Regulation of DNA-dependent protein kinase by protein ... - Nature

Oncogene (2010) 29, 6016–6026

& 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc

ORIGINAL ARTICLE

Regulation of DNA-dependent protein kinase by protein kinase CK2 in human glioblastoma cells BB Olsen, O-G Issinger and B Guerra Department of Biochemistry and Molecular Biology, University of Southern Denmark, Odense, Denmark

The DNA-dependent protein kinase (DNA-PK) is a nuclear serine/threonine protein kinase composed of a large catalytic subunit (DNA-PKcs) and a heterodimeric DNA-targeting subunit Ku. DNA-PK is a major component of the nonhomologous end-joining pathway of DNA double-strand breaks repair. Although DNA-PK has been biochemically characterized in vitro, relatively little is known about its functions in the context of DNA repair and how its kinase activity is precisely regulated in vivo. Here, we report that cellular depletion of the individual catalytic subunits of protein kinase CK2 by RNA interference leads to significant cell death in M059K human glioblastoma cells expressing DNA-PKcs, but not in their isogenic counterpart, that is M059J cells, devoid of DNA-PKcs. The lack of CK2 results in enhanced DNA-PKcs activity and strongly inhibits DNA damageinduced autophosphorylation of DNA-PKcs at S2056 as well as repair of DNA double-strand breaks. By the application of the in situ proximity ligation assay, we show that CK2 interacts with DNA-PKcs in normal growing cells and that the association increases upon DNA damage. These results indicate that CK2 has an important role in the modulation of DNA-PKcs activity and its phosphorylation status providing important insights into the mechanisms by which DNA-PKcs is regulated in vivo. Oncogene (2010) 29, 6016–6026; doi:10.1038/onc.2010.337; published online 16 August 2010 Keywords: DNA-PK; CK2; DNA repair; glioblastoma cells

autophosphorylation;

Introduction The presence of functional DNA-damage surveillance and repair pathways is fundamental for maintaining genomic integrity in normal cells. A wide variety of DNA lesions leads to activation of various cell-cycle checkpoints controlled by the PI3K-related kinases, which allow repair of DNA damage before proceeding Correspondence: Dr B Guerra, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, Odense, DK-5230 Fyn, Denmark. E-mail: [email protected] Received 8 February 2010; revised 11 June 2010; accepted 3 July 2010; published online 16 August 2010

further with cell proliferation (Harrison and Haber, 2006). Double-strand DNA breaks (DSBs) are considered the biggest threat to genomic stability because, if unrepaired or incorrectly repaired, they can lead to cell death. Two major repair pathways have been reported to be activated in response to DSBs: nonhomologous end-joining (NHEJ) and homologous recombination (Helleday et al., 2007). DNA-dependent protein kinase (DNA-PK) is a serine/threonine kinase composed of a catalytic subunit (DNA-PKcs) and a heterodimeric complex of Ku70 and Ku86, which binds to the two DNA ends in a ring conformation. This event subsequently activates DNA-PKcs, which promotes the ligation of DNA ends in the presence of other components of the NHEJ-repair machinery (Mahaney et al., 2009). Modulation of DNA-PK activity is not exclusively regulated by the coordinated assembly of Ku and DNA-PKcs on DNA ends. In this respect, it has been reported that DNA-PKcs can undergo autophosphorylation (Chan et al., 2002; Douglas et al., 2002, 2007; Chen et al., 2005). To date, it is not clear which is the exact role of this event, although some studies have reported that autophosphorylation of purified DNAPKcs results in disruption of the DNA-PK complex in vitro and loss of kinase activity in vivo (Chan and Lees-Miller, 1996; Merkle et al., 2002; Ding et al., 2003; Block et al., 2004; Douglas et al., 2007). This leaves open the possibility that dissociation of the DNA-PK complex from DNA might be necessary for facilitating subsequent repair steps by allowing the assembly of damage-responsive proteins to the site of DNA damage or activating them. A total of 16 in vitro phosphorylation sites in DNA-PKcs have been identified by mass spectrometry of which four were found phosphorylated also in vivo in okadaic acid-treated human cells (Chan et al., 2002). These findings represent a very important advance in the understanding of DSB repair mechanisms and confirm that DNA-PK implication in NHEJ may require coordinated phosphorylation at multiple sites (Weterings and Chen, 2007). Protein kinase CK2 is a serine/threonine kinase highly conserved in eukaryotic organisms so far investigated and composed of two catalytic a and/or a0 -subunits and two regulatory b-subunits. CK2 appears distributed in tetrameric complexes inside the cells but significant evidence indicates that the individual subunits do not exist exclusively within these complexes but also as free proteins (Litchfield, 2003; Bibby and Litchfield, 2005).

CK2 modulates DNA-PK autophosphorylation and activity BB Olsen et al

6017

CK2 is involved in cell-cycle progression and in the regulation of intracellular signaling pathways associated with various diseases, particularly cancer (Guerra and Issinger, 2008; Trembley et al., 2009). Current data suggest that CK2 is also involved in DNA sensing and repair. One of the most compelling evidence was reported by (Loizou et al. (2004), demonstrating that CK2 has a direct role in the DNA repair by phosphorylating XRCC1, thereby enabling the assembly and activity of DNA single-strand break repair at sites of chromosome breakage. The CK2-catalyzed phosphorylation of HP1-b, a chromatin factor bound to histone H3, was found to occur in cells exposed to DNA damage and this event was accompanied by HP1-bmobilization and subsequent alteration in chromatin structure (Ayoub et al., 2008). Although mounting evidence indicates that CK2 is implicated in the earliest cellular response to DNA breakage, the relationship between CK2 and components of the DNA damage response is yet to be thoroughly explored. In this study, by employing two human glioblastoma cell lines, M059J and M059K, which lack and express normal levels of DNA-PKcs, respectively, we report evidence that downregulation of the individual CK2 catalytic subunits leads to marked cell death in cells expressing DNA-PKcs. Furthermore, we show that knockdown of CK2 results in decreased DNA-PKcs autophosphorylation and persistent DNA damage following induction of DSBs in cells. These findings suggest that CK2 contributes to the regulation of DNA-PK in eukaryotic cells and provide important evidence of a prominent role of CK2 in the DNA damage response.

