Computational design and improvement of a broad influenza virus HA stem targeting antibody.

Broadly neutralizing antibodies (nAbs) are vital therapeutic tools to counteract both pandemic and seasonal influenza threats. Traditional strategies for optimizing nAbs generally rely on labor-intensive, high-throughput mutagenesis screens. Here, we present an innovative structure-based design framework for the optimization of nAbs, which integrates epitope-paratope analysis, computational modeling, and rational design approaches, complemented by comprehensive experimental assessment. This approach was applied to optimize MEDI8852, a nAb targeting the stalk region of influenza A virus hemagglutinin (HA). The resulting variant, M18.1.2.2, shows a marked enhancement in both affinity and neutralizing efficacy, as demonstrated both in vitro and in vivo. Computational modeling reveals that this improvement can be attributed to the fine-tuning of interactions between the antibody's side-chains and the epitope residues that are highly conserved across the influenza A virus HA stalk. Our dry-wet iterative protocol for nAb optimization presented here yielded a promising candidate for influenza intervention.