IKKγ/NEMO Facilitates the Recruitment of the IκB Proteins into the IκB Kinase Complex

IKKγ/NEMO is an essential regulatory component of the IκB kinase complex that is required for NF-κB activation in response to various stimuli including tumor necrosis factor-α and interleukin-1β. To investigate the mechanism by which IKKγ/NEMO regulates the IKK complex, we examined the ability of IKKγ/NEMO to recruit the IκB proteins into this complex. IKKγ/NEMO binding to wild-type, but not to a kinase-deficient IKKβ protein, facilitated the association of IκBα and IκBβ with the high molecular weight IKK complex. Following tumor necrosis factor-α treatment of HeLa cells, the majority of the phosphorylated form of endogenous IκBα was associated with the high molecular weight IKK complex in HeLa cells and parental mouse embryo fibroblasts but not in IKKγ/NEMO-deficient cells. Finally, we demonstrate that IKKγ/NEMO facilitates the association of the IκB proteins and IKKβ and leads to increases in IKKβ kinase activity. These results suggest that an important function of IKKγ/NEMO is to facilitate the association of both IKKβ and IκB in the high molecular weight IKK complex to increase IκB phosphorylation.

The NF-B proteins are critical for activating the expression of cellular genes that are involved in the control of the immune and inflammatory response and in protecting cells from apoptosis in response to a variety of stress stimuli (1)(2)(3)(4). NF-B is sequestered in the cytoplasm in most cells, where it is bound to a family of inhibitory proteins known as IB (2,5,6). A variety of stimuli including the cytokines TNF␣ 1 and interleukin-1, double-stranded RNA, and the viral transactivator Tax activate the NF-B pathway (4,(7)(8)(9)(10)(11). These stimuli increase the activity of two related kinases, IKK␣ and IKK␤, to result in the phosphorylation of the IB proteins (9,(12)(13)(14)(15)(16). A variety of studies using IKK␣ and IKK␤ knock-out mice indicate that IKK␤ is critical for NF-B activation in response to cytokine treatment, whereas IKK␣ is not required for this function (17)(18)(19)(20)(21)(22). The IB␣ protein is phosphorylated on serine residues 32 and 36, while IB␤ is phosphorylated on serine residues 19 and 23, and this leads to their ubiquitination and degradation by the proteasome (10,(23)(24)(25)(26)(27)(28)(29)(30)(31)(32). IB mutants in which these serine residues are changed to alanine are resistant to proteasome-mediated degradation and thus prevent the nuclear translocation of the NF-B proteins (33).
IKK␥/NEMO was initially identified in a genetic complementation assay as a factor that could restore NF-␤ activation in cells that were resistant to a variety of stimuli that normally induce the NF-B pathway (34). IKK␥/NEMO was also identified independently in biochemical studies as an essential component of the high molecular weight IKK complex (35,36). Finally, this factor was characterized as a factor known as FIP-3 that bound to the adenovirus E3 protein and could inhibit TNF␣-induced apoptosis (37). IKK␥/NEMO in conjunction with IKK␣ and IKK␤ is a component of the high molecular weight IKK complex, which migrates between 600 and 900 kDa following gel filtration chromatography (12, 14, 29, 34 -41). Biochemical fractionation and coimmunoprecipitation studies demonstrate that IKK␣, IKK␤, and IKK␥/NEMO interact in this IKK complex (35,36,40). Cells that do not express IKK␥/ NEMO are unable to assemble the high molecular weight IKK complex and increase IKK activity in response to agents that stimulate the NF-B pathway (35,36). Although IKK␥/NEMO itself does not have kinase activity, it is essential for NF-B activation (34 -36). The mechanism by which IKK␥/NEMO activates the NF-B pathway has been the subject of intense investigation.
Mutagenesis of IKK␥/NEMO has been performed in an attempt to define important functional domains (35,36,42,43). IKK␥/NEMO has a molecular mass of 48 kDa and contains a leucine zipper and two coiled-coil motifs. Residues in the aminoterminal 100 amino acids of this protein are critical for interactions with IKK␤ (43). IKK␤ preferentially associates with IKK␥/NEMO (34,36), although IKK␣ has also been shown to directly associate with IKK␥/NEMO (22,40). The coiled-coil domains in IKK␥/NEMO mediate its oligomerization, which is critical for activating IKK kinase activity (44), while its carboxyl terminus is involved in the recruitment of upstream kinases, which are critical for activating IKK (45). For example, a kinase known as RIP, which is recruited to the TNF receptor following TNF␣ treatment of cells, binds to IKK␥/NEMO and leads to the subsequent association of IKK␣ and IKK␤ (37,45). In contrast, the association of the A20 protein with IKK␥/ NEMO decreases TNF␣-mediated activation of the NF-B pathway (45). The viral transactivator Tax has also been shown to bind to IKK␥/NEMO and stimulate IKK kinase activity (42,46,47) as has the cellular protein CIKS (48). These results indicate that IKK␥/NEMO can interact with a variety of different regulatory proteins that are important in the activation of the NF-B pathway in response to various stimuli.
