IKKβ Regulates the Repair of DNA Double-Strand Breaks Induced by Ionizing Radiation in MCF-7 Breast Cancer Cells

Activation of the IKK-NFκB pathway increases the resistance of cancer cells to ionizing radiation (IR). This effect has been largely attributed to the induction of anti-apoptotic proteins by NFκB. Since efficient repair of DNA double strand breaks (DSBs) is required for the clonogenic survival of irradiated cells, we investigated if activation of the IKK-NFκB pathway also regulates DSB repair to promote cell survival after IR. We found that inhibition of the IKK-NFκB pathway with a specific IKKβ inhibitor significantly reduced the repair of IR-induced DSBs in MCF-7 cells. The repair of DSBs was also significantly inhibited by silencing IKKβ expression with IKKβ shRNA. However, down-regulation of IKKα expression with IKKα shRNA had no significant effect on the repair of IR-induced DSBs. Similar findings were also observed in IKKα and/or IKKβ knockout mouse embryonic fibroblasts (MEFs). More importantly, inhibition of IKKβ with an inhibitor or down-regulation of IKKβ with IKKβ shRNA sensitized MCF-7 cells to IR-induced clonogenic cell death. DSB repair function and resistance to IR were completely restored by IKKβ reconstitution in IKKβ-knockdown MCF-7 cells. These findings demonstrate that IKKβ can regulate the repair of DSBs, a previously undescribed and important IKKβ kinase function; and inhibition of DSB repair may contribute to cance cell radiosensitization induced by IKKβ inhibition. As such, specific inhibition of IKKβ may represents a more effective approach to sensitize cancer cells to radiotherapy.


Introduction
The IkB kinase (IKK)-nuclear factor kB (NFkB) pathway is one of the most important cellular signal transduction pathways [1]. It consists of members of the NFkB family and the family of inhibitors of NFkB (IkB), the IkB kinase (IKK) complex, and various other regulatory components. The NFkB family includes RelA (p65), RelB, c-Rel, NFkB1/p105 (p50 precursor), and NFkB2/p100 (p52 precursor); the IkB family consists of IkBa, IkBb, IkBe, Bcl-3, p100/IkBd, and p105/IkBc; and the IKK complex is composed of two catalytic subunits, IKKa and IKKb, and the regulatory subunit IKKc. Normally, members of the NFkB family form a heterodimer/homodimer that resides in the cytoplasm as an inactive complex in association with a member of the IkB family. Upon stimulation with an inflammatory stimulus, the so-called canonical or classical pathway is activated, leading to the activation of IKK complex. Activated IKKa and/or IKKb phosphorylate IkBa at S-32 and S-36. This causes IkBa ubiquitination and degradation by the S26 proteasome, thereby, allowing NFkB to translocate into the nucleus to regulate NFkB target genes. Through regulation of its target genes, NFkB can regulate various physiologic processes such as cell proliferation, migration and survival.
In addition, an increasing body of evidence suggests that activation of the IKK-NFkB pathway also play a pivotal role in the development of cancer resistance to ionizing radiation (IR) and chemotherapy [2][3][4][5]. This is because IR and many chemotherapeutic agents can activate NFkB through the atypical NFkB activation pathway by induction of DNA double-strand breaks (DSBs) [6,7]. DSBs can activate ataxia telangiectasia mutated (ATM) that in turn phosphorylates IKKc at Ser85. This leads to IKKc mono-ubiquitination and translocation into the cytoplasm, where IKKc remains associated with ATM to activate IKKa and/ or IKKb. It has been shown that activation of the IKK-NFkB pathway renders many types of tumor cells more resistant to IR and chemotherapy presumably via induction of anti-apoptotic proteins [2][3][4][5]. Therefore, inhibition of the NFkB transcriptional activity has been extensively exploited as a novel approach to sensitize cancers to radiotherapy and chemotherapy, but has achieved mixed results [2][3][4][5]. Therefore, further studies are urgently needed to gain a better understanding on how activation of the IKK-NFkB pathway regulates tumor cell sensitivity to IR and chemotherapy before a molecular targeted therapy against the IKK-NFkB pathway can be effectively employed for cancer treatment.
