Disruption of the checkpoint kinase 1/cell division cycle 25A pathway abrogates ionizing radiation-induced S and G₂ checkpoints

HeLa cells were grown in DMEM (GIBCO) supplemented with 10% FBS/100 units/ml penicillin and streptomycin/1 mM glutamine (culture medium). Normal human diploid fibroblasts (American Type Culture Collection, WI 30) were grown in minimum essential medium containing 10% FBS, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 2 mM glutamine, 100 units/ml penicillin and streptomycin.

Chk1 was detected with monoclonal (G-4) or polyclonal (FL-476) antibodies (Santa Cruz Biotechnology). Chk2 was detected with a monoclonal antibody from Neomarkers (Fremont, CA). Endogenous Cdc25A was immunoprecipitated and immunoblotted with Ab-3 (Neomarkers). Human Cdc25C and Cdc25B were detected with antibodies sc-326 and -327, respectively (Santa Cruz Biotechnology), and a polyclonal antibody was used to detect C-TAK1 (20). A polyclonal antibody was used to immunoprecipitate human cyclin B1 (21). Phospho-histone H3 antibodies were purchased from Upstate Biotechnology (Lake Placid, NY). Bound primary antibodies were detected with either horseradish peroxidase (HRP)-goat anti-mouse antibody (ICN/Jackson ImmunoResearch) or HRP-goat anti-rabbit antibody (Zymed), and proteins were visualized by chemiluminescence by using the enhanced chemiluminescence reagent (Amersham Pharmacia). Antibodies specific for human Cdc25A phosphorylated on serine-123 were generated by immunization of rabbits with the phosphopeptide CLKRSH-pS-DSLD coupled to keyhole limpet hemocyanin. Cells were lysed in mammalian cell lysis buffer (50 mM Tris⋅HCl, pH 8/150 mM NaCl/0.5% Nonidet P-40/5 mM EDTA/2 mM DTT/1 mM sodium orthovanadate/10 mM β-glycerophosphate/1 mM NaF/1 μM microcystin/1 mM PMSF/5 μg/ml leupeptin/10 μg/ml aprotinin).

RNAi was performed by using protocols supplied by Dharmacon Research (Lafayette, CO). The small interfering RNA (siRNA) oligonucleotide corresponding to nucleotides 127–147 of the human Chk1 coding region and nucleotides 82–102 of human Cdc25A were supplied by Dharmacon Research. siRNA duplexes were transfected into HeLa cells by using OLIGOFECTAMINE (Invitrogen) according to the manufacturer's recommendations. Unless indicated, cells were analyzed 36 h posttransfection. In some cases, cells were exposed to 6 or 10 Gy IR from a ⁶⁰Co source. Luciferase–siRNA was a gift from D. Dean (Washington University).

HeLa cells were mock-transfected or transfected with duplex siRNAs. Thirty-six hours later, cells were mock-irradiated or exposed to 6 Gy IR from a ⁶⁰Co source. Cells were stained with 30 μg/ml of propidium iodide and for phospho-histone H3 1 h later. The procedure for immunofluorescent detection of phosphorylated histone H3 has been described (22). Cells were analyzed by flow cytometry by using a Becton–Dickinson FACScan. The data were analyzed by using cellquest software. Chromosome spreading was performed as described (23). A minimum of 500 nuclei were counted for each sample and experiments were performed three times.

Cells were lysed in mammalian cell lysis buffer and incubated with cyclin B1-specific antibody (21). Precipitates were washed twice in incomplete kinase buffer (50 mM Tris⋅HCl, pH 7.4/10 mM MgCl₂/1 mM DTT/10 mM β-glycerophosphate) and then incubated in 50 μl of complete kinase buffer (50 mM Tris, pH 7.4/10 mM MgCl₂/1 mM DTT/20 mM β-glycerophosphate/50 μm ATP/10 μCi [γ-³²P]) containing 1 μg of histone H1 for 10 min at 30°C. ³²P incorporation into histone H1 was quantified by using the STORM 860 imaging system (Amersham Pharmacia).

