In this study we investigated the interplay of Mre11 and Sae2, both required for early steps in the processing of DNA ends prior to DNA repair, and the DNA end binding factor yKu70. The data herein demonstrate a critical role for Mre11 nuclease activity in the repair of replication-dependent DSBs and highlight an NHEJ-independent role for yKu70 in antagonizing DSB repair in S phase and G2.
We show that Mre11 nuclease activity and Sae2 are largely dispensable for cell viability in yKu70-deficient cells while the requirement for Exo1 remains. On this basis, we propose that Ku binds to and excludes Exo1 from processing DNA structures that are formed in response to CPT-induced genotoxic stress. Given that Mre11 and Sae2 act upstream of Exo1 in DSB end resection (42
), we propose that initial endonucleolytic cleavage events are catalyzed by the Mre11 nuclease and Sae2 and serve to inhibit Ku binding, thereby potentiating Exo1-mediated resection and promoting HDR-dependent repair.
The model proposed above posits that Ku deficiency facilitates the resolution of toxic DNA adducts stabilized by Mre11 nuclease and Sae2 deficiency. Implicitly, this hypothesis requires that the DNA structures to which Ku binds are normally processed by the Mre11 nuclease and Sae2 and are thus more abundant in Mre11 nuclease- and Sae2-deficient cells.
Several lines of evidence support the view that these toxic DNA structures are DSBs. First, Ku exhibits a strong preference for binding to dsDNA ends over ssDNA or circular DNA (12
). Second, CPT-induced cell cycle arrest, γ-H2AX signal, heteroallelic recombination, and chromosome breakage were all enhanced in mre11
cells and reduced to essentially wild-type levels in mre11
double mutant cells ( and ). These endpoints are each sensitive indices of DSB abundance and support the interpretation that yKu70 antagonizes HDR-mediated DSB repair in S phase and G2
in the absence of Mre11 nuclease activity. The function of Ku in this regard is not simply to promote NHEJ over HDR as inactivating NHEJ via a DNA ligase IV deficiency had no effect on mre11
CPT sensitivity (). These and other data support the view that DSB end binding by the Ku heterodimer influences mitotic DNA repair by limiting access of Exo1 for resection.
We propose that a role of the initial cleavage step at mitotic DSB ends by the Mre11 nuclease and Sae2 is to counterbalance this NHEJ-independent Ku function though the precise interaction between Mre11 and Sae2 during end processing remains unclear. In light of their role in this mechanism, the clastogen sensitivities observed in Mre11 nuclease and Sae2 deficiency are likely attributable to defects in HDR-mediated DNA repair. In this model, the initial Mre11 nuclease/Sae2-dependent cleavage step would inhibit or dismantle Ku binding at DSB ends and thereby promote resection by Exo1. Consistent with this view, Ku has a significantly higher affinity for blunt dsDNA ends over short ssDNA overhangs (). Accordingly, Ku deficiency effectively bypasses the need for the initial incision step and permits Exo1 to effect resection in mre11
mutants. In further support of this idea, Ku deficiency was recently shown to alleviate the resection defect at an HO-induced DSB in rad50Δ
). On the other hand, mre11
mutants, unlike mre11Δ
cells, did not exhibit higher detectable levels of Ku bound at an HO-induced DSB when measured by chromatin immunoprecipitation (ChIP) (60
). This discrepancy may be attributable to differential processing requirements of HO- and CPT-induced DNA ends and additionally due to the S-phase specificity of CPT-induced lesions.
We consider two nonexclusive possibilities for the molecular basis of increased DSBs in CPT-treated mre11
mutants: defects in the resolution of topoisomerase adducts and defects in DSB repair. Several lines of evidence support the view that both the Mre11 complex and Sae2 promote the resolution of Top1-DNA cleavage complexes. Impaired resolution would increase the stability of adducts and accordingly increase the risk of collision with the replisome. Increased levels of Top1-DNA adducts have been detected in Schizosaccharomyces pombe rad32mre11
nuclease and rad50S
), and S. cerevisiae mre11
mutants and murine Rad50S/S
cells are CPT sensitive () (11
). It is conceivable that the Ku heterodimer may bind to Top1-associated DSB ends. However, the flap endonuclease activity of Exo1 is the incorrect polarity to remove the Top1-associated ends (62
). Hence, this could not account for the suppression of CPT sensitivity by yKu70 deficiency in mre11
An alternative scenario is that the leading strand encounters the 5′ OH side of the nick created by the cleavage complex, leading to polymerase runoff and ultimately a single-ended DSB (57
). These DSB ends would likely be substrates for Mre11 and Sae2, as well as Ku binding. These DSBs would accumulate in mre11
cells in a Ku-dependent, Mus81-independent manner and are possibly the basis of the mre11
CPT toxicity (). In this scenario, the Exo1-dependent suppression of mre11
CPT sensitivity by yku70Δ
would reflect the promotion of DSB resection en route to HDR-mediated repair ().
Fig. 7. Antagonistic roles for Ku and Mre11 in S-phase DSB metabolism. We propose that the resection of DNA replication-associated DSBs is regulated negatively by the Ku heterodimer and positively by Mre11 nuclease- or Sae2-dependent activities. In this figure, (more ...)
Finally, a competitive relationship between Ku and the Mre11 complex has been suggested to provide a mechanism of DSB processing and repair pathway choice (41
). Limited resection in G1
cells promotes Ku- and DNA ligase IV-dependent NHEJ, whereas increased cyclin-dependent kinase (CDK) activity leads to increased resection and promotion of HDR (40
). However, the data presented herein clearly demonstrate a previously uncharacterized role for Ku in mediating DSB repair during DNA replication and the subsequent G2
. The S/G2
functions of Ku are independent of NHEJ and suggest that the heterodimer exerts a relatively broad regulatory influence on the metabolism of toxic DNA structures formed in the course of DNA replication.