Comparison of checkpoint responses triggered by DNA polymerase inhibition versus DNA damaging agents

Mutation Research 532 (2003) 215–226

Comparison of checkpoint responses triggered by DNA polymerase inhibition versus DNA damaging agents Jen-Sing Liu1 , Shu-Ru Kuo, Thomas Melendy∗ Departments of Microbiology, Immunology, and Biochemistry, and the Witebsky Center for Microbial Pathogenesis & Immunology, SUNY, School of Medicine & Biomedical Sciences, 138 Farber Hall, University at Buffalo, Buffalo, NY 14214-3000, USA Received 2 July 2003; received in revised form 21 August 2003; accepted 22 August 2003

Abstract To better understand the different cellular responses to replication fork pausing versus blockage, early DNA damage response markers were compared after treatment of cultured mammalian cells with agents that either inhibit DNA polymerase activity (hydroxyurea (HU) or aphidicolin) or selectively induce S-phase DNA damage responses (the DNA alkylating agents, methyl methanesulfonate (MMS) and adozelesin). These agents were compared for their relative abilities to induce phosphorylation of Chk1, H2AX, and replication protein A (RPA), and intra-nuclear focalization of ␥-H2AX and RPA. Treatment by aphidicolin and HU resulted in phosphorylation of Chk1, while HU, but not aphidicolin, induced focalization of ␥-H2AX and RPA. Surprisingly, pre-treatment with aphidicolin to stop replication fork progression, did not abrogate HU-induced ␥-H2AX and RPA focalization. This suggests that HU may act on the replication fork machinery directly, such that fork progression is not required to trigger these responses. The DNA-damaging fork-blocking agents, adozelesin and MMS, both induced phosphorylation and focalization of H2AX and RPA. Unlike adozelesin and HU, the pattern of MMS-induced RPA focalization did not match the BUdR incorporation pattern and was not blocked by aphidicolin, suggesting that MMS-induced damage is not replication fork-dependent. In support of this, MMS was the only reagent used that did not induce phosphorylation of Chk1. These results indicate that induction of DNA damage checkpoint responses due to adozelesin is both replication fork and fork progression dependent, induction by HU is replication fork dependent but progression independent, while induction by MMS is independent of both replication forks and fork progression. © 2003 Published by Elsevier B.V. Keywords: S-phase checkpoint; DNA damage; ATR; Chk1 kinase; Replication protein A (RPA); ␥-H2AX; Phosphorylation; Nuclear foci

1. Introduction When chromosomal DNA is damaged by natural causes or by chemotherapeutic agents, eukaryotic cells activate checkpoint pathways to preserve ∗ Corresponding author. Tel.: +1-716-829-3789; fax: +1-716-829-2158. E-mail addresses: [email protected] (J.-S. Liu), [email protected] (T. Melendy). 1 Co-corresponding author.

0027-5107/$ – see front matter © 2003 Published by Elsevier B.V. doi:10.1016/j.mrfmmm.2003.08.018

genome integrity. This leads to either cell cycle arrest and DNA repair, or programmed cell death. During S phase of the cell cycle, when the genetic material is duplicated by the DNA replication machinery, cells are especially vulnerable to genotoxic insult. While G1 and G2/M checkpoints arrest cell cycle progression at well-defined points through the inhibition of cyclin-dependent kinases [1,2], intra-S-phase checkpoints are known to occur at any time within S phase and involve multiple signaling pathways (see [3], in press).

216

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

Our previous studies of cellular DNA damage responses to an investigative anticancer drug, adozelesin, demonstrated that DNA damage responses triggered by adozelesin are dependent on replication fork progression [4]. Adozelesin is a DNA alkylating agent that binds the minor groove of A/T-rich duplex DNA and alkylates the N3 of adenine on the 3 end of the binding site. Our results showed that adozelesin-induced DNA adducts are not detected by much of the DNA damage response machinery, until collisions of replication forks with the DNA lesions result in blocked replication forks [4]. Since short-term treatment with adozelesin has been shown to not cause detectable DNA strand breaks [5], we conclude that these adozelesin-induced S-phase checkpoint responses are primarily triggered by blocked replication forks rather than by direct DNA damage. These results focused our attention on checkpoint mechanisms induced by DNA polymerase stalling and replication fork blockage. DNA replication arrest in S-phase cells can be caused by several mechanisms. Collision of DNA replication enzymes, such as helicases or polymerases, with DNA lesions can result in blockage of replication fork progression. Damage on DNA in the presence or absence of DNA strand breaks is also able to activate trans-acting mechanisms to slow down DNA synthesis on other replication forks [5] and prevent the firing of late origins [6–8]. Under these conditions, multiple checkpoint pathways, as well as DNA repair mechanisms, are activated [3], which makes it difficult to distinguish individual signaling pathways. 1.1. Intra-S-phase checkpoints Genetic and biochemical studies from yeast and mammalian cells have identified six S-phase checkpoint sensors: Rad1, Rad3, Rad9, Rad17, Rad26 and Hus1 [9]. Rad17 is a homolog of the largest subunit of replication factor C (RFC1) and is capable of forming an RFC-like complex with RFC2–5 [10,11]. Rad9, Rad1 and Hus1 can form a heterotrimer (9-1-1 complex) that is structurally similar to the proliferating cell nuclear antigen (PCNA) homotrimer [10–12]. Rad17–RFC2–5 and 9-1-1 complexes have been purified and their biological functions are currently under investigation [13]. These two complexes appear to be involved in the DNA damage checkpoint pathways of

