xi, 81 p. : ill. A print copy of this thesis is available through the UO Libraries. Search the library catalog for the location and call number. / Post-translational modification of proteins is an important cellular method of controlling various aspects of protein activity, including protein-protein interactions, half- life, and transport. An important class of post-translational modifications involves the ubiquitin family of proteins. In these modifications, a small protein, such as ubiquitin, is conjugated to a target protein through an isopeptide bond. Conjugation by a ubiquitin family member acts as a signal to regulate the activity, function, or stability of the target protein. Urm1, a ubiquitin-like protein conserved throughout all eukaryotes, was initially identified in S. cerevisiae. Loss of Urm1 leads to the disruption of a variety of cellular processes, including oxidative stress response, filamentous growth, and temperature sensitivity. This body of work comprises efforts to identify novel targets of Urm1, the mechanism by which Urm1 is attached to target proteins, and the physiological consequences of such conjugation. To gain understanding of the function and mechanism of Urm1 conjugation, the only known conjugate of Urm1, the peroxiredoxin reductase Ahp1, was examined in an effort to identify the site of modification on Ahp1 and to evaluate the physiological consequences of urmylation of Ahp1. I then completed a series of screens--a synthetic lethal screen, a two-hybrid screen, and a protein over-expression screen--to identify novel Urm1 conjugates and cellular functions dependent on Urm1. Of particular interest were genes identified in the synthetic lethal screen, namely PTC1, which encodes a protein phosphatase, and a set of genes encoding the Elongator complex, which functions in transcriptional elongation and tRNA modification. During this time period, other groups showed that thiolation of tRNAs depends on Urm1. Thus, Urm1 does not function only in protein conjugation, but also as a sulfur carrier in the thiolation of tRNA. Interestingly, I identified Elp2, a component of the Elongator complex, as a new Urm1-conjugate. Because Elp2 is also required for tRNA modification, perhaps Urm1 plays more than one role in tRNA modification. Loss of tRNA modification may disrupt many cellular functions and could explain the variety of urm1 mutant phenotypes. I have determined that all known Urm1 dependent processes are also associated with tRNA modification. / Committee in charge: Karen Guillemin, Chairperson, Biology; George Sprague, Advisor, Biology; Alice Barkan, Member, Biology; Kenneth Prehoda, Member, Chemistry; Tom Stevens, Outside Member, Chemistry

DNA is constantly subjected to chemical and mechanical damage. The ability to repair the lesions sustained is essential for all life. Double stranded DNA (dsDNA) breaks are especially toxic as both antiparallel strands of DNA are severed. The most high fidelity mechanism available to repair this damage is homologous recombination, a mechanism that uses homology from the sister chromatid to replace any lost information. Key proteins involved in maintaining genomic stability this way are conserved in all domains of life. One such component is the Mre11/Rad50 complex that is involved in the initial recognition of damage and recruitment of subsequent repair factors. Understanding the function of this DNA repair complex and any associated proteins has implications for human cancers and aging. The proteins of thermophilic archaea present an excellent opportunity to study these systems in a robust, tractable and eukaryote-like system. Archaea are in many ways biochemically unique, for example they are the only domain capable of methanogenesis. However archaea share a high level of homology with eukaryotes in many essential cellular processes such as DNA replication, homologous recombination and protein degradation. In thermophilic archaea the mre11/rad50 genes are clustered in an operon with the herA/nurA genes that form a helicase/nuclease complex. This has lead to speculation that the four proteins work together during homologous recombination to produce the 3' overhangs required by RadA to identify homology. As part of this investigation I have performed extensive bioinformatic searches of a variety of archaeal/bacterial systems. These analyses have revealed operonic linkages to other known recombinational helicase/nucleases, such as AddAB and RecBCD. These genomic linkages are especially prevalent in thermophilic organisms suggesting their functional relevance is particularly acute in organisms exposed to a high amount of genomic stress. Comparison of the evolutionary trees, constructed for each protein, makes a single genomic linkage event the most likely scenario, but cannot definitively exclude other possibilities. Exhaustive attempts were made to demonstrate an interaction between Mre11/Rad50 and HerA/NurA. Despite analysis by nickel/cobalt pulldown, immunoprecipitation, analytical gel filtration, ITC and OCTET an interaction could not be confirmed or definitively dismissed. However in the process an interesting Rad50 tetrameric assembly was identified and attempts were made to crystalize it. Hexameric helicases and translocases are key to the replication and DNA packaging of all cellular life and multiple viruses. The hexameric translocase HerA is a robust model for investigating the common features of multimeric ATPases as it is extremely stable and experimentally tractable. Here it is revealed that HerA exists in a dynamic equilibrium fluctuating between hexameric and heptameric forms with rapidly interchanging subunits. This equilibrium can be shifted to heptamer by buffering conditions or towards the hexamer by the physical interaction with the partnering nuclease NurA, raising the possibility that these alternate states may play a role in translocase assembly or function. A novel C-terminal brace, (revealed by a collaborative crystallographic structure) is investigated; as well as stabilizing the assembly, this brace reaches over the ATPase active site of its neighbouring subunit. It is seemingly involved in the conversion of energy generated by ATP hydrolysis into physical movement in the central channel of the hexamer. The regulation of homologous recombination is extremely important to prevent aberrant activity, resulting in