The Human T-cell Leukemia Virus Type-1 Tax Protein Regulates the Activity of the IκB Kinase Complex*

Two cytokine-inducible kinases, IKKα and IKKβ, are components of a 700-kDa kinase complex that specifically phosphorylates IκB. Phosphorylation of IκB by IKK leads to its ubiquitination and subsequent degradation, resulting in the nuclear translocation of NF-κB. The oncogenic protein Tax, encoded by human T-cell leukemia virus type-1 (HTLV-1), stimulates IKK activity to result in constitutive nuclear levels of NF-κB. In an attempt to gain insights into the mechanism by which Tax mediates constitutive activation of the NF-κB pathway, we analyzed the chromatographic distribution of IKK proteins using cellular extracts prepared from three T lymphocytes either lacking or containing Tax. IKK kinase activity and the distribution of proteins in the IKK complex were characterized. In extracts prepared from cells containing Tax, the activity of both IKKα and IKKβ present in the 700-kDa IKK complex were increased. Surprisingly, cell lines expressing Tax also contained an additional peak of IKKβ, but not IKKα activity, that migrated at 300 kDa rather than at 700 kDa. We noted that extracts containing Tax had extremely low levels of IκBβ, but not IκBα, and contained predominantly a truncated form of the MAP3K MEKK1. These results suggest that Tax may target several components of the NF-κB pathway leading to constitutive activation of this important regulator of cellular gene expression.

Human T-cell leukemia type-1 (HTLV-1) 1 is a retrovirus responsible for the development of human adult leukemia/ lymphoma, an aggressive and often fatal malignancy of activated CD4-positive T lymphocytes (1,2). Unlike most oncoviruses, HTLV-1 does not contain a known cellular oncogene but instead encodes a regulatory protein (Tax) which is essential for both activating viral gene expression and inducing cellular transformation (3)(4)(5). Tax-mediated transformation is manifested through at least two distinct regulatory pathways (5,6). The first pathway regulated by Tax involves its interaction with the ATF/CREB family of cellular transcription factors (7)(8)(9) and the coactivators CBP and p300 (10,11) to activate HTLV-1 gene expression. The second pathway regulated by Tax involves its ability to constitutively activate the nuclear translocation of the transcription factor NF-B (12)(13)(14).
NF-B is a heterodimeric transcription factor that is predominantly composed of 50-and 65-kDa subunits of the Rel family (for reviews, see Refs. [15][16][17]. NF-B stimulates gene expression from a number of cellular genes, including those involved in the control of the inflammatory and immune response in addition to being a key regulator of cellular growth properties (15, 18 -21). In resting cells, NF-B is largely retained in the cytoplasm by a family of cytoplasmic inhibitory proteins known as IB, which interact with NF-B to mask its nuclear localization signal (22)(23)(24). IB consists of a family of proteins including IB␣, IB␤, and IB⑀ (for a recent review, see Ref. 16). All three of these proteins contain two serine residues in their N terminus, which regulate their protein stability and a central NF-B interaction domain consisting of multiple ankyrin repeats. IB␣ and IB␤ also contain a C-terminal PEST domain that contributes to basal protein turn-over (16). Upon stimulation of cells by a variety of agents including the cytokines TNF␣ and interleukin-1, lipopolysaccharide, and the HTLV-1 Tax protein, N-terminal serine residues in IB are rapidly phosphorylated (25)(26)(27)(28)(29)(30). The IB proteins are then ubiquitinated and degraded by the 26 S proteasome, leading to the nuclear transport of NF-B and subsequent increases of expression of NF-B-responsive genes (17,27,31).
