Checkpoint Kinase 1 Down-Regulation by an Inducible Small Interfering RNA Expression System Sensitized In vivo Tumors to Treatment with 5-Fluorouracil

DNA damage checkpoints are signal transduction pathways activated on DNA damage that ensure a prompt answer to the lesion, leading to a delay or block of the cell cycle to allow the repair of the damage or, if the damage is too heavy, the induction of cell death through activation of apoptosis, necrosis, or other type of cell death. These DNA damage checkpoints have the goal to ensure high-fidelity transmission of the genetic materials to daughter cells and their dysregulation, due to mutation or altered expression of key proteins, results in genomic instability and represents an hallmark of cancer (1, 2).

Most of the anticancer agents currently used activate DNA damage transduction pathways that, depending on the phase in which the cell is hit, lead to the activation of a G₁, S, and G₂-M checkpoint. G₁ block mainly relays on p53, which is a transcriptional factor promptly activated and stabilized on DNA damage by several post-translational modifications (both phosphorylation and acetylation events) leading to the subsequent transactivation of genes important for the G₁ block of the cell cycle to allow the repair or the induction of apoptosis in case of unrepairable damage (3–5). The G₂ checkpoint relies on the maintenance inhibitory phosphorylation of the master mitotic kinase Cdk1/Cdc2 on Thr¹⁴ and Tyr¹⁵ through the inactivation of the CDC25 family phosphatases that are required to dephosphorylate those sites to allow the progression into mitosis (6, 7). Checkpoint kinase 1 (Chk1) was first identified in Schizosaccharomyces pombe as one of the main mediator of the G₂ checkpoint after DNA damage and it was later found to have a similar role in mammalians (8–10). Chk1 kinase is activated on DNA damage by ATR-mediated phosphorylation and once activated can phosphorylate, inactivating, the CDC25 family phosphatases (11, 12). These phosphatases regulate the timely activation of cyclin-dependent kinases at the G₁-S and G₂-M transitions. Chk1-mediated phosphorylation of CDC25C and CDC25B makes them sequestered into the cytoplasm by binding to the 14-3-3 family proteins, thus preventing the activation of the cyclin B/Cdc2 kinase complex responsible for the G₂ transition. Phosphorylation of CDC25A targets it to proteolysis through the ubiquitin pathway and prevents the activation of the cyclin-dependent kinase complexes necessary for the S transition (13–17). More recently, a role of Chk1 in the M transition and spindle checkpoint has been evocated (18, 19).

The unraveling of the molecular mechanisms at the basis of cell response to DNA damage in the last years has led to the identification not only of proteins important for such response (such as p53 and Chk1) but also to the discovery of how often these proteins are indeed lacking or functional inactive (due to mutation) in human tumors (20, 21). All these data have envisaged the possibility to exploit the difference between normal cells (with no mutation) and tumors cells (with many different mutations) to increase the therapeutic index of the currently used anticancer agents. In particular, impairment of the G₁ checkpoint is a frequent event in cancer cells due to mutation of tumor suppressor protein p53 rendering tumor cells dependent on the residual G₂ checkpoint to recover from DNA damage, whereas normal cells do also rely on a functional G₁ checkpoint. The possibility to abrogate the only functional residual G₂ block in tumor cells would lead to a preferential killing of tumor cells. This prediction has been validated in many cellular systems in vitro. Inhibition of Chk1, both by small interfering RNA (siRNA) against the protein and small molecules inhibiting its kinase activity, showed that it is possible to potentiate the cytotoxic activity of DNA-damaging agents and that this effect seems to be more pronounced in cells with a mutated p53 (22, 23). Previous data from our laboratory have clearly shown that in HCT-116 cellular system the inactivation of both Chk1- and p53-sensitized cells to ionizing radiation and cisplatin (DDP; ref. 24).

