IKKγ Serves as a Docking Subunit of the IκB Kinase (IKK) and Mediates Interaction of IKK with the Human T-cell Leukemia Virus Tax Protein*

The tax gene product of human T-cell leukemia virus type I induces activation of transcription factor NF-κB, which contributes to deregulated expression of various cellular genes. Tax expression triggers persistent phosphorylation and degradation of the NF-κB inhibitory proteins IκBα and IκBβ, resulting in constitutive nuclear expression of NF-κB. Recent studies demonstrate that Tax activates the IκB kinase (IKK), although the underlying mechanism remains unclear. In this report, we show that Tax physically interacts with a regulatory component of the IKK complex, the NF-κB essential modulator or IKKγ (NEMO/IKKγ). This molecular interaction appears to be important for recruiting Tax to the IKK catalytic subunits, IKKα and IKKβ. Expression of NEMO/IKKγ greatly promotes binding of Tax to IKKα and IKKβ and stimulates Tax-mediated IKK activation. Interestingly, a mutant form of Tax defective in IKK activation exhibited a markedly diminished level of NEMO/IKKγ association. These findings suggest that the physical interaction of Tax with NEMO/IKKγ may play an important role in Tax-mediated IKK activation.


The tax gene product of human T-cell leukemia virus type I induces activation of transcription factor NF-B, which contributes to deregulated expression of various cellular genes. Tax expression triggers persistent phosphorylation and degradation of the NF-B inhibitory proteins IB␣ and IB␤, resulting in constitutive nuclear expression of NF-B. Recent studies demonstrate that Tax activates the IB kinase (IKK), although the underlying mechanism remains unclear. In this report, we show that Tax physically interacts with a regulatory component of the IKK complex, the NF-B essential modulator or IKK␥ (NEMO/IKK␥). This molecular interaction appears to be important for recruiting Tax to the IKK catalytic subunits, IKK␣ and IKK␤. Expression of NEMO/IKK␥ greatly promotes binding of Tax to IKK␣ and IKK␤ and stimulates Tax-mediated IKK activation.
Interestingly, a mutant form of Tax defective in IKK activation exhibited a markedly diminished level of NEMO/IKK␥ association. These findings suggest that the physical interaction of Tax with NEMO/IKK␥ may play an important role in Tax-mediated IKK activation.
The type I human T-cell leukemia virus (HTLV-I) 1 transforms human T-cells, which is associated with the development of an acute T-cell malignancy termed adult T-cell leukemia (1). HTLV-I encodes a regulatory protein, Tax, which plays a central role in the induction of host cell transformation (1). Tax alters the expression of a large number of cellular genes involved in cell growth and survival, which appears to contribute to the oncogenic activity of this viral protein (2,3). Lacking DNA binding activity, Tax induces the target genes indirectly by modulating the activity of specific host transcription factors (3,4). Induction of many cellular genes by Tax is mediated through the transcription factor NF-B (3), a key regulator of genes involved in cell activation, proliferation, and survival (for recent reviews, see Refs. 5-7).
We and others have shown recently that the IKK is constitutively activated in HTLV-I-infected or Tax-expressing T-cells (22)(23)(24)(25). This action of Tax results in persistent degradation of IB␣ and IB␤ and constitutive nuclear expression of NF-B (26 -30). Tax has been shown to induce the catalytic activity of both IKK␣ and IKK␤, although the underlying mechanism remains unclear (17). In this paper, we show that Tax physically interacts with NEMO/IKK␥, which facilitates the recruitment of Tax to the catalytic subunits IKK␣ and IKK␤ and promotes Tax-mediated activation of IKK.

