ABSTRACT
Background: Monitoring genomic variation of SARS-CoV-2 is crucial in mitigating adaptation to the human host and developing effective treatments that safeguard global health. Bamlanivimab and etesevimab are monoclonal antibodies (mAbs) that have demonstrated potent SARS-CoV-2 neutralizing activity in both pre-clinical and clinical settings and have distinct but overlapping binding sites. Here, the selection and characterization of variants in a pre-clinical setting is presented alongside the impact of emerging variants on antibody binding affinity and viral neutralization potency. Methods: Variant selection was carried out via directed evolution of the receptor binding domain (RBD) and serial passage of authentic SARS-CoV-2 in the presence of bamlanivimab and etesevimab individually or in combination. Sequence confirmed, putative-resistance variants identified in both selection methodologies were incorporated into different assessment platforms (VSV-based SARS-CoV-2 pseudovirus neutralization, a yeast RBD display hACE2 competition, and binding affinity to mAb and hACE2) to evaluate potency loss of the selecting mAb and test activity against the mAb combination. Results: Serial passage of SARS-CoV-2 and directed evolution of the RBD protein were unable to select for resistant viral variants under the pressure of mAb combination therapy. In the same experimental paradigm, variants were identified when each mAb was evaluated alone (E484D/K/Q, F490S, Q493R, and S494P for bamlanivimab and K417N, D420N and N460K/S/T/Y for etesevimab). Neutralization and binding assessments confirmed reduced susceptibility of the variants to the single selecting mAb with 50-fold or greater reductions in potency. Importantly, aside from the Q493R variant, all other resistant viruses were neutralized by the mAb combination therapy. Conclusion: In vitro selection studies using single mAbs, bamlanivimab or etesevimab, identified key positions within the SARS-CoV-2 S-protein that have potential for viral resistance in the clinic, whereas similar studies with the mAb combination therapy were unable to select variants. Binding and competition assays confirmed the neutralization phenotyping data and indicates the mechanism of resistance is due to a reduction in binding affinity. The preclinical selection and functional characterization of resistant viral variants directly supports the observation that mAb combination therapy results in a lower frequency of treatment-emergent resistance in clinical treatment studies.