The abundance of pre-existing IGHV3-53/66-encoded Class 1 antibodies determines SARS-CoV-2 immune imprinting severity through epitope masking that suppresses de novo Omicron-specific responses, explaining why mRNA vaccine recipients exhibit stronger ancestral strain imprinting than inactivated vaccine recipients and driving regional divergence in viral evolution. This mechanism was elucidated using V(D)J-humanized mice that faithfully recapitulate human imprinting phenotypes, and validated through development of an IGHV3-53 knock-in mouse model that provides a practical tool for evaluating region-specific vaccine update strategies.

The mechanisms driving divergent SARS-CoV-2 immune imprinting in populations primed with different COVID-19 vaccines remain unclear. Recipients of inactivated vaccines readily develop Omicron-specific antibodies through repeated breakthrough infections, whereas mRNA-vaccinated individuals exhibit severe ancestral-strain imprinting that suppresses de novo Omicron-specific responses. These differences could result in distinct antibody landscapes, leading to regional epidemiological divergence and necessitating region-specific vaccine update strategies. Importantly, conventional wild-type mouse models fail to recapitulate strong human SARS-CoV-2 imprinting, which significantly hinders imprinting-related mechanistic investigation and vaccine update evaluation. Here, we surprisingly found that V(D)J-humanized mice could faithfully recapitulate human severe SARS-CoV-2 immune imprinting phenotypes. Comprehensive antibody repertoire and epitope mapping of 583 monoclonal antibodies from these models revealed that the abundance of pre-existing human IGHV3-53/66-encoded SARS-CoV-2 antibody responses determine imprinting severity following Omicron exposure through antibody-mediated masking of Omicron-specific epitopes. Both passive transfer of IGHV3-53/66 antibodies and knock-in of the human IGHV3-53 gene were sufficient to induce severe SARS-CoV-2 imprinting in wild-type mice. Concordantly, head-to-head comparison also showed that mRNA vaccine recipients retained higher IGHV3-53/66 antibody abundance and thus stronger imprinting than inactivated vaccine recipients. Consequently, compared to NB.1.8.1, XFG exhibits greater immune evasion in mRNA-vaccinated individuals but not in inactivated vaccine recipients. This explains the regional predominance of XFG in mRNA-vaccinated populations, while NB.1.8.1 prevails in inactivated vaccine-dominated countries. Together, these findings demonstrate that the V(D)J germline repertoire—even a single germline-encoded antibody response—can profoundly shape humoral imprinting severity. Accordingly, we constructed a human IGHV3-53 knock-in mouse model that can accurately recapitulate human SARS-CoV-2 antibody landscape, providing a valuable tool for guiding future COVID-19 vaccine updates.