Results Downregulation of protein kinase CK2 induces cell death in glioblastoma cells expressing DNA-PKcs We investigated the response of two human glioblastoma cell lines M059K and M059J to downregulation of the individual CK2 subunits in the presence or absence of DNA damage induced by incubation of cells with the radiomimetic drug neocarzinostatin (NCS; Povirk, 1996) for 24 h. As shown in Figure 1a, transfection of cells with small interfering RNAs (siRNAs) against the individual CK2 subunits led to a marked reduction of the CK2 protein levels in both cell lines. Additionally, the sole depletion of CK2a resulted in the downregulation of CK2b, an effect that has been previously reported (Canton et al., 2001), whereas downregulation of CK2b negatively affected the expression of CK2a0 . Reduction of CK2 protein levels caused morphological changes in M059K cells indicative of cell death (Figure 1b). A similar effect was observed when cells lacking the individual CK2 subunits were incubated with 0.5 mg/ml NCS. In the case of M059J cells, a slight morphological change was observed only in cells transfected with siRNA-CK2b in comparison to control cells. To ensure that the indicated morphological changes reflected active cell killing, flow

cytometry analysis was performed for measuring cell death (SYTOX green-positive cells) under treatment conditions indicated in Figure 1b. In M059K cells, reduction of CK2a and a0 , respectively, led to up to 33% cell killing and this percentage was slightly higher following incubation with NCS. CK2b downregulation resulted in a minor but still detectable effect attributable to the lower CK2a0 expression levels. NCS alone led to B15% cell death in comparison to control cells. In the case of M059J cells, downregulation of the individual CK2 subunits, with the exception of the CK2b-siRNA treatment, did not lead to a consistent increase in cell death with respect to control cells. The sole NCS treatment caused B33% cell death and this percentage slightly increased in the case of cells additionally depleted of CK2a (Figure 1c). The M059K cell line was employed for the subsequent experiments as downregulation of the individual CK2 catalytic subunits was most effective in inducing cell death in cells expressing DNA-PKcs (see also Supplementary Figure S1). Next, it was investigated whether knockdown of the CK2 catalytic subunits would lead to an apoptotic type of cell death in the presence and absence of NCS, respectively, by performing a western blot analysis of the caspase-3/caspase-7 substrate poly(ADP-ribose) polymerase (PARP) (Figure 2a). A slight cleavage of full-length PARP was observed in cells treated with NCS (lane 4), whereas the additional treatment with transfection reagent (lane 5) or with siRNA against the individual CK2 catalytic subunits (lanes 6 and 7) did not lead to a further increase in PARP cleavage. Results in Figure 2a were followed by a quantitative determination of the percentage of dead cells following the treatments indicated above and in the presence and absence of the broad-range caspase inhibitor zVAD-fmk or cystein cathepsin inhibitor zFA-fmk, respectively. Results reported in Figure 2b showed that the toxicity induced by cellular depletion of CK2a or -a0 was not reduced when cells were treated with zVAD-fmk or zFA-fmk (see also Supplementary Figures S2a and b). Similar results were obtained when cells were additionally incubated with NCS. These results suggest that although a small proportion of cells undergoes apoptosis (Figure 2a), both caspases and lysosomal proteases are not involved in the mechanism of cell death. Autophagy induction has been reported in malignant glioma cells exposed to ionizing radiation (Yao et al., 2003). To determine whether the aforementioned treatments induced autophagic cell death in M059K cells, cells were incubated with monodansylcadaverine (MDC), a specific fluorescent autophagosome marker (Biederbick et al., 1995), and MDC-positive cells were quantified by flow cytometry. Results in Figure 2c indicate that the proportion of MDC-positive cells increased approximately from 8% (control cells) to 15% (CK2a and -a0 siRNA treatment, respectively) in M059K cells. When cells were additionally incubated with NCS, the percentage of cells undergoing autophagy varied approximately from 14% (NCS-treated cells) to 20% (NCS treatment and downregulation of the individual CK2 subunits, also see Supplementary Oncogene


6018

Figure 1 Downregulation of CK2 leads to cell death in the human glioblastoma cell line M059K. (a) M059K and M059J cells were treated with transfection reagent only (control) or with siRNAs against the individual CK2 subunits for 72 h. Where indicated, 0.5 mg/ml NCS were added during the last 24 h of incubation with siRNA. Whole lysates were analyzed by western blot with antibodies against the indicated proteins. Detection of b-actin protein was used as a loading control. Band quantitation was obtained by analysis with ImageJ software. Values below each band were calculated assigning the value 1.0 to results from the control experiment. (b) Phase-contrast microscopy pictures of cells treated as described above showing morphological changes indicating various degrees of toxicity induced by the aforementioned treatments. Original magnification: 400. (c) Quantification of cell death was performed by flow cytometry analysis of SYTOX green-positive cells. Bars indicate the average from three independent experiments þ /s.d. *Po0.05 denotes statistically significant differences with respect to control. No statistical significant differences were found between cells depleted of CK2 in the absence or presence of NCS. In all cases, experiments were repeated at least three times and data from one representative experiment are shown. TR, transfection reagent.

