Regulation of Ikappa B kinase (IKK)gamma /NEMO function by IKKbeta -mediated phosphorylation.

The IκB kinase (IKK) complex includes the catalytic components IKKα and IKKβ in addition to the scaffold protein IKKγ/NEMO. Increases in the activity of the IKK complex result in the phosphorylation and subsequent degradation of IκB and the activation of the NF-κB pathway. Recent data indicate that the constitutive activation of the NF-κB pathway by the human T-cell lymphotrophic virus, type I, Tax protein leads to enhanced phosphorylation of IKKγ/NEMO by IKKβ. To address further the significance of IKKβ-mediated phosphorylation of IKKγ/NEMO, we determined the sites in IKKγ/NEMO that were phosphorylated by IKKβ, and we assayed whether IKKγ/NEMO phosphorylation was involved in modulating IKKβ activity. IKKγ/NEMO is rapidly phosphorylated following treatment of cells with stimuli such as tumor necrosis factor-α and interleukin-1 that activate the NF-κB pathway. By using both in vitro and in vivo assays, IKKβ was found to phosphorylate IKKγ/NEMO predominantly in its carboxyl terminus on serine residue 369 in addition to sites in the central region of this protein. Surprisingly, mutation of these carboxyl-terminal serine residues increased the ability of IKKγ/NEMO to stimulate IKKβ kinase activity. These results indicate that the differential phosphorylation of IKKγ/NEMO by IKKβ and perhaps other kinases may be important in regulating IKK activity.

The NF-B pathway is a critical regulator of the cellular response to a variety of stimuli including the cytokines, TNF␣ 1 and IL-1, bacterial and viral infection, double-stranded RNA, and the human T-cell leukemia virus transactivator protein Tax (1)(2)(3)(4). The ability to activate rapidly and subsequently silence the NF-B pathway in response to a variety of extracellular stimuli suggests that both positive and negative regulation is involved in its control. The further characterization of the mechanisms that regulate this pathway will be important for better understanding how NF-B is involved in the control of the host immune and inflammatory responses.
The members of the NF-B family of transcription factors, which include p105/50, p100/52, p65, c-Rel, and RelB, contain a Rel homology domain that mediates their heterodimerization and homodimerization properties 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 a variety of stimuli including the cytokines TNF␣ and IL-1 stimulate the activity of kinases that phosphorylate IB on amino-terminal serine residues resulting in its ubiquitination and degradation by the proteasome (3)(4)(5)(6). This process leads to the nuclear translocation of the NF-B proteins, which then bind to consensus DNA sequences located upstream of a variety of cellular genes that are involved in the control of the immune and the inflammatory response and prevent apoptosis (1)(2)(3)(5)(6)(7).
Activation of the IB kinases, which phosphorylate the IB proteins on the amino-terminal serine residues to result in their degradation, is a critical process that regulates the NF-B pathway (8 -12). The IB kinases, IKK␣ and IKK␤, are components of a 600 -900-kDa complex that is composed of these two catalytic subunits (8 -12) in addition to a scaffold protein IKK␥/ NEMO (13)(14)(15)(16). Stimulation of IKK activity by cytokine treatment has been demonstrated to involve both activation of mitogen-activated protein 3-kinases and/or IKK autophosphorylation (17,18). IKK␣ and IKK␤ have 52% amino acid identity, and their domain structure is similar with an amino-terminal kinase, leucine zipper, and helix-loop-helix motifs (8 -12). Although these kinases have a similar structure and are able to both homodimerize and heterodimerize, IKK␤ is at least 20fold more active in phosphorylation of the IB proteins as compared with IKK␣ (9,14,18,19). Studies using fibroblasts isolated from IKK␣ (20 -22) and IKK␤ (23)(24)(25) knock-out mice also demonstrate that IKK␤ is the dominant kinase in regulating NF-B activity.
IKK␥/NEMO is critical for increasing IKK activity in response to all known stimuli. This protein contains several distinct domains including an amino-terminal domain that mediates its interaction with IKK␤, a coiled-coil domain that is important in its oligomerization, a leucine zipper of as yet uncharacterized function, and the carboxyl terminus that mediates the recruitment of upstream kinases that are involved in modulating IKK activity (13-16, 26 -30). The interaction of IKK␥/NEMO with IKK␣ and IKK␤ is critical for the assembly of this high molecular weight IKK complex that leads to the recruitment of IB proteins and the stimulation of IKK␤ activity (8, 12, 14 -16, 25, 27, 31-33). Cells lacking IKK␥/NEMO are unable to assemble the high molecular weight IKK complex and exhibit severe defects in IKK activation in response to agents that stimulate the NF-B pathway (14 -16). IKK␥/ NEMO also binds to a variety of proteins other than IKKs including the adaptor protein RIP, which is involved in TNF␣mediated activation of IKK (13,30), A20 which decreases TNF␣-mediated activation of IKK (30), the human T-cell lymphotrophic virus, type I, Tax protein which stimulates IKK activity (26,34,35), and the CIKS protein which also increases IKK activity (36). Disruption of a single copy of the IKK␥/ NEMO gene, which is located on the X chromosome, in mice or in humans results in male lethality due to hepatic apoptosis, whereas females heterozygous for this defect develop a severe skin disease known as incontentia pigmentia (37)(38)(39). Thus, IKK␥/NEMO plays a central role in the activation of the NF-B pathway in response to a variety of different stimuli.
