Status of P53 in Human Cancer Cells Does Not Predict Efficacy of CHK1

Oncogene (2010) 29, 6149–6159 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc ORIGINAL ARTICLE Status of p53 in human cancer cells does not predict efﬁcacy of CHK1 kinase inhibitors combined with chemotherapeutic agents

S Zenvirt1, N Kravchenko-Balasha1 and A Levitzki

Unit of Cellular Signaling, Department of Biological Chemistry, The Hebrew University of Jerusalem, Jerusalem, Israel

DNA damage checkpoints cause cell cycle arrest, allowing Millar, 2006). An important regulator of this response DNA repair before resumption of the cell cycle. These is the tumor suppressor p53. In response to DNA checkpoints can be activated through several signaling damage, p53 is activated and stabilized by the Ser/Thr pathways. Checkpoint activators include p53, checkpoint kinases Ataxia-Telangiectasia mutated and Ataxia- kinase 1 (CHK1), checkpoint kinase 2 and/or MAPKAP Telangiectasia- and RAD3-related (Oren, 2003; Meek, kinase 2 (MK2). Many cancer cells lack p53 activity and, 2004; Harris and Levine, 2005). The activated p53 therefore, depend on alternative checkpoint activators to protein stimulates the expression of p21, which inhibits arrest the cell cycle following DNA damage. Inhibition of the cyclin-dependent kinases, resulting in cell cycle arrest these pathways is expected to specifically sensitize these at both the G1-to-S (G1/S) and the G2-to-M (G2/M) p53-deficient cells to DNA damage caused by chemother- transitions (Vogelstein et al., 2000). In the absence of apy. Using isogenic p53-proficient and p53-deficient cancer p53, the G1/S checkpoint cannot be properly activated. cell lines, we show that inactivation of CHK1, but not Intra-S and G2/M checkpoints can be activated through MK2, abrogates cell cycle arrest following chemotherapy, the Ser/Thr kinases checkpoint kinase 1 (CHK1), specifically in p53-deficient cells. However, we show that checkpoint kinase 2 (CHK2) and MAPKAP kinase 2 CHK1 is required to maintain genome integrity and cell (MK2). CHK1 is activated by Ataxia-Telangiectasia- viability, and that p53-proficient cells are no less sensitive and RAD3-related following genotoxic stress, CHK2 is than p53-deficient cells to CHK1 inhibition in the pre- activated by Ataxia-Telangiectasia mutated following sence of DNA damage. Thus, combining CHK1 inhibition double strand DNA breakage such as that caused by with DNA damage does not lead to preferential killing ionizing radiation and MK2 is activated by the p38 of p53-deficient over p53-proficient cells, and inhibiting MAP kinase following ultra-violet radiation or che- CHK1 does not appear to be a promising approach for motherapy (Abraham, 2001; Bartek and Lukas, 2003; potentiation of cancer chemotherapy. Shiloh, 2003; Manke et al., 2005; Karlsson-Rosenthal Oncogene (2010) 29, 6149–6159; doi:10.1038/onc.2010.343; and Millar, 2006; Reinhardt et al., 2007). CHK1, CHK2 published online 23 August 2010 and MK2 phosphorylate and thereby inactivate the Cdc25 phosphatases, resulting in inactivation of cyclin- Keywords: CHK1; checkpoint; cancer; p53 dependent kinases and cell cycle arrest (Donzelli and Draetta, 2003; Karlsson-Rosenthal and Millar, 2006). In about half of human cancers, the activity of p53 is Introduction compromised and the G1/S checkpoint fails (Vogelstein et al., 2000; Reinhardt et al., 2007). It has been suggested DNA damaging agents have been in use for cancer that CHK1 or MK2 inhibitors would abrogate the therapy for decades. It has been suggested that these remaining checkpoints in cancer cells lacking functional agents could be potentiated by inhibition of the DNA p53 and this would lead to preferential sensitization of damage checkpoints (Zhou et al., 2003; Kawabe, 2004; these cancer cells to chemotherapy over normal p53- Zhou and Bartek, 2004; Stark and Taylor, 2006; proficient cells (Zhou et al., 2003; Kawabe, 2004; Zhou O’Connor et al., 2007; Ashwell et al., 2008; Bucher and Bartek, 2004; Stark and Taylor, 2006; O’Connor and Britten, 2008; Lieberman, 2008). In response to et al., 2007; Ashwell et al., 2008; Bucher and Britten, DNA damage, the cell cycle is arrested at specific 2008; Lieberman, 2008). A recent publication advocated checkpoints, allowing DNA repair or, if the damage the utilization of MK2 inhibitors, and reported that the cannot be repaired, activation of programmed cell death G2/M checkpoint in p53À/À ras transformed cells (Zhou and Elledge, 2000; Karlsson-Rosenthal and required MK2 (Reinhardt et al., 2007). In view of these reports concerning the utility of CHK1 and/or MK2 Correspondence: Professor A Levitzki, Department of Biological inhibitors as sensitizers of p53À/À cells to chemother- Chemistry, Unit of Cellular Signaling, The Hebrew University of apy, we examined the sensitivity of two isogenic pairs of Jerusalem, Givat Ram, Jerusalem, Israel 91904, Israel. p53-proficient and p53-deficient cancer cell lines to E-mail: [email protected] DNA damaging agents, when CHK1 or MK2 expres- 1These authors contributed equally to this work. Received 20 January 2010; revised and accepted 6 July 2010; published sion was knocked down by small interference RNA online 23 August 2010 (siRNA) or when their catalytic activity was inhibited. p53, CHK1 and chemotherapy S Zenvirt et al 6150 We discuss our results in view of the published literature nucleation were also detected (Figure 2bV). When no and question whether there is wide utility for CHK1 or chemotherapeutic drug was administered, 5–8% of the MK2 inhibitors as anti-cancer agents. cells that were transfected with non-targeting siRNA were in mitosis in both the cell lines (Figure 2c). Knocking down CHK1 in p53-proficient U2OS cells Results resulted in a decrease in mitosis, but not in p53-deficient cells. Knockdown of MK2 had no effect in either cell CHK1 but not MK2 is responsible for cell cycle arrest line. When CHK1 and MK2 were knocked down of p53 deficient cells in response to chemotherapy together, knockdown of MK2 antagonized the effect We first examined which checkpoint kinase should be of CHK1 knockdown in the p53-proficient cells inhibited in order to specifically abrogate checkpoints in (Figure 2c). p53-deficient cancer cells undergoing chemotherapy. We DNA damage is expected to activate checkpoints, used siRNA to inhibit the expression of CHK1, MK2 or resulting in cell cycle arrest. As expected, exposure of both. To verify the role of p53 in the maintenance of the cells that were transfected with non-targeting siRNA to checkpoint, we used a pair of isogenic cell lines derived chemotherapeutic drugs dramatically reduced the num- from U2OS: U2OS-LacZÀ, a p53-proficient cell line ber of mitotic cells in both the p53-proficient and p53- stably expressing siRNA against LacZ, and U2OS-p53À, deficient U2OS cell lines, indicating the onset of cell which stably expresses siRNA against p53 (Raver- cycle arrest. In most of these arrested cells, cyclin B1 Shapira et al., 2007). To confirm that U2OS-p53À have accumulated in the cytoplasm, indicating that they were abrogated activity of p53, we measured the effect of arrested at G2 or S phase (Figure 2bIV). Inhibition of Nutlin (Mdm2 inhibitor) on induction of p53-regulated the kinases that are responsible for the onset of the genes. As shown in Figure 1a, Nutlin induced the checkpoints is expected to abrogate the cell cycle arrest expression of p21 under basal conditions as well as induced by DNA damage. We indeed found that under cisplatin (CDDP) treatment in U2OS-LacZÀ cells, knocking down CHK1 restored mitosis in the p53- whereas no induction occurred in U2OS-p53À cells. deficient cells, but not in the p53-proficient cells, Moreover, under CDDP treatment, U2OS-LacZÀ cells following treatment with CDDP or irinotecan underwent caspase-3 dependent apoptotic cell death, (Figure 2c). Thr3-phosphorylation of histone H3 is a known to be p53 mediated (Figure 1b), whereas U2OS- marker of cells entering prophase (Dai et al., 2005). p53– cells did not activate caspase-3, as expected from When cells were knocked down for CHK1 and treated cells lacking p53. with CDDP, U20S-LacZ– (p53-proficient) cells were Western blot analysis confirmed that efficient and arrested before prophase, but U20S-p53À (p53-deficient) specific knockdown of CHK1 or MK2 was achieved cells had an increase in phosphorylated H3, indicating within 48 h of transfection and lasted for at least 3 days that they were able to proceed into prophase (Figure 2a and data not shown). One day post siRNA (Figure 2d). transfection, cells were exposed to the DNA cross-linker These results indicate that CHK1 is responsible for agent CDDP, or the topoisimerase I inhibitor irinote- the chemotherapy-induced cell cycle arrest in the can, both at concentrations of 10 mM. absence of p53. In contrast, knocking down MK2 did To detect checkpoint activation and cell cycle arrest, not affect the number of cells undergoing mitosis in we used immuno-fluorescence staining for cyclin B1 and either cell line, independent of the status of p53. These 40,6-diamidino-2-phenylindole staining for heterochro- results clearly indicate that CHK1, but not MK2, is matin, 24 h after application of the chemotherapeutic required for the activation of the DNA damage drugs. Cells that had entered mitosis were identified by checkpoint following chemotherapy in these cells. In the nuclear staining of cyclin B1 and by chromosome addition, the combined knockdown of CHK1 and MK2 condensation (Figure 2bII). Cells at later stages of had a smaller effect than the knockdown of CHK1 alone mitosis were identified by the morphology of the (Figure 2c), suggesting that MK2 knockdown is dividing chromatin (Figure 2bIII). Cells with micro- antagonistic to CHK1 knockdown.

