Overexpression of the Dual Specificity Phosphatase, Cdc25c, Confers Sensitivity on Tumor Cells to Doxorubicin- Induced Cell Death

3789

Overexpression of the dual specificity phosphatase, Cdc25C, confers sensitivity on tumor cells to doxorubicin- induced cell death

Shohreh Varmeh and James J. Manfredi checkpoints in response to DNA damage results in either preventing the cells from entering the next phase of the Department of Oncological Sciences, Mount Sinai School of cell cycle to allow repair or inducing cell death if the lesion Medicine, New York, New York is beyond restoration. The two predominant checkpoints include those in the G1 and G2 phases. The former prevents Abstract the cells from entering S phase and replicating damaged Cdc25C is a dual-specificity phosphatase that is involved DNA and the latter inhibits entry into mitosis and in induction of mitosis by removal of the inhibitory phos- propagation of damaged DNA to daughter cells. It has phates from cyclin-dependent kinase 1/cyclin B. In this been proposed that the G1 checkpoint is disabled in the study, adenovirus-mediated overexpression of Cdc25C majority of cancers as a prerequisite for tumorigenesis sensitizes U2OS tumor cells to doxorubicin-induced apop- (2, 3). The G2 checkpoint, however, appears to remain tosis. U2OS cells that stably overexpress Cdc25C are also functional in the majority of cancer cells. Many conven- sensitized to doxorubicin-induced cell death. These cells tional cancer therapeutic agents exert their effects by show reduced phosphorylation of cyclin-dependent kinase causing DNA damage. Thus, retaining a functional G2 is 1onTyr15 and impaired up-regulation of p21 in response believed to confer resistance to many such therapeutic to treatment with doxorubicin. In contrast to doxorubicin, agents. To circumvent this, attempts have been made to overexpression of Cdc25C does not confer sensitivity to develop abrogators of this arrest in the G2 phase (4). apoptosis on treatment with 5-fluorouracil or hydroxyurea. Furthermore, the success of such a strategy would provide This sensitization of tumor cells to doxorubicin-induced cell the fundamental advantage of selectively targeting cancer death by overexpression of Cdc25C is not p53 dependent. cells. Thus, cancer cells unable to arrest at the G2-M border Intriguingly, nontransformed MCF10A cells are not sensi- will die in response to treatment, whereas their normal tized to doxorubicin treatment by overexpression of Cdc25C counterparts will still arrest at the G1 border. nor does the lack of Cdc25C affect cell cycle progression At the molecular level, the lack of a G1 arrest in cancer or the G2 arrest caused by doxorubicin. These results sup- cells is a result of their inability to inhibit cyclin-dependent port the idea that a combination of overexpressing Cdc25C kinase (Cdk)2, the kinase that is considered to be important with treatment with conventional genotoxic agents should for entry into S phase (1, 5). This usually arises as a con- be given serious considerations as a novel therapeutic sequence of a dysfunctional p53 and/or Rb pathway (1, 2). strategy. [Mol Cancer Ther 2008;7(12):3789–99] The target for the G2 checkpoint pathway is Cdk1 (1). Cdk1, once activated, ushers the cell into mitosis and subsequent cell division (5). Activity of Cdk1 is regulated at several Introduction levels. The final step in activation of Cdk1 is removal of the Cancer arises due to abnormalities to DNA. The major pro- inhibitory phosphates from Thr14 and Tyr15. In mammalian tective mechanisms against development of cancer are the cells, there are three related dual-specificity phosphatases, cell cycle checkpoints (1). Activation of the cell cycle Cdc25A, Cdc25B, and Cdc25C, which act to dephosphor- ylate these residues (6). To arrest the cells at the G2-M border, dephosphorylation of Cdk1 on the two inhibitory residues is prevented by inhibiting the activities of these Received 1/21/08; revised 8/29/08; accepted 9/30/08. three phosphatases by various mechanisms (1, 6–13). The Grant support: National Cancer Institute, Department of Health and Human kinase Chk1 is a major regulator of the G2 checkpoint, Services grant CA086001 and Breast Cancer Research Program, Department of Defense grant W81XWH-05-1-0305. as it is capable of inactivating all three phosphatases via The costs of publication of this article were defrayed in part by the their phosphorylation (1). payment of page charges. This article must therefore be hereby marked Many of the players in the G2 checkpoint pathway advertisement in accordance with 18 U.S.C. Section 1734 solely to including Chk1 have been proposed or tested as means to indicate this fact. abrogate the G arrest (4). In addition, other approaches Note: Current address: Cancer Genetics Program, Beth Israel Deaconess 2 Cancer Center, Harvard Medical School, 77 Avenue Louis Pasteur, Boston, have included drugs such as caffeine and UCN-01 (4). MA 02115. The major problems with these candidates have been their Requests for reprints: James J. Manfredi, Department of Oncological involvements in other cellular processes leading to un- Sciences, Mount Sinai School of Medicine, Box 1130, New York, NY 10029. Phone: 212-659-5495; Fax: 212-849-2446. wanted toxicities. A recent study described a peptide, E-mail: [email protected] CBP501, which prevents phosphorylation of Cdc25C on 216 Copyright C 2008 American Association for Cancer Research. Ser in response to DNA damage and thus abrogates G2 doi:10.1158/1535-7163.MCT-08-0838 arrest (14). Phosphorylation of this site is required for

