Protein Phosphatase 2Cβ Association with the IκB Kinase Complex Is Involved in Regulating NF-κB Activity*

The NF-κB pathway is important in the control of the immune and inflammatory response. One of the critical events in the activation of this pathway is the stimulation of the IκB kinases (IKKs) by cytokines such as tumor necrosis factor-α and interleukin-1. Although the mechanisms that modulate IKK activation have been studied in detail, much less is known about the processes that down-regulate its activity following cytokine treatment. In this study, we utilized biochemical fractionation and mass spectrometry to demonstrate that protein phosphatase 2Cβ (PP2Cβ) can associate with the IKK complex. PP2Cβ association with the IKK complex led to the dephosphorylation of IKKβ and decreased its kinase activity. The binding of PP2Cβ to IKKβ was decreased at early times post-tumor necrosis factor-α treatment and was restored at later times following treatment with this cytokine. Experiments utilizing siRNA directed against PP2Cβ demonstrated an in vivo role for this phosphatase in decreasing IKK activity at late times following cytokine treatment. These studies are consistent with the ability of PP2Cβ to down-regulate cytokine-induced NF-κB activation by altering IKK activity.

The NF-B pathway is important in the control of the immune and inflammatory response. One of the critical events in the activation of this pathway is the stimulation of the IB kinases (IKKs) by cytokines such as tumor necrosis factor-␣ and interleukin-1. Although the mechanisms that modulate IKK activation have been studied in detail, much less is known about the processes that down-regulate its activity following cytokine treatment. In this study, we utilized biochemical fractionation and mass spectrometry to demonstrate that protein phosphatase 2C␤ (PP2C␤) can associate with the IKK complex. PP2C␤ association with the IKK complex led to the dephosphorylation of IKK␤ and decreased its kinase activity. The binding of PP2C␤ to IKK␤ was decreased at early times post-tumor necrosis factor-␣ treatment and was restored at later times following treatment with this cytokine. Experiments utilizing siRNA directed against PP2C␤ demonstrated an in vivo role for this phosphatase in decreasing IKK activity at late times following cytokine treatment. These studies are consistent with the ability of PP2C␤ to down-regulate cytokine-induced NF-B activation by altering IKK activity.
The NF-B pathway is a critical regulator of the cellular response to a variety of stimuli including cytokines such as TNF␣ 1 and interleukin-1, bacterial and viral infection, and double-stranded RNA (1)(2)(3)(4)(5)(6)(7). Cytokines lead to a rapid increase in the activity of the IB kinases, and this is followed by a subsequent decrease in the activity of these kinases, suggesting both positive and negative regulation of the NF-B pathway. A better understanding of the NF-B pathway will be important in defining how these factors modulate the host immune and inflammatory response and prevent apoptosis (1)(2)(3)(4)(5)(6)(7).
The NF-B transcription factors, p105/50, p100/52, p65, c-Rel, and RelB, contain a Rel homology domain that mediates their dimerization and DNA binding properties (2). These proteins are sequestered in the cytoplasm of most cells, where they are bound to a family of inhibitory proteins known as IB (1,3). Treatment of cells with cytokines, including TNF␣ and inter-leukin-1, stimulates the activity of IB kinases that phosphorylate IB on amino-terminal serine residues, leading to its ubiquitination and degradation by the proteasome (3)(4)(5)(6)(7). This process results in the nuclear translocation of the NF-B proteins where they bind to the promoter elements of a variety of genes involved in the control of the immune and inflammatory response (1)(2)(3)(4)(5)(6)(7).
Although a number of studies have been reported on the mechanisms that lead to IKK activation, much less is known about the factors such as phosphatases that may down-regulate its activity. Previous studies suggest that the phosphatases PP2A and PP2B can negatively regulate the NF-B pathway (29 -32). However, the identity of phosphatases that control IKK activity remains to be determined. Four classes of serine/threonine phosphatases have been categorized according to their substrate specificity, divalent cation requirement, and sensitivity to inhibitors. PP1, PP2A, and PP2B (calcineurin) have ϳ40% amino acid identity in their catalytic domains, whereas PP2C does not share significant sequence homology (33). PP1, PP2A, and PP2B (calcineurin) are present in oligomeric complexes associated with their regulatory subunits and are sensitive to the phosphatase inhibitor okadaic acid. In contrast, PP2C is active as a monomer and is insensitive to okadaic acid.
In this study, we present evidence that PP2C␤ can associate with the IKK complex to result in IKK␤ dephosphorylation and reductions in its kinase activity. PP2C␤-mediated reductions in IKK␤ activity were also associated with decreases in NF-B activity. These results suggest that PP2C␤ may down-regulate the NF-B pathway at late times following cytokine stimulation.