Recently it was demonstrated (40) that the human T-cell lymphotrophic virus, type I, Tax protein, which results in the constitutive activation of the NF-B pathway, leads to constitutive phosphorylation of both IKK␤ and IKK␥/NEMO. Furthermore, IKK␤ was shown to phosphorylate IKK␥/NEMO using in vitro kinase assays. These results suggested that IKK␤ and IKK␥/NEMO could potentially regulate the function of each other. In this study, both in vivo and in vitro assays were utilized to characterize IKK␥/NEMO phosphorylation by IKK␤.
Our results indicate that IKK␤ phosphorylation of IKK␥/ NEMO appears to be important for regulating its functional properties.
Transfections and Cellular Fractionation-293T cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (37). HeLa cells were maintained in Iscove's modified Dulbecco's media and supplemented with the same components as above. Transfections were carried out using FuGENE-6 (Roche Molecular Biochemicals) as described by the manufacturer. Cytoplasmic extracts from 293T cells and HeLa cells were prepared as described previously (27).
IKK␥/NEMO Ϫ/Ϫ mouse embryo fibroblasts (37) were plated in 35-mm tissue culture wells with Dulbecco's modified Eagle's media. After 24 h, the cells were transfected with CMV expression vectors encoding either wild-type or mutant Myc-tagged IKK␥/NEMO (0.3 g), an NF-B luciferace reporter (0.1 g), and an RSV-␤-galactosidase expression vector (0.1 g). After 18 h of transfection, the cells were treated with TNF␣ (10 ng/ml) for 6 h. The cells were then treated with reporter lysis buffer (Promega), and luciferase activity was determined according to the manufacturer's protocol (Promega). The transfection efficiency was monitored by assaying ␤-galactosidase activity. All transfections were performed in triplicate and repeated in three independent experiments.
Immunoprecipitation and Immunoblotting-To determine the interactions between wild-type and mutant IKK␥/NEMO and IKK␤, wildtype or mutant CMV-IKK␤ (0.1 g) and either wild-type or mutant CMV-IKK␥/NEMO vectors (1.0 g) were transfected into 293T cells. Extracts (400 g) were then prepared in PD buffer (40 mM Tris-HCl, pH 8.0, 500 mM NaCl, 6.0 mM EGTA, 6.0 mM EDTA, 10 mM ␤-glycerophosphate, 0.5 mM dithiothreitol, 10 mM NaF, 300 M sodium vanadate, and protease inhibitors (Roche Molecular Biochemicals)), incubated for 2-4 h at 4°C with the Myc monoclonal antibody (2.0 g), directed against the Myc epitope, followed by the addition of 20 l of protein A-agarose beads for 1 h at 4°C. The immunoprecipitates were washed three times with PD buffer. Electrophoresis on a 10% SDS-polyacrylamide gel was performed, and the gel was subjected to immunoblotting with specific antibodies and developed using chemiluminescence reagents (Amersham Biosciences).
Kinase Assays-To assay IKK␤ phosphorylation of IKK␥/NEMO, 293T cells were transfected with wild-type or mutant Myc-tagged CMV-IKK␥/NEMO (4.0 g) or FLAG-tagged CMV-IKK␤ (2.0 g) and harvested 30 h post-transfection. Cytoplasmic extracts (400 g) were incubated overnight at 4°C with 1-2 g of anti-Myc or anti-FLAG monoclonal antibodies, followed by the addition of protein A-agarose (Bio-Rad) for 1-3 h at 4°C and washed three times with ELB buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 5.0 mM NaF, 5.0 mM ␤-glycerophosphate, and 1.0 mM sodium vanadate). In vitro kinase assays were performed for 30 min at 30°C in the presence of kinase buffer containing 1.0 mM dithiothreitol, 10 M ATP, and 10 Ci of [␥-32 P]ATP, and the reactions were stopped with protein sample buffer and heated at 95°C for 3 min as described previously (27). The samples were then subjected to SDS-PAGE on a 12% polyacrylamide gel and visualized by autoradiography.