No No chemotherapy Cisplatin chemotherapy Cisplatin - - - - U2OS: -LacZ- -p53- -LacZ- -p53- U2OS: -LacZ -p53 -LacZ -p53 -N +N -N +N -N +N -N +N p53 p53 p21 p21 p19 - CASP3 p17 - Tubulin Tubulin

Figure 1 Abrogated activity of p53 signaling in U2OS-p53À cells. Western blot conﬁrming depletion of p53 in U2OS-p53À cells (a) under Nutlin (N) treatment (8 mM, 24 h) without activation and following activation of p53 by 20 mM CDDP. The expression of p21 was measured. (b) Caspase-3 activation was indicated by the presence of p19 and p17 products, which were measured under 20 mM CDDP treatment. Tubulin is presented as a loading control.

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6151 U2OS-LacZ- U2OS-p53- Cisplatin (10μM) siRNA: CHK1MK2 X2 Control CHK1MK2 X2 Control No chemotherapy: I II III IV V CHK1

MK2

GAPDH

U2OS-LacZ- U2OS-p53- Mitosis Mitosis 15% 15%

10% 10%

5% 5%

0% 0% Percentage of cells No chemotherapyCisplatin Irinotecan Percentage of cells No chemotherapyCisplatin Irinotecan

U2OS-LacZ- U2OS-p53- Micronucleation Micronucleation 15% 15%

10% 10%

5% 5%

0% Percentage of cells Percentage of cells No chemotherapyCisplatin Irinotecan No chemotherapyCisplatin Irinotecan

siControl siCHK1 siMK2 siX2 No chemotherapy Cisplatin 4μM

- - - - U2OS: -LacZ -p53 U2OS: -LacZ -p53 siRNA: - CHK1 Con V - CHK1 Con V siRNA: - CHK1Con V - CHK1Con V CHK1 CHK1

