To our surprise, many of the insertion sites identified in
MDCK cells were also identified in our egg competition analysis,
indicating that, for the most part, diversity of the recovered
viruses was preserved (Fig. 3). In fact, essentially every site
identified in the HA head in MDCKs was also identified in the
egg analysis. In contrast, however, we failed to recover mutations
in the HA2 domain from the eggs (Fig. 3). We interpret this data
to mean that insertions in the head domain are relatively well
tolerated, whereas those in the stalk domain may lead to a loss
of fitness.