Mol Cancer Ther 2008;7(12). December 2008

retaining Cdc25C in the cytoplasm and thereby preventing Immunoblotting activation of Cdk1 when DNA is damaged (11, 12). CBP501 Cells were lysed in a buffer containing 50 mmol/L HEPES, selectively enhanced sensitivity of cancer cells to genotoxic 150 mmol/L NaCl, 1 mmol/L EDTA, 1% Triton X-100, cancer therapy drugs in the absence of side effects (14). 0.1 mmol/L Na3VO4, 10 mmol/L NaF, 80 mmol/L gly- Thus, Cdc25C is suggested to be a useful target when cerophosphate, 1 mmol/L phenylmethylsulfonyl fluoride, considering strategies for abrogating G2 arrest. Here, it is 20 units/mL aprotinin, 5 mg/mL leupeptin, and 5 mg/mL proposed that overexpression of Cdc25C may be another pepstatin. Lysates were clarified by centrifugation (10,000 Â g means to abrogate the G2 checkpoint. Overexpression for 10 min at 4jC)and immunoblotted. The following or lack of Cdc25C does not affect progression through the antibodies were used: anti-FLAG M2 was purchased from normal cell cycle in different cell lines in short-term or Sigma. Anti-Cdc25A (F-6), anti-Cdc25C (H-6), anti-Cdk1 long-term assays. Loss of Cdc25C also does not affect the (C-19), and anti-pT14/Y15Cdk1 were from Santa Cruz G2 arrest response to doxorubicin, a widely used anticancer Biotechnology. Anti-poly(ADP-ribose)polymerase was from drug that exerts its effects by inducing damage to genomic BP Pharmicon. Antibody to h-actin was from Oncogene DNA by a variety of proposed mechanisms, notably Research Products. Peroxidase-conjugated goat antibodies inhibition of topoisomerase II (15). Nevertheless, excess against rabbit or mouse IgG (MP Biomedicals)were used as amount of Cdc25C converts the G2 arrest induced by secondary antibodies. The signals were detected using doxorubicin to cell death in several tumor cell lines. In enhanced chemiluminescence reagents (Amersham Bioscien- contrast, overexpression of Cdc25C does not interfere with ces)and autoradiography films (Labscientific). arrests in other phases of the cell cycle caused by treatment Immunostaining with either 5-fluorouracil (5-FU)or hydroxyurea. Cells were grown on coverslips (Fisher Scientific). After fixation in 3% paraformaldehyde, cells were washed in PBS and permeabilized in a solution containing 0.5% Triton X-100, Materials and Methods 3 mmol/L MgCl2,and0.2mmol/Lsucrose.Afterwashing Cell Cultures and DrugTreatments in PBS, cells were blocked in 3% bovine serum albumin in PBS À/À Human U2OS, HCT116-p53 , and H1299 cells were for 10 min before incubation with the same buffer but grown as monolayers in DMEM (Life Technologies/ containing a 1:1,000 dilution of mouse anti-FLAG antibody Invitrogen)supplemented with 10% FCS (Life Technolo- (Upstate, Cell Signaling Solutions)for 1 h. The cells were then gies/Invitrogen)and 100 units/mL penicillin/streptomycin washed in PBS and subsequently incubated with 1:2,000 dilu- j (Life Technologies/Invitrogen)at 5% CO 2 and 37 C. tion of Alexa Fluor 545 goat anti-mouse IgG (Molecular Probes, MCF10A cells were grown in DMEM/F12 with 15 mmol/L Invitrogen Detection Technologies)in 3% bovine serum albu- HEPES, 5% horse serum, 10 Ag/mL insulin, 20 ng/mL mininPBSfor1h.DNAwasstainedwith4¶,6-diamidino-2- human epidermal growth factor, 100 ng/mL cholera toxin, phenylindole (Sigma)at a concentration of 5 mg/mL PBS. The 0.5 Ag/mL hydrocortisone, and 100 units/mL penicillin/ cells were then washed in PBS. The coverslips were mounted streptomycin. Dose-response curves were done to deter- on microscope slides (Fisher Scientific)covered with Immu- mine the optimal concentration for each drug and cell Mount oil (Thermo Electron). The coverslips were sealed with line. Exponentially growing cells were treated with various transparent nail polish. The cells were observed using a À/À drugs as follows: U2OS, HCT116p53 , and H1299 received fluorescence microscope (Nikon Eclipse E800). 0.2 Ag/mL doxorubicin, 400 Amol/L 5-FU, or 1.5 mmol/L 3-(4,5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazo- hydroxyurea. MCF10A cells received 0.05 Ag/mL doxorubi- lium Bromide and Apoptosis Assays A cin, 300 mol/L 5-FU, or 1 mmol/L hydroxyurea. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro- Construction of Recombinant Cdc25C Adenovirus mide (MTT)assays were done according to the manufac- The recombinant Cdc25C adenovirus was constructed as turer’s instructions (Sigma). For detection of apoptotic cells, described (16). Briefly, the FLAG-tagged open reading staining was done with Annexin V-PE and then analyzed frame of Cdc25C was subcloned into the pAdTrack-CMV by flow cytometry according to the manufacturer’s protocol shuttle vector. It was then linearized by digestion with (BD Pharmingen). PmeI and was cotransformed together with the adenoviral Small Interfering RNA backbone plasmid, pAdEasy-1, into the homologous Exponentially growing U2OS cells were transfected with recombination-competent Escherichia coli strain BJ5183. SMARTpool reagent containing Cdc25C or Cdc25A small The resultant recombinant was linearized with PacI and interfering RNA oligonucleotides or control oligonucleotides transfected into the adenovirus packaging cell line, 293T. (Dharmacon, RNA Technologies)using DharmaFECT rea- High titer of recombinant adenovirus was obtained by gent 1 (Dharmacon)following manufacturer’s instruction. several rounds of infection. Flow Cytometric Analysis Cells were fixed in ice-cold 70% ethanol followed by Results rinsing with PBS. DNA was stained with propidium iodide Overexpression of Cdc25C Sensitizes U2OS Tumor (20 Ag/mL; Sigma)in the presence of 1 mg/mL RNase A Cells to Doxorubicin-Induced Apoptosis (Sigma). Cell cycle analysis was done on a FACSCalibur Phosphorylation of Cdc25C by Chk1 and its subsequent and data were analyzed using CellQuest. cytoplasmic sequestration or its down-regulation by p53

