β-TrCP Mediates the Signal-induced Ubiquitination of IκBβ*

We have examined the role of β-TrCP (β-transducin repeat-containing protein) in the ubiquitination and degradation of IκBβ, one of the two major IκB isoforms in mammalian cells. We demonstrate that β-TrCP interacts specifically with IκBβ, and such interaction is dependent on prior phosphorylation of IκBβ on serines 19 and 23. Interaction with β-TrCP is also necessary for ubiquitination of IκBβ upon stimulation of cells, and deletion of the F-box in β-TrCP abolishes its ability to ubiquitinate IκBβ. Therefore, these results indicate that β-TrCP plays a critical role in the activation of NF-κB by assembling the ubiquitin ligase complex for both phosphorylated IκBα and IκBβ.

cules activated by ubiquitin-activating enzyme E1 are attached to specific lysine residues on the target protein by a ubiquitinconjugating enzyme (E2), together with a ubiquitin ligase (E3) that is specific for the substrate (10). Assembly of polyubiquitin chains on the substrate protein targets it for degradation by the 26 S proteosome (10). Recently an F-box/WD40 protein called ␤-TrCP (␤-transducin repeat-containing protein) was shown to be the substrate-recognition component of the ubiquitin ligase responsible for phosphorylation-dependent ubiquitination of IB␣ (11)(12)(13)(14)(15). ␤-TrCP recognizes IB␣ phosphorylated at Ser-32 and Ser-36 through its WD40 domain, whereas the F-box motif recruits additional proteins including Skp1 and Cullin to form the Skp1-cullin-F-box (SCF) ubiquitin ligase complex (16). ␤-TrCP belongs to a growing family of proteins containing F-boxes that are involved in assembling the SCF complex. Aside from the F-box, these proteins have another protein-protein interaction module in their C terminus, namely a WD or leucine-rich repeat (LRR) domain. These C-terminal domains mediate the interaction of SCF complexes with their substrates and determine specificity of substrate recognition. ␤-TrCP has been implicated in the ubiquitination of CD4 (through HIV protein Vpu) (17), IB␣ (11)(12)(13)(14)(15), and ␤-catenin (14, 18 -20). All these proteins share similar N-terminal inducible phosphorylation sites with the consensus sequence of DSGXS ( represents a hydrophobic residue and X represents any amino acid.). Therefore, the inducible phosphorylation of these N-terminal serine residues is the critical step that allows recruitment of ␤-TrCP and subsequent ubiquitination of these proteins (16).
IB␤, like IB␣, is also believed to undergo signal-induced phosphorylation, ubiquitination, and degradation (21)(22)(23). The critical phosphorylation sites are serines 19 and 23 because mutation of these amino acids prevents signal-induced degradation of IB␤ (21). IB␤ can also be phosphorylated in vitro by IKK␣ and IKK␤ (7). However, unlike IB␣, direct phosphorylation of IB␤ at serines 19 and 23 in vivo is yet to be demonstrated (8). Instead, one study has reported that serines 19 and 23 of IB␤ are constitutively phosphorylated in unstimulated cells, suggesting that the regulation of IB␤ might differ more fundamentally from that of IB␣ (24). Because phosphorylation of IB␤ by the IKKs does not induce a mobility shift in SDS-PAGE, and in vivo labeling experiments have been inconclusive, signal-induced phosphorylation of IB␤ remains to be unequivocally demonstrated. 2 IB␤ also differs from IB␣ in other aspects (1,22,23). For example, while IB␣ responds to all NF-B inducers, in certain cell-types IB␤ responds only to a subset of them (22,24). Also, when stimulated by the same inducers, the kinetics of IB␤ degradation is significantly slower than IB␣ (22,24). The underlying mechanism responsible for these differences is unclear although one possible explanation for the slower kinetics of IB␤ degradation might be lower efficiency of ubiquitination of phosphorylated IB␤. Understanding the details of the pathway by which phosphorylated IB␤ is ubiquitinated and degraded is therefore important for fully decoding the differential regulation of IB␣ and IB␤. The identification of ␤-TrCP as the recognition element of IB␣ ubiquitin-ligase provides an opportunity to directly test whether IB␤ also undergoes signal-induced phosphorylation * This work was supported by a grant from the National Institutes of Health (R01 AI33443) and the Howard Hughes Medical Institute. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
We report in this manuscript that ␤-TrCP specifically interacts with IB␤ in stimulated cells. This interaction requires serines 19 and 23 because mutation of these residues completely abolishes this interaction. We also demonstrate that phosphorylation-induced ubiquitination of IB␤ requires the F-box of ␤-TrCP, suggesting that both IB␣ and IB␤ appear to be ubiquitinated and degraded through the same pathway. Therefore the differential regulation of IB␣ and IB␤ is most likely because of differences in other steps in the activation pathway of NF-B.