No isogenic models have been characterized for the simultaneously deficiency in p53 and Chk1 in an in vivo setting. We here present the characterization of a tetracycline-inducible tumor system with a stable and functional expression, on the addition of tetracycline, of siRNA against Chk1 with a complete abrogation of Chk1 expression obtained in HCT-116 colon carcinoma cell and in its subclone in which p53 has been deleted by homologous recombination (25). The model not only further corroborated the in vitro concept that a greater sensitization to the cytotoxic activity of different anticancer agents can be obtained but also allowed the in vivo proof of principle that specific down-regulation of p53 and Chk1 in tumors growing in vivo render them more sensitive to the cytotoxic effects of anticancer agents without increasing the toxicity.

Cell lines and culture conditions. HCT-116 human colon cancer cell line and its sublines were maintained in IMDM (Cambrex) supplemented with 1% glutamine (Cambrex) and 10% fetal bovine serum (Sigma). The clone 379.2 (p53^−/−), kindly obtained by Prof. Vogelstein, derives from HCT-116 cell line where p53 gene has been disrupted by homologous recombination (25).

Clones stably transfected with pcDNA6/TR plasmid (Invitrogen) coding for tetracycline repressor were obtained in our laboratory from HCT-116^wt and HCT116 ^p53-/- cells and were maintained in medium supplemented with 5 μg/mL blasticidin (InvivoGen). Clones overexpressing the pSuperior siRNA against Chk1 were maintained in medium containing 5 μg/mL blasticidin and 0.17 μg/mL puromycin (Sigma).

Plasmids and stable transfections. To get cellular clones inducibly expressing the siRNAs against Chk1, the pSuperior plasmid (Oligoengine) was used. Briefly, the target siRNA sequences against Chk1, 5′-GGGAUAACCUCAAAAUCUCUU-3′ (24) and 5′-GCGUGCCGUAGACUGUCCAUU-3′, (14) were each subcloned sequentially under the control of a H1 promoter in tandem in the pSuperior plasmid. The forward and reverse siRNA oligonucleotides of each target sequence were annealed following the manufacturer's instruction (Oligoengine). The annealed oligonucleotides were cloned into the linearized vector between the BglII and the HindIII restriction sites. To clone the second cDNA target for the Chk1 siRNA in the plasmid already containing the first cDNA Chk1 target, the forward and reverse oligonucleotides of this second target had to be designed by adding, downstream to the termination sequence, the EcoRI restriction site, so that after the first step of cloning the target sequence inserted together with the H1 promoter could be excised by cutting with the EcoRI restriction enzyme and then ligated in the first pSuperior plasmid containing one cDNA Chk1 target and previously linearized with EcoRI (single site in the vector). To clone the siRNA scramble sequences, the same procedures just described were followed. The scramble siRNA target sequences were 5′-ACUAUAAGGGUCAACACCUUU-3′ and 5′-GCUUCCCAGUGGCAGGUCAUU-3′.

HCT-116 8A (p53^+/+) and 15A (p53^−/−) cells, obtained previously in our laboratory and stably expressing the repressor tetracycline coding plasmid, were transfected with pSuperior.puro double siRNA Chk1 or double siRNA scramble plasmids by using Lipofectamine 2000 reagent (Invitrogen) following the manufacturer's instructions. Forty-eight hours after transfection, cells were seeded at low density in medium containing antibiotics. Growing clones were isolated and screened for Chk1 down-regulation 72 h after the induction with 2 μg/mL doxycycline by Western blot analysis.

Transient transfection and drug treatment. HCT-116^wt and HCT116 ^p53−/− cells were transfected with siRNA duplexes (35 pmol each) using Lipofectamine 2000 reagents according to the vendor's protocol. Forty-eight hours after transfection, cells were treated for 24 h with 5-fluorouracil (5-FU), doxorubicin, etoposide (VP-16), Taxol, and the proteasome inhibitor PS341. At the end of treatment, cells were trypsinized and seeded at 200 cells/mL. Plates were allowed to grow for 1 week. Cells were then stained with crystal violet and resuspended in isopropanol and the absorbance was read on a spectrophotometer. Data are mean ± SD of at least three experiments done in triplicate.