Human diploid fibroblasts were incubated with 1 μM UCN-01 for 30 min and then either mock-irradiated or exposed to 6 Gy IR. Cells were harvested for Western blotting 1 h later. HeLa cells that had been transfected with Chk1-siRNA for 36 h were exposed to 6 Gy IR. Cycloheximide (25 μg/ml) was added, and cells were harvested at the indicated times.

Inhibition of DNA synthesis after irradiation was assessed as described (24) except for the following modifications. Cells were incubated with siRNA in the presence of 10 nCi/ml [¹⁴C]thymidine (NEN) for 24 h, then washed and incubated in DMEM for 12 h. Cells were either mock-irradiated or exposed to 5, 10, or 20 Gy IR, incubated for 1 h, and then pulse-labeled with 2.5 μCi/ml [³H]thymidine (NEN) for 15 min. Cells were harvested, washed twice in PBS, and then incubated in 100% methanol for 5 min, followed by 10% trichloroacetic acid for 5 min, followed by 0.3 N NaOH. After neutralization with HCl, radioactivity was determined by liquid scintillation counting. The resulting ratios of ³H cpm to ¹⁴C cpm, corrected for cpm that resulted from channel crossover, were a measure of DNA synthesis.

HeLa cells infected with recombinant adenovirus encoding myc-Cdc25A were incubated for 2 h in phosphate-free DMEM containing 2 mCi per ml of ³²P-labeled inorganic phosphate. Cdc25A was immunoprecipitated with anti-myc monoclonal antibody-agarose (Santa Cruz Biotechnology). GST-Chk1 was purified from baculoviral-infected Sf9 cells and 6xHis-Cdc25A and 6xHis-Cdc25A(S123A) were purified from Escherichia coli. Kinase reactions were carried out in 50 μl of complete kinase buffer for 10 min at 30°C. Proteins were resolved by SDS/PAGE, and radiolabeled Cdc25A was subjected to digestion with 2 μg/ml trypsin followed by 1 μg/ml chymotrypsin. Two-dimensional phosphopeptide mapping was performed as described (14).

HeLa cells were mock-transfected or transfected with duplex siRNAs. Thirty-six hours later, cells were pulse-labeled for 20 min with 20 μM BrdUrd (Amersham Pharmacia). Cells were stained with 30 μg/ml propidium iodide and for BrdUrd with FITC-conjugated monoclonal antibody (Caltag, South San Francisco, CA) and processed for flow cytometric analysis at the times indicated.

RNAi experiments were performed to determine whether Chk1 levels could be lowered in HeLa cells. As seen in Fig. 1A, Chk1 protein levels were significantly reduced in cells that had been incubated with Chk1-specific siRNAs (lanes 3 and 6). Chk1 levels were not reduced in cells treated with luciferase–siRNA (Fig. 1A, lanes 2 and 5), and C-TAK1 (20, 25) levels were not altered by RNAi treatment. Next, RNAi experiments were performed to determine whether Chk1 is required for the IR-induced G₂ checkpoint (Fig. 1B). The ability of G₂ cells in control and Chk1-siRNA-treated cultures were monitored for their ability to enter into mitosis within 1 h after IR. Cells were costained with propidium iodide to assess DNA content and with antiphosphohistone-H3 antibody to identify mitotic cells (22, 26). This experimental paradigm determines the percentage of G₂ cells that move into mitosis in the absence and presence of double-stranded DNA breaks (22, 26). Control (Fig. 1B) and luciferase–siRNA-treated cells (Fig. 8, which is published as supporting information on the PNAS web site) showed a 74% and 70% decrease, respectively, in the number of mitotic cells 1 h after IR. In contrast, Chk1-siRNA-treated cells showed only a 30% reduction in mitotic cells after IR. Chromosome spreading gave similar results to those obtained by phosphohistone–H3 staining, and histone H1 kinase assays revealed that Cdc2/cyclin B1 complexes were still active after exposure of Chk1-siRNA-treated cells to IR (see Figs. 9 and 10, which are published as supporting information on the PNAS web site). Cells expressing kinase-inactive Chk1 or a phosphorylation-site mutant (S315A, S345A) that fails to be activated in a checkpoint-dependent manner (14) were also compromised in their ability to delay in G₂ after IR (data not shown). Taken together these results demonstrate that Chk1 is required for the IR-induced G₂ checkpoint. Chk1-siRNA-treated cells experienced an S-phase delay, indicating an important S-phase function for Chk1 even in the absence of IR (data not shown).

