Research Inactivation of DNA-Dependent Protein Kinase Leads to … · ing the nonhomologous end joining pathway of DNA double-strand breaks repair. Here, we showed another important

Published OnlineFirst April 20, 2010; DOI: 10.1158/0008-5472.CAN-09-3362

Therapeutics, Targets, and Chemical Biology
Cancer
Research

Inactivation of DNA-Dependent Protein Kinase Leadsto Spindle Disruption and Mitotic Catastrophe withAttenuated Checkpoint Protein 2 Phosphorylationin Response to DNA Damage

Zeng-Fu Shang, Bo Huang, Qin-Zhi Xu, Shi-Meng Zhang, Rong Fan, Xiao-Dan Liu, Yu Wang, and Ping-Kun Zhou

Abstract

Authors' ABeijing Inst

Note: SupResearch O

Z-F. Shang

CorresponMedicine, 2861068213

doi: 10.115

©2010 Am

www.aacr

Downlo

DNA-dependent protein kinase catalytic subunit (DNA-PKcs) is well known as a critical component involv-ing the nonhomologous end joining pathway of DNA double-strand breaks repair. Here, we showed anotherimportant role of DNA-PKcs in stabilizing spindle formation and preventing mitotic catastrophe in responseto DNA damage. Inactivation of DNA-PKcs by small interfering RNA or specific inhibitor NU7026 resultedin an increased outcome of polyploidy after 2-Gy or 4-Gy irradiation. Simultaneously, a high incidence ofmultinucleated cells and multipolar spindles was detected in DNA-PKcs-deficient cells. Time-lapse videomicroscopy revealed that depression of DNA-PKcs results in mitotic catastrophe associated with mitoticprogression failure in response to DNA damage. Moreover, DNA-PKcs inhibition led to a prolonged G2-Marrest and increased the outcome of aberrant spindles and mitotic catastrophe in Ataxia-telangiectasiamutated kinase (ATM)–deficient AT5BIVA cells. We have also revealed the localizations of phosphorylatedDNA-PKcs/T2609 at the centrosomes, kinetochores, and midbody during mitosis. We have found that theassociation of DNA-PKcs and checkpoint kinase 2 (Chk2) is driven by Ku70/80 heterodimer. Inactivation ofDNA-PKcs strikingly attenuated the ionizing radiation–induced phosphorylation of Chk2/T68 in both ATM-efficient and ATM-deficient cells. Chk2/p-T68 was also shown to localize at the centrosomes and midbody.These results reveal an important role of DNA-PKcs on stabilizing spindle formation and preventing mitoticcatastrophe in response to DNA damage and provide another prospect for understanding the mechanismcoupling DNA repair and the regulation of mitotic progression. Cancer Res; 70(9); 3657–66. ©2010 AACR.

Introduction

Spindle checkpoint and formation stability are as impor-tant as other DNA damage checkpoints and repair mechan-isms in maintaining genomic stability and preventingcarcinogenesis (1). Centrosomes, the major microtubule or-ganizing centers in mammalian cells, play a crucial role indirecting the formation of the spindle poles during mitosis(2). In mammalian cells, centrosome duplication begins inS phase, and the two daughter centrosomes form the polesof the mitotic spindle. Immediately after mitotic exit, eachdaughter cell contains a single centrosome. The regulation

ffiliation: Department of Radiation Toxicology and Oncology,itute of Radiation Medicine, Beijing, P.R. China

plementary data for this article are available at Cancernline (http://cancerres.aacrjournals.org/).

and B. Huang contributed equally to this work.

ding Author: Ping-Kun Zhou, Beijing Institute of Radiation7 Taiping Road, Haidian District, Beijing, 100850, China. Phone:912; Fax: 861068183899; E-mail: [email protected].

8/0008-5472.CAN-09-3362

erican Association for Cancer Research.

journals.org

on July 16, 202cancerres.aacrjournals.org aded from

of centrosome duplication in mammalian cells is tightly con-trolled to maintain genomic integrity and accurate chromo-some segregation to the daughter cells (3). Centrosomeamplification is tightly associated with the formation of mul-tipolar spindles and frequently occurred in various humancancers (4–6), and it is assumed to contribute to transforma-tion or is a consequence of cancer progression. Centrosomeamplification may result from defective cell division andmultinucleation, leading to amplification of centrosomenumber during subsequent cell cycles (7–9), and can alsoarise in the cells with direct dysregulation of the centrosomeduplication cycle or when G2-M checkpoint fails and en-trance into mitosis occurs in the presence of DNA damage(10, 11). Because centrosomes usually nucleate microtubules,supernumerary centrosomes may result in multipolar spin-dles in cells, which lead to abnormal chromosome segrega-tion, and may generate cells with multiple micronuclei orbinucleated giant cells and eventually cause cells undergoingmitotic catastrophe (12, 13). This lethal phenotype was alsoshown in mammalian cells to result from premature mitosis,which was presumably related to a premature activation ofcyclin-dependent kinase-1 (14, 15) or a failure to completemitosis (16, 17). Mitotic catastrophe can be induced by DNAdamage agents, e.g., ionizing radiation (IR; refs. 17, 18),

3657

0. © 2010 American Association for Cancer Research.

http://cancerres.aacrjournals.org/

Shang et al.