Genetic studies have also been utilized to study the role of IKK␥/NEMO in regulating the NF-B pathway (49 -51). Disruption of a single copy of the IKK␥/NEMO gene, which is located on the X chromosome, results in the death of male mice in utero, while female mice develop granulocytic infiltration and both hyperproliferation and increased apoptosis of keratinocytes (49,51). The homozygous deletion of IKK␥/NEMO results in embryonic lethality in both male and female mice due to TNF␣-induced hepatic apoptosis (49,50). Fibroblasts isolated from these mice are defective in activating the NF-B pathway in response to a variety of stimulators of this pathway. In humans, mutation of a single copy of the IKK␥/NEMO gene is associated with a syndrome known as incontinentia pigmenti, an X-linked defect that results in lethality in males and a granulocytic infiltration of the skin in females (52).
Recently, another syndrome due to mutations in the putative zinc finger domain in the C terminus of IKK␥/NEMO has been described (53)(54)(55). These mutations, which impair but do not eliminate NF-B function, result in an X-linked immunodeficiency syndrome characterized by hyper-IgM production and hypohydrotic ectodermal dysplasia. Thus, both biochemical and genetic studies indicate a critical role for IKK␥/NEMO in regulating NF-B activation.
Although IKK␥/NEMO is critical for activation of the NF-B pathway, the exact mechanisms involved in its regulation remain to be elucidated. Previously, we demonstrated that interactions between IKK␥/NEMO and IKK␤ are critical for the formation of the high molecular weight IKK complex (41). In this study, we addressed the role of IKK␥/NEMO in facilitating interactions between IB and IKK␤. IKK␥/NEMO was critical for the association of IB and IKK␤ with the high molecular weight IKK complex. Furthermore, we found that IB in the high molecular weight complex was preferentially phosphorylated. Finally, we demonstrated that IKK␥/NEMO enhanced the association of the IB␣ and IKK␤ and increased IKK␤ kinase activity. Thus, the ability of IKK␥/NEMO to both stimulate the association of the IB proteins with IKK␤ and increase IKK␤ kinase activity is likely important in activating the NF-B pathway.
Transfection and Cellular Fractionation-COS, mouse embryo fibroblasts (MEFs) (a gift of Xiaodong Wong), and IKK␥/NEMO knock-out cells (a gift of Michael Karin) were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (49). HeLa cells were maintained in Iscove's modified Dulbecco's medium and supplemented with the same components as above. Transfections were carried out using Fugene-6 (Roche Molecular Biochemicals) as described by the manufacturer. For a typical cell fractionation experiment, COS cells cultured overnight in 100-mm plates were transfected with 2.0 g of each DNA construct. Cytoplasmic extracts were prepared from either 10 7 transfected COS cells or nontransfected cells including 10 8 HeLa, 2.5 ϫ 10 8 MEFs, and 2.5 ϫ 10 8 IKK␥ knock-out as detailed previously (41). TNF␣ (20 ng/ml) and MG-132 (25 M) were purchased from Roche Molecular Biochemicals and Calbiochem, respectively.
Protein Fractionation and Immunoblotting-For protein fractionation, the cytoplasmic extracts of cells were prepared according to Dignam (58) with slight modifications. Cells, washed twice with cold phosphate-buffered saline, were harvested by scraping from the culture dishes and precipitated by centrifugation at 1500 rpm for 10 min. Cell pellets were resuspended in buffer A (10 mM Hepes (pH 7.9), 1 mM EDTA, 10 mM KCl, 1 mM dithiothreitol) supplemented with phosphatase inhibitors (50 mM NaF, 50 mM glycerol phosphate, 1 mM sodium orthovanadate, 0.1 M okadaic acid) and proteinase inhibitors (Roche Molecular Biochemicals). After incubation on ice for 15 min, the cells were disrupted with 20 strokes through a 25-gauge needle and centrifuged at 14,000 rpm for 15 min. The supernatants were mixed with 0.11 volume of buffer B (0.3 M Hepes (pH 7.9), 30 mM MgCl 2 , 1.4 M KCl) and centrifuged again at 100,000 ϫ g for 60 min. The supernatants were assayed for protein concentration according to the Bradford method, and a total of 2 mg of protein was subjected to Superdex-200 column (Amersham Pharmacia Biotech) chromatography in buffer D (20 mM Hepes (pH 7.9), 0.1 M KCl, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 20% glycerol, and 0.2 mM EGTA), and 1-ml fractions were collected. Protein markers (Sigma) used for the Superdex-200 column were bovine thyroglobulin (669 kDa), horse spleen apoferritin (443 kDa), ␤-amylase (200 kDa), bovine serum albumin (66 kDa), and carbonic anhydrase (29 kDa). Western blotting was done with 30 g of protein obtained from each of the column fractions including monoclonal antibodies directed against the HA epitope (12CA5), the FLAG epitope (M2; Sigma), and the Myc epitope (Roche Molecular Biochemicals) and polyclonal antibodies directed against IB␣ (Santa Cruz Biotechnology, Inc., Santa Cruz, CA; Sc-371), phospho-IB␣ (Ser 32 ) antibody (Cell Signaling Technology), IKK␤ (Santa Cruz Biotechnology; Sc-7607), and IKK␥ (Santa Cruz Biotechnology; Sc-8330). The antibodies used are specified in the figure legends.