It has been well established that IR kills cancer cells primarily by induction of DSBs and efficient repair of DSBs is required for the clonogenic survival of irradiated cells [8,9]. Therefore, we hypothesized that activation of the IKK-NFkB pathway by IR may also promote cancer cell survival in part by regulating the repair of DSBs. To test this hypothesis, we first used BMS-345541 (BMS), a specific IKKb inhibitor [10], to selectively inhibit the IKK-NFkB pathway and found that it could significantly inhibit the repair of IR-induced DSBs in MCF-7 human breast cancer cells and H1299 and H1648 human lung cancer cells. Interestingly, the repair of IR-induced DSBs in MCF-7 cells was not affected by down-regulation of IKKa, but was significantly inhibited by IKKb knockdown. In addition, the suppression of DSB repair by knockdown or inhibition of IKKb was associated with an increased sensitivity of MCF-7 cells to IR. DSB repair function and resistance to IR were completely restored in IKKbknockdown MCF-7 cells after reconstitution with an active form of IKKb. To our knowledge, this is the first study demonstrating that activation of the IKK-NFkB pathway by IR can regulate the repair of DSBs and inhibition of IKKb activity may sensitize cancer cells to IR at least in part via inhibition of DSB repair. Therefore, specific inhibition of IKKb represents a more effective approach to sensitize cancer cells to radiotherapy.

IKKb inhibition suppresses the repair of IR-induced DSBs
Activation of NFkB by IR depends on IKKb [6]. BMS is a potent and specific IKKb inhibitor and can effectively inhibit NFkB activation induced by diverse stimuli [10]. Therefore, we treated MCF-7 cells with BMS to determine whether activation of the IKKb-NFkB pathway regulates the repair of IR-induced DSBs by cH2AX foci assay [11]. As shown in Figure 1 A, MCF-7 cells exhibited a low level of DSBs prior to exposure to IR. The basal levels of DSBs were not significantly changed after incubation with BMS (p.0.05). Exposure to IR increased DSBs in MCF-7 cells and the increases were comparable in the cells treated with vehicle or BMS 1 hr after IR when the formation of cH2AX foci reached the peak level (Figure 1 A). The numbers of cH2AX foci in vehicle-treated cells declined rapidly thereafter and were almost back to the basal level at 6 hr after IR, indicating that these cells can efficiently repair IR-induced DSBs. In contrast, the numbers of cH2AX foci in BMS-treated cells remained significantly elevated 6 hr after IR. Moreover, even up to 24 hr after IR MCF-7 cells treated with 5 mM BMS still exhibited a significant increase in cH2AX foci. Similar findings were also observed when the formation of 53BP1 foci was used as an alternative surrogate to quantify IR-induced DSBs, as 53BP1 can be rapidly recruited by c H2AX to the sites of DSBs to form 53BP1 foci (Figure 1 B) [12]. These findings demonstrate that BMS can inhibit the repair of IRinduced DSBs in MCF-7 cells. To determine whether the effect of BMS is specific to MCF-7 cells and whether other IKKb inhibitors have a similar effect as BMS, we extended the studies to two additional human lung cancer cell lines H1299 and H1648 and two other potent IKKb inhibitors SC-514 [13] and TPCA-1 [14] and observed similar results as seen in MCF-7 cells treated with BMS ( Figure 2). However, among these inhibitors examined, BMS is the most potent inhibitor of DSB repair.
BMS is equally potent as DNA-dependent protein kinase (DNA-PK) and ATM inhibitors in inhibition of DSB repair NU7026 (NU) and KU55933 (KU) are well characterized DNA-PK and ATM inhibitors, respectively [15,16]. Both of them can potently inhibit DSB repair and sensitize various tumor cells to IR. Therefore, we compared the inhibitory effect of BMS with these of NU and KU on the repair of IR-induced DSBs. As shown in Figure 3 A and B, MCF-7 cells exhibited similar increases in cH2AX and 53BP1 foci 1 hr after IR in regardless of their pretreatment. At 6 hr after IR, the majority of IR-induced DSBs were repaired in vehicle-treated MCF-7 cells, whereas significantly fewer DSBs were repaired in the cells treated with BMS, NU or KU. Even at 24 hr after IR, substantial DSBs remained unrepaired in MCF-7 cells treated with BMS and NU. These findings demonstrate that BMS is equally potent as DNA-PK and ATM inhibitors in inhibition of the repair of IR-induced DSBs.