Mol Cancer Ther 2008;7(12). December 2008

has been shown to contribute to the G2-M arrest caused (Fig. 1A and B). Interestingly, loss of the G2 arrest was bygenotoxicstress(11,12,17).Thus,theabilityof inversely proportional to the amount of the exogenous overexpression of Cdc25C to overcome this G2-M arrest levels of Cdc25C (Fig. 1C). Levels of p53 and p21 induction was examined. U2OS cells were infected with either two in the presence of high levels of Cdc25C appear decreased different amounts of a recombinant adenovirus express- (Fig. 1C), but it is unclear whether this is a direct effect or ing Cdc25C (Ad-25C)1 day before treatment with doxo- related to the use of adenoviral infection. Cells infected rubicin (0.2 Ag/mL)or empty vector with the same with these levels of Cdc25C were able to proliferate until multiplicity of infection as the highest level of Ad-25C confluency in the absence of drug treatment, indicating that (Fig. 1). At this time point, more than 95% of cells were these levels of Cdc25C do not disturb their normal infected as judged by microscopic evaluation of green proliferation (data not shown). Accumulation of cells with fluorescent protein expression (data not shown). Unin- a hypodiploid DNA content is consistent with cell death by fected U2OS cells were included as a control. The cell cycle apoptosis. This was further confirmed by examining profiles of these cells were examined 2 days after drug cleavage of poly(ADP-ribose)polymerase, a known target treatment by flow cytometry analysis (Fig. 1A and B). for apoptosis-associated caspase cleavage (Fig. 1C). Control U2OS cells or those infected with empty adeno- U2OS Tumor Cells That Are Stably Overexpressing virus were able to arrest with a 4N DNA content consistent Cdc25C Are Sensitized to Doxorubicin-Induced with a G2 arrest (Fig. 1A and B). U2OS cells overexpress- Apoptosis ing Cdc25C, however, were unable to maintain the G2 Out of concern that some of the observed effects may arrest and accumulated with a hypodiploid DNA content be a consequence of adenoviral infection, stable U2OS cell

Figure 1. Adenovirus-mediated overexpression of Cdc25C sensitizes U2OS tumor cells to doxorubicin-induced apoptosis. U2OS cells that were untreated (control) or infected with the indicated recombinant adenoviruses for 1 d were treated with doxorubicin (0.2 Ag/mL) for 2 d. A, cells were fixed and stained with propidium iodide before flow cytometry analyses. B, average cell numbers in different phases of the cell cycle from three independent experiments. C, cells were lysed and used in immunoblot analyses using the indicated antibodies. h-Actin is a loading control. Arrow, cleaved form of poly(ADP-ribose) polymerase.