Figure S2). Overall, these results suggest that the cytotoxicity is predominantly due to induction of an autophagic type of cell death and that downOncogene

regulation of the individual CK2 catalytic subunits contributes to enhance the toxicity of NCS in M059K cells.


6019

Figure 2 CK2 downregulation leads to an autophagic type of cell death. (a) M059K cells incubated with transfection reagent only (control) or transfected with siRNA against the individual CK2 catalytic subunits for 72 h and incubated in the presence or absence of 0.5 mg/ml NCS for 24 h, as indicated in the figure, were subjected to western blot analysis of PARP cleavage. Full-length PARP (116 kDa) and its cleavage product (85 kDa) are indicated. (b) Cells treated as described above were additionally incubated with 5 mM zVAD-fmk or 85 mM zFA-fmk in combination with 0.5 mg/ml NCS for 24 h. Bars indicate average from three independent experiments þ /s.d. (c) M059K cells treated as described above were stained with 50 mM MDC as indication of autophagy. Quantification of MDC-positive cells was performed by FACS analysis. Average of three independent experiments þ /s.d. is shown. *Po0.05 compared with control experiment; **Po0.01 compared with NCS-treated cells. TR, transfection reagent.

Downregulation of the individual CK2 catalytic subunits affects the activity and autophosphorylation status of DNA-PK As the downregulation of CK2 alone or in combination with NCS contributed to enhance the cytotoxicity of M059K cells, it was investigated whether these effects

were due to CK2-mediated modulation of DNA-PK activity and/or expression. Cells were treated as indicated in Figure 3a and Supplementary Figure S3, and the kinase activity of DNA-PK was analyzed by employing a specific peptide substrate. Knockdown of CK2a or -a0 led to a marked increase in DNA-PK activity (B40%) in comparison with the control experiment. As expected, the sole treatment with NCS resulted in 40% increase in DNA-PK activity with respect to control cells. The activity of DNA-PK further increased B20%, with respect to NCS-treated cells, in cells additionally lacking CK2a or -a0 . Western blot analysis of the expression levels of DNA-PKcs and the Ku subunits excluded that the observed upregulation of DNA-PK activity was due to altered protein levels of the DNA-PK complex (Figure 3b). In vitro experiments performed, employing various combinations of the indicated recombinant proteins in an increasing molar ratio, showed a direct CK2-mediated negative modulation of DNA-PK kinase activity with respect to control experiments performed with native DNA-PKcs (Figure 3c). Moreover, in vitro phosphorylation experiments performed in the presence of DNA-PKcs as a potential substrate target for CK2, excluded that the observed downregulation of DNA-PK activity was due to CK2-mediated phosphorylation of DNA-PK (data not shown). Nevertheless, both KU70 and KU86 appeared to be substrates of CK2a in vitro (Supplementary Figure S7). The ataxia telangiectasia-mutated (ATM) gene product has been shown to phosphorylate and positively regulate the kinase activity of DNA-PK (Chen et al., 2007). In order to determine whether ATM kinase contributed to enhance the activity of DNA-PK in knockdown CK2a and -a0 cells, respectively, M059K cells were left untreated or incubated with the specific ATM kinase inhibitor, KU55933. Results shown in Figures 3d and e, excluded that the enhanced DNA-PK activity observed in cells with lowered CK2a or -a0 expression levels was mediated by ATM kinase (Figure 3d), although control experiments performed by measuring the level of phosphorylation of downstream targets of ATM showed that the incubation of cells with 10 mM KU55933 for 4 h was sufficient to induce a marked inhibition of ATM in vivo (Figure 3e). Recent in vivo studies indicated that incubation of cells with DNA-damaging agents leads to autophosphorylation of DNA-PKcs at S2056. This event is accompanied by decreased DNA-PK kinase activity that is essential for the repair of DNA double-strand breaks (Chan and Lees-Miller, 1996; Merkle et al., 2002). To understand the mechanism by which the activity of DNA-PK is regulated by the individual CK2 catalytic subunits, whole extracts from M059K cells treated as indicated in Figure 4a (see also Supplementary Figure S4) were analyzed by western blot. Treatment for 1 h with 0.5 mg/ml NCS alone or with NCS and transfection reagent led to marked autophosphorylation of DNAPKcs at S2056 (lanes 4 and 5) with respect to the control experiment (lane 1). The sole depletion of CK2a or -a0 did not cause phosphorylation of DNA-PKcs (lanes 2 and 3). Surprisingly, incubation of CK2a- or Oncogene