Chk1 reportedly phosphorylates Cdc25A and Cdc25C (15, 17, 19). The fastest electrophoretic form of Cdc25A was observed to accumulate in Chk1-deficient cells suggesting a reduction in Cdc25A phosphorylation in these cells (denoted by an asterisk in Fig. 2A). Irradiation of control cells resulted in the loss of Cdc25A (Fig. 2A, compare lanes 1 and 3). In contrast, Cdc25A remained stable when Chk1-deficient cells were irradiated (Fig. 2A, compare lanes 2 and 4). Cycloheximide treatment confirmed that loss of Chk1 resulted in stabilization of Cdc25A protein after IR (see Fig. 11, which is published as supporting information on the PNAS web site). A reduction in the slower electrophoretic forms of Cdc25B was also observed in Chk1-siRNA-treated cells. Cdc25C phosphorylation status was not altered in Chk1-deficient cells and C-TAK1 was unaffected by loss of Chk1. RNAi technology was also used to determine whether Cdc25A levels could be reduced in Chk1-deficient cells. As seen in Fig. 2A (lane 6), we were able to reduce but not eliminate Cdc25A in Chk1-deficient cells. The partial reduction of Cdc25A protein observed in cells cotransfected with Chk1- and Cdc25A-siRNAs could be due to the longer half-life of Cdc25A in Chk1-deficient cells or to less efficient elimination of RNA in doubly transfected cells. However, Chk1 levels were reduced to similar levels in singly and doubly transfected cells

As an independent approach to determine whether Cdc25A is regulated by Chk1 after IR, we treated normal human diploid fibroblasts with the Chk1 inhibitor UCN-01 (10–12, 27). As seen in Fig. 2B, the fastest electrophoretic form of Cdc25A accumulated in cells where Chk1 activity was blocked by UCN-01 (lanes 2 and 4), which occurred in both the absence and presence of IR. Chk2 has been reported to phosphorylate Cdc25A after IR, thereby targeting Cdc25A for ubiquitin-mediated proteolysis (28). As seen in Fig. 2C, Chk2 levels were unaltered in Chk1-siRNA-treated cells, and Chk2 was fully phosphorylated when Chk1-siRNA-treated cells were exposed to IR (lane 4). Thus, in the absence of Chk1, Chk2 is unable to induce the degradation of Cdc25A in an IR-dependent manner.

