Prediction and surveillance of antigenic variation in influenza and SARS-CoV-2 viruses

Abstract

Influenza has long been a primary model organism for antigenically variable pathogens meaning it can escape immunity acquired through prior exposure by accumulating mutations in the key surface proteins targeted by the immune system (hemagglutinin in the case of influenza). As such, the strain contained in the influenza vaccine must be updated prior to each influenza season to match currently circulating viruses, and individuals must be re-vaccinated in order to maintain antibody-mediated protection. Ensuring that the virus strain contained in the vaccine is a close match to the viruses which will circulate in the upcoming season is important for vaccine effectiveness. Surveillance of circulating viruses and the development of methods to predict changes likely to occur in the virus population are both therefore key areas of interest for influenza vaccinology. During my PhD, SARS-CoV-2 began exhibiting antigenic variation analogous to that seen in influenza, prompting the development of similar surveillance systems and prediction methods as historically applied to influenza. Traditionally, these prediction methods have attempted to forecast which of the currently circulating genotypes will become predominant, so can only predict that an antigenic variant will reach high frequency once it has begun increasing in frequency in the population. In this thesis, I test whether measurements of the fitness effects of individual substitutions can predict which substitutions will seed future antigenic variants which are yet to arise, both for influenza and SARS-CoV-2. The method uses the basic observation that many of the key substitutions in influenza and SARS-CoV-2 evolution occur sporadically in small lineages throughout the phylogenetic tree before they form a major lineage which reaches high frequency. I further describe two pieces of work carried out to support the public health response during the SARS-CoV-2 pandemic. The first examines differences in population immunity to SARS-CoV-2 between cohorts from the UK, USA, Denmark, and China, and finds evidence of differences in the level of antibody imprinting between the cohorts which are consistent with the differences in exposure histories between the countries caused by different histories of circulating SARS-CoV-2 variants and availabilities and types of vaccines between countries. The second describes my role in contributing to and coordinating a major effort to prioritise SARS-CoV-2 variants for phenotypic characterisation, as part of the NIH’s SARS-CoV-2 Assessment of Viral Evolution Program (NIH SAVE). This selection of topics represents the nature of the study of antigenically variable pathogens. The rapid genetic and antigenic evolution of these viruses provides both a rich opportunity to study evolution in real time, and poses an ongoing threat to public health. In the final chapter, I therefore discuss both future scientific avenues for this work and how it fits in with potential future vaccine strategies.