pHis pHis

βCatenin βCatenin

Cisplatin 8μM U2OS: -LacZ- -p53- siRNA: - CHK1 Con V - CHK1 Con V CHK1

pHis

βCatenin

Figure 2 Cell cycle arrest by chemotherapy and its abrogation: p53-proficient U2OS-LacZÀ and p53-deficient U2OS-p53À cells were transfected with siRNA against CHK1, MK2, both (X2) or non-targeting siRNA (Control), as indicated. 24 h later the cells were treated with the chemotherapeutic drugs CDDP or irinotecan at 10 mM or mock treated as control. (a) Western blots confirming siRNA-mediated gene knockdown. GAPDH is presented as a loading control. Lysates from CHK1-knocked down cells were run at a distant location on the same gel. (b) Fluorescence immuno-staining for cyclin B1 (green) and DAPI staining for DNA (blue) in U2OS- LacZÀ cells one day after the administration of the CDDP. I-III) Control cells with no chemotherapy are observed at various stages of the cell cycle: (I) Cells at S or G2 phase are indicated by cytoplasmic staining of cyclin B1; (II) Mitotic entry is indicated by nuclear staining of cyclin B1 and condensed chromosomes; (III) Anaphase is indicated by degradation of Cyclin B1 and chromosome segregation. (IV) Cells treated with CDDP arrest at the S or G2 phase (Compare to I). (V) Mitotic catastrophe is indicated by micronucleation (arrow). (c) Percentage of cells in mitotic phase (see B.II and III) and of micronucleated cells (see B.V), when mock treated or treated with chemotherapeutic drugs, and stained as in B. siControl ¼ transfection with non-targeting siRNA; siCHK1 ¼ knockdown of CHK1; siMK2 ¼ knockdown of MK2; siX2 ¼ knockdown of both CHK1 and MK2. (d) Western blots measuring the phosphorylation levels of histone H3 inU2OS-LacZÀ and U2OS-p53À cells one day after the administration of 0, 4 and 8 mM of CDDP. b-catenin is presented as a loading control. Con ¼ transfection with non-targeting siRNA; CHK1 ¼ knockdown of CHK1; V- administration of transfection reagent only; 0-without treatment.

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6152 Abrogation of DNA damage checkpoints drives administration of the drug (data not shown). We then cells to undergo mitosis in the presence of damaged evaluated the short-term survival of cells by staining or improperly replicated DNA. It is expected to result with methylene blue, 3 days after the application of in mitotic catastrophe and cell death due to chemotherapy, a time that allowed the death of disturbed segregation of chromosomes during mitosis damaged cells. DNA damage checkpoints allow the (Vakifahmetoglu et al., 2008). Indeed, in U2OS-p53À repair of DNA damage before mitotic entry and thus cells knocked down for CHK1 and exposed to prevent mitotic catastrophe. Therefore, we expected that chemotherapy, we observed an increased number of knockdown of CHK1, which abrogates the checkpoints cells displaying polynucleation, a characteristic of in p53-deficient but not in p53-proficient cells, would mitotic catastrophe (Figures 2b V and c). increase cell death following DNA damage specifically in p53-deficient cells. As shown in Figure 3a, there was no significant difference in the survival of p53-deficient Checkpoint abrogation by inactivation of CHK1 does and p53-proficient cells following transfection with non- not reduce cell survival targeting siRNA as control and treatment with irinote- We next examined whether abrogation of the S and can. p53-deficient cells were slightly more sensitive than S/G2 checkpoints by CHK1 knockdown is accompanied p53-proficient cells to CDDP. Knockdown of MK2 by a decline in cell viability following chemotherapy. did not differentiate between p53-proficient and p53- Using time-lapse imaging, we observed that following deficient cells. Contrary to our expectation, following the administration of CDDP, U2OS-LacZÀ cells under- chemotherapy, the knockdown of CHK1 caused similar went cell cycle arrest, indicated by the absence of reductions in the viability of both cell lines, with no dividing cells for at least 48 h. This was accompanied by preferential sensitization of the p53-deficient cells. In cell death, most of which occurred within 48 h of the both p53-proficient and p53-deficient cell lines, knock-

No chemotherapy Cisplatin (10μM) Irinotecan (10μM) 100% U2OS-LacZ- 80% U2OS-p53- 60%

40% Viability 20%

0% siRNA: Control MK2 CHK1 x2 Control MK2 CHK1 x2 Control MK2 CHK1 x2

0 V siCon siCHK1

100 U2OS-LacZ- U2OS-p53- 80

Viability (%) 40

0 0 1.25 2.5 0 1.25 2.5 0 1.25 2.5 0 1.25 2.5 Cisplatin (μM)

U2OS: -p53- -LacZ- siRNA: - V CHK1 Con - V CHK1 Con CHK1 βCatenin

Figure 3 Survival of p53-proficient and deficient cells following chemotherapy and knockdown for CHK1 or MK2: (a) Cells treated as in Figure 1 were plated in 96-well plate (800 cells/well) and grown for 3 days, then assayed for viability with methylene blue. Viable cells are presented as percent of control, untreated cells. (b) p53-proficient U2OS-LacZÀ and p53-deficient U2OS-p53À cells were transfected with siRNA against CHK1, non-targeting siRNA (siCon) or transfection reagent only (V) as indicated. 48 h later the cells were seeded for clonogenic assay in 6-well plates (3000 cells per well) and 24 h later treated with 1.25 and 2.5 mM of CDDP for 24 h. 10 days later the cells assayed for viability with methylene blue. Viable cells are presented as percent of control, untreated cells. (c) Western blots confirming siRNA-mediated gene knockdown in the clonogenic assay. b-catenin is presented as a loading control.