3658


histone acetyltransferase depleting agents (19), and microtu-bule disruption agents (20). So, targeting cancer cells to un-dergo mitotic catastrophe is also considered to be a newefficacy strategy to overcome the drug or radiation resis-tances in cancer therapy (20, 21).The molecular and genetic mechanisms underlying the de-

velopment of centrosome amplification or multipolar spin-dles and mitotic catastrophe in the mammalian cells arenot fully defined, and a commonly accepted definition of mi-totic catastrophe is still absent currently, although some re-ports suggest that it shares several biochemical hallmarks ofapoptosis, in particular mitochondrial membrane permeabi-lization and caspase activation (22). Inactivation of a certainof cell cycle checkpoint related proteins, e.g., BRCA1, check-point kinase 2 (Chk2; ref. 23), 14-3-3δ (24), BRCA2-interactingprotein (25), and Polo-like kinase 1 (Plk1; ref. 26), can facili-tate the induction of mitotic catastrophe.DNA-dependent protein kinase catalytic subunit (DNA-

PKcs) is a member of a subfamily of proteins containing aphosphoinositol 3-kinase domain with the activity of a ser-ine/threonine protein kinase (27). DNA-PKcs is required forthe nonhomologous end joining (NHEJ) pathway of DNAdouble-strand breaks (28), V(D)J recombination of immuno-globulin genes and T-cell receptor genes (27), and telomerelength maintenance (29). More recently, DNA-PKcs has beenshown to contribute to the G2 checkpoint in response to IR(30). DNA-PKcs has also been shown to phophorylate a num-ber of Ataxia-telangiectasia mutated kinase (ATM) substratesin DNA damage response, e.g., KRAB-associated protein 1and H2AX (31). In the present study, using DNA-PKcs smallinterfering RNA (siRNA) and/or specific inhibitor NU7026to suppress DNA-PKcs in HeLa cells and ATM-deficientAT5BIVA cells, we identified an important role of DNA-PKcs in maintaining the normal spindle formation or cen-trosome stability in response to DNA damage. Inactivationof DNA-PKcs results in the phenotype of multipolar spin-dles and mitotic catastrophe upon DNA damage, which isassociated with a decreased radiation-induced phosphoryla-tion of Chk2/T68. The localizations of phosphorylatedDNA-PKcs and Chk2/T68 in centrosomes and midbodyhave also been shown. These observations may have sig-nificance in the elucidation of the mechanisms of spindlestabilization and mitotic catastrophe in response to DNAdamage as well as in the development of new radiationtherapeutic approaches.

Materials and Methods

Cell culture and irradiation. HeLa, HeLa-NC, HeLa-H1,and ATM-deficient AT5BIVA cells were maintained inDMEM containing 10% fetal bovine serum, 100 units/mLpenicillin, and 100 μg/mL streptomycin in a humidified in-cubator at 37°C in 5% CO2. HeLa-H1 and HeLa-NC weregenerated from HeLa cells by stably transfecting with aDNA-PKcs–specific siRNA construct targeting the catalyticmotif (nucleotides 11637-11655, H1) and a control siRNA con-struct (NC), respectively (32). A cobalt-60 γ-ray source was

Cancer Res; 70(9) May 1, 2010

on July 16, 202cancerres.aacrjournals.org Downloaded from

used to irradiate the cells at a dose rate of 1.74 Gy/min atroom temperature. After irradiation, the cells were harvestedeither immediately or at an indicated time of postirradiationculture and then subjected to further experiments.Antibodies. All of the antibodies are commercially available:

anti-DNA-PKcs total (sc-9051, H163, Santa Cruz), anti-phosphor-ylatedDNA-PKcs (pT2609; 19716, Rockland), anti-phosphorylatedDNA-PKcs (pT2647; ab61045, Abcam), anti-phosphorylatedDNA-PKcs (pS2056; ab18192, Abcam), anti-CHK2 (2662, CellSignaling Technology), anti-phosphorylated CHK2 (T68;2661, Cell Signaling Technology), anti-Plk1 (ab17057, Abcam),anti-ATM (Cell Signaling Technology), anti-Ku70 (ab3114,Abcam), anti-γ-tubulin (ab11317-Rabbit, ab11316-Mouse,Abcam), anti-α-tubulin (ZM-0438, Invitrogen), anti-β-actin(I-19-R, Santa Cruz), horseradish peroxidase (HRP)–conjugatedanti-rabbit IgG(H+L) (ZB-2301, Zhongshan), and HRP-conjugated anti-mouse IgG(H+L) (ZB-2305, Zhongshan).siRNAs and transfection. For the knockdown of Ku70 and

Ku80 expression, the siRNA molecules used are adopted fromprevious reports: annealed Ku70 siRNAs sense strand 5′-GGAAGAGAUAGUUUGAUUUdTdT-3′, antisense strand 5′-AAAUCAAACUAUCUCUUCCTG-3′ (33); Ku80 siRNAs sensestrand 5′-ACAAAAUCCAGCCAAGUUCdTdT-3′, antisensestrand 5′-GAACUUGGCUGGAUUUUGUTG-3′ (34); and thesilencer negative control siRNA sense strand 5′-UUCUCC-GAACGUGUCACGUTT-3′ (32). The siRNA molecules weresynthesized and purified by Shanghai GenePharma Co.;siRNA transfection was performed using Lipofectamine2000 reagent according to the procedures reported (32).The final acting concentration for each siRNA is 50 nmol/L.Cell cycle analysis by flow cytometry. After γ-ray irradia-

tion, the cells were harvested either immediately or at theindicated time of postirradiation culture and fixed with75% ethanol. The cells were resuspended in PBS plus 0.1%saponin and 1 μg/mL RNase A (Sigma), incubated for 20 minat 37°C, and stained with 25 μg/mL propidium iodide (Sigma).The cell cycle distribution was evaluated by flow cytometry,counting >10,000 cells per sample.Confocal immunofluorescence microscopy. Cells were

grown on poly-D-lysine–coated culture slides (BD Pharmin-gen), washed in PBS, fixed in PBS containing 4% or 0.5%paraformaldehyde for 15 minutes, and permeabilized in Tri-ton buffer (0.5% Triton X-100 in PBS). The fixed cells wereblocked in blocking solution (2% bovine serum albumin,0.1% Tween, PBS) for 30 minutes at 37°C in a humidifiedchamber. Immunostaining was performed using an anti-phosphorylated DNA-PKcs (pT2609 or pT2647) antibody(1:100), an anti-phosphorylated CHK2/T68 antibody (1:200),an anti-Plk1 antibody (1:200), an anti-α-tubulin antibody(1:200), or a mouse/rabbit anti-γ-tubulin antibody (1:1,000)for 2 hours at room temperature in a humidified chamberand washed thrice in blocking buffer. The cells were incu-bated with anti-mouse/rabbit Rhodamine (1:200) and anti-mouse/rabbit FITC (1:200) secondary antibodies. DNAswere stained with 4′,6-diamidino-2-phenylindole (DAPI) inmounting solution. Confocal immunofluorescence micro-scopy was performed using an LSM 510 laser-scanningconfocal microscope (Zeiss).