cells in S phase as well as in other phases of the cell cycle. Rad3 and Rad26 are homologs of the mammalian ataxia telangiectasia mutated (ATM)–Rad3-related protein kinase (ATR) and ATR-interacting-protein (ATRIP), respectively [14]. ATR, together with the ATM kinase and DNA-dependent protein kinase (DNA-PK), belong to the family of phosphatidylinositol 3-kinases. DNA-PK is activated by the presence of double-strand DNA breaks (DSB) [15], and plays an important role in the repair of DSBs [16]. ATM is primarily responsible for the activation of cellular responses to DNA double strand breaks in all phases of the cell cycle [17,18]. Although ATR contributes to checkpoint activity induced by many types of DNA damage, one major function of ATR is to serve as a DNA replication checkpoint sensor. ATR and its downstream effector kinase, Chk1, are essential and cause embryonic lethality in mice when either is genetically deleted [19,20]. ATR is a DNA binding protein with higher affinity for UV-damaged than undamaged DNA [21]. ATR forms foci at sites of stalled replication forks in response to DNA replication arrest [22]. Furthermore, studies in Xenopus and budding yeast suggest that ATR (and its yeast homolog Mec1) may associate with replication forks constitutively to act as a DNA replication checkpoint during S phase in undamaged cells [23,24]. In response to DNA damage or DNA replication arrest, ATR is activated to phosphorylate and activate Chk1 kinase [20,25,26]. Cells with a functional defect in Chk1 (or its yeast analogs Cds1 or Rad53) are highly sensitive to replicational stress [27–29]. And it has been suggested that Chk1 is required for the maintenance of replication complex stability at stalled replication forks [27–30]. Genetic studies have also identified a number of DNA replication proteins as participants in S-phase checkpoint responses. They are the primase subunits of the DNA polymerase ␣–primase complex, DNA polymerase ε, replication protein A (RPA), and the four smaller subunits of RFC (RFC2–5) [31–36]. RFC2–5 s checkpoint activity may be associated with their requirement for Rad17 function as described above. Studies using a Xenopus cell-free assay system demonstrated that the primase activity of polymerase ␣–primase complex is required for aphidicolin-induced Chk1 phosphorylation in vitro [37]. Recent studies showed that RPA is required for

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

chromatin association of ATR and ATRIP [38–40]. The RPA checkpoint mutant protein (RPA-t11) from budding yeast is unable to load Ddc2 (the budding yeast ATRIP) onto single-stranded DNA (ssDNA) [39]. In addition, when RPA-t11 is bound to damaged DNA, the normal switching between RPA and the strand exchange protein, Rad51, does not occur, thereby preventing homologous recombination [41]. In response to DNA damage, RPA forms foci on damaged DNA within minutes (our unpublished data), and the 32 kDa subunit of RPA becomes hyper-phosphorylated [42,43]. RPA’s high affinity for ssDNA as well as damaged double-stranded DNA [44,45] suggests a surveillance function in DNA damage response for this relatively abundant nuclear protein. Histone H2AX is a histone H2A isoform that is non-essential but plays an important role in connecting DNA damage sensors and effectors [46]. H2AX is required for the DNA damage-induced focus formation of Brca1, 53BP1, NBS1 and Mre11 [47–49]. It has been shown that H2AX is phosphorylated at serine 139 by either the ATM or ATR kinase in cells with damaged DNA or under replication stress [50,51]. This phosphorylated H2AX (referred to as ␥-H2AX) is detectable within 1 min of ␥-radiation as foci on damaged DNA [51]. Induction of ␥-H2AX has been used extensively as a very early signal of DNA damage. In this study, we used several well-studied agents that are known to induce checkpoint responses either specifically or preferentially in S phase and compared their abilities to trigger various checkpoint and DNA damage responses. We have focused specifically on agents that cause replication fork pausing without inducing DNA strand breaks.

2. Materials and methods 2.1. Chemicals and antibodies Aphidicolin, 4,6-diamidino-2-phenylindole (DAPI), hydroxyurea (HU) and methyl methanesulfonate were purchased from Sigma. Adozelesin, generously supplied by Dr. Terry Beerman and Pharmacia Upjohn Co. (Kalamazoo, MI), was dissolved in dimethylacetamide (2 mg/ml) and further diluted in dimethylsulfoxide prior to its addition into culture medium.

217

Monoclonal antibodies against ␥-H2AX and human Chk1 (G-4) were purchased from Upstate Biotech and Santa Cruz Biotech, respectively. The monoclonal and affinity purified polyclonal antibodies against RPA32 were as described [4,52]. Fluorescein conjugated goat–anti-mouse and Alexa 568-conjugated goat–anti rabbit antibodies were purchased from Vector Laboratory Inc. and Molecular Probes, Inc., respectively. Bromouracil deoxyriboside (BUdR) and monoclonal antibody against BUdR were purchased from BD Pharmingen. 2.2. Cell culture and treatment Monolayer cultured HeLa cells (ATCC) were maintained in DMEM with 10% FBS. Human fibroblast cells transfected with wild-type ATR (ATR-wt) or kinase-dead mutant of ATR (ATR-kd), which dramatically reduces, but does not wholly eliminate ATR function (gifts from Dr. K. Cimprich) [53], were maintained in DMEM with 15% FBS and 400 ␮g/ml G418. ATR-wt or ATR-kd was induced by overnight incubation with 1 ␮g/ml doxycycline. The selected concentrations of aphidicolin, HU, adozelesin and MMS used in this study were based on titrations of each reagent into the cell lines used, such that the level of each agent used was sufficient to block the vast majority, if not all, BUdR incorporation after 30 min treatment (data not shown). 2.3. Indirect immunofluorescent staining Monolayer cultured cells treated with drugs as indicated were rinsed with PBS and harvested by trypsinization. Cells were permeabilized in PBS containing 0.25% Triton X-100 for 5 min at room temperature, followed by spinning onto a poly-l-lysine-coated cover glass through 1 M sucrose in PBS and fixed with 3% paraformaldehyde in PBS. The fixed cell nuclei were pre-blocked with 10% normal goat serum and 3% BSA in PBS, and incubated with anti-␥-H2AX monoclonal antibody and antigen-purified polyclonal antibody against RPA32 (in PBS containing 3% normal goat serum, 3% BSA and 0.5% Triton X-100) at room temperature for 1 h. Fluorescein-conjugated goat–anti-mouse and Alexa 568-conjugated goat–anti rabbit antibodies were then used as secondary antibodies at room temperature for 1 h. DAPI (2 ␮M)

218

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

was used to counterstain DNA. Total DNA (blue), RPA foci (red), and ␥-H2AX foci (green) were examined using a Leiz Orthoplan 2 epifluorescent microscope with a SPOT-RT digital camera and software. Adobe Photoshop was used for image processing and printing.

tions, and the membranes were probed with monoclonal antibodies against RPA32, Chk1 and ␥-H2AX at room temperature for 1 h. Peroxidase-conjugated goat–anti-mouse IgG (Pierce) was used as the secondary antibody. The membranes were then incubated with Supersignal ECL reagent (Pierce) and exposed to X-ray film.