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6153 down of CHK1 significantly reduced cell survival, was maximal at 8 h, as indicated by accumulation of regardless of whether chemotherapy was applied, cdc25a in the cells and reduced after 16 h and 24 h. The suggesting that CHK1 is required for full viability of reduction of cdc25a might be accounted for activation these cells. In addition, similarly to the effect on cell of overlapping regulatory mechanisms such as CHK2 cycle arrest, the combined CHK1 and MK2 knockdown phosphorylation (Falck et al., 2001). had a smaller effect than the knockdown of CHK1 We further confirmed our results by using another alone, again suggesting an antagonistic effect of MK2 isogenic pair of p53-proficient (HCT116 p53 þ / þ )and knockdown on CHK1 knockdown. p53-null (HCT116 p53À/À) cell lines (Bunz et al., 1998). A long-term survival clonogenic assay corroborated Time-course analysis of CHK1 inhibition with the this result. Knockdown of CHK1 decreased the survival selective CHK1 inhibitor SB-218078 revealed that wild of U2OS-LacZÀ and U2OS-p53Àcells under CDDP type HCT116 p53 þ / þ and HCT116 p53À/À cells survived treatment to similar extents (Figure 3b). to similar extents under CDDP treatment as measured To confirm these findings, we used the non-selective by clonogenic assay (Figures 5 b and c). CHK1 inhibitor UCN-01, which had been previously shown to abrogate the S and G2/M checkpoints Inactivation of CHK1 induces DNA damage and induce mitotic catastrophe in p53-deficient cells Our results show that inhibition of CHK1 activity (Wang et al., 1996; Shao et al., 1997; Hirose et al., 2001; results in a decline in mitosis even without chemo- Tse and Schwartz 2004; Chen et al., 2006; Levesque therapy in U20S-lacZ– cells, but not in U20S-p53À cells et al., 2008). Like siRNA-mediated knockdown of (Figures 2c and d). This cell cycle arrest might be CHK1, UCN-01 by itself reduced cell viability because of checkpoint activation by p53, suggesting that even without chemotherapeutics (Figures 4a and b). inhibition of CHK1 by itself induces DNA damage. We Similarly to the siRNA-mediated knockdown of CHK1, examined this by western blot using an antibody against we found no difference between the survival of the p53-deficient and p53-proficient cells following com- the phosphorylated form of histone H2A.X, (g-H2A.X), a marker for DNA damage (Rogakou et al., 1998). bined treatment with UCN-01 and chemotherapeutics. siRNA was used to knock down CHK1, CHK2, MK2, When cells were treated with any one of a variety of chemotherapeutic drugs, application of UCN-01 or all three together, in both p53-deficient and p53- proficient U2OS cell lines (Figure 6a). As expected, reduced long-term survival by an extent similar to the CDDP or irinotecan induced g-H2A.X accumulation in reduction observed without chemotherapy. We did not both cell lines (Figure 6b). Knockdown of either CHK2 observe the expected significant cooperation between or MK2 had little or no effect on the amount of g- CHK1 inhibition and chemotherapy in the p53-deficient H2A.X. In contrast, knockdown of CHK1 induced cells. Similar results were obtained using another pair of considerable accumulation of g-H2A.X, irrespective of p53-proficient and p53-deficient isogenic cell lines: Control, stably transfected with an empty vector as whether chemotherapeutics were applied, indicating that CHK1 inhibition causes DNA damage (Figure 6b). control and MCF7-p53–, which stably expresses siRNA Similar results were obtained using p53-proficient and against p53 (Brummelkamp et al., 2002). In both cell lines, concurrent application of UCN-01 with che- p53-deficient MCF7-derived cells. These cells were knocked down for CHK1, MK2 or both together, as motherapeutics resulted in reduced cell viability com- verified by western blot (Figure 6c and data not shown). pared with cells treated with chemotherapeutic agents As for U20S-derived cells (Figure 2b), knockdown of alone, with no significant difference between CHK1 in the p53-deficient MCF7 cell line abrogated the p53-proficient and the p53-deficient cell lines the cell cycle arrest induced by CDDP. This can be seen (Figure 4c). MCF7-p53À cells were more sensitive to by the depletion in the inactive, phosphorylated form irinotecan than MCF7-Control, and concurrent treat- of Cdc2 (Figure 6c); Dephosphorylation of Cdc2 is ment with UCN-01 caused a further reduction in cell required for cell cycle progression. Transfection of viability, to similar extents in both cell lines. non-targeting siRNA slightly elevated the amount of Tse and Schwartz (2004) reported that application of UCN-01 after the topoisomerase I inhibitor SN38 g-H2A.X, indicating that the transfection procedure causes some DNA damage. Knockdown of CHK1, but caused more cell death by mitotic catastrophe in p53- not of MK2, resulted in a substantial increase in the deficient HCT116 cells, than did their concurrent administration (Hirose et al., 2001). We, therefore, amount of g-H2A.X (Figure 6c). Again, knockdown of MK2 antagonized the effects of CHK1 knockdown. examined whether application of UCN-01 subsequent to These results indicate that ablation of CHK1 leads to chemotherapy might achieve selective sensitization of DNA damage. This DNA damage might account for the p53-deficient cells to chemotherapy. The sequential cell death caused by CHK1 inhibition. treatment did not enhance the death of either p53- deficient cells or p53-proficient cells (Figure 4c). Application of the CHK1 selective inhibitor SB-218078 to U2OS cells corroborated these results. Discussion Time-course inhibition of CHK1 decreased the survival of U2OS-LacZÀ and U2OS-p53À cells under CDDP In this study, we examined whether the inhibition of treatment to similar extents as measured by clonogenic CHK1 or MK2 or of both can be utilized for the assay (Figures 5a and c). Interestingly, the inhibition selective sensitization of p53-deficient tumor cells to

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6154 - 100% U2OS-LacZ U2OSU2OS-p53- 80%

60%

Viability 40%

20%

0% UCN-01: - - + + - - + + - - + + - - + + - - + + - - + + - - + + - - + + (50nM) No Cisplatin 5-fluorouracil Irinotecan Methotrexate Doxorubicin Etoposide Taxol (5nM) chemotherapy (10μM) (5nM) (10μM) (2μM) (10nM) (2μM)

U2OS-LacZ- U2OS-p53- 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% Percentage of cells

0% Percentage of cells 0% 0105 0 5 10 Etoposide (μM) Etoposide (μM)

U2OS-LacZ- U2OS-p53- 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% Percentage of cells Percentage of cells 0% 0% 02010 010 20 Cisplatin (μM) Cisplatin (μM) No UCN-01 100nM UCN-01

MCF7-LacZ- 100% MCF7-p53- 80% 60%

viability 40% 20% 0% UCN-01: --C C S S --C C S S --C C S S (100nM) No chemotherapy Cisplatin (3.3μM) Irinotecan (11.7μM) Figure 4 The effect of UCN-01 on cell survival: (a) U2OS-LacZÀ and U2OS-p53À cells were plated in 96-well plates, and 24 h later were treated with the indicated chemotherapeutic drugs along with 50 nM UCN-01. After 3 days, the cells were assayed for cell viability with methylene blue. (b) Dose response of U2OS-LacZ- and U2OS-p53À cells to 48-h treatment with etoposide or CDDP, with or without UCN-01 (100 nM). Viability was assayed with methylene blue. (c) p53-proﬁcient MCF7-Control and p53-deﬁcient MCF7-p53À cells were treated with the indicated chemotherapeutic drugs. For concurrent treatment (C), UCN-01 (100 nM) was added at the same time as the chemotherapeutic drugs, 24 h after plating the cells. For sequential treatment (S), chemotherapeutic drugs were added 24 h after plating the cells and UCN-01 (100 nM) was added 24 h later. For control cells without chemotherapy, UCN-01 was added at the same time as for cells with chemotherapy, that is 24 h (C) or 48 h (S) after plating. Cell viability was assayed using methylene blue, 3 days after the administration of the chemotherapeutic drugs.

