Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 Dec 9.
Published in final edited form as: Mol Cell. 2011 Dec 9;44(5):785–796. doi: 10.1016/j.molcel.2011.09.026

p50 (NF-κB1) is an effector protein in the cytotoxic response to DNA methylation damage

Adam M Schmitt 1,4, Clayton D Crawley 1,4, Shijune Kang 1, David R Raleigh 1, Xiaohong Yu 1, Joshua S Wahlstrom 1, David J Voce 1, Thomas E Darga 2, Ralph R Weichselbaum 2, Bakhtiar Yamini 1
PMCID: PMC3240852  NIHMSID: NIHMS338292  PMID: 22152481

SUMMARY

The functional significance of the signaling pathway induced by O6-methylguanine (O6-MeG) lesions is poorly understood. Here, we identify the p50 subunit of NF-κB as a central target in the response to O6-MeG and demonstrate that p50 is required for SN1-methylator-induced cytotoxicity. In response to SN1-methylation, p50 facilitates the inhibition of NF-κB-regulated anti-apoptotic gene expression. Inhibition of NF-κB activity is noted to be an S-phase specific phenomenon that requires the formation of O6-MeG:T mismatches. Chk1 associates with p50 following SN1-methylation and phosphorylation of p50 by Chk1 results in the inhibition of NF-κB DNA binding. Expression of an un-phosphorylateable p50 mutant blocks inhibition of NF-κB-regulated anti-apoptotic gene expression and attenuates SN1-methylator-induced cytotoxicity. While O6-MeG:T-induced, p50-dependent signaling is not sufficient to induce cell death, this pathway sensitizes cells to the cytotoxic effects of DNA breaks.

INTRODUCTION

The DNA damage response (DDR) is a highly conserved signal transduction pathway activated following the recognition of cellular stress. In this hierarchical pathway, it is proposed that damage sensors pass information on to transducers that control the cellular response via effectors (Harper and Elledge, 2007). Effector proteins facilitate the downstream response by regulating the expression of specific genes. It is generally accepted that the DDR acts to maintain genomic integrity by activating cellular programs that either result in damage repair or removal of the damaged cell. Genotoxic chemotherapeutics often mediate their therapeutic effect by exploiting the DDR to induce cytotoxicity. In this regard, loss or depletion of an effector protein can render a cell resistant to chemotherapeutic-induced cytotoxicity.

O6-MeG, the primary cytotoxic lesion formed by SN1-type methylating agents such as temozolomide (TMZ), induces cytotoxicity by a mechanism involving mismatch repair (MMR). Under physiological conditions, O6-MeG lesions that are not repaired by O6-methylguanine-DNA methyltransferase (MGMT) can mispair with deoxythymidine residues to form O6-MeG:T mismatches. Following mismatch recognition, a signaling response is activated that involves the apical transducer, ATR, and its downstream kinase, Chk1 (Stojic et al., 2004). Although this pathway is specifically activated in response to DNA damage (Yoshioka et al., 2006), the functional significance of the response is poorly understood.

A connection between DNA damage and NF-κB has been clearly established (Campbell et al., 2004; Wu and Miyamoto, 2007 and references therein; Yamini et al., 2007). Mammalian NF-κB proteins act as dimers and include: p50 (NF-κB1, p105), p52 (NF-κB2, p100), p65 (RelA), c-Rel, and RelB. Subunit dimerization and DNA binding are mediated by residues within the conserved Rel homology domain (RHD) in the N-terminal region of each protein. At rest, NF-κB is retained in the cytoplasm through interaction with the inhibitor-κB (IκB) proteins (Vallabhapurapu and Karin, 2009). Following IκB degradation, NF-κB dimers translocate to the nucleus where they bind specific κB-sites in the promoter region of genes involved in cell survival, apoptosis and the immune system (www.NF-kB.org). While IκB protein degradation represents the primary control point for NF-κB activation, NF-κB-dependent gene expression is also regulated by post-translational modification of individual subunits (Perkins, 2006). Although NF-κB is best known as a stimulus-induced transcription factor, even in resting cells NF-κB activity is seen, a finding likely due to the basal shuttling of NF-κB dimers between the cytoplasm and nucleus (Birbach et al., 2002; Carlotti et al., 2000).

A significant body of work has elucidated the mechanism by which O6-MeG lesions induce cytotoxicity. However, studies examining the link between upstream damage repair proteins and downstream cellular effects are rare. Here, we identify p50 as a central target in the damage response to O6-MeG and show that p50-mediated signaling is a necessary factor in MMR-dependent cytotoxicity.

RESULTS

Inhibition of NF-κB by SN1-methylating agents requires p50

We previously noted that SN1-methylators inhibit the transcriptional activity of NF-κB by a mechanism similar to the pathway by which they induce cytotoxicity (Yamini et al., 2007). Based on this observation, it was hypothesized that a target in the NF-κB dimer that enables inhibition of NF-κB activity is also necessary for induction of apoptosis. As p65 and p50 make up the NF-κB dimer at rest (Figure S1A), these subunits were studied. In both wild type (wt) and p65−/− mouse embryonic fibroblasts (MEFs), TMZ inhibits NF-κB activity as assessed by a luciferase reporter bearing three repeats of the immunoglobulin k-light chain κB-site upstream of a minimal IFN-β promoter (Yamini et al., 2007) (Figure 1A). However, in p50−/− (nfkb1−/−) MEFs, inhibition of NF-κB by TMZ is attenuated (Figure 1B), as is inhibition of NF-κB by the related SN1-methylator, methylnitrosourea (MNU) (data not shown). This finding is recapitulated in human glioma cells that have undergone siRNA-mediated knockdown of p65 or p50 (Figure S1B and C). Of note, TMZ blocks both endogenous and TNFα-induced NF-κB activity (Yamini et al., 2007 and data not shown). Importantly, re-expression of p50 in p50−/− MEFs reconstitutes the ability of TMZ to inhibit NF-κB (Figure 1C). The above findings suggest that p50 is necessary for inhibition of NF-κB activity by TMZ. As a further control, a luciferase reporter controlled by the human BclXL promoter with a functional κB-site was used. TMZ inhibits expression from the wt-BclXL promoter/reporter in wt but not p50−/− MEFs and mutation of the κB-element (ΔκB BclXL) blocks the effect of TMZ demonstrating the importance of the κB-site for p50-mediated inhibition of NF-κB by TMZ (Figure 1D).

Figure 1. p50 is required for inhibition of NF-κB activity by SN1-methylators.

Figure 1

Open in a new tab

(A–C) NF-κB (Ig-κB luc)-dependent luciferase assays. Data show mean NF-κB-dependent luciferase, relative to renilla, normalized to value without TMZ, ± SD of triplicate samples.

(A) Wt and p65−/− MEFs were pre-treated with vehicle or TMZ (16 hrs) followed by 10 ng/ml TNFα. *p < 0.005. (Inset: immunoblot with anti-p50 or p65).

(B) Wt and p50−/− MEFs were treated with TMZ or vehicle (16 hrs). (Inset: Immunoblot). *p < 0.03.

(C) p50−/− MEFs co-transfected with pCMV-p50 (p50) or pCMV (vector) and Ig-κB luc were treated with TMZ or vehicle. *p < 0.05.