EXPERIMENTAL PROCEDURES
Cell Cultures, Antibodies, and Reagents-293, HeLa, and COS cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. The anti-flu mouse monoclonal antibody (12CA5) was produced and purified in this laboratory. Anti-flag monoclonal antibody M5 and anti-flag M2 affinity gel were purchased from Sigma. All other antibodies were purchased from Santa Cruz Biotechnology. Protein A-Sepharose was purchased from Amersham Pharmacia Biotech.
Cloning of Human ␤-TrCP and ␤-TrCP⌬F-Total RNA was isolated from HeLa cells using TRIzol™ Reagent (Life Technologies, Inc.) and used for RT-PCR to amplify ␤-TrCP cDNA with appropriate 5Ј-and 3Ј-primers. A FLAG-epitope coding sequence was inserted after the starting codon. The 1.7-kilobase PCR product was subcloned into BamHI and XbaI sites of expression vector pCDNA3 (Invitrogen) and sequenced.
To delete the F-box from ␤-TrCP, two internal primers were designed to flank the boundary sequences outside of F-box region and used individually with the 5Ј-or 3Ј-primers described above to amplify the N terminus or C terminus of ␤-TrCP. The 500-base pair N-terminal and 1.1-kilobase C-terminal PCR products were then used as templates in the sequential PCR with 5Ј-and 3Ј-primers. The resulting ␤-TrCP⌬F was subcloned into BamHI-XbaI sites of pCDNA3, and its sequence was confirmed by DNA sequencing.
Luciferase Assay-Subconfluent 293 cells were transfected with 500 ng of pBIIX luciferase reporter construct, along with different amounts of ␤-TrCP or ␤-TrCP⌬F constructs. The total transfected DNA amounts were equalized with empty pCDNA3 vector. After 36 h, cells were treated with or without 20 ng/ml TNF-␣ for 4 h before harvest for luciferase assay (Promega).
Transfection, Immunoprecipitation, and Immunoblotting-Cells were grown in 10 centimeter plates to 40% confluence and transfected with indicated DNA using FuGENE™ 6 (Roche Molecular Biochemicals). After incubation for 36 h, cells are treated with or without TNF-␣ (10 ng/ml) for 30 min before being lysed with TNT lysis buffer (200 mM Tris-HCl, pH 8.0, 200 mM NaCl, 1% Triton-100) supplemented with protease inhibitors. In immunoprecipitation experiments, cell lysates were incubated with 20 l of anti-flag M2 affinity gel for 3 h or 10 l of anti-IB␤ (C-20) along with 20 l of protein A-Sepharose for 4 h at 4°C. Immobilized immuno-complexes were washed with TNT three times, boiled in SDS loading buffer, and resolved on 10% SDS-PAGE. Proteins were transferred to Immobilon transfer membrane (Millipore Corp.) and blotted with indicated primary antibody for 3 h at room temperature and appropriate secondary antibody for 1 h. Immunoreactive bands were visualized by ECL.

RESULTS AND DISCUSSION
␤-TrCP Is Involved in NF-B Activation and ␤-TrCP⌬F Acts As a Dominant Negative Regulator of NF-B Activation-Human ␤-TrCP was cloned using RT-PCR from HeLa cells with primers designed according to the published protein sequence. An F-box deletion mutant, ␤-TrCP⌬F, was generated by deleting the F-box region from leucine 148 to leucine 192 (Fig. 1) (19). Both wide type and ⌬F mutant of ␤-TrCP were flag-tagged at their N terminus, cloned into the expression vector pcDNA3, and their integrity verified by DNA sequencing and in vitro expression.
We first investigated the role of ␤-TrCP in NF-B activation. We transfected 293 cells with ␤-TrCP or ␤-TrCP⌬F construct, together with a luciferase reporter gene pBIIX-luc, which harbored two NF-B binding sites in its promoter region. As shown

FIG. 1. ␤-TrCP, wild type and a mutant form lacking the F-box.
Shown are schematic domain structures of ␤-TrCP and ␤-TrCP⌬F used in this study. Both constructs were tagged with the flag-epitope at the N terminus.