The discovery of two IB kinases, IKK␣ and IKK␤, whose activity is markedly increased by cytokine treatment of cells (32)(33)(34)(35)(36) or the presence of the HTLV-1 Tax protein (37)(38)(39)(40), was critical in defining cellular components which regulate the NF-B pathway. Within a high molecular weight IKK complex that migrates between 600 and 900 kDa (33,(41)(42)(43)(44)(45), several additional cellular components have been identified that may regulate IKK activity. These include two members of the MAP3K family, MEKK1 (33) and NIK (43); a scaffold protein, IKAP (43); and a regulatory protein, NEMO/IKK␥/IKKAP1 (44 -46). Several lines of evidence suggest that MEKK1 and NIK are upstream activators of IKKs and function to mediate cytokine induction of the NF-B pathway (32,(47)(48)(49). IKK activity is induced in cells overexpressing either MEKK1 or NIK (47,49). In addition, NIK associates with IKK both in vivo and in vitro (32,36,43). Finally, recombinant MEKK1 induces in vitro phosphorylation of IKK␣ and IKK␤ (37,47). These results suggest that NIK and MEKK1 kinases can modulate IKK activity. The exact role of NEMO/IKK␥/IKKAP1 is not clear. However, the direct interaction of this protein with IKK␤ and its ability to activate IKK kinase activity suggest that this protein mediates association of IKK with the upstream activators of the NF-B pathway (44 -46). IKAP may be a scaffold protein that binds to NIK and IKKs to assemble them into an active kinase complex (43).
Tax-mediated nuclear translocation of NF-B is likely controlled by at least two distinct mechanisms. Tax is able to bind directly to the p100 and p105 precursors of NF-B that function as cytoplasmic inhibitors of the NF-B protein, facilitating NF-B nuclear translocation (50). More importantly, Tax increases the activity of cellular signal transduction pathways that increase the phosphorylation of IB␣ and IB␤ at Nterminal serine residues (29,39). Recent studies demonstrate that Tax interaction with MEKK1 (37), IKKs (38,40), or NIK (40) stimulates IKK activity (37)(38)(39)(40). NEMO or IKK␥, a component of the 700-kDa IKK complex, is critical for modulating IKK activity, and has recently been demonstrated to function in mediating Tax activation of the NF-B pathway (44,51,52). These results suggest that the same set of signaling components that are necessary for the rapid cytokine-induced phosphorylation of IB are also likely utilized for the persistent activation of the NF-B pathway by Tax.
In contrast to the effects of most cytokines, which lead to a rapid but transient increase in IKK activity, Tax leads to a persistent increase in IKK activity (38). This leads to the constitutive accumulation of NF-B in the nucleus of HTLV-1transformed T lymphocytes (53). Persistent decreases in IB␤ levels by Tax have been suggested to be a potential mechanism by which Tax leads to constitutive activation of the NF-B pathway (29). IB␤ gene expression, in contrast to that of IB␣, is not up-regulated by NF-B (29,54), which may explain the ability of Tax to lead to persistent decreases in IB␤ but not IB␣ protein levels.
In an attempt to better understand the role of Tax on constitutive activation of IKK kinase activity, we analyzed IKK activity and protein composition following chromatographic fractionation of extracts prepared from cells either containing or lacking Tax. We demonstrate that in HTLV-1-infected Tcells, IKK activity in the high molecular weight IKK complex is significantly increased compared with its activity in cells lacking Tax. In addition, we detected a lower molecular mass complex that contains increased IKK␤ activity in cells containing Tax. Tax was also found to modify the composition of several other components that regulate the NF-B pathway. These results suggest that Tax mediates activation of the NF-B pathway by affecting several components of this important regulatory loop.

EXPERIMENTAL PROCEDURES
Cells-SLB (55) and JPX-9 cells (56) were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum (HyClone Laboratories), 2 mM L-glutamine, and antibiotics (penicillin-streptomycin). Jurkat cells were maintained in the same medium, except that the serum (HyClone) containing low levels of lipopolysaccharide were used.
Preparation of Cytoplasmic Extracts-Cytoplasmic extracts were prepared according to Dignam (57) with slight modifications. Cells were harvested from culture medium by centrifugation for 10 min at 2000 rpm (Beckman bench-top centrifuge, CH3.7 rotor). Pelleted cells were washed twice with cold phosphate-buffered saline and were resuspended in 5 volumes of buffer A (10 mM Hepes (pH 7.9), 1.5 mM MgCl, 10 mM KCl, 0.5 mM dithiothreitol) supplemented with phosphatase inhibitors (50 mM NaF, 50 mM glycerophosphate, 0.125 M okadaic acid, 1 mM sodium orthovanadate) and proteinase inhibitors at the level suggested by the manufacturer (Roche Molecular Biochemicals). After incubation for 15 min on ice, the cells were lysed with 40 strokes of a Kontes all-glass Dounce homogenizer (B type pestle). The nuclei were pelleted by centrifugation at 2000 rpm. The supernatant was mixed with 0.11 volume of buffer B (0.3 M Hepes (pH 7.9), 0.03 M MgCl, and 1.4 M) and then centrifuged at for 60 min at 100,000 ϫ g. The super-natant was dialyzed for 5-8 h against 20 volumes of buffer D (20 mM HEPES (pH 7.9), 0.1 M KCl, 0.5 mM dithiothreitol, 0.5 mM PMSF, 20% glycerol, and 0.2 mM EDTA).