(D) Wt BclXL or ΔκB BclXL luciferase reporter assays in MEFs treated with 100 μM TMZ or vehicle (mean ± SD of triplicate samples). *p < 0.02 relative to untreated.

See also Figure S1.

Inhibition of NF-κB by TMZ is dependent on cellular MGMT activity (Yamini et al., 2007). Wt and p50−/− MEFs express similar MGMT activity, and 6-benzylguanine (BG), a specific MGMT inhibitor (Dolan et al., 1990), obliterates the activity in both cell lines (Figure S1D). BG does not sensitize p50−/− MEFs to inhibition of NF-κB by TMZ (Figure S1E) indicating that loss of p50 attenuates inhibition of NF-κB by a mechanism not involving changes in MGMT activity.

SN1-methylators attenuate NF-κB-dependent gene expression by inhibiting DNA binding of p50-containing NF-κB

NF-κB DNA binding was next assessed by gel shift assay (EMSA). TMZ blocks basal NF-κB binding in wt but not p50−/− MEFs (Figure 2A). Consistent with the prior observation that TMZ inhibits NF-κB via a post-translocational effect (Yamini et al., 2007), TMZ does not alter Iκbα degradation kinetics in wt or p50−/− cells (Figure S2A). To examine binding in human cells, stable U87 glioma transfectants expressing p50 (p105) or control shRNA were constructed (Figure 2B). As p50 is formed from the N-terminus of p105, an shRNA targeting the C-terminus of p105 was used to allow re-expression of p50 (Figure S2B). Supershift studies show that in these p50-depleted cells, p52 and p65 are present in the NF-κB dimer (Figure S2C), a finding also noted in p50−/− MEFs (Hoffmann et al., 2003). TMZ inhibits NF-κB binding in glioma cells expressing control but not p50 shRNA (Figure 2B, lower).

Figure 2. SN1-methylators block the DNA binding of p50-contaning NF-κB.

Figure 2

Open in a new tab

(A) EMSA with Ig-κB probe. Top panel- MEFs were treated with 100 μM TMZ for the indicated time. Bottom panel- MEFs were treated with TMZ for 16 hrs. SC- specific competitor. NS- non-specific competitor.

(B) Top panel- Immunoblot with the indicated antibody. Bottom panel- EMSA of stable transfectants following treatment with 100 μM TMZ or vehicle (16 hrs).

(C) Nuclear extracts were isolated from cells either unstimulated (lanes 1 and 6) or stimulated with TNFα (30 min). Extracts were treated with 10 μM TMZ for the indicated time (min). 15 ng wt-p50 was given to sh-p50 extracts prior to treatment where indicated. Nonspecific (NS) and specific (SC) competitors and p50 supershift (α-p50) were used as shown. The sample was divided following treatment and EMSA (upper panel) or IB (lower panel) of the indicated lanes was performed. Data demonstrate inhibition of NF-κB DNA binding by TMZ is reconstituted in sh-p50 extract following addition of wt-p50 without loss of p50 protein.

(D) qChIP following treatment with vehicle or 100 μM TMZ (16 hrs). IP performed with anti-p65 or -RNA Pol II and qPCR with BclXL promoter specific primers. Data represent enrichment of p65 or RNA Pol II, relative to IgG, ± SEM of three independent experiments. *p < 0.01 relative to untreated.

(E) qPCR of endogenous mRNA in sh-control (c) and sh-p50 (p50) U87 cells treated with 100 μM TMZ or vehicle (16 hrs). Data show relative mRNA expression, normalized to untreated, ± SD of triplicate samples from three separate experiments. *p < 0.05 relative to untreated.

See also Figure S2.

Next, to examine the effect of p50 re-expression, nuclear extracts were isolated and treated in vitro. Although TMZ inhibits NF-κB binding in extracts from p50-proficient cells (sh-c), inhibition of NF-κB in p50-depleted extracts (sh-p50) is only seen after supplementing with p50 protein (Figure 2C). Importantly, nuclear extracts contain the MMR machinery necessary for inhibition of NF-κB (Yamini et al., 2007) and have been shown to undergo DNA replication (Krokan et al., 1975), a further requirement for inhibition by SN1-methylators (see Figure 4).

Figure 4. Inhibition of NF-κB by O6-MeG is S-phase dependent.

Figure 4

Open in a new tab

(A) Ig-κB luciferase assay in U87 cells serum starved and treated with vehicle or TMZ (16 hrs).

(B) Upper- FACS analysis of DNA content in U87 cells at the indicated time (hours) following release from double thymidine block. US- unsynchronized cells. Lower- diagram of treatment schedule for (C).

(C) Ig-κB luciferase assay at indicated time after thymidine release (x-axis) following treatment with vehicle (white bars) or 100 μM TMZ (black bars). *p < 0.001 relative to vehicle.

(D) Ig-κB EMSA in unsynchronized (US) and synchronized cells (0, 2, 4, 6 and 16 hours after release) following treatment with vehicle or 100 μM TMZ. SC- specific competitor, NS- nonspecific competitor.

(E) Ig-κB EMSA. 20 nM duplex substrate was administered to nuclear extracts of U87 cells and EMSA performed. SC and NS- as above.

(F and G) Ig-κB luciferase assay 6 hours after transfection with indicated duplex substrate. (F) U87 cells. *p < 0.005 relative to G:C.

(G) The indicated cell line was treated with 20 ng duplex. *p < 0.05 relative to G:C.

(H) Ig-κB luciferase assay in wt and p50−/− MEFs treated with 6-TG (16 hours). *p < 0.05 relative to untreated.

Luciferase assay data represent mean ± SD of triplicate samples.

See also Figure S4.

As TMZ specifically blocks NF-κB DNA binding, recruitment of NF-κB to an endogenous promoter was next examined using ChIP. We previously noted that TMZ attenuates recruitment of p65 to the BclXL promoter (Yamini et al., 2007). As anticipated, in MMR-proficient cells (TK6), TMZ also attenuates recruitment of p50 to this promoter with similar kinetics to the inhibition of NF-κB activity (Figure S2D). To examine the effect of p50 loss on NF-κB recruitment, quantitative ChIP was performed using an antibody directed against p65. Although TMZ reduces binding of p65 to the BclXL promoter in control cells, TMZ does not alter p65 binding in p50-deficient cells (Figure 2D). These findings indicate that p50 is required for inhibition of NF-κB binding by TMZ in vivo and suggest that TMZ inhibits the entire NF-κB complex via an effect specifically on p50.

The importance of p50 for inhibition of NF-κB raises the question of whether p50 is necessary for inhibition of NF-κB-regulated genes. TMZ inhibits the basal expression of several NF-κB-regulated, anti-apoptotic genes without affecting two NF-κB-independent genes, thioredoxin (TXN) and STAT1 (Figure 2E). Similar inhibition is seen in murine cells (Figure S2E). Depletion of p50, in both human and mouse cells, blocks the down-regulation of the anti-apoptotic genes (Figure 2E and 6B), suggesting that p50 is necessary for inhibition of endogenous NF-κB-regulated genes by TMZ.

