Regulation Mechanisms of the Ubiquitin C-terminal Hydrolases UCH-L5 and BAP1

The covalent attachment of the small protein ubiquitin to target proteins, ubiquitination, is a major type of modification in cells. This modification can have a vast array of outcomes that allow cells to tightly control processes. Ubiquitination is reversed by deubiquitinating enzymes (DUBs) allowing these enzymes to regulate all processes affected by ubiquitination. Because of this powerful property, they themselves need to be kept in check by a variety of regulatory mechanisms that exist in cells.

In chapter 2 we describe the currently known mechanisms of DUB regulation. These mechanisms are diverse and can be exerted from outside the DUB, by external factors or post-translational modifications, or from within DUBs by domains specialized in recruiting targets or activating the catalysis. The different types of regulation do not stand on their own though, but instead cooperate. This is exemplified by the fact that many DUBs are subject to multiple types of regulation simultaneously. A particular type of DUBs that are regulated are ubiquitin C-terminal hydrolases (UCH) family. These enzymes are all characterized by an catalytic UCH domain responsible for the catalytic activity. The family members UCH-L5 and BAP1 also contain an additional ULD domain at their C-termini that is thought to be important for their regulation. For both proteins, the regulation is thought to be performed by proteins containing DEUBiquitinase ADaptor (DEUBAD) domains. BAP1 and UCH-L5 play important roles in basic cellular processes such as DNA repair and gene regulation and are mutated in several cancers. UCH-L5 presents a remarkable case of DUB regulation since it can be activated by the DEUBAD domain of RPN13 but inhibited by the DEUBAD domain of INO80G.

In chapter 3 we show how this regulation works. Using crystal structure and functional analyses we demonstrate that RPN13 activates UCH-L5 by increasing the affinity for substrates through correct positioning of the ULD domain and active site cross over loop. Conversely, INO80G inhibits UCH-L5 by dramatically decreasing the affinity for substrates and achieves this by blocking the ubiquitin binding site via large conformational changes.

In chapter 4 we investigate the mechanism of H2A deubiquitination by the BAP1/ASXL1 complex, a complex homologous to the UCH-L5/RPN13 complex. We show that the BAP1/ ASXL1 complex specifically deubiquitinates the Polycomb site (K119) and is inactive on the DNA damage site (K13/15). Like RPN13, the DEUBAD domain of ASXL1 activates BAP1 H2A deubiquitination by increasing the affinity for the ubiquitinated substrate. Activation of BAP1 by the DEUBAD domain of ASXL1 is however not the only requirement to deubiquitinate nucleosomal H2A. This also requires the C-terminal extension of BAP1 that binds to nucleosomes. An interesting feature of the BAP1/ASXL1 complex is that it is an asymmetric complex consisting of two molecules of BAP1 and one molecule of ASXL1. Moreover BAP1 itself forms homo-oligomers.

In chapter 5 we characterize the oligomeric state of BAP1 and the BAP1/ASXL1DEU complex and examine the consequence of different oligomerization states on the enzymatic activity. We find that the C-terminal two helices of BAP1 promote higher order oligomerization and that ASXL1DEU prevents this by promoting a 2:1 BAP1/ASXL1DEU complex. We further show that in this complex both BAP1 molecule can be activated to deubiquitinate H2A despite the presence of only one ASXL1DEU molecule. We end with a general discussion presented in chapter 6 where we place our findings in a broader context.