Western Blot Analysis-Equal volumes of proteins (50 -100 l) from each of the Superdex-200 fraction were resolved by SDS-polyacrylamide gel electrophoresis, transferred onto nitrocellulose membranes, and probed with specific antibodies. The membrane-bound immune complexes were analyzed using an enhanced chemiluminescence system (Amersham Pharmacia Biotech).

TNF␣ Activation of IKK Kinase
Activity-IKK is present in a large molecular mass complex of approximately 700 kDa (41,42), and its activity is stimulated by cytokine treatment of cells (32,33,35,36,58). Before determining the effects of Tax on the activity and mobility of IKK complexes, we first analyzed the chromatographic distribution IKK␣ and IKK␤ in extracts prepared from mock-treated and TNF␣-stimulated Jurkat cells. Following Superdex-200 fractionation of these extracts, immunoprecipitation of column fractions with either IKK␣-or IKK␤specific antibodies was performed and kinase activity was assayed using a GST fusion with the N-terminal 54 amino acids of IB␣. Increased phosphorylation of IB by IKK␣ (Fig. 1A) and IKK␤ (Fig. 1C) was observed in fractions derived from TNF␣-treated cells when compared with fractions derived from mock-treated cells (Fig. 1, A and C). There was no phosphorylation of a GST-IB␣ fusion protein containing substitution of serine residues 32 and 36 with alanine (data not shown). The majority of this activity was found migrating between 600 and 700 kDa. There were similar levels of IKK␣ and IKK␤ proteins in mock and TNF␣-treated cells as determined by Western blot analysis (Fig. 1, B and D). These data confirm previous results that TNF␣ stimulates IKK activity present in a high molecular weight complex (33).

IKK Migration in Extracts Prepared from HTLV-1-infected T-Cells-
To determine the role of Tax on modulating the activity and composition of the IKK complex, we compared IKK complexes in extracts isolated from the HTLV-1-infected T-cell line, SLB (55), and the uninfected T-cell line, Jurkat. The fractionation scheme and analysis was repeated utilizing three separate sets of Jurkat and SLB extracts, and the results obtained were similar to those shown. Cytoplasmic extracts from these cells were first step-eluted with 0.3 M KCl on a Q-Sepharose column and then fractionated on a Superdex-200 gel filtration column. Fractions eluted from the Superdex-200 column were analyzed for kinase activity and the presence of specific proteins using Western blot analysis. Expression of the Tax protein in the SLB cells was confirmed using immunoblotting analysis with a Tax monoclonal antibody (37). The 40-kDa Tax protein was observed only in SLB extract and peaked at two positions corresponding to a molecular mass of 500 -600 kDa and 60 -100 kDa ( Fig. 2A) (59,60). As expected, no Tax was detected in fractions isolated from Jurkat cells ( Fig. 2A).
Since extracts prepared from different T-lymphocyte cell lines were analyzed, it was important to compare the activity of other cellular kinases in these cells. The cyclin H-dependent kinase CDK7, which phosphorylates the RNA polymerase II CTD was analyzed (61,62). This kinase has no known role in activating the NF-B pathway. CDK7 was immunoprecipated from each of these column fractions, and kinase reactions were performed using a GST fusion with the CTD of RNA polymerase II as a substrate (63). The level and chromatographic position of CDK7 kinase activity was similar in extracts prepared from Jurkat and SLB cells (Fig. 2B). In addition, Western blot analysis indicated that CDK7 protein was present at a similar chromatographic position and in similar amounts in extracts prepared from the two cell lines (Fig. 2C).
Fractions of these cell extracts isolated following Superdex-200 chromatography were then immunoprecipitated with antibodies directed against either IKK␣ or IKK␤, and kinase activity was assayed. In fractions containing Jurkat extract, IKK␣ activity was found to migrate predominantly at a molecular mass of approximately 600 kDa, although a small amount of the activity was also evident at ϳ70 kDa (Fig. 3A). In fractions isolated from SLB extract, there was increased IKK␣ kinase activity relative to that seen in fractions isolated from Jurkat cells, and this activity corresponded to a complex with a molecular mass of ϳ 600 kDa (Fig. 3A). IKK␣ protein levels and distribution were comparable in extracts prepared from the two cell lines (Fig. 3B). These results indicated that the increased amount of IKK activity in HTLV-1-infected cells was not due to differences in the level of the IKK proteins.
