<i>Supporting data for “</i>B Cell Receptor Sequencing Guided Screening and Optimization of Broadly Neutralizing Antibodies against SARS-CoV-2<i>”</i>

<p dir="ltr">The COVID-19 pandemic, driven by the continuous evolution of severe acute respiratory syndrome coronavirus 2 (SARS‑CoV‑2 and the emergence of immune-evasive variants, remains a global threat, elevating reinfection risks and challenging existing therapeutics. I, therefore, conducted a comprehensive study on SARS-CoV-2 antibodies, under the hypothesis that antibody engineering strategies such as bispecific antibodies could overcome such immune evasion and integrating B cell receptor (BCR) sequencing with functional screening assay would enable efficient antibody discovery. This study focused on two main aspects: the engineering of broad-neutralizing antibodies to combat immune evasion and the development of efficient strategies for screening potent antibodies.</p><p><br></p><p dir="ltr">For the first aspect, following the fifth wave of COVID-19 in Hong Kong, I did a serological survey (n=36) to assess herd immunity against emerging variants. Using neutralization assays, I demonstrated that convalescents from the third and fourth waves infected by B.1.1.63 and B.1.36 showed significantly weaker responses to Omicron sublineages as compared with those infected with BA.2/BA.5 during the fifth wave. These results indicated a higher susceptibility to reinfection among patients previously exposed to earlier-waves. Moreover, I found that breakthrough infections elicited stronger neutralizing responses than infection alone. This finding underscored the role of hybrid immunity for better protection. Subsequently, to overcome the immune escape of BA.4/5 against the previously identified broadly neutralizing antibody (bnAb) ZCB11, I engineered bispecific antibodies in DVD-Ig format by fusing the class I ZCB11 with class III neutralizing antibodies P2D9/P3E6. My results showed that these bispecific antibodies successfully restored neutralization activities against BA.4/5, although with reduced potency. I found higher IC₅₀ values (ZCB11-P2D9: 0.5746 μg/mL; ZCB11-P3E6: 0.1639 μg/mL) than those of parental monoclonal antibodies (P2D9: 0.0753 μg/mL; P3E6: 0.0743 μg/mL) against BA.4/5. Structure-guided design targeting the F486V-driven disruption of a hydrophobic interface failed to yield functional gain-of-binding mutants, underscoring the challenges of rational affinity maturation. These results indicated that the pairing between class I and class III neutralizing antibodies is unlikely a good strategy for constructing potent bispecific broadly neutralizing antibodies, probably due to structural hinders.</p><p><br></p><p dir="ltr">For the second aspect, I tried to optimize antibody screening by integrating BCR sequencing with functional validation. A total of 146 BCR sequences were selected and tested via phylogenetic and similarity-based criteria from the total BCR repertoire derived from a well-defined bnAb donor by sequencing 3395 single B cell clones. None of them, however, showed neutralization activities. Concurrently, several ultrapotent broadly neutralizing antibodies were isolated from this donor using conventional single B cell sorting method. Unexpectedly, identical BCR clones were not found from the repertoire sequenced. This result indicated the low frequency of ultrapotent bnAbs in the donor. Lastly, I adopted a method of linking B cell receptor to antigen specificity through sequencing (LIBRA-seq). I successfully identified 20 cross-reactive antibodies from memory B cells with the top candidate showing broad but weak neutralization. In conclusion, my findings not only revealed polyclonal antibody responses against SARS-CoV-2 but indicated useful platforms of technology for engineering of bispecific antibodies and a promising sequence-guided screening framework for rapid antibody discovery.</p>