TMZ creates DNA adducts that lead to distinct and temporally separate forms of DNA damage. By following the formation of this damage after TMZ exposure and comparing its appearance with activation of the DNA damage response, we here show that the Chk1 phosphorylation previously linked only to TMZ-induced MMR-dependent DNA damage in TMZ-sensitive cells in fact occurs in all GBM cells examined long before the activation of Chk2, creation of MMR-dependent TMZ-induced DNA damage, and activation of the G2 checkpoint. As such these results re-define the TMZ-induced DNA damage response and show that Chk1 is likely a suboptimal biomarker of TMZ-induced drug action.
In the present studies, Chk1 was phosphorylated/activated at early time points (<12 hrs post TMZ exposure) in multiple GBM cell lines, prior to DNA DSB formation and regardless of MGMT or MMR status. These results stand in contrast to those of previous studies which, because they focused on later time points or used super-physiologic doses of methylating agents, reached very different conclusions. In studies in which cells were exposed to physiologic, near IC50 concentrations of temozolomide or the related SN1 methylating agent MNNG, and in which effects were monitored 48 hrs after drug exposure, Chk1 phosphorylation was associated with events linked to MMR-dependent processing of drug-induced O6MG lesions and drug-induced G2 arrest 
. In contrast, in studies in which concentrations of MNNG some 150-fold greater than the IC50 were used, Chk1 phosphorylation was shown to be independent of cellular MMR capacity but was associated with the creation of DNA double strand breaks, H2AX activation, and cell cycle arrest 
. In either instance, Chk1 phosphorylation was thought to be closely associated with DNA DSB (generated either by futile MMR of O6MG-thymine mispairs or by base excision repair of high-density N-methylations), cell cycle arrest, and drug sensitivity. The results of the present study, which examined the effects of clinically achievable, IC50 concentrations of TMZ at time points prior to the formation of DNA DSB, clearly show that Chk1 phosphorylation is independent of both MMR and DNA DSB, can be separated from drug-induced cytotoxicity, and rather is driven by single-stranded forms of DNA damage resulting from non-O6MG lesions. As such these studies more clearly define the early events that occur following exposure of GBM cells to clinically achievable doses of a relevant therapeutic methylating agent.
Although our results clearly show that early Chk1 phosphorylation is a distinct entity, unrelated to MMR processing of O6MG and divorced from activation of the G2 checkpoint, the lesions that induced Chk1 activation at early time points following TMZ exposure are not well defined. In addition to O6MG, TMZ induces N7G and 3 meA lesions. These lesions do not mispair and lead to MMR-dependent DNA DSB and cytotoxicity 
. 3 meA and N7G also do not lead to artifactual DNA strand breakage under the alkaline conditions used in the Comet assay, 3 meA being stable while N7G is converted to a ring-opened FAPy form that is also highly stable under alkaline conditions 
. N7G and 3 meA do, however, both give rise to apurinic sites and DNA single-strand breaks as consequences of their repair by the base excision repair system, and are both processed in a time frame (2–12 hours) that parallels the appearance/disappearance of ALD and Chk1 activation 
. The density of these lesions following IC50 exposures to TMZ was not sufficient to generate DNA DSB as neither H2AX foci, DNA damage detectable under neutral pH conditions, nor cell cycle arrest (not shown) were apparent in any cell type at the early time points that corresponded with early Chk1 phosphorylation. DNA single-strand breaks generated by 3 meA or N7G can, however, activate ATR as noted in the present study, and can lead to phosphorylation of Chk1 
although if this occurs in TMZ-treated cells, it is unclear why a more immediate cell cycle arrest does not follow. One possible explanation is that full Chk1 activity requires phosphorylation of ser 345 as well as ser 317 
. Although TMZ-induced Chk1 ser345 phosphorylation was shown in this study to occur in the absence of a functional MMR system, other studies in cells incubated with a similar methylating agent (MNNG) showed that Chk1 ser 317 phosphorylation occurred only in MMR-proficient cells 
. It may therefore be that the early TMZ-induced phosphorylation of Chk1 noted in the present study can occur in the absence of an activated MMR system, but is limited to ser 345, and while sufficient to lead to phosphorylation of cdc25C, is insufficient to initiate cell cycle arrest. Nonetheless, the lesions that induce Chk1 phosphorylation at early time points following TMZ exposure may be either AP sites generated by spontaneous depurination of N7G (and which appear as ALD following cleavage under alkaline assay conditions), frank DNA single-strand breaks generated following apurinic endonuclease-mediated cleavage of 3 meA- or N7G-associated AP sites, or a combination of the two. While all of these lesions are associated with the creation of single strand DNA and single-strand breaks that can cause Chk1 activation, the present study is the first to show that ser345 pChk1 activation is related to these non-O6MG TMZ-induced lesions and the subsequent single-stranded DNA damage they cause, is not unique to TMZ-sensitive cells, and as such is not likely a useful biomarker of response.