There was much less IKK␤ kinase activity in Jurkat cells than that found in SLB cells (Fig. 3C). In contrast to our results with IKK␣, the majority of IKK␤ kinase activity in extracts prepared from SLB cells was found in a 300-kDa complex rather than in the 600 -700-kDa complex (Fig. 3C). This "shifted" distribution profile of IKK␤ kinase activity between Jurkat and SLB extracts correlated with a shift in the peak, but not Extracts were prepared from these cells as described under "Experimental Procedures" and subjected to chromatography on Superdex-200 column. The fractions derived from the indicated cells were immunoprecipitated with an anti-IKK␣ polyclonal antibody (Santa Cruz, sc-7218) and assayed for phosphorylation of a GST/IB␣ (aa 1-54) substrate. Phosphoproteins were resolved on a 12% SDS-polyacrylamide gel and analyzed by autoradiography. Fraction numbers and molecular mass markers are indicated at the bottom and the top of the figures, respectively. B, fractions derived from the indicated cells were subjected to Western blotting using the same antibody was used as in A. C, IKK␤ kinase activity was assayed using these same fractions by first performing immunoprecipitation with IKK␤ antibody (Santa Cruz, sc-7607) and then assaying kinase activity with the GST/IB␣ (aa 1-54) protein. D, Western blot analysis was also performed on these fractions using the IKK antibody used in C.

FIG. 2. Chromatographic distribution of Tax and CDK7 isolated from Jurkat and SLB extract.
A, equivalent amounts of proteins isolated following Superdex-200 gel filtration were resolved on a 12% SDS-polyacrylamide gel and were analyzed by Western blot analysis using anti-Tax monoclonal antibody. B, kinase activity was performed with these column fractions derived from the indicated cell lines following immunoprecipitation with CDK 7 antibody (Santa Cruz, sc-723) utilizing GST/CTD as a substrate. C, Western blot analysis with CDK7 antibody was used to analyze these fractions. Molecular weight markers for the column are indicated at the top of the figure, and column fractions are indicated at the bottom of the figure. the level, of the IKK␤ protein in these extracts (Fig. 3D). These results demonstrate that there appears to be an additional lower mobility complex containing IKK␤ in HTLV-1-infected cells as compared with extracts prepared from Jurkat cells.
NIK and MEKK1 Activity in the Presence and Absence of Tax-NIK and MEKK1 are kinases that can stimulate IKK activity by phosphorylation of IKK␣ and IKK␤ (47)(48)(49). Thus, it was important to address whether Tax could increase the activity or alter the chromatographic behavior of either NIK or MEKK1. NIK was immunoprecipitated from either Jurkat or SLB column fractions that had been subjected to Superdex-200 chromatography and were analyzed in Fig. 2. NIK kinase activity was assayed using a recombinant kinase-defective mutant of IKK␣ (Lys 3 Met) as the substrate. This IKK␣ substrate was used because the Lys 3 Met mutation eliminated IKK␣ autophosphorylation. Although there was a slight increase in NIK kinase activity in SLB fractions (Fig. 4A), the chromatographic distribution of NIK protein (Fig. 4B) was similar in extracts prepared from Jurkat and SLB cells. The significance of this slight increase in NIK activity remains to be determined. These results suggest that NIK be not markedly altered by Tax expression.