To assay IKK␤ phosphorylation of wild-type and mutant GST-IKK␥/ NEMO proteins, wild-type or mutant CMV-IKK␤ (2.0 g) was transfected into 293T cells, and at 30 h post-transfection, the cells were harvested and lysed in PD buffer. The extracts were immunoprecipitated with the M2 FLAG monoclonal antibody, and following the addition of protein A-agarose, in vitro kinase assays with the GST-IKK␥/ NEMO substrate (10.0 g) were performed as described above. Finally, to assay increases in IKK␤ activity by wild-type and mutant IKK␥/ NEMO, 293T cells were cotransfected with CMV expression vectors encoding IKK␥/NEMO (0.4 g) and IKK␤ (0.01 g), respectively. At 30 h post-transfection, cellular lysates (200 g) were prepared in PD buffer, and in vitro kinase assays with a GST-IB␣ substrate (10.0 g) were performed as described above.
In Vivo Phosphorylation-For in vivo labeling, HeLa cells at 60% confluence were grown in Dulbecco's modified Eagle's medium lacking either phosphate or methionine (Invitrogen) in the absence of serum for 4 h. At that time, either 50 Ci of [ 32 P]orthophosphate (50 Ci/ml) or [ 35 S]methionine (50 Ci/ml) (PerkinElmer Life Sciences) was added for 4 h. The cells were then treated with either TNF␣ (20 ng/ml) (Roche Molecular Biochemicals) or IL-1 (20 ng/ml) (Roche Molecular Biochemicals) for the indicated times and then harvested in PD buffer. The cellular lysates (200 l) were incubated overnight at 4°C with a monoclonal antibody directed against IKK␥/NEMO (BD PharMingen); the immunoprecipitates were isolated following the addition of 20 l of protein A-agarose and washed with PD and then RIPA buffer, and the labeled IKK␥/NEMO proteins were resolved on a 10% SDS-polyacrylamide gel and visualized by autoradiography.
Mass Spectrometry-In vitro kinase assays were performed with immunopurified IKK␤ and GST-IKK␥/NEMO in the presence of [␥-32 P]ATP. The 32 P-labeled GST-IKK␥/NEMO was subjected to SDS-PAGE, and the 32 P-labeled GST-IKK␥/NEMO species was excised from the gel and subjected to trypsin digestion overnight. The trypsin-digested IKK␥/NEMO protein was applied to a reverse-phase high pressure liquid chromatography (Applied Biosystems, 130 A Separation system), and the fractions were collected. The majority of the counts were found in fractions 20 and 21, and the phosphopeptides isolated from these high pressure liquid chromatography fractions were analyzed by a matrix-assisted laser desorption ionization time-of-flight mass spectrometer (Voyager-DE TM , Biospectrometry Workstation, Perspective Biosystems), and the amino acid sequence was analyzed by protein microsequencing using an Applied Biosystems 494 protein sequencer.
Phosphoamino Acid Analysis-Cytoplasmic extract (200 g) was prepared from 293T cells transfected with Myc-tagged CMV-IKK␥/NEMO and FLAG-tagged CMV-IKK␤ and immunoprecipitated with Myc antibody to isolate the IKK␥/NEMO and IKK␤. In vitro kinase assays were then performed with [␥-32 P]ATP, and the reactions were subjected to electrophoresis on a 10% SDS-polyacrylamide gel and transferred to a polyvinylidene difluoride membrane, and the 32 P-labeled IKK␥/NEMO species was isolated and subjected to hydrolysis in 6.0 N HCl at 110°C for 1 h (42). The 32 P-labeled IKK␥/NEMO residues and the unlabeled phosphoserine, phosphothreonine, and phosphotyrosine standards were analyzed by one-dimensional electrophoresis using thin layer cellulose chromatography. The cellulose TLC plate was then stained with 2% ninhydrin (Sigma) followed by autoradiography.
Because the kinetics of IKK␥/NEMO phosphorylation appeared to correlate with increases in IKK activity, we asked whether either wild-type, constitutively active, or kinase-defective constructs of IKK␤ or IKK␣ could result in changes in IKK␥/NEMO phosphorylation. In vivo [ 32 P]orthophosphate labeling was performed on 293T cells that were transfected with either Myc-tagged IKK␥/NEMO, wild-type, or mutant IKK constructs or both IKK␥/NEMO and IKK. The samples were then analyzed following immunoprecipitation with a Myc monoclonal antibody and SDS-PAGE and autoradiography. There was little phosphorylation of IKK␥/NEMO alone (Fig. 1B, lane 2, top panel) but increased phosphorylation in the presence of the wild-type and constitutively active (S177E/S181E) IKK␤ (Fig. 1B, lanes 3 and 5, top panel). In contrast, there was no detectable phosphorylation of IKK␥/NEMO with IKK␤ (S177A/ S181A) (Fig. 1B, lane 7, top panel). The epitope-tagged wildtype and constitutively active IKK␤ proteins that coimmunoprecipitated with IKK␥/NEMO were also highly phosphorylated (Fig. 1B, lanes 3 and 5, top panel). Neither wild-type, constitutively active (S176E/S180E), or kinase-defective (S176A/S180A) IKK␣ mutants were able to stimulate IKK␥/ NEMO phosphorylation (Fig. 1B, lanes 9 -14, top panel). Western blot analysis indicated that IKK␣, IKK␤, and IKK␥/NEMO were expressed at similar levels in each of these transfections (Fig. 1B, lower panels). Thus, there was a marked enhancement of IKK␥/NEMO phosphorylation in the presence of wildtype and constitutively active IKK␤ but not IKK␣.