Figure 6. Phosphorylation of p50 mediates SN1-methylator-induced cytotoxicity.

Figure 6

Open in a new tab

(A) qChIP in U87 cells stably expressing GFP-tagged empty vector, p50wt, p50S329A or p50S329D. Cells were treated with vehicle or 100 μM TMZ for 16 hrs and IP performed with the indicated antibody. qPCR was performed with BclXL promoter specific primers. Data represent promoter enrichment of p50 or histone H1 relative to IgG control ± SEM of 3 separate experiments. *p < 0.05 relative to untreated p50wt sample. Inset: Immunoblot with anti-p50 demonstrates equal expression of GFP-p50 mutants in stable clones.

(B) qPCR of endogenous mRNA in wt MEFs and p50−/− stable clones expressing empty vector (EV), p50wt or p50S329A following treatment with 100 μM TMZ or vehicle (16 hrs). Data show mean, normalized to untreated, ± SD of triplicate samples from three experiments. *p < 0.05 relative to untreated.

(C) Colony forming assay in wt and p50−/− MEF stable clones treated with TMZ or vehicle. Cells include parental wt and p50−/− MEFs, and p50−/− MEFs expressing empty vector (EV), p50wt or p50S329A. Data show mean surviving fraction ± SD of duplicate samples, repeated thrice. *p < 0.02 relative to similarly treated p50S329A expressing cells. Lower panel: Immunoblot with anti-p50.

Data from stable transfectants are representative of >1 clone.

See also Figures S6.

Loss of p50 results in resistance to SN1-methylator-induced apoptosis

Given that p50 is required for inhibition of anti-apoptotic genes by TMZ, the requirement of p50 for TMZ-induced apoptosis was next examined. Depletion of p50, but not p65, attenuates TMZ-induced annexin V binding (Figure 3A). Similarly, loss of p50 blocks induction of apoptosis by the SN1-methylator, MNU (Figure 3B). Interestingly, p50 loss does not affect viability following doxorubicin or ionizing radiation (IR) (Figure S3A), suggesting that p50 depletion does not lead to general resistance to damage-induced cell death. Re-expression of p50 sensitizes p50−/− MEFs to induction of apoptosis by TMZ (Figure 3C). These findings suggest that p50 is necessary for SN1-methylator induced apoptosis.

Figure 3. Loss of p50 results in resistance to SN1-methylator-induced apoptosis.

Figure 3

Open in a new tab

(A–C) FACS analysis of annexin V binding 72 hours following treatment (mean ± SD of triplicate samples, repeated at least twice with similar results).

(A) U87 cells transfected with control, p50- or p65-siRNA were treated with vehicle or 100 μM TMZ. *p > 0.3.

(B) MEFs were treated with MNU. *p < 0.001 relative to vehicle.

(C) p50−/− MEFs co-transfected with empty vector (EV) or wt-p50 (p50) and GFP were treated with 100 μM TMZ. Data represent the GFP-positive population.

(D and E) Colony forming assays in wt and p50−/− MEFs (mean ± SD of duplicate samples shown, repeated thrice with similar results).

(D) Cells were treated with the indicated concentrations of TMZ. *p < 0.03. **p < 0.005.

(E) Cells were pretreated with 20 μM BG (2 hours) and then with TMZ. *p < 0.02.

(F) Induction of γH2AX foci in sh-c and sh-p50 cells untreated or treated with 100 μM TMZ.

Left: representative merged image of γH2AX foci and nuclei. Right: mean γH2AX foci per nucleus, ± SD of 200 cells per group. *p < 0.0005 relative to untreated. See also Figure S3.

Although loss of p50 attenuates TMZ-induced apoptosis, it is possible that pathways such as senescence or mitotic catastrophe eventually result in replicative arrest. However, deletion of p50 attenuates the reduction of colony formation induced by TMZ (Figure 3D) and MNU (Figure S3B) suggesting that p50 is a genuine mediator of SN1-methylator-induced cytotoxicity.

Proliferation rate and MGMT activity also contribute to the response to SN1-methylators. Wt and p50-depleted cells have similar proliferation rates as determined by doubling time (Figure S3C), and BG does not sensitize p50−/− cells to killing by TMZ (Figure 3E). These results suggest that the resistance to apoptosis mediated by loss of p50 is not due to a change in proliferation rate or MGMT activity. Decreased formation of double strand breaks (DSBs) is also a potential cause for the reduced cytotoxicity seen with loss of p50. However, the observation that similar levels of γH2AX foci are induced in wt and p50-depleted cells by TMZ (Figure 3F) suggests that the resistance to cytotoxicity with loss of p50 is not due to decreased DSB formation or increased DSB repair.

In addition to cytotoxicity, the MMR-dependent response to SN1-methylation involves G2/M cell cycle arrest (Stojic et al., 2004). However, TMZ causes similar levels of G2/M arrest in wt and p50-depleted glioma cells (Figure S3D) and wt and p50−/− MEFs (Figure S3E). These findings suggest that p50 specifically mediates SN1-methylator-induced cytotoxicity and not MMR-dependent G2/M arrest.

Inhibition of NF-κB by SN1-methylation is S-phase dependent

Inhibition of NF-κB by TMZ is attenuated when cell turnover is slowed by serum starvation (Figure 4A and S4A), suggesting that progression through the cell cycle is necessary for the inhibitory effect. To examine this hypothesis, U87 cells were synchronized at G1 by double-thymidine block and at various times following release treated for 3 hours with TMZ (Figure 4B). Although 3-hour TMZ does not inhibit NF-κB in unsynchronized cells, 3-hour TMZ exposure inhibits NF-κB in cells that have been released for 4–6 hours, a time point that correlates with passage through the first S-phase after release (Figure 4C). Inhibition of NF-κB is attenuated when 3-hour TMZ is given either at release or 16 hours after release, when most cells are in G1. EMSA studies mirror the luciferase data showing that 2-hour TMZ inhibits NF-κB DNA binding when treatment is initiated 4–6 hours after thymidine release (Figure 4D).

The requirement of S-phase suggests that it is specifically O6-MeG:T mismatches that block NF-κB. To examine this hypothesis, oligonucleotide (oligo) duplex substrates bearing a unique central bp (O6-MeG:T, O6-MeG:C, G:T or G:C) were constructed (Figure S4B). These substrates are similar to ones previously reported (Yoshioka et al., 2006) and were biotinylated at both 5′ and 3′ ends to discourage their degradation. Initially, duplexes were administered to nuclear extracts to investigate their effect on DNA binding. O6-MeG:T substrate inhibits NF-κB binding substantially more than control duplexes (Figure 4E). Next, substrates were introduced into live cells in the presence of an NF-κB-dependent reporter. O6-MeG:T, but not control, substrate decreases NF-κB activity (Figure 4F), a finding noted only in p50-proficient glioma cells (Figure 4G) and MEFs (Figure S4C). To examine p50 re-expression, nuclear extracts of sh-p50 cells were supplemented with p50 protein. Inhibition of NF-κB binding by O6-MeG:T is reconstituted in the presence of wt-p50 (p50wt) (Figure 5K, lane 5). These results suggest that O6-MeG:T mismatches act via p50 to inhibit NF-κB DNA binding and activity. The lack of an effect by non-mismatched, O6-MeG:C, substrate supports the hypothesis that S-phase is required for inhibition. In addition, isolation of O6-MeG within the duplex strongly supports the contention that O6-MeG specifically, and not another SN1-methylator-induced lesion, inhibits NF-κB, a finding initially proposed using cells with varying MGMT levels (Yamini et al., 2007).

