NF-κB-inducing Kinase and IκB Kinase Participate in Human T-cell Leukemia Virus I Tax-mediated NF-κB Activation*

The tax gene product of human T-cell leukemia virus I induces aberrant expression of various cellular genes, which contributes to transformation of host cells. Induction of many Tax target genes is mediated through transcription factor NF-κB. Here we show that Tax triggers activation of cellular protein kinases, IκB kinase α (IKKα) and IKKβ, which phosphorylate the NF-κB inhibitory protein IκBα, resulting in its degradation and NF-κB activation. Constitutive IKK activation occurs in both Tax-transfected and human T-cell leukemia virus I-infected T cells. We further demonstrate that Tax-mediated NF-κB signaling also requires the NF-κB-inducing kinase (NIK). Consistently, inactive forms of either IKKs or NIK attenuate Tax-mediated NF-κB activation. Therefore, Tax activates NF-κB by targeting cellular signaling molecules, including both IKKs and NIK.

The type I human T-cell leukemia virus (HTLV-I) 1 is the etiologic agent of an acute T-cell malignancy, termed adult T-cell leukemia (1)(2)(3). HTLV-I encodes a regulatory protein, Tax, which plays a central role in HTLV-I-induced host cell transformation (4 -6). Tax not only serves as a transactivator of the HTLV-I viral long terminal repeat (LTR) but also alters the expression of a large number of cellular genes (7). Deregulation of cellular genes appears to contribute to HTLV-I-induced host cell transformation (7,8).
Lacking DNA binding activity, Tax induces the target genes indirectly by modulating the activity of specific host transcription factors (7,9). Tax activates HTLV-I LTR by physical interaction with the CREB/ATF family of proteins, which specifically bind to the cAMP response element-like sequences present in HTLV-I LTR (10 -12). Tax interacts with the basic leucine zipper region of the CREB/ATF factors, enhancing their dimerization and DNA binding activities (13)(14)(15). Induction of many cellular genes by Tax is mediated through the cellular transcription factor NF-B (7), a key regulator of genes involved in cell activation and proliferation (16). NF-B activity is normally regulated through its cytoplasmic retention by specific inhibitors, including IB␣ and related proteins (17). Activation of NF-B by cytokines is mediated by signal transduction cascades, leading to activation of the IB kinases, IKK␣ and IKK␤ (18 -22). These kinases phosphorylate the IBs, a modification that triggers their ubiquitination and proteolysis, allowing the released NF-B dimers to enter the nucleus and activate target genes (23). In HTLV-I-infected cells or cells transfected with Tax cDNA expression vector, NF-B is constitutively nuclear (24). The deregulated nuclear expression of NF-B is essential for Tax-induced host cell transformation (25). Although precisely how Tax induces the nuclear translocation of NF-B is not fully understood, recent studies have shown that Tax induces the phosphorylation and degradation of IBs (26 -33).
In this paper, we show that Tax triggers activation of the IKKs, and this action of Tax requires the recently identified NF-B-inducing kinase (NIK) (34).
Immunoblotting Assays-Human 293 kidney carcinoma cells were seeded in 0.1% gelatin-treated 24-well plates (2.5 ϫ 10 4 cells/well) and transfected using DEAE-dextran (44) with 0.12 g of HA-IB␣ and other indicated expression vectors. Jurkat-Tag cells (5 ϫ 10 6 ) were transfected using the same protocol with 1.6 g of HA-IB␣ and other indicated expression vectors. After 40 h, whole-cell extracts were prepared and analyzed by Western blotting (46).

