Chk1 is a key effector of checkpoint responses to a wide range of genotoxic stresses in vertebrate cells, including the G2/ M and S/ M arrests which delay the onset of mitosis whilst DNA is damaged or replication incomplete (Bartek and Lukas, 2003
) . In many situations these responses are triggered in a rapid and acute fashion, suggesting that Chk1 function must be subject to stringent regulation. DNA damage or DNA synthesis inhibition triggers ATM/ ATR-mediated phosphorylation of Chk1 on two SQ motifs, S317 and S345, within the C-terminal regulatory domain, however it remains unclear exactly how these modifications affect Chk1 biochemical and biological function. Measurements of Chk1 kinase activity after exposure to genotoxic stress have produced conflicting results; some studies have documented modest increases during checkpoint activation (Feijoo et al., 2001
; Zhao and Piwnica-Worms, 2001
), whereas others have failed to detect any change (Kaneko et al., 1999
). An alternative view holds that subcellular localisation play a key role in Chk1 regulation. In its most extreme form this hypothesis suggests that Chk1 is constitutively active and that the role of regulatory ATM/ ATR phosphorylation is simply to dissociate the protein from chromatin and enable it to migrate to other subcellular locations such as the cytoplasm or centrosome (Kramer et al., 2004
; Shimada et al., 2008
; Smits et al., 2006
Here, we show that DNA damage and replication arrest do stimulate endogenous Chk1 kinase activity, and that activation closely parallels S317 and S345 phosphorylation. We also found that a previously unrecognised SQ phosphorylation site, S366, is modified after DNA damage and replication arrest, and that the extent and duration of its modification is similar to the better-characterised S345 and S317 sites.
Interestingly, the kinetics of Chk1 regulatory phosphorylation and kinase activation were surprisingly transient in comparison to the duration of downstream checkpoint responses. The G2/ M and S/ M checkpoint delays elicited by DNA damage or replication arrest are sustained for at least 8 hours in DT40 cells under these conditions (Zachos et al., 2003
; Zachos et al., 2005
), however in each case Chk1 activation and SQ site phosphorylation peaked within 1 hour and then declined substantially by 3 hours. Presumably therefore either a brief pulse of Chk1 activity can rapidly set in motion biochemical events which lead to prolonged mitotic delay, or alternatively, once established, the checkpoints can be maintained by much lower levels of Chk1 activity.
Additional complexity was revealed when we investigated the cell cycle dependence of Chk1 activation. Even in unperturbed cell cultures it was apparent that Chk1 activity varied in a cell cycle dependent fashion, with much higher activity in fractions enriched for mid to late-S/ G2 phase cells and much lower levels in fractions containing predominantly cells in G1 and early S. Remarkably, a very similar pattern of variation was observed after either irradiation or aphidicolin treatment when the overall levels of Chk1 kinase activity were substantially amplified, and in all three situations the extent of S317, S345, and S366 regulatory phosphorylation closely paralleled kinase activation.
Irradiation will cause DNA damage in every exposed cell whereas aphidicolin should only stall replication forks in cells actively engaged in DNA synthesis, yet both induce very similar cell cycle phase-specific increases in Chk1 activity. Why should this be? We did not observe major cell cycle fluctuations in the level of Claspin, an adaptor protein required for ATR-mediated Chk1 activation (Mailand et al., 2006
; Peschiaroli et al., 2006
), in our elutriation experiments (data not shown). One possible explanation might be that DNA damage activates the ATR-Chk1 pathway only in S and G2 cells, as recently proposed by Jazayeri and colleagues (Jazayeri et al., 2006
). Alternatively, recent evidence that Cdk activity, which increases with cell cycle progression, is required for efficient checkpoint signalling downstream of damage recognition could also be consistent with our observations (Bonilla et al., 2008
). Further work will be required to clearly define the processes responsible for this pattern of cell cycle phase-specific Chk1 activation.
We also evaluated the consequences of substituting individual SQ residues with non-phosphorylatable alanine residues or, in the case of S317 and S345, with phospho-mimicking aspartic acid residues for G2/ M and S/ M checkpoint proficiency (Zachos et al., 2003
). Strikingly, whereas substitution of S345 with either alanine or aspartic acid rendered Chk1 completely incapable of restoring either checkpoint response, alanine substitution mutations of S317 and S366 had more subtle and complex effects. Each of these mutants was able to partially restore the G2/ M and S/ M checkpoint arrests triggered by DNA damage and replication arrest. Notably, the S317A and S317D mutants were fully able to suppress the onset of premature mitosis in cells with unreplicated DNA (). We believe that these phenotypes are most likely a consequence of altered SQ site phosphorylation, although we recognise that such substitution mutations might also affect protein structure directly.
To gain further insight into why mutations of individual SQ phosphorylation sites within the Chk1 SCD differentially affect checkpoint proficiency, we determined how the kinase activity of each mutant responded to genotoxic stress. Whereas the biologically inactive Chk1 S345A was completely refractory to induction in response to both DNA damage and replication arrest, the partially active S317A and S366A mutants remained weakly inducible. Interestingly, much lower levels of S345 phosphorylation occurred within the S317A and S366 mutant proteins, which could underlie the partial checkpoint phenotypes.
Our findings differ significantly from a previous study (Niida et al., 2007
), which evaluated the consequences of substituting alanine or aspartic acid for S317 and S345 in endogenous Chk1 in mouse embryonic stem cells (ES cells). In this study both residues were found to be essential for the G2/ M and S/ M checkpoints yet no differences in the kinase activity of the mutant proteins was detected. S317 was also reported to be essential for DNA damage and replication checkpoint proficiency in human cells, although kinase activity was not investigated (Wilsker et al., 2008
). The basis for these differences are currently unclear; cell type-specific factors may be involved, particularly since Chk1 is essential in human and mouse cells but not in DT40. Further work will be required to clarify this issue.
Finally, we sought to gain insight into the mechanism through which SQ phosphorylation stimulated Chk1 kinase activity. In an attempt to distinguish between effects on protein conformation and altered association with potential regulatory factors, we investigated the effect of washing immunoprecipitates of soluble Chk1 formed under standard conditions with a more stringent buffer (RIPA) prior to performing a kinase assay. Remarkably, this simple manipulation greatly increased the specific activity of Chk1 obtained from unstimulated cells, which bore little or no SQ site phosphorylation, but had relatively less effect on the more active, SQ-phosphorylated Chk1 obtained from aphidicolin-stimulated cells. We considered the possibility that this effect might reflect abrogation of the intramolecular inhibitory interaction which occurs between the C-terminal regulatory region and the kinase domain (Chen et al., 2000
; Oe et al., 2001
), however this interaction was not dissociated by RIPA washing. Remarkably, Chk1 S345A also became highly active after RIPA washing even though it was wholly refractory to activation under native non-stringent conditions. We conclude that S345 phosphorylation is not essential for high levels of Chk1 kinase activity per se
, even though it is required for biological function in vivo
and for stimulation of kinase in vitro
when the protein is purified under native conditions. These observations suggest that the role of S345 phosphorylation is to relieve repression of latent Chk1 catalytic activity under conditions of checkpoint activation, which might be achieved via altered configuration or association with as yet unknown trans-regulatory molecules. In future it will be important to elucidate the molecular mechanism that underlies this regulation.