Figure 5. Phosphorylation of p50 mediates inhibition of NF-κB.

Figure 5

Open in a new tab

(A–C) Ig-κB luciferase assays following pre-treatment with vehicle or 100 μM TMZ (16 hours) followed by 10 ng/ml TNFα (mean ± SD of triplicate samples)

(A) U20S cells stably expressing tet-on ATR-wt and ATR-kd. *p < 0.002.

(B) ATM+/+ and ATM−/− cells.

(C) U87 cells expressing myc-tagged wt or kinase dead Chk1. *p < 0.01.

(D) Co-IP study. U87 cells were treated with vehicle or 100 μM TMZ (16 hours) and nuclear extracts isolated. Left- IP with the indicated antibody and IB with anti-p50 or anti-Chk1. Right- immunoblot of input nuclear sample.

(E) Chk1 consensus phosphorylation sequence (human p50 ser329- underlined). *phosphorylated residue.

(F) Kinase assay using immunoprecipitated FLAG-tagged wt-Chk1 (wt) or kd-Chk1 (kd) and p50wt (wt) or p50S329A (mt). Autoradiogram (32P) and IB of the same blot shown.

(G) Kinase assay using purified recombinant Chk1 and p50. IB was performed with the indicated antibodies, including anti-phospho-ser329-p50 (p-p50). Bottom panel- EMSA of the same blot.

(H) Kinase assay following incubation with G:C or O6-MeG:T (O6:T) substrate using nuclear extracts from sh-p50 cells supplemented with p50wt or p50S329A (S329A). Chk1 inhibitor, Gö6976 (Gö). Autoradiogram (32P) and IB of the same blot shown.

(I) In vivo p50 ser329 phosphorylation. Left- U87 cells were treated with vehicle or 100 μM TMZ. Right- U87 cells incubated with siRNA (48 hrs) were treated with vehicle or 100 μM TMZ (16 hrs). Immunoblot was performed with anti-phospho-S329-p50 (p-p50) and the stripped membrane re-probed with anti-p50 and/or anti-Chk1.

(J) Ig-κB luciferase assay in p50−/− MEFs co-transfected with Ig-κB luc and either empty vector, wt-p50 or p50S329A. TMZ (0, 25, 100, 250 μM) was given for 16 hours (mean ± SD of triplicate samples shown). *p < 0.005 relative to 0 μM TMZ.

(K) Ig-κB EMSA of sh-p50 nuclear extracts supplemented with p50wt or p50S329A and treated with duplex substrate as in figure 5H. SC and NS- specific and non-specific competitor, respectively. Blot demonstrates that under kinase conditions, O6-MeG:T but not control substrate blocks the DNA binding of p50wt, not p50S329A.

(L) Immunoblot (left) and EMSA (right) of p50wt and p50S329D protein.

(M) Ig-κB luciferase assay in U87 cells co-transfected with Ig-κB-luc/renilla and either empty vector (EV), p50wt, p50S329A or p50S329D (mean ± SD of triplicate samples shown). *p < 0.01. See also Figures S5.

The requirement of mismatched bases is consistent with the hypothesis that MMR plays a role in the inhibition of NF-κB; an observation demonstrated using paired MMR-proficient and deficient cell lines (Yamini et al., 2007) and supported by knock-down of a specific MMR subunit (Figure S4D). The need for MMR suggests that non-SN1-methylating agents that act via MMR-dependent signaling, such as 6-thioguanine (6-TG) (Swann et al., 1996), may also block NF-κB. 6-TG inhibits NF-κB activity, and as with SN1-methylators, inhibition by 6-TG requires active MMR (data not shown) and is attenuated in cells depleted of p50 (Figure 4H). In addition, loss of p50 renders cells resistant to 6-TG (Figure S4E) a finding that supports the role of p50 in the general MMR-dependent cytotoxic response.

SN1-methylators inhibit NF-κB activity by Chk1-mediated phosphorylation of p50

The kinetics of NF-κB inhibition are similar to those of MMR-induced damage signaling (Stojic et al., 2004; Yoshioka et al., 2006) suggesting that DDR intermediates may be involved in the inhibitory effect. The general ATM/ATR inhibitor, caffeine, attenuates inhibition of NF-κB by TMZ (Figure S5A) as does expression of kinase-dead ATR (Figure 5A). However, TMZ inhibits NF-κB equally in isogenic ATM−/− and ATM+/+ cells (Figure 5B). Together, these results suggest that inhibition of NF-κB involves ATR not ATM, an observation consistent with the finding that ATR is activated within 12 hours of SN1-methylation whereas ATM is activated later (Stojic et al., 2004), after inhibition of NF-κB has already occurred.

The importance of ATR suggests that Chk1 may also be necessary for inhibition of NF-κB and, consistent with previous reports (Stojic et al., 2004), TMZ activates Chk1 in MMR-proficient (TK6) cells (Figure S5B). Over-expression of kinase dead Chk1 (Figure 5C), or addition of Chk1 inhibitor (Figure S5C), attenuates inhibition of NF-κB by TMZ. To examine whether Chk1 and p50 associate following TMZ treatment, co-IP studies were next performed. Reciprocal co-IP demonstrates that TMZ induces the association of Chk1 with p50 in the nucleus at a time point consistent with the inhibition of NF-κB (Figure 5D).

Human p50 contains a conserved Chk1 consensus phosphorylation site at serine 329 (S329) (Hutchins et al., 2000) (Figure 5E). A second minimum consensus site is present at S338. However, as phosphorylation of S338 enhances p50 DNA binding (Hou et al., 2003), we focused on S329 and examined the ability of Chk1 to phosphorylate this residue. Immunopurified wt-Chk1, but not kd-Chk1, incorporates radiolabeled phosphorous into p50wt (Figure 5F). Phosphorylation is substantially decreased when S329 is mutated to alanine (p50S329A) or when Chk1 inhibitor (Gö6976) is added. To exclude potential contaminants in the immunoprecipitated Chk1, purified recombinant Chk1 was also used. Purified Chk1 phosphorylates p50wt at S329 as assessed both by autoradiography and by a phospho-specific anti-p50-S329 antibody (Figure 5G). Moreover, incubation of p50wt, but not p50S329A, with Chk1 inhibits p50 binding on EMSA (Figure 5G, lower panel). Taken together, the above findings suggest that Chk1 phosphorylates p50 at S329 and further, that this phosphorylation blocks p50 DNA binding. Kinase assay was also performed using duplex substrates. Nuclear lysates from p50-depleted cells were supplemented with p50wt or p50S329A and incorporation of 32P examined following the addition of duplexes. Of note, O6-MeG:T substrate has previously been shown to activate Chk1 (Yoshioka et al., 2006) a finding recapitulated in our hands (data not shown). O6-MeG:T, but not control substrate, induces phosphorylation of p50wt but not p50S329A, an effect blocked by Gö6976 (Figure 5H), suggesting that O6-MeG:T mismatches induce Chk1-dependent S329 phosphorylation.

