Deep mutational scanning to understand the evolution of SARS-CoV-2 spike.

Abstract

SARS-CoV-2 emerged as a zoonosis in 2019, causing a pandemic that led to significant global mortality, and devastating economic impact. SARS-CoV-2’s pathogenicity was related to its novelty in an immunologically naïve population. With increasing population immunity through vaccination and/or natural infection attenuating the disease, societies have been able to return to a semblance of normality, however SARS-CoV-2 has persisted to become endemic. With endemicity SARS-CoV-2 has continued to adapt and evolve, initially to optimise transmission and latterly to escape immune responses. To predict the future evolution, a deep mutagenesis scanning platform was developed. Deep mutagenesis scanning allows the phenotypic effects of thousands of mutations to be explored in a high-throughput manner. Using whole trimeric Alpha spike displayed on mammalian cells provided a physiologically relevant model and allowed the identification of mutations that increase ACE2 binding (the receptor for SARS-CoV-2 spike) and immune escape, which subsequently appeared in the Omicron lineages. Using this novel deep mutagenesis platform, the evolutionary trajectory of the SARS-CoV-2 receptor-binding domain can be seen to be restricted by epistasis. Vaccine induced immune responses against the receptor binding-domain are found to be remarkably focused on one or two residues despite being polyclonal and these residues have been repeatedly selected for in a variety of variants. From work exploring the antigenic effects of receptor-binding domain mutations, it becomes apparent the N-terminal domain contributes significantly to the immune escape seen with Delta and BA.1. This effect of the N-terminal domain does not appear to be mediated by escape from N-terminal domain directed antibodies, but by making the receptor-binding domain more difficult to neutralise. The plasticity of and focused immune response on the receptor-binding domain make further SARS-CoV-2 antigenic drift inevitable. Work described here suggests the most dramatic changes in antigenicity requires changes in both the N-terminal domain and receptor-binding domain.