MATERIALS AND METHODS
Plasmids and Antibodies-The pcDNA-HA vector was constructed by inserting a copy of the influenza hemagglutinin (HA) epitope tag (YPY-DVPDYA) together with the translation initiation codon (ATG) into the mammalian expression vector pcDNA3.1 (Invitrogen). To generate pcDNA-HA-NEMO/IKK␥, the murine NEMO cDNA (provided by Drs. S. Yamaoka and A. Isrëal (17)) was subcloned into the pcDNA-HA vector downstream of the HA tag. Truncation mutants of NEMO/IKK␥ were generated by restriction digestion and designated by the specific amino acid residues retained in the mutant proteins. For example, NEMO/IKK␥-(1-312) contains the N-terminal 312 amino acids. pCMV4-HA-NIK was generated by inserting HA-tagged human NIK cDNA (provided by Dr. David Wallach (20)) in the pCMV4 vector (31). pCMV4-HA-IB␣ was described previously (32). The pcDNA-HA-IKK␣, pcDNA-HA-IKK␤, and pcDNA-HA-MEKK1 were provided by Dr. M. Karin. The pCMV4-Tax, pCMV4-TaxM22, and pCMV4-TaxM47 were provided by Dr. W. C. Greene (33). The anti-HA and anti-IKK␥ monoclonal antibodies were from Roche Molecular Biochemicals and Imgenex Corp., respectively. The anti-Tax monoclonal antibody was prepared from a hybridoma (168B17-46-34) provided by the AIDS Research and Reference Program, NIAID, National Institutes of Health. All the other antibodies were purchased from Santa Cruz Biotechnology, Inc.
Immunoprecipitation (IP) and Immunoblotting Assays-Human 293 kidney carcinoma cells were seeded in 0.1% gelatin-treated six-well plates (1 ϫ 10 5 cells/well) and transfected using DEAE-dextran with the indicated cDNA expression vectors. The DNA amounts used for the transfections were normalized based on the expression efficiency of each of the expression vectors: 50 ng for IKK␣, IKK␤, MEKK1, and NIK; 50 -200 ng for NEMO/IKK␥ and its truncation mutants; 0.5 g for Tax and its mutants. After 40 h, recipient cells were lysed in RIPA buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 0.01 volume of a protease inhibitor mixture (34)). The RIPA buffer was also used to prepare whole-cell extracts from the HTLV-I-infected SLB-1 cells (24). IP was performed as described previously (34), and the precipitated proteins were analyzed by SDS-polyacrylamide gel electrophoresis followed by immunoblotting. For immunoblotting analyses of IB␣ phosphorylation, 293 cells were transfected in 24-well plates. Whole-cell extracts were prepared in ELB buffer (34) supplemented with phosphatase inhibitors and analyzed by immunoblotting (34).

NEMO/IKK␥ Serves as a Primary Target of Tax in the IKK
Complex-Prior studies suggest that Tax interacts with the IKK catalytic subunits IKK␣ and IKK␤ (23) as well as the upstream kinase MEKK1 (22). However, since the IKK components are present in a large complex in intact cells, it remains unknown which component serves as the primary target of Tax. To address this question, co-IP was performed using a high stringency binding buffer (RIPA) to assess the relative binding affinity of Tax with the different molecular components of the IKK complex. Tax was transiently expressed in 293 cells either alone or together with each of the known IKK components, including IKK␣, IKK␤, NEMO/IKK␥, MEKK1, and NIK, all of which were tagged with the HA antigenic epitope. In the absence of an IKK component, Tax was not precipitated by the anti-HA antibody (Fig. 1A, lane 3), although this protein was readily precipitated by the anti-Tax antibody (lane 2), demonstrating the specificity of the antibodies. Under these condi-tions, a small amount of Tax was coprecipitated with IKK␣ and NIK (lanes 4 and 8). Coprecipitation of Tax with IKK␤ and MEKK1 was also detected after an extended exposure time (data not shown). Importantly, the IP assays revealed a remarkably higher amount of Tax coprecipitated with NEMO/ IKK␥ (Fig. 1A, lane 6). Thus, while Tax may weakly bind to various components of the IKK complex, NEMO/IKK␥ appears to be a major target of this viral transactivator protein. To address the physiological relevance of this finding, we examined the interaction between Tax and NEMO/IKK␥ in HTLV-I-infected T-cells. As expected, Tax was readily detected from the immune complex precipitated by the anti-Tax antibody (Fig. 1C, lane 2), but not from that precipitated by a preimmune serum (lane 1). More importantly, a significant amount of Tax was coprecipitated with IKK␥ by an IKK␥-specific antibody (lane 3). These findings clearly demonstrated that Tax stably binds to NEMO/IKK␥ in both transfected and HTLV-Iinfected cells.
