We are interested in how the rapid evolution of influenza viruses comes about, and how this evolution contributes to the ecology of the virus within the wide range of hosts it infects. We use molecular and classical virology techniques, single cell approaches, mathematical modeling and in vivo infection models. Most of our current projects focus on influenza A virus ressortment, the process by which two viruses that co-infect the same cell exchange intact gene segments. Reassortment between influenza A viruses of distinct lineages and/or adapted to distinct host species is frequently implicated in the emergence of zoonotic and pandemic strains. Reassortment among co-circulating human influenza viruses has furthermore led to the formation of novel epidemic strains and facilitated the spread of antiviral drug resistant viruses. In short, reassortment is important in the evolution and epidemiology of influenza viruses. Our work seeks to improve our understanding of reassortment at a fundamental level by defining the underlying factors that dictate the frequency of reassortment, and the implications of reassortment for IAV evolution.

Efficiency of influenza A virus reassortment
How frequently does reassortment occur in a host co-infected with two influenza A viruses? This simple question has been challenging to answer due to the wide range of phenotypes exhibited by reassortant viruses. The differing replication rates of parental and reassortant viruses makes reassortment difficult to quantify in an unbiased fashion. To circumvent this problem, we devised a strategy to measure reassortment between two viruses of the same strain that differ only by silent genetic tags (Marshall et al. 2012). We call these parental viruses “wildtype” and “variant”. The genotypic and phenotypic similarity of the parental viruses ensures that all 256 possible progeny genotypes formed are of equivalent fitness. For this reason, genotyping of viruses produced during co-infection gives an accurate measure of reassortment efficiency. We are applying this strategy to track viral genetic diversification in cell culture and in vivo under a range of co-infection conditions. Results to date suggest that reassortment is surprisingly efficient in a population of co-infected cells and within an intact host (Tao et al. 2015). Current efforts on this project are aimed at assessing species-specific differences in natural hosts.

Sources of constraint in influenza virus reassortment
When two very different influenza A viruses reassort, most progeny genotypes suffer fitness defects due to sub-optimal interactions among viral components (Phipps et al. 2017). This phenomenon is a type of epistasis, referred to as “segment mismatch” in the influenza field. Importantly, it is a major factor determining the outcome of mixed infections. The precise causes of segment mismatch are poorly defined and are difficult to tease apart when all are manifested simultaneously in a standard co-infection system. To simplify the problem, and better understand the underlying mechanisms leading to segment mismatch, we are using our wildtype/variant system (described above) as a starting point to generate parental viruses that differ only in targeted ways. Current efforts are focused on determining the extent to which sequence divergence in the viral RNA packaging signals disfavors the formation of reassortant genotypes (White et al. 2017).

Incomplete genomes as drivers of influenza virus diversification
For decades it has been known that a large proportion of virions that influenza and many other RNA viruses produce cannot initiate productive infection. The biological significance of these particles has, however, remained unclear. We recently found that semi-infectious influenza virus particles, which deliver an incomplete genome to the site of replication and comprise 90-99% of virions, significantly augment reassortment (Fonville et al. 2015). This finding suggests that these “defective” particles are not simply an unfortunate byproduct of rapid replication, but rather play an important role in viral evolution. We are currently working to understand how incomplete genomes arise and test the hypothesis that the presence of incomplete genomes accelerates influenza virus evolution under selection.