Sequence signatures of two public antibody clonotypes that bind SARS-CoV-2 receptor binding domain

Abstract

Since the COVID-19 pandemic onset, the antibody response to SARS-CoV-2 has been extensively characterized. Antibodies to the receptor binding domain (RBD) on the spike protein are frequently encoded by IGHV3-53/3-66 with a short complementarity-determining region (CDR) H3. Germline-encoded sequence motifs in heavy chain CDRs H1 and H2 have a major function, but whether any common motifs are present in CDR H3, which is often critical for binding specificity, is not clear. Here, we identify two public clonotypes of IGHV3-53/3-66 RBD antibodies with a 9-residue CDR H3 that pair with different light chains. Distinct sequence motifs on CDR H3 are present in the two public clonotypes that seem to be related to differential light chain pairing. Additionally, we show that Y58F is a common somatic hypermutation that results in increased binding affinity of IGHV3-53/3-66 RBD antibodies with a short CDR H3. These results advance understanding of the antibody response to SARS-CoV-2.

Fig. 1. Two major clonotypes of IGHV3-53/3-66 antibodies to SARS-CoV-2 RBD.

a The number of IGHV3-53/3-66 RBD antibodies that use the same light chain with the same CDR H3 are tabulated. The two most common combinations are IGKV1-9 pairing with 9 aa CDR H3 and IGKV3-20 pairing with 9 aa CDR H3, denoted as clonotype 1 and clonotype 2, respectively. b Sequence logos for the CDR H3 regions of IGHV3-53/3-66 antibodies that pair with IGKV1-9 or IGKV3-20. A sequence logo for the CDR H3 regions of 203,175 IGHV3-53/3-66 antibodies from Observed Antibody Space database that have a CDR H3 length of 9 aa is shown for reference (repertoire). The position of each residue is labeled on the x-axis based on Kabat numbering. c IGHJ gene usage for clonotypes 1 and 2, as well as 203,175 IGHV3-53/3-66 antibodies from Observed Antibody Space database that have a CDR H3 length of 9 aa (repertoire), are shown as pie charts. For antibodies in clonotypes 1 and 2, only those with nucleotide sequence information available were analyzed. d Amino acid sequences for different IGHJ genes are shown. Source data are available in the Source data file.

Fig. 2. Structural analysis of sequence signatures in CDR H3 of clonotypes 1 and 2.

a Interaction of L96, V98, and G100 (Kabat numbering) in CDR H3 of CC12.1 (PDB 6XC2) with the IGKV1-9 light chain of the antibody, and SARS-CoV-2 RBD. b Interaction of F96, D98 and Y100 (Kabat numbering) in CDR H3 of CC12.3 (PDB 6XC4) with the IGKV3-20 light chain of the antibody, and SARS-CoV-2 RBD. Gray: RBD; light blue: heavy chain; pink: light chain. V_H and V_L indicate residues belong to the heavy and light chain of the antibody, respectively.

Fig. 3. Specific pairing of CDR H3 and light chain is critical for IGHV3-53/3-66 antibody binding to SARS-CoV-2 RBD.

a Binding of different Fabs to SARS-CoV-2 RBD was measured by biolayer interferometry with RBD loaded onto the biosensor and Fab in solution. Y-axis represents the response. Dissociation constant (K_D) for each Fab was obtained using a 1:1 binding model, which is represented by the red curves. COV107-23 belongs to clonotype 1, whereas COVD21-C8 belongs to clonotype 2. b Fab crystal structures of wild-type (WT) COV107-23 and COV107-23 heavy chain pairing with COVD21-C8 light chain are compared. Left panel: structural alignment using residues 1–90 of the heavy chain. Right panel: Zoomed-in view for the CDR H3. c Conformations at the tips of the CDR H3s in WT COV107-23 and COV107-23 heavy chain pairing with COVD21-C8 light chain are shown. A β-turn is observed in CDR H3 of WT COV107-23, with V_H G97 and V_H G99 at the i and i + 2 positions, respectively. V_H indicates residues belong to the heavy chain.

Fig. 4. Binding and expression profiling of 143 CDR H3 variants in B38 antibody.

a For each of the 143 CDR H3 variants, the enrichment in occurrence frequencies after FACS selections for binding to RBD and expression level are shown. Blue: CDR H3 variants that are derived from IGHV3-53/3-66 RBD antibodies that use IGKV1-9. Black: CDR H3 variants that are derived from IGHV3-53/3-66 RBD antibodies that do not use IGKV1-9. Shaded area indicates log₁₀ enrichment in binding >0. Data are from the average of n = 2 independent biological replicates. b The amino-acid length distribution of CDR H3 variants that are positively enriched in binding (log₁₀ enrichment in binding >0) or not (log₁₀ enrichment in binding ≤0) is shown. c Sequence logos are shown for CDR H3 variants with 9 aa (Kabat numbering) that are positively enriched or not enriched. d Comparison of log₁₀ enrichment in binding for CDR H3 variants from IGHV3-53/3-66 RBD antibodies that use IGKV1-9 and those that do not use IGKV1-9. e Comparison of log₁₀ enrichment in expression for CDR H3 variants from IGHV3-53/3-66 RBD antibodies that use IGKV1-9 and those that do not use IGKV1-9. d, e Two-tailed Student’s t-test was used to compute the p-value. Error bars represent standard deviations. The center horizontal bars represent the means. Enrichment in expression and binding, as well as count data are provided in Supplementary Data 2. Data are from the average of n = 2 independent biological replicates. Source data are provided in the Source data file.

Fig. 5. Y58F is a signature somatic hypermutation in IGHV3-53/3-66 RBD antibodies with a short CDR H3.

a IGHV3-53/3-66 RBD antibodies are categorized based on their CDR H3 length (Kabat numbering). Occurrence frequencies of individual somatic hypermutations in different categories were quantified and shown as a heatmap. The number of antibodies in each category is indicated on the right of the heatmap. b Both Y58 and F58 variants were constructed for four IGHV3-53 antibodies. Binding kinetics of each of these antibodies as Fab format to SARS-CoV-2 RBD was measured by biolayer interferometry with RBD loaded on the biosensor and Fab in solution. Y-axis represents dissociation constants (K_D) that were obtained using a 1:1 binding model. Of note, the WTs of COV107-23, COVD21-C8, and CC12.3 contain F58, whereas the WT of COVA2-20 contains Y58. The light chain gene usage and the number of amino acid (aa) residues in the CDR H3 region of each antibody are indicated. Source data are provided in the Source data file.