Cancer Research



DNA-PKcs Is Required for Spindle Stability


Coimmunoprecipitation and immunoblotting assay. Forgeneral cell lysis and coimmunoprecipitation products ofCHK2 and DNA-PKcs, one experiment used an anti-CHK2antibody to coimmunoprecipitate DNA-PKcs, and anotherexperiment used an anti-DNA-PKcs antibody to coimmuno-precipitate from HeLa cell extracts. The coimmunoprecipita-tions were performed using the immunoprecipitation kit(Protein A/G, Roche Molecular Biochemicals) according tothe manufacturer's instructions. Briefly, cells were washedtwice with ice-cold PBS and collected by centrifugation.The cell pellets were resuspended in prechilled lysis buffer[50 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl, 1% NonidetP40, 0.5% sodium deoxycholate, and the requisite amount ofthe protease inhibitor tablet provided in the kit] and homog-enized. The supernatants were collected by centrifugation at12,000 × g for 10 minutes at 4°C to remove debris and thensubjected to immunoprecipitation. After being preclearedwith protein A/G–agarose, the supernatants were incubatedfor 3 hours with 2 μg of the anti-DNA-PKcs antibody or anti-CHK2 antibody at 4°C, followed by an overnight incubationwith protein A/G–agarose at 4°C. The immunoprecipitateswere collected by centrifugation, washed twice with washbuffer 1 [50 mmol/L Tris-HCl (pH 7.5), 500 mmol/L NaCl,1% Nonidet P40, and 0.05% sodium deoxycholate] and oncewith wash buffer 2 [10 mmol/L Tris-HCl (pH 7.5), 0.1% Non-idet P40, and 0.05% sodium deoxycholate]. The immunopre-cipitates were denatured by heating to 100°C for 3 minutes ingel-loading buffer and centrifuged at 12,000 × g for 20 s toremove the protein A/G–agarose.For Western blots, the above coimmunoprecipitation pro-

ducts were denatured, resolved by SDS-PAGE, and subjectedto Western blot analyses. Otherwise, the cells were harvestedand washed twice in ice-cold PBS. Cell pellets were treatedwith lysis buffer indicated above (one protease inhibitorcocktail tablet in a 50-mL solution), and total protein wasisolated. The protein (50 μg) was resolved using SDS-PAGE(8%) and then transferred onto a polyvinylidene difluoridemembrane for Western blot analysis.Fluorescence time-lapse imaging. To analyze the dynam-

ics of mitosis in DNA-PKcs–deficient HeLa-H1 cells after 4-Gy irradiation, the culture dish was incubated in a chamberwith 5% CO2 for the indicated period of time and placed onthe stage of an inverted Andor Revolution xD Live cell con-focal fluorescence microscope equipped with a YokogawaCSU-X1 spinning disc confocal scanner, iXON EM CCD cam-era (512 × 512 pixel), and 488-nm and 561-nm laser lines. Tofacilitate the in vivo time-lapse analysis, HeLa-H1 cells werecotransfected with H2B-GFP and HSPC300-RFP expressionplasmids to visualize nuclear and cell morphology, respec-tively. Twenty-four hours posttransfection, the cells were ir-radiated with 4 Gy as described above and imaged 24 hourspostirradiation. The microscope view fields were recorded at60× magnification every 2 minutes for 20 hours. The imagesfrom the z-stacks were combined in a maximum projection,and the time series of image projections were exported to atime-lapse video with eight frames per second. The videosand still images taken from the videos were processed usingAndor IQ confocal software.

www.aacrjournals.org


Results

Inactivation of DNA-PKcs results in aberrant spindlesand multinucleated cells upon DNA damage. We have first-ly observed an increased proportion of polyploidy cells and aprolonged G2-M arrest in the siRNA-mediated DNA-PKcsknockdown HeLa-H1 cells after 4 Gy γ-ray irradiation in flowcytometric analysis (Fig. 1B; Supplementary Fig. 1). This phe-notype of increased induction of polyploidy upon DNA dam-age has also been shown in DNA-PKcs–deficient M059J cellsand HepG2-H1 cells, another siRNA-mediated DNA-PKcsknockdown cell line derived from HepG2 cells (data notshown). Furthermore, the inactivation of DNA-PKcs by its spe-cific inhibitor NU7026 also resulted in an increased proportionof polyploidy cells in response to 4-Gy irradiation (Fig. 1B; Sup-plementary Fig. S1). A prolonged G2-M arrest and increasedinduction of polyploidy was further observed in DNA-PKcsknockdown HeLa-H1 cells compared with control HeLa-NCcells at 2 Gy of lower dose irradiation (Supplementary Fig. S2).We then investigated the stability of spindle construction

in response to DNA damage. The centrosomes and spindlestructures were detected in DNA-PKcs–deficient HeLa-H1cells and control HeLa-NC cells after 4-Gy irradiation usingconfocal immunofluorescence microscopy. By staining withantibodies against α-tubulin and γ-tubulin, the spindles werevisualized, and multicentrosomes or aberrant spindle pat-terns, as exemplified by asymmetrical dipolar, tripolar, or tet-rapolar spindles, and centrosome breakage were detected inirradiated HeLa-H1 cells (Fig. 1C). A significantly increasedproportion of multipolar spindle cells was observed inHeLa-H1 cells 24 to 48 hours postirradiation, which took>60% of the mitotic HeLa-H1 cells 48 hours postirradiation(Fig. 1D). Even without irradiation, an increased proportionof mitotic cells with aberrant spindles was also displayed inDNA-PKcs–deficient HeLa-H1 cells. However, no obviouscentrosome amplification during the interphase wasobserved in DNA-PKcs–deficient cells (data not shown).Suppression of DNA-PKcs sensitizes HeLa cells to mitotic

catastrophe induced by IR. After staining with an antibodyagainst α-tubulin and staining nuclear DNA with Hochest33258, a vast pattern of aberrant nuclears were observed,including micronuclei, and multinucleated giant cells' multi-polar spindles in irradiated HeLa-H1 cells (Fig. 2A). Theproportion of multinucleated cells increased with timepostirradiation and reached ∼35% 48 hours postirradiation(Fig. 2B).The formation of multinuclear cells and multipolar spindle

in cells after DNA damage is the characteristic of cells under-going mitotic catastrophe (18, 35, 36). Using time-lapse videomicroscopy, we aimed to detect the association betweenDNA-PKcs deficiency and mitotic catastrophe in responseto DNA damage. We monitored mitotic progression and mi-totic catastrophe in DNA-PKcs–depleted HeLa-H1 cells after4-Gy irradiation. Together with the observation of confocalimmunofluorescence microscopy, we found that the mitoticcatastrophe is a major type of cell death in DNA-PKcs–deficient HeLa-H1 postirradiation. A representative cellundergoing mitotic catastrophe is shown in Fig. 2C and