MEKK1 was previously demonstrated to bind to Tax, and this binding was found to correlate with Tax activation of the NF-B pathway (37). MEKK1 activity was assayed following immunoprecipitation with antibody directed against the C terminus of MEKK1. GST-MEK4 was used as a substrate in these studies since MEK4 is a well characterized substrate of MEKK1 (64,65). The majority of MEKK1 kinase activity migrated between 200 and 450 kDa and exhibited a similar mobility in extracts prepared from SLB and Jurkat cells (Fig. 5A). Western blot analysis demonstrated that full-length MEKK1 was readily detected in Jurkat extract, but no full-length MEKK1 was detected in SLB fractions (Fig. 5B). However, a faster mobility species was detected in Western blot analysis with the C-terminal MEKK1 antibody in the same chromatographic positions in both cell extracts (Fig. 5B). Peptide blocking indicated that both the full-length and the faster-migrating proteins detected by this MEKK1 antibody were MEKK1-related proteins (data not shown). An N-terminal specific MEKK1 antibody also could detect the full-length MEKK1 in Jurkat cells, while the lower molecular weight species was not observed in either Jurkat or SLB fractions (data not shown). These results indicate that the faster-migrating protein was likely a cleaved product of MEKK1 that lacks N-terminal coding sequences. These results suggest that the activity and distribution of NIK and MEKK1 were not markedly altered in the presence of Tax, although differences in the proportion of MEKK1 protein cleavage appeared to be correlated with Tax expression.

NF-B and IB Distribution in the Presence and Absence of Tax-Column fractions isolated from SLB and Jurkat cells
were also analyzed for the presence of the NF-B and IB proteins. The p50 (Fig. 6A), p65 (Fig. 6B), IB␣ (Fig. 6C), and IB␤ (Fig. 6D) proteins were predominantly distributed in a high molecular fraction corresponding to a molecular mass of 450 -600 kDa. This was similar to the position that the majority of IKK␣ was found in extracts prepared from both cell lines. As expected, the amount of p65 was lower in SLB extract than that seen in Jurkat extract due to the fact that only the cytoplasmic components of these cells were analyzed. IB␣ levels were similar in Jurkat and SLB extracts (Fig. 6C), while IB␤ levels were much lower in SLB extract than in Jurkat extract (Fig. 6D). It is important to note that the IB␣ levels in SLB cells contain two species, which reflect phosphorylated and unphosphorylated IB␣ species when assayed with phosphospecific IB␣ antibody (Ref. 66 and data not shown). The low level of IB␤ in SLB cells is in agreement with a previous study that demonstrates that this protein helps to mediate the constitutive nuclear expression of NF-B in Tax-containing cells (29).
Analysis of IKK Activity in Cells with Inducible Tax Expression-Our data demonstrate that increased IKK␣ and IKK␤ kinase activity and the presence of a novel 300-kDa IKK␤ complex are found in extracts prepared from HTLV-1-transformed cells. To determine whether the increase in IKK activity was due to the presence of Tax or other HTLV-1-encoded proteins expressed in the SLB cells, a Jurkat cell line (JPX-9) (56) in which Tax expression is under the control of the metallothionein promoter was assayed. The role of Tax could be assessed in this cell line in the absence of other viral proteins following cadmium induction, which induces Tax expression (50,56). Cadmium induction of JPX-9 cells results in a marked induction of NF-B binding activity when assayed by EMSA (Ref. 50 and data not shown). Extracts prepared from either untreated or cadmium (Cd)-treated JPX-9 cells were subjected to chromatography on a Superdex-200 column and assayed for IKK activity and protein distribution.
IKK␣ activity was increased in extracts prepared from Cdtreated JPX-9 cells as compared with extracts prepared from untreated JPX-9 cells (Fig. 7A). The position of IKK␣ kinase activity in Cd-treated JPX-9 cells was similar to the chromatographic distribution of IKK␣ seen in SLB cells (Fig. 7A). No obvious differences were seen in IKK␣ protein levels between uninduced and Cd-induced JPX-9 cells (Fig. 7B). In Cd-induced JPX-9 cells, IKK␤ activity was increased in the column fractions migrating between ϳ600 and 700 kDa and also in fractions migrating at 300 kDa (Fig. 7C). These results are similar to the findings in extracts prepared from SLB cells and suggest that Tax leads to the activation of IKK␤ kinase activity in complexes migrating at both 700 kDa and at 300 kDa. There was no obvious change in the distribution of the IKK␤ protein between extracts prepared from non-induced and Cd-induced JPX-9 cells, as was noted with extracts prepared from SLB cells (Fig. 7D).