2.4. S-phase cell double immunostaining For double staining in BUdR-labeled HeLa nuclei, the protocol was modified from that previously described [4,54] to optimize both the BUdR and RPA signal. Briefly, HeLa cells were labeled with 20 ␮M BUdR for 15 min. The cells were washed and fresh culture medium added before addition of the second agent. Treatment with the second agent was for 1 h. The cells were harvested and spun onto cover glass slides as described above and the slides were treated with 2 N HCl for 90 s to denature the DNA. Monoclonal antibody against BUdR and polyclonal antibody against RPA32 were used as primary antibodies. Fluorescein-conjugated goat–anti-mouse and Alexa 568-conjugated goat–anti rabbit antibodies were then used as secondary antibodies, and DAPI (2 ␮M) was used to counterstain DNA. Total DNA (blue), RPA foci (red), and BUdR incorporation (green) were examined using either a Leiz Orthoplan 2 epifluorescent microscope (Fig. 3A) or a Bio-Rad MRC-1024 confocal imaging system (Fig. 3B). The confocal images of BUdR and RPA were taken separately and Adobe Photoshop was used to create merged images.

3. Results and discussion 3.1. Reversible replication fork stalling Aphidicolin and hydroxyurea (HU) are commonly used in cell culture to synchronize cells in S phase. Subsequent removal of these agents from the culture medium allows cells to recover and reenter the cell cycle. Aphidicolin is an inhibitor of the replicative DNA polymerases ␣, ␦ and ε. Aphidicolin treatment stalls DNA replication during either initiation or elongation by blocking DNA polymerase action. In cells treated with aphidicolin, ATR-dependent Chk1 phosphorylation is induced to stabilize replication complexes on stalled forks (Fig. 1, lane 2) [24,27,37]. However, in the same cells, there is no detectable focus formation or phosphorylation of RPA or H2AX (Figs. 2–4) [4]. Since both RPA and H2AX are present at DNA replication forks and can be phosphorylated by ATR, this

2.5. Immunoblotting Mock- or drug-treated cells (1 × 106 ) were washed with PBS and lysed directly in 50 ␮l of SDS sample buffer (20 mM Tris–HCl, pH 7.5; 2% SDS; 1 M 2-mercaptoethanol). Total protein from an equal number of cells (∼5 × 104 ) was resolved by electrophoresis on 10 or 15% (w/v) SDS–polyacrylamide gels. The acrylamide/bisacrylamide ratio used was optimized for the detection of damage-induced hyper-phosphorylated RPA32, but the partly phosphorylated RPA32 species seen in S and G2 phases of an unperturbed cell cycle does not resolve under the gel conditions used. The gels were transferred to Hybond-P membrane using NovaBlot (Amersham Pharmacia Biotech) as per the manufacturer’s instruc-

Fig. 1. Induction of Chk1 phosphorylation. HeLa cells (1 × 106 ) were either mock treated or treated with either: 5 ␮M aphidicolin, 2 mM hydroxyurea (HU), 20 nM adozelesin, or 0.033% methyl methanesulfonate (MMS) for 1 h. The cells were then harvested and immediately lysed in 50 ␮l of SDS sample buffer. Ten microliters of each sample were separated by electrophoresis on a 10% polyacrylamide gel, transferred to a HyBond-P membrane and immunoblotted using a monoclonal antibody against human Chk1 protein. Peroxidase-conjugated goat–anti-mouse IgG antibody, enhanced chemiluminescent agents and X-ray film were used to detect the Chk1. Migration of phosphorylated Chk1 is slower than under-phosphorylated form and is indicated on the right side of the panel.

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

Fig. 2. Induction of RPA32 and H2AX phosphorylation. HeLa cells (1 × 106 ) with (even numbered lanes) or without (odd numbered lanes) pre-treatment with 5 ␮M aphidicolin for 1 h were further treated with nothing (mock), 2 mM HU, 20 nM adozelesin (Adozel), or 0.033% MMS for 2 h before harvesting and lysis in 50 ␮l of SDS sample buffer as in Fig. 1. Two microliters of each sample were separated on a 15% polyacrylamide gel and immunoblotted with monoclonal antibodies specific to the 32 kDa subunit of RPA (RPA32) or the ␥-H2AX. The immunoblot was performed as described in Fig. 1. Migration of RPA32, hyper-phosphorylated RPA32 (RPA32-Pi) and ␥-H2AX are indicated to the right of the panel.

suggests a missing connection between ATR and these two substrates. It has been shown that efficient Chk1 and Rad17 phosphorylation by ATR is dependent on Rad9 of the 9-1-1 complex [55]. Claspin (and its yeast homologue Mrc1) is also required for Chk1 phosphorylation by ATR (or Cds1 or Rad53 phosphorylation in yeasts) [38,56,57]. Claspin and the 9-1-1 complex were suggested to be adaptor proteins for ATR and its substrates [58]. However, phosphorylation of H2AX by ATR is independent of the 9-1-1 complex [50]. Whether these adaptors are functioning as switches to determine which down-stream pathways of ATR will

Fig. 3. Induction of RPA foci in S-phase nuclei. HeLa cells were labeled with bromouracil deoxyriboside (BUdR) for 15 min, the media was then replaced, and the cells were either mock-treated or treated with 5 ␮M aphidicolin (Aphid), 2 mM HU, 20 nM adozelesin (Adozel), or 0.033% MMS for 1 h. The cells were lysed in PBS with Triton X-100. The nuclei were spun onto glass cover slips, fixed with paraformaldehyde and treated with 2 N HCl for 90 s to denature the DNA. A monoclonal antibody against BUdR and antigen-purified polyclonal antibody against RPA32 were used as primary antibodies, and fluorescein-conjugated goat–anti-mouse and Alexa 568-conjugated goat–anti rabbit IgG were used as secondary antibodies. DAPI (2 ␮M) was used to counterstain DNA. Total DNA (blue), RPA foci (red) and BUdR incorporation (green) were examined using either: (A) a Leiz Orthoplan 2 microscope with a SPOT-RT digital camera and software, or (B) a Bio-Rad MRC-1024 confocal imaging system. In part (B), the RPA and BUdR images were merged such that co-localization is represented as yellow.

