IFIT3 and IFIT2/3 promote IFIT1-mediated translation inhibition by enhancing binding to non-self RNA

Renata C Fleith; Harriet V Mears; Edward Emmott; Stephen C Graham; Daniel S Mansur; Trevor R Sweeney

The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/261776 doi: bioRxiv preprint complex closely resembles the IFIT1:IFIT2:IFIT3-containing 150-200 kDa complex previously identified in HeLa cell lysates (29) , supporting its biological relevance. Unlike the IFIT1:IFIT3 tetramer, the IFIT1:IFIT2:IFIT3 tetramer was unstable and precipitated during incubation at 4 ï‚°C. In contrast, the heterotrimer was stable and enabled us to investigate the impact of IFIT2:IFIT3 complexing on the translation inhibition activity of IFIT1 (discussed below). A recent report described the crystal structure of wild type IFIT1 and revealed a dimerisation interface between Cterminal TPRs (BioRxiv: https://doi.org/10.1101/152850), responsible for the concentration dependent homo dimerisation of the protein, confirmed here. Importantly, mutation of this interface had no impact on the translation inhibition activity of IFIT1 (BioRxiv: https://doi.org/10.1101/152850 and Figure 5E ). We observed that IFIT3 shares multiple key residues involved in the IFIT1 'YxxxL' dimerisation motif, also in a C-terminal TPR. We therefore hypothesized that this motif is the site of interaction between IFIT1 and IFIT3. Mutational analysis of IFIT1 confirmed that this motif is indeed required for interaction with IFIT3 ( Figure 5 ). The double-mutant IFIT1-Y460E/L464E did not interact with IFIT3 when analysed by SEC. Since the wild type IFIT1 homodimer is less stable than the IFIT1:IFIT3 complex, as shown by our SEC-MALS analysis, we propose that the functionally-important role of the 'YxxxL' motif in humans is mediating interaction between IFIT1 and IFIT3 (shown schematically in Figure 6 ). Importantly, the crystal structure of IFIT5 reveals that, although critical interacting residues may be conserved, they are buried in an interface with a terminal helix not present in IFIT1 (8) (9) (10) , explaining why IFIT5 does not interact with IFIT3 ((12) and Figure 1 ). While the 'YxxxL' motif is conserved in mouse Ifit1b1, Ifit3 is truncated such that the 'YxxxL' motif is absent. In mice and other rodents, the Ifit1b1 gene has been duplicated (6) and it is possible that these extra IFITs could compensate for the disrupted murine Ifit1b1-Ifit3 interaction. This suggests species differences in the role of this motif, and in IFIT oligomerisation in general, that must be considered when examining phenotypes of small animal models used to examine the impact of IFIT depletion.