(A) General workflow for measuring the mutation rate using mutant ΔHA-GFP viruses. Parallel cultures of MDCK-HA cells were infected with passage one stocks of mutant ΔHA-GFP viruses at low multiplicity. The time for initial replication was varied to allow for a number of replicated viruses and revertants adequate to measure the mutation rate for a given class. Supernatants were transferred to 96-well plates of MDCK cells and incubated for 14 hr to allow for infection and GFP expression in target cells. The mutation rate for each mutant ΔHA-GFP virus and class was calculated as described in the methods and text based on the initial and final titer (Ni and Nf, anti-GFP positive infected cells) and proportion of cultures with no revertants (P0, wells without green fluorescence). (B–D) Specificity of the reversion to fluorescence assay. The (B) A to G, (C) G to A, and (D) G to C mutation rates for A/Puerto Rico/8/1934 H1N1 were measured at 37°C in cells pretreated with 0.625 μM 5-azacytidine (AzaC), 15 μM 5-fluorouracil (5 FU), or 2.5 μM ribavirin (Riba). No data are shown for G to C with 2.5 μM ribavirin because large titer decreases upon drug treatment prohibited measurements. Filled symbols represent measurements in which P0 is between 0.1 and 0.69, where the assay is most precise. Open circles represent data with P0 between 0.7 and 0.9. Arithmetic means are indicated. A one-way ANOVA with a Dunnett’s correction for multiple comparisons was used for each mutation class to compare each drug treatment to no drug treatment. *p<0.05, **p<0.01, ***p<0.005. Plotted data are in Figure 3—source data 1.