chemotherapy. Using siRNA mediated gene knockdown knockdown of MK2 antagonizes this effect (Xiao et al., in two pairs of isogenic cell lines differing only in the 2006). Both our results and those of Xiao et al. differ expression of p53, we showed that in the absence of p53, from those obtained for ras-transformed murine em- the cells are dependent on CHK1, but not on MK2, for bryonic fibroblasts, in which knockdown of p53 makes checkpoint activation and cell cycle arrest following these cells dependent on MK2 for activation of the G2 administration of CDDP or irinotecan. Moreover, we checkpoint (Reinhardt et al., 2007). These differences found that MK2 knockdown reduces the effects of might stem from the fact that both our studies and those CHK1 knockdown on cell cycle arrest, cell survival and of Xiao et al. used human cancer cells, whereas the formation of spontaneous DNA damage. These Reinhardt et al. (2007) used ras transformed mouse results confirm data from Xiao et al. (2006), which embryonic fibroblasts. showed that knockdown of CHK1 abrogated cell cycle Selective abrogation of the checkpoints in p53- arrest induced by camptothecin or doxorubicin in HeLa deficient cells, by inhibition of CHK1, was expected to and H1299 cell lines, which are p53-deficient, whereas sensitize cells to DNA damage, whereas the unperturbed

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6155 5 M SB-218078 U2OS-LacZ- 140 U2OS-LacZ- CDDP 1.25 μM U2OS-LacZ- CDDP 2.5 μM 120 U2OS-p53- - 100 U2OS-p53 CDDP 1.25 µM U2OS-p53- CDDP 2.5 µM 80

60 Viability (%) 40

0 0 4 8 16 24 (hours)

HCT116 p53+/+ HCT116 p53+/+ CDDP 1.25 μM HCT116 p53+/+ CDDP 2.5 μM 5 M SB-218078 HCT116 p53-/- 120 HCT116 p53-/- CDDP 1.25 μM HCT116 p53-/- CDDP 2.5 μM 100

Viability (%) 40

0 0 4 8 16 24 (hours)

U2OS: p53- LacZ- HCT116: p53-/- p53+/+

5 μM SB (hours): 0 4 8 16 24 0 4 8 16 24 5 μM SB (hours): 0481624 0481624 cdc25a cdc25a

βCatenin βCatenin

Figure 5 The effect of SB-218078 on cell viability: p53-proficient U2OS-LacZÀ and p53-deficient U2OS-p53À cells (a) and HCT116 carcinoma cells carrying wild type p53 (p53 þ / þ ) and p53 null (p53À/À)(b) were mock treated (for the 0 h treatment) or supplemented with 5 mM SB-218078. 1 h later CDDP was added at 1.25 and 2.5 mM for 0 h, 4 h, 8 h, 16 h and 24 h. Then the cells were plated in 6 well plates (800 cells per well) for clonogenic assay. 14 days later (for the U2OS cells), or 7 days later (for the HCT116 cells), the cells were assayed for viability with methylene blue. All the colonies were stained regardless of colony size. Viable cells are presented as percent of control, untreated cells. (c) Western blots confirming the CHK1 inhibition. Cdc25a accumulation indicates the inhibition of CHK1. b-catenin is presented as a loading control.

p53-dependent checkpoint(s) in p53-proficient cells were status of p53 (Figure 3). Other studies found, as we did, expected to allow DNA repair and thereby to prevent that the effect of UCN-01 and SB-218078 on cell the sensitization of normal cells to DNA damage (Zhou survival was not dependent on the status of p53 (Hirose et al., 2003; Kawabe, 2004; Zhou and Bartek, 2004; et al., 2001; Tse and Schwartz 2004; Levesque et al., Bucher and Britten, 2008; Lieberman, 2008). As 2008). These studies employed several p53-active and predicted by this model, we observed (Figure 2c) that p53-inactive cancer cell lines. To explain these findings, knockdown of CHK1 in p53-depleted cells induced it was suggested that the p53 proficient cell lines that mitotic catastrophe and polynucleation, which results were used had defective p53 signaling, which could not from mitotic entry of cells with damaged DNA properly activate cell cycle arrest, so that, when CHK1 (Vakifahmetoglu et al., 2008). However, unexpectedly, was inhibited, these cells underwent aberrant mitosis we found that short-term cell survival, as well as the without cytokinesis and died like p53-deficient cells long-term ability to form colonies under CHK1 deple- (Levesque et al., 2008; Zhang et al., 2008). In our study, tion was similar in the isogenic cell lines regardless of the however, the p53-proficient U2OS cells had intact p53

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6156

U2OS-LacZ- U2OS-p53-

Control Control

CHK1 CHK2 CHK2 CHK1

MK2 MK2 X3 siRNA : X3

CHK1

CHK2

MK2

GAPDH

Cisplatin Irinotecan No chemotherapy Control Control Control CHK1 CHK2 CHK1 CHK2 CHK1 CHK2 MK2 MK2 MK2 X3 X3

siRNA: X3

U2OS γ -H2A.X -LacZ- GAPDH

U2OS γ -H2A.X -p53- GAPDH

MCF7-p53- MCF7 MCF7 - - -LacZ -p53 siRNA : CHK1 MK2 X2 Control N p53 CHK1

GAPDH MK2

γ No -H2A.X chemotherapy GAPDH

γ -H2A.X Cisplatin (15μM) CDC2 Y15P

GAPDH

Figure 6 Measurement of DNA damage: (a, b) U2OS-LacZÀ and U2OS-p53À cells were transfected with siRNA against CHK1, CHK2, MK2, all three (X3) or non-targeting siRNA (Control). 24 h later the cells were treated with the chemotherapeutic drugs CDDP or irinotecan at 10 mM.(a) Western blots confirming the gene knockdowns. (b) Western blots showing the phosphorylated form of histone H2A.X (g-H2A.X) as a marker for DNA damage, and GAPDH as a loading control. (c) MCF7-Control and MCF7-p53 cells transfected with the indicated siRNA against CHK1, MK2, both together (X2), or non-targeting siRNA (Control), as well as non- transfected cells (N). One day post transfection cells were treated with CDDP (15 mM) for additional 24 h. Left: Western blot confirming depletion of p53 in MCF7-p53 cells compared with MCF7-LacZ cells treated with 10 uM CDDP. Right: Western blots confirming knockdowns of CHK1 or MK2; g-H2A.X as a marker for DNA damage indicating induction of DNA damage by knockdown of CHK1; Phosphorylation of CDC2 at tyrosine 15 as an indication of cell cycle arrest; GAPDH as a loading control.