219

be activated, is an open question and will require further investigation. HU is an inhibitor of ribonucleotide reductase. HU treatment depletes the intra-cellular pool of deoxyribonucleotide triphosphates, the building blocks for DNA, thereby causing DNA polymerases to stall. As in the case of aphidicolin, HU-induced Chk1 phosphorylation (Fig. 1, lane 3) is dependent on ATR [50]. However, HU treatment differs from aphidicolin, in that it is able to induce RPA and ␥-H2AX foci and

220

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

Fig. 4. Induction of RPA and ␥-H2AX foci. HeLa cells with (right panels) or without (left panels) pre-treatment with 5 ␮M aphidicolin for 1 h, were further treated with 20 nM adozelesin (Adozel), 2 mM HU, 0.033% MMS, or mock-treated for 1 h. The cells were harvested, lysed, spun onto glass slides, fixed, and immunostained for RPA (red) and ␥-H2AX (green), using the monoclonal antibody against ␥-H2AX, essentially as described in Fig. 3A. Conditions for microscopy were the same as in Fig. 3A.

RPA32 hyper-phosphorylation (Figs. 2–4) [50,59]. As shown in Fig. 5 and in agreement with previously published results [50], HU-induced ␥-H2AX foci appear to be dependent on the kinase activity of ATR. How-

ever, HU-induction of RPA foci is not. This is consistent with recent findings that association of ATR with stalled replication forks happens after RPA binds to ssDNA [38–40].

Fig. 5. HU-induced RPA and ␥-H2AX foci in cells transfected with wild-type or dominant-negative mutant ATR. Since ATR is essential, ATR knockout cell lines are not available. Hence, a human fibroblast cell line was transfected with an expression plasmid carrying a dominant-negative kinase-dead mutant of ATR (ATR-kd) under the control of an inducible promoter. In the presence of the inducer (doxycyclin), expression of ATR-kd dramatically reduces ATR function in transfected cells. The same fibroblast cell line transfected with a wild-type ATR (ATR-wt) expression plasmid was used as a control. In cells expressing ATR-wt (left panels), 2 mM HU treatment for 1 h was able to induce RPA (red) and ␥-H2AX (green) foci, as shown above for HeLa cells (Fig. 4). However, in cells expressing ATR-kd (right panels), the HU-induced ␥-H2AX foci (green) were substantially reduced, while the RPA foci (red) remained at the same levels as in the control cells (left panels and Fig. 4).

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

In cells pulse-labeled with bromouracil deoxyriboside (BUdR) (to identify nuclei in S phase)2 , HU-induced RPA foci appear only in BUdR-positive cells (Fig. 3A), with a staining pattern similar to that of BUdR (Fig. 3B, upper panels)3 . This result suggests that the HU-induced RPA foci may occur at sites of DNA replication, consistent with the idea that HU induces a response at replication forks. One possible interpretation is that HU treatment stalls DNA polymerases more efficiently than helicases, so long stretches of ssDNA are generated which recruit large amounts of RPA. As shown previously (and in Figs. 2 and 4), adozelesin-induced focus formation of RPA and ␥-H2AX, as well as phosphorylation of RPA32, can be blocked by pre-treating the cells with aphidicolin [4]. If HU-induced cellular checkpoint responses are dependent on polymerase progression, then we should also be able to block RPA and H2AX phosphorylation and focalization by pre-treating the cells with aphidicolin. We found, however, that aphidicolin pre-treatment slightly enhanced HU-induced RPA and H2AX phosphorylation, and showed no inhibition of HU-induced RPA and ␥-H2AX foci (Figs. 2 and 4). This suggests that the combined effect of aphidicolin and HU may result in somewhat more efficient polymerase inhibition than either alone, thereby inducing somewhat stronger cellular responses. However, as aphidicolin alone efficiently blocks replicative polymerases, pre-treatment with aphidicolin followed by HU treatment should result in no more ssDNA than HU treatment alone. Therefore, aphidicolin pre-treatment followed by HU treatment would be expected to show reduced RPA focus formation. Since HU is still able to induce focus forma2 These experiments and exposures were optimized to detect damage-induced RPA staining, which is several-fold higher than normal S-phase RPA staining. RPA staining due to recruitment to DNA replication foci can be seen by the faint RPA staining in the mock-treated S-phase nuclei (in Fig. 3, note that only those nuclei labeled with BUdR are positive for RPA). Staining of RPA in undamaged S-phase nuclei is light due to the procedure and exposure used. 3 Cells were pulse-labeled with BUdR for 15 min. Between the time of BUdR removal and addition of fresh culture medium and the time that the second agent was added to the cultures took an estimated 10–15 min. This delay is likely responsible for the slight offset between the patterns of BUdR and RPA staining in the HU treated cells.

221

tion and phosphorylation of RPA and H2AX in the presence of aphidicolin, we conclude that the inability of aphidicolin to induce these cellular responses is not because aphidicolin necessarily blocks these responses, but rather that the treatment with aphidicolin may not rise to a level sufficient to induce these responses. Furthermore, these results suggest that HU triggers responses through a second mechanism that is fork-dependent but not dependent on polymerase progression. One possibility would be that, on aphidicolin-stalled replication forks, HU is able to induce more ssDNA through a pathway independent of DNA polymerase progression. An alternate possibility is that the amount of ssDNA generated at replication forks by uncoupling of unwinding from polymerization depends on the dNTP pool usage, which would be high in cells treated with aphidicolin alone and low in cells treated with aphidicolin plus HU. Another possibility is that HU treatment of aphidicolin-stalled replication forks could trigger an additional adaptor of ATR kinase to phosphorylate additional ATR substrates, other than just Chk1. These hypotheses are currently under investigation. 3.2. Replication fork blocking During S phase of the cell cycle, damage to DNA activates checkpoint pathways through trans-acting mechanisms to slow down DNA synthesis and prevent the firing of late replication origins. If damaged DNA is not removed and replication forks collide with damaged sites, DNA helicases or polymerases are generally physically impeded by these DNA lesions, resulting in blocked replication forks. For DNA lesions that are poorly detected or unresolved by cellular DNA repair pathways, the primary DNA damage response to these lesions will likely be triggered by replication fork blockage. Several anti-cancer agents selectively target cells in S phase through this mechanism. Camptothecin (CPT) is one of the best studied of these agents [60]. CPT is a topoisomerase I inhibitor that traps topoisomerase I as a covalent enzyme-nicked DNA complex. These “cleavable complexes” are not recognized by DNA damage checkpoint mechanisms until replication forks collide with them and create strand breaks (see [61], in press). It has been shown that in the presence of aphidicolin (the absence of replication fork progression), CPT-induced DSBs, as