6020

Figure 3 Positive modulation of DNA-PKcs activity following CK2 downregulation. (a) M059K cells were treated as described in Figure 1a. Whole-cell lysates were employed for determination of endogenous DNA-PKcs activity. DNA-PKcs activity is expressed as percentage of control. The activity of DNA-PKcs contained in extracts from M059J treated with 0.5 mg/ml NCS for 24 h was also included in the assay as a control. Bars indicate mean þ /s.d. of three independent experiments. *Po0.01 compared with control experiment; **Po0.01 compared with NCS-treated cells. (b) Whole lysate from cells treated as indicated above was subjected to western blot analysis of various proteins as indicated in the figure. (c) Bar graph showing the influence of increasing amount of various types of CK2 on DNA-PKcs activity measured in phosphorylation assays in the presence of a peptide substrate specific for DNA-PKcs (Achari and Lees-Miller, 2000). Kinase activity is expressed as percentage of control. The ratio between CK2 and DNA-PKcs is expressed in picomol. (d) Kinase activity of endogenous DNA-PKcs in whole extract from cells incubated with transfection reagent or transfected with siRNAs against the individual CK2 catalytic subunits. Where indicated, cells were incubated with 10 mM KU55933 (ATM inhibitor) for 4 h. Bars indicate means þ /s.d. of two independent experiments. (e) Western blot analysis of crude extracts from M059K cells left untreated or incubated with KU55933 for verifying the effectiveness of inhibition of endogenous ATM. The phosphorylation and protein levels of downstream substrate targets of ATM are shown in the figure. TR, transfection reagent.

-a0 -knockdown cells with NCS caused a marked decrease in DNA-PKcs autophosphorylation (lanes 6 and 7) while its expression levels remained unchanged. In order to see Oncogene

whether CK2 kinase activity had a role in the modulation of DNA-PKcs autophosphorylation, cells were incubated with 40 mM resorufin, a highly specific CK2


6021

Figure 4 Lack of CK2a or -a0 leads to decreased DNA-PKcs autophosphorylation upon DNA damage. (a) M059K cells were treated as described in Figure 1a. Whole extracts were employed for western blot analysis of the indicated proteins. Experiments were performed three times obtaining similar results. Data from one representative experiment are shown. (b) Protein extracts from cells left untreated, incubated with 40 mM resorufin for 24 h, 0.5 mg/ml NCS for 1 h or a combination of resorufin and NCS were analyzed for CK2 kinase activity. The activity of endogenous CK2 is expressed as percentage of control. (c) Western blot analysis of whole lysate from cells treated as described above was performed employing antibodies against the indicated proteins.

kinase inhibitor (Sandholt et al., 2009), for 24 h in the presence or absence of 0.5 mg/ml NCS for 1 h. Incubation of cells with resorufin inhibited B40% of endogenous CK2 activity (Figure 4b). Moreover, the presence of resorufin did not affect the autophosphorylation of DNA-PK induced by NCS (Figure 4c; lanes 3 and 4). Overall, these results suggest that the mechanism by which CK2 modulates the activity and the autophosphorylation of DNA-PKcs is not caused by CK2mediated phosphorylation or changes in the expression levels of DNA-PKcs. Downregulation of CK2 inhibits DNA-PK-dependent DNA double-strand break repair Next it was investigated whether CK2-knockdownmediated inhibition of DNA-PKcs autophosphorylation impairs DNA DSB repair. The analysis of cells incubated with 0.5 mg/ml NCS for the indicated times and subsequently stained with anti-phospho-histone H2AX(S139) antibody for visualizing nuclear foci of DNA damage (Kuo and Yang, 2008) revealed a significant enhancement in residual DNA damage in CK2a- or -a0 -knockdown cells with respect to cells treated with only NCS (Figure 5a). Quantification of the number of p-H2AX-positive cells revealed that although

40% of the total number of cells was still presenting DNA damage 24 h after the initial incubation with NCS, B80% of cells lacking the individual CK2 catalytic subunits were still p-H2AX positive after 24 h (Figure 5b). Western blot analysis of whole extracts from cells treated as described above further confirmed the results (Figure 5c and Supplementary Figure S5). To further analyze the ability to repair DNA damage in cells lacking CK2a or -a0 , a previously established plasmid-based (that is, pEGFP-N1) end-joining assay was utilized (Wang et al., 2006). Quantification of green fluorescent protein (GFP)-positive cells by fluorescenceactivated cell sorting analysis showed that there were B40% less GFP-positive cells in CK2-knockdown cells in comparison with the control assay (Figure 5d). The results of these experiments clearly support the idea that CK2 might contribute to the autophosphorylation of DNA-PKcs at S2056, which occurs in vivo following DNA damage, by bringing DNA-PKcs molecules in close proximity. Hence, it was investigated whether CK2 interacts with DNA-PKcs in cells left untreated or exposed to DNA damage by performing an in situ proximity ligation assay that enables the detection and quantification of protein–protein interactions in native cells (So¨derberg et al., 2006). Results shown in Figure 6a clearly indicated the presence of prominent Oncogene


6022

Figure 5 Downregulation of CK2 inhibits DNA-PKcs-dependent DNA double-strand break repair. (a) Photomicrographs of cells treated with NCS or a combination of NCS and CK2-siRNAs analyzed at various time points as indicated in the figure. Cells were labeled with anti-phospho-histone H2AX antibody and with a FITC-conjugated secondary antibody. Cellular DNA was counterstained with DAPI (blue fluorescence). Intense green fluorescence indicates large numbers of foci of DNA double-strand breaks. (b) Quantification of the number of green fluorescence-positive cells, labeled as described above, was performed by using ImageJ software and expressed as percentage of the total number of cells in each sample. Bars indicate means þ /s.d. of four independent experiments. *Po0.05 indicates significant enhancement of number of green fluorescence-positive cells in CK2 siRNAtreated cells exposed to NCS compared with the corresponding control experiments. (c) Detection of DNA-PKcs autophosphorylation and histone H2AX phosphorylation by western blot analysis of crude extracts from cells treated as described above. A representative of two independent experiments is shown. (d) In vivo end-joining assay performed employing pEGFP-N1-plasmid DNA. End-joining activity was calculated as the ratio between the number of GFP-positive cells following electroporation with the HindIII-digested plasmid and the amount of GFP-positive cells after electroporation with the circular plasmid, respectively. The quantification was performed by assigning a value of 100% to the ratio calculated for the control experiment. Quantification of the number of green fluorescence-positive cells was performed by FACS analysis. Bars represent mean þ /s.d. of three independent experiments. *Po0.05 denotes statistically significant differences. TR, transfection reagent.

