SARS-CoV-2 BA.1 live virus resistance passaging

Monoclonal antibodies targeting conserved regions of the spike protein, such as the RBD class 4 region and S2 fusion machinery, exhibit broad sarbecovirus neutralization. However, their clinical effectiveness against COVID-19 has been limited by SARS-CoV-2 spike escape mutations and suboptimal neutralization potency. This project utilized structure-conditioned, machine learning-guided protein design to optimize two broadly neutralizing sarbecovirus antibodies to restore the neutralizing activity of a class 4 anti-RBD antibody against Omicron variants and improve the potency of an anti-S2 stem helix antibody. These optimized antibodies demonstrated enhanced neutralization in both in vitro and in vivo models while also exhibiting a higher barrier to resistance in combination. Resistance passaging experiments, in which SARS-CoV-2 BA.1 live virus was repeatedly exposed to increasing concentrations of the antibodies as single agents and in combination, were conducted to assess resistance development. Illumina-based deep sequencing of viral genomes throughout the experiment provided insights into the selective pressures exerted by these antibodies. This work underscores the successful optimization of antiviral antibodies against evolving pathogens and highlights the potential of such approaches in developing robust antibody combinations for epidemic and pandemic preparedness.