Role of Checkpoint Kinase 1 in Preventing Premature Mitosis in Response to Gemcitabine

Introduction

2′,2′-Difluoro-2′-deoxycytidine (gemcitabine), a deoxycytidine analogue, is a potent radiation sensitizer both in vitro and clinically ( 1– 3). Gemcitabine requires phosphorylation to produce its radiosensitizing and cytotoxic activity. Deoxycytidine kinase catalyzes a series of sequential phosphorylations converting gemcitabine to the monophosphorylated, diphosphorylated, and triphosphorylated metabolites (dFdCMP, dFdCDP, and dFdCTP, respectively). dFdCTP can interfere with DNA synthesis by competition with endogenous dCTP and incorporation into replicating DNA. dFdCDP is a potent inhibitor of ribonucleotide reductase, reducing the synthesis of deoxynucleotide triphosphates, primarily dATP ( 1). This causes cells to redistribute into the early S-phase of the cell cycle. Correlative studies have suggested that simultaneous depletion of dATP pools (through ribonucleotide reductase inhibition) and accumulation in the S-phase of the cell cycle are required to achieve radiosensitization by gemcitabine ( 1, 4, 5).

Given the critical role of cell cycle arrest in radiosensitization, it seemed important to understand the mechanism underlying gemcitabine-induced S-phase arrest. In response to DNA damage, the cell cycle halts, which prevents the propagation of cells with damaged DNA. The two predominant signaling pathways regulating cell cycle progression following DNA damage are the ataxia telangiectasia mutated (ATM)/checkpoint kinase (Chk) 2 and ataxia telangiectasia related (ATR)/Chk1 signal transduction pathways. The ATM/Chk2 and ATR/Chk1 pathways elicit control at the G₁, S, and G₂ cell cycle checkpoints ( 6, 7). In response to ionizing radiation–induced DNA double-strand breaks, the ATM/Chk2 pathway prevents DNA synthesis and cell cycle progression ( 8). The ATR/Chk1 pathway is induced in response to agents that inhibit DNA replication either directly (aphidicolin; hydroxyurea) or indirectly (ultraviolet radiation; refs. 9, 10). The ATM/Chk2 and ATR/Chk1 signaling pathways converge on the Cdc25 phosphatase family. Within the Cdc25 family, Cdc25A promotes G₁-S transition through stimulation of cyclin E/A-cyclin-dependent kinase (Cdk) 2 complexes ( 11) and promotes G₂-M transition through cyclin B-Cdk1 activation ( 12, 13). Cdc25A is degraded after phosphorylation by either Chk1 or Chk2 in a ubiquitin- and proteosome-dependent manner in response to DNA damage. Although radiation-induced Cdc25A degradation tends to depend on Chk2 ( 8) and hydroxyurea- or ultraviolet radiation-induced degradation of Cdc25A on Chk1 ( 14), crossover pathways do exist ( 15, 16).

Because of the importance of cell cycle redistribution in gemcitabine-mediated radiosensitization, we designed a study to assess the role of Chk1 and Chk2 in gemcitabine-induced cell cycle arrest. We hypothesized that gemcitabine, as an inhibitor of ribonucleotide reductase or as DNA chain terminator, might induce both Chk1 and Chk2 signal transduction pathways. When we found that both Chk1 and Chk2 were phosphorylated and Cdc25A down-regulated in response to gemcitabine, we used small interfering RNA (siRNA) techniques to directly assess the roles of Chk1 and Chk2 in gemcitabine-induced cell cycle arrest. We hypothesized that Chk1 and/or Chk2 might directly mediate DNA synthesis inhibition by gemcitabine or, alternatively, that Chk1 and/or Chk2 might act to coordinate DNA synthesis with the rest of the cell cycle machinery. We tested this hypothesis by simultaneously monitoring DNA content and entry into mitosis of cells treated with gemcitabine and depleted of Chk1 or Chk2.

Materials and Methods

Cell culture. We chose a panel of cell lines with varying abilities to be radiosensitized by gemcitabine. A549, BxPC-3, Panc-1, RKO, and SW620 cells were grown in DMEM (Panc-1 and SW620) or RPMI supplemented with 10% fetal bovine serum (FBS) and 2 mmol/L glutamine. In addition, RKO cells were supplemented with 1.5 g/L sodium bicarbonate, 0.1 mmol/L nonessential amino acids, and 1.0 mmol/L sodium pyruvate. Cells were passaged weekly. Cells were routinely screened for Mycoplasma contamination.