Loss of Chk1 interferes with the ability of cells to eliminate Cdc25A after IR (Fig. 2), and loss of Chk1 causes bypass of the IR-induced G₂ checkpoint (Fig. 1). To determine whether elevated levels of Cdc25A contributed to checkpoint abrogation in Chk1-deficient cells, we tested whether the IR-induced G₂ checkpoint could be restored by reducing Cdc25A levels. Reduction in Cdc25A levels was observed in cells transfected with Cdc25A-siRNA (Fig. 3A, lane 3). As noted earlier, the fastest electrophoretic form of Cdc25A accumulated in cells transfected with Chk1-siRNA (Fig. 3A, lane 2), and this form was stable after irradiation (Fig. 3A, lane 6). We were able to reduce but not eliminate Cdc25A in irradiated Chk1-deficient cells by Cdc25A-siRNA cotransfection (Fig. 3A, lane 8). Cdc25C levels and phosphorylation status were not altered under any of the experimental conditions. Next, the ability of these cells to delay in G₂ after IR was examined (Fig. 3B). Fewer mitotic cells were observed in Cdc25A-siRNA-treated cells relative to control cells. Control cells and cells transfected with Cdc25A-siRNA showed an ≈80 and an ≈70% decrease, respectively, in the number of mitotic cells after IR, whereas cells transfected with Chk1-siRNA showed only a 22% decrease. Importantly, cells transfected with siRNAs for both Chk1 and Cdc25A partially regained their ability to delay in G₂ as a 50% decrease in the number of mitotic cells was observed after IR (Fig. 3B). Thus, lowering Cdc25A levels in Chk1-deficient cells is sufficient to partially restore G₂ cell cycle delay after IR. The incomplete restoration of G₂ delay in Cdc25A-siRNA-treated cells may be due to the remaining Cdc25A protein (Fig. 3A, lane 8) or to other pathways that are deregulated in the absence of Chk1.

To determine whether Chk1 is also required for the IR-induced S-phase checkpoint, [³H]thymidine incorporation into DNA was monitored 1 h after exposure of cells to various doses of IR (Fig. 4). Control cells had an intact checkpoint as indicated by the dose-dependent decrease in [³H]thymidine incorporation into newly synthesized DNA. In contrast, cells treated with Chk1-siRNA continued to initiate DNA replication as indicated by the radio-resistant DNA synthesis at all doses of IR. Furthermore, cyclin E-associated kinase activity was not reduced in Chk1-siRNA-treated samples after IR (data not shown). Importantly, treatment of Chk1-deficient cells with Cdc25A-siRNA partially rescued the IR-induced S-phase checkpoint.

The fastest electrophoretic form of Cdc25A accumulates in Chk1-deficient cells (Figs. 2A and 3A). Labeling experiments were carried out in vivo to monitor the phosphorylation status of this form of Cdc25A. As seen in Fig. 5A, multiple-phosphorylated forms of Cdc25A that run as a smear on SDS gels were observed in control cells (lane 1). In contrast, a faster electrophoretic form of Cdc25A was seen in Chk1-deficient cells, and this form was poorly phosphorylated in vivo (Fig. 5A, lane 2). These results suggest that Chk1 either directly or indirectly regulates the phosphorylation of Cdc25A in vivo. Phosphorylation of Cdc25A on S123 facilitates the proteasome-dependent degradation of Cdc25A (19, 28). We generated an antibody specific for Cdc25A when it is phosphorylated on serine-123 (Fig. 5B) and used this antibody to determine whether Cdc25A is phosphorylated on S123 in the absence of DNA damage (Fig. 5C). Lysates prepared from asynchronously growing HeLa cells were resolved directly by SDS/PAGE (Fig. 5C, lane 1) or were first precipitated with protein A beads alone (Fig. 5C, lane 2) or with antibody specific for Cdc25A (Fig. 5C, lane 3). Western blotting was performed with antibodies specific for Cdc25A (Fig. 5C, Upper) or with the S123 phospho-specific antibody (Fig. 5C, Lower). As seen in Fig. 5C, Cdc25A was phosphorylated on S123 in the absence of checkpoint activation and this likely contributes its instability during an unperturbed cell cycle. Next, labeling experiments were carried out in vitro and in vivo to determine whether Cdc25A is a substrate of Chk1 and whether Chk1 directly phosphorylates Cdc25A on serine-123. Chk1 phosphorylated Cdc25A on S123 in vitro (Fig. 5, D and E) and Cdc25A was phosphorylated on S123 in vivo (Fig. 5F). The phosphorylation pattern of Cdc25A labeled in vivo (Fig. 5F) was remarkably similar to that of Cdc25A phosphorylated by Chk1 in vitro (Fig. 5D), suggesting that Chk1 is a key Cdc25A-kinase in vivo.