Next, the ability of TMZ to phosphorylate p50 in vivo was examined using the phospho-S329 antibody. TMZ induces phosphorylation of S329 at a time point consistent with the inhibition of NF-κB activity (Figure 5I). S329 phosphorylation is blocked by Chk1-specific siRNA (Figure 5I- right panel) and Chk1-inhibitor (data not shown) suggesting that TMZ-induced S329 phosphorylation is Chk1-dependent. Specificity of the S329 antibody was confirmed by competition assay using the phospho-peptide immunogen compared to the identical non-phospho peptide (Figure S5D). In addition, TMZ induces the activation of Chk1 in p50−/− cells (Figure S5E) supporting the hypothesis that Chk1 is activated upstream of p50.

As TMZ induces S329 phosphorylation, the functional affect of S329A mutation on inhibition of NF-κB was next examined. Expression of p50S329A does not reconstitute the inhibition of NF-κB activity by TMZ in p50−/− MEFs (Figure 5J) or sh-p50 glioma cells (Figure S5F). Similarly, p50S329A does not reconstitute the inhibition of NF-κB DNA binding by O6-MeG:T in sh-p50 nuclear extract (Figure 5K, compare lanes 9 and 5). Importantly, p50S329A is a functional p50 analog, as it efficiently binds κB-DNA (Figure S5G) and when it is co-expressed with p65, p50S329A induces NF-κB-dependent activity to a similar extent as p50wt and p65 (Figure S5H). Taken together, these findings indicate that TMZ signals via Chk1, to induce p50 phosphorylation at S329 resulting in inhibition of NF-κB DNA binding and activity.

The significance of S329 phosphorylation was also examined using a phospho-mimetic p50 mutant, p50S329D. Purified p50S329D binds kB-DNA substantially less than p50wt (Figure 5L) and expression of p50S329D results in a decrease in NF-κB-dependent transcription compared to p50wt (Figure 5M). These findings support the observation that phosphorylation at S329 directly inhibits p50 DNA binding and activity.

Phosphorylation of p50 mediates SN1-methylator-induced cytotoxicity

Given that S329 phosphorylation is required for inhibition of NF-κB DNA binding, the role of S329 in p50 promoter recruitment was next examined. GFP-tagged p50wt, p50S329A or p50S329D was stably expressed in U87 cells and BclXL promoter-enrichment assessed by ChIP. Although p50wt and p50S329A are recruited to the promoter to a similar extent, p50S329A is resistant to inhibition by TMZ (Figure 6A). In addition, p50S329D binds the promoter substantially less than p50wt and is not affected by TMZ. These findings suggest that phosphorylation of S329 blocks the basal recruitment of p50 to the BclXL gene promoter. Of note, in GFP-p50 transfectants, GFP pull-down yields similar findings to those seen with anti-p50 IP (data not shown).

To link S329 to the downstream response, stable transfectants expressing p50wt or p50S329A were constructed using p50−/− MEFs. Initially, anti-apoptotic gene expression was examined. p50wt, but not p50S329A, restores the ability of TMZ to inhibit BclXL and COX2 expression suggesting that phosphorylation of S329 is important for inhibition of anti-apoptotic genes by TMZ. Next, cell viability was studied. Expression of p50wt, but not p50S329A, reconstitutes the sensitivity of p50−/− MEFs to TMZ both on colony formation (Figure 6C) and short-term viability assay (Figure S6A). These results, in conjunction with the finding that p50S329A does not alter surviving fraction following IR (Figure S6B), suggest that p50 S329 mediates a pathway-specific cytotoxic response. As an additional control, p50 was stably knocked-down in an immortal human fibroblast cell line, MSU1.1. Re-expression of p50wt, but not p50S329A, re-constitutes killing by TMZ in these cells (Figure S6C). BG sensitizes only the p50wt expressing cells to TMZ, a finding that further supports the observation that p50 mediates SN1-methylator-induced killing by an MGMT-independent mechanism.

The p50-mediated response to O6meG:T sensitizes cells to DNA strand breaks

An important question raised by our data is whether a p50-dependent pathway is sufficient to induce the cytotoxic response to O6-MeG. As O6-MeG:T substrate can be introduced into intact cells and induces p50-dependent signaling (Figure 5), we examined whether this duplex can elicit the downstream response as assessed by clonogenic assay. O6-MeG:T substrate, used at a concentration sufficient to induce p50-mediated signaling, does not decrease clonogenic survival relative to control in glioma cells (Figure 7A) or MEFs (Figure 7C). In addition, O6-MeG:T substrate does not induce cell cycle arrest (Figure S6D). These findings suggest that O6-MeG-induced signaling, involving p50-mediated regulation of NF-κB, is not sufficient to induce the downstream damage response. While it is possible that the lack of effect is due to limited oligo uptake and insufficient signaling, this is unlikely as the magnitude of inhibition of NF-κB by O6-MeG:T is equivalent to the magnitude of inhibition induced by cytotoxic concentrations of TMZ. As apoptosis induced by SN1-methylating agents is thought to be the result of MMR-induced DSBs (Kaina et al., 2007), a more feasible explanation is that the O6-MeG:T substrate is not converted into the type of lesion necessary for killing. This hypothesis is substantiated by the fact that similar substrates are not extensively modified (Yoshioka et al., 2006) and that O6-MeG:T substrate does not increase the amount of γH2AX foci compared to control (data not shown).

Figure 7. O6-MeG:T mismatches sensitize to DNA damage-induced cell death.

Figure 7

Open in a new tab

Colony formation assays. Cells were un-transfected (UT) or transfected with 20 ng duplex substrate (16 hrs) and then un-irradiated (−IR) or irradiated (+IR). Data for the −IR cells are normalized to −IR UT and data for +IR are normalized to +IR UT. 0.5 Gy causes no change in surviving fraction and 2 Gy IR causes 25–30 % decrease relative to −IR. Data show mean value ± SD of duplicate samples, repeated at least twice with similar results.

(A) U87 cells, treated with 0.5 Gy. *p < 0.0005

(B) U87 cells, treated with 2 Gy. *p < 0.02.

(C) Wt and p50−/− MEFs pretreated with 20 μM BG and then treated as in (B). *p < 0.002.

(D) U87 cells treated with 0.5 Gy IR and either un-transfected (UT) or transfected with the indicated duplex substrate. Treatment was administered in the order indicated. *p < 0.002 relative to G:C. See also Figure S6.

The above findings, when considered with the observation that p50 is at least necessary for efficient O6-MeG-induced killing, suggest that O6-MeG:T-induced, p50-dependent signaling may act to sensitize cells to cytotoxicity that is induced by lesions such as DSBs. To examine this hypothesis, duplex substrates were introduced into cells and subsequently DNA strand breaks induced with IR. Whether IR is given at a dose that causes no loss of viability on clonogenic assay (0.5 Gy), or at a higher, cytotoxic, dose (2 Gy), pretreatment with O6-MeG:T substrate reduces the surviving fraction relative to control (Figure 7A and B). Moreover, the decrease in surviving fraction is noted only in p50-proficient cells (Figure 7C). Similar findings are seen with induction of apoptosis in wt and p50-depleted cells (data not shown).