signals in fixed M059K cells, indicative of the existence of endogenous interacting proteins, in control cells. Interestingly, the number of signals markedly increased Oncogene

in cells exposed to 0.5 mg/ml NCS for 1 h (see also Supplementary Figure S6). The morphological evaluation obtained by fluorescence microscopy was


6023

Figure 6 Detection of endogenous CK2-DNA-PKcs complexes by in situ proximity ligation assay (PLA). (a) Complexes between endogenous CK2a0 and DNA-PKcs were visualized by staining M059K cells left untreated (control) or incubated with 0.5 mg/ml NCS for 1 h with antibodies against CK2a0 and DNA-PKcs. Nuclei were counterstained with Hoechst (blue). No Ab, refers to control assay where cells were incubated with only secondary antibodies. Each red dot represents an interaction detected by the assay between CK2a0 and DNA-PKcs targeted by two primary antibodies. (b) Quantification of the number of signals per cell as distinct fluorescent red spots was performed by computer-assisted image analysis as reported in the Materials and methods. Mean þ /s.d. of three independent experiments is shown. *Po0.001 indicates statistically significant differences.

accompanied by a quantitative analysis of red fluorescent-positive cells and results shown in Figure 6b demonstrate that the number of signals per cell increased by a factor of 2.5 in cells treated with NCS with respect to control experiment.

Discussion In eukaryotic cells, the repair process of DSBs comprises two mechanisms namely the homologous recombination and NHEJ. NHEJ is the predominant repair pathway in

mammalian cells and can be operational in all phases of the cell cycle. The DNA-PK complex and, in particular, DNA-PKcs, has a key role in NHEJ by localizing rapidly to DNA breaks and phosphorylating itself. The role of CK2 with respect to the DNA damage response has been analyzed in several model systems. In this study, we aimed to investigate the possible role of CK2 as an essential regulatory component of DNA-PKcs. The analysis of two isogenic human glioblastoma cell lines, M059K and M059J, which express and lack DNAPKcs, respectively, revealed pronounced differences in the response to cellular depletion of protein kinase CK2. Cell death was induced in M059K but not in M059J cells upon treatment with CK2-siRNA against the individual catalytic subunits. Lack of significant cell death in M059J cells permitted to exclude a direct responsibility of CK2, although this kinase has been shown to be essential for viability and proliferation in many cell systems (St-Denis and Litchfield, 2009). The results presented here provide for the first time in vivo evidence that CK2 is essential for modulating DNA-PK activity in normal and DNA-damaged cells by physical interaction and modulation of DNA-PKcs autophosphorylation. In the absence of DNA damage, targeting CK2a or -a0 by siRNA strongly enhances DNA-PKcs activity. As we did not observe CK2-mediated phosphorylation of DNA-PKcs in vitro, this suggests that phosphorylation is not involved in the modulation of DNA-PK activity. Furthermore chemical inhibition of CK2 with resorufin did not affect the autophosphorylation of DNA-PKcs, excluding a role mediated by CK2 kinase activity. However, both KU proteins were found phosphorylated by CK2a in vitro. Although the physiological effects of these phosphorylations remain to be elucidated, results indicate that CK2 might modulate DNA-PK activity via the KU proteins. Data obtained in vivo by the application of the in situ proximity ligation assay and in vitro by using recombinant proteins indicated that CK2 does interact with DNA-PKcs and negatively modulates its kinase activity. Cells lacking CK2 have higher DNA-PKcs activity and undergo cell death in support of the notion of a pro-cell death role of DNA-PK possibly via modulation of p53 as reported previously (Wang et al., 2000; Woo et al., 2002). Following induction of DNA damage, downregulation of the individual CK2 catalytic subunits led to extensive cell death in M059K cells indicating that the lack of CK2 sensitized cancer cells to NCS treatment. This was accompanied by a marked decrease of DNAPKcs autophosphorylation at S2056 with respect to cells treated solely with NCS. Moreover, CK2 downregulation resulted in a significant impairment of DNA DSBs repair as indicated by the persistent phosphorylation of histone H2AX in CK2-knockdown cells. Overall, these results provide evidence that block of DNA repair might be the consequence of the CK2-mediated cellular inhibition of DNA-PKcs. Having shown that downregulation of the individual CK2 catalytic subunits lowers DNA-PKcs autophosphorylation, the mechanism by which this effect is achieved remains to be elucidated. Further experiments are necessary to Oncogene


