Speaker
Description
Several SARS-CoV-2 Variants of Concern (VOCs), including Alpha, Omicron, and more recently, BA.2.86/JN.1, are thought to have emerged from chronically infected individuals, including immunocompromised or immunosuppressed individuals. This hypothesis is supported by several lines of evidence: (1) Immediate ancestors of these VOCs have not been sampled; (2) These VOCs exhibited higher-than-expected divergence when they first emerged; (3) These VOCs carry a large number of nonsynonymous mutations in the Spike gene, particularly in epitope regions. Given the likely role of chronic infections in VOC evolution, we developed a within-host model that explores patterns of adaptive viral evolution and the drivers that shape these patterns. Our model builds on the tunably rugged Rough Mount Fuji fitness landscape model by incorporating synonymous mutations as well as three types of nonsynonymous mutations: phenotypic mutations that only impact within-host replication fitness, immune escape mutations, and pleiotropic mutations that simultaneously impact replication fitness and immune escape. We explore various parameterizations of this model to determine what factors drive within-host viral adaptation. We first find that an intermediate degree of ruggedness is consistent with observations of parallel substitutions across chronically infected individuals and with observations of multiple co-circulating within-host lineages. This finding supports empirical studies that have found evidence for epistatic interactions within the Spike gene. Second, we find that pleiotropic mutations can either facilitate or impede viral adaptation with respect to replication fitness, depending on the ruggedness of the landscape and the extent to which a mutation results in immune escape. Together, these results help explain observed patterns of viral evolution within prolonged SARS-CoV-2 infections and may help to project phenotypes that might be seen in future VOCs.
