Proteomic and functional studies of the influenza A virus PA-X protein

Abstract

Viral RNA endonuclease activity is required for influenza A virus (IAV) replication. This activity resides in the polymerase acidic (PA) protein, which assembles into viral RNA-dependent RNA polymerase (RdRp) complexes and cleaves nascent host pre-mRNAs proximal to 5’-m7G caps, creating primers for viral mRNA synthesis. A rare (+1) ribosomal frameshifting event during translation of the PA open reading frame (ORF) creates the polymerase acidic-X (PA-X) protein. PA-X retains the amino-terminal PA RNA endonuclease domain, but contains a novel short carboxy-terminus, dubbed the X-ORF. Accumulating evidence indicates PA-X is a host shutoff protein functioning in the nucleus, selectively cleaving RNAs transcribed by host RNA polymerase II (pol II) while sparing RNA pol I, III, and viral transcripts. The molecular mechanism for this specificity remains to be elucidated. I hypothesize that PA-X gains access to target RNAs by X-ORF-mediated interaction with host proteins. In this study, I used a proximity labeling proteomic method known as BioID to identify host proteins that interact with the X-ORF. In BioID, fusion of the bait protein to a promiscuous biotin ligase allows efficient biotinylation of lysine residues on nearby proteins. X-ORF baits subjected to BioID included a 61-amino acid variant from A/Puerto Rico/8/1934 (H1N1) and a truncated 41-amino acid variant from A/California/7/2009 (H1N1), as well as a mutant X-ORF lacking basic residues required for nuclear localization. Affinity-purified proteins were trypsinized, subjected to reductive dimethylation with stable isotope tags, and identified by mass spectrometry. Using quantitative analysis, 29 high-confidence candidate X-ORF-interacting proteins were identified. X-ORF interacting proteins were validated using a luciferase-based functional assay in cells where each candidate host gene was silenced by short-hairpin RNAs. Through this study, the cleavage factor I (CFIm) complex proteins, cleavage and polyadenylation specificity factor subunit 5 (CPSF5) and CPSF6, were identified as required for PA-X function. The CFIm complex is poorly characterized but is known to influence site selection for mRNA 3’-end cleavage and polyadenylation. My observations are concordant with the emerging model for PA-X host shutoff activity, which has been shown to require canonical mRNA 3’-end processing mechanisms.