EXPERIMENTAL PROCEDURES
Analysis of IKK-associated Proteins-CMV expression vectors encoding FLAG-tagged IKK␥/NEMO and Myc-tagged IKK␤ were transfected into 293 cells. At 48 h post-transfection, the cells were harvested and homogenized in Tris-buffered saline (50 mM Tris-HCl, pH 7.4, and 250 mM NaCl). After centrifugation at 12,000 ϫ g for 10 min, the supernatant was applied to the M2 FLAG affinity gel column (Sigma), and the bound FLAG-tagged proteins were washed extensively with Tris-buffered saline and eluted with FLAG peptide (Sigma). The eluted proteins were then dialyzed against buffer D (20 mM Hepes, pH 7.9, 20% glycerol, 100 mM KCl, 0.2 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM dithiothreitol).
The affinity-purified FLAG-IKK␥/NEMO and associated proteins were precipitated with trichloroacetic acid and resuspended in 100 mM ammonium bicarbonate containing 5% acetonitrile. The purified proteins were reduced, alkylated with dithiothreitol and iodoacetamide, and digested with trypsin. The peptide mixture was loaded onto an on-line capillary HPLC system (Waters, Milford, MA), equilibrated in 0.5% acetic acid, and the peptides were eluted using a linear gradient of 0 -40% acetonitrile over 60 min followed by 40 -60% over 10 min at a flow rate of 0.3 ml/min. The eluted peptides were analyzed using an LCQ-DECA ion trap mass spectrometer (Finnegan, San Jose, CA) (34). All tandem spectra were searched against the University of Washington human data base using the SEQUEST algorithm (35). Data processing of the SEQUEST files to identify proteins associated with IKK␥/NEMO was then performed (36).
DNA Constructs-The human PP2C␤ cDNA was isolated from total HeLa RNA followed by reverse transcriptase-PCR using a SuperScript kit (Invitrogen) as suggested by the manufacturer's protocol. The primers used for the cloning of the PP2C␤ cDNA were 5Ј-GTTCCGAAGCT-TATGGGTGCATTTTTGGATAAACC-3Ј and 5Ј-GCCTAGTCTAGAT-CATATTTTTTCACCACTCATCT-3Ј. The resulting PCR product was cloned into the HindIII and XbaI cloning sites of the pCMV-FLAG and pCMV-Myc expression vectors and confirmed by DNA sequencing. The phosphatase-defective PP2C␤ (Arg 3 Gly) mutant was constructed by substitution of the arginine residue at position 179 with glycine using oligonucleotide-directed mutagenesis with a QuikChange kit (Stratagene) and confirmed by DNA sequencing. FLAG-tagged wild type IKK␣, IKK␤, IKK␥/NEMO, and the constitutively active IKK␤ Ser-Ser 3 Glu-Glu mutant, in which serine residues 177 and 181 were converted to glutamate, were each expressed from a pCMV5 construct as previously described (9,23). The GST-IB␣ fusion protein extending from amino acid 1 to 54 was described previously (37).
Cell Lines and Transfections-293T and HeLa cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% (v/v) fetal bovine serum (Invitrogen). Transfections with NF-B luciferase and Rous sarcoma virus-␤-galactosidase constructs into 293T cells were performed using Gene-Juice (Invitrogen) as previously described (38).
CMV expression vectors containing a neomycin resistance gene with or without the FLAG-PP2C␤ cDNA were transfected into 293 cells to derive G418-resistant PP2C␤ and control cell lines. Clones were selected in the presence of 0.5 mg/ml G418 (Invitrogen), and FLAG-tagged PP2C␤-stably expressing cells were identified by immunoblot analysis with the FLAG antibody.
For siRNA transfection of HeLa cells, cells at 30 -40% confluence were transfected using Oligofectamine (Invitrogen) with 21-mer doublestranded RNA oligonucleotides (40 nM) corresponding to either PP2C␤ (sense 5Ј-GGGAAAAGGAGCGAAUCCATT-3Ј) or the control HTLV-1 Tax (sense 5Ј-GAUGGACGCGUUACGGCUTT-3Ј). At 48 h post-transfection, the cells were either untreated or treated with TNF␣ (10 ng/ml) for different times as described and lysed in PD buffer for kinase assays and Western blot analysis or in Trizol for quantitative real time PCR.
