Predicting the evolution of influenza virus’s defective interfering genomes

Influenza A virus (IAV) rapidly evolves between seasons, enabling the virus to reinfect previously infected hosts. To inform vaccine strain selection, considerable research has therefore focused on predicting the genetic and antigenic evolution of influenza’s hemagglutinin protein, the major target of our immune response. In contrast, little attention has been placed on characterizing the evolutionary dynamics of other components of influenza’s genomic diversity. This genomic diversity encompasses fully-infectious particles, semi-infectious particles that express an incomplete set of essential viral genes, and defective interfering particles (DIPs) that interfere with the replication of the wild-type IAV. Understanding the evolutionary dynamics of DIPs is important given recent work showing that IAV strains can differ in their rates of DIP generation and that these differences affect the severity of infection outcome. Here, we detail recent work from a set of experimental passage studies that addresses the predictability of DIP evolution. Specifically, we statistically interface mechanistic models with data from these passage studies, including deep sequencing data. We find evidence for both selection and drift in driving DIP evolutionary dynamics, with predictable DIP evolution on the largest gene segments emerging from what appears as transitive competitive relationships between the DIPs. Unlike for antigenic phenotypes, where being different is the primary factor impacting fitness, this finding indicates that some DIPs are fitter than others likely as a consequence of either more efficient replication or packaging in coinfected cells. These findings are encouraging for efforts that focus on the development of therapeutic interfering particles for influenza disease control. This is collaborative work between Molly Gallagher/Jeremy Harris/Koelle and Fadi Alnaji/Brigitte Martin/Chris Brooke at UIUC.