Susceptibility of broad reactivity nanobodies to resistance mutations in the S2 domain of SARS-CoV-2 predicted by yeast display deep mutational scanning

The rapid evolution of SARS-CoV-2 has led to the erosion of vaccine induced serum neutralisation and monoclonal antibody efficacy. As such, interest is inevitably moving towards more conserved regions of the SARS-CoV-2 spike protein like the S2 domain. Resistance mutations continue to be a major obstacle for the development of antivirals and vaccines which target the RBD but what extent these will be a problem for S2 binding antibodies is not known. We have developed a yeast display deep scanning mutagenesis platform which allows an unbiased prospective assessment of millions of single and double mutations for their effects on antibody binding to the S2 domain. We have compared the mutational resistance of a panel of five nanobodies mapped to four distinct non-competing epitopes within the conserved fusion peptide, stem helix and heptad repeat 2 elements of the S2 domain. Yeast display deep mutational scanning predicted reduced binding of C303, G223, G225, and G142 to naturally occurring resistance mutations which was further experimentally confirmed on SARS-CoV-2 variants. Our study shows that resistance mutations in conserved elements of the S2 domain may still pose a challenge to the development of monoclonal antibodies and subunit vaccines.