Cancer Res; 70(9) May 1, 2010 3659



Shang et al.

3660


Supplementary Fig. S3 (video); this cell initiated mitosis andchromosome congression but was unable to segregate sisterchromatids to drive the chromosomes toward the spindlepoles and eventually undergone mitotic catastrophe.Phosphorylated DNA-PKcs localizes at centrosomes and

midbody. It is obvious that DNA-PKcs deficiency can lead todisruption of spindles and multinucleated cells in responseto DNA damage, which eventually results in mitotic celldeath. This suggests that the DNA-PKcs could be an impor-tant regulator in the overall progression of mitosis and spin-dle formation, especially after DNA damage. It is necessary todetermine the subcellular localization of the DNA-PKcs dur-ing mitosis. Plk1 is a well-known regulator of cell division ineukaryotic cells, and the most striking feature of Plk1 is itsdifferential localization to various subcellular structures dur-ing mitosis (26). Plk1 initially associates with centrosomesduring prophase, accumulates at the kinetochores in prome-taphase and metaphase, is then recruited to the central spin-dle in anaphase, and finally becomes enriched in themidbody during cytokinesis. Therefore, we used Plk1 as amolecular marker for detecting the subcellular localization



of the DNA-PKcs. Using immunofluorescence with an anti-body against the nonphosphorylated DNA-PKcs protein, wefound that the DNA-PKcs had a diffuse nuclear distribution(data not shown), and we could not detect any specific local-ization. However, a clear localization of the phosphorylatedDNA-PKcs/T2609 or DNA-PKcs/T2647 at the centrosomeswas seen when an antibody specific for phosphorylatedDNA-PKcs/T2609 or DNA-PKcs/T2647 was used for the im-munofluorescence staining of HeLa cells (Fig. 2D, I and II).The phosphorylated DNA-PKcs/T2609 was also detected atthe kinetochores in metaphase (Fig. 2D, I) and accumulationat the midbody while cell is approaching the end of cytoki-nesis (Fig. 2D, IV). These localization features of DNA-PKcs/pT2609 have also been shown in A549 cells (data not shown).Interestingly, the phosphorylated Chk2/pT68 also localizeswith Plk1 at the centrosomes (Fig. 2D, III). During cytokine-sis, Chk2/T68 localizes at both centrosomes and midbody(Fig. 2D, V), whereas DNA-PKcs/pT2609 only localizes atthe midbody (Fig. 2D, IV).Phosphorylation of Chk2 at T68 upon DNA damage is

DNA-PKcs dependent. As a DNA damage response protein,

Figure 1. Deficiency of DNA-PKcs leads to increased outcome of polyploidy and aberrant spindles in response to IR. A, Western blotting analysisshowing the siRNA-mediated suppressed expression of DNA-PKcs but normal expression of its partner Ku70 and Ku80 in HeLa-H1 compared with thecontrol HeLa-NC cells. B, time-dependent induction of polyploidy in DNA-PKcs–deficient HeLa-H1 cells, HeLa-NC cells treated with the DNA-PKcs inhibitorNU7026, and control HeLa-NC cells after 4-Gy γ-ray irradiation. The data are presented as the mean and SD from three independent experiments(*, P < 0.05; #, P < 0.01 compared with HeLa-NC cells at the same time point). C, representative images showing the aberrant spindle patterns in irradiatedHeLa-H1 cells. Twenty-four hours after 4-Gy irradiation, the cells were fixed, costained with antibodies against α-tubulin and γ-tubulin, and analyzedusing confocal immunofluorescence assay. Nuclei were visualized with DAPI staining. Arrows, centrosome breakage. D, quantitative analysis of aberrantmitotic cells with aberrant spindles or multicentrosomes. The data are presented as the mean and SD of three independent experiments (*, P < 0.05;#, P < 0.01 compared with HeLa-NC cells at the same time point).

Cancer Research





Chk2 has been recently shown to involve in the spindle dam-age response through phosphorylating BRCA1 at S988 site,which is necessary for the inhibiting microtubule nucleationactivation of BRCA1 (37). To test the hypothesis that DNA-PKcs is functionally associated with Chk2 in maintaining thestability of spindle formation upon DNA damage, we havecompared the phosphorylation levels of p-Chk2/T68 betweenDNA-PKcs–suppressed HeLa-H1 cells and the efficient HeLa-NC cells. As shown in Fig. 3A, the T68-phosphorylated formof Chk2 significantly increased after 4-Gy irradiation in thecontrol HeLa-NC. In contrast, a remarkable reduction ofIR-induced phosphorylated Chk2 was shown in the DNA-PKcs–suppressed HeLa-H1 cells. This result suggests thatphosphorylation of Chk2 at T68 in response to DNA damageis contributed by DNA-PKcs. To support these data, wefurther investigated IR-induced phosphorylation of Chk2 atT68 in HeLa cells in the presence of NU7026, a DNA-PKcs–specific inhibitor. As shown in Fig. 3B, inactivation of DNA-PKcs by NU7026 significantly inhibited 4-Gy IR–induced



phosphorylation of Chk2 at T68. In addition, we have alsoobserved an increased expression of cyclin B1 protein andPlk1 and decreased expression of cyclin E protein but nochange on cyclin-dependent kinase 1 in HeLa-H1 cells com-pared with HeLa-NC cells after 4-Gy irradiation (Fig. 3C).ATM is considered as a major kinase for T68 phosphory-