Immunoprecipitation and Immunoblotting-Cytoplasmic lysates of COS cells containing 300 g of protein were incubated with either 2.5 g of the 12CA5 monoclonal antibody directed against the HA epitope, the M2 monoclonal antibody directed against the FLAG epitope, or the Myc monoclonal antibody. Following this, 20 l of protein A-agarose beads were added and mixed for 1 h at 4°C, and the immunoprecipitates were washed three times with modified buffer D (20 mM Hepes (pH 7.9), 100 mM KCl, 200 mM NaCl, 0.2 mM EDTA, 20% glycerol, 0.5 mM dithiotheitol, 0.06% Nonidet P-40) containing proteinase inhibitors (Roche Molecular Biochemicals). Electrophoresis on a 6% SDS-polyacrylamide gel was performed, and the gel was subjected to immunoblotting with specific antibodies and chemiluminescence reagents (Amersham Pharmacia Biotech).

IKK␥/NEMO Stimulates IKK␤ Phosphorylation of IB␣ and
IB␤-To better understand the mechanism by which IKK␥/ NEMO stimulates IKK␤ kinase activity, we transfected expression vectors encoding wild-type epitope-tagged IKK␤ and a variety of epitope-tagged IKK␥/NEMO constructs into COS cells (Fig. 1A). Following immunoprecipitation of IKK␤, we assayed its ability to phosphorylate GST fusion proteins with either IB␣ and IB␤ in in vitro kinase assays. Wild-type IKK␥/NEMO markedly increased IKK␤ kinase activity for the GST-IB␣ substrate, as did the C-terminal IKK␥/NEMO deletion mutant (Fig. 1B, top panel). In contrast, the N-terminal IKK␥/NEMO deletion mutant did not increase IKK␤ kinase activity (Fig. 1B, top panel). The same pattern of results was seen with these extracts when IKK␤ was assayed with the GST-IB␤ substrate (Fig. 1B, second panel). Western blot analysis confirmed the expression of the epitope-tagged IKK␤ and IKK␥/NEMO proteins (Fig. 1B, third and fourth panels). These results indicate that the N terminus of IKK␥/NEMO is critical for its ability to stimulate IKK␤ phosphorylation of both IB␣ and IB␤.
IKK␥/NEMO Increases IB␣ and IB␤ Association with the High Molecular Weight IKK Complex-One potential step that may be regulated by IKK␥/NEMO to activate the NF-B pathway is to facilitate the association of IB␣ and IB␤ with IKK␤ in the high molecular weight IKK complex. This may lead to increased IKK␤ phosphorylation of IB␣ and IB␤ and thus result in enhanced IB degradation and NF-B nuclear translocation. Previous studies indicate that IKK␣, IKK␤, and IKK␥/ NEMO interact in the high molecular weight IKK complex (35,36,40). Our previous studies indicate that the high molecular weight IKK complex can be formed in COS cell extracts prepared following transfection of expression vectors encoding epitope-tagged IKK␣, IKK␤, and IKK␥/NEMO (41). This high molecular weight complex could also be formed by transfecting IKK␤ and IKK␥ without IKK␣ and contained high levels of IKK␤ kinase activity.