Immunoprecipitation and Immunoblotting-To determine the interactions between PP2C␤ and the IKKs, Myc epitope-tagged wild type PP2C␤ (2.0 g) was transfected into 293T cells with FLAG epitopetagged CMV vectors (2.0 g) encoding either IKK␣, IKK␤, or IKK␥/ NEMO. Extracts (400 g) were prepared in PLC buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1.5 mM MgCl, 1.0 mM EGTA, 10.0 mM NaPP i , 100 mM NaF, and 0.5 mM dithiothreitol) and protease inhibitors (Roche Applied Science). These extracts were incubated for 2 h at 4°C with the FLAG antibody (2.0 g) followed by the addition of 20 l of protein A-agarose (Bio-Rad) for 1.5 h at 4°C. The immunoprecipitated complexes were washed three times with PLC buffer, subjected to electrophoresis on a 10% SDS-polyacrylamide gel, immunoblotted with specific antibodies, and developed using chemiluminescence reagents (Amersham Biosciences).
Anti-PP2C␤ antibody was generated using a genetic immunization technique by the Center for Biomedical Invention at UT Southwestern Medical Center (39). In brief, a codon-optimized human cDNA sequence encoding amino acids 381-480 of PP2C␤ was synthesized and inserted into the expression vector pBQAP-TT. A mixture of expression vectors including pBQAP-TT-PP2C␤, mouse granulocyte-macrophage colonystimulating factor, and mouse Flt3L (2:1:1) was imbedded on gold particles and delivered into the ears of CD1 mice using the Helios gene gun. The mice were boosted every 3 weeks for a total of four immunizations, and serum was collected 4 weeks after the last immunization. This serum was used in Western blot and immunoprecipitation analysis at 1:1000 and 1:200 dilutions, respectively.
Kinase and Phosphatase Assays-CMV expression vectors encoding either FLAG-tagged wild type IKK␤ (0.5 g) or the constitutively active IKK␤ Ser-Ser 3 Glu-Glu construct (0.5 g) were transfected either alone or in the presence of Myc-tagged wild type or Arg 3 Gly mutant of PP2C␤ (3.0 g) into 293T cells and harvested 30 h post-transfection. Extracts (200 g) in PD buffer were immunoprecipitated overnight at 4°C with 1-2 g of FLAG antibody, followed by the addition of protein A-agarose for 1-3 h at 4°C, and extensively washed with PD buffer. In vitro kinase assays were performed for 20 min at 30°C in kinase buffer containing 1.0 mM dithiothreitol, 10 M ATP with 10 Ci of [␥-32 P]ATP and the GST-IB␣ substrate (5.0 g) followed by analysis by SDS-PAGE and autoradiography.
To assay the ability of PP2C␤ to dephosphorylate IKK␤, FLAG-IKK␤ protein was immunoprecipitated from extracts with the FLAG antibody, and autophosphorylation assays were carried out in kinase buffer containing [␥-32 P]ATP. The radiolabeled IKK␤ protein was washed extensively and incubated with or without FLAG affinity column-purified wild type or Arg 3 Gly mutant of PP2C␤ at 30°C for 30 min followed by SDS-PAGE and autoradiography.
In Vivo Phosphorylation Assay-HeLa cells at 60% confluence were grown in Dulbecco's modified Eagle's medium lacking phosphate (Invitrogen) in the presence of 50 Ci of [ 32 P]orthophosphate (5.0 mCi/ml) (PerkinElmer Life Sciences) for 4 h. The cells were then treated with TNF␣ (10 ng/ml) (Roche Applied Science) for the indicated times and harvested in PD buffer. The extracts (200 l) were immunoprecipitated with antibody directed against IKK␣/␤, and the radiolabeled IKK␤ proteins were resolved on a 10% SDS-polyacrylamide gel and visualized by autoradiography.
Quantitative Real Time PCR-Quantitative PCR was utilized to evaluate the efficiency of siRNA-mediated knock-down of PP2C␤ (40). cDNAs were prepared from the control and RNAi-transfected HeLa cells using primers for PP2C␤ forward (5Ј-CTACCGACAACTTCTG-GAGGAG-3Ј) and reverse (5Ј-TCGAAGAAGTAGCTGTGGCAG-3Ј) and 18 S RNA forward (5Ј-AGGAATTGACGGAAGGGCAC-3Ј) and reverse (5Ј-GGACATCTAAGGGCATCACA-3Ј). Each PCR was carried out in triplicate in a 20-l volume using Sybr Green Mastermix (Applied Biosystems) in the ABI Prism 7700 Sequence Detection System (40). Quantitation of PP2C␤ mRNA levels was determined using the ABI dissociation curve and normalized to the amount of 18 S RNA present in each sample.