Gemcitabine. Gemcitabine was obtained from Eli Lilly (Indianapolis, IN). Gemcitabine was dissolved in PBS and stored at −20°C. Cells were treated with 10 to 300 nmol/L gemcitabine for 24 hours.

Irradiation. Cells were irradiated at room temperature using a Pantak DXT300 orthovoltage unit with 6 Gy at a dose rate of ∼3 Gy/min. Dosimetry was carried out using an ionization chamber connected to an electrometer system that was directly traceable to a National Institute of Standards and Technology calibration.

Clonogenic survival. Cells were subcultured into 100 mm² dishes and treated for 24 hours with gemcitabine. Cells were then irradiated and processed for clonogenic survival according to the method described previously ( 17). Cell survival curves were fitted using the linear quadratic equation, and the mean inactivation dose was calculated according to the method of Fertil et al. ( 18). The radiation enhancement ratio was calculated by dividing the mean inactivation dose under control conditions by the mean inactivation of gemcitabine-treated cells. Cytotoxicity was measured by the surviving fraction of gemcitabine-treated cells to untreated control cells.

Small interfering RNA. SMARTpool Chk1, Chk2, and nonspecific control pool (−pool) siRNAs were purchased from Dharmacon (Lafayette, CO). Panc-1 cells were transfected with siRNA using Oligofectamine transfection reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's protocol. Cells in Opti-MEM reduced serum medium (Invitrogen) were treated with 100 nmol/L siRNA and Oligofectamine. After 4 hours, medium was adjusted to contain 10% FBS. On the next day after transfection, cells were replated and then exposed to drug 24 hours later. SW620 cells were transfected with siRNA using LipofectAMINE 2000 (Invitrogen) according to the manufacturer's protocol. Cells were transfected in the presence of 10% FBS. Twenty-four hours after transfection, medium was exchanged with normal growth medium. On the next day, SW620 cells were replated and then exposed to drug 48 hours later.

Flow cytometry. Cells were harvested and fixed in 70% ethanol. For DNA content flow cytometry, cells were stained with a solution of 0.018 mg/mL propidium iodide (PI) and 0.04 mg/mL RNase A. For flow cytometry with P-histone H3, cells were permeabilized with 0.25% Triton-X 100 and then resuspended in P-histone H3 antibody (Upstate Biotechnology, Lake Placid, NY) followed by a FITC-conjugated anti-rabbit secondary antibody as described previously ( 19). Cells were then stained with PI (0.033 mg/mL). Premature and normal mitoses were separated by first defining normal mitosis under control conditions as P-histone H3–positive cells with 4N DNA content and then applying these parameters to treated samples. Premature mitosis was defined as the P-histone H3–positive cells with <4N DNA content. For bromodeoxyuridine (BrdUrd) flow cytometry, cells were exposed to 30 μmol/L BrdUrd for 15 minutes and processed as described previously ( 20) using an antibody recognizing BrdUrd (PharMingen, San Diego, CA) followed by a FITC-conjugated goat anti-mouse secondary antibody (Sigma Chemical Co., St. Louis, MO). In each experiment, a control sample without BrdUrd was processed to determine the background signal. Human lymphocytes or trout erythrocyte nuclei (BioSure, Grass Valley, CA) were included as internal standards. Cells were analyzed by counting 10,000 events on a Beckman Coulter (Fullerton, CA) Epics Elite (University of Michigan Flow Cytometry Core) or 40,000 events on a Becton Dickinson FACScan (San Jose, CA).

Immunoblotting. Whole-cell lysates were prepared in 10 mmol/L Tris (pH 7.4), 2% SDS, 1× Complete protease inhibitor cocktail (Roche, Mannheim, Germany), 1 mmol/L NaF, 2 mmol/L Na₃VO₄, and 1 mmol/L Na₂PO₇. Protein concentration was determined with the BCA Protein Assay Reagent (Pierce, Rockford, IL). Samples were diluted in 1× loading buffer [0.32 mol/L Tris-HCl, 10% glycerol, 2% SDS, 0.2% bromophenol blue, 4% 2-mercaptoethanol (pH 6.8)] before loading onto 7.5% or 10% polyacrylamide gels. Separated proteins were transferred to polyvinylidene difluoride membranes and hybridized overnight at 4°C with antibodies recognizing P-Chk1 (S317 or S345), P-Chk2 (T68; Cell Signaling Technology, Beverly, MA), Chk2 (N-17), Cdc25A (F-6; Santa Cruz Biotechnology, Santa Cruz, CA), or Chk2 (clone 7; Upstate Biotechnology). Membranes were probed with secondary antibodies, incubated with Enhanced Chemiluminescence Plus (Amersham Biosciences, Little Chalfont, United Kingdom) and then exposed to film.

