The Deubiquitinating Enzyme USP11 Controls an IκB Kinase α (IKKα)-p53 Signaling Pathway in Response to Tumor Necrosis Factor α (TNFα)*

Post-translational modification and degradation of proteins by the ubiquitin-proteasome system are key regulatory events in cellular responses to various stimuli. The NF-κB signaling pathway is controlled by the ubiquitin-mediated proteolysis. Although mechanisms of ubiquitination in the NF-κB pathway have been extensively studied, deubiquitination-mediated regulation of the NF-κB signaling remains poorly understood. The present studies show that a deubiquitinating enzyme, USP11, specifically regulates IκB kinase α (IKKα) among the NF-κB signaling molecules. Knocking down USP11 attenuates expression of IKKα in the transcriptional, but not the post-translational, level. However, down-regulation of USP11 dramatically enhances NF-κB activity in response to tumor necrosis factor-α, indicating that IKKα does not require activation of NF-κB. Instead, knock down of USP11 or IKKα is associated with abrogation of p53 expression upon exposure to tumor necrosis factor-α. In concert with these results, silencing of USP11 is associated with transcriptional attenuation of the p53-responsive genes, such as p21 or Bax. Importantly, the ectopic expression of IKKα into cells silenced for USP11 restores p53 expression, demonstrating that USP11 functions as an upstream regulator of an IKKα-p53 signaling pathway.

These post-translational modifications are critical for "on-off" switch in the signaling cascade. For example, one member of IB, designated as IB␣, is phosphorylated in response to TNF␣ or interleukin-1 and is thereby subjected to ubiquitination and degradation by the 26 S proteasome (5). The NF-B essential modifier (NEMO, also known as IKK␥) is a key to activating IB kinase (IKK) complex and is polyubiquitinated by TRAF6 (6). In contrast, deubiquitinating mechanisms that target to ubiquitinated NF-B signaling molecules are largely unclear. Certain insights have been derived from the finding that CYLD was identified by its association with NEMO and through systematic screening for deubiquitinating enzymes that impede NF-B signaling (7)(8)(9). Overexpression of CYLD represses NF-B activation in response to various stimuli, including TNF␣. Conversely, inactivation of CYLD by RNA interference increases inducible NF-B activity. These findings thus suggest that CYLD acts as a negative regulator of the NF-B signaling pathway. Little is otherwise known about direct deubiquitinating enzymes targeting to ubiquitinating molecules that are involved in NF-B signaling.
A recent study mapped a protein interaction network for TNF␣/NF-B pathway components by means of an integrated approach, including tandem affinity purification, liquid chromatography tandem mass spectrometry, network analysis, and directed functional perturbation studies using RNA interference (10). This study identified numerous previously unknown interactors, including a deubiquitinating enzyme, USP11. However, the role for USP11 in the NF-B signaling pathway remains obscure.
Our recent studies demonstrated that IKK␣, but not IKK␤, is activated and translocates into the nucleus in response to oxidative stress (11). Upon exposure to oxidative stress, IKK␣ activation does not contribute to NF-B activation; instead, nuclear IKK␣ regulates the transcription activity of the p53 tumor suppressor, indicating that the IKK␣ 3 p53 signaling pathway is associated with the cellular response to oxidative stress. In this study, we show that USP11 regulates IKK␣ expression in a transcriptional level. Importantly, down-regulation of IKK␣ does not contribute to NF-B activation in response to TNF␣. Instead, USP11-mediated regulation of IKK␣ functions to control p53 transcription activity. These findings support a novel mechanism in which the deubiquitinating enzyme USP11 contributes to TNF␣-induced IKK␣ 3 p53 signaling pathway.