lation of Chk2 in response to DNA damage (38), and DNA-PKcs depletion results in the downregulation of ATM (39).To confirm the dependency of phosphorylation of Chk2/T68 on DNA-PKcs, we further investigated IR-induced T68phosphorylation of Chk2 in ATM-deficient AT5BIVA cells. Al-though ATM is deficient in AT5BIVA cells (Fig. 4A), the auto-phosphorylation form of DNA-PKcs/pS2056 is obviouslyinduced after 4-Gy irradiation (Fig. 4B), indicating thatDNA-PKcs is activated upon DNA damage in this cell line.Interestingly, T68 phosphorylation of Chk2 is also significant-ly induced by 4 Gy in AT5BIVA, and inactivation of DNA-PKcs by NU7026 can effectively inhibit IR-induced T68phosphorylation of CHK2 (Fig. 4C). These results further

Figure 2. DNA-PKcs deficiency leads to formation of multinucleated cells and mitotic cell death, and the subcellular localizations of phosphorylatedDNA-PKcs and CHK2 during mitosis. A, representative images showing multinucleated cells (arrowheads), micronuclei (arrow), and aberrant spindlepatterns in irradiated HeLa-H1. The cells were fixed, stained with an antibody against α-tubulin, and analyzed using confocal microscopy. Nuclei werevisualized with Hochest 33258 staining. B, quantitative analysis of multinucleated cells from irradiated HeLa-H1 and control HeLa-NC cells. The data arepresented as the mean and SD of three independent experiments (*, P < 0.05; #, P < 0.01 compared with HeLa-NC cells at the same time point).C, a representative mitotic cell death observed using time-lapse fluorescence microscopy. The irradiated DNA-PKcs–deficient HeLa-H1 cells enteredmitosis and initiated chromosome congression but were unable to segregate sister chromatids and drive the chromosomes toward the spindle poles andthen eventually underwent mitotic catastrophe. D, colocalization analyses of phosphorylated DNA-PKcs/T2609/T2647 and phosphorylated Chk2/T68with Plk1. The cells were fixed, costained with antibodies against pT2609-DNA-PKcs or pT2647-DNA-PKcs or pT68-Chk2 and Plk1, and analyzed usingconfocal microscopy. Nuclei were visualized with DAPI staining. IV and V, arrow indicates the centrosome, and arrowhead indicates the midbody.

Cancer Res; 70(9) May 1, 2010 3661



Shang et al.

3662


confirmed the dependency of IR-induced phosphorylation ofChk2/T68 on DNA-PKcs.The association of DNA-PKcs with Chk2 is driven by Ku

70/80 heterodimer. To identify the interaction betweenDNA-PKcs and Chk2, we have firstly performed the immuno-precipitation assay using an antibody against DNA-PKcs orChk2 to pull down each other in HeLa cells extracts. Theresult indicates that DNA-PKcs and Chk2 can coimmunno-precipitate each other (Fig. 3D; Supplementary Fig. S4A).Although DNA-PKcs is suppressed in HeLa-H1 cells, itspartner Ku70/Ku80 is efficient (Fig. 1A). It is well known thatKu70/80 heterodimer, as a regulatory component, associateswith DNA-PKcs to form DNA-dependent protein kinasecomplex whereas Ku70/80 was also shown to interact withChk2 (40). Therefore, it was then asked whether DNA-PKcsinteracts with Chk2 directly or indirectly via mediated by



its partner Ku heterodimer. As shown in Fig. 3D, siRNA-mediated knockdown of both Ku70 and Ku80 deprives theassociation of DNA-PKcs with Chk2 in HeLa cells, whichsuggests that the interaction of DNA-PKcs and Chk2 ismediated by Ku heterodimer. Moreover, the associationof DNA-PKcs and Chk2 in the presence of Ku70/80 isaugmented by 4-Gy irradiation (Supplementary Fig. S4B).Inactivation of DNA-PKcs augments the induction

of multipolar spindles and mitotic catastrophe in ATM-deficient cells by IR. After 2-Gy or 4-Gy irradiation, we observeda G2-M phase arrest in ATM-deficient AT5BIVA cells, and thisarrest is completely released 24 hours post-IR (Fig. 5A and B).In contrast, an increased accumulation of AT5BIVA cells inG2-M phase still remained even at 24 to 48 hours postirradia-tion when the cells were cotreatedwith DNA-PKcs–specific in-hibitor NU7026 (Fig. 5A and B), suggesting that inactivation of

Figure 3. DNA-PKcs indirectly interacts with Chk2, and inactivation of DNA-PKcs attenuates IR-induced phosphorylation of Chk2/T68. A, comparisonof 4 Gy–induced phosphorylation of Chk2/T68 between DNA-PKcs–deficient HeLa-H1 and control HeLa-NC cells by Western blotting analysis.B, attenuation of IR-induced phosphorylation of Chk2/T68 by DNA-PKcs inhibitor NU7026. C, effect of DNA-PKcs knockdown by siRNA on the expressionof a number of cell cycle–related proteins. D, DNA-PKcs interacts with Chk2 in the presence of Ku70/80 heterodimer. Immunoprecipitations wereperformed using an antibody against DNA-PKcs to pull down Chk2 in HeLa cells with normal or siRNA-mediated depressed expression of Ku70/80heterodimer. Immunoprecipitation of IgG was used for control.

Cancer Research





the DNA-PKcs also leads to a prolonged G2-M phase arrest inAT5BIVA cells. Finally, we have found that inactivation ofDNA-PKcs by NU7026 augmented the occurrence of multi-centrosomes or multipolar spindles and other forms of aber-rant spindles, as well as multinucleated cells in AT5BIVA cellsafter 2-Gy (Fig. 5D) or 4-Gy (Fig. 5C) irradiation, which even-tually underwent mitotic catastrophe.