Although BMS is a selective inhibitor of IKKb, it is not know whether it inhibits DNA-PK and ATM [10]. Therefore, we examined the effects of BMS on DNA-PK and ATM in in vitro kinase assays. As shown in Figure 3 C and D, 5 mM NU and KU almost completely inhibited the kinase activities of DNA-PK and ATM, respectively. However, the same concentration of BMS had no such effect. Even at a higher concentration (10 mM), the kinase activities of DNA-PK and ATM remained unaffected by BMS. This result suggests that the inhibition of DSB repair by BMS is unlikely attributed to a non-specific inhibition of DNA-PK and ATM.

IKKb is essential for efficient repair of IR-induced DSBs
To further explore the requirement of IKK in efficient DSB repair, we generated stable IKKa and/or IKKb knockdown cell lines, e.g. IKKa(-), IKKb(-), and IKKa/b(-) cells, by transfection of MCF-7 cells with lentiviral short hairpin RNAs (shRNAs) that specifically target IKKa and/or IKKb mRNA. As shown in Figure 4 A-C, down regulation of IKKa and/or IKKb expression significantly inhibited IR-induced NFkB activation. Interestingly, IKKb(-) and IKKa/b(-) cells, but not IKKa(-) cells, exhibited a significant reduction in the repair of IR-induced DSBs (Figure 4 D and F). These findings suggest that IKKb, but not IKKa is essential for DSB repair, which was confirmed by the observations from IKKa and/or IKKb knockout mouse embryonic fibroblasts (MEFs) as well ( Figure S1).
To further validate the essential role of IKKb in DSB repair and determine whether the kinase activity of IKKb is required for the regulation, we reconstituted MCF-7/IKKb(-) cells with wildtype IKKb (WT-IKKb), kinase-dead IKKb (K44M-IKKb), and constitutively active IKKb (SSEE-IKKb) [17] by lentiviral transfection. The expression of these respective transgenes was confirmed by Western blot as shown in Figure 5 A. As reported in a previous study the cells transfected with WT-IKKb exhibited constitutive activation of the NFkB pathway as those transfected with SSEE-IKKb [17] and the activation could not be augmented by IR ( Figure 5  These results along with the data from BMS experiments confirmed that IKKb is critical for DSB repair and its kinase activity is indispensable for this function.

Inhibition of IKKb sensitizes MCF-7 cells to IR
Since efficient repair of IR-induced DSBs is required for the clonogenic survival of irradiated cells [8,9], we hypothesized that suppression of DSB repair via inhibition of IKKb kinase activity can sensitize tumor cells to IR. To test this hypothesis, we exposed MCF-7, MCF-7/IKKa(-), MCF-7/IKKb(-), and MCF-7/IKKa/ b(-) cells to 0, 1, 2 and 3 Gy IR, which led to a dose-dependent reduction in their survival rate (Figure 6 A). The reduction was greater in MCF-7/IKKb(-) and MCF-7/IKKa/b(-) cells than that in MCF-7 and MCF-7/IKKa(-) cells. Reconstitution of MCF-7/ IKKb(-) with WT-IKKb or SSEE-IKKb restored their resistance to IR, whereas MCF-7/IKKb(-)/K44M-IKKb cells remained equally sensitive to IR as vector-transfected MCF-7/IKKb(-) cells (Figure 6 B). In addition, pharmacological inhibition of IKKb kinase activity with BMS also sensitized MCF-7 cells to IRinduced clonogenic cell death (Figure 6 C). These findings suggest that inhibition of IKKb activity sensitizes MCF-7 cells to IR at least in part via inhibition of DSB repair. Altogether our data support the notion that activation of IKKb promotes the repair of DSBs and suppression of IKKb activity inhibits the repair of IRinduced DSBs and sensitizes certain cancer cells to IR-induced cell death. Discussion IR is one of the most widely used therapeutic modalities for cancer. Unfortunately, many tumor cells are inherently more resistant to IR or can acquire radioresistance shortly after radiotherapy, which inevitably leads to treatment failure and relapse of the disease [4]. An accumulating body of evidence suggests that constitutive activation of the IKK-NFkB pathway can contribute to cancer development, progression and resistance to cancer therapy [2,3], whereas activation of this pathway by IR can also render tumor cells more resistant to radiotherapy [4]. Therefore, inhibition of the IKK-NFkB pathway has the potential to increase the therapeutic index of radiotherapy [4].
