Abstract
Influenza virus has a high mutation rate, and this low replicative fidelity contributes to its capacity for rapid evolution. Clonal sequencing and fluctuation tests have suggested that the mutation rate of influenza A virus is 7.1 × 10−6 − 4.5 × 10−5 substitutions per nucleotide per cell infection cycle and 2.7 × 10−6 − 3.0 × 10−5 substitutions per nucleotide per strand copied (s/n/r). However, sequencing assays are biased toward mutations with minimal impacts on viral fitness and fluctuation tests typically investigate only a subset of the twelve mutational classes. We developed a fluctuation test based on reversion to fluorescence in a set of virally encoded mutant green fluorescent proteins. This method allowed us to measure the rates of selectively neutral mutations representative of all 12 mutational classes in the context of an unstructured RNA. We measured an overall mutation rate of 1.8 × 10−4 s/n/r for PR8 (H1N1) and 2.5 × 10−4 s/n/r for Hong Kong 2014 (H3N2). The replication mode was linear. The mutation rates of these divergent strains are significantly higher than previous estimates and suggest that each replicated genome will have an average of 2-3 mutations. The viral mutational spectrum is heavily biased toward A to G and U to C transitions, resulting in a transition to transversion bias of 2.7 and 3.6 for the two strains. These mutation rates were relatively constant over a range of physiological temperatures. Our high-resolution analysis of influenza virus mutation rates will enable more refined models of its molecular evolution.
Significance The rapid evolution of influenza virus is a major problem in public health. A key factor driving this rapid evolution is the virus’ very high mutation rate. We developed a new method for measuring the rates of all 12 mutational classes in influenza virus, which eliminates some of the biases of existing assays. We find that the influenza virus mutation rate is much higher than previously reported and is consistent across two distinct strains and a range of temperatures. Our data suggest that influenza viruses replicate at their maximally tolerable mutation rates, highlighting both the virus’ evolutionary potential and its significant constraints.