to C) Crystal structure of SARS-CoV-2 spike RBD in complex with VHH E and VHH U at 1.87 Å (A) and detailed interaction interface of RBD (in white) with VHH E (B) and RBD with VHH U (C), respectively. (D and E) Crystal structure of SARS-CoV-2 spike RBD in complex with VHH V at 2.55 Å (D) and detailed interaction interface of RBD with VHH V (E). Escape mutants (see Fig. 5 and tables S5 to S8) in the RBD are highlighted in teal and labeled with asterisks. (F) Overview of binding sites of three neutralizing nanobodies on the RBD and their overlap with ACE2, based on PDB ID 6M0J (67). Steric clashes with VHH E are indicated within the dashed circles. N-glycans at N322 and N546 of ACE2 are depicted as yellow spheres. All structural analyses of VHH U and VHH E in complex with RBD were based on one of the two copies in the asymmetric unit with closer alignment to the localized reconstructions of VHH E with RBD and VHH VE with RBD using cryo-EM. Single-letter abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr.

Fig. 3. Cryo-EM structures reveal that VHHs stabilize SARS-CoV-2 spike trimers with RBDs in the up conformation.

(A to D) Cryo-EM reconstructions [(A) and (C)] and atomic models [(B) and (D)] of VHH E [(A) and (B)] and VHH V [(C) and (D)] in complex with trimeric SARS-CoV-2 spike. Frequencies of the identified complexes as well as total numbers of considered particles are noted. (A and B) VHH E (in green) binds to SARS-CoV-2 in a 3-up conformation in the most abundant complex; the resolution is 3.3 Å [0.143 Fourier shell correlation (FSC)]. (C and D) VHH V (in red) binds to SARS-CoV-2 in a 2-up conformation with all VHH binding sites occupied at a resolution of 3.0 Å (0.143 FSC). In the most abundant complex, VHH V binds to the RBD in the up or the down conformation. (E and F) HEK 293 cells inducibly expressing SARS-CoV-2 S Δ18 and either eGFP or tagRFP-t were seeded into microscopy-grade 96-well plates in a 1:1 ratio and induced with 1 μg/ml doxycycline for 20 hours. Cells were treated with 1 μM of the indicated VHHs, and microscopy images were recorded every 20 min for 14 hours at 37°C. (E) Fusion was quantified by calculating Pearson correlation coefficients (PCC) between eGFP and tagRFP-t. Average values from four fields of view of an experiment representative of three independent experiments are displayed. (F) Representative images of cells 12 hours after treatment are displayed (also see fig. S13 and movies S8 to S13). Scale bars, 100 μm.

Fig. 4. Multivalent VHH fusions potentiate neutralization of SARS-CoV-2.

(A) SARS-CoV-2 spike–pseudotyped VSV ΔG eGFP was incubated with twofold serial dilutions of 0.25 μM VHH E, 1 μM VHH U, 1 μM VHH V, or the indicated combinations containing 50% of each VHH at 37°C for 1 hour. Vero E6 cells were subsequently incubated with the mixtures, and infection was quantified as in Fig. 1B. Normalized values from three independent experiments ± SEM are plotted. (B) Caco-2 cells were infected with mNeonGreen-expressing infectious-clone-derived SARS-CoV-2 (icSARS-CoV-2-mNG) in the presence of the indicated nanobody concentrations. Cells were fixed 48 hours postinfection and stained for DNA, and infection was quantified by microscopy. Normalized values from three independent experiments ± SEM are plotted. (C and G) SARS-CoV-2 spike–pseudotyped VSV ΔG eGFP was incubated with the indicated concentrations of HA-tagged single, bivalent, or trivalent VHHs at 37°C for 1 hour, followed by infection of Vero E6 cells as in Fig. 1C. Normalized values from three independent experiments ± SEM and IC₅₀ values are plotted. (D and H) SARS-CoV-2 was incubated with the indicated concentrations of HA-tagged VHHs, followed by plaque assay on Vero E6 cells as in Fig. 1D. Normalized values of three independent experiments ± SEM and IC₅₀ values are plotted. (E and F) Cryo-EM reconstruction (E) and atomic models (F) of VHH VE in complex with trimeric SARS-CoV-2 spike. Frequencies of the identified complexes as well as total number of considered particles are noted. The biparatopic VHH binds to SARS-CoV-2 in the 3-up conformation; the resolution is 2.62 Å (0.143 FSC). (I and J) HEK 293 cell lines inducibly expressing SARS-CoV-2 S Δ18 and either eGFP or tagRFP-t were treated with the indicated VHHs and analyzed as in Fig. 3, E and F, displaying representative images after 12 hours (I), as well as quantified fusion (J) (also see fig. S22 and movies S14 to S22). Scale bars, 100 μm. (K) HEK 293T cells expressing ACE2-tagRFP-t were incubated with DyLight 488–labeled SARS-CoV-2 spike RBD in the presence of the indicated concentrations of nanobodies. RBD bound to ACE2-positive cells was quantified by flow cytometry. Normalized data from three independent experiments ± SEM are plotted. Data presented in fig. S2E and Fig. 4K are from the same experiments, and values for VHHs E, V, and LaM-4 in fig. S2E are plotted for reference.

Fig. 5. Simultaneous targeting of two independent epitopes with neutralizing VHHs prevents viral escape.

(A) Genome structure of VSV SARS-CoV-2 S Δ18 eGFP. UTR, untranslated region. (B to E) Evolution experiment. (B) Replication-competent VSV SARS-CoV-2 S Δ18 eGFP at a multiplicity of infection (MOI) of 0.5 was incubated with different concentrations of the indicated VHHs and allowed to replicate on Vero E6 cells in 12 wells for 4 days. The fraction of infected (eGFP-positive) cells and the cytopathic effect (CPE) were estimated microscopically and are plotted according to the indicated color code. (C) Cleared supernatants from the wells indicated with a circle in (B) were diluted with four volumes of fresh infection medium (1:5 dilution) and used for a second round of replication on Vero E6 cells in the presence of the indicated VHH concentrations. Cleared supernatants were harvested as in (B). (D and E) Cell lysates from the wells indicated by circles in (B) and (C) [corresponding to (D) and (E), respectively] were subjected to targeted resequencing of the RBD to identify variants that had emerged at the interfaces to VHH E (interface E) or to VHH U, V, or W (interface UVW) and to quantify their allelic distribution (see tables S5 and S6 for details on detected variants). (F and G) Wild-type VSV SARS-CoV-2 spike eGFP or plaque-purified escape mutants of VHH E (S1-1f, Spike S494P), VHH U (S1-2h, Spike S371P), VHH V (S1-3b, Spike K378Q), VHH W (S1-4a, Spike S371P), and VHH LaM-4 (S1-10a, WT spike) at an MOI of 0.5 were incubated with the indicated VHH concentrations (F) or 1 μM of the indicated VHH (G) and used for Vero E6 infection experiments as in Fig. 1B. Infection was quantified by flow cytometry; normalized data from three independent experiments ± SEM are plotted. n.d., not detected.