Discussion

DNA-PKcs, ATM, and the ATM and Rad3-related kinase(ATR) are three major members of the phosphatidylinositol3-kinase–related kinase family, which play a proximal role inthe cellular responses to DNA damage. ATM and ATR areconsidered as initiators of DNA damage checkpoint signal-ing. In response to double-strand breaks, ATM and ATRblock cell cycle progression through activating checkpointsignaling, whereas both of them initiate two distinct check-point signaling pathways dependent on the type of inducedDNA damage. When DNA double-strand breaks are inducedby IR in the cells, the activated ATM initiates a cell cyclecheckpoint signal through phosphorylating the Chk2 at T68site (41, 42). Alternatively, in response to UV-induced DNAdamage or double-strand breaks associated with replicationforks, ATR plays a dominant role in initiating a checkpointsignaling pathway through phosphorylating Chk1 at S317(43, 44). In contrast, DNA-PKcs has been identified as animportant enzyme in the NHEJ of DNA double-strand breakrepair and in triggering double-strand break–inducedapoptosis. A recent report showed its function on the G2

checkpoint in response to IR (30). In this study, we haveshown another important role of DNA-PKcs in maintainingthe stability of spindle formation and negatively regulatingmitotic catastrophe in response to IR. We showed that inac-



tivation of DNA-PKcs by siRNA or specific inhibitor NU7026in HeLa cells results in a striking outcome of polyploidy, mul-tinucleated cells and multipolar spindles and accelerates themitotic catastrophe in response to IR. A number of DNAdamage response proteins have been shown associated withthe stabilization of centrosome and spindle construction andnegative regulation of mitotic cell death in response to DNAdamage, including Chk 2 protein, which was shown to benecessary for the centrosome stability and against IR-induced mitotic catastrophe via phosphorylating BRCA1/T988 (37). As expected, we have observed an increased phos-phorylation of Chk2/T68 induced by 4-Gy IR in HeLa cells.Moreover, a remarkable phosphorylation of Chk2/T68 in-duced by IR was also detected in the ATM-deficient AT5BIVAcells. Our result further showed that inactivation of DNA-PKcs by specific siRNA and/or inhibitor Nu7026 sharply de-creased the IR-induced phosphorylation of Chk2/T68 in HeLacells as well as AT5BIVA cells. These results are consistentwith previous reports showing that DNA-PKcs regulatesChk2 phosphorylation in ATM-deficient cells (31) or ATMand DNA-PKcs wild-type cells (40). Recently, Tomimatsuand colleagues have also observed the phosphorylation ofChk2/T68 induced by 8 Gy of higher-dose IR in AT5BIVA cellline, and this IR-induced phosphorylation of Chk2 was sug-gested by both ATR and DNA-PKcs (31). Our result showedthat 4 Gy of lower dose–induced phosphorylation of Chk2/T68 can be effectively attenuated by DNA-PKcs–specific in-hibitor NU7026, suggesting that phosphorylation of Chk2/T68 upon stress of a relative low-dose IR may be predomi-nantly dependent on DNA-PKcs. Upon DNA damage, Ku70/Ku80 heterodimer associates with DNA-PKcs to form theDNA-PK complex, and this complex plays an important rolein the pathway of NHEJ pathway of DNA double-strandbreaks (27). Li and Stern reported that Chk2 constitutivelyinteracts with Ku70/Ku80 (40). In present study, our results

Figure 4. Effect of DNA-PKcsinactivation by NU7027 onIR-induced phosphorylation ofChk2/T68 in ATM-deficientAT5BIVA cells. A, Westernblotting analysis of ATM proteinexpression. B, activation ofautophosphorylation ofDNA-PKcs/S2056 by IR inAT5BIVA cells, and the attenuationby NU7026. C, attenuation ofIR-induced phosphorylation ofChk2/T68 by DNA-PKcs inhibitorNU7026 in AT5BIVA cells.

Cancer Res; 70(9) May 1, 2010 3663



Shang et al.

3664


show that DNA-PKcs and Chk2 can be coimmunocipitatedby either antibody in HeLa cell extracts. However, the asso-ciation of DNA-PKcs and Chk2 is completely deprived whenKu70/80 heterodimer is knocked down by siRNA molecules.It is obvious that the interaction of DNA-PKcs and Chk2 isdriven by Ku heterodimer.Our observation also shows that Chk2/pT68 colocalizes

with Plk1 at the centrosomes and the midbody, which is con-sistent with previous reports (45). In addition, we showedthat the phosphorylated DNA-PKcs also localizes at the cen-trosomes and the midbody. During cytokinesis, Chk2/pT68localizes at both midbody and centrosome in cytoplasm ofdaughter cells, whereas DNA-PKcs only localizes at themidbody. In addition, the phosphorylated DNA-PKcs coloca-lizes at the kinetochores with Plk1 in metaphase. These sub-cellular localization features further support that DNA-PKcsplays an important role in regulation of spindle formationand mitosis.



The formation of multinucleated cells and multipolar spin-dles in cells after DNA damage is a characteristic of cells un-dergoing mitotic catastrophe (18, 35, 36). A hallmark ofmitotic catastrophe is the premature entry of cells into mi-tosis or failure of cytokinesis despite the presence of dam-aged DNA and which results in the activation of apoptosisor necrosis (18, 22). Our data show that inactivation ofDNA-PKcs increased the induction of multinucleated cellsand multipolar aberrant spindles in response to IR. The ob-servation of time-lapse video microscopy further confirmedthe mitotic catastrophe due to the failure of mitosis pro-gression. Again, it was reported that inhibition of Chk2 bya dominant-negative Chk2 mutant or a chemical inhibitor,debromohymenialdesine, attenuated the arrest of multi-nuclear HeLa syncytia at the G2-M boundary and ratherentered mitotis (premature mitotic entry) and subsequentlyundergone mitotic catastrophe (23). Taken together the im-portant role of phosphorylated Chk2/T68 in regulation of

Figure 5. Effect of DNA-PKcs inactivation by NU7026 on cell cycle progression of ATM-deficient cells in response to IR. A, flow cytometric histogramsof cell cycle progression of ATM-deficient AT5BIVA cells after 4-Gy irradiation with or without the treatment of DNA-PKcs inhibitor NU7026. B, quantitativeanalysis of cell cycle distribution of AT5BIVA cells after 2-Gy or 4-Gy irradiation with or without NU7026 treatment. C and D, observation of aberrantnuclei morphology and aberrant spindles of 2-Gy (C) and 4-Gy (D) irradiated AT5BIVA cells treated with NU7036 by confocal immunofluorescence assay.The cells were fixed and costained with antibodies against γ-tubulin and α-tubulin.