In order to test the possibility that IKK␥/NEMO may be involved in recruiting IB into the high molecular weight IKK complex, expression vectors encoding FLAG-tagged IB␣, Myctagged IKK␤, and either wild-type Myc-tagged IKK␥/NEMO or ⌬N or ⌬C mutants were transfected in various combinations into COS cells (Fig. 2). S100 extracts were prepared from the transfected COS cells and subjected to chromatography on a Superdex-200 column in order to analyze the migration of these proteins based on their molecular mass. Western blot analysis indicated that when the FLAG-tagged IB␣ protein was transfected alone, it peaked at a molecular mass of ϳ200 kDa ( Fig.  2A). Similar migration of FLAG-tagged IB␣ protein was seen when it was expressed with Myc-tagged IKK␥/NEMO (Fig. 2B). When Myc-tagged IKK␤ was transfected with the FLAGtagged IB␣, the majority of the IB␣ still migrated at 200 kDa although a portion of IB␣ migrated with IKK␤ between 600 and 700 kDa (Fig. 2C). However, when the epitope-tagged IKK␤, IKK␥/NEMO, and IB proteins were expressed together, a larger portion of the IB␣ was found to comigrate with IKK␤ and IKK␥ in the high molecular weight fractions (Fig. 2D). Identical results were noted when an epitope-tagged IKK␣ was included with IKK␤, IKK␥/NEMO, and IB␣ in these transfections (data not shown). When the C-terminal truncated form of IKK␥/NEMO was cotransfected with the FLAG-tagged IB␣ and the Myc-tagged IKK␤, there was less FLAG-tagged IB␣ present in the high molecular weight fractions (Fig. 2E). Transfection of the N-terminal deleted form of IKK␥/NEMO with the IB␣ and IKK␤ resulted in several effects. First, the majority of IKK␤ migrated at a lower molecular weight than in the presence of wild-type IKK␥/NEMO (Fig. 2F). Furthermore, the majority of the IB␣ protein was found to migrate at about 200 kDa, which was similar to the results seen when IB␣ was transfected alone. These results suggest that IKK␥/NEMO is involved in the association of both IKK␤ and IB␣ in a high molecular weight IKK complex.
Next we assayed the effects of IKK␥/NEMO expression on IB␤ association with the high molecular weight IKK complex. COS cells were transfected with various combinations of FLAG-tagged IB␤, Myc-tagged IKK␤, and wild-type or mutant Myc-tagged IKK␥/NEMO. Again S100 extracts were prepared, and following Superdex 200 chromatography, Western blot analysis was performed on the different column fractions using epitope-specific monoclonal antibodies. IB␤ was found to migrate in a very broad peak, which ranged from 200 to 500 kDa (Fig. 3A). The expression of either IKK␥/NEMO or IKK␤ did not significantly alter IB␤ migration (Fig. 3, B and C). In contrast, the expression of wild-type IKK␥/NEMO with IKK␤ and IB␤ resulted in the peak of both IKK␤ and IB␤ shifting to 600 -700 kDa (Fig. 3D). The C-terminal deleted form of IKK␥/NEMO resulted in decreased IB␤ present in the high molecular weight IKK fraction as compared with that seen with wild-type IKK␥/NEMO (Fig. 3E), while the N-terminal deleted form of IKK␥/NEMO resulted in the migration of the majority of IKK␤ and IB␤ in lower molecular weight fractions (Fig. 3F). These results suggest that IKK␥/NEMO is important not only to facilitate the recruitment of IKK␤ into a high molecular weight IKK complex but also to facilitate the association of IB␣ and IB␤.
IKK␤ Kinase Activity Is Necessary for IB␣ Recruitment by IKK␥/NEMO-Recent observations suggest that the incorporation of IKK␤ into high molecular weight IKK complex by IKK␥/NEMO is not dependent on the kinase activity of IKK␤ (41). To determine whether IKK␥/NEMO required an active IKK␤ kinase to facilitate IB␣ association, FLAG-tagged IB␣ and the FLAG-tagged catalytically defective IKK␤ mutant, IKK␤ (K/M), were cotransfected into COS cells in the presence or absence of Myc-tagged IKK␥/NEMO. Similar to the results observed with extracts containing wild-type IKK␤ and IB␣ in the absence of IKK␥/NEMO (Fig. 2C), the majority of the IB␣ was present in lower molecular weight fractions migrating at 200 kDa when IKK␤ (K/M) was coexpressed with IB␣ ( Fig.  4A). However, when IKK␤(K/M) was expressed together with the epitope-tagged IKK␥/NEMO and IB␣ proteins, the majority of IKK␤ (K/M) but not IB␣ was associated with IKK␥/ NEMO in the high molecular weight fractions (Fig. 4B). These results suggest that an active IKK␤ kinase is necessary for the chromatographic shift of IB␣ by IKK␥/NEMO. IB␣ Association with IKK␤ and IKK␥/NEMO Is Independent of Its Phosphorylation State-The previous data indicate that IKK␥/NEMO facilitates the association of both IKK␤ and IB␣ in a high molecular weight IKK complex and that IKK␥/ NEMO requires intact IKK␤ kinase activity to result in IB␣ association with this complex. To determine whether IB␣ phosphorylation was critical for its association with IKK␥/ NEMO and IKK␤ in the high molecular weight fractions, a degradation-resistant IB␣ mutant was transfected with wildtype IKK␤ in either the presence or absence of IKK␥/NEMO. In this mutant, IB␣ (SS/AA), serine residues 32 and 36, which are sites for IKK␤ phosphorylation, were changed to alanine residues in order to prevent its phosphorylation and subse-quent degradation in response to activators of the NF-B pathway (33). When IB␣(SS/AA) was expressed with IKK␤, IB␣ (SS/AA) was predominantly present at a molecular mass of ϳ200 kDa with a portion of this protein also present at 600 -700 kDa (Fig. 5C). The fact that a portion of this IB␣ protein migrated at 600 -700 kDa in the absence of transfected IKK␥/ NEMO probably reflects the increased protein stability of this mutant as compared with wild-type IB␣ and its binding to low levels of endogenous IKK␥/NEMO. The expression of IKK␥/ NEMO with IKK␤ and IB␣ (SS/AA) resulted in the peak of IB␣(SS/AA) shifting to between 600 and 700 kDa (Fig. 5D). This result is similar to the effects of IKK␥/NEMO that were seen when it was expressed with IKK␤ and wild-type IB␣ (Fig. 5, A and B). Taken together, these results suggest that IB␣ association in the high molecular weight complex is not dependent on its N-terminal phosphorylation, but it does require IKK␥/NEMO and active IKK␤.