Different Sequences of NEMO/IKK␥ Are Involved in Binding to Tax and IKK␣-NEMO/IKK␥ is known to form a stable complex with IKK␣ and IKK␤ in vivo (17,18). To map the regions of NEMO/IKK␥ involved in binding to the IKK catalytic subunits and Tax, progressive truncations were generated from both the N and C termini of NEMO/IKK␥. Deletion of the N-terminal 110 amino acids of NEMO/IKK␥ did not affect its binding to IKK␣ (Fig. 2A, lane 3) or Tax (lane 10). Deletion of up to 292 amino acids from the C terminus also did not affect the IKK␣ association (lanes 4 -6), although further removal of 8 amino acids generated a mutant (1-112) that no longer bound IKK␣ (lane 7). In contrast to that seen with IKK␣, Tax inter-  1-7) or Tax (lanes 8 -14). Cell extracts were subjected to IP with anti-IKK␣ (lanes 1-7) or anti-HA (lanes 8 -14) followed by immunoblotting (IB) using anti-HA (lanes 1-7) or anti-Tax (lanes 8 -14). Since the full-length NEMO/IKK␥ migrated closely with the immunoglobulin heavy chain (IgH), lanes 1 and 2 were run in a separate gel. The coprecipitated NEMO/IKK␥ proteins are labeled with arrowheads. B, extracts used in the corresponding lanes of A were subjected to direct immunoblotting using anti-HA to monitor the expression of the full-length and truncated forms of NEMO/IKK␥. action with NEMO/IKK␥ required the C-terminal sequences of NEMO/IKK␥. Removal of 100 amino acids from this end of the molecule significantly reduced, although not abolished, its interaction with Tax ( Fig. 2A, lane 11). This low level binding activity was also detected between Tax and an IKK␥ mutant lacking its C-terminal 157 amino acids (lane 12). However, further deletion of 135 or more amino acids from this end generated IKK␥ mutants (1-120 and 1-112), which completely lost Tax-binding activity (lanes 13 and 14). Of note, IKK␥-(1-120) still retained its IKK␣ binding activity. Thus, while the N-terminal 120 amino acids of NEMO/IKK␥ are sufficient for binding to IKK␣, the C-terminal region is required for strong interaction with Tax.
NEMO/IKK␥ Facilitates Binding of Tax to the Catalytic Subunits of IKK and Promotes Tax-mediated IKK Activation-To assess the role of the interaction between Tax and NEMO/IKK␥ in IKK activation, we examined the effect of NEMO/IKK␥ on Tax interaction with the catalytic subunits of IKK and on Tax-mediated IKK activation. For these studies, Tax was coexpressed with HA-tagged IKK␣ or IKK␤ together with either the full-length NEMO/IKK␥ or its C-terminal truncation mutant, 1-312. We found that the low level interaction of IKK␣ and IKK␤ with Tax could be more readily detected when the IP was performed using the anti-Tax antibody (Fig.  3A, lanes 2 and 5). More importantly, the interaction of Tax with both IKK␣ and IKK␤ was remarkably enhanced by the full-length NEMO/IKK␥ (lanes 3 and 6). Parallel immunoblotting assays showed that NEMO/IKK␥ did not affect the expression level of the IKKs (data not shown). Furthermore, this stimulatory effect was only weakly detected with the NEMO/ IKK␥-(1-312) (lanes 4 and 7), a C-terminal truncation mutant exhibiting reduced affinity in Tax binding (see Fig. 2A).
We then examined the effect of NEMO/IKK␥ on Tax-mediated IKK activation by measuring the in vivo phosphorylation of its substrate IB␣. In this regard, the in vivo IB␣ phosphorylation could be detected by immunoblotting since the phosphorylated IB␣ migrates more slowly on SDS gels (27,(35)(36)(37). As we reported previously (24), expression of IKK␤ alone or together with Tax in 293 cells did not induce appreciable IB␣ phosphorylation (Fig. 3B, upper panel, lanes 1 and 2). However, when these cells were cotransfected with the full-length NEMO/IKK␥, the IKK␤ kinase activity was markedly induced, resulting in the potent phosphorylation of the substrate IB␣ (upper panel, lanes 5 and 6, IB␣-P). Degradation of IB␣ became evident when a higher dose of NEMO/IKK␥ was used (lane 6, upper panel). We also observed that Tax induces the autophosphorylation of IKK␤ in the presence of NEMO/IKK␥ (middle panel, lanes 5 and 6, IKK␤-P). These functional effects were not significantly detected with the NEMO/IKK␥-(1-312) (lanes 3 and 4). Furthermore, NEMO/IKK␥ was unable to activate IKK␤ in the absence of Tax (Fig. 3C). Together, these results strongly suggest that NEMO/IKK␥ promotes association of Tax with the IKK catalytic subunits and stimulates Tax-mediated IKK activation.