RESULTS
Catalytically Inactive IKKs Inhibit Tax-mediated NF-B Activation-To determine whether the recently cloned IKKs are required for activation of NF-B by Tax, we examined the effects of catalytically inactive forms of IKK␣ (IKK␣(K44M)) and IKK␤ (IKK␤(K44A)) (20, 21) on Tax-mediated transactivation of a B-driven luciferase reporter gene (B-TATA-luc). The B-TATA-luc reporter gene was transfected into Jurkat T cells either alone or together with a Tax expression vector along with increasing amounts of IKK␣(K44M) or IKK␤(K44A). As expected, Tax potently activated the B enhancer, resulting in marked induction of luciferase activity (Fig. 1A, column 2). More importantly, expression of increasing amounts of catalytically inactive IKK␣ or IKK␤ led to a dose-dependent inhibition of reporter induction (columns 3-6). At the concentrations used, the inactive IKKs only modestly affected RelA-mediated B-TATA-luc transactivation (Fig. 1B). Thus, the IKKs appeared to be required for Tax-mediated NF-B activation but not the nuclear function of NF-B. Parallel experiments performed with the WT forms of IKK␣ and IKK␤ revealed that IKK␣ potentiated Tax-mediated induction of B-dependent transcription in a dose-dependent manner (Fig. 1C, columns 2-4). At low concentrations, IKK␤ also significantly enhanced Tax-mediated transactivation (columns 5 and 6), although ex-pression of higher amounts of IKK␤ caused a reduction in reporter gene expression (column 7). At the amounts transfected, neither IKK␣ nor IKK␤ significantly affected reporter expression in Tax's absence (columns 8 and 9). These results further suggested that these IB kinases participate in Taxmediated NF-B activation. The different dose requirement for IKK␣ and IKK␤ to potentiate Tax activity may be due to differential expression of the endogenous IKKs in Jurkat cells. However, it may also be due to their different properties (19,21,22).
Inactive IKKs Interfere with Tax-induced IB␣ Phosphorylation-To determine whether the effect of the transdominant IKK mutants on Tax-mediated NF-B activation was due to the inhibition of IB phosphorylation, immunoblotting was used to examine phosphorylation of IB␣ in transiently transfected cells. When coexpressed with IB␣, Tax induced the appearance of a slower migrating form of IB␣ (Fig. 2, lane 2), which upon incubation with alkaline phosphatase regained faster electrophoretic mobility (data not shown). As demonstrated (29,48), this electrophoretic mobility change of IB␣ resulted from its phosphorylation at serines 32 and 36, since mutation of these two serines within IB␣ abolished this change (data not shown). We then examined the effect of the inactive IKKs on Tax-induced IB␣ phosphorylation. Expression of either IKK␣(K44M) or IKK␤(K44A) led to a dose-dependent inhibition of the Tax-induced appearance of the phosphorylated form of IB␣, IB␣-P (Fig. 2, lanes 3-9, middle panel). Under these conditions, the level of Tax expression was not significantly influenced (lower panel), suggesting that the dominant negative effect of the inactive IKKs was specifically on Tax-induced signal transductions. These findings strongly suggest that IKK␣ and IKK␤ are key components of Tax-mediated induction of both IB␣ phosphorylation and subsequent NF-B activation.
Tax Activates IKKs in both Transfected and HTLV-I-infected Human T Cells-To examine whether Tax causes IKK activation, we transfected Jurkat T cells with cDNA expression vectors encoding hemagglutinin (HA)-tagged IKK␣ or IKK␤ in the presence or absence of the Tax expression vector. The transiently expressed IKK␣ and IKK␤ were immunoprecipitated using an anti-HA monoclonal antibody, and their kinase activity was measured by an in vitro kinase assay (20, 21) using a substrate containing the N-terminal portion of IB␣ fused to glutathione S-transferase (GST-IB␣-(1-54)). When expressed alone, IKK␣ exhibited low basal kinase activity (Fig. 3A, lane 1,  top). However, Tax coexpression led to marked enhancement in IKK␣ catalytic activity (lane 2, top). Additionally, Tax enhanced autophosphorylation of IKK␣ (lane 2, top). We similarly examined whether IKK␤ also serves as a target of Tax. As previously reported (21,22), IKK␤ exhibited higher levels of basal kinase activity than IKK␣ (Fig. 3A, lane 3, top). However, this isoform of IKK appeared to be less responsive to Tax than IKK␣. Nevertheless, expression of Tax in the cells did result in significant enhancement of IKK␤ catalytic activity (lane 4).