IKK␥/NEMO Facilitates IKK␤ Association with IB␣ and IB␤-In order to further explore the role of IKK␥/NEMO on facilitating interactions between IKK␤ and IB␣, expression vectors encoding FLAG-tagged IB␣, HA-tagged IKK␤, and Myc-IKK␥/NEMO were transfected in various combinations into COS cells. Following immunoprecipitation of either HA-IKK␤ or Myc-IKK␥/NEMO, Western blot analysis with epitopespecific antibodies directed against either IB␣, IKK␥, or IKK␤ was performed (Fig. 6A). Immunoprecipitation of extracts transfected with FLAG-IB␣ and Myc-IKK␥/NEMO with the 12CA5 antibody directed against the HA-epitope indicated that there was little nonspecific association of FLAG-IB␣ (Fig. 6A,  lane 2, top panel). When FLAG-IB␣ and HA-IKK␤ were cotransfected, immunoprecipitation of the HA-IKK␤ resulted in the association of a small amount of FLAG-IB␣ (Fig. 6A, lane  3, top panel). However, transfection of Myc-IKK␥/NEMO, FLAG-IB␣, and HA-IKK␤ resulted in the increased association of FLAG-IB␣ with the immunoprecipitated HA-tagged IKK␤ (Fig. 6A, lane 4, top panel). Transfection of a C-terminal truncated form of Myc-IKK␥ also resulted in increased association of FLAG-IB␣ with the immunoprecipitated HA-IKK␤ (Fig. 6A, lane 5, top panel). Transfection of an N-terminal truncated form of Myc-IKK␥ did not significantly increase FLAG-IB␣ association with HA-IKK␤ (Fig. 6A, lane 6, top  panel). Both the wild-type and the C-terminal deletion of IKK␥/ NEMO, but not the N-terminal deletion, coimmunoprecipitated with IKK␤ and IB␣ (Fig. 6A, second panel, lanes 4 -6). These results suggest that IKK␤, IKK␥/NEMO, and IB␣ form a complex.
The Myc antibody was next used to immunoprecipitate the Myc-IKK␥/NEMO protein from these extracts (Fig. 6A, third  panel). IKK␥/NEMO alone did not result in the association of IB␣ (Fig. 6A, lane 2, third panel). However, the presence of either the wild-type or the C-terminal truncation of IKK␥/ NEMO, but not the N-terminal truncation, resulted in the association of both IB␣ and HA-IKK␤ (Fig. 6A, lanes 4 -6,  third and fourth panels). Western blot analysis demonstrated the expression of each of these proteins following transfection (Fig. 6A, bottom three panels). These results also indicate that the expression of IKK␥/NEMO with IKK␤ increases the association of IB␣.
Similar transfection assays were performed with FLAG-IB␤, HA-IKK␤, and Myc-IKK␥/NEMO in order to determine the role of IKK␥/NEMO and IKK␤ in the association with IB␤ (Fig. 6B). When antibody to the HA epitope was used in the immunoprecipitation assays, there was little nonspecific association of IB␤ with IKK␥/NEMO in the absence of HA-tagged IKK␤ (Fig. 6B, lane 2, top panel). FLAG-IB␤ was able to associate with HA-IKK␤ alone (Fig. 6B, lane 3, top panel), and this association was increased following the expression of Myc-IKK␥/NEMO (Fig. 6B, lane 4, top panel). The expression of either the N-or C-terminal IKK␥/NEMO constructs resulted in less association of FLAG-IB␤ with HA-IKK␤ than did wild- type IKK␥/NEMO (Fig. 6B, lanes 5 and 6, top panel). The wild-type and C-terminal deletion of IKK␥/NEMO, but not the N-terminal deletion, also coimmunoprecipitated with IKK␤ and IB␤ (Fig. 6B, lanes 4 -6, second panel).