Cancer Research





centromsome stability and spindle formation (22, 37), atten-uation of Chk2 phosphorylation at least partially attributes tothe augmentation of IR-induced aberrant spindles and mito-tic catastrophe mediated by inactivating DNA-PKcs. On theother hand, deficiency of DNA-PKcs causes a decreased DNArepair capability and results in more residual DNA damage inthe cells, including the mitotic cells, which prompts the mi-totic catastrophe.Together, these data show that DNA-PKcs is necessary for

stabilizing centrosomes and spindle formation and prevent-ing mitotic catastrophe in response to DNA damage. UponDNA damage, Ku70/80 heterodimer drives the associationof DNA-PKcs and Chk2, and DNA-PKcs exerts its effectregulating spindle formation and mitosis through, at leastpartially, phosphorylating Chk2/T68. Furthermore, the locali-zations of phosphorylated DNA-PKcs at centrosomes, kineto-chores, and midbody imply that DNA-PKcs may play a directrole in regulating mitotic progression. As DNA-PK is wellknown for its function in the NHEJ pathway of DNA double-



strand break repair, its involvement in mitotic responses toDNA damage provides another interesting prospect forunderstanding the mechanism coupling DNA repair andthe regulation of cell cycle progression, as well as strategyin the development of new radiation therapeutic approaches.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Grant Support

National Key Basic Research Program of MOST, China grant 2007CB914603(973 Program), Chinese National Natural Science Foundation grants 30371232and 30970677, and Outstanding Youth Scientist Foundation of NFSC, Chinagrant 30825011.

The costs of publication of this article were defrayed in part by the paymentof page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received 09/09/2009; revised 01/12/2010; accepted 02/02/2010; publishedOnlineFirst 04/20/2010.

References
1. Rao CV, Yamada HY, Yao Y, Dai W. Enhanced genomic instabilities
caused by deregulated microtubule dynamics and chromosomesegregation: a perspective from genetic studies in mice. Carcino-genesis 2009;30:1469–74.

2. Heald R, Tournebize R, Habermann A, Karsenti E, Hyman A. Spindleassembly in Xenopus egg extracts: respective roles of centrosomesand microtubule selforganization. J Cell Biol 1997;138:615–28.

3. Hinchcliffe EH, Sluder G. “It takes two to tango”: understanding howcentrosome duplication is regulated throughout the cell cycle. GenesDev 2001;15:1167–81.

4. Lingle WL, Barrett SL, Negron VC, et al. Centrosome amplificationdrives chromosomal instability in breast tumor development. ProcNatl Acad Sci U S A 2002;99:1978–83.

5. Pihan GA, Purohit A, Wallace J, et al. Centrosome defects andgenetic instability in malignant tumors. Cancer Res 1998;58:3974–85.

6. Rieder CL, Faruki S, Khodjakov A. The centrosome in vertebrates:more than a microtubule organizing center. Trends Cell Biol 2001;11:413–9.

7. Meraldi P, Honda R, Nigg EA. Aurora-A overexpression reveals tetra-ploidization as a major route to centrosome amplification in p53−/−cells. EMBO J 2002;21:483–92.

8. Borel F, Lohez OD, Lacroix FB, Margolis RL. Multiple centrosomesarise from tetraploidy checkpoint failure and mitotic centrosomeclusters in p53 and RB pocket protein-compromised cells. Proc NatlAcad Sci U S A 2001;99:9819–24.

9. Roh M, Gary B, Song C, et al. Overexpression of the oncogenickinase pim-1 leads to genomic instability. Cancer Res 2003;63:8079–84.

10. Ko MJ, Murata K, Hwang DS, Parvin JD. Inhibition of BRCA1 inbreast cell lines causes the centrosome duplication cycle to bedisconnected from the cell cycle. Oncogene 2006;25:298–303.

11. Wang X, Wang RH, Li W, et al. Genetic interactions between Brca1and Gadd45a in centrosome duplication, genetic Stability, and neuraltube closure. J Biol Chem 2004;279:29606–14.

12. Wang Q, Hirohashi Y, Furuuchi K, et al. The centrosome in normaland transformed cells. DNA Cell Biol 2004;23:475–89.

13. Weinert TA, Hartwell LH. The RAD9 gene controls the cell cycleresponse to DNA damage in Saccharomyces cerevisiae. Science1988;241:317–22.

14. Fotedar R, Flatt J, Gupta S, et al. Activation-induced T-cell death iscell cycle dependent and regulated by cyclin B. Mol Cell Biol 1995;15:932–42.

15. Castedo M, Perfettini JL, Roumier T, Kroemer G. Cyclin-dependent

kinase-1: linking apoptosis to cell cycle and mitotic catastrophe.Cell Death Differ 2002;9:1287–93.

16. Chan TA, Hermeking H, Lengauer C, Kinzler KW, Vogelstein B. 14-3-3sigma is required to prevent mitotic catastrophe after DNA damage.Nature 1999;401:616–20.

17. Nitta M, Kobayashi O, Honda S, et al. Spindle checkpoint function isrequired for mitotic catastrophe induced by DNA-damaging agents.Oncogene 2004;23:6548–58.

18. Huang X, Tran T, Zhang L, Hatcher R, Zhang P. DNA damage-induced mitotic catastrophe is mediated by the Chk1-dependentmitotic exit DNA damage checkpoint. Proc Natl Acad Sci U S A2005;102:1065–70.

19. Ha GH, Kim HS, Lee CG, et al. Mitotic catastrophe is the pre-dominant response to histone acetyltransferase depletion. CellDeath Differ 2009;16:483–97.

20. Dowling M, Voong KR, Kim M, Keutmann MK, Harris E, Kao GD.Mitotic spindle checkpoint inactivation by trichostatin a definesa mechanism for increasing cancer cell killing by microtubule-disrupting agents. Cancer Biol Ther 2005;4:197–206.