6024

substantiate the findings reported here, yet the enhanced physical interaction between CK2 and DNA-PKcs observed in cells treated with NCS strongly suggests that CK2 might be necessary to bring DNA-PKcs molecules in close proximity, by creating a docking site that facilitates intermolecular autophosphorylation of DNA-PKcs. Previously, it has been suggested that autophosphorylation at more than one site is likely to be required for DNA-PKcs function in NHEJ in vivo (Uematsu et al., 2007). Phosphorylation of DNA-PKcs at S2056 as well as at T2609 has been shown to occur in vivo in response to DNA damage (Chan et al., 2002; Douglas et al., 2002). Compelling evidence based on location, model systems and response to various DNA-damaging agents indicates that these amino-acid residues may be distinctly regulated. Although it is not yet clear whether the phosphorylation of the aforementioned amino acids reflects their different roles in the DNA-PKcs-mediated NHEJ repair, it certainly merits to further investigate the role of CK2 with respect to T2609 phosphorylation. Moreover, the identification of potential p-S2056-binding proteins in control and in CK2-depleted cells, respectively, would certainly contribute to unveil the DNA-PKcs-mediated molecular mechanisms operating in the NHEJ-repair pathway. In summary, we show that CK2 contributes to regulate the kinase activity and autophosphorylation of DNA-PKcs in cells. Strong evidence indicates that CK2 interacts in vivo with DNA-PKcs in normal growing cells and that the association increases in cells exposed to DNA damage. The lack of CK2 in cells treated with NCS is accompanied by marked decrease in DNA-PKcs autophosphorylation; this suggests that CK2 is essential for facilitating the interaction of DNA-PKcs molecules. Downregulation of CK2 sensitizes M059K cells to NCS treatment. Thus, we believe that CK2 may be an excellent target to overcome resistance of DNA-PKcs-proficient brain tumor cells toward radiation therapy.

M059K cells with 40 mM resorufin (Sigma, Brondby, Denmark) for 23 h before the addition of 0.5 mg/ml NCS for 1 h. Endogenous ATM was inhibited by incubating M059K cells with 10 mM KU55933 (Calbiochem, Nottingham, UK) for 4 h. To determine the mode of cell death, M059K cells were incubated with 0.5 mg/ml NCS for 24 h in combination with 5 mM z-Val-Ala-Asp-fluoromethylketone (zVAD(OMe)-fmk; Calbiochem) or 85 mM z-Phe-Ala-fmk (zFA-fmk; Calbiochem) for inhibition of caspases or cathepsins, respectively. Autophagy was analyzed by flow cytometry using MDC as reported (Zhang et al., 2009). Further details are provided in Supplementary Materials and Methods. Western blot analysis Whole-cell extracts and immunoblotting were as described (Achari and Lees-Miller, 2000; Olsen and Guerra, 2008a). The antibodies used were the following: mouse monoclonal antiCK2b, -CK2a and -p53 (all from Calbiochem); rabbit polyclonal anti-CK2a0 obtained by immunizing rabbits with a specific peptide (SQPCADNAVLSSGTAAR) of human CK2a0 ; mouse monoclonal anti-b-actin (Sigma); mouse monoclonal anti-PARP (BD Pharmingen, San Diego, CA, USA), rabbit polyclonal anti-KU86 and anti-BID, mouse monoclonal anti-KU70 and anti-DNA-PKcs (all from Santa Cruz Biotechnology, Santa Cruz, CA, USA); rabbit polyclonal anti-Chk2 (p-T68), anti-p53 (p-S15) and anti-H2AX (p-S139) (all from Cell Signaling Technology, Beverly, MA, USA); rabbit polyclonal anti-DNA-PKcs (p-S2056) (Abcam, Cambridge, UK); and rabbit polyclonal anti-BID (p-S78) (Bethyl Laboratories, Montgomery, TX, USA) and mouse monoclonal anti-CHK2 (Upstate Biotechnology, Frederikssund, Denmark). Purification of recombinant proteins CK2a, CK2a0 , CK2a2b2 and CK2a0 2b2 were expressed and purified as previously described (Niefind et al., 2000; Olsen et al., 2006, 2008b). DNA-PK was purified from K562 cells essentially as described (Matsumoto et al., 1997). In vitro protein kinase assay The activity of DNA-PK in high-salt whole extracts from M059K and M059J cells and kinase assays employing purified DNA-PK and recombinant CK2 were measured as described (Achari and Lees-Miller, 2000). Further details are provided in Supplementary Materials and Methods.

Materials and methods Cell culture The human glioblastoma cell lines M059K and M059J were obtained from the American Type Culture Collection (Rockville, MD, USA). The cells were cultured in Dulbecco’s modified Eagle’s medium (Invitrogen, Taastrup, Denmark) supplemented with 10% fetal bovine serum and maintained at 37 1C under 5% CO2. Transfection and cell treatments Cells were transfected with a set of four siRNA duplexes directed against CK2a, CK2a0 and CK2b (ON-TARGET plus SMARTpools), respectively, using Dharmafect I (Dharmacon, Lafayette, CO, USA) for 72 h according to the manufacturer’s instructions. Where indicated, cells were incubated with the radiomimetic drug NCS (a gift from Dr Hiroshi Maeda, Kumamoto University, Japan). Dead cells were measured with a FACSCalibur flow cytometer following the SYTOX greenbased assay (Invitrogen) as previously described (Yde et al., 2007). Endogenous CK2 activity was inhibited by incubating Oncogene

Immunofluorescence and in situ proximity ligation assay M059K cells grown on coverslips were incubated with rabbit polyclonal anti-H2AX (p-S139) antibody (Cell Signaling Technology) followed by incubation with biotinylated swine anti-rabbit IgG (Dako, Glostrup, Denmark) and streptavidin-conjugated fluorescein-isothiocyanate (Dako) as previously described (Olsen and Guerra, 2008a). Cells were counterstained with 40 ,6-diamidino-2-phenylindole, analyzed on a DMRBE microscope ( 630 magnification) equipped with a Leica DFC420C camera (Leica, Herlev, Denmark) and processed using ImageJ software (NIH, Bethesda, MD, USA). Cells grown on coverslips were analyzed for DNA-PKcs/CK2a0 interactions using a monoclonal anti-DNA-PKcs antibody (Santa Cruz Biotechnology) and a rabbit polyclonal antibody against CK2a0 according to the manufacturer’s instructions (Olink Biosciences, Uppsala, Sweden). The number of in situ proximity ligation signals was counted using the freeware software Blobfinder (http://www. cb.uu.se/Bamin/BlobFinder). Nuclei were visualized by Hoechst staining. Further details are provided in Supplementary Materials and Methods.


