The Fanconi Anemia pathway is required in mammals to repair DNA interstrand crosslinks. The pathway is coordinated by a single monoubiquitin signal on two proteins, FANCD2 and FANCI. The assembly and the removal of this signal are both required for repair of the DNA lesion.
The site of attachment of the ubiquitin signal is critical, and is Lysine561 in FANCD2. The signal is removed by the deubiquitinase USP1, which requires its cofactor UAF1 for activity. Recently we reported that USP1-mediated removal of ubiquitin from K561 of FANCD2 is dependent on a substrate-targeting element in the N-terminus of USP1 (Arkinson et al., 2018). This element is required for site-specific removal, as USP1 missing this element will deubiquitinate other modified lysines, but not K561. Intriguingly, USP1 is similar to both USP12 and USP46 in requiring UAF1 for full activity, while USP12 requires a second beta-propeller cofactor, WDR20 (Li et al., 2016). In the case of USP1, collaboration with UAF1 leads to a 30-fold increase in Kcat (Cohn et al., 2007). How USP1-UAF1 deubiquitinate FANCD2, and how UAF1 promotes USP1’s catalytic activity is unknown.
We have determined the structure of USP1-UAF1 alone and in complex with Ubiquitin. We find that in contrast to USP12 and USP46, only one beta-propeller module is associated with each USP1 molecule. We identify residues required to allow ubiquitin access to the active site, and residues required for USP1 activation. Our structures provide a molecular rationale for the lack of requirement for a second co-factor.
Through reconstitution of the FANCD2-Ub/FANCI-Ub complex (Chaugule et al., 2019a; Chaugule et al., 2019b), we find that FANCI-Ub regulates USP1’s activity towards FANCD2-Ub. The FANCD2-Ub/FANCI-Ub-DNA complex is resistant to deubiquitination, however, compromising ubiquitin recognition on the FANCI-site permits USP1 to deubiquitinate FANCD2-Ub. These structural insights allow us to propose a model for USP1-UAF1-mediated deubiquitination.