Constitutive Activation of TAK1 by HTLV-1 Tax-dependent Overexpression of TAB2 Induces Activation of JNK-ATF2 but Not IKK-NF-κB*

HTLV-1 Tax oncoprotein induces persistent activation of the transcription factor NF-κB and CREB (cAMP-response element-binding protein)/ATF. Transforming growth factor-β-activated kinase 1 (TAK1) has been shown to play a critical role in these transcription factors. Here, we found that TAK1 was constitutively activated in Tax-positive HTLV-1-transformed T cells. Tax induced persistent overexpression of TAK1-binding protein 2 (TAB2), but not TAB3, which is essential for TAK1 activation. Surprisingly, TAK1 was not involved in the activation of NF-κB. On the other hand, JNK and p38 mitogen-activated protein kinases were activated by TAK1. In addition, ATF2, but not CREB, was a target for the TAK1-JNK pathway, and p38 negatively regulated TAK1 activity through TAB1 phosphorylation. These results indicate that Tax-mediated TAK1 activation is important for the activation of ATF2 rather than NF-κB.

(ATLL). Infection with this virus results in the activation of several transcriptional factors including NF-B and CREB in host CD4ϩ T cells. In particular, the activation of NF-B correlates with the expression of HTLV-1-derived oncoprotein Tax (1). It has been reported that Tax associates with IKK␥ to activate IB kinase (IKK) complex (2).
Various alternations of intracellular functions by infecting HTLV-1 and Tax expression result in the onset of ATLL; however, the molecular mechanisms of these alternations are still unclear. The present study investigated whether TAK1 is involved in Tax-dependent NF-B activation and other signaling pathways in HTLV-1-transformed cells.

* This work was supported in part by Grant-in-aid (C)17590055 for Scientific
Research; the 21st Century Center of Excellence Program; and the Cooperative Link of Unique Science and Technology for Economy Revitalization from the Ministry of Education, Culture, Sports, Science and Technology, Japan. 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. 1  Cell Culture-HTLV-1-transformed cells (19) were maintained in RPMI 1640 supplemented with 10% fetal calf serum, 100 units/ml penicillin, and 100 g/ml streptomycin at 37°C in 5% CO 2 .
Immunoblotting-Cell lysates, prepared as described previously (6), were resolved by SDS-PAGE and transferred to Immobilon-P nylon membrane (Millipore). The membrane was treated with BlockAce (Dainippon Pharmaceutical Co. Ltd., Suita, Japan) overnight at 4°C and probed with primary antibodies as described above. Antibodies were detected using horseradish peroxidase-conjugated anti-rabbit, anti-mouse, anti-goat, and anti-sheep IgG (Dako), and visualized with the ECL system (Amersham Biosciences).
Electrophoretic Mobility Shift Assay-Nuclear extracts were prepared from HuT-102 cells as described previously (21) and probed with 32 P-labeled consensus B and Oct-1 sites. The pro-tein-DNA complexes were separated on 4% PAGE as described previously (21).