Finally, we addressed the amino acid residues in IKK␥/ NEMO that were phosphorylated by IKK␤. Expression vectors encoding FLAG-tagged IKK␤ and Myc-tagged IKK␥/NEMO were cotransfected into 293T cells, and following immunoprecipitation of the epitope-tagged IKK␥/NEMO and the associated IKK␤ in vitro kinase assays were performed. The reactions were then subjected to SDS-PAGE and transferred to a polyvinylidene difluoride membrane followed by autoradiography. Analysis of the 32 P-labeled IKK␥/NEMO species indicated that IKK␥/NEMO was phosphorylated predominantly on serine residues (Fig. 1C). In summary, phosphorylation of IKK␥/ NEMO was enhanced in response to treatment of cells with TNF␣ and IL-1. Furthermore, IKK␤ expression results in enhanced IKK␥/NEMO phosphorylation.
Domains in IKK␥/NEMO That Are Phosphorylated by IKK␤-Because IKK␥/NEMO phosphorylation was increased by cytokines that stimulate IKK␤ activity and also by the transfection of wild-type or constitutively active IKK␤, we attempted to identify the regions in IKK␥/NEMO that were phosphorylated by IKK␤. The amino acid sequence of IKK␥/NEMO is indicated as are the positions of the carboxyl-terminal deletion mutants and potential sites of IKK␤ phosphorylation ( Fig.  2A). In addition, the structural domains in IKK␥/NEMO and a schematic of the various deletion mutants that were assayed in in vitro kinase assays are shown (Fig. 2B).
To address further the domains in IKK␥/NEMO that were phosphorylated by IKK␤, CMV expression vectors encoding Myc-tagged wild-type and mutant IKK␥/NEMO constructs and FLAG-tagged IKK␤ were cotransfected into 293T cells. Following immunoprecipitation of either epitope-tagged IKK␤ (Fig.  3B, top panel) or IKK␥/NEMO (Fig. 3B, middle panel) with monoclonal antibodies directed against these epitopes, in vitro kinase assays were performed, and the samples were analyzed by SDS-PAGE and autoradiography. Wild-type IKK␤, but not the IKK␤ (K/M) mutant, resulted in phosphorylation of IKK␥/ NEMO and was itself autophosphorylated (Fig. 3B, lanes 2 and  12, top and middle panels). Similar to the results seen with IKK␤ and the GST-IKK␥/NEMO mutants, there was a marked reduction in the ability of IKK␤ to phosphorylate IKK␥/NEMO when the regions between amino acid residues 358 and 394 and 137 and 180 were deleted (Fig. 3B, lanes 4 and 8, top and middle panels). Western blot analysis revealed similar levels of epitope-tagged IKK␥/NEMO and IKK␤ proteins in these transfections (Fig. 3B, bottom panels). These results indicate that at least two regions of IKK␥/NEMO are phosphorylated by IKK␤.
Sites in IKK␥/NEMO That Are Targets of IKK␤ Phosphorylation-Next mass spectrometry was utilized to identify domains in GST-IKK␥/NEMO that were phosphorylated in vitro by IKK␤. Following in vitro kinase assays with IKK␤ and GST-IKK␥/NEMO in the presence of [␥-32 P]ATP, SDS-PAGE was performed, and the 32 P-labeled GST-IKK␥/NEMO species was excised from the polyacrylamide gel and subjected to digestion with trypsin. The trypsin-digested IKK␥/NEMO protein was then applied to a reverse-phase high pressure liquid chromatograph, and the fractions were collected. Two fractions (fractions 20 and 21), which contained the majority of the 32 P-labeled peptides, were analyzed by matrix-assisted laser desorption ionization time-of-flight and then subjected to peptide microsequencing. This analysis indicated that fraction 20 was composed entirely of a peptide corresponding to amino acids 353-378 of IKK␥/NEMO, whereas fraction 21 was composed of a peptide corresponding to amino acids 144 -159 of IKK␥/NEMO (data not shown).