Cell growth and clonogenic survival assays. Cells with or without doxycycline were counted every 24 h until 96 h from seeding. Doxycycline was added every 48 h. For clonogenic assays, cells were harvested from log-phase growing cultures and plated at 200 cells/mL into six-well plates in medium containing or not doxycycline. After 72 h, cells were treated with different doses of anticancer agents, and at the end of treatment, cells were incubated for 14 days in a humidified incubator at 37°C, allowing colonies formation. Colonies were then stained with crystal violet solution. The plates were quantified and analyzed by using the Image-Pro Plus software program (Media Cybernetics). Three random fields were counted for each triplicate samples and average values were presented as mean ± SD.

Cell cycle analysis. Cells were fixed immediately after the end of 24 h of treatment with 5 μmol/L 5-FU and after 24 h from recovery. The 5-FU treatment was done following 72 h of induction or not with 2 μg/mL doxycycline. To stain with phosphorylated γH2AX, cells were fixed at indicated time points with methanol-free formaldehyde in PBS 1% for 30 min on ice and then in 70% ethanol. The phosphorylated γH2AX and DNA staining procedure was done as already described (26).

Western blot analysis. Briefly, protein extracts were obtained by using a lysis buffer containing 10 mmol/L Tris-HCl (pH 7.4), 150 mmol/L NaCl, 0.1% NP-40, 5 mmol/L EDTA, and 50 mmol/L NaF in the presence of protease inhibitors. Total cellular proteins (40 μg) were separated on 8% SDS-polyacrylamide gels and electrotransferred to nitrocellulose membrane (Protran, Schleicher & Schuell). Immunoblotting was carried out with anti-Chk1 (G-4), β-actin (C-11), p53 (DO-1), Chk2 (H-300), P-Cdc2 (Thr¹⁵, Tyr¹⁶), Cdc2(p34), P-Cdc25C-Ser²¹⁶, Cdc25C (C-20), and geminin (FL-209) purchased from Santa Cruz Biotechnology, whereas anti-Cdc25A was purchased by NeoMarkers.

Immunofluorescence. Cells either induced or not at 72 h from the first induction were fixed with 4% paraformaldehyde for 20 min following permeabilization with Triton X-100 (0.5% in PBS) for 5 min and blocking in 2% BSA and 0.2% Triton X-100 in PBS (blocking buffer) for 1 h. Cells were subsequently incubated for 1 h with the anti-Chk1 primary antibody (clone DCS310, Sigma) diluted 1:500 in blocking buffer and then FITC-conjugated secondary antibody Alexa Fluor 488 donkey anti-mouse (Molecular Probes) was added to the cells diluted 1:500. Cells were incubated with Hoechst 33342 (trihydrochloride, trihydrate, Invitrogen) solution (final concentration: 1 μg/mL) for 20 min and mounted with the mounting media Fluor Save (Calbiochem), and then observed in a Zeiss Axiophot photomicroscopy equipped for epifluorescence (Carl Zeiss).

Immunohistochemistry. Tumor tissues were quickly excised from sacrificed mice and fixed in formalin 4% for 24 h. Paraffin-embedded tissue sections were dewaxed in xylene, rehydrated through a graded ethanol series, and then washed in tap water. Antigen retrieval was done for 30 min at 95°C in 1 mmol/L EDTA (pH 8.0) in a water bath. Endogenous peroxidase was inhibited by incubating the slides in 1% H₂O₂ for 10 min. The slides were placed in a humidified staining chamber at room temperature and blocked with normal goat serum (Vector) diluted 1:10 in PBS-Tween 20 for 1 h and then incubated for 1 h with Chk1 antibody (Novocastra; diluted 1:100). The slides were washed in PBS-Tween 20, incubated with post-primary block for 20 min in humidifier chamber, washed again, and then incubated for 30 min with a polymer. The chromogen diaminobenzidine was then applied for a time no longer than 5 min to avoid background signal and the reaction was stopped in water. The sections were counterstained with Mayer's hematoxylin for 10 s, dehydrated, and mounted under coverslips with Eukitt and visualized with a BX60 microscope (Olympus).

