Fig. 4: Isolation and characterization of cross-neutralizing antibodies to sarbecoviruses.

A Schematic diagram of the antibody source and quantity (left). Based on spike-binding assays, 370 mAbs were categorized into five types: S1-binding mAbs (red in pie charts), S2-binding mAbs (blue), S1- and S2-binding mAbs (green), extracellular domain of the spike protein (S-ECD) but not S1- and S2-binding mAbs (pink), and no binding (gray). The percentages of neutralizing (red) and non-neutralizing (gray) mAbs were evaluated by SARS-CoV-2 pseudotyped particle assays at an antibody concentration of 10 μg/mL. B–D The gene region usage, CDR3 lengths, and SHM were analyzed for 116 nAbs using IgBLAST. Pie chart showing the germline gene region usage of these nAbs. The five gene regions most used by heavy (left) and light (right) chains are given (B). Distributions of H-CDR3 and L-CDR3 length over a total of 116 nAbs (C). Percent identities of heavy and light chains to the inferred germline among 116 nAbs. Data are presented as medians ± IQRs (25–75%), and the error bars indicate the medians with IQRs (D). E Comparison of the lengths of H-CDR3 (left) and L-CDR3 (right) for nAbs from four distinct cohorts. F Comparison of the SHM levels of heavy (left) and light (right) chains. In panels E and F, Kruskal–Wallis tests were used with post hoc Dunn’s multiple comparisons tests. Significance levels are denoted as follows: *, P < 0.05; ***, P < 0.001; ****, P < 0.0001. A two-tailed test with P < 0.05 was considered statistically significant. ns, no significant. Conva, COVID-19 convalescents. Conva+Vac, COVID-19 convalescents after vaccination. Vac, individuals who received a two-dose vaccination. Vac+Booster, individuals who received a three-dose vaccination.