An examination of the sequences in these two domains indicated that between residues 353 and 378 there was a threonine residue at position 356 and serine residues at positions 369 and 375, whereas between residues 144 and 159 there was a threonine residue at position 147 and serine residues at positions 148, 156, and 158 ( Fig. 2A). These data and our mutagenesis studies, which indicated that a region of IKK␥/NEMO between 358 and 394 was a target for IKK␤ phosphorylation, suggested that serine residues 369 and 375 may be targets for IKK␤mediated phosphorylation. These serine residues were mutated to alanine. Mutagenesis of IKK␥/NEMO also indicated that another region between residues 137 and 180 was a target for IKK␤ phosphorylation. Thus, the serine and threonine residues contained within the phosphopeptides 144 -159 identified by mass spectrometry were also mutated to alanine. Fi-  11 and 12). B, CMV expression vectors alone (lane 1) or encoding Myc-tagged wild-type (WT) or mutant IKK␥/NEMO constructs (4.0 g) were transfected into 293T cells with 2.0 g of either wild-type IKK␤ (lanes 2-10) or wild-type IKK␤ alone (lane 11). In addition, IKK␤ (K44M) was transfected either with wild-type IKK␥/NEMO (lane 12) or alone (lane 13). Extracts were prepared and immunoprecipitated (IP) with FLAG monoclonal antibody (top panel) or Myc monoclonal antibody (middle panel) followed by in vitro kinase assays, SDS-PAGE, and autoradiography. Western blot analysis was performed on a portion of these extracts with Myc and FLAG monoclonal antibodies to detect the epitope-tagged IKK␥/NEMO and IKK␤ proteins, respectively (bottom panels).
nally, residues 147, 148, 156 and 158, and 369 and 375 were all mutated to alanine in the context of a GST-IKK␥/NEMO fusion protein (Fig. 4A). In vitro kinase assays using these GST-IKK␥/ NEMO fusion proteins indicated that substitution of serine residues 369 and 375 with alanine markedly reduced its phosphorylation by IKK␤ (Fig. 4A, lanes 1 and 2, top panel). In contrast, there was little change in phosphorylation of GST-IKK␥/NEMO by IKK␤ when threonine residue 147 and serine residues 148, 156, and 158 were substituted with alanine (Fig.  4A, lane 3, top panel). A GST-IKK␥/NEMO protein containing a combination of these two sets of mutations exhibited a marked reduction in phosphorylation by IKK␤ similar to that seen with the mutation of serine residues 369 and 375 (Fig. 4A,  lane 4, top panel). These results suggested that either residues 369 and 375 alone or in combination were the predominant sites in IKK␥/NEMO for phosphorylation by IKK␤.
It was important to address whether mutation of either residues 147, 148, 156, and 158 altered IKK␤ phosphorylation of GST-IKK␥/NEMO. For these studies, we utilized a GST-IKK␥/NEMO fusion protein extending between residues 1 and 180 to assay the role of these residues in the absence of the carboxyl-terminal phosphorylation sites. Site-directed mutants in which alanine was substituted for threonine residue 147 and serine residue 148 (Fig. 4B, lane 2, top panel), serine residues 156 and 158 (Fig. 4B, lane 3, top panel), or all of these residues (Fig. 4B, lane 4, top panel) resulted in decreased phosphorylation by IKK␤. These results suggested that there were likely multiple residues that were phosphorylated by IKK␤ in this region of IKK␥/NEMO. Amino Acid Residues in IKK␥/NEMO Phosphorylated by IKK␤-To confirm the previous GST-IKK␥/NEMO mutagenesis results and also address the specific residues in the carboxyl terminus of IKK␥/NEMO that were phosphorylated by IKK␤, expression vectors encoding Myc-tagged wild-type and mutant IKK␥/NEMO and FLAG-tagged IKK␤ were transfected into 293T cells. Following immunoprecipitation with epitope-specific antibodies directed against either IKK␤ (top panel) or IKK␥/NEMO (middle panel), in vitro kinase assays were performed and analyzed following SDS-PAGE and autoradiography. IKK␤ phosphorylation of the IKK␥/NEMO mutant in which serine residues 369 and 375 were changed to alanine was reduced as compared with that seen with wild-type IKK␥/ NEMO (Fig. 5A, lanes 2 and 3, top and middle panels). IKK␤ phosphorylated the IKK␥/NEMO mutant containing substitutions of residues 147, 148, 156, and 158 with alanine to a similar level as seen with wild-type IKK␥/NEMO (Fig. 5A,  lane 4, top and middle panels). Combining these mutations resulted in decreased IKK␥/NEMO phosphorylation by IKK␤ similar to the results seen with the mutation of residues 369 and 375 (Fig. 5A, lane 5, top and middle panels). Thus, IKK␤ phosphorylation of IKK␥/NEMO residues between 144 and 159 was only detected in the absence of the carboxyl terminus of IKK␥/NEMO.