RESULTS
PP2C␤ Is Associated with the IKK Complex-First, we investigated whether proteins in addition to IKK␣, IKK␤, IKK␥/ NEMO, and Hsp90/Cdc37 were associated with the IKK complex (8 -13, 15-17). For these studies, 293 cells were cotransfected with CMV expression vectors encoding FLAGtagged IKK␥/NEMO and Myc-tagged IKK␤, and extracts were prepared. The FLAG-tagged IKK␥/NEMO was then isolated using FLAG affinity chromatography. After extensive washing of the column to remove nonspecific associated proteins, the remaining proteins were eluted with FLAG peptide and subjected to trypsin digestion followed by HPLC and LCQ-DECA ion trap mass spectrometry. In addition to IKK␣ and IKK␤, three peptides corresponding to the serine/threonine protein phosphatase 2C␤ (PP2C␤) were identified. These results suggested that PP2C␤ could potentially associate with IKK␥/ NEMO and IKK␤ (Table I).
Characterization of IKK Interactions with PP2C␤-To confirm the association of PP2C␤ with the IKK complex, coimmunoprecipitation experiments were performed utilizing G418-resistant 293 cell lines that either stably expressed FLAG-tagged PP2C␤ or did not express this epitope-tagged protein. Western blot analysis with the FLAG antibody confirmed the presence of FLAGtagged PP2C␤ in the stably transfected cell line and its absence in the control cell line (Fig. 1A, lanes 1 and 2).
To analyze the association of the endogenous PP2C␤ with components of the IKK complex (Fig. 1D), extracts prepared from the HeLa cells were immunoprecipitated with either normal mouse IgG (Fig. 1D, lane 1) or a murine polyclonal antibody directed against human PP2C␤ (Fig. 1D, lane 2) followed by Western blot analysis with the antibodies directed against components of the IKK complex. PP2C␤ was associated with components of the IKK complex including IKK␣ (Fig. 1D, top), IKK␤ (Fig. 1D, middle), and IKK␥/NEMO (Fig. 1D, bottom). There was no association of PP2C␤ with these proteins when the extracts were immunoprecipitated with normal mouse IgG (Fig. 1D, lane 1, top). The murine antibody directed against PP2C␤ reacted with both endogenous and transiently overex-pressed FLAG-PP2C␤ (Fig. 1D, bottom). These results demonstrate that both endogenous and overexpressed PP2C␤ interacts with one or more components of the IKK complex.
PP2C␤ Dephosphorylates IKK␤ in Vivo and in Vitro-Cytokine treatment increases the activity of the MAP3 kinase TAK1, which has been demonstrated to function as an upstream kinase that stimulates IKK␤ and activates the NF-B pathway (40,41). PP2C␤ has previously been demonstrated to associate with TAK1 and dephosphorylate this kinase to result in reduced stress-activated protein kinase activity (41). Since IKK␤ is the critical kinase involved in cytokine-induced NF-B activation and can associate either directly or indirectly with PP2C␤, we addressed whether PP2C␤ might dephosphorylate IKK␤ and thus reduce NF-B activity.
HeLa cells were transfected with either wild type or the Arg 3 Gly mutant of PP2C␤ and labeled in vivo with [ 32 P]orthophosphate either in the presence or absence of TNF␣ ( Fig. 2A). Following immunoprecipitation of endogenous 32 P-labeled IKK␤, SDS-PAGE and autoradiography were performed. There was no detectable IKK␤ phosphorylation in untreated cells ( Fig. 2A, lanes 1-3, top), but a marked increase in IKK␤ and likely IKK␣ phosphorylation TABLE I Amino acid sequences of PP2C␤ determined by mass spectroscopy Peptides associated with IKK␥/NEMO were identified by data-dependent liquid chromatography/tandem mass spectrometry analysis on an LCQ-DECA ion trap mass spectrometer and included the peptides that are found in human PP2C␤ (GP: H55801, 53 kDa). was noted following TNF␣ treatment (Fig. 2A, lane 4, top). The phosphorylation of IKK␤ and probably IKK␣ was significantly reduced in TNF␣-treated cells transfected with wild type PP2C␤ ( Fig.  2A, lane 5, top) but not with the PP2C␤ Arg 3 Gly (R/G) mutant ( Fig. 2A, lane 6, top). Western blot analysis of a portion of the unlabeled cellular extracts demonstrated similar levels of endogenous IKK␣/␤ and the transfected Myc-tagged PP2C␤ proteins ( Fig.  2A, bottom).