222

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

well as focus formation and phosphorylation of RPA and H2AX are all prevented [4,62]. We have recently shown that cellular DNA damage responses induced by the DNA alkylating agent adozelesin can also be blocked by pre-treating cells with aphidicolin. However, in contrast to CPT, adozelesin does not cause DNA strand breaks during short-term treatment [5]; therefore, the adozelesin adduct-blocking of replication forks appears to be the trigger for the checkpoint responses induced by adozelesin. Methyl methanesulfonate (MMS) is another DNA alkylating agent that is commonly used to study S-phase DNA damage checkpoint responses. While it has been reported that MMS treatment induces Rad53 phosphorylation in yeast [63] and Chk1 phosphorylation in Xenopus oocyte extracts [64], we did not detect Chk1 phosphorylation in HeLa cells for up to 4 h following MMS treatment (Fig. 1, lane 5). MMS-induced RPA32 hyper-phosphorylation only occurs after a long delay, and is barely detectable following a 2-h incubation (Fig. 2 and data not shown). However, it is striking that such high levels of ␥-H2AX are induced within the 2-h MMS treatment (Fig. 2, lane 7). While MMS-induced RPA foci can be seen in cells outside of S phase, the intensity of RPA staining is stronger in BUdR-positive (S-phase) cells (Fig. 3A). This indicates that S-phase cells are preferentially subjected to damage caused by MMS treatment. However, unlike the RPA staining observed after treatment with adozelesin or HU, the staining pattern of RPA induced by MMS did not match the BUdR staining pattern (Fig. 3B), suggesting that the preferential S-phase damage caused by MMS does not occur at sites of DNA replication, and is not replication fork associated. Further, treatment of cells with aphidicolin prior to the addition of MMS did not block the formation of either RPA or ␥-H2AX foci (Fig. 4). Under these conditions, aphidicolin pre-treatment actually enhanced MMS-induced RPA32 phosphorylation slightly (Fig. 2, lanes 7 and 8). Together, these results indicate that MMS induction of DNA damage responses, although S-phase stimulated, does not depend on polymerase progression. A possible explanation for this is that methylation of DNA by MMS is readily detectable by DNA repair pathways, which then can directly activate DNA damage checkpoint mechanisms. A corollary to this would be that MMS induction of checkpoint responses is not dependent

Table 1 Cellular responses to stalled replication forks

Phospho-Chk1 Phospho-RPA32 ␥-H2AX RPA foci ␥-H2AX foci Apoptosis

Aphidicolin

HU

Adozelesin

MMS

+ − − − − −

+ + + + + −

+ + + + + +

− Delayed +++ + + +

on replication fork blockage. This lack of replication fork dependence may explain why we did not detect an induction of Chk1 phosphorylation by MMS treatment, since Chk1 acts directly at compromised replication forks. 3.3. Conclusions We have compared cellular responses to DNA polymerase inhibition by aphidicolin or HU with responses to polymerase blockage by adozelesin or MMS. The results are summarized in Table 1. 3.3.1. Polymerase stalling In mammalian cells treated with aphidicolin, DNA replication is arrested, and ATR and Chk1 are activated to stabilize DNA replication forks associated with stalled polymerases. The polymerase stalling induced by aphidicolin may be similar to the stalling that occurs during normal S phase [23]. ATR/Mec1 associates with replication complexes at replication forks during S phase, and it is believed that ATR and Chk1 function to stabilize paused replication complexes [23,24]. This is consistent with the idea that ATR functions as a DNA replication checkpoint sensor. It has been reported that low levels of phospho-Chk1 can be detected in normal cycling cells [65], which is consistent with the idea that ATR and Chk1 act during an unperturbed S phase to stabilize paused replication forks. Since aphidicolin affects the DNA polymerases at all replication forks, this likely explains why aphidicolin treatment results in a substantial increase in levels of phospho-Chk1. Since HU treatment triggers additional cellular damage responses beyond those seen with aphidicolin alone, HU must induce something other than just stalled polymerases. HU treatment is able to induce

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

RPA and ␥-H2AX foci at replication forks already stalled by aphidicolin; therefore HU is likely to affect additional factors associated with the DNA replication complex. The association of ribonucleotide reductase (the target of HU) with S-phase checkpoint mechanisms leads to the possibility that ribonucleotide reductase may function directly in a DNA replication checkpoint. This possibility is supported by the finding that the ribonucleotide reductase inhibitor, Sml1, acts directly downstream of Mec1 [66]. Another possibility is that dNTP pools, which are diminished by HU but not by aphidicolin, regulate other S-phase checkpoint responses, including those leading to RPA and H2AX phosphorylation and focus formation. Yet another possibility is that HU treatment leads to more extensive fork unwinding than aphidicolin treatment, thereby generating a stronger signal that activates RPA and H2AX responses. These possibilities are not mutually exclusive. 3.3.2. Polymerase blocking The DNA alkylating agents adozelesin and MMS are both able to damage DNA directly by forming DNA adducts. We have shown in this study that these two drugs preferentially induce DNA damage responses in cells in S phase. However, RPA and H2AX focus formation and phosphorylation induced by adozelesin, but not MMS, are dependent on active replication fork progression, as shown by their different sensitivities to aphidicolin pre-treatment. Similar to HU treatment, adozelesin-induced RPA and ␥-H2AX foci exhibit staining patterns that match that of incorporated BUdR (Fig. 3B and data not shown), these responses are likely due to replication checkpoint pathways at replication forks, and likely involve the induction of Chk1 activity to stabilize stalled replication complexes. Conversely, the small methyl groups on DNA induced by MMS are readily detectable by cellular repair pathways, and these pathways are likely capable of activating S-phase DNA damage checkpoint pathways independently of replication fork checkpoints. MMS-induced checkpoint responses are independent of replication fork progression. This is shown by the insensitivity of MMS-induced ␥-H2AX and RPA focalization and phosphorylation to aphidicolin. The lack of co-localization of RPA and BUdR in MMS-treated cells (Fig. 3B), which is seen even at very low levels of MMS down to 0.01%, suggests