Among various inhibitors of the IKK-NFkB pathway, IKKb inhibitors have emerged as the most promising anti-tumor agents and novel tumor sensitizers for IR and chemotherapy [13]. However, the mechanisms of their action have not been well studied but are presumably attributed to the inhibition of NFkB activity, which can increase tumor cell apoptosis by reducing the expression of anti-apoptotic proteins. The results from our studies reveal that IKKb inhibitors can also inhibit the repair of IRinduced DSBs. This effect is not due to a non-specific inhibition of DNA-PK and ATM but specific inhibition of IKKb, because DSB repair was also significantly inhibited by silencing IKKb expression but not by IKKa knockdown and the repair function was restored after reconstitution of a functional IKKb. Therefore, our results revealed a previously undescribed and important IKKb kinase function, e.g. regulation of DSB repair.
It has been well established that IR kills cancer cells primarily by induction of DSBs. DSBs are considered the most detrimental DNA lesions and a single unrepaired DSB is sufficient to kill a cell. Therefore, targeted inhibition of the DSB repair pathways has been actively pursued as a way to sensitize tumor cells to IR and other chemotherapeutic agents [8,9]. KU-55933 and NU-7026 are two well studied tumor sensitizers that inhibit DSB repair by targeting ATM and DNA-PK, respectively [15,16]. We found that the potency of the IKKb inhibitor BMS in inhibiting DSB repair is comparable to that of KU and NU. Since IKKb inhibitors such as BMS can inhibit not only DSB repair but also NFkB-mediated induction of anti-apoptotic proteins [2,3,18], they are potentially more advantageous than ATM and DNA-PK inhibitors as a radiosensitizer. Interestingly, even though BMS is cytotoxic to some tumor cells and can sensitize MCF-7 human breast cancer cells to IR, it is a relatively safe agent that does not cause noticeable normal tissue damage in vivo [19][20][21]. These findings highlight the therapeutic potential of IKKb inhibitors as an antitumor agent and a tumor sensitizer.
The mechanisms by which IKKb regulates DSB repair have yet to be elucidated. Our preliminary data showed that selective inhibition of the NFkB transcriptional activity by ectopical expression of a mutant IkBa or down-regulation of RelA by RNAi had no effect on the repair of IR-induced DSBs ( Figure  S2), indicating that the induction of NFkB-RelA activity is not required for the regulation of DSB repair. However, it remains to be determined if activation of the other members of the NFkB family by IKKb, such as c-Rel, may be involved in the regulation of DSB repair. For example, a recent report showed that activation of IKKb up-regulates the expression of Claspin via c-Rel [22]. Claspin can regulate DNA damage-activated checkpoint response by promoting ataxia telangiectasia and Rad3related protein (ATR)-mediated Chk1 phosphorylation and activation [23,24]. However, it may not be unexpected to find that IKKb may regulate DSB repair independent of NFkB, because several non-IkB targets of IKKb have been identified recently [25,26]. For example, it has been shown that IKKb can directly phosphorylate Aurora kinase A to regulate its stability for the maintenance of bipolar sindle assembly and genomic stability [23]. In addition, a recent study showed that IKKb translocates to the nucleus following UV irradiation [27]. It is plausible that IKKb enters the nucleus following IR treatments to assist DSB repair processes. Alternatively, it will be interesting to determine if IKKb-dependent DSB repair could be initiated by a mechanism involving the cytoplasmic IKKb-ATM axis [6,7,28]. Identification of IKKb substrate(s) required for DSB repair and elucidation of the mechanisms by which IKKb regulates DSB repair will therefore opens up a new model of DNA damage response in mammalian cells, which will be investigated in our future