Mechanism of Ribosome Rescue by Alternative Release Factor B

Abstract

Rescue of ribosomes stalled on non-stop mRNAs is essential in bacteria and in the mitochondria of all eukaryotes. The lack of a stop codon causes ribosomes to stall at the 3’ end of mRNA with an empty A site, and specialized rescue mechanisms are required to release the truncated translation product and allow ribosome recycling. Of the known bacterial ribosome rescue systems, only ArfB has intrinsic peptidyl-tRNA hydrolysis activity, and is conserved in all eukaryotic mitochondria. The precise mechanism of ArfB-mediated ribosome rescue is not well understood. In this study, we use rapid kinetics in conjunction with FRET- and anisotropy-based methods to construct a detailed kinetic model for ArfB-mediated ribosome rescue. We find that ArfB binds to the ribosome rapidly regardless of mRNA length, and that the association rate for the majority of ArfB molecules in the ensemble is close to diffusion controlled. This is likely due to the flexibility of its unstructured C-terminal tail, which allows ArfB to associate with the ribosome in different orientations. A slow engagement step follows, which allows ArfB to discriminate between stalled ribosomes with and without mRNA extending past the P site. The engaged state of ArfB involves specific interactions that strongly increase the affinity of ArfB for the ribosome, and is the active state in which ArfB performs peptidyl-tRNA hydrolysis. ArfB dissociates slowly from the post-hydrolysis ribosome, which leads to a low turnover rate that can be increased somewhat by the presence of ribosome recycling factors. Cryo-electron microscopy structures of ArfB bound to two different substrates, one with 3’ mRNA extensions and one without, provide structural snapshots of ArfB-ribosome complexes along the rescue pathway. By superimposing the NMR structure of ArfB, in which the C-terminal tail is unstructured, on the pre-hydrolysis stalled ribosome, we model possible initial binding complexes that support the kinetic data on multiple binding rates. The 2.6 Å structure of ArfB in the active state show an extensive network of specific interactions between ArfB and the ribosome, all of which involve residues that were previously found to be functionally important in a mutational study. These structures provide explanations for the high affinity and slow dissociation of ArfB. Our study demonstrates the role of intrinsic disorder in protein-ribosome interactions and provides a basis for the understanding of ArfB-like proteins in mitochondria.