To assess the physiologic relevance of our findings with transfected cells, we next examined whether IKKs are also activated in HTLV-I-infected human T cells. For these studies, we used cell lines from different stages of HTLV-I-induced transformation, including E55, N1185, N1186, HUT102, SLB-1, C8166, and MT-2. The E55, N1185, and N1186 cells are infected with HTLV-I but have not progressed to the IL-2independent stage, and the growth of these cells requires IL-2 (40). The HUT102, SLB-1, C8166, and MT-2 cells are IL-2independent HTLV-I-transformed T cells (36 -39). Both the IL-2-dependent and IL-2-independent HTLV-I-infected cell lines express high levels of Tax, which is associated with constitutive NF-B activation (26,28,33). 2 The C8166 cells are particularly interesting, since these cells carry a defective HTLV-I that selectively expresses Tax but not other viral proteins (36). To examine whether IKK was activated in these Tax-expressing cells, endogenous IKK␣ was immunoprecipi-tated from equal amounts of protein extracts of either control Jurkat cells or the C8166 cells. In vitro KA revealed that while IKK␣ from Jurkat cells had no detectable kinase activity, IKK␣ isolated from C8166 cells exhibited substantial kinase activity (Fig. 3B, middle) as well as autophosphorylation activity (top). The presence of constitutive IKK␣ kinase activity in C8166 cells was not due to an elevated IKK␣ expression. In fact, the C8166 IKK␣ immune complex contained significantly less IKK␣ than the Jurkat IKK␣ immune complex (Fig. 3B, bottom). Interestingly, when C8166 cells were incubated with tosylphenylalanyl chloromethyl ketone, an alkylating agent that inhibits Tax-induced IB␣ phosphorylation and NF-B activation (26), IKK␣ activity was largely abolished (lane 3, middle). Similar KA demonstrated that IKK was also constitutively activated in all of the other HTLV-I-infected cell lines (Fig. 3C) and a T-cell clone expressing Tax in the context of a herpes saimiri vector (Fig. 3C, lane 2, Tax1; Ref. 6). Immunoblotting revealed that all of these Tax-expressing cell lines had a lower IKK protein level than the Jurkat cells (data not shown), suggesting that the IKK activity induction associated with Tax expression was due to activation of the kinase. To further confirm that constitutive IKK activation resulted from HTLV-I infection, we infected human peripheral blood T cells with HTLV-I in vitro and compared the IKK␣ activity between the uninfected and the bulk HTLV-I-infected T cells. As expected, the uninfected human T cells exhibited undetectable levels of IKK␣ activity, which was activated upon mitogen stimulation (Fig. 3D, lanes 1 and 2). Remarkably, upon HTLV-I infection, the IKK␣ kinase activity was dramatically induced (lane 3). Together, these results demonstrate that Tax activates the IKKs in both Tax-expressing and HTLV-I-infected T cells.
Activation of IKKs by Tax Correlates with Its Ability to Activate NF-B-Prior studies led to identification of Tax mutants that are either defective or effective in NF-B activation (49). To correlate Tax-mediated activation of IKK␣ with induction of NF-B, we examined the effect of two well characterized Tax mutants, M22 and M47, on IKK␣ activity. Consistent with its deficiency in NF-B activation (49), M22 did not activate IKK␣ (Fig. 4, lane 3). In contrast, M47, which is fully effective in NF-B activation (49) induced IKK␣ activity was specific for serine 32 and serine 36 of IB␣, since substitution of these two serines with alanines in the substrate abolished 32 P incorporation (Fig. 4, lane 6, middle). Similar results were obtained with IKK␤, whose catalytic activity was enhanced by both WT Tax and TaxM47 but not influenced by TaxM22 (lanes 8 -10). We noticed that TaxM47 consistently induced stronger IKK activity than WT Tax (compare the middle part of lanes 2 and 8 and lanes 4 and 10). While the underlying mechanism remains to be further investigated, this result is consistent with the finding that M47 is a stronger NF-B activator compared with WT Tax (49). Parallel immunoblotting assays revealed that the kinase activity of IKKs was well correlated with phosphorylation of IB␣ in intact cells, indicated by its retarded electrophoretic mobility (data not shown).