Mol Cancer Ther 2008;7(12). December 2008

lines overexpressing Cdc25C were established. Four con- Cdc25C and Cdk1 (Fig. 3A)as well as p53 itself (Fig. 3B). trol clones were established and shown to behave like the Taken together, these results are consistent with abrogation parental U2OS cells when treated with doxorubicin (data of the G2 arrest leading to subsequent entry into mitosis not shown). The results from one representative clone at with damaged DNA (18, 19). the indicated time points are presented (Fig. 2). Consistent U2OS CellsThat Overexpress Cdc25C Have Reduced with results obtained from cells infected with Ad-25C, Phosphorylation of Cdk1 on Tyr15 and Impaired Up- Cdc25C-overexpressing clones were unable to maintain Regulation of p21 in Response to Treatment with an arrest at the G2-M phases and this occurred in a man- Doxorubicin ner dependent on the levels of Cdc25C. The results of Phosphorylation of Cdk1 on Thr14 and Tyr15 has been two representative clones expressing different amount shown to be important for induction of G2-M arrest (1). of Cdc25C are shown (Fig. 2A and B). Twenty-four hours Thus, the status of phosphorylation of these residues was after drug treatment, only 1% to 2% of the control cells examined in immunoblotting assays. As shown in Fig. 3A, were found with a 2N DNA content representing the in control cells, phosphorylation of these residues increased G1 phase. The majority of the cells were found to have a 1 day after drug treatment. In Cdc25C-overexpressing cells, DNA content of 4N (Fig. 2A). At the same time point, the level of Cdc25C appeared to be a determining factor. In higher levels of Cdc25C were associated with increases clone 4 that expresses lower levels of Cdc25C, an increase in the number of cells found in the G1 and S phases and in phosphorylation of Cdk1 on these residues was observed consequently less in the G2-M phases (Fig. 2A). At later at 1 day, suggesting that these levels of Cdc25C do not time points, in a manner that correlates with the level of affect significantly the initial phosphorylation of Cdk1 on exogenous Cdc25C, cells overexpressing Cdc25C were these residues. However, in this clone, phosphorylation of unable to maintain their G2 arrest and subsequently died these residues was not maintained as long as in control by apoptosis as judged by the increased number of cells cells (Fig. 3A). Higher amount of Cdc25C in clone 18, with a sub-G1 DNA content (Fig. 2A and B). Excess amount however, showed little phosphorylation of these residues of Cdc25C did not affect other behaviors of these cells (Fig. 3A). These results are consistent with the results of that were examined such as proliferation rate or mor- flow cytometry analyses at these time points (Fig. 2). phology compared with the parental cells even after several To further confirm that overexpression of Cdc25C over- passages (data not shown). This suggests that these levels comes the G2 arrest, immunoblotting using antibody of Cdc25C are tolerated by U2OS cells. against phosphorylated histone H3 on Ser10 (pSer10H3), a To examine the effects of Cdc25C overexpression on a marker of mitosis, was done (Fig. 3A). In control cells, little single-cell basis, immunofluorescence analyses were done pSer10H3 was detected at any examined time points after using clone 18, which expresses higher levels of Cdc25C drug treatment (Fig. 3A). In Cdc25C-overexpressing cells, (see Fig. 3A). In exponentially growing cells, exogenous the higher the amount of Cdc25C, the more pSer10H3 was Cdc25C was found to be both cytoplasmic and nuclear detected (Fig. 3A). Levels of total p53 as well as the amount (Fig. 2C). p53 is a substrate for the ATM and ATR kinases of p53, which is phosphorylated on Ser15, were determined. that are activated in response to DNA damage, with Ser15 Overexpression of Cdc25C did not affect the level of p53 on p53 as a primary site for modification by these kinases. induction (Fig. 3B). Consistent with the immunofluores- The intensity and frequency of positivity for p53 that is cence results, the amount of p53 that is phosphorylated on phosphorylated on Ser15 is similar between the control and Ser15 was unchanged (Fig. 3B). Mdm2 is a target for p53 clone 18. In agreement with the flow cytometry analyses, transcriptional up-regulation. High levels of Cdc25C did at 4 days after drug treatment, the clone 18 cells that not interfere with the amount of Mdm2 that is up-regulated overexpress Cdc25C were found to display morphologic in response to doxorubicin treatment (Fig. 3B). Intriguingly, evidence of apoptosis (Fig. 2C; Supplementary Fig. S1).1 however, the increase in levels of another p53 target, the Some of these cells appear to have reduced staining Cdk inhibitor p21, was delayed and impaired in cells with with the anti-FLAG antibody. This is likely due to cells high levels of Cdc25C (Fig. 3B). As p21 has been that are in the late stages of apoptosis showing reduced characterized as playing a key role in cell cycle arrest expression of multiple proteins including the FLAG-tagged checkpoints, the lower levels of p21 in the Cdc25C- expressing cells may contribute to the failure to maintain the G2 checkpoint. Indeed, p21-null cells are exquisitely 1 Supplementary material for this article is available at Molecular Cancer sensitive to DNA damage even in the absence of Cdc25C Therapeutics Online (http://mct.aacrjournals.org/). overexpression (Supplementary Fig. S2).1

