Mechanism underlying linkage-selective polyubiquitin recognition

• Main Author: Kristariyanto, Yosua Adi
• Format: Dissertation
• Language: English
• Published: University of Dundee 2017

Ubiquitylation is a post-translational modification that conjugates ubiquitin (Ub) onto proteins. When Ub itself is ubiquitylated, eight types of polyubiquitin (polyUb) chains can be formed: M1, K6, K11, K27, K29, K33, K48 and K63. Each one of these different chain types can couple modified proteins to diverse cellular signalling pathways. For example, proteins modified with K48 polyUb chains are targeted for proteasomal degradation. In contrast, modification with K63 polyUb chains has non-proteolytic roles in DNA damage response and immune signalling pathway. However, the cellular roles of K29 and K33 polyUb were poorly understood. Central to the myriad of polyUb signalling are Ub-binding domains (UBDs). Some UBDs can differentiate between types of polyUb chains and thus, are essential for the specificity of Ub signalling, but these mechanisms of linkage-selective polyUb recognition are less known. The work presented in my thesis centres on understanding the mechanisms underlying linkage-selective polyUb recognition by UBDs. I started this study by developing methods to enzymatically assemble large quantity of K29 and K33 polyUb chains in vitro. These allowed me to obtain milligram amounts of K29 and K33 chains, which were instrumental in characterising these chain types biochemically and biophysically. I identified the first NZF domain (NZF1) of TRABID as a UBD that selectively binds to K29 and K33 chains. To understand the molecular basis for TRABID NZF1 specific binding, I determined the crystal structure of K29 diUb in complex with TRABID NZF1. I found that TRABID NZF1 binds to the hydrophobic patch only on the distal Ub. Binding to K29 polyUb is achieved by additional interactions of the NZF with the unique surface on the proximal Ub moiety and explains the linkage-selective binding of TRABID NZF1 to K29 and K33 chains. Furthermore, I established methods to isolate K29 chains from cells using TRABID NZF1. I discovered that K29 chains may be present in heterotypic chains, containing other linkage types such as K48. I was then interested in identifying other small UBDs that can selectively bind to other polyUb chains. During this endeavour, I discovered an uncharacterised protein FAM63A containing a tandem MIU (motif interacting with Ub) that selectively binds to K48 chains. I discovered that the linkage-selective binding is mediated by the second MIU (MIU2) motif in FAM63A. The crystal structure of tMIU in complex with K48-linked polyUb chains reveals the mechanism of linkage-selective binding. FAM63A MIU2 contains three distinct surfaces that bind to polyUb in a conformation that only K48-linked triUb can accommodate. Our laboratory recently discovered that in addition to tMIU, FAM63A also contains a catalytic domain of deubiquitylating enzymes (DUBs) that is highly specific in cleaving K48 chains. DUBs regulates Ub signalling by removing Ub from the modified proteins. In the last chapter of my thesis, I characterised the DUB activity of FAM63A beyond its selectivity in cleaving K48 chains. I discovered that FAM63A acts as a chain-trimming enzyme that cleaves polyUb chains from the distal end. The work described in this thesis covers three key areas of polyUb signalling: assembly, recognition and disassembly. In spite of this, the overarching theme of my work is to understand how UBDs selectively recognise polyUb chains. In addition to the mechanistic insights, the linkage-selective UBDs characterised in this study can be further exploited as tools to delineate the functional cellular roles of different polyUb signals.