Next, immunoprecipitation of Myc-IKK␥/NEMO was performed. In extracts containing wild-type Myc-IKK␥/NEMO and FLAG-IB␤, immunoprecipitation revealed no association of IB␤ with IKK␥/NEMO (Fig. 6B, lane 3, third panel). In the presence of wild-type Myc-IKK␥/NEMO, HA-IKK␤, and FLAG-IB␤, both IKK␤ and IB␤ were immunoprecipitated with Myc-IKK␥ antibody (Fig. 6B, lane 4, third panel). The expression of either the C-terminal or N-terminal IKK␥/NEMO deletions with IKK␤ and IB␤ resulted in decreased association of IB␤ (Fig. 6B, lanes 5 and 6, third panel). IKK␤ associated with both the wild-type and the C-terminal truncation of IKK␥/ NEMO but not the N-terminal truncation (Fig. 6B, lanes 4 -6,  fourth panel). Western blot analysis of these extracts demonstrated the expression of these proteins (Fig. 6B, bottom three panels). These results indicate that IKK␥/NEMO increases the association of both IB␣ and IB␤ with IKK␤.
TNF␣ Treatment Increases the Amount of Phosphorylated IB␣ in the IKK Complex-The previous studies were performed following transfection of wild-type and mutant IB␣, IKK␤, and IKK␥/NEMO proteins to characterize the roles of each of these proteins in the assembly of the IKK complex. In order to further address the relevance of these observations, we next characterized the components of the endogenous IKK complex following chromatographic fractionation of S100 extracts prepared from untreated and TNF␣-treated HeLa cells (Fig. 7). Western blot analysis indicated that the majority of the endogenous IKK␤ and IKK␥/NEMO proteins were present in a complex migrating between 600 and 700 kDa in extracts prepared from both untreated and TNF␣-treated HeLa cells (Fig. 7A). In addition to this peak, a smaller amount of IKK␥/NEMO was also detected in a lower molecular weight complex migrating between 200 and 300 kDa (Fig. 7A). Quantitation of these gels by densitometry indicated that more than 50% of the IKK␤ and IKK␥/NEMO proteins were present in column fractions 8 and 9. In extracts prepared from unstimulated HeLa cells, the peak of endogenous IB␣ was found to migrate at ϳ500 kDa. However, IB␣ was not detected in extracts prepared from TNF␣-stimulated cells using either a rabbit polyclonal IB␣ antibody or a phosphospecific IB␣ antibody that recognized IB␣ phosphorylated on serine residue 32 (Fig. 7A). IKK kinase activity was assayed using a GST-IB␣ substrate and was found to be maximal in column fractions 8 and 9 that contained the majority of the IKK␤ and IKK␥/NEMO proteins (Fig. 7A).
In an attempt to analyze the presence of the phosphorylated form of IB␣ following TNF␣ stimulation, the proteasome inhibitor MG-132 was added to HeLa cells for 1 h prior to treatment with TNF␣ in order to prevent IB␣ degradation by the proteasome (Fig. 7B). The addition of MG-132 did not change the distribution of IKK␤ and IKK␥/NEMO with more than 60% of these proteins present in column fractions 8 and 9 in extracts prepared from untreated and TNF␣-treated HeLa cells (Fig.  7B). However, in the presence of MG-132 there was a 4-fold increase in the amount of phosphorylated IB␣ in TNF␣treated HeLa extracts. Nearly 50% of this phosphorylated form of endogenous IB␣ was detected in column fractions 8 and 9 versus only 30% in extracts prepared in the absence of TNF␣ (Fig. 7B). The presence of phosphorylated IB␣ correlated with column fractions that contained increased IKK activity, suggesting that the presence of IKK␤, IKK␥/NEMO, and IB␣ in the high molecular weight complex correlated with TNF␣induced increases in IB␣ phosphorylation.
Similar experiments were also performed to determine the distribution of IKK␤, IKK␥/NEMO, and IB␣ in extracts prepared from untreated and TNF␣-treated parental mouse embryo fibroblasts (MEF) and IKK␥/NEMO-deficient cells (Fig. 8). As expected, based on our previous results, there was a 4-fold increase in the amount of phosphorylated IB present in extracts prepared from TNF␣-treated MEF cells as compared with untreated cells, and this correlated with the peak of IKK activity (Fig. 8A). More than 55% of the phosphorylated IB␣ was detected in column fractions 8 and 9 in extracts prepared from TNF␣-treated MEF cells versus only 30% in extracts prepared from untreated MEF cells. There was no change in the distribution of either IKK␤ or IKK␥/NEMO following TNF␣-treatment (Fig. 8A). Interestingly, in extracts prepared from MEF cells, the majority of IKK␤ was found to migrate at ϳ200 kDa, while only a small portion of IKK␤ was detected at 600 -700 kDa. Furthermore, IKK␤ migrated in this high mo- -transfection, the cells were harvested, and equal amounts of protein from the S100 extracts were immunoprecipitated with antibodies directed against either the HA (IKK␤) (top two panels) or Myc (IKK␥) epitopes (middle two panels), and the amounts of the associated proteins were determined by immunoblotting. The expression level of IB␣, IKK␤, and IKK␥/NEMO in the lysates before immunoprecipitation was determined by immunoblotting with antibodies directed against FLAG, HA, or Myc, respectively (lower three panels). lecular fraction at a slightly higher molecular weight than that found in lower molecular weight fractions and might correspond to a phosphorylated form of IKK (Fig. 8A). These results suggest that the presence of IKK␤, IKK␥/NEMO, and IB␣ in the high molecular weight complex leads to TNF␣-mediated increases in IKK␤ kinase activity and IB␣ phosphorylation. In extracts prepared from IKK␥/NEMO-deficient cells (34,49), endogenous IKK␥/NEMO was not detected, and neither IB␣ nor IKK␤ was detected in the high molecular weight fractions (Fig. 8B). These results support a role for IKK␥/NEMO in facilitating the association of IB␣ and IKK␤ in high molecular weight fractions to result in increased IKK␤ phosphorylation of IB␣.