Results

To determine whether radiosensitization by gemcitabine is accompanied by perturbation of the cell cycle, we examined BrdUrd incorporation and cell cycle distribution in BxPC-3, Panc-1, A549, RKO, and SW620 cells in response to gemcitabine. We investigated both noncytotoxic and cytotoxic conditions that would produce radiosensitization ( Table 1 ) and found that S-phase arrest accompanied radiosensitization. In these experiments, S-phase arrest was characterized by the presence of BrdUrd-positive cells with a cessation of cell doubling. In BxPC-3 and Panc-1 cells, treatment with gemcitabine under radiosensitizing conditions (10-30 and 100-300 nmol/L, respectively) resulted in arrest of the cells in the early S-phase of the cell cycle ( Fig. 1 ; Table 2 ). In BxPC-3 and Panc-1 cells, S-phase arrest and radiosensitization were produced by noncytotoxic concentrations of gemcitabine (10 and 100 nmol/L, respectively). Under conditions of moderate radiosensitization (10 nmol/L), A549 cells accumulated in both early and middle S-phases. A subpopulation of A549 cells arrested in early S-phase was visible in the histogram of propidium iodide–stained cells (data not shown). However, greater sensitization was caused by a higher concentration (30 nmol/L) that produced S-phase arrest. In RKO cells, a lower, nonradiosensitizing concentration of gemcitabine (10 nmol/L) did not arrest cell cycle progression, whereas a higher cytotoxic concentration of gemcitabine (20-30 nmol/L) caused arrest and radiosensitization. In SW620 cells, conditions that produced radiosensitization and cytotoxicity (30-100 nmol/L) also caused early S-phase arrest. These results provide a correlation between radiosensitization and redistribution of cells into S-phase by gemcitabine.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Effect of gemcitabine on cell cycle distribution. BxPC-3, Panc-1, A549, RKO, and SW620 cells were treated for 24 hours with low-dose or high-dose gemcitabine (Gem). BrdUrd incorporation and PI staining were analyzed by flow cytometry. Representative of three independent experiments.

To begin to understand the mechanism of the S-phase arrest by gemcitabine, we examined the DNA damage–induced cell cycle checkpoint proteins, Chk1 and Chk2. We measured the levels of Chk1 and Chk2 phosphorylated at sites required for enzymatic activity. Cells treated with gemcitabine under radiosensitizing conditions showed an accumulation of S317 and S345 phosphorylated Chk1 (P-Chk1 S317 and P-Chk1 S345, respectively) in all five cell lines ( Fig. 2A and B ) with no change in the total level of Chk1 protein ( Fig. 2C). In BxPC-3 and Panc-1 cells, P-Chk1 accumulation was observed as early as 8 hours after the start of gemcitabine treatment and persisted during the 24-hour gemcitabine treatment, whereas in A549 and RKO cells the induction occurred later (16-24 hours; data not shown). We also measured the levels of T68 phosphorylated Chk2 (P-Chk2 T68) in gemcitabine-treated cells. Radiosensitizing concentrations of gemcitabine also produced an accumulation of P-Chk2 (T68) in BxPC-3, Panc-1, A549, RKO, and SW620 cells ( Fig. 2D) without change in total Chk2 protein levels ( Fig. 2E). These results suggest that both Chk1 and Chk2 are activated by the conditions associated with S-phase arrest and radiosensitization.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Effect of gemcitabine on Chk1 and Chk2. A-E, cells were treated for 24 hours with gemcitabine (Gem 1, BxPC-3, A549, and RKO, 10 nmol/L; Panc-1, 100 nmol/L; SW620, 30 nmol/L; Gem 2, BxPC-3, A549, and RKO, 30 nmol/L; Panc-1, 300 nmol/L; SW620, 100 nmol/L) or for 8 hours with hydroxyurea (HU) or left untreated (C). Additionally, cells were treated with 6 Gy irradiation alone (6Gy) or following 24 hours of gemcitabine treatment (Gem 1 + 6Gy). Irradiated cells were incubated for an additional 30 minutes after irradiation. Levels of Chk1 or Chk2 in whole-cell lysates were detected by immunoblotting with antibodies that specifically recognize S317 (A) or S345 (B) phosphorylated Chk1, total Chk1 (C), T68 phosphorylated Chk2 (D), or total Chk2 (E). β-actin is shown as a loading control. Representative of at least three independent experiments.