Figure 2. U2OS tumor cells that are stably overexpressing Cdc25C are sensitized to doxorubicin-induced apoptosis. A, U2OS cells stably transfected with a pcDNA3 vector or one containing a cDNA for Cdc25C were treated with 0.2 Ag/mL doxorubicin. Cells were harvested at the indicated time points. Representative clone transfected with empty vector and two clones overexpressing different amount of Cdc25C. Cells were fixed and stained with propidium iodide before flow cytometry analyses. B, average cell numbers in different phases of the cell cycle at 4 d from three independent experiments. C, a control clone transfected with empty vector and a clone overexpressing Cdc25C were treated with 0.2 Ag/mL doxorubicin. Cells were fixed at the indicated time points and immunostained with antibodies, which recognize either the FLAG epitope (red) or phosphorylated Ser15 on p53 (green). DNA was stained with 4¶,6-diamidino-2-phenylindole (blue). Arrows, examples of apoptotic cells.

Mol Cancer Ther 2008;7(12). December 2008

Figure 3. U2OS cells that overexpress Cdc25C have reduced phosphorylation of Cdk1 on Tyr15 and impaired up- regulation of p21 in response to treatment with doxorubicin. A and B, a control clone transfected with empty vector and two clones overexpressing different amounts of Cdc25C were trea- tedwith0.2Ag/mL doxorubicin for various indicated times. Cell lysates were prepared and used in immunoblot analyses with the indicated antibodies. h-Actin is a loading control. C, a control clone transfected with empty vector and two clones overexpressing different amounts of Cdc25C were treated with 400 Amol/L 5-FU. At the indicated time points, cells were harvested and used in flow cytometry analyses. Representative experiment. Columns, average cell numbers in different phases of the cell cycle at 2 d from three independent experiments. D, a control clone transfected with empty vector and two clones overexpressing different amounts of Cdc25C were treated with 1.5 mmol/L hydroxyurea. At the indicated time points, cells were harvested and used in flow cytometry analyses. Representative experiment.

Overexpression of Cdc25C in U2OS Tumor Cells Does of excess amount of Cdc25C on cell cycle arrest in G1-S or S Not Confer Sensitivity to Apoptosis onTreatment with phase of the cell cycle, Cdc25C-overexpressing U2OS 5-FU or Hydroxyurea clones or control cells were treated with either of two The role of Cdc25C overexpression in other cell cycle agents that are known to induce arrests at the G1-S-phase checkpoints was then examined. To investigate the effect border (20). 5-FU and hydroxyurea are inhibitors of

Mol Cancer Ther 2008;7(12). December 2008

pyrimidine synthesis or both purine and pyrimidine with Cdc25C abrogating the G2 arrest, high levels of synthesis, respectively (20). Flow cytometry analyses of Cdc25C caused a reduction in the number of cells with cell cycle profiles of U2OS cells treated with these drugs 4N DNA content and a corresponding increase in that of at the indicated time points are shown in Fig. 4C and D. cells with 2N DNA content in a manner that correlated with Cell cycle profile of the control cells in response to 5-FU Cdc25C level of expression (23% in clone 4 and 16% in was consistent with both G1-S and G2 arrests (Fig. 4C). In clone 18; Fig. 4C). Under these conditions, the cells that f these cells, 50% of cells were found to have a DNA escaped the G2 arrest did not die but become arrested in the content of 4N by 2 days (Fig. 4C). Interestingly, consistent G1 phase.

Figure 4. Sensitization of tumor cells to doxorubicin-induced apoptosis by overexpression of Cdc25C is not p53 dependent. The p53-null tumor cell lines HCT116- p53À/À (A) or H1299 (B) were infected with the indicated adenoviruses 1 d before drug treatments. Cells were incubated with either 0.2 Ag/mL doxorubicin, 400 Amol/L 5-FU, or 1.5 mmol/L hydroxyurea for 2 d. Cells were then prepared for flow cytometry analyses. In each case, cells that were not infected with adenovirus were included as an additional control.