The distribution of Tax in non-induced and Cd-induced JPX-9 cells was also analyzed. Tax was detected in extract prepared from the Cd-induced but not non-induced JPX-9 cells and was detected exclusively in a chromatographic position corresponding to ϳ600 kDa (Fig. 7E). The relative chromatographic distribution of IKK␥/NEMO was also assessed in noninduced and induced JPX-9 cells. Since IKK␥/NEMO can interact with IKK␤ and regulate IKK activity (51,52), the chromatographic profile of IKK␥/NEMO should be similar to that of the IKK␤. Western blot analysis demonstrated that the majority of IKK␥/NEMO was present in similar column fractions with IKK␤ (Fig. 7F). These results suggest that Tax itself can increase IKK␣ and IKK␤ kinase activity and lead to the generation of two IKK complexes, each of which has increased ability to phosphorylate IB␣.
Finally, we assayed whether Tax would alter IKK␤ kinase activity in COS cells transfected with an epitope-tagged IKK␤ cDNA alone or in the presence of a Tax cDNA. Following transfection of COS cells with these constructs, cytoplasmic extracts were prepared and subjected to chromatography on a Superdex-200 column. The column fractions were assayed for both IKK␤ kinase activity (Fig. 7G) and IKK␤ protein levels in Western blot analysis (Fig. 7H)  the 300-kDa fraction (Fig. 7G). In the presence of Tax, Western blot analysis demonstrated that the predominant amount of IKK␤ kinase was present in the 300-kDa fraction (Fig. 7H). These results are consistent with a direct role for Tax to increase IKK␤ kinase activity present in the same 300-kDa fraction detected in HTLV-1-infected cells. DISCUSSION The analysis presented in this study focuses on the role of the HTLV-1 Tax protein on modulating the activity and the composition of the IKK complex (37)(38)(39)(40). Using extracts prepared from the HTLV-1-infected cells (SLB) and the Tax-inducible cells, JPX-9, we demonstrate that Tax increases the kinase activity of both IKK␣ and IKK␤ within a ϳ700-kDa complex. Interestingly, Tax also induces an appreciable increase of IKK␤ kinase activity, which is present within a lower molecular weight complex migrating at ϳ300 kDa. A similar size IKK␤ complex has also been described in non-induced and TNF␣treated extracts (45). In TNF␣-treated cells, majority of IKK␤ kinase activity was in the 700-kDa fraction with only low levels of kinase activity in the 300-kDa fraction. The activation of IKK activity by Tax is not at due to increased synthesis of the IKK proteins as the overall protein levels of these kinases were not changed in Tax-expressing and non-expressing cells.
Our data suggest that stimulation of IKK activation by Tax likely occurs by modulating the activity of a preformed high molecular weight complex (41). However, changes in the composition of such a complex may occur during cytokine induction or Tax expression. The distribution of variety of proteins involved in NF-B regulation, including p50, p65, NIK, IB␣, IB␤, and IKK␥, overlap the 700-kDa complex containing activated IKKs. The chromatographic profile of these proteins was similar in cells either containing or lacking Tax. Consistent with the notion that a preformed IKK complex can be activated by various stimulators of the NF-B pathway is the previous observation that IKK is present in an inactive form in the ϳ700-kDa IKK complex isolated from uninduced HeLa cells and can be activated by the addition of recombinant MEKK1 (42).
We observed a predominance of different forms of MEKK1 in extracts prepared from Jurkat and SLB cells. The full-length form of MEKK1 was detected in Jurkat cells, and its activity profile for a well characterized substrate MEK4 correlated with IKK activity. MEK4 was used as a substrate in this assay, rather than IKK␤, due to the high background seen with this later substrate. An N-terminal truncated form of the MEKK1 protein was present in both Jurkat and SLB cells and migrated FIG. 7. Chromatographic distribution of the IKK complex in JPX-9 and COS extract. Extracts were prepared from JPX-9 cells that were either untreated or treated with 20 M cadmium. IKK␣ and IKK␤ kinase activity (A and C, respectively) and protein distribution (B and D) were analyzed on Superdex-200 column fractions with the antibodies described in Fig. 1. E, Western blot analysis of Tax expression was analyzed in column fractions isolated from extracts prepared from untreated or Cd-treated cells. F, Western blotting of the IKK␥ protein from the column fractions was performed using IKK␥ antibody (Santa Cruz, sc-8330). Positions of the proteins are indicated at the far right. Molecular mass markers for the Superdex-200 column are indicated at the top, and the fraction numbers are indicated at the bottom of each panel. G and H, COS cells (10 8 ) were transfected with a Flag-tagged IKK␤ cDNA alone or in the presence of Tax. Cytoplasmic extracts were prepared from these cells and subjected to Superdex 200 chromatography. G, the M2 monoclonal antibody was utilized to immunoprecipitate the Flag-tagged IKK␤ protein in each of these fractions and kinase assays were performed with the GST-IB␣ protein (aa 1-54), followed by SDS-polyacrylamide gel electrophoresis and autoradiography. H, the M2 monoclonal antibody was used in Western blot analysis to detect the Flag-tagged IKK␤ protein.