The ability of PP2C␤ to dephosphorylate another kinase such as MEKK1, which has been implicated in activating the NF-B pathway, was next addressed. Expression vectors encoding HA-tagged IKK␤ or MEKK1 alone or together with either Myc-tagged wild type or the Arg 3 Gly mutant of PP2C␤ were transfected into 293T cells followed by in vivo labeling with [ 32 P]orthophosphate. Immunoprecipitation of HA-tagged IKK␤ demonstrated that its phosphorylation was markedly reduced in the presence of wild type PP2C␤ (Fig. 2B, lanes 2 and 3, top) but not in the presence of PP2C␤ Arg 3 Gly mutant (Fig. 2B,  lane 4, top). There were no changes in the phosphorylation of HA-tagged MEKK1 in either the presence of wild type or the Arg 3 Gly mutant of PP2C␤ (Fig. 2B, lanes 6 and 7, top). Western blot analysis demonstrated similar expression levels of epitope-tagged IKK␤, MEKK1, and PP2C␤ (Fig. 2B, bottom). Similar studies indicated that there was also no effect of PP2C␤ on the in vivo phosphorylation of NIK, another kinase implicated in activating the NF-B pathway (data not shown).
Finally, we addressed whether PP2C␤ could dephosphorylate IKK␤ in in vitro assays. FLAG-tagged IKK␤ expressed in 293T cells was immunoprecipitated and autophosphorylated in vitro by incubation with [␥-32 P]ATP. The 32 P-labeled IKK␤ protein was then incubated with either FLAG affinity-purified wild type or the Arg 3 Gly mutant of PP2C␤ and analyzed following SDS-PAGE and autoradiography (Fig. 2C). There was markedly reduced phosphorylation of IKK␤ in the presence of wild type PP2C␤ (Fig. 2C, lane 2, top) but not in the presence of the PP2C␤ Arg 3 Gly mutant (Fig. 2C, lane 3, top). Western blot analysis demonstrated similar expression of IKK␤ and PP2C␤ (Fig. 2C, bottom). Taken together, both in vivo and in vitro assays demonstrated that PP2C␤ could dephosphorylate IKK␤.
PP2C␤ Inhibits IKK␤ Kinase Activity-Treatment with cytokines such as TNF␣ and interleukin-1 leads to increases in the phosphorylation of serine residues 177 and 181 in the IKK␤ activation loop to stimulate its kinase activity (28,38). Mutation of serine residues 177 and 181 to alanine reduces IKK␤ kinase activity, whereas substitution of these residues with glutamates, which mimics phosphorylation, results in a constitutively active kinase (42). Next we addressed whether PP2C␤ could dephosphorylate serine residues 177 and 181 in IKK␤ to decrease its kinase activity. For these studies, 293T cells were transfected with FLAG-tagged wild type IKK␤ (Fig. 3A) or the constitutively active IKK␤ Ser-Ser 3 Glu-Glu (SS/EE) mutant (Fig. 3B) either alone or together with the Myc-tagged wild type or Arg 3 Gly mutant of PP2C␤. FLAG-tagged IKK␤ was immunoprecipitated from these extracts and assayed in in vitro kinase assays with a GST-IB␣ substrate. The kinase activity of wild type IKK␤ was markedly reduced in the presence of wild type PP2C␤ (Fig. 3A, lane 3, top) but not by the PP2C␤ Arg 3 Gly mutant (Fig. 3A, lane 4, top). In contrast, the kinase activity of IKK␤ Ser-Ser 3 Glu-Glu mutant was not significantly altered in the presence of either wild type or the Arg 3 Gly mutant of PP2C␤ (Fig. 3B, lanes 2-4, top). Western blot analysis demonstrated similar levels of expression of IKK␤ and the IKK␤ Ser-Ser 3 Glu-Glu mutant in addition to the wild type and Arg 3 Gly mutant of PP2C␤ (Fig. 3, A and B, lower panels). These results suggested that the PP2C␤-mediated reductions in IKK␤ kinase activity could potentially be explained by its ability to dephosphorylate serine residues in the IKK␤ activation loop. siRNA Directed against PP2C␤ Increases TNF␣-induced IKK Activity-In order to address whether endogenous PP2C␤ is involved in regulating TNF␣-induced IKK activity, siRNA directed against PP2C␤ was utilized to determine whether it altered IKK activity following TNF␣ treatment. HeLa cells were transfected with Oligofectamine alone or Oligofectamine containing siRNAs directed against either PP2C␤ or the HTLV-1 tax gene as a control. At 48 h post-transfection, cells were treated with TNF␣ and harvested at 0, 5, 15, 30, 60, or 120 min. To determine the ability of siRNA to reduce PP2C␤ mRNA, RNA prepared from these cells was analyzed by quantitative real time PCR (Fig. 4A). Real time PCR analysis demonstrated an ϳ70% inhibition of PP2C␤ mRNA levels in the presence of PP2C␤ siRNA (Fig. 4A, lanes 7-12) as compared with cells transfected with Oligofectamine alone or Tax siRNA (Fig. 4A, lanes 1-6 and 13-18). These results suggested that siRNA transfection could efficiently decrease the amount of PP2C␤ mRNA in HeLa cells.
