SUMO-Targeted Ubiquitin Ligases and Their Functions in Maintaining Genome Stability

Abstract

Small ubiquitin-like modifier (SUMO)-targeted E3 ubiquitin ligases (STUbLs) are specialized enzymes that recognize SUMOylated proteins and attach ubiquitin to them. They therefore connect the cellular SUMOylation and ubiquitination circuits. STUbLs participate in diverse molecular processes that span cell cycle regulated events, including DNA repair, replication, mitosis, and transcription. They operate during unperturbed conditions and in response to challenges, such as genotoxic stress. These E3 ubiquitin ligases modify their target substrates by catalyzing ubiquitin chains that form different linkages, resulting in proteolytic or non-proteolytic outcomes. Often, STUbLs function in compartmentalized environments, such as the nuclear envelope or kinetochore, and actively aid in nuclear relocalization of damaged DNA and stalled replication forks to promote DNA repair or fork restart. Furthermore, STUbLs reside in the same vicinity as SUMO proteases and deubiquitinases (DUBs), providing spatiotemporal control of their targets. In this review, we focus on the molecular mechanisms by which STUbLs help to maintain genome stability across different species.

Keywords: RNF111; RNF4; STUbL; SUMO; Slx5/Slx8; genome stability; ubiquitin.

Figure 1

Overview of STUbLs. (a) The domain structures of STUbLs from Saccharomyces cerevisiae, Saccharomyces pombe, Drosophila melanogaster, and Homo sapiens are shown. Domain information was obtained from UniProt (https://www.uniprot.org/ accessed on 15 March 2021) or NCBI Protein (https://www.ncbi.nlm.nih.gov/protein/ accessed on 15 March 2021) databases. For STUbLs for which SIM information was unavailable, the predicted SIM sequences (SIM*) were obtained from GPS-SUMO [41,42]. (b) Sequence alignment of the RING domains of STUbLs. Amino acids are highlighted in blue based on their similarity. The seven conserved cysteine residues and one histidine residue in a canonical RING sequence are highlighted in orange in the consensus sequence. Sequence alignment was performed using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/ accessed on 15 March 2021) and Jalview [43]. (c) The crystal structure of RNF4 RING (pink)-UbcH5A (green)~ubiquitin (gold) (Protein Data Bank (PDB) code 4AP4 [44]). (d) The crystal structure of RNF111 RING (Ark2C, pink)-UbcH5B (green)~donor ubiquitin (gold) (PDB code 5D0K [45]). A second ubiquitin (Ub*, orange) that directly binds to the back side of RNF111 RING is shown. (e) The crystal structure of Ubc13 (green)~ubiquitin (gold)-RNF4 RING (pink)-Mms2 (blue) (PDB code 5AIT [46]). The priming ubiquitin (Ub*, orange) is shown. The structures were prepared with the Chimera program (http://www.cgl.ucsf.edu/chimera/ accessed on 15 March 2021). Abbreviations: SIM, SUMO-interacting motif; RING, really interesting new gene E3 ligase domain; aa, amino acids; UBZ4, ubiquitin-binding zinc finger type 4 domain; ARM, arginine-rich motif; RKK, arginine–lysine–lysine motif; SOB, SUMO one binding motif; M domain, middle domain.

Figure 2

Relocalization of damaged DNA to the nuclear periphery by STUbLs. (a) Relocalization of irreparable DSBs to the nuclear periphery in S. cerevisiae. Irreparable DSBs that are mono-SUMOylated are relocated to Mps3, whereas DSBs that are poly-SUMOylated are relocated to the NPC in an Slx5/Slx8-dependent manner. Once DSBs are at the NPC, Slx5/Slx8 promotes the degradation of proteins blocking late HR steps, enabling the loading of Rad51 and subsequent repair [219,220,221]. (b) Relocalization of eroded telomeres to the nuclear periphery in S. cerevisiae. In telomerase-positive cells (left), telomeres are associated with NE via the interaction between telomerase and Mps3. In the absence of telomerase, Slx5/Slx8 recognizes SUMOylated proteins that bind to critically short telomeres and relocates them to NPC, at which type II recombination occurs to extend telomere lengths [222,223,224]. (c) Relocalization of collapsed forks to the NPCs in S. cerevisiae. Replication forks are prone to collapse at the expanded CAG repeats. Collapsed forks are processed and bound by RPA, Rad52, and Rad59, which are mono-SUMOylated by Smc5/Smc6 and Mms21. Collapsed forks relocate exclusively to the NPCs in an STUbL-dependent manner. Slx5/Slx8 also facilitates the degradation of Rad52 at the NPCs, allowing the association of Rad51 and subsequent fork restart, potentially with the assistance of Srs2 helicase [97,225]. (d) Relocalization of arrested forks to the nuclear periphery in S. pombe. Arrested forks caused by replication fork barriers are poly-SUMOylated by Pli1 SUMO E3 ligase. Rad51 loading and activity are required before relocalization by Rfp1/Rfp2/Slx8. At the NPC, both Ulp1 SUMO-specific peptidase and STUbL-proteasome can remove poly-SUMO chains, which are inhibitory to fork restart [172]. (e) Key to proteins and compartments shown in (a–d). Figures were created with BioRender.com. Abbreviations: STUbL, SUMO-targeted E3 ubiquitin ligase; Mps3, mono polar spindle 3; DSB, DNA double-strand break; NPC, nuclear pore complex; HR, homologous recombination; NE, nuclear envelope; RPA, replication protein A; Smc5/Smc6, structural maintenance of chromosome protein 5/6; Mms21, methyl methanesulfonate sensitivity 21; CMG, Cdc45-Mcm2-7-GINS; PCNA, proliferating cell nuclear antigen; Siz1, SAP and mIZ-finger domain 1; Srs2, suppressor of Rad6 2.