signaling (Figure 1), were arrested following DNA reported to potentiate the toxicity of DNA damaging damage and did not enter mitosis even in the absence agents (Wang et al., 1996; Shao et al., 1997; Hirose et al., of CHK1 (Figures 2c and d). 2001; Tse and Schwartz 2004; Chen et al., 2006; Contrary to these results, other groups reported that Levesque et al., 2008). We observed significant toxicity UCN-01 and other CHK1 inhibitors, when combined caused by CHK1 inhibition, which appeared to be with DNA damaging agents, reduced the efficacy of additive to the toxicity of the chemotherapeutic drugs. colony formation of p53-inactive cell lines more CHK1 toxicity may be more pronounced in the cell lines drastically than that of their isogenic p53-active cell that we used in this study. Our data showed that CHK1 lines (Wang et al., 1996; Shao et al., 1997; Chen et al., inhibition reduces viability regardless of the status of 2006). In these and other studies, UCN-01 and other p53 and does not cooperate with chemotherapeutic CHK1 inhibitors, as well as CHK1 knockdown, were drugs to selectively reduce viability of p53 deficient cells.

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6157 The question remains why checkpoint abrogation by cancer cells. Inactivation of CHK1 abrogates cell cycle CHK1 inhibition, which is selective for p53-deficient arrest following chemotherapy in p53-deficient cells but cells, does not reduce their short and long-term not in p53-proficient cells and this checkpoint abrogation survival compared with p53-proficient cells following results in cell death by mitotic catastrophe. However, p53- DNA damage. In both cell lines studied here, knock- deficient cells are not more sensitive than p53-proficient down of CHK1 led to a reduction in cell viability. cells to the combination of CHK1 inhibition and chemo- This was accompanied by DNA damage in all cell therapeutic agents. CHK1 is necessary for checkpoint lines even without chemotherapeutic drugs and by a activation and consequent repair in p53-deficient cells, reduction in the number of mitotic entries in p53- but is also necessary for the recovery of p53-proficient proficient cells, as expected to the event of DNA cells following cell cycle arrest. Depletion of p53 changes damage. These findings suggest that in addition to its the mechanism of cell death, but does not cooperate with role in checkpoint activation, CHK1 is involved in CHK1 inhibition to increase the degree of cell death. maintenance of DNA integrity and is indispensable Most notably, inhibition of CHK1 does not cooperate for cell survival. Supportive of this view, Cho et al. with DNA damaging agents. Rather, CHK1 is necessary (2005) found that CHK1 is required for the resumption for cell viability and recovery from checkpoint arrest of mitosis after checkpoint activation by hydroxyurea. and CHK1 inhibition causes additional DNA damage, This requirement for CHK1 to re-enter the cell cycle which might provide CHK1 inhibitors with non- might reflect its role in the stabilization of stalled selective therapeutic activity, like other DNA damaging replication forks during the S phase arrest. In the agents. Indeed, clinical trials have revealed that UCN-01 absence of CHK1, replication forks collapse and has some therapeutic effect when not combined with replication cannot be resumed correctly, causing spon- chemotherapy (Senderowicz, 2003). We conclude that taneous DNA damage and cell death (Cho et al., 2005; CHK1 inhibitors cannot selectively target cancer cells, Petermann and Caldecott, 2006). Thus, when CHK1 is based on their deficiency in p53. This conclusion is inhibited and DNA is damaged, p53-proficient cells are mirrored by the view of radiation oncologists that the not more viable than p53-depleted cells: p53-deficient status of p53 does not correlate with the clinical outcome cells cannot activate checkpoints in the absence of of radiation therapy and chemotherapy (Finkel, 1999). CHK1 and die by mitotic catastrophe. p53-proficient cells can indeed activate DNA-damage checkpoints through the p53 pathway, but cannot re-enter the cell Materials and methods cycle in the absence of CHK1. Therefore, both p53- proficient and p53-deficient cells eventually die. Supportive Cell culture and transfections of this suggestion, it has been reported that when treated All cells were maintained at 37 1C in a 5% CO2 humidified with UCN-01, p53-proficient U87MG and HCT116 cells incubator. MCF7 human breast cancer cells, stably transfected undergo senescence following DNA damage, which indi- with empty pSUPER vector (MCF7-Control) or with short cates their exit from the cell cycle (Hirose et al., 2001; Tse hairpin RNA against p53 (MCF7-p53À) (Brummelkamp et al., and Schwartz, 2004). A recent study by Tao et al. (2009) 2002), were a gift from R Agami (Amsterdam, The Netherlands). that examined the response of HCT116 colon carcinoma U2OS human osteosarcoma cells stably expressing short hairpin cells to the selective CHK1 inhibitor Chir-124 following RNA against lacZ (U2OS-LacZÀ)orp53(U2OS-p53À) (Raver- ionizing radiation (Tao et al., 2009) confirmed the results Shapira et al., 2007) were kindly provided by Professor M Oren of our study. These authors showed that the effect of (Weizmann Institute, Israel). These cells were grown as monolayers in Dulbecco’s modified Eagle’s medium (DMEM; Biolo- Chir-124 radiosensitization was similar in p53-deficient gical Industries, Kibbutz Beit Haemek, Israel) with 10% fetal and p53-proficient cells and operate through distinct bovine serum (FBS, Sigma-Aldrich, St Louis, MO, USA) and 1% pathways, namely mitotic catastrophe in p53-deficient penicillin-streptomycin (Biological Industries), supplemented with and another pathway in p53-proficient cells (Tao et al., 1 mg/ml puromycin (Sigma-Aldrich) for the MCF7-derived cells, 2009), similarly to our observations. or 2 mg/ml puromycin for the U2OS-derived cells. HCT116- In our study, the p53-proficient cells underwent p53 þ / þ and the p53-null derivative HCT116-p53À/À (Bunz et al., caspase-3-dependent apoptosis, whereas the p53-depleted 1998) were kindly provided by Professor B Vogelstein (Johns cells underwent caspase independent cell death (Figure 1). Hopkins Kimmel Cancer Center, Baltimore). These cells were These results suggest that the status of p53 might determine grown as monolayers in McCoy’s medium (Biological Indus- the mechanism of cell death and survival, but does not tries) with 10% fetal bovine serum (FBS, Sigma-Aldrich) and 1% penicillin-streptomycin (Biological Industries). necessarily affect the degree of cell survival under CHK1 Transfection of siRNA (siGENOME SMARTpool, Dharmacon, inhibition. It has been shown earlier that inhibition of hub Lafayette, CO, USA) was performedwithDharmaFECT1reagent proteins may change the mechanismofcelldeath,butis according to the manufacturer’s instructions. not necessarily coupled to the extent of cell death, upon application of genotoxic stress (Kravchenko-Balasha et al., 2009). In the current study, although CHK1 inhibition Western blotting and antibodies Cells were washed with phosphate buffered saline (PBS) and selectively abrogates cell cycle arrest following chemo- harvested by scraping in boiling sample buffer. Total cell extracts therapy only in p53-deficient cells, the degree of cell (40 mg protein) were resolved on SDS-10% polyacrylamide gel, death and viability is not affected by p53 status. electroblotted to a nitrocellulose membrane (Watman, Protran, In conclusion, CHK1, but not MK2, appears to be Maidstone, Kent, UK) and reacted with the appropriate anti- the gatekeeper of checkpoint activation in p53-deficient bodies. Bands were visualized by ECL chemiluminescence.