Plasmids-Constructs of IKK␣ and IKK␤ are described elsewhere (12). IKK␥ cDNA was amplified by PCR from a human fetal brain cDNA library and cloned into the pcDNA3-FLAG vector as described elsewhere (13). USP11 cDNA (14) was cloned into the pcDNA3-FLAG vector. An RNA interferenceresistant form of USP11 was constructed by introducing silent mutations using PCR-based site-directed mutagenesis. The sequences of oligonucleotide primers are as follows: 5Ј-CAT-ACCGATTCaATaGGgCTAGTATTGCGC-3Ј and 5Ј-GCG-CAATACTAGcCCtATtGAATCGGTATG-3Ј. Lowercase letters represent silent mutations. Mutations were confirmed by sequencing. HA-ubiquitin plasmid has been reported previously (15). Cell transfection was performed as described (16,17). The total DNA concentration was kept constant by inclusion of an empty vector.
RT-PCR Analysis for Gene Expression-Total cellular RNA was extracted using the RNeasy kit (Qiagen). First-stand cDNA synthesis and the following PCR reactions were performed with 500 ng of total RNA using SuperScript one-step RT-PCR system (Invitrogen) according to the manufacturer's protocol. The reaction products were resolved on a 2% agarose gel.
Reporter Gene Assays-293 cells stably transfected with pNF-B-luc and pTK-hyg (Panomics) were transfected with a variety of siRNAs followed by the treatment with TNF␣. The luciferase activity was determined with a Bright-Glo luciferase assay system (Promega) according to the manufacturer's protocol.
Chromatin Immunoprecipitation Assays-Cells were harvested and washed with chilled phosphate-buffered saline once followed by incubation in 1% formaldehyde for 15 min at room temperature for chromatin cross-linking. The cells were then collected and washed with chilled phosphate-buffered saline again. After centrifugation, the cell pellets were resuspended in SDS lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl, pH 8.0, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 mM Na 3 VO 4, 10 g/ml aprotinin, 1 g/ml leupeptin, and 1 g/ml pepstatin A), and the lysates were sonicated to obtain DNA fragments 200 -500 bp in length. After centrifugation, 50 l of the supernatant was used as an input, and the remainder was diluted 2-2.5-fold in washing buffer (50 mM Tris-HCl, pH 7.6, 150 mM NaCl, 0.1% Nonidet P-40, and protease inhibitors as described above). This diluted fraction was subjected to immunoprecipitation with 2 g of indicated antibodies for 2 h to overnight at 4°C with rotation. The immunocomplexes were collected with 30 l of protein A-Sepharose beads (Santa Cruz Biotechnology) for 1-2 h at 4°C with rotation. The beads were then pelleted by centrifugation and washed sequentially with 300 l of the following buffers: wash buffer I (500 mM NaCl, 0.1% SDS, 2 mM EDTA, and 20 mM Tris-HCl, pH 8.0), wash buffer II (250 mM LiCl, 1 mM EDTA, 10 mM Tris-HCl, pH 8.0, and 1% deoxycholate), and then twice with Tris-EDTA buffer. Precipitated chromatin complexes were removed from the beads by shaking with 150 l of elution buffer (1% SDS and 0.1 M NaHCO 3 ) for 15 min, and this step was repeated. All the eluates were collected, and then the cross-linking was reversed by adding NaCl to a final concentration of 200 mM. The mixture was allowed to stand overnight at 65°C. The remaining proteins were digested with the extraction buffer (50 mM Tris-HCl, pH 6.8, 10 mM EDTA, and 40 g/ml proteinase K) for 1 h at 45°C. DNA was recovered by phenol/chloroform/isoamyl alcohol (25/24/1) extraction and precipitated with 0.1 volume of 3 M sodium acetate and 2.5 volumes of ethanol. PCR amplification was performed in chromatin immunoprecipitated fragments using oligonucleotide pairs as described elsewhere (20).
Subcellular Fractionation-Subcellular fractionation was performed as described previously (21)(22)(23). Purity of the fractions was monitored by immunoblot analysis with anti-PCNA and anti-tubulin.

