UCSF

Protein synthesis is fundamental to all life. The ribosome is central to this process and is aided at every stage by numerous translation factors. Innate defects within the translation apparatus, as well as external environmental stressors, may cause the ribosome to stall. This threatens to expose the cell to potentially toxic, partially synthesized protein products, while sequestering the ribosome in an unproductive state. A number of ribosome-sensing pathways have begun to be elaborated. Of these, the HBS1L/Pelota complex and the E3 ligase ZNF598 constitute the best studied initiators into ribosome quality control. HBS1L/Pelota preferentially sense ribosomes stalled with an empty A site, and ZNF598 recognizes ribosome collisions. While many quality control pathways have been studied in the context of mRNA or translation inhibitor-mediated stalls, the cellular response to ribosomes specifically stalled with an occluded A site has not been systematically studied.

In Chapter 1, we describe the use of a natural product derivative, ternatin-4, as a chemical probe to identify novel ribosome-sensing quality control pathways. Ternatin-4 traps the eEF1A/GTP/aminoacyl-tRNA ternary complex at the ribosome A site, preventing accommodation. We made the serendipitous discovery that ternatin-4 causes near complete degradation of eEF1A in a translation and ubiquitin-dependent manner. Using a fluorescent reporter-based CRISPRi screen, we identify two E3 ligases, RNF14 and RNF25, as central players in this pathway. Characterization of proteome-wide, ternatin-induced, and RNF14/25-dependent ubiquitination marks revealed eEF1A, RPS27A, and GCN1 as key targets for ubiquitination. We define a role for RNF14 in direct ubiquitination of eEF1A, whereas RNF25 directly ubiquitinates RPS27A. The ribosome collision sensor GCN1 binds RNF14, likely through the RNF14 RWD domain, and is required for eEF1A degradation. Ubiquitination of RPS27A is also prerequisite for eEF1A degradation. Thus, two combined signaling inputs, collision sensing via GCN1 and RPS27A ubiquitination by RNF25, allow RNF14 activation and ultimately, clearance of stalled eEF1A from the ribosome.

In Chapter 2, we expand the scope of the RNF14/RNF25/GCN1 pathway to additional activators and substrates. We first examine differences in ternatin-induced eEF1A degradation which arise between the two isoforms of eEF1A, eEF1A1 and eEF1A2. Next, we expand the pathway to a poorly understood translational GTPase, DRG1, which may represent a constitutive RNF14 substrate. We show that the RNF pathway recognizes not only translation elongation factors at the A site, but also the termination factor eRF1 when stalled by a small molecule, SRI-41315, or a hydrolysis-deficient AAQ mutation. We finally examine the role of environmental stresses in activating the RNF14/RNF25/GCN1 pathway and find that ultraviolet light, which likely causes decoding defects through mRNA crosslinks, stimulates ubiquitination of eEF1A, ribosomal proteins, and GCN1 in an RNF14-dependent manner. Collectively, this thesis presents a mechanism for a ribosome-sensing pathway which responds to multiple disruptions to translation factors at the A site.