Structural basis and mode of action for two broadly neutralizing nanobodies targeting the highly conserved spike stem-helix of sarbecoviruses including SARS-CoV-2 and its variants

Fig 1. Screening and identification of SARS-CoV-2 S2-specific neutralizing nanobodies. (A) Schematic diagram showing the construction of the VHH phage library specific for the SARS-CoV-2 S2 (T1076-Q1208) antigen. The process includes immunizing an alpaca with the SARS-CoV-2 S2 (T1076-Q1208) antigen, isolating lymphocytes from peripheral blood mononuclear cells (PBMC), constructing a VHH phage library, panning and screening S2-specific VHH phages, Sanger sequencing, and VHH protein expression. This figure was created using BioRender.com. (B) ELISA-binding profiles of the S2 (T1076-Q1208) protein and the indicated nanobodies. Emission OD450 values are plotted as histograms. (C) Binding capacity of the indicated nanobodies to the SARS-CoV-2 S protein on the surface of HEK-293T cells in flow cytometry. (D) Inhibition of SARS-CoV-2 S-mediated syncytium formation in the presence of the indicated nanobodies. NC, negative control: only HEK-293T/EGFP/S cells. PC, positive control: syncytia formation induced by mixing HEK-293T/EGFP/S cells with HEK-293T-hACE2 cells in the absence of nanobodies. Fused syncytia are indicated with white arrows. Scale bar is 100 μm.

Fig 2. Identification of epitopes recognized by nanobodies H17 and H145. (A-B) The binding ability of H17 and H145 to a series of truncated S2 proteins which fused with GST at the N-terminus using Western Blot (WB). (C-D) Multi-concentration ELISA-binding profile of H17 and H145 to the indicated S2 antigen. OD450 values are plotted as curves. Data are means ± SD of triplicate samples. (E-F) Binding ability of the indicated S2 antigen to H17 and H145 analyzed by biolayer interferometry (BLI). Biotinylated H17 and H145 were immobilized to SA chip. Single association and dissociation curves were detected. b., binding. n.b., no binding. (G) Summary of the binding features of H17 and H145 to the denatured antigen-truncates by WB, ELISA, and BLI.

Fig 3. Neutralizing efficacy of H17 and H145 agaisnt SARS-CoV-2 variants. (A) Cross-neutralization activity of H17 and H145 against the SARS-CoV-2 prototype and its major variants in the pseudovirus neutralizing assay. H6 was used as a negative control. Data are expressed as means ± SD. Experiments were performed in triplicate. (B) Table summarizing the neutralizing IC50 values. (C) Sequence conservation of the stem helix epitope spanning DPLQPELDSFKEEL among diverse SARS-CoV-2 variants. The sequence corresponds to spike residues 1139 to 1152 of the SARS-CoV-2 prototype virus. Sequences were downloaded from the GISAID database and https://covariants.org.

Fig 4. Cross-neutralizing capacity of H17 and H145 against other sarbecoviruses. (A) Binding capacity of H17 and H145 to sarbecoviral S proteins expressed on HEK-293T cells by flow cytometry. Data are means ± SD of duplicate samples. (B) Cross-neutralization activity of H17 and H145 against sarbecoviruses detected by pseudovirus neutralization assay. Data are means ± SD of triplicate samples. H6 was used as a negative control. (C) Table summarizing the binding EC50 value and the neutralizing IC50 value by flow cytometry and pseudovirus assay, respectively. (D) Epitope sequence alignment among sarbecoviruses, including SARS-CoV-2- and SARS-CoV-related lineages. Sequences were downloaded from the GISAID database and https://covariants.org.

Fig 5. Complex structure of H145 bound to the epitope SH-peptide. (A) Cartoon representation of the H145/SH-peptide complex structure. Side view (upper) and top view (lower) of the complex structure are presented. H145 is in cyan, and SH-peptide in yellow. CDRs of H145 are shown in orange, magenta, and blue, respectively. (B-C) Residues involved in paratope recognition (the distance cutoff is 4.5 Å). Amino acids of H145 (B) and SH-peptide (C) involved in the contacts are shown. H145 is present in surface and the involed residues are coloured in magenta. SH-peptide is present in sticks and coloured in yellow. (D-E) Hydrogen-bond and salt-bridge interactions between H145 and SH-peptide (distance cutoff is 3.2 Å). Structures are presented and shown in cartoon and sticks. H145 is shown in cyan, and SH-peptide in yellow. CDRs of H145 are shown in orange, magenta, and slate, respectively.

Fig 6. Structural basis for nanobody neutralization. (A) H145/SH-peptide complex structure superimposed on the reported SARS-CoV-2 spike trimer structures in both the prefusion (PDB: 6XR8) and the postfusion (PDB: 6XRA) states. Superimposition was done by aligning the complex of H145 (red) and SH-peptide (green) with the corresponding part in one of spike protomers (SH shown as yellow). (B) Binding modes of anti-SH nNbs and nAbs onto the stem helix (D1139-P1162) trimer structure in the SARS-CoV-2 prefusion state (PDB: 6XR8). The stem helix is shown in cartoon representation. One stem helix binding by nanobodies or antibodies is colored in red, and the other two, which clash with anti-SH nNbs or nAbs, are colored in blue. The light chain of each Fab of nAb is colored gray, and the heavy chain is in a different color for distinction. The Fab of nAbs is shown in surface. (C) Epitopes recognized by the two nanobodies and the reported anti-SH nAbs labeled on the SARS-CoV-2 stem helix (D1139-P1162, yellow) sturctures both in the prefusion (PDB: 6XRA) and postfusion (PDB: 6XR8) states. (D) Real-time BLI binding profiles of H145 and S2P6 (left), H145 and CV3-25 (middle), and S2P6 and CV3-25…

Fig 7. H17 and H145 bind to viral S in an acidification-insensitive manner. (A) S2 (T1076-Q1208) protein was biotinylated and immobilized. Slow-on/slow-off kinetic data are analyzed by the 1:1 binding model. Recorded binding profiles and calculated kinetic parameters are shown. (B) Binding capacity of the indicated nanobodies to SARS-CoV-2 spike on membrane of HEK-293T cells at different pH conditions (7.4 or 5.4). The percentage of H145 positive cells was detected by flow cytometry.

Fig 8. Extended taking-action window of H17 and H145 for virus neutralization. (A-B) Quantitative analysis of the inhibitory activity for anti-SH nNbs (H17 and H145) and anti-RBD nNbs (Nb20 and Nb007) by the syncytium-formation inhibition assay (A) and the pseudovirus neutralization assay (B). Time of adding nNbs was adjusted to two different time points: in the Pre-1 h group, nNbs were pre-incubated with HEK-293T/S cells or SARS-CoV-2 pseudovirus for 1 hour at 37°C and then mixed with HEK-293T-hACE2 cells, respectively; in the Post-1 h group, nNbs were added after HEK-293T/S and HEK-293T-hACE2 cells were pre-incubated for 1 h at 37°C or after SARS-CoV-2 pseudovirus and HEK-293T-hACE2 cells were pre-incubated for 1 h on ice, respectively. (C) Binding affinity of anti-SH and anti-RBD nNbs to the prefusion locked SARS-CoV-2 S trimer (F817P, A892P, A899P, A942P, K986P, V987P, R683A, and R685A) as the biotinylated and immobilized protein at pH 7.4 detected by BLI assay. Slow-on/slow-off kinetic data are analyzed by the 1:1 binding model. Recorded binding profiles and calculated kinetic parameters are shown.