USP11 Is Associated with IKK␣ Expression-Previous studies
have implicated a number of deubiquitinating enzymes in several signaling cascades, including the NF-B pathway. In particular, a recent study suggested a potential involvement of USP11 in NF-B signaling. To determine whether USP11 affects the expression of NF-B signalosome, U2OS cells were transfected with a scramble siRNA or a siRNA that targets USP11. We examined the expression of seven gene products that are directly involved in the NF-B signaling pathway (Fig.  1A). Among these products, the expression of IKK␣ was significantly reduced in cells silenced for USP11. To exclude the possibility that knocking down USP11 is associated with off-target effects, siRNAs that target different USP11 sequences (USP11 siRNA-a and siRNA-b) were transfected into U2OS cells. Transfection of USP11 siRNA-a or siRNA-b resulted in similar levels of attenuated IKK␣ expression, indicating that USP11 specifically down-regulates IKK␣ expression (see Fig. 4A). To confirm the regulation of IKK␣ expression by USP11, U2OS cells were transfected with the USP11 siRNA followed by the transfection of the FLAG-USP11 mutant that is resistant for the USP11 siRNA by introducing silent mutations. The results demonstrated that IKK␣ expression was restored by forced expression of USP11 into U2OS cells silenced for USP11 (Fig.  1B). To investigate whether suppression of IKK␣ by USP11 silencing is regulated by the ubiquitin-proteasome pathway, FLAG-tagged IKK families were transfected into U2OS cells together with HA-tagged ubiquitin. Previous studies demonstrated that IKK␥ is degraded by the ubiquitin-proteasome system (4). In concert with these results, overexpression of IKK␥ was sufficient for substantial ubiquitination, regardless of TNF␣ treatment (Fig. 1C). By contrast, there was little if any ubiquitination on IKK␣ or IKK␤, indicating that IKK␣ and IKK␤ function independently of ubiquitination-mediated regulation. These results thus suggest that USP11 regulates IKK␣ by a ubiquitin-independent manner. Another possibility is that regulation of IKK␣ by USP11 is mediated by undefined molecules that are controlled by the ubiquitin-deubiquitin mechanism. To further define USP11-mediated regulation of NF-B signalosome, total RNA from U2OS cells in the presence or absence of USP11 was subjected to RT-PCR analysis. Knock down of USP11 specifically attenuated IKK␣ expression, suggesting that regulation of IKK␣ by USP11 occurs at the transcriptional level (Fig. 2A). In this context, a previous study showed that IKK␣ is positively regulated by a transcription factor, Ets-1 (20). Thus, it is plausible that USP11 affects Ets-1 by the ubiquitin-deubiquitin mechanism to control IKK␣ expression. To examine this possibility, U2OS cells were transfected with the scramble siRNA or USP11 siRNA followed by treatment with or without TNF␣. The expression level of Ets-1 remained unchanged even in USP11-silencd cells, clearly indicating that Ets-1 is not targeted by USP11, at least post-translational level (Fig. 2B).
To further define the USP11-independent regulation of Ets-1, we performed chromatin immunoprecipitation assays with anti-Ets-1 antibody. As reported previously (20), Ets-1 occupied IKK␣ promoter, and this occupancy was increased after TNF␣ stimulation in U2OS cells (Fig. 2C). Importantly, the finding that Ets-1 occupancy on the IKK␣ promoter in cells silenced for USP11 was comparable with that in control cells indicates that Ets-1 transcriptional activity is regulated in an USP11-independent matter (Fig. 2C). Similar results were obtained in MCF-7 cells (data not shown). Taken together,

FIGURE 2. USP11 regulates IKK␣ expression at the transcriptional level.
A, U2OS cells transfected with scramble siRNA or USP11 siRNA were treated with TNF␣. Total RNA was analyzed by semiquantitative RT-PCR using specific primers targeting as indicated. B, U2OS cells were transfected as described above and left untreated (C) or treated with TNF␣ for 1 h (T). Cell lysates were analyzed by immunoblotting (IB) with the indicated antibodies. C, U2OS cells were transfected as described above and left untreated or treated with TNF␣ for 8 h. Chromatin immunoprecipitation assays were performed by using anti-Ets-1. Immunoprecipitated chromatin was analyzed by PCR using primer sequences containing IKK␣ promoter region. these findings support an essential role for USP11 in the regulation of IKK␣ expression at a transcriptional level.
