Speaker
Description
All viruses must solve the problem of ensuring their genetic material is recognised and packaged by a viral capsid. This problem is especially intricate in viruses with segmented genomes, such as influenza viruses. Interchangeability of packaging signals in different influenza A subtypes remains a key unanswered question in determining strains' pandemic potential. The packaging signals' specificity makes them candidates for drug targets with low risk of host cross-reactivity.
We have recently refined a computational technique for finding regions of high conservation in RNA virus coding regions. We applied this technique to eight different haemagglutinin/neuraminidase/host combinations of influenza A (983,552 gene sequences analysed). We then used the predicted high conservation regions as input to a RNA structural prediction algorithm.
Many of our predictions better delineate regions of RNA crucial to influenza A packaging. One such region is conserved between subtypes, but with markedly different primary sequence in the different subtypes, raising the possibility that segments from different subtypes may not be fully interchangeable. Our technique also predicts regions of conservation beyond packaging signals. Other predictions include an association between avian strains and conservation around the initiation site of an N-terminal truncated form of PB1; the reason for this conservation is unclear but may represent a form of host adaptation that needs to switch when infecting different hosts.
Our work provides computational insights into RNA structure and function across the influenza A genome. The resultant structural predictions explain a number of the functional results of previous packaging signal mutational analyses. They naturally lead to predictions for experiments that
