S-phase checkpoint prevents leading strand degradation from strand-associated nicks at stalled replication forks.

The S-phase checkpoint is involved in coupling DNA unwinding with nascent strand synthesis and is critical to maintain replication fork stability in conditions of replicative stress. However, its role in the specific regulation of leading and lagging strands at stalled forks is unclear. By conditionally depleting RNaseH2 and analyzing polymerase usage genome-wide, we examine the enzymology of DNA replication during a single S-phase in the presence of replicative stress and show that there is a differential regulation of lagging and leading strands. In checkpoint proficient cells, lagging strand replication is down-regulated through an Elg1-dependent mechanism. Nevertheless, when checkpoint function is impaired we observe a defect specifically at the leading strand, which was partially dependent on Exo1 activity. Further, our genome-wide mapping of DNA single-strand breaks reveals that strand discontinuities highly accumulate at the leading strand in HU-treated cells, whose dynamics are affected by checkpoint function and Exo1 activity. Our data reveal an unexpected role of Exo1 at the leading strand and support a model of fork stabilization through prevention of unrestrained Exo1-dependent resection of leading strand-associated nicks after fork stalling.

Exo1 phosphorylation inhibits exonuclease activity and prevents fork collapse in rad53 mutants independently of the 14-3-3 proteins.

The S phase checkpoint is crucial to maintain genome stability under conditions that threaten DNA replication. One of its critical functions is to prevent Exo1-dependent fork degradation, and Exo1 is phosphorylated in response to different genotoxic agents. Exo1 seemed to be regulated by several post-translational modifications in the presence of replicative stress, but the specific contribution of checkpoint-dependent phosphorylation to Exo1 control and fork stability is not clear. We show here that Exo1 phosphorylation is Dun1-independent and Rad53-dependent in response to DNA damage or dNTP depletion, and in both situations Exo1 is similarly phosphorylated at multiple sites. To investigate the correlation between Exo1 phosphorylation and fork stability, we have generated phospho-mimic exo1 alleles that rescue fork collapse in rad53 mutants as efficiently as exo1-nuclease dead mutants or the absence of Exo1, arguing that Rad53-dependent phosphorylation is the mayor requirement to preserve fork stability. We have also shown that this rescue is Bmh1-2 independent, arguing that the 14-3-3 proteins are dispensable for fork stabilization, at least when Exo1 is downregulated. Importantly, our results indicated that phosphorylation specifically inhibits the 5' to 3'exo-nuclease activity, suggesting that this activity of Exo1 and not the flap-endonuclease, is the enzymatic activity responsible of the collapse of stalled replication forks in checkpoint mutants.