The p97 cofactors UBXN7 and UBXN8 interact with and modulate the function of cullin-RING complexes and Fanconi anaemia proteins FANCD2/FANCI, respectively

• Main Author: Bandau, Susanne
• Published: University of Dundee 2014
• Subjects: 571.6
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.613617

The ATPase p97 plays a role in diverse cellular activities such as cell cycle progression, membrane fusion, DNA repair and ubiquitin-dependent protein degradation (including endoplasmic reticulum-associated protein degradation, ERAD). The functional diversity of p97 is achieved by its association with a large number of cofactors including the 13 human UBX-domain containing proteins. My PhD projects focused on two of the UBX-domain proteins, UBXN7 and UBXN8. Among human UBA-UBX proteins, UBXN7 is the most proficient in interacting with CRL (cullin-RING E3 ligase) subunits, in particular CUL2, and it was assumed that these interactions were indirect, mediated by ubiquitylated substrates. However, we show that UBXN7 interaction with CUL2 is independent of ubiquitin- and substrate binding. Instead, it involves the direct docking of the ubiquitin interaction motif (UIM) in UBXN7 onto the neddylated cullin. Furthermore, we found that UBXN7 overexpression keeps the E3 ligase CUL2 (a member of the CRL2 complex) in its neddylated form and causes the accumulation of non-ubiquitylated HIF1a (CRL2 substrate). Both effects are strictly UIM-dependent and occur only when UBXN7 contains an intact UIM. We also show that HIF1a carrying long ubiquitin-chains can interact with an alternative ubiquitin-binding protein, which is independent from p97’s segregase activity. We therefore propose that UBXN7 negatively regulates the ubiquitin-ligase activity of CRL2 by sequestering CUL2 in its neddylated form and that this might prevent recruitment of ubiquitin-receptors other than p97 to HIF1a. The mass spectrometry analysis of Flag-UBXN8 immunoprecipitates identified several DNA damage-related proteins, including the Fanconi anaemia proteins FANCD2 and FANCI. I could show that homodimeric UBXN8 interacts directly with non-ubiquitylated FANCD2 and FANCI. The direct binding of UBXN8 to the non-ubiquitylated FA proteins supports the notion that UBX-only proteins interact with substrates in an ubiquitin-independent manner. Furthermore, FANCD2 and FANCI are released from UBXN8 upon DNA damage, that in case of FANCI requires p97 binding to UBXN8. My data indicate that UBXN8 acts as a negative regulator in the DNA damage response, because UBXN8 silenced cells show increased resistance to ICL-inducing agents. This phenotype was supported by my finding that UBXN8 silenced cells have increased levels of FANCD2 and FANCI mono-ubiquitylation as well as increased dimer formation and FANCD2 foci formation in the presence and absence of DNA damage. UBXN8 overexpression had the opposite effect on FANCD2 and FANCI modification and dimerization. I therefore propose that UBXN8 has an inhibitory effect on FANCD2 and FANCI that may prevent their ectopic activation in the absence of DNA damage.