While these results suggest that p50-mediated signaling sensitizes cells to DNA damage-induced killing, an alternative interpretation would be that O6-MeG:T lesions cause a level of damage that is sub-lethal but when combined with IR results in additive cytotoxicity. In such a scenario, the temporal relationship between O6-MeG:T-mediated signaling and IR would be predicted to be irrelevant. However, the reduction in surviving fraction is only seen when O6-MeG:T substrate is administered before IR and not after (Figure 7D). Lack of an inducing effect by O6-MeG:T substrate when given after IR, or in p50-depleted cells, suggests that the sensitizing action of these oligos is not due to differential cellular uptake compared to controls. Based on these findings, we propose a model in which O6-MeG:T mismatches initiate an MMR/ATR/Chk1-dependent signaling pathway that targets p50 to regulate NF-κB-dependent transcription that sensitizes cells to DNA strand break-induced cytotoxicity.

DISCUSSION

The data presented suggest that the p50 subunit of NF-κB acts downstream of Chk1 in the cytotoxic DNA damage response. p50 facilitates cell death directly and not by an indirect effect on mediators of O6-MeG-induced killing. In such a capacity, p50 can be considered to be an effector protein that enables the upstream damage signal to regulate changes in downstream gene expression. As a central subunit of a highly conserved transcription factor, p50 is ideally situated to mediate damage signaling throughout the cell. Although the role of effector protein is most often ascribed to p53, SN1-methylation-induced cytotoxicity is known to involve p53-indpendent mechanisms (Kaina et al., 2007). In addition, whereas p53 mediates both damage-induced cytotoxicity and cell cycle arrest, p50 appears to be involved only in cytotoxicity. This finding suggests that there is a split in the methylator-induced response downstream of Chk1 such that p50 specifically mediates a cytotoxic pathway. Interestingly, a separation of methylator-induced killing from checkpoint response has also been described for the p53 ortholog, p73 (Li et al., 2008).

While an MMR-directed process is necessary for the response to O6-MeG, the mechanism by which cytotoxicity is induced remains unresolved. In the ‘futile cycle’ theory it is proposed that as MMR targets the newly synthesized DNA strand, methylated adducts are left in place setting up futile rounds of damage repair that lead to DNA nicks and subsequently DSBs (Karran et al., 1993). The ‘signaling’ hypothesis posits that no processing is required and that the MMR-machinery directly signals the cytotoxic response (Fishel, 2001). The observations that damage signaling is directly activated by O6-MeG:T lesions but not by undamaged G:T mismatches (Yoshioka et al., 2006) and that point mutations of MMR subunits lead to a separation of repair and damage signaling (Yang et al., 2004) support the importance of a signaling pathway to the damage response. However, there has been no verification that O6-MeG-induced signaling is in fact able to activate the downstream damage response. Furthermore, it was noted that the damage response is abrogated when O6-MeG lesions are removed after the time when signaling would be expected to have occurred (Mojas et al., 2007). Although this latter finding supports the contention that signaling is not sufficient to induce the damage response, the question of whether a signaling pathway is necessary for MMR-induced cytotoxicity remains unanswered.

Identification of p50 as a target downstream of O6-MeG has enabled us to demonstrate that a signaling pathway although not sufficient, is nevertheless necessary for efficient MMR-dependent killing. We propose a model whereby p50-dependent effects lead to inhibition of cell survival gene expression resulting in a decrease in the cellular cytotoxic threshold. We hypothesize that this pathway ‘primes’ the cell so that a lethal event can be induced by less DNA damage than is necessary without p50-dependent signaling. Such a model takes into account the well-described requirement of DSBs for MMR-dependent cytotoxicity. Furthermore, this model is temporally consistent with the overall response to SN1-methylation in that following the formation of O6-MeG lesions, MMR-induced signaling is induced after the first S-phase while DSBs are formed after the second S-phase (Kaina et al., 2007; Mojas et al., 2007; Stojic et al., 2004). If a cell either repairs the induced DNA damage or blocks p50-mediated signaling then apoptosis is attenuated. The finding that a p50-dependent effect results in inhibition of BclXL is also consistent with the prior observation that the mitochondrial pathway is necessary for O6-MeG-induced apoptosis (Hickman and Samson, 2004).

The data presented show that SN1-methylators act via p50 to inhibit NF-κB activity. However, the NF-κB dimer present in the nucleus at rest is comprised of both p50 and p65 and we previously noted that SN1-methylators inhibit p65 DNA binding (Yamini et al., 2007). These findings, when considered with the observation that TMZ inhibits p65 promoter-recruitment in a p50-dependent manner (Figure 2D), suggest that inhibition of the p50/p65 dimer is involved in mediating TMZ-induced cytotoxicity. However, p65 is dispensable for both inhibition of NF-κB and induction of apoptosis (Figures 1A and 3A) indicating that it is not inhibition of p50/p65 per se that is necessary for TMZ-induced cytotoxicity but p50-containing dimers in general. In addition, the observation that TMZ does not block p65 promoter-recruitment in p50-depleted cells suggests that SN1-methylators do not act on p52 to inhibit NF-κB as in these cells p52 cross-compensates for p50 to form p52/p65 dimers (Figure S2C) (Hoffmann et al., 2003).

Although NF-κB subunit modification is a common mechanism for regulation of NF-κB activity, p50 specific changes have been relatively under-examined (Perkins, 2006). While phosphorylation of p50 is reported to increase its DNA binding (Hou et al., 2003), p50 phosphorylation has not been shown to inhibit NF-κB activity. p50 S329 is not a residue that makes direct DNA contact suggesting that inhibition of DNA binding following S329 phosphorylation may be the result of a conformational change in p50 or the NF-κB dimer. Inhibition of NF-κB by SN1-methylators is somewhat surprising given the propensity of DNA-damage to activate NF-κB (Wu and Miyamoto, 2007). Nevertheless, cisplatin has been reported to inhibit p65 activity by inducing Chk1-mediated phosphorylation of threonine 505 (Campbell et al., 2006), while UV light and daunorubicin convert p65 into an active repressor of gene expression (Campbell et al., 2004). Although these agents attenuate NF-κB activity without affecting DNA binding, doxorubicin has been reported to reduce p65 promoter binding (Ho et al., 2005). Interestingly, a recent report examining the effect of replication stress on NF-κB noted that ATR, induced by agents such as hydroxyurea, suppresses NF-κB as assessed by EMSA (Wu and Miyamoto, 2008). While the authors demonstrate that the inhibition of NF-κB by ATR involves the interaction of ATR with NEMO (inhibitor-κB kinase γ), it is possible that an ATR/Chk1/p50 mediated effect also contributes to their findings.

This work identifies p50 as an intermediate in the cytotoxic response to DNA damage. From a practical standpoint, these data suggest that patients with decreased p50 expression may be relatively resistant to chemotherapeutics like TMZ or 6-TG. Although complete loss of p50 is rarely seen clinically, reduced p50 expression is not uncommon and has been associated with an NFKB1 promoter polymorphism (Karban et al., 2004). These observations suggest that p50 expression may be an important factor in the pharmacogenomic stratification of patients to certain chemotherapeutic agents.