Increasing Amounts of IKK␥/NEMO Facilitate IKK␤ Association with IB␣ and Increased IKK␤ Kinase Activity-The results presented suggested that IKK␥/NEMO facilitates the association of IKK␤ and IB␣ and that this process was important in the ability of IKK␥/NEMO to stimulate IKK␤ phosphorylation of IB␣. Next we determined whether the transfection of increasing amounts of IKK␥/NEMO would increase the association of IB␣ and IKK␤ and stimulate IKK␤ kinase activity. Expression vectors encoding Myc-tagged IKK␤ and a FLAG-tagged IB␣ (SS/AA) mutant were transfected into COS cells with increasing concentrations of an HA-tagged IKK␥/NEMO expression vector. This IB␣ mutant was utilized in these studies in order to quantitate IB␣ association in the absence of potential IKK␤-mediated degradation. S100 extracts were prepared, and a monoclonal antibody directed against the Myc epitope was used to immunoprecipitate IKK␤ followed by Western blot analysis with the M2 monoclonal antibody directed against the FLAG-tagged IB␣ mutant (Fig.  9A). These results demonstrated that the expression of increasing amounts of IKK␥/NEMO resulted in enhanced association of IB␣ with IKK␤ (Fig. 9A). Densitometry revealed a 6-fold increase in the amount of IB␣ associated with IKK␤ following the transfection of increasing concentrations of IKK␥/NEMO. Western blot analysis revealed that the levels of the epitopetagged IKK␤ and IB␣ in the S100 extracts were unchanged when increasing amounts of IKK␥/NEMO were transfected (Fig. 9A).
Next the ability of increasing amounts of IKK␥/NEMO to stimulate IKK␤ kinase activity was determined. An IKK␤ expression vector and increasing amounts of an expression vector encoding IKK␥/NEMO were transfected into COS cells. S100 extracts were prepared and immunoprecipitated with the Myc antibody that recognizes the Myc-tagged IKK␤. In vitro kinase assays with a GST-IB␣ substrate were then performed and indicated that transfection of increasing concentrations of IKK␥/NEMO stimulated IKK␤ phosphorylation of the IB␣ substrate ϳ18-fold (Fig. 9B). These results suggest that IKK␥/ NEMO can both facilitate the association of IB␣ with IKK␤ and increase IKK␤ kinase activity for the IB␣ substrate. DISCUSSION IKK␥/NEMO is essential for the activation of the NF-B pathway in response to a variety of stimuli (34 -41, 49 -51). For example, cell lines deficient in IKK␥/NEMO expression are defective in NF-B activation (34, 49 -51). The defects in NF-B signaling in the absence of IKK␥/NEMO probably result from two defects. First, the high molecular weight IKK complex cannot be assembled in the absence of IKK␥/NEMO (34, 49 -51). Second, IKK␥/NEMO is probably critical for the recruitment of upstream kinases, such as RIP, that are involved in stimulating IKK activity (37,45) and for the interaction with IKK␣ and IKK␤ (43). In this study, we further explore the mechanisms by which IKK␥/NEMO stimulates the NF-B pathway and demonstrate that IKK␥/NEMO also plays a role in the association of IKK␤ with the IB proteins in the high molecular weight IKK complex and increases IKK␤ activity to result in enhanced IB phosphorylation.
Both wild-type and a C-terminal deleted form of IKK␥/ NEMO increased IKK␤ phosphorylation of IB␣ and IB␤ in in vitro kinase assays. However, an N-terminal deleted form of IKK␥/NEMO that is unable to bind to IKK␤ (12,36,41,42) did not stimulate IKK␤ phosphorylation of the IB proteins. These results indicate that IKK␥/NEMO binding to IKK␤ correlates with increases in IKK␤-mediated phosphorylation of the IB.
Whether the effect of IKK␥/NEMO was due strictly to a catalytic effect on IKK␤ or potentially to IKK␥/NEMO enhancement of IKK␤ binding to IB was next addressed.