Knocking Down USP11 Enhances TNF␣-induced NF-B Activation-To assess the effects of USP11 on the activity of NF-B, 293 cells stably transfected with the luciferase-reporter vector containing NF-B response elements were silenced for USP11 or RelA/p65. As shown previously, silencing of RelA/ p65 was associated with pronounced inhibition of NF-B activity in response to TNF␣ (Fig. 3A). By contrast, knocking down USP11 enhanced TNF␣-induced NF-B activation (Fig. 3B). To further define the involvement of USP11 in NF-B activation, U2OS cells were transfected with the scramble siRNA or USP11 siRNA followed by TNF␣ treatment. Immunoblot analysis of whole cell lysates demonstrated substantial attenuation of resynthesized IB␣ in the absence of USP11 (Fig. 3B). In accordance with this result, immunoblotting of nuclear lysates with anti-RelA/p65 revealed that at 2 h after TNF␣ treatment, 73% of nuclear RelA/p65 was exported from the nucleus in cells transfected with the scramble siRNA (Fig. 3B). In contrast, 83% of RelA/p65 remained in the nucleus in cells silenced for USP11 at the same time point (Fig. 3B), indicating that nuclear export of RelA/p65 was abrogated in cells silenced for USP11 (Fig. 3B). These findings provide a potential model in which, in USP11silencing cells, RelA/p65 remains in the nucleus because of attenuation of resynthesized IB␣, resulting in sustained activation of NF-B. In addition, given the data from reporter assays that imply enhancement of NF-B activity in the absence of USP11, the confirmed results that IKK␣ expression is downregulated by silencing USP11 further suggested that NF-B activation occurs independently of IKK␣ under our experimental conditions (Fig. 3B). Taken together, these findings indicate that USP11 negatively regulates NF-B activity in response to TNF␣ by a yet undefined mechanism.
USP11 Controls the IKK␣ 3 p53 Signaling in Response to TNF␣-Our recent study showed that the protein kinase C ␦ 3 IKK␣ signaling pathway stabilizes and activates p53 in response to oxidative stress (11). These finding led us to examine whether USP11 controls p53 through IKK␣ upon exposure to TNF␣. As shown previously, knocking down USP11 in U2OS cells was associated with attenuation of IKK␣ expression (Fig.  4A). More importantly, steady-state levels of p53 expression were also down-regulated in the absence of USP11 (Fig. 4A). Although U2OS cells express p53 at low but detectable levels (13), p53 expression is little if any in unstressed MCF-7 cells. In this regard, we examined the effect of TNF␣ on p53 expression by MCF-7 cells. In concert with accumulating studies (24 -27), p53 was stabilized after treatment of cells with TNF␣ for 8 h, and its expression was sustained at least 24 h following TNF␣ treatment (Fig. 4B). In these experimental conditions, TNF␣induced p53 expression was completely abrogated in cells silenced for USP11 (Fig. 4C). In concert with this result, mRNA expression of p53-responsive genes, such as Bax and p21, were also attenuated in the absence of USP11 (Fig. 4C). By contrast, mRNA of p53 was constant regardless of USP11 expression, suggesting that USP11 regulates p53 via a post-translational, not a transcriptional, mechanism (Fig. 4C). To further define whether regulation of p53 by USP11 is associated with ubiquitin-proteasome system, MCF-7 cells transfected with scramble siRNA or USP11 siRNA were pretreated with proteasome inhibitor, MG132, followed by treatment with TNF␣. Immunoblot analysis with anti-p53 revealed that, even in MG132treated cells, knocking down USP11 reduced the expression levels of p53, suggesting that regulation of p53 by USP11 occurs independently of ubiquitin-proteasome proteolysis (Fig. 4D).