at a position between the large IKK complex and the lower molecular weight IKK␤ complex. Differences between the distribution of MEKK1 protein and its kinase activity may be due to different MEKK isoforms such as MEKK2 and MEKK3 that can be immunoprecipitated with the C-terminal MEKK1 antibody and phosphorylate the MEK4 substrate (67,68). Whether both the full-length and the truncated forms of MEKK1 are involved in activation of IKKs is not known. However, a variety of data suggest that MEKK1 in addition to MEKK2 and MEKK3 can stimulate IKK activity (37,47,68,69). HTLV-1 infection or perhaps Tax itself may induce the cleavage of the N terminus of MEKK1 (70 -72) or perhaps lead to differential splicing of the MEKK1 gene. Finally, the truncation of MEKK1 in SLB cells may simply be a reflection of the HTLV-1 transformation. Whether these changes in MEKK1 are related to Tax-mediated activation of IKK activity will require further investigation.
A striking feature that we observed following chromatography of extracts prepared from HTLV-1-infected (SLB) and Taxexpressing cells (JPX-9) was the generation of a faster mobility complex containing high levels of IKK␤ activity. This IKK␤ complex appears to correlate with Tax expression and may result from the removal of an activated form of IKK␤ from the large molecular weight IKK complex. This notion is supported by several observations. First, in extracts prepared from SLB cells as compared with Jurkat cells, there is a greater abundance of IKK␤ protein detected in the lower molecular weight fractions. This is in sharp contrast to IKK␤ distribution in extracts prepared from TNF␣-treated Jurkat cells, where the majority of IKK␤ activity is associated with the high molecular weight IKK complex. We did not detect the same degree of shift of IKK␤ protein in Cd-induced JPX-9 cells as was seen in extracts prepared from SLB cells. However, both extracts exhibited increased IKK␤ activity in the lower molecular weight IKK complex. As SLB cells express about 20-fold higher levels of Tax than Cd-induced JPX-9 cells (data not shown), a detectable shift in the IKK␤ protein to the lower molecular weight IKK complex might not readily be seen in induced JPX-9 extracts due to the lower levels of Tax. Finally, we demonstrated that Tax increases the activity of IKK␤ present in the 300-kDa IKK complex following transfection of Tax and IKK␤ cDNA constructs. Further analysis will need to be performed to determine components of this lower molecular weight IKK complex in addition to IKK␤.
Tax appears to utilize a distinct mechanism to initiate and maintain IKK activation to result in the constitutive nuclear accumulation of NF-B. This may result from Tax-induced modification of the composition of the IKK complex. Recent reports indicate that IKK␤, but not IKK␣, is the active kinase necessary for stimulation of the NF-B pathway (73)(74)(75)(76)(77). Our observations suggest that the lower molecular weight IKK complex containing IKK␤ may be important for the persistent Tax activation of the NF-B pathway seen with Tax. Once IKK is activated, the ability of HTLV-1 to maintain IKK activity would be a potential mechanism to establish a persistent infection. The ability of Tax to stimulate IKK activity in both high and lower molecular weight IKK complexes may contribute to the constitutive activation of NF-B pathway seen in the presence of Tax (78).
The mechanism that Tax utilizes to result in persistent increases in IKK activity remains to be determined. If Tax is responsible for formation of a lower molecular weight complex containing IKK␤, then Tax may not need to be physically incorporated into the lower molecular weight complex. As Tax itself is unable to interact directly with IKKs (37, 52), it is likely that Tax acts transiently on other components of the IKK complex via interaction with upstream kinases such as NIK or MEKK1 (37). Other possible candidates for Tax modulation of IKK activity include the recently identified scaffold protein IKAP (43) and the Tax-interacting protein IKK␥/NEMO (44 -46, 51, 52). The identification of cellular factors that are modified by Tax will provide further insights into the mechanism by which Tax expression leads to persistent activation of the NF-B pathway and its ability to transform T lymphocytes.