A Systems-Predicted Neutralizing Antibody Assay Identifies Compartment-Specific Features Correlating with Neutralization Breadth

Quantifying neutralizing antibody levels across anatomical sites and understanding correlates of neutralization breadth is paramount for monoclonal antibody treatments and next-generation vaccination strategies. The systems serology platform allows for characterization of antibody responses at the isotype and subclass level to an array of analytes simultaneously, generating an antibody signature that can be correlated with an outcome. We sought to expand our platform through the development of a multiplexed neutralizing antibody quantitation assay that can be incorporated into our existing systems serology framework to identify antibody features that may be driving neutralization responses. We therefore developed a multiplex-based systems-predicted neutralizing antibody (SNAb) assay capable of quantifying antibody neutralization levels to dozens of targets simultaneously in a high-throughput manner. Through the incorporation of SNAb into our existing systems serology workflow, we have found that antibody signatures correlating with neutralization are highly influenced by preexisting immunity, anatomical compartment, and vaccination route. For H1-lineage human influenza virus and H5-lineage avian influenza virus, serum-circulating neutralizing antibodies generated through vaccination in nonhuman primates were strongly correlated with Fc-gamma receptor-binding antibodies, which have historically been linked with effector function. This is in contrast to mucosal-resident neutralization signatures that were largely driven by IgG and IgA binding antibodies. We further validated our system-based neutralization approach by correlating antibody features to other conventionally employed and regulatory-approved neutralization outputs, such as microneutralization (MN) and hemagglutination inhibition (HAI) assays. These assays yielded almost identical antibody signatures correlating with neutralization compared to our SNAb assay, highlighting that none of these approaches imparts bias into predictive neutralization signatures. Collectively, we propose that neutralizing antibody titers can be quantified to an array of targets simultaneously and in distinct compartments for viruses such as influenza, SARS-CoV-2, and HIV-1 in a sample-sparing manner. Moreover, the antibody repertoire contributing to neutralization appears to be heavily influenced by the local immune milieu.