In addition, we wished to assess the interaction between gemcitabine and radiation on Chk1 and Chk2. In four of five cell lines (BxPC-3, Panc-1, RKO, and SW620), radiation treatment (6 Gy) alone induced P-Chk1 protein levels at 30 minutes after irradiation ( Fig. 2A and B). Treatment with gemcitabine before irradiation produced an increase in P-Chk1 protein levels in BxPC-3, Panc-1, and SW620 cells compared with treatment with radiation or gemcitabine alone. As anticipated, radiation treatment resulted in induction of P-Chk2 protein in all of the cell lines ( Fig. 2D). This effect was similar to that observed in cells treated with gemcitabine alone or with gemcitabine before irradiation. These findings suggest that gemcitabine and radiation might have overlapping abilities to activate Chk1 and Chk2 in some cell lines but not in others.

Checkpoint kinase 1 regulates Cdc25A protein levels in gemcitabine-treated cells. To further explore the mechanism of S-phase arrest by gemcitabine, we investigated whether Chk1 or Chk2 activation could produce Cdc25A degradation. Treatment of cells with gemcitabine under radiosensitizing conditions caused a reduction in Cdc25A protein in all five cell lines ( Fig. 3A ). In addition, radiation also resulted in a reduction of Cdc25A protein levels in both the presence and the absence of gemcitabine pretreatment. These results suggest that the induction of Chk1 and/or Chk2 in response to gemcitabine or radiation produces a downstream response, which initiates a reduction in Cdc25A protein levels.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Effect of gemcitabine on Cdc25A protein levels. A, cells were treated as described in Fig. 2. Levels of Cdc25A were measured by immunoblotting. B, cells were either mock treated (−) or treated with −pool, Chk1, or Chk2 siRNA as indicated. After 24 hours, gemcitabine (100 nmol/L) was added to cells for an additional 24 hours. Levels of Cdc25A, Chk1, Chk2, and β-actin proteins were measured in the cell lysates. Representative of three independent experiments.

To determine if the reduction in Cdc25A protein in response to gemcitabine was mediated by Chk1 or Chk2, we used siRNA to deplete Chk1 or Chk2 from Panc-1 cells that were well sensitized under noncytotoxic conditions. In initial experiments, we confirmed that cells depleted of Chk1 were not able to induce P-Chk1 protein in response to gemcitabine. Likewise, cells depleted of Chk2 were no longer able to induce P-Chk2 protein. Following treatment with Chk1 or Chk2 siRNA, Panc-1 cells were treated with gemcitabine and Cdc25A protein levels were assessed. Under control conditions (−pool), gemcitabine produced the expected reduction in Cdc25A protein ( Fig. 3B). However, in cells depleted of Chk1, Cdc25A protein levels persisted despite treatment with gemcitabine. It is of interest to note that the basal levels of Cdc25A were increased in cells depleted of Chk1. In addition, we found that gemcitabine still produced a reduction in Cdc25A protein levels in cells depleted of Chk2. These results provide evidence that Chk1, but not Chk2, regulates Cdc25A protein in both untreated and gemcitabine-treated cells.

Checkpoint kinase 1 prevents premature mitosis in gemcitabine-treated cells. Next, we wished to understand how Chk1 and Chk2 were involved in the cell cycle redistribution produced by gemcitabine. We postulated two means by which Chk1 or Chk2 might regulate the cell cycle. We hypothesized that Chk1 and/or Chk2 could be the cause of DNA synthesis inhibition by gemcitabine or, alternatively, that Chk1 and/or Chk2 might be activated by DNA synthesis inhibition and thus act to coordinate DNA synthesis with the rest of the cell cycle machinery. We expected that if Chk1 or Chk2 were causative in DNA synthesis inhibition by gemcitabine, then depleting Chk1 or Chk2 would result in new DNA synthesis; however, if induction of Chk1 or Chk2 were the effect of DNA synthesis inhibition, then depleting Chk1 or Chk2 would permit cell cycle progression despite incompletely replicated DNA. To address these two hypotheses, in cells depleted of Chk1 or Chk2, we monitored DNA content by PI staining and progression of the cell cycle by P-histone H3 staining. We found that the depletion of Chk1 and Chk2 from cells did not prevent the accumulation of cells with a S-phase DNA content induced by gemcitabine ( Fig. 4A ). These results suggest that neither Chk1 nor Chk2 is the cause of DNA synthesis arrest in gemcitabine-treated cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