EXPERIMENTAL PROCEDURES

Cells and reagents

U87 cells were cultured as described (Yamini et al., 2007). Primary wt and p50−/− MEFs were harvested at embryonic day 13.5, and primary p65−/− MEFs were provided by Dr. A Lin (University of Chicago). 3T3 immortalized wt and p50−/− MEFs were obtained from Dr. Y-X Fu (University of Chicago). Tetracycline-inducible, U2OS cells expressing ATR-wt and ATR–kd were obtained from Dr. P Nghiem (University of Washington) (Nghiem et al., 2002) and cultured in 1 μg/ml doxycycline for 48 hours prior to use. ATM+/+ and ATM−/− cells were purchased from Coriell Cell Repositories. TMZ was obtained from the Drug Synthesis and Chemistry Branch, National Cancer Institute, NIH.

Antibodies and plasmids

The polyclonal phospho-specific p50 S329 antibody was raised in rabbit following immunization with the phospho-peptide: CPKYKDINITKPA(pSer)V (GenScript corp.). Other antibodies are described in the supplemental data. Ig-κB luc has been described (Yamini et al., 2007). BclXL and ΔκB BclXL reporters were provided by Professor R Hay (University of Dundee). pCMV-p50 was provided by Dr. R Ricciardi (University of Pennsylvania). p50S329A and p50S329D were constructed by site-directed mutagenesis using QuikChange II XL (Strategene), based on pCMV-p50. GFP-p50 constructs were made by cloning cDNA for the various p50 mutants into the KpnI site of pEGFP-C1 (Clontech). For stable expression, p50 was subcloned into pcDNA3.1+ Hygro. Stable MEF transfectants were made using p50−/− 3T3 MEFs (3T3 MEFs have the same response as primary MEFs with respect to inhibition of NF-κB).

Oligonucleotide substrates

Two 40 bp biotin end-labeled, PAGE-purified DNA oligos bearing a unique central base, G or O6-MeG, were acquired (Sigma). The complementary strands bearing the central bases T or C were also obtained. The oligos were annealed to form 4 duplex substrates with the central base pairs: O6-MeG:T, O6-MeG:C, G:T and G:C.

Bacterial p50 expression

The RHDs (aa 1–366) of human p50wt, p50S329A and p50S329D were subcloned into pET45b+ (Novagen) containing a 5′ poly-histidine (His) tag. His-tagged p50 was isolated from BL21 E. coli using Qiagen Ni-NTA spin columns. Protein content was quantified and purity confirmed by PAGE and coomasie staining. Binding of purified proteins to κB-DNA was assessed by EMSA.

Luciferase, ChIP, EMSA and Kinase assays

See supplemental data for oligo and primer details. Luciferase assay transfection was normalized using Renilla reniformis and experiments repeated thrice. For luciferase assays with oligos, duplex substrates were introduced 24 hours after Ig-κB/renilla using Oligofectamine (Invitrogen). For quantitative ChIP, IP was performed with anti-p65 (Millipore, #06-418), anti-p50, anti-GFP, anti-RNA polymerase II (Millipore, # 05-623B), anti-Histone H1 and normal IgG. qPCR was performed with human BclXL promoter specific primers. The change in Ct for the BclXL promoter amplification was determined for p65 (or p50) IP relative to the input DNA and similar calculations were made for the anti-RNA Pol (or Histone H1) IP and IgG IP samples. Next, the anti-p65 (p50) or anti-RNA Pol (Histone H1) values were subtracted from the anti-IgG value to control for non-specific binding.

EMSA was performed using the Ig-κB probe. For p50 replacement EMSA, 5 μg of sh-p50 nuclear extract was incubated with 15 ng p50wt or p50S329A (determined by immunoblot to be similar to the level of p50 in sh-c cells) for 30 min on ice prior to treatment. For Chk1 IP kinase assay, 293T cells were transfected with FLAG-Chk1 or FLAG-kd-Chk1 (D130A) (Dr. Jiri Lukas, Center for Genotoxic Stress Research, Denmark). Also, 50 ng active recombinant Chk1 (Active Motif) was used where indicated. For kinase assay with duplex substrates, 5 μg of sh-p50 nuclear extract was incubated with 15 ng of p50wt or p50S329A and 20 nM duplex substrate under kinase conditions. EMSA and kinase assays are representative of three experiments.

Statistical analyses

Statistical significance was taken as p < 0.05 using a 2-tailed Student’s t-test.

Supplementary Material

01

HIGHLIGHTS.

p50 is phosphorylated by Chk1 in response to SN1-methylation/O6-MeG:T mismatches. Phosphorylation of p50 blocks NF-κB DNA binding and transcriptional activity. Phospho-p50 mediates inhibition of NF-κB-regulated anti-apoptotic gene expression. p50 pathway is necessary for efficient O6-methylguanine-induced cytotoxicity.

Acknowledgments

We are grateful to N.D. Perkins, R.T. Hay, R. Ricciardi, A. Lin, H Piwnica-Worms and J. Lukas for plasmid constructs; to P. Nghiem, J. McCormick, G. Marra, Y-X. Fu and A. Lin for cell lines; to M. Abe, B. Strauss and members of the Weichselbaum lab; to Shannon Delaney for the MGMT assay and M. Eileen Dolan for intellectual contributions. This work was supported by Novartis Oncology provided through the CALGB Foundation and NIH 1R01CA136937 (BY), NIH R01CA111423 (RRW), a gift from the Foglia Family, the Center for Radiation Therapy and the Ludwig Center for Metastasis Research.

Footnotes

The authors declare no conflicts of interest.