6025 In vivo DNA end-joining assay The plasmid-based end-joining assay was performed as previously described (Wang et al., 2006). pEGFP-N1 (Clontech-Takara Bio Europe, Saint-Germain-en-Laye, France) was linearized using HindIII (Fermentas, St Leon-Rot, Germany), which cuts between the promoter and the coding region of GFP. M059K cells were transfected with siRNA against CK2a or CK2a0 . At 48 h after transfection, cells were electroporated (NEON Tranfection system, Invitrogen) with linear or circular pEGFP-N1 according to the manufacturer’s instructions, and at 24 h after electroporation, the GFP intensity was detected by flow cytometry. Further details are provided in Supplementary Materials and Methods. Statistical and densitometrical analysis The statistical significance of differences between the means of two groups was evaluated by the two-tailed t-test and the level of significance was set as indicated in the figure legends.

Quantification of the intensity of protein bands on western blots was performed with the ImageJ software (NIH).

Conflict of interest The authors declare no conflict of interest.

Acknowledgements This study was supported by grants from the Danish Cancer Society (DP08152) and the Danish Natural Science Research Council (272–07–0258) to BG. BBO and OGI were supported by grants from the Danish Cancer Society (DP06083, DP07109).

References Achari Y, Lees-Miller SP. (2000). Detection of DNA-dependent protein kinase in extracts from human and rodent cells. Methods Mol Biol 99: 85–97. Ayoub N, Jeyasekharan AD, Bernal JA, Venkitaraman AR. (2008). HP1-beta mobilization promotes chromatin changes that initiate the DNA damage response. Nature 453: 682–686. Bibby AC, Litchfield DW. (2005). The multiple personalities of the regulatory subunit of protein kinase CK2: CK2 dependent and CK2 independent roles reveal a secret identity for CK2beta. Int J Biol Sci 1: 67–79. Biederbick A, Kern HF, Elsa¨sser HP. (1995). Monodansylcadaverine (MDC) is a specific in vivo marker for autophagic vacuoles. Eur J Cell Biol 66: 3–14. Block WD, Yu Y, Merkle D, Gifford JL, Ding Q, Meek K et al. (2004). Autophosphorylation-dependent remodeling of the DNAdependent protein kinase catalytic subunit regulates ligation of DNA ends. Nucleic Acids Res 32: 4351–4357. Canton DA, Zhang C, Litchfield DW. (2001). Assembly of protein kinase CK2: investigation of complex formation between catalytic and regulatory subunits using a zinc-finger-deficient mutant of CK2beta. Biochem J 358: 87–94. Chan DW, Chen BP, Prithivirajsingh S, Kurimasa A, Story MD, Qin J et al. (2002). Autophosphorylation of the DNA-dependent protein kinase catalytic subunit is required for rejoining of DNA doublestrand breaks. Genes Dev 16: 2333–2338. Chan DW, Lees-Miller SP. (1996). The DNA-dependent protein kinase is inactivated by autophosphorylation of the catalytic subunit. J Biol Chem 271: 8936–8941. Chen BP, Chan DW, Kobayashi J, Burma S, Asaithamby A, Morotomi-Yano K et al. (2005). Cell cycle dependence of DNAdependent protein kinase phosphorylation in response to DNA double strand breaks. J Biol Chem 280: 14709–14715. Chen BP, Uematsu N, Kobayashi J, Lerenthal Y, Krempler A, Yajima H et al. (2007). Ataxia telangiectasia mutated (ATM) is essential for DNA-PKcs phosphorylations at the Thr-2609 cluster upon DNA double strand break. J Biol Chem 282: 6582–6587. Ding Q, Reddy YV, Wang W, Woods T, Douglas P, Ramsden DA et al. (2003). Autophosphorylation of the catalytic subunit of the DNA-dependent protein kinase is required for efficient end processing during DNA double-strand break repair. Mol Cell Biol 23: 5836–5848. Douglas P, Cui X, Block WD, Yu Y, Gupta S, Ding Q et al. (2007). The DNA-dependent protein kinase catalytic subunit is phosphorylated in vivo on threonine 3950, a highly conserved amino acid in the protein kinase domain. Mol Cell Biol 27: 1581–1591.