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6158 The following antibodies were used: Rabbit anti-MK2 (#3042), Cell survival assay CHK1 (#2345), P15Y-cdc2 (#9111), GAPDH-14C10 (# 2118) or Cells were plated in 96-well plates at 3000 cells/well. After 1 phospho-Histone H3 (Thr3) (#9714) (Cell Signaling Technology); day, drugs were applied as indicated. After 3 additional days, Mouse anti-cdc25a (#7389) (Santa Cruz Biotechnology, Santa the methylene blue assay was performed (Dent et al., 1995). Cruz, CA, USA), Mouse anti-g-histone H2A.X (Upstate Biotechno- Briefly, cells were fixed with 0.5% glutaraldehyde, stained with logy, Lake Placid, NY, USA); Rabbit anti-CHK2 (#8108) (Abcam, 1% methylene blue in 0.01 M borate buffer (pH 8.5), washed Cambridge, MA, USA) Rabbit anti-p53-FL393 (#6243) (Santa and dried and then incubated with 100 mlof1N HCl for 1 h at Cruz Biotechnology); Fluorescein (FITC)-conjugated species- 37 1C. Optical density was measured with an ELISA plate specific Donkey anti-Mouse, peroxidase-conjugated AffinPure reader at 630 nm. Donkey anti-Mouse and peroxidase-conjugated AffinPure Goat anti-Rabbit (Jackson ImmunoResearch Laboratories Inc., West Clonogenic assay Grove, PA, USA). Cells were transfected as indicated and 48 h later were plated in 6-well plates at 800 cells/well. After 1 day, CDDP was applied Immunofluorescence as indicated. After 24 h, the medium was replaced by fresh Cells growing on 12 mm coverslips were fixed with ice-cold medium and after 10 additional days, the methylene blue methanol for 10 min. After rehydration with PBS, cells were assay was performed. All colonies were stained regardless of permeabilized with 0.5% Triton X-100 in PBS, washed with colony size. PBS and then blocked with 5% bovine serum albumin in PBST (PBS þ 0.05% Tween 20). Samples were incubated for 1 h with primary antibody, washed and incubated with secondary anti- Conflict of interest body, and 40,6-diamidino-2-phenylindole for 30 min in the dark, then washed and sealed with Fluorescent Mounting Medium (DakoCytomation, Glostrup, Denmark) and visualized with an The authors declare no conflict of interest. Olympus IX70 inverted microscope. In each treatment, 80–300 cells from at least two separate fields were examined. Acknowledgements p53 activity assay Cells were plated in 6-well plates at 2800 cells/well. After 1 day, We would like to thank Dr Shoshana Klein for discussions and Nutlin and CDDP were applied as indicated for 24 h, and then editing. This work was partially supported by the Lady Davis cell lysate was taken for western blot analysis. Fellowship Trust and by Algen Biopharmaceuticals, Ltd.