In vivo studies. Female athymic NCr-nu/nu mice (6 weeks old) were obtained from Charles River. Mice were maintained under pathogen-free conditions and provided with food and water ad libitum. Procedures involving animals and their care are conducted in conformity with the institutional guidelines that are in compliance with national (D.L. n.116,G.U., suppl.40, 18 Febbraio 1992, Circolare No. 8, G.U., July 14, 1994) and international laws and policies (EEC Council Directive 86/609, OJ L 358,1, December 12, 1987; Guide for the Use of Laboratory Animals, U.S. National Research Council, 1996). Exponentially growing cells of the four clones were expanded in vitro and injected 5 × 10⁶ cells/mouse s.c. Each mouse received on the left flank cells expressing scramble vector and on the contralateral flank cells expressing siRNA against Chk1. Antitumor activity of 5-FU was evaluated in mice either in the presence or in the absence of Tetracycline (Sigma) in the drinking water. Each group consisted of at least six mice. 5-FU (Sigma) was dissolved in sterile water and administrated i.v. at the dose of 50 mg/kg for each treatment (q7d × 4) in a volume of 10 mL/kg body weight. The length (L) and the width (W) of the tumor mass were measured by caliper twice weekly and the tumor volume (TV) was calculated as TV = (L × W²) / 2, being W < L. Animals were sacrificed 30 days after tumor implant and tumors were fixed in formalin 4% for immunohistochemistry.

Statistical analysis. The unpaired Student's t test was used to analyze the in vitro clonogenic assay results, whereas in vivo results were compared by ANOVA analysis. Statistical analysis were done with GraphPad Prism program.

Transient transfection with siRNA against Chk1 in HCT-116 p53^wt and HCT-116 p53^−/−. Previous data from our laboratory have shown that transfection of HCT-116 p53^wt and p53^−/− cells with siRNA against Chk1 was able to completely down-regulate Chk1 protein level and to sensitize cells to ionizing radiation and DDP treatment (24). To understand if abrogation of Chk1 could sensitize cells to the cytotoxic activity of other anticancer agents and if this effect could be amplified in a p53^−/− background, we transiently transfected HCT-116 p53^wt and p53^−/− cells with siRNA against Chk1 and treated with antitumor agents with different mechanism of action (either DNA-damaging and non-DNA-damaging agents) as detailed in Materials and Methods. As shown in Supplementary Fig. S1, a clear down-regulation of the protein in observed with this experimental protocol. Figure 1A and B showed the dose-response curve of mock and siRNA against Chk1-treated cells after treatment with different anticancer agents. A clear sensitization could be observed with 5-FU, VP-16, and doxorubicin, all drugs able to interfere with DNA synthesis. Such sensitization was even more evident in a p53^−/− background, corroborating previous data in which cells with a compromised G₁ checkpoint can be further sensitized to the cytotoxic activity of DNA-damaging agents by abrogating the residual G₂ checkpoint. Interestingly enough, agents such as PS341 and Taxol, not directly interfering with DNA or DNA synthesis, were equally cytotoxic in the presence or absence of Chk1 expression.