21. Sekhar KR, Sonar VN, Muthusamy V, et al. Novel chemicalenhancers of heat shock increase thermal radiosensitization througha mitotic catastrophe pathway. Cancer Res 2007;67:695–701.

22. Castedo M, Perfettini JL, Roumier T, et al. Mitotic catastropheconstitutes a special case of apoptosis whose suppression entailsaneuploidy. Oncogene 2004;23:4362–70.

23. Castedo M, Perfettini JL, Roumier T, et al. The cell cycle checkpointkinase is a negative regulator of mitotic catastrophe. Oncogene2004;23:4353–61.

24. Vakifahmetoglu H, Olsson M, Tamm C, Heidari N, Orrenius S,Zhivotovsky B. DNA damage induces two distinct modes of celldeath in ovarian carcinomas. Cell Death Differ 2008;15:555–66.

25. Meng X, Fan J, Shen Z. Roles of BCCIP in chromosome stability andcytokinesis. Oncogene 2007;26:6253–60.

26. Petronczki M, Lenart P, Peters JM. Polo on the rise-from mitoticentry to cytokinesis with Plk1. Dev Cell 2008;14:646–59.

27. Jackson SP. Sensing and repairing DNA double-strand breaks.Carcinogenesis 2002;23:687–96.

28. Burma S, Chen BP, Chen DJ. Role of non-homologous end joining(NHEJ) in maintaining genomic integrity. DNA Repair (Amst) 2006;5:1042–8.

29. Bailey SM, Brenneman MA, Halbrook J, Nickoloff JA, Ullrich RL,Goodwin EH. The kinase activity of DNA-PK is required to protectmammalian telomeres. DNA Repair (Amst) 2004;3:225–33.

Cancer Res; 70(9) May 1, 2010 3665



Shang et al.

3666


30. Arlander SJ, Greene BT, Innes CL, Paules RS. DNA-PK-dependentG2 checkpoint revealed following knockdown of ATM in humanmammary epithelial cells. Cancer Res 2008;68:89–97.

31. Tomimatsu N, Mukherjee B, Burma S. Distinct roles of ATR andDNA-PKcs in triggering DNA damage responses in ATM-deficientcells. EMBO Rep 2009;10:629–35.

32. An J, Yang DY, Xu QZ, et al. DNA-dependent protein kinase catalyticsubunit modulates the stability of c-Myc oncoprotein. Mol Cancer2008;7:32.

33. Martinez JJ, Seveau S, Veiga E, Matsuyama S, Cossart P. Ku70, acomponent of DNA-dependent protein kinase, is a mammalianreceptor for Rickettsia conorii. Cell 2005;123:1013–23.

34. Chang I-Y, Youn C-K, Kim H-B, et al. Oncogenic H-Ras up-regulatesexpression of Ku80 to protect cells from γ-ray irradiation in NIH3T3cells. Cancer Res 2005;65:6811–9.

35. Dodson H, Wheatley SP, Morrison CG. Involvement of centrosomeamplification in radiation-induced mitotic catastrophe. Cell Cycle2007;6:364–70.

36. Wonsey DR, Follettie MT. Loss of the forkhead transcription factorFoxM1 causes centrosome amplification and mitotic catastrophe.Cancer Res 2005;65:5181–9.

37. Chabalier-Taste C, Racca C, Dozier C, Larminat F. BRCA1 is



regulated by CHK2 in response to spindle damage. Biochim BiophysActa 2008;1783:2223–33.

38. Buscemi G, Perego P, Carenini N, et al. Activation of ATM and Chk2kinases in relation to the amount of DNA strand breaks. Oncogene2004;23:7691–700.

39. Peng Y, Woods RG, Beamish H, et al. Deficiency in the catalyticsubunit of DNA-dependent protein kinase causes down-regulationof ATM. Cancer Res 2005;65:1670–7.

40. Li J, Stern DF. Regulation of CHK2 by DNA-dependent protein kinase.J Biol Chem 2005;280:1204–50.

41. Niida H, Nakanishi M. DNA damage checkpoints in mammals.Mutagenesis 2006;21:3–9.

42. CuadradoM,Martinez-Pastor B, Fernandez-Capetillo O. ATR activationin response to ionizing radiation: still ATM territory. Cell Div 2006;1:7.

43. Osborn AJ, Elledge SJ, Zou L. Checking on the fork: the DNA-replication stress-response pathway. Trends Cell Biol 2002;12:509–16.

44. Cimprich KA, Cortez D. ATR: an essential regulator of genomeintegrity. Nat Rev Mol Cell Biol 2008;9:616–27.

45. Tsvetkov L, Xu X, Li J, Stern DF. Polo-like kinase 1 and CHK2 interactand co-localize to centrosomes and the midbody. J Biol Chem 2003;278:8468–75.

Cancer Research



2010;70:3657-3666. Published OnlineFirst April 20, 2010.Cancer Res Zeng-Fu Shang, Bo Huang, Qin-Zhi Xu, et al. DamageCheckpoint Protein 2 Phosphorylation in Response to DNASpindle Disruption and Mitotic Catastrophe with Attenuated Inactivation of DNA-Dependent Protein Kinase Leads to

Updated version

10.1158/0008-5472.CAN-09-3362doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2010/04/19/0008-5472.CAN-09-3362.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/70/9/3657.full#ref-list-1

This article cites 45 articles, 16 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/70/9/3657.full#related-urls

This article has been cited by 11 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/70/9/3657To request permission to re-use all or part of this article, use this link

on July 16, 2020. © 2010 American Association for Cancer Research.cancerres.aacrjournals.org Downloaded from


http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-09-3362

http://cancerres.aacrjournals.org/content/suppl/2010/04/19/0008-5472.CAN-09-3362.DC1

http://cancerres.aacrjournals.org/content/70/9/3657.full#ref-list-1

http://cancerres.aacrjournals.org/content/70/9/3657.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/70/9/3657


Documents

Research Inactivation of DNA-Dependent Protein Kinase Leads to … · ing the nonhomologous end joining pathway of DNA double-strand breaks repair. Here, we showed another important