Mol Cancer Ther 2008;7(12). December 2008

In response to hydroxyurea, the majority of the control staining assay was then used as a more sensitive means cells accumulated with a S-phase DNA content at 2 days of detecting apoptotic cells to see whether overexpression following drug treatment (Fig. 4D). This was more of Cdc25C affects death in MCF10A cells. Control and Ad- predominant in Cdc25C-overexpressing clones in a manner 25C-infected cells were treated with 0.05 Ag/mL doxoru- dependent on the levels of exogenous Cdc25C (Fig. 4D). bicin and harvested at the same time points as described for These results show that excess amounts of Cdc25C only the MTT assays. Percentages of apoptotic Ad-25C-infected override the G2 arrest caused by doxorubicin and 5-FU and cells at 3 days post-treatment with doxorubicin were have no effect on the G1-S arrest caused by treatment with comparable with that found in control cells (Fig. 5C 5-FU and hydroxyurea. Further, there is no evidence of and D). This further confirms that overexpression of apoptosis in the Cdc25C-expressing cells in response to Cdc25C does not cause enhanced toxicity in nontrans- either of these agents. formed MCF10A cells. Treatment of either the controls or Sensitization of Tumor Cells to Doxorubicin-Induced cells infected with empty adenovirus with 5-FU or hydro- Apoptosis by Overexpression of Cdc25C Is Not p53 xyurea caused a G1-S accumulation that was not affected Dependent on infection with Ad-25C (Fig. 5A). Thus, the ability of high U2OS cells are osteosarcoma cells that express wild-type levels of Cdc25C to sensitize cells to apoptosis correlates p53. To determine the role of p53 status in the ability of with the transformation phenotype of the cells being overexpression of Cdc25C to overcome the G2 arrest, examined. This suggests that overexpression of Cdc25C two cell lines, H1299 (lung carcinoma)and HCT116- may selectively enhance doxorubicin-induced cell death in À À p53 / (colon carcinoma), which lack p53 expression, were tumor cells rather than in nontransformed cells. tested. Both cell lines were treated with similar proto- Lack of Cdc25C Does Not Affect Cell Cycle Progres- cols. Infection of these cells with Ad-25C did not influence sion or the G2 Arrest Caused by Doxorubicin cell cycle profiles of these cells as observed by flow Mice lacking Cdc25C develop normally and their cell cytometry analyses (Fig. 4A and B). The expression from cycle checkpoints appear to be intact (21, 22). To address Ad-25C was confirmed by microscopic analyses of expres- the effects of loss of Cdc25C expression, Cdc25C was sion of green fluorescent protein (data not shown). eliminated by the means of a small interfering RNA Infection with Ad-25C, however, abolished the G2 arrest approach in U2OS cells. Manipulation of the related in both cell lines (Fig. 4A and B). As observed in U2OS cells, Cdc25A was included as a control. As shown in Fig. 6A, elevated levels of Cdc25C did not affect the G1-S-phase the majority of Cdc25C and Cdc25A were successfully arrest caused by 5-FU or hydroxyurea (Fig. 4A and B). down-regulated. As judged from flow cytometry analysis, Thus, excess amounts of Cdc25C are capable of abolishing lack of Cdc25A resulted in accumulation of cells with a 4N G2 arrest in cell lines from different origins and regardless DNA content with a concomitant decrease of cells with 2N of p53 status. DNA content (Fig. 6B). Lack of Cdc25C, however, did Nontransformed MCF10A Cells Are Not Sensitized to not affect cell cycle progression during the course of study Doxorubicin-Induced Apoptosis by Overexpression of (5 days; Fig. 6B). The phosphorylation status of Cdk1 on Cdc25C Thr14 and Tyr15 by immunoblotting was then examined. MCF10A cells are a nontransformed breast epithelial line. Consistent with the flow cytometry analysis, absence of These cells were infected with Ad-25C to investigate the Cdc25C did not affect the levels of phosphorylation on effect of overexpression of Cdc25C on cells that were not these residues in comparison with control cells (Fig. 6A). of tumorigenic origin. Ad-25C infection did not affect the Lack of Cdc25A, on the other hand, led to an increase in the normal cell cycle profile of these cells (Fig. 5A). This phosphorylated form of Cdk1 on these residues (Fig. 6A). same amount of Ad-25C, however, was capable of Next, the effect of absence of Cdc25C on the response overcoming the accumulation of cells with a 4N DNA of U2OS cells to doxorubicin was examined. The number content caused by treatment with doxorubicin showing an of cells accumulated with a 4N DNA content was almost increase in cells in the G1 phase and a corresponding the same as in control cells (Fig. 6C). decrease in cells with 4N DNA content (Fig. 5A). Notably, these cells showed no evidence of an apoptotic response. These cells were then examined before and after doxoru- Discussion bicin treatment using the MTT assay. Viability of cells Evidence provided in this study supports overexpression infected with either empty adenovirus or Ad-25C for of Cdc25C as a bona fide abrogator of the G2 checkpoint. 4 days did not differ significantly from that of control cells It is noteworthy that there is little effect of high levels (Fig. 5B). Similarly, reduction in cell growth following of Cdc25C on the G1- and S-phase checkpoints (4). Here, 3-day doxorubicin treatment was comparable in cells it is shown that excess amounts of Cdc25C abrogates the infected with empty adenovirus or Ad-25C and the controls G2 arrest caused by two common chemotherapeutic (Fig. 5B). These results suggest that elevated amounts agents, doxorubicin and 5-FU, whereas it does not affect of Cdc25C do not impose a negative effect on viability of the G1-S- and S-phase arrests caused by 5-FU and hydro- MCF10A cells during this time course. Further, these xyurea in several cell lines. Interestingly, these observa- data suggest that MCF10A cells arrest in G1-S rather than tions were made in cell lines derived from different tissues. G2-M on expression of excess Cdc25C. The Annexin V-PE This suggests that Cdc25C may have utility to abrogate the