To further investigate the role of NIK in Tax-mediated NF-B signaling, we analyzed the effect of WT NIK on Taxinduced IB␣ phosphorylation. In 293 cells, Tax induced only a modest increase in IB␣ phosphorylation (Fig. 5C, lane 2). As previously reported (19), overexpression of IKK␤ or IKK␤ plus NIK produced certain levels of IB␣ phosphorylation activity (data not shown). However, at lower levels, neither IKK␤ alone nor IKK␤ plus NIK induced IB␣ phosphorylation (Fig. 5C,  lanes 5 and 6). Furthermore, when coexpressed with Tax, IKK␤ induced only weak IB␣ phosphorylation (lane 4), suggesting the lack of additional factors. Interestingly, when Tax was cotransfected with both IKK␤ and NIK, a markedly higher level of IB␣ phosphorylation was detected (lane 8). Similar results were obtained with IKK␣ (data not shown). These results strongly suggest that Tax-induced IB␣ phosphorylation involves both IKKs and NIK. DISCUSSION We have shown that Tax activation of NF-B is mediated by an indirect mechanism, which requires cellular signaling machinery. This mechanism differs from the one that mediates Tax activation of the CREB/ATF factors, which involves direct interaction between Tax and these basic leucine zipper-containing cellular DNA binding proteins (13)(14)(15). Our data demonstrate that the recently cloned IB kinases, IKK␣ and IKK␤, are among the cellular signaling molecules participating in Tax-induced NF-B activation. Catalytically inactive IKKs interfere with both the phosphorylation of IB␣ and subsequent NF-B activation induced by Tax ( Figs. 1 and 2). Consistently, Tax triggers the activation of both IKK␣ and IKK␤, and this is tightly associated with the site-specific phosphorylation of IB␣ in intact cells. The importance of IKK for Tax-mediated NF-B induction is further supported by experiments performed with Tax mutants. A mutant form of Tax (M22), defective in NF-B activation, is unable to trigger IKK activation. On the other hand, M47, which harbors a mutation selectively affecting the CREB/ATF pathway, is fully active upon IKK activation (Fig. 4). Thus, Tax-mediated IKK activation is well correlated with its ability to activate NF-B. Notably, IKK activation occurs not only in Tax-transfected cells but also in various HTLV-I-infected human T-cell lines as well as in human peripheral blood T cells freshly infected with HTLV-I (Fig.  3), thus confirming the physiological relevance of our findings. Our findings support the previous reports that Tax may induce the phosphorylation and degradation of IBs (26 -31). We believe that this is a major pathway by which Tax activates NF-B.
Tax has also been suggested to directly bind to NF-B proteins and enhance their transactivation function in the nucleus (50). Although this proposal does not explain how Tax induces the nuclear translocation of NF-B, such a mechanism may exist to amplify the cytoplasmic action of Tax, i.e. activation of IKKs and induction of NF-B nuclear expression. We have noticed that the catalytically inactive IKKs and NIK have more dramatic inhibitory effects on Tax-induced IB␣ phosphorylation (Figs. 2 and 5A) than on Tax-mediated B reporter activation (Figs. 1A and 5B). Although this difference could be due to the higher sensitivity of the luciferase assay compared with immunoblotting, it may also suggest that Tax has a direct transactivation function in the nucleus. Nevertheless, our results indicate that activation of IKKs plays a central and critical role in Tax-induced nuclear expression of NF-B in both Tax-transfected and HTLV-I-infected T cells.
The mechanism by which Tax affects IKK activity remains to be further investigated. However, it appears that IKKs are not the sole target for Tax. The dominant-negative inhibition assays suggest that NIK is involved in Tax-induced IKK signaling functions. Recent studies suggest that NIK serves as an upstream kinase involved in IKK activation (18,19,51,52). NIK physically interacts with IKKs and stimulates their kinase activity when these proteins are overexpressed together in 293 cells (19). It remains to be tested whether Tax enhances or stabilizes the interaction between NIK and IKKs. Neverthe- less, the ability of Tax to induce sustained IKK activation clearly explains the basis for constitutive IB phosphorylation and NF-B activation in Tax-expressing and HTLV-I-infected T cells. Given the critical role of NF-B in cell growth and survival (53)(54)(55)(56)(57), the activation of these IB kinases may be an important mechanism by which HTLV-I transforms host cells. As such, Tax provides a useful model for how a retroviral regulatory protein transforms host cells by targeting cellular signal transduction pathways.
During the revision of this paper, similar findings were reported by Yin et al. (58). However, unlike our study, this report suggests that IKK␤, but not IKK␣, is required for Tax activation of NF-B. This discrepancy is likely due to the different experimental strategies and conditions used in these two studies. For example, Yin et al. used HIV-1 LTR, instead of the B enhancer, as an NF-B reporter. Since the HIV-1 LTR also contains other enhancers, such as NF-AT, which may respond to Tax (59), the result obtained with this reporter may not exactly reflect the activity of NF-B. Nevertheless, we have performed both reporter gene assays (using the B reporter, Fig. 1) and in vivo and in vitro IB␣ phosphorylation assays (Figs. 2-4), which clearly demonstrate that both IKK␣ and IKK␤ are involved in Tax activation of NF-B.