Viral evolution prediction identifies broadly neutralizing antibodies to existing and prospective SARS-CoV-2 variants

Abstract

Monoclonal antibodies (mAbs) targeting the SARS-CoV-2 receptor-binding domain are used to treat and prevent COVID-19. However, the rapid evolution of SARS-CoV-2 drives continuous escape from therapeutic mAbs. Therefore, the ability to identify broadly neutralizing antibodies (bnAbs) to future variants is needed. Here we use deep mutational scanning to predict viral receptor-binding domain evolution and to select for mAbs neutralizing both existing and prospective variants. A retrospective analysis of 1,103 SARS-CoV-2 wild-type-elicited mAbs shows that this method can increase the probability of identifying effective bnAbs to the XBB.1.5 strain from 1% to 40% in an early pandemic set-up. Among these bnAbs, BD55-1205 showed potent activity to all tested variants. Cryogenic electron microscopy structural analyses revealed the receptor mimicry of BD55-1205, explaining its broad reactivity. Delivery of mRNA-lipid nanoparticles encoding BD55-1205-IgG in mice resulted in serum half-maximal neutralizing antibody titre values of ~5,000 to XBB.1.5, HK.3.1 and JN.1 variants. Combining bnAb identification using viral evolution prediction with the versatility of mRNA delivery technology can enable rapid development of next-generation antibody-based countermeasures against SARS-CoV-2 and potentially other pathogens with pandemic potential.