To determine whether USP11 regulation of p53 involves IKK␣, MCF-7 cells were transfected with scramble siRNA, USP11 siRNA, or IKK␣ siRNA and then treated with or without TNF␣. The results demonstrated that p53 expression was substantially siRNA were treated with TNF␣. Whole cell lysates or nuclear lysates were subjected to immunoblot analysis with the indicated antibodies. In nuclear lysates, a percentage of nuclear RelA reduction was calculated by the densitometric analysis using NIH Image. The amount of nuclear RelA at 30 min after TNF␣ exposure was defined as 100%. The data represent the mean of three independent experiments. PCNA, proliferating cell nuclear antigen. attenuated in USP11-or IKK␣-deficient cells (Fig. 4E). Importantly, the ectopic expression of IKK␣ in USP11-silenced cells restored p53 expression (Fig. 4E), suggesting the possibility that effect of USP11 on p53 expression is regulated exclusively through IKK␣. In this regard, recent studies have demonstrated that IKK␣ specifically translocates into the nucleus and phos-phorylates histone H3 (28,29). Moreover, IKK␣ forms a complex with transcription coactivators such as AIB1/SRC-3 to modulate gene expression (30). Other studies have suggested that IKK␣ is recruited to the chromatin to derepress the silencing mediator for retinoic acid and thyroid hormone receptor (SMRT) (31,32). Taken together, these findings indicate a pivotal role for nuclear IKK␣ as a regulator of transcription factors in response to TNF␣. Given the previous finding that IKK␣ also stabilizes p53 by Ser-20 phosphorylation (11), these results thus support a mechanism by which USP11 regulates the expression of IKK␣, which in turn controls p53 in response to TNF␣.
Regulation of IKK␣ 3 p53 Signaling Pathway by USP11-Previous work showed that protein kinase C ␦ activates IKK␣ in response to oxidative stress (11). Such activation was associated with increased p53 expression, indicating that the protein kinase C ␦ 3 IKK␣ signaling pathway functions as a positive regulator of p53. In this context, it should be clarified whether the protein kinase C ␦ 3 IKK␣ signaling is also involved in TNF␣-induced p53 expression. Nevertheless, the present study identified another regulator of IKK␣, USP11. Interestingly, despite the fact that USP11 is a deubiquitinating enzyme, USP11 regulation of IKK␣ was independent of ubiquitin-deubiquitin status. Instead, USP11 controlled IKK␣ at the transcriptional level by unknown mechanism. Importantly, USP11 also controlled the stabilization and activation of p53 through IKK␣ regulation at a steady-state level and in response to TNF␣. In this regard, there is indeed another possibility that USP11 directly deubiquitinates p53 to stabilize its expression. However, other work demonstrated that overexpression of USP11 has no obvious effect on the levels of p53 ubiquitination or stabilization of p53 (33). The present findings that ectopic expression of IKK␣ in USP11-depleted MCF-7 cells markedly increased p53 expression also support an indirect role for USP11 in the regulation of p53. Although the precise mechanism by which IKK␣ controls p53 remains unclear, available evidence indicated that IKK␣ phosphorylates p53 for its stabilization and activation in response to oxidative stress. In this context, activation of p53 by TNF␣ may be, at least in part, involved in IKK␣-mediated phosphorylation.
A recent study has demonstrated that p53 expression in IKK␣ Ϫ/Ϫ knock-out mouse embryonic fibroblast is comparable with that in wild type mouse embryonic fibroblast (34). The result also demonstrated that p53 expression was substantially high even in the steady-state level, and no increase was observed after TNF␣ exposure. In this regard, given the present study demonstrating that p53 expression was attenuated in human cancer cells silenced for IKK␣ (Fig. 4E), there is indeed an apparent discrepancy to be solved in the future work. However, regulation of expression for human p53 is considerably different from that for mouse p53 (35), suggesting the possibility that human IKK␣ is specifically associated with the regulation of human p53 expression. Obviously, further studies are needed to clarify this issue.
In summary, the present studies demonstrate that USP11 controls IKK␣ by a deubiquitination-independent mechanism. Regulation of IKK␣ by USP11 is associated with p53 stabilization and activation upon exposure to TNF␣ (Fig. 5). These findings indicate that USP11 controls the IKK␣ 3 p53 signaling pathway in response to TNF␣.