Although non-O6MG lesions appear to initiate the DNA damage response following TMZ exposure, unrepaired O6MG lesions appear to be responsible for the persistent Chk1 phosphorylation. In support of this idea, pChk1 levels returned to control values within 12 hours of TMZ exposure in base excision repair/MGMT-proficient cells, but remained elevated for at least 24 hours in those cells lacking MGMT. While it is possible that O6MG lesions were acted on by the same base excision repair system that repairs N7G and 3 meA lesions and activates Chk1, O6MG does not undergo spontaneous or alkali-induced depurination, and is a poor substrate for apurinic endonuclease or MPG-mediated removal 
. It rather appears that these O6MG lesions are processed by an alternative mechanism into at least alkali-cleavable AP sites and perhaps into unresolved DNA single-strand break intermediates. Furthermore, this system appears to be relatively efficient at converting unrepaired O6MG into ALD but relatively inefficient at fully resolving the lesions as the associated ALD persists at least 24 hours after TMZ exposure created the O6MG lesions. Thymine or guanine-thymine specific thymine glycosylases may act on unrepaired O6MG lesions, perhaps slowly or unsuccessfully, leading to prolonged Chk1 activation in MGMT-deficient cells 
. Conversely, the late Chk1 activation may simply be a result of incomplete processing of O6MG-induced mismatches by the MMR system. These results suggest that while Chk1 is initially activated in response to N7G/3 meA lesions, mechanisms that recognize and process unrepaired and unmispaired O6MG can prolong Chk1 activation.
The present studies, in addition to more clearly defining the DNA damage response to TMZ and the role of Chk1 in the process, also have implications for the use of TMZ and the monitoring of its action. Because Chk1 phosphorylation/activation has previously been associated with TMZ-induced DNA DSB damage and cell cycle arrest, both of which are precursors to TMZ-induced cell death, Chk1 activation has been suggested to be a biomarker of TMZ activity 
. The results of the present study, however, show that while ser345 pChk1 may be a marker of TMZ delivery to the cell and of TMZ-induced DNA damage, it is not per se an indicator of therapeutic response. Nonetheless the early Chk1 phosphorylation following physiologic exposures to TMZ is of biologic consequence as the monitoring of the timing of Chk1 activation could potentially provide information as to the functionality of repair systems in the cell, with early and transient ser345 Chk1 phosphorylation being a marker of functional base excision repair, and prolonged ser345 Chk1 phosphorylation being a marker of lack of MGMT. Additionally, the differences in Chk1 phosphorylation noted between the TMZ-sensitive and TMZ-resistant cells used in this study have recently been shown to be related to changes in pyruvate metabolism detectable by non-invasive magnetic spectroscopy imaging 
. In this way knowledge of the sequence of events that occur following exposure to therapeutically relevant concentrations of TMZ may not only improve our understanding of the links between the DNA damage response and the control of cellular metabolism, but also contribute to the development of methods that allow real-time monitoring of TMZ response, and improved care for individuals with brain tumors.