mutations and genome reorganization. In eukaryotic organisms, it is well established that post-translational modifications and protein turnover at the proteosome play important roles in this control. In particular, there is significant interest currently in the ubiquination-proteasome destruction pathway as a mechanism for extracting DNA repair components from chromatin at the termination of the DNA repair process. To date no Ubiquitin proteins have been identified in the Archaea, however related proteins URMs/SAMPs (Ubiquitin Related Modifier/Small Archaeal Modifier Protein) have previously been identified. URMs are thought to have evolved from a common antecedent to eukaryotic ubiquitin and likely represent an evolutionary 'missing link' in the adaption of sulphur transfer proteins for covalent modifications. There has been speculation that Urm1 may play a similar role to ubiquitin in the proteasome degradation pathway and we have recently provided evidence to corroborate this. Here the potential for modification of Mre11/Rad50/HerA/NurA by Urm1 was investigated. Indeed Rad50 shows evidence of clear urmylation both in vivo and in vitro. Western blotting and mass spec analysis confirmed the covalent attachment of Urm1 to Rad50. Furthermore I present preliminary evidence that this urmylation can lead to the destruction of Rad50 via a direct physical interaction with the proteasome. This is the first evidence of such a regulatory system for Rad50. Investigating the urmylation and destruction of Rad50 was closely linked to investigating the archaeal proteasome, a close homologue of the eukaryotic proteasome. To date the majority of archaeal core proteasomes examined were believed to consist of only two subunits; alpha and beta. The subunits are arranged into heptameric rings, which then form an alpha/beta/beta/alpha stack with a single channel running through the centre of all four rings. Here we reveal that in Sulfolobus species the inner catalytic chambers are made up of mixed beta rings composed of two subunits. The first plays a crucial structural role but appears catalytically inert, while the second conveys catalytic activity. Here we investigate an inactive complex, containing only the structural beta subunit, and an active complex, containing both beta subunits. First, electron microscopy was performed on both complexes revealing the expected four-layered toroidal stack. Both complexes were subsequently investigated crystallographically. A 3.8 Å structure was determined for the inactive complex. As well as being one of the few archaeal core proteasome structures, this is also an important first step towards structurally investigating the novel three-subunit proteasome. The discovery of active and inactive beta subunits in the archaea brings them even closer to eukaryotic proteasomal systems, making the archaea even more valuable as model systems.

Posttranslational modification (PTM) of proteins is essential to maintain homeostasis and viability in all eukaryotic cells. Hence, besides the sequence and 3D folding of a polypeptide, modification by multiple types of PTMs, ranging from small molecular groups to entire protein modules, adds another layer of complexity to protein function and regulation. The ubiquitin-like modifiers (UBLs) are such a group of evolutionary conserved protein modifiers, which by covalently conjugating to target proteins can modulate the subcellular localization and activity of their targets. One example of such a UBL, is the Ubiquitin related modifier 1 (Urm1). Since its discovery in 2000, Urm1 has been depicted as a dual function protein, which besides acting as a PTM, in addition functions as a sulfur carrier during the thio-modification of a specific group of tRNAs. Due to this dual capacity, Urm1 is considered as the evolutionary ancestor of the entire UBL family. At present, it is well established that Urm1, with help of its dedicated E1 enzyme Uba4/MOCS3, conjugates to multiple target proteins (urmylation) and that Urm1 thus plays important roles in viability and the response against oxidative stress. The aim of this thesis has been to, for the first time, investigate the role of Urm1 and Uba4 in a multicellular organism, utilising a multidisciplinary approach that integrates Drosophila genetics with classical biochemical assays and proteomics. In Paper I, we first characterized the Drosophila orthologues of Urm1 (CG33276) and Uba4 (CG13090), verified that they interact physically as well as genetically, and that they together can induce urmylation in the fly. By subsequently generating an Urm1 null Drosophila mutant (Urm1n123), we established that Urm1 is essential for viability and that flies lacking Urm1 are resistant to oxidative stress. Providing a molecular explanation for this phenotype, we demonstrated an involvement of Urm1 in the regulation of JNK signaling, including the transcription of the cytoprotective genes Jafrac1 and gstD1. Besides the resistance to oxidative stress, we have moreover (Manuscript IV) made an in-depth investigation of another phenotype displayed by Urm1n123 mutants, an overgrowth of third instar larval neuromuscular junctions (NMJs), a phenotype which is shared also with mutants lacking Uba4 (Uba4n29). To increase the understanding of Urm1 in the fly, we next employed a proteomics-based approach to identify candidate Urm1 target proteins (Paper II). Using this strategy, we identified 79 Urm1-interacting proteins during three different stages of fly development. Of these, six was biochemically confirmed to interact covalently with Urm1, whereas one was found to be associated with Urm1 by non-covalent means. In Manuscript III, we additionally identified the virally encoded oncogene Tax as a target of Urm1, both in Drosophila tissues and mammalian cell lines. In this study, we established a strong correlation between Tax urmylation and subcellular localization, and that Urm1 promoted a cytoplasmic accumulation and enhanced signalling activity of Tax, with implications for a potential role of Urm1 in Tax-induced oncogenesis. Taken together, this thesis provides a basic understanding of the potential roles and targets of Urm1 in a multicellular organism. The four studies included cover different aspects of Urm1 function and clearly points towards a highly dynamic role of protein urmylation in fly development, as well as in adult life.