Biophysical principles predict fitness of SARS-CoV-2 variants

Abstract

SARS-CoV-2 is under constant selective pressure from antibodies while requiring efficient binding to host cells for successful infection. Here we focus specifically on the receptor binding domain (RBD) of the SARS-CoV-2 spike protein, a key determinant of viral entry into host cells. Through a comprehensive approach, comprising large-scale sequence analysis of SARS-CoV-2 variants and the formulation of a fitness function based on protein folding and binding thermodynamics, we unravel the relationship between the epidemiological fitness contribution of the RBD and its biophysical properties. We developed a biophysical model mapping the phenotype space, characterized by binding constants to cell receptors and antibodies, onto the fitness landscape for variants ranging from the ancestral Wuhan Hu-1 to the Omicron BA.1. Using equilibrium statistical mechanics, we show how fitness can be expressed as a function of binding constants to cell receptors and antibodies and validate our findings through experimentally measured binding affinities and population data on frequencies of variants. This forms the basis for a comprehensive epistatic map, relating the genotype space to fitness. Our study thus delivers a tool for predicting the fitness of variants harboring previously unseen mutations, and sheds light on the impact of specific mutations on viral fitness. These insights offer not only greater understanding of viral evolution but also potentially aid in guiding public health decisions in the battle against COVID-19 and future pandemics.