Because serine residues 369 and 375 appeared to be critical sites in IKK␥/NEMO for IKK␤ phosphorylation, alanine substitutions were introduced into each of these two serine residues individually. Each of these epitope-tagged IKK␥/NEMO constructs was transfected with IKK␤ into 293T cells and immunoprecipitated with either Myc or FLAG monoclonal antibodies, and in vitro kinase assays were performed. This analysis revealed that mutation of serine residue 369 resulted in a 60% reduction of phosphorylation by IKK␤ (Fig. 5B, lanes 1-3,  top and middle panels). In contrast, mutation of serine residue 375 resulted in only a 20% decrease in IKK␥/NEMO phosphorylation by IKK␤ (Fig. 5B, lane 4, top panel). When both serine residues 369 and 375 were changed to alanine, there was a 50 -70% decrease in IKK␥/NEMO phosphorylation (Fig. 5B,  lane 5, top and middle panels). Thus, the carboxyl terminus of IKK␥/NEMO is the major site of phosphorylation by IKK␤.
Finally, we addressed whether alterations in IKK␥/NEMO phosphorylation affected its interaction with IKK␤. CMV expression vectors encoding epitope-tagged wild-type and mutant IKK␥/NEMO and either wild-type IKK␤ (Fig. 5C, top panel) or the IKK␤ (S177A/S181A) mutant (Fig. 5C, lower panel) were transfected into 293T cells. Following immunoprecipitation of Myc-tagged IKK␥/NEMO, Western blot analysis was performed with FLAG antibody to detect the epitope-tagged IKK␤. These results indicated that there were similar levels of interaction between wild-type and mutant IKK␥/NEMO with either wild-type or kinase-defective IKK␤ (Fig. 5C). These results suggest that changes in IKK␥/NEMO phosphorylation did not dramatically alter its interaction with IKK␤.
IKK␥/NEMO Phosphorylation Regulates Both IKK␤ and NF-B Activation-Previously, cotransfection assays were utilized to demonstrate that IKK␥/NEMO resulted in the recruitment of IKK␤ into the high molecular weight IKK complex and stimulated the ability of IKK␤ to phosphorylate IB␣ (27,33). It was important to address whether alterations in the phosphorylation of IKK␥/NEMO led to changes in its ability to enhance IKK␤ phosphorylation of IB␣ and/or stimulate NF-B luciferase activity. For these studies, expression vectors encoding IKK␤ alone or together with either wild-type or mutant IKK␥/NEMO were transfected into 293T cells. The FLAGtagged IKK␤ was immunoprecipitated and assayed in in vitro kinase assays with a GST-IB␣ substrate. Wild-type IKK␥/ NEMO markedly enhanced IKK␤ phosphorylation of IB␣ (Fig.  6A, lanes 2 and 9, top panel). Next we addressed whether mutation of threonine residue 356 in IKK␥/NEMO, which did not appear to be a site of IKK␤ phosphorylation, altered the ability of IKK␥/NEMO to stimulate IKK␤. This IKK␥/NEMO mutant resulted in similar levels of IKK␤ activity to that seen with wild-type IKK␥/NEMO (Fig. 6A, lane 3, top panel). Surprisingly, mutation of serine residue 369 resulted in a 3-fold increase in IKK␥/NEMO stimulation of IKK␤ activity as compared with wild-type IKK␥/NEMO (Fig. 6A, lane 4, top panel), whereas mutation of residue 375 resulted in a 2.5-fold increase in IKK␥/NEMO stimulation of IKK␤ activity (Fig. 6, lane 5, top  panel). Mutation of both serine residues 369 and 375 resulted in approximately a 6-fold increase in the ability of IKK␥/NEMO to stimulate IKK␤ activity as compared with wild-type IKK␥/ NEMO (Fig. 6A, lane 6, top panel). Finally, mutation of residues 147, 148, 156, and 158 resulted in a 2-3-fold increase in the ability of IKK␥/NEMO to stimulate IKK␤ activity as compared with wild-type IKK␥/NEMO (Fig. 6A, lane 7, top panel), whereas a combination of these mutations and the carboxylterminal mutations resulted in more than a 6-fold increase in IKK␥/NEMO stimulation of IKK␤ activity (Fig. 6A, lane 8, top  panel). These results were seen in three independent experiments. Western blot analysis indicated similar levels of expression of both IKK␥/NEMO and IKK␤ in these assays (Fig. 6A,  lanes 1-9, middle and lower panels). Similar results with these mutants were seen when IKK␥/NEMO and IKK␤ were immunoprecipitated with Myc antibody directed against IKK␥/ NEMO, and IKK␤ activity was assayed in in vitro kinase assays with a GST-IB␣ substrate (data not shown). Thus, these data suggest that phosphorylation of IKK␥/NEMO likely reduces its ability to stimulate IKK␤ activity.