Stable transfection with a tetracycline-inducible pSuperior siRNA against Chk1. To get an experimental model that would allow us to translate these data in an in vivo setting, we decided to stably transfect a tetracycline-responsive pSuperior plasmid coding for siRNA against Chk1, as detailed in Material and Methods, in HCT-116^wt and HCT-116 p53^−/− cells. After having cloned the desired sequences in the pSuperior plasmid, this plasmid was transfected in both HCT-116 p53^wt and p53^−/− clones expressing the tetracycline repressor. Different clones were obtained able to show a specific down-regulation of the Chk1 target protein on the addition of doxycycline in the culture medium. pSuperior expressing the scrambles sequences were also constructed and transfected in the same tetracycline repressor-expressing sublines. Four clones were selected for further characterization (one pair for each HCT p53^+/+ and HCT p53^−/− cells expressing scramble and siRNA Chk1 plasmids). As depicted in Fig. 2A, a clear and almost complete down-regulation of Chk1 could be obtained starting from 48 h after the addition of doxycycline and persisted up to 144 h (data not shown) in pSuperior siRNA Chk1-expressing clones, whereas no modification of the protein levels was detected in scrambles expressing clones. These data were also confirmed in a cell-to-cell analysis: indeed, immunofluorescence analysis clearly showed that the down-regulation was uniform in all the cells (Supplementary Fig. S2). The protein down-regulation was specific for Chk1, as the levels of Chk2 protein were not affected by the addition of doxycycline in the culture medium (Fig. 2A).

We have already reported that down-regulation of Chk1 protein in HCT-116 cell line and in its subline devoid of p53 by transient transfection with siRNA against Chk1 did not interfere with cell growth (24). Again, no difference in cell growth could be observed on the addition of doxycycline in both p53^+/+ and p53^−/− cells (Fig. 2B), suggesting that in these cell lines Chk1 is dispensable for cell growth, contrary to what has been reported for other experimental systems (9). In addition, even maintaining the cells to continuous exposure to doxycycline for >2 months with a continuous repression of Chk1, no modification of the cell doubling times was recorded (data not shown) corroborating the fact that in this cell system Chk1 is truly dispensable for cell growth. We also studied the downstream effects of Chk1 ablation in unstressed condition, and as it can be seen in Supplementary Fig. S3, no modification of phosphorylation of Ser²¹⁶ Cdc25C and its total level and Cdc25A levels were observed in both p53^+/+ and p53^−/− cells after 72 h of doxycycline induction, corresponding to a time point where Chk1 protein is ablated. Phosphorylation of Cdc2 on Thr¹⁴ and Tyr¹⁵ was analyzed by Western blot analysis with an antibody recognizing phosphorylated Thr¹⁴ and Tyr¹⁵ in both p53^+/+ and p53^−/− cells transfected with siRNA against Chk1 and a light decreased in the phosphorylation of Cdc2 could be observed in p53^+/+ and p53^−/− cells transfected with siRNA against Chk1 induced with doxycycline, corroborating the fact that Chk1 is regulating the phosphorylating status of Cdc2 as already reported (6). However, in unstressed condition, the observed effect seems to interfere neither with cell cycle distribution or cell growth as suggested by cell growth curves (Fig. 2B) and cell cycle phases distribution (Fig. 3). We also tested the level of geminin, which is not a direct Chk1 target but whose levels vary depending on the cell phase being absent during G₁ phase and accumulating through S, G₂, and M phases, and again no difference in its level could be observed in p53^+/+ and p53^−/− in the presence or absence of doxycycline (Supplementary Fig. S3).

Characterization of the selected inducible clones. Having set up the experimental conditions, we tested the cytotoxic activity of different anticancer agents in the presence or absence of doxycycline (Fig. 4). As an almost complete down-regulation of Chk1 protein levels was always observed at 72 h after the induction with doxycycline (Fig. 2A), this time point was chosen to perform the treatment with cytotoxic agents. We focus on two DNA-damaging agents (5-FU and DDP, able to interact with DNA or DNA synthesis) and on the microtubule-interacting agent Taxol. The activity of DDP was increased in the absence of Chk1 (in doxycycline-inducible clones), and such effect was more evident in p53-deficient cells (Fig. 4A). Indeed, a shift from 25 to 21 μmol/L was observed in p53^+/+ but was more marked in p53^−/− cells (from 17.9 to 4.5 μmol/L). 5-FU sensitization was observed only when both Chk1 and p53 were knocked down (Fig. 4B). The activity of all the tested drugs was not modified in scramble clones on the addition of doxycycline. Taxol cytotoxic activity was not influenced by either the level of p53, Chk1, or both (Fig. 4C).