Mol Cancer Ther 2008;7(12). December 2008

G2 arrest in cancer cells from various origins. Of great confirming the role of Cdc25C in the G2 checkpoint. This interest is the observation that the same amount of Cdc25C is a desirable result as proliferation of cells harboring DNA that is capable of overcoming the G2 arrest does not affect damage following chemotherapy is an unwanted out- the normal cell cycle progression of the examined cell lines come. Loss of a G2 arrest has been reported to lead to a including one that is nontumorigenic. To date, there is no phenomenon called mitotic catastrophe, which results in reported evidence of overexpression of Cdc25C in human formation of large micronucleated cells with decondensed tumors in contrast to the other two related phosphatases, chromatin (24). Examination of MCF10A cells overexpress- Cdc25A and Cdc25B (23). Furthermore, as observed here, ing Cdc25C and treated with doxorubicin revealed no Cdc25C needs to be overexpressed at considerably high morphologic changes consistent with mitotic catastrophe. levels (when endogenous Cdc25C is not detectable)for This, however, does not exclude death as a result of it to overcome the G2 arrest. This strongly suggests min- apoptosis or mitotic catastrophe in other types of nontrans- imal toxicity caused by overexpression of Cdc25C even formed cells when exposed to elevated levels of Cdc25C in a nontumorigenic line such as MCF10A. Importantly, and doxorubicin. Cdc25C did not sensitize this latter nontransformed cell In this study, lack of Cdc25C did not affect cell cycle line to doxorubicin-induced cell death. In this case, cells progression of U2OS cells or their response to doxorubicin. failed to arrest in G2-M at the examined time points, further Furthermore, mice lacking Cdc25C develop normally and

Figure 5. Nontransformed MCF10A cells are not sensitized to doxorubicin-induced apoptosis by overexpression of Cdc25C. A, nontransformed breast epithelial MCF10A cells were infected with the indicated adenoviruses 1 d before drug treatments. Cells were incubated with either 0.05 Ag/mL doxorubicin, 300 Amol/L 5-FU, or 1 mmol/L hydroxyurea for 2 d. Cells were then prepared for flow cytometry analyses. Uninfected cells were included as an additional control. B and C, MCF10A cells were infected with either empty adenovirus or Ad-25C 1 d after plating. One day later, one set of cells was treated with doxorubicin (0.05 Ag/mL). Three days later, viability and apoptosis of all cells were measured by MTT (B) and Annexin V-PE (C) staining assays, respectively.

Mol Cancer Ther 2008;7(12). December 2008

Figure 6. Lack of Cdc25C does not affect cell cycle progression or the G2 arrest caused by doxorubicin. Cdc25C or Cdc25A were depleted in U2OS cells after transfection with small interfering RNA oligonucleotides targeting either Cdc25C or Cdc25A. A, cell lysates were sub- jected to immunoblotting using either Cdc25C or Cdc25A antibodies 3 d after transfection (left). Cell lysates were also immunoblotted using antibodies against Cdk1 or its Thr14- Tyr15 phosphorylated form (right). h-Actin is used as a loading control. B, parallel dishes of cells were used in flow cytometry analyses. Represen- tative experiment. Columns, average cell numbers in different phases of the cell cycle from three independent experiments. C, 1 d after transfection with small interfering RNA oligonucleotides that target Cdc25C, cells were treated with doxorubicin (0.2 Ag/mL). Two days later, cells were harvested and used in flow cytometry analyses. Representative experiment. Columns, average cell numbers in different phases of the cell cycle from two independent experiments.

the integrity of the DNA damage checkpoints of their opportunity for reducing the level of drug actually used embryonic fibroblasts remains intact (21, 22). Based on and hence toxicity. These results support the idea that a these observations and a lack of alteration of Cdc25C in combination of overexpressing Cdc25C with treatment many human tumors, it is reasonable to postulate that with conventional genotoxic agents should be given serious Cdc25C is unlikely to be involved in any other cellular considerations as a novel therapeutic strategy. processes and, if so, it is likely to be a redundant player. Two major obstacles encountered with many chemotherapeutic agents include their adverse side effects and the Disclosure of Potential Conflicts of Interest resistance developed against them. In addition, they often No potential conflicts of interest were disclosed. need to be used in combination for a positive therapeutic outcome. Often, however, this further increases undesirable Acknowledgments side effects. Thus, it is expected that a combination of high We thank Drs. Stuart Aaronson, Zhen-Qiang Pan, and Lois Resnick- levels of Cdc25C and such drugs would provide the Silverman for advice.