References

Abraham RT. (2001). Cell cycle checkpoint signaling through the Finkel E. (1999). Does cancer therapy trigger cell suicide? Science 286: ATM and ATR kinases. Genes Dev 15: 2177–2196. 2256–2258. Ashwell S, Janetka JW, Zabludoff S. (2008). Keeping checkpoint Harris SL, Levine AJ. (2005). The p53 pathway: positive and negative kinases in line: new selective inhibitors in clinical trials. Expert Opin feedback loops. Oncogene 24: 2899–2908. Investig Drugs 17: 1331–1340. Hirose Y, Berger MS, Pieper RO. (2001). Abrogation of the Chk1- Bartek J, Lukas J. (2003). Chk1 and Chk2 kinases in checkpoint mediated G(2) checkpoint pathway potentiates temozolomide- control and cancer. GGGGZ 3: 421–429. induced toxicity in a p53-independent manner in human glioblas- Brummelkamp TR, Bernards R, Agami R. (2002). A system for stable toma cells. Cancer Res 61: 5843–5849. expression of short interfering RNAs in mammalian cells. Science Karlsson-Rosenthal C, Millar JB. (2006). Cdc25: mechanisms 296: 550–553. of checkpoint inhibition and recovery. Trends Cell Biol 16: Bucher N, Britten CD. (2008). G2 checkpoint abrogation and 285–292. checkpoint kinase-1 targeting in the treatment of cancer. Br J Kawabe T. (2004). G2 checkpoint abrogators as anticancer drugs. Mol Cancer 98: 523–528. Cancer Ther 3: 513–519. Bunz F, Dutriaux A, Lengauer C, Waldman T, Zhou S, Brown JP Kravchenko-Balasha N, Mizrachy-Schwartz S, Klein S, Levitzki A. et al. (1998). Requirement for p53 and p21 to Sustain G2 Arrest (2009). Shift from apoptotic to necrotic cell death during human After DNA Damage. Science 282: 1497–1501. papillomavirus-induced transformation of keratinocytes. J Biol Chen Z, Xiao Z, Gu WZ, Xue J, Bui MH, Kovar P et al. (2006). Chem 284: 11717–11727. Selective Chk1 inhibitors differentially sensitize p53-deﬁcient cancer Levesque AA, Fanous AA, Poh A, Eastman A. (2008). Defective p53 cells to cancer therapeutics. Int J Cancer 119: 2784–2794. signaling in p53 wild-type tumors attenuates p21waf1 induction and Cho SH, Toouli CD, Fujii GH, Crain C, Parry D. (2005). Chk1 is cyclin B repression rendering them sensitive to Chk1 inhibitors that essential for tumor cell viability following activation of the abrogate DNA damage-induced S and G2 arrest. Mol Cancer Ther replication checkpoint. Cell Cycle 4: 131–139. 7: 252–262. Dai J, Sultan S, Taylor SS, Higgins JM. (2005). The kinase haspin is Lieberman HB. (2008). DNA damage repair and response proteins as required for mitotic histone H3 Thr 3 phosphorylation and normal targets for cancer therapy. Curr Med Chem 15: 360–367. metaphase chromosome alignment. Genes Dev 19: 472–488. Manke IA, Nguyen A, Lim D, Stewart MQ, Elia AE, Yaffe MB. Dent MF, Hubbold L, Radford H, Wilson AP. (1995). The methylene (2005). MAPKAP kinase-2 is a cell cycle checkpoint kinase that blue colorimetric microassay for determining cell line response to regulates the G2/M transition and S phase progression in response growth factors. Cytotechnology 17: 27–33. to UV irradiation. Mol Cell 17: 37–48. Donzelli M, Draetta GF. (2003). Regulating mammalian checkpoints Meek DW. (2004). The p53 response to DNA damage. DNA Repair through Cdc25 inactivation. EMBO Rep 4: 671–677. (Amst) 3: 1049–1056. Falck J, Mailand N, Syljua˚sen RG, Bartek J, Lukas J. (2001). The O’Connor MJ, Martin NM, Smith GC. (2007). Targeted cancer ATM–Chk2–Cdc25A checkpoint pathway guards against radio- therapies based on the inhibition of DNA strand break repair. resistant DNA synthesis. Nature 410: 842–847. Oncogene 26: 7816–7824.

Oncogene p53, CHK1 and chemotherapy S Zenvirt et al 6159 Oren M. (2003). Decision making by p53: life, death and cancer. Cell Tse AN, Schwartz GK. (2004). Potentiation of cytotoxicity of Death Differ 10: 431–442. topoisomerase i poison by concurrent and sequential treatment Petermann E, Caldecott KW. (2006). Evidence that the ATR/Chk1 with the checkpoint inhibitor UCN-01 involves disparate mechan- pathway maintains normal replication fork progression during isms resulting in either p53-independent clonogenic suppression or unperturbed S phase. Cell Cycle 5: 2203–2209. p53-dependent mitotic catastrophe. Cancer Res 64: 6635–6644. Raver-Shapira N, Marciano E, Meiri E, Spector Y, Rosenfeld N, Vakifahmetoglu H, Olsson M, Zhivotovsky B. (2008). Death through a Moskovits N et al. (2007). Transcriptional activation of miR-34a tragedy: mitotic catastrophe. Cell Death Differ 15: 1153–1162. contributes to p53-mediated apoptosis. Mol Cell 26: 731–743. Vogelstein B, Lane D, Levine AJ. (2000). Surfing the p53 network. Reinhardt HC, Aslanian AS, Lees JA, Yaffe MB. (2007). p53-deficient Nature 408: 307–310. cells rely on ATM- and ATR-mediated checkpoint signaling Wang Q, Fan S, Eastman A, Worland PJ, Sausville EA, through the p38MAPK/MK2 pathway for survival after DNA O’Connor PM. (1996). UCN-01: a potent abrogator of G2 damage. Cancer Cell 11: 175–189. checkpoint function in cancer cells with disrupted p53. J Natl Rogakou EP, Pilch DR, Orr AH, Ivanova VS, Bonner WM. (1998). Cancer Inst 88: 956–965. DNA double-stranded breaks induce histone H2AX phosphoryla- Xiao Z, Xue J, Sowin TJ, Zhang H. (2006). Differential roles of tion on serine 139. J Biol Chem 273: 5858–5868. checkpoint kinase 1, checkpoint kinase 2, and mitogen-activated Senderowicz AM. (2003). Novel direct and indirect cyclin-dependent protein kinase-activated protein kinase 2 in mediating DNA kinase modulators for the prevention and treatment of human damage-induced cell cycle arrest: implications for cancer therapy. neoplasms. Cancer Chemother Pharmacol 52(Suppl 1): S61–S73. Mol Cancer Ther 5: 1935–1943. Shao RG, Cao CX, Shimizu T, O’Connor PM, Kohn KW, Pommier Zhang WH, Poh A, Fanous AA, Eastman A. (2008). DNA damage- Y. (1997). Abrogation of an S-phase checkpoint and potentiation of induced S phase arrest in human breast cancer depends on Chk1, camptothecin cytotoxicity by 7-hydroxystaurosporine (UCN-01) in but G2 arrest can occur independently of Chk1, Chk2 or human cancer cell lines, possibly influenced by p53 function. Cancer MAPKAPK2. Cell Cycle 7: 1668–1677. Res 57: 4029–4035. Zhou BB, Elledge SJ. (2000). The DNA damage response: putting Shiloh Y. (2003). ATM and related protein kinases: safeguarding checkpoints in perspective. Nature 408: 433–439. genome integrity. Nat Rev Cancer 3: 155–168. Zhou BB, Anderson HJ, Roberge M. (2003). Targeting DNA Stark GR, Taylor WR. (2006). Control of the G2/M transition. Mol checkpoint kinases in cancer therapy. Cancer Biol Ther 2: Biotechnol 32: 227–248. S16–S22. Tao Y, Leteur C, Yang C, Zhang P, Castedo M, PierreÁ et al. (2009). Zhou BB, Bartek J. (2004). Targeting the checkpoint kinases: Radiosensitization by Chir-124, a selective CHK1 inhibitor: effects chemosensitization versus chemoprotection. Nat Rev Cancer 4: of p53 and cell cycle checkpoints. Cell Cycle 15: 1196–1205. 216–225.

Oncogene