Investigation of the FAT10 conjugation pathway

Abstract

Posttranslational modifications are important means to alter a proteins’ activity, function, stability or its intracellular localization. The post-translational conjugation of one or more molecules of Ub and ubiquitin-like proteins (UBLs) to selected substrates, namely ubiquitination, is one of the most important and multifaceted regulatory mechanisms in biology and requires the sequential interaction of a 3-step enzyme cascade. The initiating step for ubiquitin or UBL conjugation to its target proteins is the formation of an energy-rich thioester at its C-terminus. This activation takes places through an ubiquitin activating E1 enzyme. The activated ubiquitin is then transferred in a trans-thioesterification cascade to one of multiple E2 conjugating enzymes. The ubiquitin-charged E2 enzyme and a specific substrate protein are then both bound by a ubiquitin protein ligase (E3), which catalyzes the transfer of the activated ubiquitin between its carboxy-terminal hydroxyl-group onto the ε-amino-side chain of an internal lysine residue of the acceptor protein. Canonical ubiquitin-like proteins (UBLs) such as ubiquitin, SUMO, NEDD8, and ISG15 are transferred by a similar E1-E2-E3 multi-enzyme cascade to its targets. For the ubiquitin-like modifier FAT10, the enzyme cascade has not yet been characterized completely.The IFN-γ and TNF-α inducible modifier FAT10 is a young member of ubiquitin-like proteins, which can be conjugated to target proteins via its C-terminal diglycine motif. Moreover, it is to date the only identified ubiquitin-like protein, which can assign substrate proteins, in an ubiquitin-independent manner, for proteasomal degradation.Recently, the FAT10 activating enzyme (E1) UBA6 and a FAT10 conjugating enzyme (E2), namely USE1 was identified, which interestingly, was at the same time the first known substrate for FAT10, as it was auto-FAT10ylated in cis.The aim of this thesis was the characterization of the FAT10 conjugation pathway, starting with the identification of UBA6 interacting proteins in a yeast two-hybrid screen.Given that at the beginning of the doctoral thesis no FAT10 conjugating E2 enzymes were known so far, the focus here was the identification of potential FAT10 E2 enzymes. In a yeast two-hybrid approach, a direct interaction of UBA6 and the C-terminal end of BRUCE, containing the entire highly conserved ubiquitin conjugating domain, could be shown. This interaction was further characterized with a construct encoding for full length BRUCE via co-immunoprecipitation experiments, especially in terms of exhibiting a putative FAT10 E2 function.A further aim of this doctoral thesis was the identification of putative FAT10 substrates. In a yeast two-hybrid screen with the putative FAT10 E3 ligase TRIM11 and a cDNA-library from human thymus, a specific interaction between TRIM11 and JunB as well as with Ambra1 could be observed. The RING finger containing E3 ligase TRIM11 was previously identified in a yeast two-hybrid screen with FAT10 and a cDNA-library from human thymus. For both JunB and Ambra1, a specific interaction with TRIM11 could be verified in co-immunoprecipitation assays. Moreover, it could be demonstrated that JunB becomes covalently linked to either ubiquitin or FAT10, but not in the presence of a FAT10 mutant, lacking the di-glycine motif, which is required for isopeptide linkages to substrate proteins. This result points to a conjugate formation between JunB with the ubiquitin-like modifier ubiquitin and FAT10, which indicates, that JunB is a ubiquitin as well as FAT10 specific substrate. Moreover, a non-covalent linkage of JunB with either ubiquitin or FAT10 could be detected. Furthermore, cycloheximide experiments revealed evidence that not only the isopeptide linked form of JunB and FAT10 became assigned for proteasomal degradation but also JunB, which was non-covalently modified with FAT10. Proteasome inhibition with MG132 led to an accumulation of the JunB-FAT10 conjugate, which would be expected for a FAT10 substrate. The role of TRIM11 as a FAT10 specific E3 ligase could not be solved definitely, due to the fact that JunB, in presence of recombinant FAT10, UBA6 (E1), USE1 (E2) and TRIM11 (putative FAT10 E3) did not become FAT10ylated in vitro. TRIM11 overexpression resulted in a decreased protein level of ubiquitin, FAT10, JunB and also the JunB-ubiquitin and JunB-FAT10 conjugates.A clear co-localization of FAT10 and JunB at the nuclear membrane could be detected by means of confocal microscopy. MG132 treatment caused the translocation of JunB into the cytosol. As a presumable functional consequence, JunB FAT10ylation led to a reduced JunB trans-activating capacity in reporter assays.In case of Ambra1, a non-covalent interaction with ubiquitin and FAT10 could be verified in co-immunoprecipitation experiments. Addition of MG132 led to an accumulation of over-expressed Ambra1, indicating that Ambra1 becomes degraded by the proteasome. However, no accumulation after proteasome inhibition was observable, when Ambra1 was ectopically co-expressed with FAT10, suggesting that the non-covalent interaction with FAT10 led to a different degradation mechanism other than the proteasome. Experiments with confocal microscopy illustrate an unambiguous co-localization of FAT10 and Ambra1 in punctuated structures, suggesting that the interaction of Ambra1 and FAT10 led to translocation of both proteins into punctuated structures.