Mol Cancer Ther 2008;7(12). December 2008

References 14. Sha SK, Sato T, Kobayashi H, et al. Cell cycle phenotype- based optimization of G -abrogating peptides yields CBP501 with a 1. Kastan MB, Bartek J. Cell-cycle checkpoints and cancer. Nature 2004; 2 unique mechanism of action at the G checkpoint. Mol Cancer Ther 2007; 432:316 – 23. 2 6:147 – 53. 2. Levine AJ. p53, the cellular gatekeeper for growth and division. 15. Gewirtz DA. A critical evaluation of the mechanisms of action Cell 1997;88:323 – 31. proposed for the antitumor effects of the anthracycline antibiotics 3. Massague J. G1 cell-cycle control and cancer. Nature 2004;432: Adriamycin and daunorubicin. Biochem Pharmacol 1999;57:727 – 41. 298 – 306. 16. Varmeh-Ziaie S, Manfredi JJ. The dual specificity phosphatase 4. Kawabe T. G2 checkpoint abrogators as anticancer drugs. Mol Cancer cdc25b, but not the closely related cdc25c, is capable of inhibiting cellular Ther 2004;3:513 – 9. proliferation in a manner dependent upon its catalytic activity. J Biol Chem 5. Morgan DO. Principles of cdk regulation. Nature 1995;374:131 – 4. 2007;282:24633 – 41. 6. Donzelli M, Draetta GF. Regulating mammalian checkpoints through 17. St Clair S, Giono L, Varmeh-Ziaie S, et al. DNA damage-induced cdc25 inactivation. EMBO Rep 2003;4:671 – 7. downregulation of cdc25c is mediated by p53 via two independent mechanisms: one involves direct binding to the cdc25c promoter. Mol Cell 7. Xiao Z, Chen Z, Gunasekera AH, et al. Chk1 mediates S and G2 arrests 2004;16:725 – 36. through Cdc25a degradation in response to DNA-damaging agents. J Biol Chem 2003;278:21767 – 73. 18. Bunz F, Dutriaux A, Lengauer C, et al. Requirement for p53 and p21 to sustain G2 arrest after DNA damage. Science 1998;282:1497 – 501. 8. Bartek J, Lukas J. Mammalian G1- and S-phase checkpoints in response to DNA damage. Curr Opin Cell Biol 2001;13:738 – 47. 19. Roninson IB, Broude EV, Chang BD. If not apoptosis, then what? Treatment-induced senescence and mitotic catastrophe in tumor cells. 9. Falck J, Mailand N, Syljuasen RG, Bartek J, Lukas J. The Atm-Chk2- Drug Resist Updat 2001;4:303 – 13. 25a checkpoint pathway guards against radioresistant DNA synthesis. Nature 2001;410:842 – 7. 20. Longley DB, Harkin DP, Johnston PG. 5-Fluorouracil: mechanisms of action and clinical strategies. Nat Rev Cancer 2003;3:330 – 8. 10. Cazales M, Schmitt E, Montembault E, Dozier C, Prigent C, Ducommun B. Cdc25b phosphorylation by Aurora-A occurs at the G2/M 21. Chen MS, Hurov J, White LS, Woodford-Thomas T, Piwnica-Worms transition and is inhibited by DNA damage. Cell Cycle 2005;4:1233 – 8. H. Absence of apparent phenotype in mice lacking cdc25c protein phosphatase. Mol Cell Biol 2001;21:3853 – 61. 11. Peng CY, Graves PR, Thoma RS, Wu Z, Shaw AS, Piwnica-Worms H. Mitotic and G2 checkpoint control: regulation of 14-3-3 protein binding by 22. Ferguson AM, White LS, Donovan PJ, Piwnica-Worms H. Normal cell phosphorylation of cdc25c on serine-216. Science 1997;277:1501 – 5. cycle and checkpoint responses in mice and cells lacking cdc25b and 12. Sanchez Y, Wong C, Thoma RS, et al. Conservation of the chk1 cdc25c protein phosphatases. Mol Cell Biol 2005;25:2853 – 60. checkpoint pathway in mammals: linkage of DNA damage to cdk 23. Kristjansdottir K, Rudolph J. Cdc25 phosphatases and cancer. Chem regulation through cdc25. Science 1997;277:1497 – 501. Biol 2004;11:1043 – 51.

13. Bulavin DV, Higashimoto Y, Popoff IJ, et al. Initiation of a G2/M 24. Castedo M, Perfettini JL, Roumier T, Andreau K, Medema R, Kroemer checkpoint after ultraviolet radiation requires p38 kinase. Nature 2001; G. Cell death by mitotic catastrophe: a molecular definition. Oncogene 411:102 – 7. 2004;23:2825 – 37.

Mol Cancer Ther 2008;7(12). December 2008

Downloaded from mct.aacrjournals.org on September 25, 2021. © 2008 American Association for Cancer Research. Overexpression of the dual specificity phosphatase, Cdc25C, confers sensitivity on tumor cells to doxorubicin-induced cell death

Shohreh Varmeh and James J. Manfredi

Mol Cancer Ther 2008;7:3789-3799.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/7/12/3789

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2008/12/08/7.12.3789.DC1

Cited articles This article cites 24 articles, 10 of which you can access for free at: http://mct.aacrjournals.org/content/7/12/3789.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/7/12/3789. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.