Douglas P, Sapkota GP, Morrice N, Yu Y, Goodarzi AA, Merkle D et al. (2002). Identification of in vitro and in vivo phosphorylation sites in the catalytic subunit of the DNA-dependent protein kinase. Biochem J 368: 243–251. Guerra B, Issinger OG. (2008). Protein kinase CK2 in human diseases. Curr Med Chem 15: 1870–1886. Harrison JC, Haber JE. (2006). Surviving the breakup: the DNA damage checkpoint. Annu Rev Genet 40: 209–235. Helleday T, Lo J, van Gent DC, Engelward BP. (2007). DNA doublestrand break repair: from mechanistic understanding to cancer treatment. DNA Repair (Amst) 6: 923–935. Kuo LJ, Yang LX. (2008). Gamma-H2AX—a novel biomarker for DNA double-strand breaks. In vivo 22: 305–309. Litchfield DW. (2003). Protein kinase CK2: structure, regulation and role in cellular decisions of life and death. Biochem J 369: 1–15. Loizou JI, El-Khamisy SF, Zlatanou A, Moore DJ, Chan DW, Qin J et al. (2004). The protein kinase CK2 facilitates repair of chromosomal DNA single-strand breaks. Cell 117: 17–28. Mahaney BL, Meek K, Lees-Miller SP. (2009). Repair of ionizing radiation-induced DNA double-strand breaks by non-homologous end-joining. Biochem J 417: 639–650. Matsumoto Y, Suzuki N, Sakai K, Morimatsu A, Hirano K, Murofushi H. (1997). A possible mechanism for hyperthermic radiosensitization mediated through hyperthermic lability of Ku subunits in DNA-dependent protein kinase. Biochem Biophys Res Commun 234: 568–572. Merkle D, Douglas P, Moorhead GB, Leonenko Z, Yu Y, Cramb D et al. (2002). The DNA-dependent protein kinase interacts with DNA to form a protein-DNA complex that is disrupted by phosphorylation. Biochemistry 41: 12706–12714. Niefind K, Guerra B, Ermakowa I, Issinger OG. (2000). Crystallization and preliminary characterization of crystals of human protein kinase CK2. Acta Crystallogr D 56: 1680–1684. Olsen BB, Boldyreff B, Niefind K, Issinger OG. (2006). Purification and characterization of the CK2alpha0 -based holoenzyme, an isozyme of CK2alpha: a comparative analysis. Protein Expr Purif 47: 651–661. Olsen BB, Guerra B. (2008a). Ability of CK2beta to selectively regulate cellular protein kinases. Mol Cell Biochem 316: 115–126. Olsen BB, Rasmussen T, Niefind K, Issinger OG. (2008b). Biochemical characterization of CK2alpha and alpha’ paralogues and their derived holoenzymes: evidence for the existence of a heterotrimeric CK2alpha’-holoenzyme forming trimeric complexes. Mol Cell Biochem 316: 37–47. Povirk LF. (1996). DNA damage and mutagenesis by radiomimetic DNA-cleaving agents: bleomycin, neocarzinostatin and other enediynes. Mutat Res 355: 71–89. Oncogene


6026 Sandholt IS, Olsen BB, Guerra B, Issinger OG. (2009). Resorufin: a lead for a new protein kinase CK2 inhibitor. Anticancer Drugs 20: 238–248. St-Denis NA, Litchfield DW. (2009). Protein kinase CK2 in health and disease: from birth to death: the role of protein kinase CK2 in the regulation of cell proliferation and survival. Cell Mol Life Sci 66: 1817–1829. So¨derberg O, Gullberg M, Jarvius M, Ridderstra˚le K, Leuchowius KJ, Jarvius J et al. (2006). Direct observation of individual endogenous protein complexes in situ by proximity ligation. Nat Methods 3: 995–1000. Trembley JH, Wang G, Unger G, Slaton J, Ahmed K. (2009). Protein kinase CK2 in health and disease: CK2: a key player in cancer biology. Cell Mol Life Sci 66: 1858–1867. Uematsu N, Weterings E, Yano K, Morotomi-Yano K, Jakob B, Taucher-Scholz G et al. (2007). Autophosphorylation of DNAPKCS regulates its dynamics at DNA double-strand breaks. J Cell Biol 177: 219–229. Wang HC, Chou WC, Shieh SY, Shen CY. (2006). Ataxia telangiectasia mutated and checkpoint kinase 2 regulate BRCA1 to promote the fidelity of DNA end-joining. Cancer Res 66: 1391–1400. Wang S, Guo M, Ouyang H, Li X, Cordon-Cardo C, Kurimasa A et al. (2000). The catalytic subunit of DNA-dependent

protein kinase selectively regulates p53-dependent apoptosis but not cell-cycle arrest. Proc Natl Acad Sci USA 97: 1584–1588. Weterings E, Chen DJ. (2007). DNA-dependent protein kinase in nonhomologous end joining: a lock with multiple keys? J Cell Biol 179: 183–186. Woo RA, Jack MT, Xu Y, Burma S, Chen DJ, Lee PW. (2002). DNA damage-induced apoptosis requires the DNA-dependent protein kinase, and is mediated by the latent population of p53. EMBO J 21: 3000–3008. Yao KC, Komata T, Kondo Y, Kanzawa T, Kondo S, Germano IM. (2003). Molecular response of human glioblastoma multiforme cells to ionizing radiation: cell cycle arrest, modulation of the expression of cyclin-dependent kinase inhibitors, and autophagy. J Neurosurg 98: 378–384. Yde CW, Gyrd-Hansen M, Lykkesfeldt AE, Issinger OG, Stenvang J. (2007). Breast cancer cells with acquired antiestrogen resistance are sensitized to cisplatin-induced cell death. Mol Cancer Ther 6: 1869–1876. Zhang Y, Wu Y, Tashiro S, Onodera S, Ikejima T. (2009). Involvement of PKC signal pathways in oridonin-induced autophagy in HeLa cells: a protective mechanism against apoptosis. Biochem Biophys Res Commun 378: 273–278.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene