Discovery and Characterization of a Pan-betacoronavirus S2-binding antibody

Abstract

Three coronaviruses have spilled over from animal reservoirs into the human population and caused deadly epidemics or pandemics. The continued emergence of coronaviruses highlights the need for pan-coronavirus interventions for effective pandemic preparedness. Here, using LIBRA-seq, we report a panel of 50 coronavirus antibodies isolated from human B cells. Of these antibodies, 54043-5 was shown to bind the S2 subunit of spike proteins from alpha-, beta-, and deltacoronaviruses. A cryo-EM structure of 54043-5 bound to the pre-fusion S2 subunit of the SARS-CoV-2 spike defined an epitope at the apex of S2 that is highly conserved among betacoronaviruses. Although non-neutralizing, 54043-5 induced Fc-dependent antiviral responses, including ADCC and ADCP. In murine SARS-CoV-2 challenge studies, protection against disease was observed after introduction of Leu234Ala, Leu235Ala, and Pro329Gly (LALA-PG) substitutions in the Fc region of 54043-5. Together, these data provide new insights into the protective mechanisms of non-neutralizing antibodies and define a broadly conserved epitope within the S2 subunit.

Figure 1.. Identifying broadly reactive mAbs using a diverse CoV spike LIBRA-seq panel

(A) A panel of antibodies identified by LIBRA-seq (rows), with corresponding V- and J-genes and percent nucleotide identities, CDR lengths and amino acid sequences, and isotype (columns). LIBRA-seq scores for each antigen are shown alongside each antibody as a heatmap from −2 (tan) to 2 (purple). HA-NC99 was included as a negative control antigen. (B) 50 IgG and IgA cells identified by LIBRA-seq are shown as circles, with their respective LIBRA-seq scores (LSSs) for SARS-CoV-2 spike (x axis), SARS-CoV spike (y axis), and MERS-CoV spike (color heatmap). (C) Monoclonal antibodies were produced by microexpression and tested for binding by ELISA to a panel of human coronavirus spike proteins. ELISA area under the curve (AUC) values were calculated from binding curves in Figure S1A and are shown as heatmaps from minimum (white) to maximum (blue) binding. Antibodies that bound to the SARS-CoV-2 spike are shown. ELISA controls are described in Figure S1.

Figure 2.. mAb 54043-5 is an ultrabroad S2 antibody

(A) Phylogenetic tree of coronavirus spikes from select members of the Orthocoronavirinae subfamily. Members of the alpha- (α) beta- (β) and deltacoronavirus (δ) genera are indicated by color. The betacoronavirus genus contains five lineages: A-D and the Hibecovirus lineage (H). Spikes included in LIBRA-seq experiments and binding assays are underlined, and those included exclusively in binding assays are boldly labeled. Scale bar denotes amino acid phylogenetic distance. (B) ELISA curves for mAb 54043-5 binding to spikes from selected CoVs (left) and associated ELISA AUC values for each (right). An “*” denotes a non-human coronavirus (C) ELISA curves for mAb 54043-5 binding to spikes from SARS-CoV-2 variants (left) and associated ELISA AUC values for each (right) (D) ELISA AUCs for 54043-5 binding to the S1 and S2 domains of spikes from SARS-CoV-2 and MERS-CoV, or positive control mAbs CR3022 and 1F8. ELISA controls described in figure S1 (E) SPR sensorgram for 54043-5 Fab binding to the S2 subunit of the SARS-CoV-2 spike. Binding curves are colored black, and data fit to a 1:1 binding model is colored red.

Figure 3.. Cryo-EM structure of Fab 54043-5 bound to the SARS-CoV-2 S2 subunit

(A) Side and top-down views of the 3.0 Å 3D reconstruction of Fab 54043-5 bound to S2. The S2 protomers are colored green, blue, or salmon. The 54043-5 heavy and light chains are colored purple and white, respectively. (B) 54043-5 binds an epitope at the apex of S2, spanning the junction between heptad repeat 1 (HR1) and the central helix (CH). The S2 subunit bound to one Fab is shown as cartoons (left), with two protomers underneath the corresponding EM map and colored gray. One protomer is colored according to the gene schematic (bottom) and 54043-5 is colored as in (A). Zoomed in views of the 54043-5 Fab-S2 interface (right) are shown as cartoons, with important residues shown as sticks. Phe100F is shown with a partially transparent surface to illustrate space filling in a hydrophobic pocket within the epitope. Oxygen atoms are colored red, nitrogen atoms are colored blue, sulfur atoms are colored yellow, and hydrogen bonds are shown as light blue dashed lines. (C) Top and side views of the SARS-CoV-2 spike protein, with the S2 subunit shown as cartoons and colored as in (A). The 54043-5 epitope on one protomer is colored in purple. The S1 subunit, shown as a partially transparent gray surface, almost entirely covers the apex of S2 and the 54043-5 epitope when present.

Figure 4.. Antibody sequence feature uniqueness and conservation of epitope across CoVs

(A) Sequence feature analysis of mAb 54043-5 compared to coronavirus antibodies in the CoVAbDab. The plot displays the percent amino acid identity of the CDRH3 (x-axis) and CDRL3 (y-axis) of a subset of antibodies to 54043-5. Antibodies shown have at least one identical variable (V) gene or ≥50% CDR3 sequence identity. Colors denote shared V gene usage with 54043-5. (B) The count of similar public clones for each S2-binding antibody with characterized epitopes, based on epitope group. Each antibody is represented as a point, with its x-axis coordinate reflecting the number of antibodies in the CoVAbDab sharing both heavy and light V genes and having a CDRH3 with ≥50% amino acid sequence identity. The y-axis coordinate corresponds to the number of antibodies with the same V genes and a CDRL3 sequence with ≥50% identity. (C) Sequence alignment of Betacoronavirus strains, focused on the epitope bound by mAb 54043-5. Structurally buried residues are in bold, with those significantly buried (≥25 Ų buried solvent accessible surface area) enclosed in boxes. S2P mutation residues are highlighted in red, and strains with complete conservation of significantly buried residues are listed in bold. (D) Sequence conservation within the S2 subunit of the spike, mapped onto a surface representation of the S2 subunit and colored based on sequence alignment in (C) (left). The 54043-5 epitope is outlined in yellow. A zoomed in view of the S2 apex is shown as cartoons, with 54043-5 epitope residues shown as sticks. (E) Relative epitope buried surface area (BSA) and conservation across the Orthocoronavirinae genera. Two conserved and structurally characterized S2-epitopes are shown compared to mAb 54043-5, with relative BSA shown as a color gradient from white (0 BSA) to dark green (most BSA within the row). The percent conservation of each epitope residue among the four genera are shown as a line graph.

Figure 5.. Fc effector functional characteristics of lead candidates in bead-based and cell-based assays

(A) Lead cross-reactive mAb candidates, 54041-1, 54043-4, 54043-5, and 54042-13 were tested for their ability to mediate antibody-dependent cellular phagocytosis (ADCP) for SARS-CoV-2, compared to antigen-positive control CR3022 and negative control palivizumab (an anti-RSV antibody). AUC shown was calculated based on the phagocytosis score in figure S4E. (B) 54041-1, 54043-4, and 54043-5 were tested for their ability to mediate antibody-dependent monocyte phagocytosis (ADMP) for SARS-CoV-2, compared to a no-antibody negative control. CR3022 was used as an antigen-positive control. Percent phagocytosis calculation is detailed in the methods section. (C) Cross-reactive mAbs were tested for their ability to mediate antibody-dependent cellular trogocytosis (ADCT) for SARS-CoV-2, compared to antigen-positive control CR3022 and negative control palivizumab. AUC shown was calculated based on the trogocytosis score in figure S4G. (D) Cross-reactive mAbs were tested for their ability to mediate antibody-dependent cellular cytotoxicity (ADCC) against SARS-CoV-2, compared to antigen-positive control CR3022 and negative control palivizumab. AUC shown was calculated based on the cytotoxicity score in figure S4F. (E). Cross-reactive mAbs were additionally tested against OC43 for their ability to mediate antibody-dependent cellular phagocytosis compared to the OC43-specific positive control, 54044–5, and negative control palivizumab. AUC shown was calculated based on the phagocytosis score in figure S4H. All Fc effector data is shown as mean ±SDs.

Figure 6.. Treatment of mice with 54043-5 as pre- or post-exposure treatment of SARS-CoV-2 infection

Mice were treated, according to group, either 24 hours prior to (prophylactic) or following (therapeutic) intanasal SARS-CoV-2 challenge. Results are expressed as absolute mean values plus SEM. (A, B) Body weight (A) and survival (B) was monitored daily for K18-hACE2 mice treated prophylactically with PBS buffer (vehicle), wild-type 54043-5, 54043-5 LALA-PG, an isotype control (#1664), or a neutralizing SARS-CoV-2 antibody (S309) for 14 days following challenge with SARS-CoV-2 (WA1/2020). Statistical significance of the body weight differences between the mAb testing groups and the vehicle group at days 5–7 p.i. are reported in Figure S6B. (C) Three mice per group in the prophylactic study were sacrificed 3 days post-infection (p.i.) and viral titers in the lung tissue were measured. Lung viral titer is expressed as the logarithm of the viral genomic equivalents (GEq) per mL of homogenized lung tissue. (D, E) Body weight (D) and survival (E) were monitored daily for BALB/c mice treated therapeutically with PBS buffer (vehicle), wild-type 54043-5, 540435-LALA-PG, an isotype control antibody (DENV-2D22), or a neutralizing SARS-CoV-2 antibody (COV2–2381) for five days following challenge with SARS-CoV-2 (MA10). (F) Four mice per group in the therapeutic study were sacrificed on days 2 and 5 p.i. and viral titers in the lung tissue were measured. Lung viral titer is expressed as the logarithm of the 50% tissue culture infectious dose of virus (TCID50) per gram of lung tissue. **, p<0.001.