223

that MMS-triggered checkpoint responses are independent of replication forks. This may explain why MMS did not result in detectable Chk1 phosphorylation, since Chk1 acts at stalled replication forks to stablilize them [27–30]. We envision two possible explanations for why aphidicolin pre-treatment does not result in decreased checkpoint activation due to MMS-induced DNA adducts. One is that the efficient recognition of MMS adducts by DNA repair pathways would act to recruit much of the cellular RPA, PCNA, DNA polymerase ␦, and other proteins shared by the DNA replication and repair pathways. This could sequester these essential replication proteins away from replication forks, resulting in many fewer active replication forks, and much less chance of fork collisions with MMS adducts. A second possibility is that the MMS-activated repair pathways may efficiently induce a trans-acting mechanism to suppress DNA replication, so there is very little opportunity for the forks to collide with MMS adducts. Whichever model is correct, the lack of fork collision with MMS adducts would explain the lack of Chk1 phosphorylation in MMS-treated cells. One would expect that the cellular responses to polymerase stalling and blocking should be very different. Indeed, while treatment of cells with adozelesin or MMS for 4 h results in a high percentages of cells entering programmed cell death, prolonged treatment (up to 18 h) with aphidicolin or HU does not induce any detectable apoptosis (data not shown). However, the differences we see among the immediate early responses to polymerase stalling and blocking do not fall into obvious categories. Aphidicolin, HU, and adozelesin treatments are all able to activate ATR and Chk1. The responses to treatment with HU (a polymerase stalling agent) and adozelesin (a polymerase blocking agent) presented in this study are very similar. The differences in cell survival must come from downstream effects. While ATR may be a primary replication checkpoint sensor, there are secondary mechanisms to determine which ATR pathway(s) will be activated. We hypothesize that the adaptor proteins, 9-1-1, Claspin and possibly others, are likely involved in these decisions. Based on the variety of checkpoint response patterns seen to date, we anticipate that additional adaptor proteins will soon be identified.

224

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

Acknowledgements We would like to thank Dr. Karlene Cimprich for the ATR cell lines, Dr. Yueh-Ming Loo for comments on this manuscript and Gabriel Martins for the assistance with the confocal microscopy. This work was supported by National Institutes of Health research Grant CA 89259 (to TM). SRK was supported by an US Army Breast Cancer Research Program Postdoctoral Traineeship, DAMD17-00-1-0418, and TM was supported by a National Institutes of Health Independent Scientist Award, K02 AI01686.

References [1] J.M. Sidorova, L.L. Breeden, Precocious G1/S transitions and genomic instability: the origin, connection, Mutat. Res. 232 (2003) 5–19. [2] P.R. Andreassen, O.D. Lohez, R.L. Margolis, G2 and spindle assembly checkpoint adapation, and tetraploidy arrest: Implications for intrinsic and chemically-induced genomic instability, Mutat. Res. 232 (2003) 245–253. [3] M.P. Longhese, M. Clerici, G. Lucchini, The S-phase checkpoint and its regulation, Mutat. Res. 232 (2003) 41–58. [4] J.-S. Liu, S.-R. Kuo, T.A. Beerman, T. Melendy, Induction of DNA damage responses by Adozelesin is S phase-specific and dependent on active replication forks, Mol. Cancer Ther. 2 (2003) 41–47. [5] R.J. Cobuzzi Jr., W.C. Burhans, T.A. Beerman, Inhibition of initiation of simian virus 40 DNA replication in infected BSC-1 cells by the DNA alkylating drug adozelesin, J. Biol. Chem. 271 (1996) 19852–19859. [6] A. Di Leonardo, S.P. Linke, K. Clarkin, G.M. Wahl, DNA damage triggers a prolonged p53-dependent G1 arrest and long-term induction of Cip1 in normal human fibroblasts, Genes Dev. 8 (1994) 2540–2551. [7] J.C. Game, DNA double-strand breaks and the RAD50– RAD57 genes in Saccharomyces, Seminars Cancer Biol. 4 (1993) 73–83. [8] A.J. Osborn, S.J. Elledge, L. Zou, Checking on the fork: the DNA-replication stress-response pathway, Trends Cell Biol. 12 (2002) 509–516. [9] F. al-Khodairy, A.M. Carr, DNA repair mutants defining G2 checkpoint pathways in Schizosaccharomyces pombe, EMBO J. 11 (1992) 1343–1350. [10] M. Rauen, M.A. Burtelow, V.M. Dufault, L.M. Karnitz, The human checkpoint protein hRad17 interacts with the PCNA-like proteins hRad1, hHus1, and hRad9, J. Biol. Chem. 275 (2000) 29767–29771. [11] C. Venclovas, M.P. Thelen, Structure-based predictions of Rad1, Rad9, Hus1 and Rad17 participation in sliding clamp and clamp-loading complexes, Nucl. Acids Res. 28 (2000) 2481–2493.

