Clonal interference and changing selective pressures shape the escape of SARS-CoV-2 from hundreds of antibodies.

SARS-CoV-2 has evolved increased resistance to human polyclonal antibody responses. But, how it escaped individual monoclonal antibodies from these responses has not been thoroughly explored. Cao et al. used deep mutational scanning to identify mutations that allow SARS-CoV-2 to escape individual antibodies, doing so for hundreds of different antibodies. Here, we use these data to reconstruct how the virus escaped each antibody in nature. For each antibody, we predict how levels of escape changed in the global SARS-CoV-2 population over time. For many antibodies, these levels dramatically fluctuated due to escape mutations being displaced by clade-turnover events. We validate predicted patterns using pseudovirus neutralization data. Fitness effects estimated from natural sequences suggest that mutations are displaced due to clonal interference between clades and that the order in which mutations arose is shaped by changing selective pressures. Overall, this work suggests that SARS-CoV-2 evaded polyclonal responses via complex evolutionary dynamics.