The PP2A phosphatase counteracts the function of the 9-1-1 axis in checkpoint activation

DNA damage elicits a checkpoint response depending on the Mec1/ATR kinase, which detects the presence of single-stranded DNA and activates the effector kinase Rad53/CHK2. In Saccharomyces cerevisiae, one of the signaling circuits leading to Rad53 activation involves the evolutionarily conserved 9-1-1 complex, which acts as a platform for the binding of Dpb11 and Rad9 (referred to as the 9-1-1 axis) to generate a protein complex that allows Mec1 activation. By examining the effects of both loss-of-function and hypermorphic mutations, here, we show that the Cdc55 and Tpd3 subunits of the PP2A phosphatase counteract activation of the 9-1-1 axis. The lack of this inhibitory function results in DNA-damage sensitivity, sustained checkpoint-mediated cell-cycle arrest, and impaired resection of DNA double-strand breaks. This PP2A anti-checkpoint role depends on the capacity of Cdc55 to interact with Ddc1 and to counteract Ddc1-Dpb11 complex formation by preventing Dpb11 recognition of Ddc1 phosphorylated on Thr602.

Two checkpoint complexes are independently recruited to sites of DNA damage in vivo

The Ddc1/Rad17/Mec3 complex and Rad24 are DNA damage checkpoint components with limited homology to replication factors PCNA and RF-C, respectively, suggesting that these factors promote checkpoint activation by "sensing" DNA damage directly. Mec1 kinase, however, phosphorylates the checkpoint protein Ddc2 in response to damage in the absence of all other known checkpoint proteins, suggesting instead that Mec1 and/or Ddc2 may act as the initial sensors of DNA damage. In this paper, we show that Ddc1 or Ddc2 fused to GFP localizes to a single subnuclear focus following an endonucleolytic break. Other forms of damage result in a greater number of Ddc1-GFP or Ddc2-GFP foci, in correlation with the number of damage sites generated, indicating that Ddc1 and Ddc2 are both recruited to sites of DNA damage. Interestingly, Ddc2 localization is severely abrogated in mec1 cells but requires no other known checkpoint genes, whereas Ddc1 localization requires Rad17, Mec3, and Rad24, but not Mec1. Therefore, Ddc1 and Ddc2 recognize DNA damage by independent mechanisms. These data support a model in which assembly of multiple checkpoint complexes at DNA damage sites stimulates checkpoint activation. Further, we show that although Ddc1 remains strongly localized following checkpoint adaptation, many nuclei contain only dim foci of Ddc2-GFP, suggesting that Ddc2 localization may be down-regulated during resumption of cell division. Lastly, visualization of checkpoint proteins localized to damage sites serves as a useful tool for analysis of DNA damage in living cells.