Investigating alternative AUG codon usage in avian influenza A virus segment 2

Influenza A viruses (IAVs) have a segmented, negative sense RNA genome. PB1-F2 is an IAV accessory protein encoded by segment 2, in the +1 reading frame. Avian IAVs predominantly encode full length PB1-F2s, whereas human IAVs often have stop codons resulting in C-terminus truncations or ablation of PB1-F2 expression. One reported function of PB1-F2 is innate immune antagonism, which requires C-terminal motifs. Full length PB1-F2 is translated from AUG 4 of segment two, however there are often one or more in frame downstream AUGs (AUGs 7, 8, or 9). Although there have been previous reports of expression from AUGs 7-9, no one has matched a protein product to a specific AUG codon. There have also been reports of an increase in N-terminal truncations of PB1-F2 in recent years. The aim of this PhD was to assess AUG 7-9 usage and possible phenotypes of the products, and assess whether the increase in N-terminal truncations allowed continued expression of possibly functional C-terminal fragments. Bioinformatic analysis of avian IAV segment 2 sequences indicates conservation of open reading frames encoding the PB1-F2 C-terminus, particularly in domestic birds, despite some subtypes acquiring N-terminal stop codons that persist through several years. Conservation of the C-terminus from AUGs 7 and 9 leads to the hypothesis that these serve as independent initiation codons for the C-terminus. C-terminal fragment expression from specific AUG codons has been shown in 293T cells using tagged proteins. Persistence of N-terminal truncations suggests that only producing C-terminal fragments has no detrimental effects on the virus. This could be a method of host-specific adaptation of the virus. Segment 2 mutants were generated in an avian H5N1 IAV background, which differed in the presence or absence of the AUG start codons or stop codon positions in PB1-F2. Significant differences in viral polymerase activity, measured using mini-replicon assays in avian cells, were observed for some stop codons, but none of the AUG mutants. It was also found that any C-terminal expression of PB1-F2 is sufficient to antagonise the poly(I:C) induced IFN response in avian cells, raising questions on the true minimal requirements for PB1-F2 function. Annotation of segment 2 sequences may need to be adapted to account for continued expression of C-terminal fragments from sequences with an N-terminal truncation