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Birbach A, Gold P, Binder BR, Hofer E, de Martin R, Schmid JA. Signaling molecules of the NF-kappa B pathway shuttle constitutively between cytoplasm and nucleus. J Biol Chem. 2002;277:10842–10851. doi: 10.1074/jbc.M112475200. [DOI] [PubMed] [Google Scholar]
  2. Campbell KJ, Rocha S, Perkins ND. Active repression of antiapoptotic gene expression by RelA(p65) NF-kappa B. Mol Cell. 2004;13:853–865. doi: 10.1016/s1097-2765(04)00131-5. [DOI] [PubMed] [Google Scholar]
  3. Campbell KJ, Witty JM, Rocha S, Perkins ND. Cisplatin mimics ARF tumor suppressor regulation of RelA (p65) nuclear factor-kappaB transactivation. Cancer Res. 2006;66:929–935. doi: 10.1158/0008-5472.CAN-05-2234. [DOI] [PubMed] [Google Scholar]
  4. Carlotti F, Dower SK, Qwarnstrom EE. Dynamic shuttling of nuclear factor kappa B between the nucleus and cytoplasm as a consequence of inhibitor dissociation. J Biol Chem. 2000;275:41028–41034. doi: 10.1074/jbc.M006179200. [DOI] [PubMed] [Google Scholar]
  5. Dolan ME, Moschel RC, Pegg AE. Depletion of mammalian O6-alkylguanine-DNA alkyltransferase activity by O6-benzylguanine provides a means to evaluate the role of this protein in protection against carcinogenic and therapeutic alkylating agents. Proc Natl Acad Sci U S A. 1990;87:5368–5372. doi: 10.1073/pnas.87.14.5368. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Fishel R. The selection for mismatch repair defects in hereditary nonpolyposis colorectal cancer: revising the mutator hypothesis. Cancer Res. 2001;61:7369–7374. [PubMed] [Google Scholar]
  7. Harper JW, Elledge SJ. The DNA damage response: ten years after. Mol Cell. 2007;28:739–745. doi: 10.1016/j.molcel.2007.11.015. [DOI] [PubMed] [Google Scholar]
  8. Hickman MJ, Samson LD. Apoptotic signaling in response to a single type of DNA lesion, O(6)-methylguanine. Mol Cell. 2004;14:105–116. doi: 10.1016/s1097-2765(04)00162-5. [DOI] [PubMed] [Google Scholar]
  9. Ho WC, Dickson KM, Barker PA. Nuclear factor-kappaB induced by doxorubicin is deficient in phosphorylation and acetylation and represses nuclear factor-kappaB-dependent transcription in cancer cells. Cancer Res. 2005;65:4273–4281. doi: 10.1158/0008-5472.CAN-04-3494. [DOI] [PubMed] [Google Scholar]
  10. Hoffmann A, Leung TH, Baltimore D. Genetic analysis of NF-kappaB/Rel transcription factors defines functional specificities. Embo J. 2003;22:5530–5539. doi: 10.1093/emboj/cdg534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hou S, Guan H, Ricciardi RP. Phosphorylation of serine 337 of NF-kappaB p50 is critical for DNA binding. J Biol Chem. 2003;278:45994–45998. doi: 10.1074/jbc.M307971200. [DOI] [PubMed] [Google Scholar]
  12. Hutchins JR, Hughes M, Clarke PR. Substrate specificity determinants of the checkpoint protein kinase Chk1. FEBS Lett. 2000;466:91–95. doi: 10.1016/s0014-5793(99)01763-9. [DOI] [PubMed] [Google Scholar]
  13. Kaina B, Christmann M, Naumann S, Roos WP. MGMT: key node in the battle against genotoxicity, carcinogenicity and apoptosis induced by alkylating agents. DNA Repair (Amst) 2007;6:1079–1099. doi: 10.1016/j.dnarep.2007.03.008. [DOI] [PubMed] [Google Scholar]
  14. Karban AS, Okazaki T, Panhuysen CI, Gallegos T, Potter JJ, Bailey-Wilson JE, Silverberg MS, Duerr RH, Cho JH, Gregersen PK, et al. Functional annotation of a novel NFKB1 promoter polymorphism that increases risk for ulcerative colitis. Hum Mol Genet. 2004;13:35–45. doi: 10.1093/hmg/ddh008. [DOI] [PubMed] [Google Scholar]
  15. Karran P, Macpherson P, Ceccotti S, Dogliotti E, Griffin S, Bignami M. O6-methylguanine residues elicit DNA repair synthesis by human cell extracts. J Biol Chem. 1993;268:15878–15886. [PubMed] [Google Scholar]
  16. Krokan H, Bjorklid E, Prydz H. DNA synthesis in isolated HeLa cell nuclei. Optimalization of the system and characterization of the product. Biochemistry. 1975;14:4227–4232. doi: 10.1021/bi00690a012. [DOI] [PubMed] [Google Scholar]
  17. Li LS, Morales JC, Hwang A, Wagner MW, Boothman DA. DNA mismatch repair-dependent activation of c-Abl/p73alpha/GADD45alpha-mediated apoptosis. J Biol Chem. 2008;283:21394–21403. doi: 10.1074/jbc.M709954200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Mojas N, Lopes M, Jiricny J. Mismatch repair-dependent processing of methylation damage gives rise to persistent single-stranded gaps in newly replicated DNA. Genes Dev. 2007;21:3342–3355. doi: 10.1101/gad.455407. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Nghiem P, Park PK, Kim YS, Desai BN, Schreiber SL. ATR is not required for p53 activation but synergizes with p53 in the replication checkpoint. J Biol Chem. 2002;277:4428–4434. doi: 10.1074/jbc.M106113200. [DOI] [PubMed] [Google Scholar]
  20. Perkins ND. Post-translational modifications regulating the activity and function of the nuclear factor kappa B pathway. Oncogene. 2006;25:6717–6730. doi: 10.1038/sj.onc.1209937. [DOI] [PubMed] [Google Scholar]
  21. Stojic L, Mojas N, Cejka P, Di Pietro M, Ferrari S, Marra G, Jiricny J. Mismatch repair-dependent G2 checkpoint induced by low doses of SN1 type methylating agents requires the ATR kinase. Genes Dev. 2004;18:1331–1344. doi: 10.1101/gad.294404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Swann PF, Waters TR, Moulton DC, Xu YZ, Zheng Q, Edwards M, Mace R. Role of postreplicative DNA mismatch repair in the cytotoxic action of thioguanine. Science. 1996;273:1109–1111. doi: 10.1126/science.273.5278.1109. [DOI] [PubMed] [Google Scholar]
  23. Vallabhapurapu S, Karin M. Regulation and function of NF-kappaB transcription factors in the immune system. Annu Rev Immunol. 2009;27:693–733. doi: 10.1146/annurev.immunol.021908.132641. [DOI] [PubMed] [Google Scholar]
  24. Wu ZH, Miyamoto S. Many faces of NF-kappaB signaling induced by genotoxic stress. J Mol Med. 2007;85:1187–1202. doi: 10.1007/s00109-007-0227-9. [DOI] [PubMed] [Google Scholar]
  25. Wu ZH, Miyamoto S. Induction of a pro-apoptotic ATM-NF-kappaB pathway and its repression by ATR in response to replication stress. Embo J. 2008;27:1963–1973. doi: 10.1038/emboj.2008.127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Yamini B, Yu X, Dolan ME, Wu MH, Darga TE, Kufe DW, Weichselbaum RR. Inhibition of nuclear factor-kappaB activity by temozolomide involves O6-methylguanine induced inhibition of p65 DNA binding. Cancer Res. 2007;67:6889–6898. doi: 10.1158/0008-5472.CAN-06-4496. [DOI] [PubMed] [Google Scholar]
  27. Yang G, Scherer SJ, Shell SS, Yang K, Kim M, Lipkin M, Kucherlapati R, Kolodner RD, Edelmann W. Dominant effects of an Msh6 missense mutation on DNA repair and cancer susceptibility. Cancer Cell. 2004;6:139–150. doi: 10.1016/j.ccr.2004.06.024. [DOI] [PubMed] [Google Scholar]
  28. Yoshioka K, Yoshioka Y, Hsieh P. ATR kinase activation mediated by MutSalpha and MutLalpha in response to cytotoxic O6-methylguanine adducts. Mol Cell. 2006;22:501–510. doi: 10.1016/j.molcel.2006.04.023. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01
NIHMS338292-supplement-01.doc (6.3MB, doc)

RESOURCES