One Thousand SARS-CoV-2 Antibody Structures Reveal Convergent Binding and Near-Universal Immune Escape

Abstract

Since the emergence of SARS-CoV-2, understanding how antibodies recognize and adapt to viral evolution has been central to vaccine and therapeutic developments. To date, over 1,100 SARS-CoV-2 antibody structures--16% of all known antibody-antigen complexes--have been resolved, marking the largest structural biology effort towards a single pathogen. Here, we present a comprehensive analysis of this landmark dataset to investigate the principles of antibody recognition and immune escape. Human immunoglobulins (IgGs) and camelid single-chain antibodies dominate the dataset, collectively mapping 99% of the receptor-binding domain surface. Despite remarkable sequence and conformational diversity, antibodies exhibit striking convergence in their paratope structures, revealing evolutionary constraints in epitope selection. Structural and functional analyses reveal near-universal immune escape of antibodies, including all clinical monoclonals, by advanced variants such as KP3.1.1. On average, over one-third of antibody epitope residues are mutated. These findings support pervasive immune escape, underscoring the need to effectively leverage multi-epitope targeting strategies to achieve durable immunity.