A detailed cell cycle analysis was done in the selected clones with or without doxycycline in the medium and after 5-FU treatment (5 μmol/L) to test whether Chk1 down-regulation was associated with a different cell cycle perturbation. As it can be seen in Fig. 3, at the end of 24 h treatment with 5-FU in all the clones not induced with doxycycline, a substantial increase of cells in early, mid, and late S phase could be detected compared with control cells; in the presence of doxycycline, which leads to a complete down-regulation of Chk1 levels in clones expressing the siRNA against Chk1, the cell cycle distribution appears similar to cells treated with 5-FU in the absence of doxycycline (compare in Fig. 3A, lane 4 with lane 8). These data would suggest that at this time point 5-FU treatment causes an S-phase slowing and/or delay and the lack of Chk1 in both a p53^+/+ and a p53^−/− background seems to have no effect. When the same analysis is done 24 h later, after having removed 5-FU and drug-free medium was added, p53^+/+ and p53^−/− cells in the absence of doxycycline were recovering from the S phase slowing as suggested by the decrease number of cells in S phase and the increase proportion of cells in G₂-M phase. In the presence of doxycycline, this effect is even more marked (particularly in a p53^−/− background) with DNA histogram profiles very similar to untreated control ones, suggesting a faster exit from S phase in the absence of Chk1 (Fig. 3B, compare lane 4 with lane 8). We have also analyzed the γH2AX induction, herein taken as a marker of DNA double-strand break, in controls and 5-FU-treated cells in the absence or presence of doxycycline. No increase in γH2AX could be detected in all the four clones in the presence and absence of doxycycline (Fig. 3A, compare lane 1 with lane 5) in unstressed condition, again suggesting that in this experimental system the lack of Chk1 does not cause any damage. At the end of 5-FU treatment, a marked increase in γH2AX is observed in the absence of doxycycline (Fig. 3A, compare lane 1 with lane 3) with a decrease at 24 h recovery (Fig. 3B, compare lane 1 with lane 3); a less marked increase of γH2AX could be observed in cells treated with 5-FU in the presence of doxycycline (Fig. 3A and B, compare lane 5 and lane 7). These data would suggest that at 24 h recovery the lack of Chk1 abrogates the S-phase checkpoint induced by 5-FU treatment with a more marked effect in cells with a p53^−/− background, data that correlate with a greater toxic effect of 5-FU treatment when both Chk1 and p53 are missing.

In vivo experiment. Clones were transplanted s.c. in nude mice as described in Materials and Methods. As depicted in Fig. 5A, although p53^−/− derived clones (both scramble and siRNAChk1-expressing cells) grew slower than tumors derived from p53^+/+ cells, no difference in tumor growth could be observed in the presence or absence of tetracycline in the drinking water. The percentage of tumor takes was similar among tetracycline-induced and noninduced groups (data not shown). To test the effective down-regulation of Chk1, we first analyzed, by Western blot analysis, the protein level on whole tumor lysates. Unfortunately, the Chk1 antibodies we tested were not sufficiently specific for human species, as they did recognize also the murine protein (data not shown), not allowing a correct quantification of Chk1 protein levels by this technique. We then decided to set up the conditions for an immunohistochemistry analysis to evaluate the Chk1 in situ expression. As shown in Fig. 5B, whereas Chk1 protein could be detected in slides of tumors from both HCT-116 p53^+/+ and p53^−/− scramble and Chk1 siRNA clone in the absence of tetracycline, the addition of tetracycline did cause a down-regulation of Chk1 protein only in tumor cells overexpressing the pSuperior plasmid coding for siRNA against Chk1 in both a p53^+/+ and a p53^−/− background. Chk1 staining was generally within the nucleus. No modification of the immunohistochemistry pattern was observed in scrambled clones grown in animals supplemented with tetracycline. These immunohistochemistry expression data did validate the experimental clones in an in vivo setting. In addition, as no modification on tumor growth or tumor takes could be observed in the presence and absence of tetracycline, it is suggested that down-regulation of Chk1 in HCT-116 cell system does not interfere also with the in vivo tumor take and growth.