The extracts from this experiment were also analyzed for the levels of phospho-IB␣, IB␣, endogenous PP2C␤, IKK␤, and actin (Fig. 4C). There was enhanced phosphorylation of IB␣ at 5 min post-TNF␣ treatment in control and PP2C␤ and Tax siRNA-treated cells (Fig. 4C, lanes 2, 8, and 14, top). At 15 min post-TNF␣ treatment, phospho-IB␣ levels were decreased in the control as well as in the siRNA-treated cells (Fig. 4C, lanes  3, 9, and 15, top). However, there was no detectable phosphorylation of IB␣ in the extracts prepared from the control and Tax siRNA-transfected cells between 30 and 120 min post-TNF␣ treatment (Fig. 4C, lanes 4 -6 and 16 -18, top). In contrast, there were significant levels of phospho-IB␣ in the extracts prepared from the PP2C␤ siRNA-transfected cells at these times (Fig. 4C, lanes 10 -12, top). Total IB␣ levels were decreased at 15 min post-TNF␣ treatment in extracts prepared from both the control and the siRNA-treated cells and increased by 60 min post-TNF␣ treatment (Fig. 4C, middle). A slight increase in IB␣ levels at 60 min post-TNF␣ treatment was consistently seen in extracts prepared from the PP2C␤ siRNA-treated cells as compared with that seen in the control and Tax siRNA-treated cells. The PP2C␤ siRNA resulted in a 70% reduction in endogenous PP2C␤ levels (Fig. 4C, lanes  7-12, middle) as compared with control and Tax siRNA-transfected cells (Fig. 4C, lanes 1-6 and 13-18, middle). Similar levels of IKK␤ and actin were noted (Fig. 4C, bottom). These results, which were seen in three independent experiments, suggested that siRNA directed against PP2C␤ could result in a prolonged increase in TNF␣-mediated IKK␤ activity.
PP2C␤ Decreases NF-B-directed Gene Expression-Next we addressed whether overexpression of PP2C␤ altered the kinetics of TNF␣-mediated IB␣ degradation and NF-B-regulated gene expression. Parental cells and 293 cells stably expressing FLAG-tagged PP2C␤ were treated with TNF␣ for various times, and extracts were analyzed by Western blot for changes in the levels of phospho-IB␣ and total IB␣ levels. The cells stably expressing the FLAG-tagged PP2C␤ demonstrated slightly reduced levels of phospho-IB␣ at both early (Fig. 5A,  lanes 2 and 9, top) and later times (Fig. 5A, lanes 5-7 and  12-14, top) post-TNF␣ treatment as compared with parental cells. The cells stably expressing FLAG-tagged PP2C␤ also exhibited reduced degradation of IB␣ at both early and late times post-TNF␣ treatment (Fig. 5A, lanes 3-7 and 10 -14,  middle). Western blot analysis demonstrated similar expression of IKK␣, IKK␤, FLAG-PP2C␤, and actin. These results indicated that overexpression of PP2C␤ reduced phospho-IB␣ levels and IB␣ degradation and was also associated with decreased resynthesis of IB␣.
To determine whether PP2C␤ overexpression altered NF-Bmediated gene expression, the parental cells and cells stably expressing FLAG-tagged PP2C␤ were cotransfected with an NF-B luciferase reporter and a Rous sarcoma virus-␤-galactosidase expression vector. At 18 h post-transfection, the cells were either untreated or treated with TNF␣ for 6 h prior to assaying luciferase activity (Fig. 5B). NF-B reporter activity was increased in both TNF␣-treated parental cells and cells stably expressing FLAG-tagged PP2C␤ (Fig. 5B). However, TNF␣-mediated increases in NF-B activity were reduced by 60% in cells stably expressing FLAG-tagged PP2C␤ as compared with parental cells. These experiments were repeated three times, and the average of triplicate samples is shown. These studies indicated that PP2C␤ decreases TNF␣-mediated NF-B activation.