[12] J.D. Griffith, L.A. Lindsey-Boltz, A. Sancar, Structures of the human Rad17-replication factor C and checkpoint Rad 9-1-1 complexes visualized by glycerol spray/low voltage microscopy, J. Biol. Chem. 277 (2002) 15233–15236. [13] L.A. Lindsey-Boltz, V.P. Bermudez, J. Hurwitz, A. Sancar, Purification and characterization of human DNA damage checkpoint Rad complexes, Proc. Natl. Acad. Sci. U.S.A. 98 (2001) 11236–11241. [14] D. Cortez, S. Guntuku, J. Qin, S.J. Elledge, ATR and ATRIP: partners in checkpoint signaling, Science 294 (2001) 1713– 1716. [15] S. Martensson, O. Hammarsten, DNA-dependent protein kinase catalytic subunit: structural requirements for kinase activation by DNA ends, J. Biol. Chem. 277 (2002) 3020– 3029. [16] G.C. Smith, N. Divecha, N.D. Lakin, S.P. Jackson, DNAdependent protein kinase and related proteins, Biochem. Soc. Symp. 64 (1999) 91–104. [17] M.B. Kastan, D.S. Lim, The many substrates and functions of ATM, Nat. Rev. Mol. Cell Biol. 1 (2000) 179–186. [18] Y. Shiloh, ATM and ATR: networking cellular responses to DNA damage, Curr. Opin. Genet. Dev. 11 (2001) 71–77. [19] E.J. Brown, D. Baltimore, ATR disruption leads to chromosomal fragmentation and early embryonic lethality, Genes Dev. 14 (2000) 397–402. [20] Q. Liu, S. Guntuku, X.S. Cui, S. Matsuoka, D. Cortez, K. Tamai, G. Luo, S. Carattini-Rivera, F. DeMayo, A. Bradley, L.A. Donehower, S.J. Elledge, Chk1 is an essential kinase that is regulated by Atr and required for the G(2)/M DNA damage checkpoint, Genes Dev. 14 (2000) 1448–1459. [21] K. Unsal-Kacmaz, A.M. Makhov, J.D. Griffith, A. Sancar, Preferential binding of ATR protein to UV-damaged DNA, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 6673–6678. [22] R.S. Tibbetts, D. Cortez, K.M. Brumbaugh, R. Scully, D. Livingston, S.J. Elledge, R.T. Abraham, Functional interactions between BRCA1 and the checkpoint kinase ATR during genotoxic stress, Genes Dev. 14 (2000) 2989–3002. [23] R.S. Cha, N. Kleckner, ATR homolog Mec1 promotes fork progression, thus averting breaks in replication slow zones, Science 297 (2002) 602–606. [24] M. Hekmat-Nejad, Z. You, M.C. Yee, J.W. Newport, K.A. Cimprich, Xenopus ATR is a replication-dependent chromatin-binding protein required for the DNA replication checkpoint, Curr. Biol. 10 (2000) 1565–1573. [25] H. Zhao, H. Piwnica-Worms, ATR-mediated checkpoint pathways regulate phosphorylation and activation of human Chk1, Mol. Cell Biol. 21 (2001) 4129–4139. [26] T.P. Heffernan, D.A. Simpson, A.R. Frank, A.N. Heinloth, R.S. Paules, M. Cordeiro-Stone, W.K. Kaufmann, An ATRand Chk1-dependent S checkpoint inhibits replicon initiation following UVC-induced DNA damage, Mol. Cell Biol. 22 (2002) 8552–8561. [27] C. Feijoo, C. Hall-Jackson, R. Wu, D. Jenkins, J. Leitch, D.M. Gilbert, C. Smythe, Activation of mammalian Chk1 during DNA replication arrest: a role for Chk1 in the intra-S-phase checkpoint monitoring replication origin firing, J. Cell Biol. 154 (2001) 913–923.

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226 [28] J.A. Tercero, J.F. Diffley, Regulation of DNA replication fork progression through damaged DNA by the Mec1/Rad53 checkpoint, Nature 412 (2001) 553–557. [29] Y. Wang, T.A. Beerman, D. Kowalski, Anti-tumor drug adozelesin differentially affects active and silent origins of DNA replication in yeast checkpoint kinase mutants, Cancer Res. 61 (2001) 3787–3794. [30] G. Zachos, M.D. Rainey, D.A.F. Gillespie, Chk1-deficient tumour cells are viable but exhibit multiple checkpoint and survival defects, EMBO J. 22 (2003) 713–723. [31] M.P. Longhese, H. Neecke, V. Paciotti, G. Lucchini, P. Plevani, The 70 kDa subunit of replication protein A is required for the G1/S and intra-S DNA damage checkpoints in budding yeast, Nucl. Acids Res. 24 (1996) 3533–3537. [32] G. D’Urso, B. Grallert, P. Nurse, DNA polymerase alpha, a component of the replication initiation complex, is essential for the checkpoint coupling S phase to mitosis in fission yeast, J. Cell Sci. 108 (1995) 3109–3118. [33] T. Naiki, T. Shimomura, T. Kondo, K. Matsumoto, K. Sugimoto, Rfc5, in cooperation with rad24, controls DNA damage checkpoints throughout the cell cycle in Saccharomyces cerevisiae, Mol. Cell Biol. 20 (2000) 5888– 5896. [34] K. Sugimoto, S. Ando, T. Shimomura, K. Matsumoto, Rfc5, a replication factor C component, is required for regulation of Rad53 protein kinase in the yeast checkpoint pathway, Mol. Cell Biol. 17 (1997) 5905–5914. [35] H.S. Kim, S.J. Brill, Rfc4 interacts with Rpa1 and is required for both DNA replication and DNA damage checkpoints in Saccharomyces cerevisiae, Mol. Cell Biol. 21 (2001) 3725– 3737. [36] T.A. Navas, Z. Zhou, S.J. Elledge, DNA polymerase epsilon links the DNA replication machinery to the S-phase checkpoint, Cell 80 (1995) 29–39. [37] W.M. Michael, R. Ott, E. Fanning, J. Newport, Activation of the DNA replication checkpoint through RNA synthesis by primase, Science 289 (2000) 2133–2137. [38] J. Lee, A. Kumagai, W.G. Dunphy, Claspin, a Chk1regulatory protein, monitors DNA replication on chromatin independently of RPA, ATR, and Rad17, Mol. Cell 11 (2003) 329–340. [39] L. Zou, S.J. Elledge, Sensing DNA damage through ATRIP recognition of RPA-ssDNA complexes, Science 300 (2003) 1542–1548. [40] V. Costanzo, D. Shechter, P.J. Lupardus, K.A. Cimprich, M. Gottesman, J. Gautier, An ATR- and Cdc7-dependent DNA damage checkpoint that inhibits initiation of DNA replication, Mol. Cell 11 (2003) 203–213. [41] N. Kantake, T. Sugiyama, R.D. Kolodner, S.C. Kowalczykowski, The recombination-deficient mutant RPA (rfa1-t11) is displaced slowly from single-stranded DNA by Rad51 protein, J. Biol. Chem. 278 (2003) 23410–23417. [42] M.S. Wold, Replication protein A: a heterotrimeric, singlestranded DNA-binding protein required for eukaryotic DNA metabolism, Ann. Rev. Biochem. 66 (1997) 61–92. [43] C. Iftode, Y. Daniely, J.A. Borowiec, Replication protein A (RPA): the eukaryotic SSB, Crit. Rev. Biochem. Mol. Biol. 34 (1999) 141–180.

