Orderly progression through S-phase requires dynamic ubiquitylation and deubiquitylation of PCNA.

Living cells tolerate DNA damage during S-phase to prevent the risk of irreversible DNA replication fork collapse1. DNA damage tolerance is based on translesion synthesis (TLS), which involves a dual mechanism, primarily mediated by specialized low fidelity DNA polymerases called TLS-polymerases. These TLS polymerases can be mutagenic because they induce an error-prone process that causes mutations. The second mechanism depends on template switching, the error-free component of the bypass that involves sister-strand pairing between nascent chains within the same replication fork. It is well established that both mechanisms efficiently prevent replisome stalling at damaged sites. Eukaryotic cells regulate the choice of alternative pathways/mechanisms to bypass DNA lesions during S-phase through the ubiquitylation of proliferating-cell nuclear antigen (PCNA), a processivity factor for DNA synthesis2,3. The monoubiquitylation of PCNA (ubPCNA) at Lys164 enhances the affinity of error-prone TLS DNA polymerases, and further polyubiquitylation of Lys164-monoubiquitylated PCNA promotes template switching2,3,4,5,6. The evolutionarily conserved Lys164-ubiquitylation of PCNA and its crucial role in DNA damage tolerance are well understood7,8,9. However, little is known regarding the significance of PCNA deubiquitylation in eukaryotes. Work from the past decade has identified mammalian Usp1, human Usp10, and budding yeast Ubp10 as major deubiquitylating enzymes (DUBs) for PCNA1,10,11,12. Usp1 has been identified as a DUB that deubiquitylates mono-ubPCNA and mono-ubFANCD2 in human cells10. Upon UV light-induced DNA damage, Usp1 undergoes autocleavage, and PCNA therefore becomes ubiquitylated, suggesting that Usp1 continuously deubiquitylates PCNA in the absence of DNA damage. However, ubPCNA accumulation does not correlate with Usp1 (auto-induced) proteolysis when cells are exposed to other genotoxic agents, such as MMS and mitomycin C13,14, or the DNA replication blocking agent HU15, suggesting either that another human PCNA DUB exists or that Usp1 activity is regulated in a different manner in response to other DNA damaging agents. Therefore, the relevance of USP1 in reverting PCNA ubiquitylation when confronted with different DNA-damaging agents remains unclear. This is an important issue, particularly because PCNA ubiquitylation is required for mammalian cell survival not only after UV irradiation but also upon exposure to HU and MMS16,17. However, consistent with a role as a PCNA DUB, the depletion of chicken USP1 in DT40 cells or in murine Usp1−/− MEFs increases PCNA (and also FANCD2) mono-ubiquitylation14,18. More recently, it has been shown that upon UV-mediated DNA damage, HeLa cells rely on USP10 to deubiquitylate ISGylated-PCNA12. We have previously shown that the ubiquitin protease Ubp10 deubiquitylates K164 mono-and di-ubiquitylated PCNA during S phase in the yeast Saccharomyces cerevisiae11. Furthermore, we demonstrated that Ubp10 forms a complex with PCNA in vivo, as expected for an enzyme-substrate complex. Additionally, in agreement with a direct role as a PCNA DUB, we found that only catalytically active Ubp10 reverts PCNA ubiquitylation. However, despite the identification of Usp1, Usp10, and Ubp10 as PCNA DUBs, little is known regarding the deubiquitylation of ubPCNA in any other model organisms. In the fission yeast S. pombe, a key model organism in understanding the cell division cycle, the proteases that deubiquitylate ubPCNA remain unknown. Potential candidates in fission yeast are 20 genes that encode different deubiquitylating enzymes or DUBs including 14 ubiquitin-specific proteases (USPs), 2 ubiquitin C-terminal hydrolases (UCHs), 2 ovarian tumour proteases (OTUs) and 2 JAB1/MPN/Mov34 metalloenzymes (JAMMs)19. Here, we show that Ubp2, Ubp15, and Ubp16 ubiquitin proteases, likely with the help of Ubp12, revert PCNA ubiquitylation in the fission yeast S. pombe during S-phase. All the DUBs involved in PCNA deubiquitylation in this model system belong to the same subfamily of ubiquitin-specific proteases (USPs/UBPs). We found that each of these ubiquitin proteases is dedicated to the deubiquitylation of a specific subnuclear fraction or a particular type of ubiquitylated PCNA. Our data suggest that the dynamic ubiquitylation and deubiquitylation of PCNA occurs during S-phase and ensures proper DNA replication progression in S. pombe. Consequently, we propose that excessive DNA replication bypass interferes with the normal progression of DNA replication forks during S-phase.