Finally, we addressed whether the interactions of FLAG-PP2C␤ and IKK␤ were altered following TNF␣ stimulation. Parental cells and cells stably expressing FLAG-tagged PP2C␤ were treated with TNF␣ for various times, and extracts were prepared and immunoprecipitated with IKK␣/␤ antibody or normal rabbit IgG followed by Western blot analysis with FLAG antibody (Fig. 5C). Significant interactions between FLAG-PP2C␤ and IKK␣/␤ were demonstrated in the absence of TNF␣ (Fig. 5C, lane 6, top). However, these interactions were  (Fig. 5C, lanes 9 and 10, top). There was no detectable interaction between FLAG-PP2C␤ and IKK in parental cells or when normal rabbit IgG was used to immunoprecipitate IKK␣/␤ from the cells stably expressing FLAGtagged PP2C␤ (Fig. 5C, top and middle). Western blot analysis demonstrated similar expression of FLAG-PP2C␤, IKK␣, and IKK␤ (Fig. 5C, bottom) and TNF␣-mediated IB␣ degradation (Fig. 5C, bottom). These results indicated that the association of PP2C␤ and IKK␣/␤ was regulated by TNF␣ treatment and that the kinetics of this association appeared to correlate with changes in IKK activity. DISCUSSION In this study, we present evidence suggesting that PP2C␤ negatively regulates the NF-B pathway post-TNF␣ treatment by dephosphorylating IKK␤ and thus reducing its kinase activity. Several lines of evidence substantiate these conclusions. First, we found that both endogenous and overexpressed PP2C␤ interacted with the IKK complex. Second, we observed that PP2C␤ dephosphorylated both endogenous and transiently expressed IKK␤ but not other kinases including MEKK1 and NIK. Third, PP2C␤ reduced TNF␣-mediated increases in wild type IKK␤ activity while not changing the activity of the constitu-tively active IKK␤ Ser-Ser 3 Glu-Glu mutant. Fourth, siRNA directed against PP2C␤, but not the control tax gene, prolonged cytokine-induced IKK␤ activity to result in increased phospho-IB␣ levels. Finally, overexpression of PP2C␤ reduced TNF␣mediated IB␣ degradation and resynthesis as well as the levels of phospho-IB␣, leading to decreases in NF-B reporter activity. Collectively, these data are consistent with a role for PP2C␤ in down-regulating NF-B activity at late times post-TNF␣ treatment by associating with the IKK␣/␤ complex and dephosphorylating IKK␤.
IKK␤ exhibits maximum kinase activity within 5 min following cytokine stimulation, and its activity is decreased by 30 -60 min post-TNF␣ stimulation. The mechanism by which IKK␤ activity is decreased following cytokine stimulation has not been totally elucidated. In the current study, mass spectrometry and protein interaction studies demonstrated that IKK subunits can associate with the serine/threonine phosphatase PP2C␤. The fact that PP2C␤ alters the activity of wild type IKK␤ but not the constitutive IKK␤ Ser-Ser 3 Glu-Glu mutant suggests that PP2C␤ probably acts on IKK itself or an upstream kinase to down-regulate the NF-B pathway. PP2C␤ has been reported to dephosphorylate the upstream kinase TAK1, which stimulates IKK activity (40,41). However, the effects of PP2C␤ were relatively specific in that this phospha- FIG. 4. PP2C␤ siRNA alters the kinetics of TNF␣-induced IKK␤ activity. A, HeLa cells were transfected with either Oligofectamine (lanes 1-6) or Oligofectamine containing 40 nM annealed sense and antisense 21-mer siRNA oligonucleotides directed against PP2C␤ (lanes 7-12) or the HTLV-I tax gene (lanes [13][14][15][16][17][18]. At 48 h post-transfection, HeLa cells were either untreated (lanes 1, 7, and 13) or treated with TNF␣ (10 ng/ml) for 5 (lanes 2, 8, and 14), 15 (lanes 3, 9,  and 15), 30 (lanes 4, 10, and 16), 60 (lanes 5, 11, and 17) or 120 (lanes 6, 12, and 18) min. Total RNA was prepared from each sample, and quantitative real time PCR was utilized to analyze the expression of the PP2C␤ gene and normalized to the expression levels of 18 S RNA. These experiments were repeated three times, and the average of triplicate samples is shown, with the error bars denoting the S.E. B, HeLa extracts (200 g) were prepared from these cells and immunoprecipitated (IP) with an IKK␣/␤ antibody followed by in vitro kinase assays with the GST-IB␣ substrate and SDS-PAGE and autoradiography (top). Extracts were also analyzed by Western blot for IKK␣ and IKK␤ expression (bottom). C, Western blot analysis was performed on the extracts from B using antibodies directed against phospho-IB␣, IB␣, endogenous PP2C␤, IKK␤, and actin. tase did not alter the phosphorylation of two other kinases, MEKK1 and NIK, reported to be involved in IKK activation. Thus, in addition to IKK␤, only a subset of upstream kinases that have been reported to activate the NF-B pathway may potentially be targets for PP2C␤. In summary, the results from both siRNA studies and overexpression of PP2C␤ implicate this phosphatase in regulating the NF-B pathway in response to TNF␣ treatment.