DNA content in response to gemcitabine in cells depleted of Chk1 or Chk2. A, Panc-1 cells were treated with −pool, Chk1, or Chk2 siRNA. After transfection, gemcitabine was added for an additional 24 hours. Cell cycle distribution was determined by the PI-stained DNA content. B, immunoblots of Chk1, Chk2, and β-actin proteins in Panc-1 cells treated with siRNA. Representative of three independent experiments.

To test the alternative hypothesis of whether Chk1 and/or Chk2 act to coordinate DNA synthesis with the cell cycle, we monitored entry into mitosis by staining for P-histone H3 ( 19). As anticipated, treatment with gemcitabine markedly reduced the number of cells entering mitosis ( Fig. 5A and B ). The percentage of cells in mitosis was reduced from 1.9% to 0.5% in Panc-1 cells and from 2.5% to 0.7% in SW620 cells by gemcitabine. Depletion of Chk2 in gemcitabine-treated cells had no effect on the fraction of cells entering mitosis. However, depletion of Chk1 in gemcitabine-treated Panc-1 cells resulted in the entry of cells into mitosis despite incomplete DNA synthesis (4.2%). In SW620 cells, a small percentage of cells were observed in premature mitosis (0.4%) in response to gemcitabine and nonspecific siRNA (−pool + gemcitabine). However, depletion of Chk1 in gemcitabine-treated SW620 cells produced a marked increase in the percentage of cells prematurely entering mitosis (1.7%). Depletion of Chk1 or Chk2 from cells in the absence of gemcitabine treatment did not produce any consistent alterations in the fraction of mitotic cells. Although Chk2 was substantially reduced by siRNA ( Figs. 4B and 5C), it was not completely suppressed. Therefore, we cannot exclude the possibility that Chk2 is involved in cell cycle progression following gemcitabine. Nonetheless, we conclude that Chk1, but not Chk2, prevents premature mitosis in gemcitabine-treated cells and acts to coordinate DNA synthesis with the rest of cell cycle machinery.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Entry into mitosis in response to gemcitabine in cells depleted of Chk1 or Chk2. Panc-1 (A) and SW620 (B) cells were treated as in Fig. 4. Entry into mitosis was assessed by staining with an antibody recognizing P-histone H3. DNA content was assessed by PI staining. Immunoblots of Chk1, Chk2, and β-actin proteins in SW620 cells treated with siRNA (C). Representative of three to eight independent experiments. The mean number ± SE of cells in premature mitosis was 0.3 ± 0.1% or 4.6 ± 0.3% (n = 3) for Panc-1 cells treated with gemcitabine and −pool siRNA or gemcitabine and Chk1 siRNA, respectively, and 0.5 ± 0.1% or 2.8 ± 0.3% (n = 8), respectively, for SW620 cells.

Discussion

In this study, we have found that treatment with gemcitabine under radiosensitizing conditions produces arrest in early S-phase that is associated with accumulation of the phosphorylated forms of Chk1 and Chk2. Although both Chk1 and Chk2 were phosphorylated, the subsequent degradation of Cdc25A in response to gemcitabine was mediated by Chk1. Neither Chk1 nor Chk2 were directly responsible for the inhibition of DNA synthesis induced by gemcitabine. However, Chk1 negatively regulated entry into mitosis in gemcitabine-treated cells. These findings suggest that gemcitabine stimulates Chk1 to initiate a G₂-M cell cycle checkpoint. Furthermore, these data imply that Chk1 acts to coordinate the cell cycle with DNA synthesis, thus preventing premature mitotic entry in gemcitabine-treated cells ( Fig. 6 ).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Cell cycle effects of gemcitabine. Inhibition of DNA synthesis by gemcitabine, either through DNA incorporation or dATP pool depletion, leads to the induction of P-Chk1. Chk1 in turn triggers the degradation of Cdc25A and this ultimately renders cyclin B-Cdk1 complexes inactive, thus preventing the progression of cells into mitosis. In gemcitabine-treated cells depleted of Chk1 Cdc25A protein persists, cyclin B-Cdk1 complexes are active and cells are driven into mitosis despite incompletely replicated DNA.