␤-TrCp in IB␤ Degradation
previously (11)(12)(13)(14)(15), introduction of ␤-TrCP⌬F into the cells significantly inhibited NF-B activation in a concentration-dependent manner (data not shown). The F-box motif has been found to be important for associating with Skp1, which in turn binds to Cullin and an E2 enzyme to form a functional ubiquitin-conjugating complex. Therefore the ␤-TrCP⌬F construct would bind to phosphorylated IB but fail to recruit the other components, thus inhibiting the degradation of IB and activation of NF-B. Surprisingly, we also found that transfection of wild type ␤-TrCP also inhibited NF-B activation, although to a lesser extent than the F-box deletion mutant (data not shown). The explanation for this observation is unclear, but one possibility is that overexpression of ␤-TrCP results in the accelerated degradation of some other component that is required for activation and nuclear translocation of NF-B.
␤-TrCP Binds to Phosphorylated IB␤-To examine whether ␤-TrCP directly interacts with phosphorylated IB␤, we conducted immunoprecipitation experiments in transfected cells. IB␣ and IB␤ were transfected into 293 and HeLa cells respectively, along with either FLAG-tagged wild type ␤-TrCP or FLAG-tagged TrCP⌬F construct. Cells were incubated with the proteosome inhibitor, calpain inhibitor 1, before treatment with TNF-␣. Cell lysates were precipitated with anti-flag affinity gel, and the immobilized immuno-complex was immunoblotted for IB␣ or IB␤. The experiment confirmed that interaction between IB␣ and ␤-TrCP is only observed in TNF␣stimulated cells (Fig. 2A). Similarly, IB␤ failed to associate with ␤-TrCP in unstimulated cells, but bound efficiently to ␤-TrCP in TNF-␣ stimulated cells (Fig. 2B). In both instances, deletion of the F-box deletion did not affect the ability of ␤-TrCP to interact with IB␣ or IB␤ (Fig. 2, A and B). This result is therefore consistent with the notion that ␤-TrCP interacts with its phosphorylated substrate through its WD domain, and this binding is independent of the F-box.
Treatment with TNF-␣ has been shown to cause the phosphorylation of IB␣ at the N-terminal serine residues 32 and 36 (21). It has been implied, but not directly demonstrated, that IB␤ is also inducibly phosphorylated on serines 19 and 23 (2). To ascertain whether ␤-TrCP binds only to IB␤ phosphorylated at serines 19 and 23, we used an IB␤ mutant, IB␤19/23 SS/AA, in which serines 19 and 23 were replaced with nonphosphorylatable alanines. As shown in Fig. 2C, IB␤19/23 SS/AA mutant failed to associate with ␤-TrCP even upon TNF-␣ treatment. This confirmed that the interaction between IB␤ and ␤-TrCP is contingent upon prior phosphorylation of IB␤ at the N terminus. Interestingly, an IB␤19/23 SS/DD mutant in which the two serines were substituted by aspartic acid, to mimic the phosphorylated state, also failed to bind to ␤-TrCP or ␤-TrCP⌬F. This experiment demonstrates the stringent substrate specificity of ␤-TrCP for phosphate groups in the IB proteins. Similar specificity of interaction had been observed in earlier studies where it was demonstrated that only a phosphopeptide encompassing the IB␣ degradation motif, but not a serine-to-glutamate-substituted peptide, could compete with intact IB␣ for ubiquitin conjugation (15).
␤-TrCP Promotes Ubiquitination of Phosphorylated IB␤ in Vivo-To directly assess whether ␤-TrCP is a component of the ubiquitin ligase for IB␤, we transfected COS cells with wild type or ⌬F mutant ␤-TrCP along with HA(flu)-tagged IKK␤. COS cells appear to lack certain components in the signaling FIG. 3. ␤-TrCP, but not ␤-TrCP⌬F, promotes IKK-dependent ubiquitination of IB␤. A, COS cells were transfected with flu-tagged wild type IB␤, pcDNA3 (Ϫ) or ␤-TrCP constructs as indicated, along with flu-tagged IKK␤ in lanes 2, 4, and 6. Cell lysates were incubated with anti-flag M2 affinity gel, and the immunoprecipitated complex was resolved on SDS-PAGE and blotted with anti-IB␤ in upper panel or anti-ubiquitin monoclonal antibody in middle panel. IKK␤ expression was checked in lower panel by Western blot with anti-flu antibody. B, the experiment was similar to that in Fig. 3A, lanes 3 and 4. COS cells were transfected as indicated, immunoprecipitated with anti-flag M2 affinity gel, and immunoblotted with anti-IB␤(C-20). After ECL reaction, the blot was exposed for significantly longer time than that for panel A. C, COS cells were transfected with IB␤, IKK␤, and ␤-TrCP constructs as indicated. Immunoprecipitation was carried out with protein A-Sepharose and anti-IB␤(C-20). Bound proteins were immunoblotted with antiubiquitin antibody (upper panel). Protein expression in the transfected cells was analyzed, and immunoblotting results are shown in middle (␤-TrCP, immunoblotted with anti-flag antibody) and lower (IB␤, immunoblotted with anti-flu antibody) panels.