To assess the role of upstream kinases in the IKK␥-dependent induction of IKK␤ by Tax, we examined the effect of a dominant-negative NIK (DN-NIK) on the in vivo phosphorylation of IB␣ triggered by Tax (Fig. 3D). As expected, phosphorylated form of IB␣ was readily detected in cells transfected with IKK␤ together with NEMO/IKK␥ and Tax (lane 1). Cotransfection of the DN-NIK led to a partial inhibition of the IB␣ phosphorylation. Thus, it is likely that Tax-mediated activation of IKK may involve both its physical interaction with IKK and the participation of upstream kinases.
A Mutant Form of Tax, Defective in IKK Activation, Is Inefficient in NEMO/IKK␥ Binding-To further assess the functional importance of Tax interaction with NEMO/IKK␥, we examined the ability of two well characterized Tax mutants, M22 and M47 (33), to interact with NEMO/IKK␥. The M22 is defective in IKK activation and induction of NF-B nuclear expression, while M47 remains competent in these functions (22)(23)(24)(25). Interestingly, co-IP assays revealed that while the wild type Tax and M47 both strongly interacted with NEMO/ IKK␥ (Fig. 4, upper panel, lanes 2 and 4), only a marginal NEMO/IKK␥ binding activity was detected with M22 (lane 3). Thus, the physical interaction of Tax with NEMO/IKK␥ is well correlated with its ability to activate the IKK catalytic subunits. DISCUSSION HTLV-I-encoded Tax protein is a potent activator of the cellular transcription factor NF-B. Tax activation of NF-B involves phosphorylation and degradation of the NF-B inhibitory proteins IB␣ and IB␤ (26 -30). Tax activates the catalytic activity of IKK in both transiently transfected and HTLV-I-transformed cells (22)(23)(24)(25). The mechanism by which Tax activates IKK remains elusive. A recent study suggests that the NEMO/IKK␥ subunit of IKK is essential for Tax-mediated NF-B activation, although it is unclear how this noncatalytic signaling protein participates in Tax activation of IKK (17). Our current data demonstrate that NEMO/IKK␥ physically interacts with Tax in both transfected 293 cells and HTLV-Iinfected T-cells. Under stringent IP conditions, NEMO/IKK␥ exhibited markedly higher Tax binding activity than various other IKK components (Fig. 1). The functional importance of this physical interaction in Tax-mediated activation of IKK is supported by the experiments performed with Tax mutants (Fig. 4). The M22 mutant, known to be defective in IKK activation (22)(23)(24)(25), exhibits strikingly reduced binding to NEMO/ IKK␥. How could the binding of NEMO/IKK␥ to Tax contribute to Tax activation of IKK? One potential mechanism is that NEMO/IKK␥ serves as an adaptor molecule to recruit Tax to the IKK catalytic subunits. As shown in Fig. 2, the C-terminal region of NEMO/IKK␥ is required for its strong binding to Tax, whereas the N-terminal region of NEMO/IKK␥ mediates interaction with IKK␣. Although further studies are required to demonstrate that NEMO/IKK␥ can form a heterotrimer with Tax and IKK␣ (or IKK␤), our current data show that NEMO/ IKK␥ indeed promotes the binding of Tax to IKK␣ and IKK␤ as well as stimulates the Tax-mediated IKK activation (Fig. 3). Furthermore, these functions were not significantly detected with a NEMO/IKK␥ C-terminal truncation mutant, 1-312, which exhibits markedly reduced Tax binding activity.
It remains to be further studied how the recruitment of Tax to IKK complex triggers IKK activation. A recent study suggests that Tax also physically interacts with MEKK1 and activates the catalytic activity of this IKK-activating kinase (22). Thus, an intriguing model is that Tax recruits upstream kinases, such as MEKK1, to the IKK complex, thus triggering phosphorylation-dependent IKK activation. However, interaction of Tax with MEKK1 appears to be relatively weak under the stringent IP conditions used in this study (Fig. 1). We have shown previously that NIK is functionally involved in Tax activation of IKK, although this kinase also does not strongly bind to Tax. It remains to be examined whether the weak interaction of Tax with MEKK1 or NIK is sufficient to recruit these upstream kinases to IKK. Nevertheless, we have shown that the IKK␥-dependent IKK activation by Tax is at least partially sensitive to DN-NIK. This finding suggests that Tax activation of IKK may involve both Tax/IKK physical interaction and activation or recruitment of upstream kinases.