There are two possible scenarios for how DNA synthesis might be arrested in response to DNA damage. Firstly, arrest of DNA synthesis could be directly mediated through checkpoint activation. Alternatively, arrest might occur for other reasons, such as chain termination, which activates checkpoints. Others have found that Chk1 and Chk2 cause the S-phase arrest in response to topoisomerase I poisons. For example, Xiao et al. ( 13) found that Chk1 mediates the S-phase checkpoint induced by camptothecin, whereas Yu et al. ( 21) observed that Chk2 mediates the camptothecin-induced S-phase checkpoint. These studies showed that depletion of Chk1 or Chk2 abrogated the camptothecin-induced S-phase arrest. In contrast to camptothecin, gemcitabine could directly arrest DNA synthesis through the depletion of deoxynucleotide triphosphate pools and incorporation into DNA. In turn, arrested DNA synthesis might then stimulate cell cycle checkpoint activation. For camptothecin, inhibition of DNA synthesis might not be mediated by the topoisomerase I cleavage complexes alone but is likely dependent on induction of a cell cycle checkpoint ( 22, 23). Taken together, these findings suggest that drugs that directly affect DNA synthesis (such as the antimetabolites) may show a different pattern of checkpoint responses from drugs that affect other cellular functions, such as DNA unwinding. In addition, although our studies showed that neither Panc-1 nor SW620 cells treated with gemcitabine resume DNA synthesis in response to Chk1 depletion, Pan et al. ( 24) have found that Chk1 depletion allows gemcitabine-treated U2OS cells to resume DNA synthesis. The difference between our findings and theirs might be due to inherent differences among Panc-1, SW620, and U2OS cells or differences in their metabolism of gemcitabine.

Our finding that depletion of Chk1 results in the premature entry of gemcitabine-treated cells into mitosis could be a result of the elevated expression of Cdc25A in Chk1-depleted cells. Overexpression of Cdc25A has been shown to drive cells into mitosis even in the presence of incompletely synthesized DNA ( 12, 25). Furthermore, Molinari et al. ( 26) showed that Cdc25A overexpression caused cells with arrested DNA synthesis (by hydroxyurea) to enter mitosis. These findings suggest that Chk1 depletion or Cdc25A overexpression can drive the cell cycle machinery independent of DNA synthesis and lead cells into premature mitosis.

Although the focus of this study has been on assessing the role of Chk1 and Chk2 in early S-phase arrest, it will be necessary to study subsequent cellular events. For instance, it is unclear why Panc-1 and BxPC-3 cells tolerate Chk1 and Chk2 activation and DNA synthesis inhibition for up to 24 hours with minimal cytotoxicity, whereas A549, RKO, and SW620 cells show substantial cytotoxicity under the same conditions. The most obvious factor that might influence cytotoxicity is p53 status. For instance, p53 has been shown to influence cell death in response to antimetabolites through Fas-mediated apoptosis ( 27), and we have shown previously that RKO cells with disrupted p53 function are more resistant to gemcitabine-mediated cytotoxicity ( 28). Furthermore, in the current study, p53 mutant cells (BxPC-3 and Panc-1) tolerate gemcitabine-induced S-phase arrest with minimal cytotoxicity, whereas p53 wild-type cells (A549 and RKO) do not ( Table 1). However, in SW620 cells, which are p53 mutant, cell cycle arrest is accompanied by cytotoxicity. Therefore, p53 is unlikely to be the only factor regulating cytotoxicity in S-phase–arrested cells. It will be important to distinguish the mechanism by which some cells are capable of tolerating S-phase arrest in future studies. Furthermore, although we have begun to understand the roles of Chk1 and Chk2 in gemcitabine-mediated cell cycle arrest, their influence on cytotoxicity and radiosensitization have not yet been investigated. It is possible that Chk1 depletion will enhance gemcitabine-mediated cytotoxicity and radiosensitization, thus providing a therapeutic potential. Studies of not only the initial activation of checkpoints but also the consequences of checkpoint activation in subsequent cellular events, such as survival, are required.