225

[44] S.M. Patrick, J.J. Turchi, Replication protein A (RPA) binding to duplex cisplatin-damaged DNA is mediated through the generation of single-stranded DNA, J. Biol. Chem. 274 (1999) 14972–14978. [45] Y. Lao, X.V. Gomes, Y. Ren, J.S. Taylor, M.S. Wold, Replication protein A interactions with DNA. Part III. Molecular basis of recognition of damaged DNA, Biochemistry 39 (2000) 850–859. [46] C. Redon, D. Pilch, E.P. Rogakou, O. Sedelnikova, K. Newrock, W.M. Bonner, Histone H2A variants H2AX and H2AZ, Curr. Opin. Genet. Dev. 12 (2002) 162–169. [47] A. Celeste, S. Petersen, P.J. Romanienko, O. FernandezCapetillo, H.T. Chen, O.A. Sedelnikova, B. Reina-San-Martin, V. Coppola, E. Meffre, M.J. Difilippantonio, C. Redon, D.R. Pilch, A. Olaru, M. Eckhaus, R.D. Camerini-Otero, L. Tessarollo, F. Livak, K. Manova, W.M. Bonner, M.C. Nussenzweig, A. Nussenzweig, Genomic instability in mice lacking histone H2AX, Science 296 (2002) 922–927. [48] C.H. Bassing, K.F. Chua, J. Sekiguchi, H. Suh, S.R. Whitlow, J.C. Fleming, B.C. Monroe, D.N. Ciccone, C. Yan, K. Vlasakova, D.M. Livingston, D.O. Ferguson, R. Scully, F.W. Alt, Increased ionizing radiation sensitivity and genomic instability in the absence of histone H2AX, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 8173–8178. [49] A. Peng, P.L. Chen, NFBD1, like 53BP1, is an early and redundant transducer mediating Chk2 phosphorylation in response to DNA damage, J. Biol. Chem. 278 (2003) 8873– 8876. [50] I.M. Ward, J. Chen, Histone H2AX is phosphorylated in an ATR-dependent manner in response to replicational stress, J. Biol. Chem. 276 (2001) 47759–47762. [51] E.P. Rogakou, C. Boon, C. Redon, W.M. Bonner, Megabase chromatin domains involved in DNA double-strand breaks in vivo, J. Cell Biol. 146 (1999) 905–916. [52] S. Din, S.J. Brill, M.P. Fairman, B. Stillman, Cell-cycleregulated phosphorylation of DNA replication factor A from human and yeast cells, Genes Dev. 4 (1990) 968–977. [53] J. Dziegielewski, T.A. Beerman, Cellular responses to the DNA strand-scission enediyne C-1027 can be independent of ATM, ATR, and DNA-PK kinases, J. Biol. Chem. 277 (2002) 20549–20554. [54] D.S. Dimitrova, I.T. Todorov, T. Melendy, D.M. Gilbert, Mcm2, but not RPA, is a component of the mammalian early G1-phase pre-replication complex, J. Cell Biol. 146 (1999) 709–722. [55] L. Zou, D. Cortez, S.J. Elledge, Regulation of ATR substrate selection by Rad17-dependent loading of Rad9 complexes onto chromatin, Genes Dev. 16 (2002) 198–208. [56] K. Tanaka, P. Russell, Mrc1 channels the DNA replication arrest signal to checkpoint kinase Cds1, Nat. Cell Biol. 3 (2001) 966–972. [57] A.A. Alcasabas, A.J. Osborn, J. Bachant, F. Hu, P.J. Werler, K. Bousset, K. Furuya, J.F. Diffley, A.M. Carr, S.J. Elledge, Mrc1 transduces signals of DNA replication stress to activate Rad53, Nat. Cell Biol. 3 (2001) 958–965. [58] C. Claudia, S. Chini, J. Chen, Human claspin is required for replication checkpoint control, J. Biol. Chem. 278 (2003) 30057–30062.

226

J.-S. Liu et al. / Mutation Research 532 (2003) 215–226

[59] G.S. Brush, C.W. Anderson, T.J. Kelly, The DNA-activated protein kinase is required for the phosphorylation of replication protein A during simian virus 40 DNA replication, Proc. Natl. Acad. Sci. U.S.A. 91 (1994) 12520–12524. [60] L.F. Liu, S.D. Desai, T.K. Li, Y. Mao, M. Sun, S.P. Sim, Mechanism of action of camptothecin, Ann. N. Y. Acad. Sci. 922 (2000) 1–10. [61] Y. Pommier, C. Redon, V.A. Rao, J.A. Seiler, O. Sordet, H. Takemura, S. Antony, L.H. Meng, Z.Y. Liao, G. Kohlhagen, H.L. Zhang, K.W. Kohn, Repair of and checkpoint response to topoisomeras I-mediated DNA damage, Mutat. Res. 232 (2003) 173–203. [62] D. Strumberg, A.A. Pilon, M. Smith, R. Hickey, L. Malkas, Y. Pommier, Conversion of topoisomerase I cleavage complexes on the leading strand of ribosomal DNA into 5 -phosphorylated DNA double-strand breaks by replication runoff, Mol. Cell Biol. 20 (2000) 3977–3987.

[63] M. Chang, M. Bellaoui, C. Boone, G.W. Brown, A genome-wide screen for methyl methanesulfonate-sensitive mutants reveals genes required for S-phase progression in the presence of DNA damage, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 16934–16939. [64] P.J. Lupardus, T. Byun, M.C. Yee, M. Hekmat-Nejad, K.A. Cimprich, A requirement for replication in activation of the ATR-dependent DNA damage checkpoint, Genes Dev. 16 (2002) 2327–2332. [65] K. Jiang, E. Pereira, M. Maxfield, B. Russell, D.M. Goudelock, Y. Sanchez, Regulation of Chk1 includes chromatin association and 14-3-3 binding following phosphorylation on Ser-345, J. Biol. Chem. 278 (2003) 25207–25217. [66] X. Zhao, R. Rothstein, The Dun1 checkpoint kinase phosphorylates and regulates the ribonucleotide reductase inhibitor Sml1, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 3746– 3751.