The Israeli Acute Paralysis Virus IRES captures host ribosomes by mimicking a ribosomal state with hybrid tRNAs

Francisco Acosta-Reyes; Ritam Neupane; Joachim Frank; Israel S. Fernández

The SL-III seems to play an important role in orienting the IRES unit formed 295 by the ASL/SL-III to a position that perfectly matches that of a canonical P 296 site tRNA (Fig. 6D) . Nucleotides belonging to the SL-III establish contacts 297 with several residues of ribosomal proteins uL5 and uL16 as well as with the 298 28S rRNA nucleotide A4255, all components of the large ribosomal subunit 299 (60S) (Fig. 6C, right) . 300 In canonical translocation, the movement of tRNAs and mRNA has to be 301 coordinated in order to vacate the ribosomal A site for the next incom- icking a pre-translocation state of the ribosome with tRNAs (Fig. 7, top) . In The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/606236 doi: bioRxiv preprint center of the 40S, the IAPV-IRES has to move (translocate) from the A to the P site. We were able to gain structural information of this transi- (Fig. 6C) . Additionally, featuring extremely dynamic capabilities, a 388 single-stranded region of the IAPV-IRES termed the VRL (Fig. 2D) is able 389 to modify its configuration in a context-specific manner. In the context of Recombinant eRF1* was purified according to a previously described protocol [3] . The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/606236 doi: bioRxiv preprint templates were generated using density data obtained from a screening ses-72 sion on a F20 microscope that had employed Gaussian picking. Using these 73 templates particle picking was performed using GAUTOMACH and a par-74 ticle diameter value of 320Ã…. The picked particles were manually screened 75 on the micrographs to remove problematic regions. All 2D and 3D classifica-76 tions and refinements were performed using RELION [9] . The picked particles 77 were binned 4 times and subjected to a 2D classification to separate the 40S 78 and 80S particles. We then employed 3D Refine to generate initial consensus 79 models from both the 40S and 80S particle sets (Fig. S1 ). We then used Models for the mammalian ribosome and eRF1* were docked into the maps