Previous results suggest that IKK␥/NEMO interacts predominantly with IKK␤ rather than IKK␣ to assemble the high molecular weight IKK complex (34,36,41). Since IKK␤ is the critical kinase involved in activating the NF-B pathway (17)(18)(19)(20)(21)(22)59), it was important to address whether IKK␥/NEMO could facilitate the association of the IB proteins with IKK␤ in the high molecular weight IKK complex. We found that, following transfection of expression vectors encoding IKK␤, IKK␥/ NEMO, and IB␣, IKK␤ itself could result in limited association of the IB proteins in the high molecular weight fractions obtained following column chromotography. In contrast, IKK␥/ NEMO alone did not result in enhanced IB association with the high molecular weight IKK complex. However, the presence of both IKK␥/NEMO and IKK␤ significantly enhanced the association of the IB proteins in the high molecular weight fractions. In contrast, an N-terminal truncation of IKK␥/ NEMO that did not bind to IKK␤ was unable to enhance IB association into the high molecular weight fractions. These results indicate that the presence of IKK␥/NEMO results in the association of both IB␣ and IKK␤ in the high molecular weight complex.
Next we investigated the requirements of IKK␥/NEMO to facilitate IB association with the high molecular fractions containing IKK. IKK␥/NEMO could recruit a kinase-dead IKK␤ mutant into the high molecular weight fractions, but IKK␥/NEMO and this mutant did not facilitate IB association. However, a degradation-resistant IB mutant could associate with IKK␤ and IKK␥/NEMO in the high molecular weight fractions, suggesting that phosphorylation of IB is not a requirement for its association with the high molecular weight IKK complex. Thus, IKK␥/NEMO and an active IKK␤ kinase are required for optimal association of IB in the high molecular weight IKK complex.
Analysis of column chromatography of extracts prepared from untreated and TNF␣-treated HeLa, MEF, and IKK␥/ NEMO knock-out cells was performed. These studies indicated that TNF␣ treatment results in marked increases in the amount of phosphorylated IB␣ in the high molecular weight fractions containing IKK␥/NEMO and IKK␤. Moreover, the presence of phosphorylated IB␣ in these fractions correlated with TNF␣-mediated increases in IKK activity. Cells lacking IKK␥/NEMO did not generate a high molecular weight IKK complex, and phosphorylated IB␣ was not detected. These results indicate that the high molecular weight complex that contains IKK␥/NEMO, IKK␤, and IB is the target for TNF␣mediated increases in IKK␤ kinase activity and IB␣ phosphorylation. However, we cannot state conclusively that IB␣ is preferentially phosphorylated in the high molecular weight complex. In summary, the results with the transfected IKK␤, IKK␥/NEMO, and IB␣ expression vectors correlate with the results obtained by analysis of the endogenous IKK complex.
Finally, coimmunoprecipitation studies were performed to assay for the in vivo interactions between IKK␥/NEMO, IKK␤, and IB. Although the IB proteins could associate with IKK␤ alone, they could not associate with IKK␥/NEMO alone. However, these studies demonstrate that the presence of IKK␥/ NEMO could enhance the association of IKK␤ with the IB proteins. Furthermore, transfection of increasing amounts of IKK␥/NEMO resulted in enhanced association of IKK␤ with IB␣ and increased IKK␤ kinase activity. These results suggest that IKK␥/NEMO probably has at least two major functions. First, it stimulates IKK␤ kinase activity either by stimulating IKK autophosphorylation or by recruitment of upstream kinases (37,45). Second, it recruits the IKK proteins into a high molecular weight complex and facilitates the interactions between IKK␤ and IB␣. Thus, IKK␥/NEMO appears to function as an adaptor protein to increase the interactions of key factors required for NF-B activation.
The mechanisms by which IKK␥/NEMO leads to recruitment of the IB␣ proteins into the high molecular weight complex remains to be determined. IKK␥/NEMO binding to IKK␤ may alter IKK␤ conformation to enhance the association of the IB␣ and IB␤ substrates in the high molecular weight IKK complex. Alternatively, IKK␥/NEMO and IKK␤ may both directly interact with IB␣. Finally, it is possible that IKK␥/NEMO recruits additional cellular proteins, which are required to facilitate the interactions between IKK␤ and IB. The enhanced association of the IB proteins and IKK␤, which is mediated by IKK␥/NEMO, probably results in the increased phosphorylation of IB␣ in response to various activators of the NF-B pathway. Consistent with these findings, TNF␣ treatment did not alter IB␣ migration following chromatography of extracts prepared from IKK␥/NEMO-deficient cells. Thus, one defect in activating the NF-B pathway in the absence of the IKK␥/ NEMO proteins may be due to the failure of IKK␤ to associate with the IB proteins to enhance its phosphorylation and degradation. Another defect in cells lacking IKK␥/NEMO is probably due to the failure to activate IKK␤ kinase activity. Further studies will be necessary to better define the interactions of these proteins and how they lead to activation of the NF-B pathway.