studies. In conclusion, for the first time, we demonstrate that IKKb regulates the repair of IR-induced DSBs. Moreover, IKKb, but not IKKa, is primarily responsible for promoting survival of certain tumor cells after IR at least in part by facilitating DSB repair. Surprisingly, NFkB-RelA is dispensable for IKKbdependent repair of DSBs. Therefore, IKKb inhibition or critical processes involved in the IKKb-dependent DSB repair pathway may be exploited as a novel therapeutic strategy to increase the sensitivity of tumor cells to IR. Cell lines MCF-7, H1299 and H1648 cell lines were originally obtained from ATCC (Manassas, VA). They were selected for our study because IR is a common therapeutic modality for breast and lung cancer and these cell lines have been extensively used in radiation research. MCF-7 cells stably transfected with the dominant-negative mutant IkBa (mIkBa or IkBa A32/36) were kindly provided by Dr. Jian Jian Li (University of California at Davis, Sacramento, CA). Immortalized wide type, IKKa, IKKb, and IKKa/b double knockout mouse embryonic fibroblasts were kindly provided by Dr. Shigeki Miyamoto (University of Wisconsin-Madison, Madison, WI) with the permission of Dr. Inder Verma (Salk Institute, La Jolla, CA). All these cells were maintained in Dulbecco's modified Eagle's minimum (DMEM) supplemented with 10% fetal bovine serum (HyClone, Logan, UT), penicillin (100 U/ml) and streptomycin (100 mg/ml) in a humidified incubator (95% air/5% CO 2 ) at 37uC.

DNA-PK kinase assay
The kinase activity of DNA-PK was measured using a SignaTECT DNA-Dependent Protein Kinase Assay System (Cat# V7870, Promega, Madison, WI). The biotinylated peptide substrate was incubated with 50 units purified DNA-PK (Cat# V5811, Promega) and (c-32 P)ATP in the presence or absence of BMS (5 or 10 mM) or NU (5 mM) for 5 min at 30uC according to the manufacturer's instructions. The biotinylated substrate was captured on a streptavidin membrane, washed and quantified by a Storm 860 Phosphorimager (Molecular Dynamics, Sunnyvale, CA).

ATM kinase assay
ATM was purified from irradiated MCF-7 cells by immunoprecipitation with anti-ATM antibody (Cat#A300-135A, from BETHYL Laboratories, Montgomery, TX) as previously described [30]. Aliquots of the purified ATM were incubated with 500 ng of recombinant H2AX (kindly provided by Dr. Benjamin Chen, University of Texas Southwestern Medical Center, Dallas, TX), 2 ml of 100 mM ATP and 10 mCi of (c-32 P)ATP in the presence or absence of BMS (5 or 10 mM) or KU (5 mM) at 30uC for 10 min. After SDS-polyacrylamide gel electrophoresis, c-32 P-H2AX was visualized by autoradiography.
Knockdown of IKKa and/or IKKb with short hairpin RNA (shRNA)
NFkB RelA DNA-binding activity assay Nuclear proteins were extracted from cells using the Nuclear Extract Kit (from Active Motif, Carlsbad, CA) per the manufacturer's protocol and were not contaminated by cytoplasmic elongation factor 2 (EF2) based on the result of Western blot using an antibody against EF2. NFkB-RelA DNA-binding activity was determined by a TransAM TM NFkB-RelA kit (Active Motif) using 5 mg of nuclear extract proteins according to the manufacturer's instructions.

Down-regulation of RelA with small interference RNA (siRNA)
To down-regulate the expression of RelA with siRNA, 5610 4 MCF-7 cells were seeded into a well of six-well plates. After overnight incubation, the medium was removed and then replaced with transfection media containing control (siGENOME Non-Targeting siRNA, D-001210-02-05) or RelA (siGENOME SMARTpool siRNA, M-003533-02-0005) siRNA (final concentration 50 nmol/L) along with DharmaFECT4 transfection reagent (Dharmacon) according to the manufacturer's protocol. After 24 h incubation, the transfection medium was removed and replaced with cell culture medium. The cells were allowed to grow for an additional 48 h to achieve maximal knockdown of RelA as shown by real-time RT-PCR and Western blot analyses.

