RESUMO
The COVID-19 pandemic has highlighted how viral variants that escape monoclonal antibodies can limit options to control an outbreak. With the emergence of the SARS-CoV-2 Omicron variant, many clinically used antibody drug products lost in vitro and in vivo potency, including AZD7442 and its constituent, AZD1061 [VanBlargan2022, Case2022]. Rapidly modifying such antibodies to restore efficacy to emerging variants is a compelling mitigation strategy. We therefore sought to computationally design an antibody that restores neutralization of BA.1 and BA.1.1 while simultaneously maintaining efficacy against SARS-CoV-2 B.1.617.2 (Delta), beginning from COV2-2130, the progenitor of AZD1061. Here we describe COV2-2130 derivatives that achieve this goal and provide a proof-of-concept for rapid antibody adaptation addressing escape variants. Our best antibody achieves potent and broad neutralization of BA.1, BA.1.1, BA.2, BA.2.12.1, BA.4, BA.5, and BA.5.5 Omicron subvariants, where the parental COV2-2130 suffers significant potency losses. This antibody also maintains potency against Delta and WA1/2020 strains and provides protection in vivo against the strains we tested, WA1/2020, BA.1.1, and BA.5. Because our design approach is computational--driven by high-performance computing-enabled simulation, machine learning, structural bioinformatics and multi-objective optimization algorithms--it can rapidly propose redesigned antibody candidates aiming to broadly target multiple escape variants and virus mutations known or predicted to enable escape.
RESUMO
1Minimizing the human and economic costs of the COVID-19 pandemic and of future pandemics requires the ability to develop and deploy effective treatments for novel pathogens as soon as possible after they emerge. To this end, we introduce a unique, computational pipeline for the rapid identification and characterization of binding sites in the proteins of novel viruses as well as the core chemical components with which these sites interact. We combine molecular-level structural modeling of proteins with clustering and cheminformatic techniques in a computationally efficient manner. Similarities between our results, experimental data, and other computational studies provide support for the effectiveness of our predictive framework. While we present here a demonstration of our tool on SARS-CoV-2, our process is generalizable and can be applied to any new virus, as long as either experimentally solved structures for its proteins are available or sufficiently accurate homology models can be constructed.