Other serine/threonine protein phosphatases including PP2A and PP2B have been implicated in the negative regulation of signaling pathways including the NF-B pathway (43). For example, treatment with okadaic acid, which is an inhibitor of PP2A, can activate the NF-B pathway (32,44). Furthermore, PP2A binding to IKK␥/NEMO inhibits IKK activity, and this effect can be reversed by the interaction of PP2A with the HTLV-1 Tax protein to result in constitutive IKK activity (30). PP2A has also been demonstrated to interact with and dephosphorylate RelA to inhibit the NF-B activity (29), whereas PP2B (calcineurin) has been shown to dephosphorylate IB and decrease NF-B activity following growth factor stimulation (31). Thus, multiple phosphatases have been implicated in down-regulating NF-B activity by a variety of different mechanisms.
At least seven distinct PP2C gene products (2C␣, 2C␤, 2C␥, 2C␦, 2C⑀, Wip1, and Ca 2ϩ /calmodulin-dependent protein kinase phosphatase) have been described in mammalian cells (45) in addition to a variety of spliced variants (46 -49). Of the six different members of the PP2C family, three (PP2C␣, PP2C␤, and Wip1) have been shown to be involved in the negative regulation of signaling pathways including stressactivated protein kinase (41,47,50,51). For example, the ectopic expression of mouse PP2C␣ and PP2C␤-1 can inhibit the stress-activated MKK3/6-p38 and MKK4/7-c-Jun N-terminal kinase pathways probably via dephosphorylation of TAK1, but they do not alter the mitogen-activated MKK1-ERK1 pathway (45,47,50). PP2C␣ and PP2C␤ have also been demonstrated to dephosphorylate the cyclin-dependent kinases Cdk2 and Cdk6 and inhibit their activity (52). However, PP2C family members do not always function as negative regulators of signaling pathways. For example, PP2C␣ can function as a positive regulator of Wnt signaling by dephosphorylating Axin (53). Thus, the PP2C family members can play both positive and negative roles in regulating signal transduction pathways. There are likely multiple mechanisms that can down-regulate IKK activity following cytokine treatment. First, phosphatases such as PP2C␤ and PP2C⑀ may function to inhibit the activity of upstream kinases such as TAK1 that stimulate IKK  7 and 14) min. Western blot analysis was performed on extracts prepared from these cells using antibodies directed against phospho-IB␣ (top) or IB␣, IKK␣, IKK␤, the FLAG epitope, and actin (bottom). B, parental cells or cells stably expressing FLAG-tagged PP2C␤ were cotransfected with an NF-B luciferase reporter vector (0.2 g) and a Rous sarcoma virus-␤-galactosidase expression vector (0.2 g). At 18 h post-transfection, the cells were either left untreated or treated with TNF␣ (5 ng/ml) and harvested 6 h later. Luciferase and ␤-galactosidase activity was then determined. These experiments were repeated three times, and the average of triplicate samples is shown with the error bars denoting the S.E. C, parental cells and cells stably expressing FLAG-tagged PP2C␤ were either untreated (lanes 1 and 6) or treated with TNF␣ (10 ng/ml) for 5 (lanes 2 and 7), 15 (lanes 3 and 8), 30 (lanes 4 and 9), or 60 (lanes 5 and 10) min. Extracts were immunoprecipitated (IP) with an IKK␣/␤ antibody (top) or normal rabbit IgG (middle), and Western blot analysis was performed with FLAG antibody to detect PP2C␤ (lanes 1-11; top and middle). Western blot analysis was performed on extracts prepared from these cells using either FLAG, IKK␣, IKK␤, or IB␣ antibodies (bottom). (40,41,45). Second, phosphatases such as PP2A can interact with IKK␥/NEMO to inhibit IKK activity (8,30). Third, IKK can undergo autophosphorylation at multiple sites in its carboxyl terminus to down-regulate its activity (28). Finally, we demonstrate in this study that PP2C␤ associates with the IKK complex and leads to reduced IKK activity at late times following TNF␣ treatment. These results suggest that multiple mechanisms, including phosphatases such as PP2C␤, are probably involved in both maintaining basal IKK activity and downregulating IKK activity following cytokine treatment.