Real-time PCR
Real-time PCR was done as previously described using the following primers: RelA, forward 59-CCTTCCTCATCC-CATCTT TG-39 and reverse 59-CCTCAATGTCCTCTTT-CTGC-39; and GAPDH, forward 59-CCC CAC ACA CAT GCA CTT ACC-39 and reverse 59-CCT ACT CCC AGG GCT TTG ATT-39. The threshold cycle (Ct) value for RelA was normalized to the Ct value of GAPDH. The relative RelA mRNA expression was calculated using the comparative C T (2 -DDCt ) method as previously described [29].
Clonogenic survival assay MCF-7 cells were seeded into wells of 12-well plates at 1610 4 cells/well. After overnight incubation, they were exposed to various doses (0, 1, and 3 Gy) of IR with or without pretreatment as indicated in individual experiments. The cells were allowed to grow for additional 12 days to form colonies before stained with 0.1% crystal purple. Colonies with more than 50 cells were counted. Survival fraction was calculated according to the plating efficiency of control cultures.

Statistical analysis
The data were analyzed by analysis of variance (ANOVA). In the event that ANOVA justified post hoc comparisons between group means, these were conducted using the Student-Newman-Keuls test for multiple comparisons. For experiments in which only single experimental and control groups were used, group differences were examined by unpaired Student t test. Differences were considered significant at P,0.05. All of these analyses were done using GraphPad Prism from GraphPad Software (San Diego, CA). Figure S1 IKKb but not IKKa knockout inhibits the repair of IR-induced DSBs in mouse embryonic fibroblasts. Mouse embryonic fibroblasts (MEF) from wild-type (WT), IKKa, IKKb, and IKKa/b knockout mice were exposed to 2 Gy IR. DSBs were analyzed by cH2AX and 53BP1 immunofluorescent staining at 1 h and 6 h after IR. Un-irradiated cells were included as controls (CTL). Representative photomicrographs (1006 magnifications) of cH2AX (red) and 53BP1 (green) immunofluorescent staining and nucleic counterstaining with Hoechst-33342 (blue) are shown in (A) and the average numbers of cH2AX and 53BP1 foci/cell from three independent experiments are presented (B) and (C) as mean 6 SE. * p,0.05, ** p,0.01, and *** p,0.001, vs. WT MEFs. (TIF) Figure S2 IKKb regulates DSB repair in a NFkB-RelA independent manner. (A) and (B) Ectopic expression of mIkBa inhibits IR-induced phosphorylation of IkBa and NFkB activation in MCF-7 cells. The levels of phosphorylated IkBa (p-IkBa) and total IkBa in the lysates from vector-or mIkBa-transfected MCF7 cells before (CTL) or 30 min after IR (2 Gy) were analyzed by Western blots. NFkB activation was analyzed by quantification of the levels of RelA in the nuclear extracts from vector-or mIkBatransfected MCF7 cells before (CTL) or 30 min after IR (2 Gy) by an ELISA assay. The data presented in (B) are mean 6 SE (n = 3). *** p,0.001, vs. vehicle. (C) Ectopic expression of mIkBa has no effect on the repair of IR-induced DSBs in MCF-7 cells. DSBs were analyzed by cH2AX immunofluorescent staining at 1 and 6 h after vector-or mIkBa-transfected MCF7 cells were exposed to 2 Gy IR. Un-irradiated cells were included as a control (CTL). The average numbers of cH2AX foci/cell from three independent experiments are presented as mean 6 SE. (D) Down-regulation of RelA mRNA expression by siRNA was confirmed by real-time PCR. The expression of RelA and GAPDH mRNA in RelA siRNA-treated cells was expressed as a percentage of that in control siRNA-treated cells. The data are presented as mean 6 SE (n = 3). *** p,0.001, vs. control siRNA treatment. (E) Downregulation of RelA expression by siRNA was confirmed by Western blot in MCF-7 cells transfected with control (CTL) or RelA siRNA. Un-transfected MCF-7 cells (Control) were included as a control. (F) Down-regulation of RelA expression by siRNA has no effect on the repair of IR-induced DSBs in MCF-7 cells. DSBs were analyzed by cH2AX immunofluorescent staining at 1 and 6 h after control (CTL siRNA) or RelA siRNA-transfected MCF7 cells were exposed to 2 Gy IR. Un-irradiated cells were included as controls (CTL). The average numbers of cH2AX foci/ cell from three independent experiments are presented as mean 6 SE.