␤-TrCp in IB␤ Degradation
pathways leading to NF-B activation and hence do not respond to traditional inducers of NF-B. 3 Therefore to help bypass this difficulty, we co-transfected HA(flu)-tagged IKK␤, FLAG-tagged wild type or ⌬F mutant ␤-TrCP, and IB␤. The IB␤ bound to ␤-TrCP was analyzed by immunoprecipitation of ␤-TrCP with anti-FLAG antibody, followed by immunoblotting with IB␤ antisera. Under these conditions, where transfection of IKK␤ presumably led to continuous phosphorylation of IB proteins, multiple IB␤ bands with increasing molecular weights were detected in ␤-TrCP transfected cells. (Fig. 3A,  upper panel, lane 4). In contrast, only a single band corresponding to IB␤ is observed in cells transfected with the dominant negative ␤-TrCP⌬F construct (Fig. 3A, upper panel, lane 6). Although we could detect forms of IB␤ containing one or two ubiquitin molecules using the IB␤ antibody, we did not observe polyubiquitinated forms under these experimental conditions. Because the levels of polyubiquitinated IB forms are very low, probably because they are rapidly degraded, we repeated the experiment and exposed the ECL blot for significantly longer periods. Under these conditions, we could detect low amounts of higher molecular weight forms of IB␤ that probably represent polyubiquitinated forms of the protein (Fig.  3B). To further characterize the higher molecular weight IB␤ immunoreactive bands, we immunoblotted the anti-flag-immunoprecipitated complexes with ubiquitin antibody. In contrast to the immunoblot with the IB␤ antibody, the slower migrating bands were readily detected with the ubiquitin antibody (Fig. 3A, middle panel, lane 4). The explanation for why the ubiquitin antibody detects the higher molecular weight species more readily is probably because of the far greater number of epitopes that are presented by the polyubiquitinated forms. As expected, deletion of the F-box in ␤-TrCP (␤-TrCP⌬F) almost completely blocked the formation of ubiquitin-IB␤ conjugates (Fig. 3A, middle panel, lane 6). Therefore in cells transfected with IKK␤, ␤-TrCP is directly involved in IB␤ ubiquitination.
To further confirm that ␤-TrCP promotes the ubiquitination of IB␤, whereas the ␤-TrCP⌬F suppresses it, we examined the state of IB␤ in ␤-TrCP and ␤-TrCP⌬F transfected cells by directly immunoprecipitating IB␤ itself. Transfection of either IB␤ or IB␤ along with ␤-TrCP does not lead to significant ubiquitination of IB␤ (Fig. 3C, upper panel, lanes 1 and 2). However, upon activation by IKK␤ transfection, IB␤ is polyubiquitinated (Fig. 3C, upper panel lane 3). Transfection of wild type ␤-TrCP significantly increased the level of ubiquitination of IB␤ (Fig. 3C, lane 4), whereas cells transfected with ␤-TrCP⌬F failed to generate ubiquitinated forms of IB␤ (Fig.  3C, lane 5). The ubiquitination of IB␤ is dependent on serines 19 and 23 because a mutant IB␤ containing alanines in these positions could not be ubiquitinated (Fig. 3C, lane 6). Therefore these observations are in agreement with earlier results examining the binding of mutant IB␤ with ␤-TrCP (Fig. 2, B and C).
In summary, we report that ␤-TrCP binds specifically to the inducibly phosphorylated IB␤ and promotes its ubiquitination. Our findings further help establish the role of ␤-TrCP as a component of the ubiquitin ligase for IB proteins, and demonstrate that the signal-induced phosphorylation of IB␤ by IKKs is a critical step that precedes their ubiquitination and degradation. Therefore, differences in the regulation of IB␣ and IB␤ must be because of differences in other steps in the pathway. For example, it is possible that IB␤ complexes contain an additional regulatory component that determines the rate of degradation of ubiquitinated IB␤ proteins, thus explaining their slower rate of degradation. Alternatively, such an associated regulatory protein may influence the ability of IB␤ to be efficiently phosphorylated by IKK. Finding the answers to these questions remains a challenge for the future.