We then checked if the antitumor activity of the anticancer agents could be different in this in vivo model, where we could specifically down-regulate the Chk1 tumor level by adding tetracycline in the drinking water. We choose to perform the experiment only in the p53^−/− HCT-116 cells and to evaluate the activity of 5-FU. 5-FU was selected as it is clinically used in colon cancer tumors and our in vitro data suggested that its activity was potentiated by Chk1 down-regulation in a p53^−/− background. The selected schedule was 5-FU i.v. weekly starting from a palpable tumor mass. This schedule was not very active in this HCT-116 p53^−/− clones transfected with pSuperior and siRNA against Chk1 in the absence of tetracycline (normal Chk1 protein levels); however, a clear activity could be observed in the overexpressing pSuperior siRNA Chk1 tumor in animals receiving tetracycline in the drinking water (Fig. 6). Indeed, a statistically significant inhibition could be observed at day 36 and persisted until the end. A summary of different variables is reported in Table 1.

Cell cycle checkpoints not only ensure the correct progression of the cell cycle but also rapidly activate signal transduction pathways in the presence of damage. Defects due to mutations or aberrant expression of checkpoint proteins are frequently found in human tumors conferring growth advantage and playing a causative role in tumor formation and progression (21).

In the last years, the unraveling of the molecular mechanisms at the basis of the cellular response to anticancer treatment has envisaged the possibility to exploit the tumor checkpoint deficiencies to increase the therapeutic index of chemotherapy. One of the most frequently mutated protein in human tumors is p53 resulting in a nonfunctional G₁ checkpoint in tumor cells. This means that tumor cells with mutated p53 will rely only on S and G₂ checkpoints to cope with a damage. Abrogating the residual G₂ checkpoint should significantly improve the cytotoxic effects of the current used anticancer agents that have been shown to activate the G₁, S, and G₂ checkpoints. The identification of a “druggable” target (the Chk1 kinase) has made possible to test this hypothesis (22). Chk1 has been originally described as a key protein in the G₂ DNA damage checkpoint even if more recent data have clearly shown its role in the S, G₂-M, and spindle checkpoints (9, 10, 19, 27). Also, Chk2 has been described to be involved in the DNA damage signal transduction of the G₁ and G₂-M checkpoints but, differently to Chk1, it has been clearly shown that its lack, caused by both siRNAs or targeting gene disruption, does not sensitize cells to anticancer agents (24, 28).

The experimental evidence that abrogating G₂ checkpoint after anticancer drug treatments increases their cytotoxic activity has been generated using both siRNA against Chk1 and Chk1 inhibitors in system with wild-type and mutated p53 (24, 29, 30). We wanted to set up an experimental system that could be used for long-term experiments and more importantly could be translated in an in vivo setting. Previous data from our laboratory showed that using HCT-116 cellular system it was possible to sensitize cells to ionizing radiation and DDP when both p53 and Chk1 functions were ablated (24). We first extended these data and tested the activity of anticancer agents, with different mechanism of action. We found that down-regulation of Chk1 by siRNA against Chk1 do sensitize HCT-116 p53^wt cells to the DNA-interfering agents 5-FU, doxorubicin, and VP-16. Such effect was greater after 5-FU and VP-16 in a p53^−/− background. However, no potentiation could be observed after treatment with non-DNA-damaging drugs, such as Taxol and PS341, a proteasome inhibitor. The lack of sensitization to Taxol of Chk1-depleted cells is in keeping with recent data obtained on DT40 cells Chk1^−/−, which were shown to be more resistant to this treatment than the wild-type DT40 cell line. The Taxol resistance in DT40 Chk1^−/− cells was correlated with the lack of apoptosis induction following drug treatment (19). It was, however, reported that Chk1-depleted HT299 and HeLa cells were 10-fold more sensitive to Taxol treatment than control cells. This effect was partially explained by the fact that inhibition of Chk1 caused a facilitated M-phase entry, an inhibition of mitotic exit and a stronger activation of caspase-8-mediated apoptosis (31). We have not studied in detail in our system the activity of Taxol in the presence or absence of Chk1, but it might be that different cell lines and different experimental conditions can account for these discrepancies.