Next we assayed whether IKK␥/NEMO mutants exhibited changes in their ability to activate an NF-B luciferase reporter construct. Transfection of an NF-B reporter alone or in the presence of CMV expression vectors encoding wild-type or mutant IKK␥/NEMO was performed using IKK␥/NEMO Ϫ/Ϫ mouse embryo fibroblasts (37). An RSV-␤-galactosidase expression vector was included in these assays to control for changes in transfection efficiency. Following transfection, the cells were treated with TNF␣ for 6 h prior to harvesting and assays of luciferase activity. These studies demonstrated that wild-type IKK␥/NEMO stimulated NF-B reporter activity ϳ2.5-fold (Fig. 6B, lanes 1 and 2), whereas an IKK␥/NEMO construct that contained a mutation of serine residue 369 to alanine resulted in a 4.5-fold increase in NF-B reporter activity (Fig.  6B, lanes 1 and 4). Mutation of serine residues 369 and 375 in IKK␥/NEMO to alanine resulted in a 6-fold increase in NF-B luciferase activity (Fig. 6B, lane 6). Mutation of residues 147, 148, 156, and 158 either alone or in combination with mutation of residues 369 and 375 also resulted in the increased ability of IKK␥/NEMO to stimulate NF-B luciferase activity as compared with wild-type IKK␥/NEMO (Fig. 6B, lanes 7 and 8). It is interesting to note that mutation of all the potential IKK␤ phosphorylation sites in IKK␥/NEMO did not result in further enhancement of NF-B luciferase activity as compared with mutation of either residues 369 and 375 or 147, 148, 156, and 158 alone. Western blot analysis of these lysates revealed similar levels of IKK␥/NEMO expression (Fig. 6B, lower panel). These results, which were repeated in three independent experiments, indicated that mutation of serine residues 369 and 375 increased the ability of IKK␥/NEMO to stimulate NF-B activity. DISCUSSION In this study, the role of IKK␤-mediated phosphorylation of IKK␥/NEMO on regulating its function was explored. IKK␤ was found to phosphorylate IKK␥/NEMO predominantly on serine residue 369 in the carboxyl terminus of IKK␥/NEMO using both in vitro and in vivo assays. Mutation of serine residues 369 and 375 increased the ability of IKK␥/NEMO to stimulate IKK␤ phosphorylation of IB␣ and activate an NF-B reporter construct. Mutation of the putative phosphorylation sites at residues 147, 148, 156, and 158 did not markedly alter IKK␥/NEMO phosphorylation unless the carboxyl terminus was eliminated. These results suggest that phosphorylation of IKK␥/NEMO by IKK␤ and perhaps other kinases may be important in modulating its function. IKK␥/NEMO is required for NF-B activation in response to a variety of different signals (13)(14)(15)(16). Thus, the mechanisms by which IKK␥/NEMO leads to activation of the NF-B pathway has been the subject of intense investigation. The ability of IKK␥/NEMO to regulate the NF-B pathway is likely mediated via its differential interactions with factors that regulate the NF-B pathway. Transient expression assays of IKK␥/NEMO indicate that IKK␥/NEMO results in the recruitment of IKK␤ into the high molecular weight IKK complex and stimulate its kinase activity (27,33). Furthermore, IKK␥/NEMO can enhance the association of the IB proteins with the IKK complex to facilitate its phosphorylation and subsequent degradation (33). IKK␥/NEMO has also been demonstrated to be able to increase IKK activity in a reconstituted yeast system to result in the assembly of the high molecular weight IKK complex (44). IKK␥/NEMO can bind to other proteins including the adaptor protein RIP at early times following TNF␣ binding to the TNF receptor (13,30) and the NF-B inhibitory protein A20 at later times post-TNF␣ treatment to reduce TNF␣-mediated effects on the NF-B pathway (30). These findings suggest that the differential association of IKK␥/NEMO with cellular regulatory proteins is important for modulating TNF␣-induced activation of the NF-B pathway. Thus IKK␥/NEMO interacts with a variety of cellular proteins that are likely critical in modulating IKK activity.
