Urm1 in tRNA thiolation and protein modification

Abstract

The eukaryotic ubiquitin-related modifier Urm1 is a conserved member of the ubiquitin (Ub) family. Like other members of this family, Urm1 adopts a β-grasp fold and terminates with a diglycine motif. Urm1 was identified a decade ago by its homology to two prokaryotic members of the Ub family, MoaD and ThiS. We demonstrate that, unlike Ub and other Ub-like protein modifiers, the C-terminal glycine of Urm1 is derivatized to a thiocarboxylate by the addition of sulfur. Thiocarboxylation of Urm1 requires the ATP-dependent activation of its terminal glycine by the E1-like domain of the enzymatic activity MOCS3 (Uba4 in yeast). The activated glycine is subsequently modified with sulfur by the sulfurtransferase domain of MOCS3. This mechanism is reminiscent of a strategy used by prokaryotes to modify the sulfur carriers MoaD and ThiS with a C-terminal thiocarboxylate. We demonstrate that thiocarboxylated Urm1 functions as sulfur donor in the thio-modification of certain tRNAs. Urm1 binds the thiouridylase ATPBD3 (Ncs6 in yeast), which mobilizes sulfur from the C-terminus of Urm1 and transfers it to the wobble uridine of tRNALys(UUU), tRNAGlu(UUC), and tRNAGln(UUG). A deficiency in the Urm1 pathway suppresses thiolation of tRNAs in both S. cerevisiae and mammals, indicating that this pathway is conserved across eukaryotes. Early studies in yeast suggested that Urm1 can also be covalently attached to proteinaceous substrates (urmylation). However, the underlying mechanism of conjugation remained unexplored and only one target of urmylation was identified. We show that an oxidizing environment promotes the conjugation of Urm1 to its substrates both in vitro and in vivo. Oxidant-induced urmylation requires the C-terminal thiocarboxylate of Urm1 and resembles many aspects of ubiquitylation, including the formation of a thioester intermediate with MOCS3 and a covalent lysine-linked adduct. Using a proteomic approach, several new substrates for urmylation were identified, including components of the Urm1 pathway itself and proteins implicated in nuclear translocation. The functional consequences of a defect in the Urm1 pathway were studied in a transgenic mice that express a dominant negative variant of Urm1 in a doxycycline-inducible manner. This variant lacks the C-terminal glycine of Urm1 (Urm1 ΔG) and consequently fails to be thiocarboxylated. Urm1 ΔG is robustly expressed in several tissues upon doxycycline addition, and reduces tRNA thiolation by 75%. While adult mice that express Urm1 ΔG appear healthy, they fail to generate offspring when maintained on doxycycline, unlike their wild-type counterparts. Finally, we investigated the expression and function of an alternate isoform of Urm1 (Urm1B). The N-terminal region of Urm1B is identical to Urm1, but its C-terminus is distinct, lacks a terminal diglycine motif, and consequently fails to be thiocarboxylated. Urm1B is rapidly degraded by the proteasome, unless stabilized by co-expression of ATPBD3. The dual role of Urm1 in tRNA thiolation and protein modification underscores the notion that Urm1 is an evolutionary intermediate that integrates the functions of both prokaryotic sulfur carriers and eukaryotic protein modifiers. Further experiments may reveal additional components of the Urm1 pathway and determine the impact of this pathway on translation, stress response, and embryonic development.