As the system did behave as predicted, we worked to obtained a tetracycline-inducible HCT-116 system (in a p53^+/+ and p53^−/− background) where on the addition of tetracycline it could be possible to simultaneously induce the synthesis of two siRNAs directed against Chk1. This approach has been already successfully applied. Different stable transfected clones were obtained and a clear and persistent down-regulation of Chk1 protein could be obtained after doxycycline addition in the culture medium. The down-regulation was observed both by Western and immunofluorescence techniques. Again, no modification of cell growth was found either in HCT-116 cell line p53^+/+ or in its subline p53^−/− after down-regulation of Chk1 protein. More importantly, the continuous exposure of cells to doxycycline for >2 months did not alter the growth rate of the cells. These data are important as in somatic cells, the lack of Chk1 has been reported to yield different results. Chk1-deficient mice are embryonically lethal and Chk1-deficient mammary gland cells undergo apoptosis when induced to proliferate (9, 32). The Chk1 gene has been successfully knocked down in an DT40 cell line with no great effect on cell growth, with only a slight increase in apoptosis rate and slowdown of the cell growth observed (10). In short-term assays, the use of siRNA against Chk1 rarely interfered with cell growth in many different tumor cell lines (29), even if with some exceptions. Indeed, in U2OS cells with Chk1 depleted, an aberrant increased initiation of DNA replication coupled with generation of DNA strand breaks and H2AX phosphorylation correlated with a potent inhibition of cell growth (33, 34).

When the activity of DNA-damaging and non-DNA-damaging agents was tested in the selected clones in the absence and in the presence of doxycycline, again it was found that a sensitization could be observed in a p53^−/− and Chk1^−/− background after DDP and 5-FU, whereas Taxol activity was not influenced by the presence of p53 and Chk1. The analysis of cell cycle perturbation after 5-FU treatment suggests that 24 h after the end with 5-FU an abrogation of S-phase checkpoint is observable and is much more marked in a p53^−/− and Chk1^−/− background. The observed effects are different from the ones reported by other studies assessing the effects of 5-FU treatment together with both pharmacologic (35) and siRNA inhibition of Chk1 (36) and in the DT40 cellular system Chk1^−/− (37). In the former cases, abrogation of Chk1 after 5-FU treatment caused a rapid onset of apoptosis without cell cycle progression specifically in cells with a S-phase DNA content or a progression to early M phase and induction of apoptosis, whereas in the latter cells undergo an impeded replication in the presence of 5-FU and then were completely blocked in the subsequent G₁ phase for a long time. All these data reinforce the fact that Chk1 abrogation can vary according to the experimental system used.

The system has been successfully translated in vivo. Tumors xenografted in mice showed a Chk1 down-regulation following induction of tetracycline in drinking water. The lack of Chk1 interfered neither with tumor takes nor with tumor growth, again corroborating the finding that in this cellular system Chk1 is dispensable for growth. More importantly, treating mice with 5-FU, a clear activity, expressed as tumor growth delay, could be observed in those tumors lacking both p53 and Chk1. These findings clearly show that in vivo, particularly in tumors not expressing p53, Chk1 depletion increases the therapeutic index of 5-FU. The possibility to increase the antitumor activity with no increased toxicity is attractive and can be exploited in clinic by combining specific Chk1 synthetic inhibitors with anticancer agents such as 5-FU in tumors lacking functional p53.

No potential conflicts of interest were disclosed.

We thank the Italian Association for Cancer Research for the generous contribution.