Post-translational modifications such as phosphorylation are important in altering protein function (45). In this study, we addressed the ability of IKK␤ to phosphorylate IKK␥/NEMO and then determined whether IKK␤-mediated phosphorylation of IKK␥/NEMO altered its ability to stimulate IKK␤ activity. First, we demonstrated that endogenous IKK␥/NEMO was rapidly phosphorylated following treatment of cells with activators of the NF-B pathway including TNF␣ and IL-1. It is interesting to note that the kinetics of IKK␥/NEMO phosphorylation correlate with the increases in IKK␤ activity in response to these agents. These results in conjunction with data obtained from transient expression assays and in vitro kinase assays with IKK␤ and IKK␥/NEMO suggest that IKK␤ can directly phosphorylate IKK␥/NEMO. The level of phosphorylated IKK␥/NEMO increased rapidly following cytokine stimulation. The rapid decrease in both IKK␥/NEMO phosphorylation and IKK␤ activity following cytokine treatment suggests that both of these proteins could potentially be targets of phosphatases such as protein phosphatase 2A which reduces IKK activity following cytokine treatment (8). The transient phosphorylation of IKK␥/NEMO following cytokine stimulation may result in its differential interactions with a variety of cellular proteins including IKK␤, IB, A20, and RIP (9, 13-16, 30, 33, 36).
Our studies suggest that at least two regions of IKK␥/NEMO are phosphorylated by IKK␤. However, additional sites of IKK␤ phosphorylation of IKK␥/NEMO are indeed likely. Other kinases such as protein kinase C␣ have also been demonstrated to phosphorylate serine residues in IKK␥/NEMO at positions that differ from those identified using IKK␤ (43). In contrast to the results seen with mutation of residues in IKK␥/NEMO that are targets for IKK␤-mediated phosphorylation, mutation of serine residues 85 and 141 that are phosphorylated by protein kinase C␣ result in the reduced ability of IKK␥/NEMO to stimulate IKK␤ phosphorylation of IB␣ (43). Given the specific association of IKK␥/NEMO and IKK␤ and the temporal relationship between the increased IKK␤ activity and IKK␥/ NEMO phosphorylation, our data and that of a previous study (40) suggest that IKK␤ is an important, although not likely the only, kinase involved in phosphorylating IKK␥/NEMO.
The major site of IKK␤ phosphorylation in the carboxyl terminus of IKK␥/NEMO is serine residue 369, although serine residue 375 may also potentially be phosphorylated by IKK␤. IKK␤ was also found to phosphorylate IKK␥/NEMO in a region between amino acids 137 and 180 at potential sites including residues 147, 148, 156, and 158. It is likely that IKK␤ predominantly phosphorylates the carboxyl terminus of IKK␥/NEMO when not bound to other proteins. However, the carboxyl terminus of IKK␥/NEMO is capable of binding to a variety of proteins such as A20 and RIP (30), and it is possible that this binding may lead to conformational changes that make additional domains in IKK␥/NEMO more accessible to phosphorylation by IKK␤. It is also possible that phosphorylation of IKK␥/ NEMO may be temporally regulated with the carboxylterminal residues initially phosphorylated by IKK␤ and then either IKK␤ or other kinases phosphorylating additional residues in IKK␥/NEMO. The results of this study suggest that IKK␤-mediated phosphorylation of IKK␥/NEMO leads to its decreased ability to stimulate IKK␤ and activate an NF-B reporter construct. Initially, the nonphosphorylated form of IKK␥/NEMO may preferentially bind to IKK␤ to stimulate its ability to phosphorylate IB␣. Subsequent phosphorylation of IKK␥/NEMO by IKK␤ may perhaps alter its conformation, its ability to oligomerize (46), or its interaction with other cellular proteins. This can potentially decrease IKK activity and/or lead to IKK␥/ NEMO association with other regulators of the NF-B pathway. Changes in IKK␥/NEMO phosphorylation may in part be responsible for its ability to both activate and inhibit NF-Bdependent gene expression under different conditions (30). For example, mutagenesis of IKK␤ indicated that in certain instances reduction in IKK␤ binding to IKK␥/NEMO can result in increased IKK␤-mediated activation of NF-B reporter constructs (28). One possible explanation for this finding is that the decreased interaction of IKK␤ and IKK␥/NEMO can lead to reduced IKK␥/NEMO phosphorylation and thus decreased activation of IKK␤. Finally, we noted that although the majority of IKK␥/NEMO is located in the cytoplasm, a small portion of IKK␥/NEMO is also present in the nucleus. Thus phosphorylation of IKK␥/NEMO may also serve to target the nuclear import of IKK␥/NEMO. Further studies will be needed to address the kinetics of phosphorylation of IKK␥/NEMO and identify additional kinases and/or phosphatases that are able to modulate its phosphorylation state. The slightly different effects of the IKK␥/NEMO phosphorylation site mutants on activating IKK␤ and stimulating NF-B luciferase activity suggest that these different IKK␥/NEMO phosphorylation sites may in fact play multiple roles in IKK␥/NEMO regulation of the NF-B pathway. In summary, the results presented in this study suggest that phosphorylation of IKK␥/NEMO by IKK␤ likely plays an important role in IKK␥/NEMO regulation of the NF-B pathway.