Tax-induced Activation of NF-B and CREB/ATF-We first
investigated the relationship between the activation of intracellular signals and Tax expression in HTLV-1 infected cells. As shown in Fig. 1A, Tax expression correlated with activation of the IKK-NF-B pathway. Constitutive IKK kinase activity correlated with the phosphorylation of p65 at Ser-536 and IB␣ at Ser-32/Ser-36 (Fig. 1A). In addition, JNK, p38, CREB, ATF1, and ATF2 were also activated in Tax-expressing T cells (Fig.  1B).
Tax-dependent TAK1 Activation and TAB2 Expression-Although Tax interacts with IKK␥ (2), it remains unknown how it activates the IKK complex. Here, we hypothesized that Tax recruits the upstream molecule to activate IKK. TAK1 is a potential candidate for this molecule as it has been shown to be an IKK kinase in several cells including splenic T lymphocytes. We therefore investigated whether Tax could activate TAK1 in HTLV-1-transformed cells. As shown in Fig. 2A, the phosphorylation of TAK1 at Thr-187, a critical site for TAK1 activation, was promoted selectively in Tax-positive cells (Hut-102 and MT-2). Interestingly, in these cells, the expression of TAB2, but not TAB3, increased at the both protein and mRNA levels (Fig.  2, A and B). In addition, the shifted band of TAB1 on SDS-PAGE was detected in Tax-positive cells, although the protein expression levels were comparable. Immunoprecipitation with anti-TAB2 or TAB3 antibodies revealed increased formation of the active TAK1 complex with TAB1 and TAB2 but not TAB3. Furthermore, knockdown of Tax by siRNA decreased TAB2 expression as well as TAK1 activation in HuT-102 cells (Fig.  2C). In addition, down-regulation of TAB2 led to decreased phosphorylation of TAK1 without affecting Tax protein expression. These results indicate that Tax activates TAK1 by increasing TAB2 expression at the endogenous level.
TAK1 Activation Is Important for the Activation of MAPKs but Not of NF-B-To examine whether activated TAK1 causes NF-B activation, we evaluated the effect of 5Z-7-oxozeaenol, a specific inhibitor for TAK1, on the downstream signaling pathways. Although treatment with the inhibitor resulted in a rapid decrease in the phosphorylation of TAK1 at 3-12 h (Fig. 3A), the phosphorylation of NF-B p65 was not affected in HuT-102 and MT-2 cells (Fig. 3B). On the other hand, TAK1 inhibitor suppressed the activation of JNK and p38 (Fig. 3B). In contrast, a specific IKK␤ inhibitor blocked selectively the phosphorylation of NF-B p65 (Fig. 3B). This correlated with the fact that TAK1 complex did not physiologically associate with Tax-IKK complex (Fig. 3C). Similarly, RNA interference (RNAi) of TAK1 suppressed the activation of JNK and p38 but not NF-B (Fig.  3D). The constitutive IKK kinase activity and the DNA binding activity of NF-B were not affected by inactivation of TAK1 by RNAi (Fig. 3, D and E). Moreover, although IKK inhibitor induced a severe apoptotic response, a higher concentration of TAK1 inhibitor did not induce cell death (Fig. 3F). These results clearly demonstrated that Tax-dependent TAK1 activation is independent of NF-B activation but is important for the activation of stress-activated MAPKs.
Activation of ATF2 by TAK1 through JNK1-To characterize the physiological role of activated TAK1 in HTLV-1-transformed cells, we next investigated the effects of TAK1 kinase activity blockade on the CREB/ATF family of transcription factors. Both 5Z-7-oxozeaenol and TAK1 siRNA suppressed the constitutive phosphorylation of ATF2, but not of CREB and ATF1 (Fig. 4, A and B). A JNK-specific inhibitor, SP600125, also inhibited the signal leading to ATF2. We further clarify the role of JNK1 and JNK2 by RNAi and found that JNK1 transduced mainly the TAK1 signal to ATF2 activation (Fig. 4D). These results indicate that TAK1 regulates the activation of ATF2 via JNK1 activation. TAK1 Is Negatively Regulated by p38-mediated TAB1 Phosphorylation-On the other hand, a p38 inhibitor, SB203580, and siRNA against p38␣ enhanced the JNK-ATF signal (Fig. 4, C and E). It has been reported that the association   (5 M). After incubation at 37°C in 5% CO 2 for 60 h, cell lysates were extracted and immunoblotted with the indicated antibodies. IKK activity was measured by in vitro kinase assay using purified IKK complex with anti-IKK␥ antibody. E, HuT-102 cells were electroporated with each siRNA (5 M). After incubation at 37°C in 5% CO 2 for 60 h, nuclear extracts were prepared. An electrophoretic mobility shift assay (EMSA) was carried out using 32 P-labled oligonucleotide probes containing consensus B and Oct-1 sites. F, HuT-102 cells were treated with 5Z-7-oxozeaenol (0.1 or 0.9 M) or 3 M IKK inhibitor VI when seeded in a 3.5-cm dish. After incubation at 37°C, 5% CO 2 for 54 h, cell viability was measured using Cell Titer-Glo kit.
of TAB1 with p38␣ negatively regulates TAK1 kinase activity by phosphorylating TAB1. We therefore hypothesized that the inhibition of p38 promoted the activation of TAK1 by canceling TAB1-dependent feedback inhibition. As expected, p38 inhibitor increased TAK1 activation with decreased TAB1 phosphorylation (Fig. 4F). These results suggest that p38 maintains the level of TAK1 activity by controlling the TAB1-mediated negative feedback mechanism.

DISCUSSION
Constitutive activation of NF-B is a typical pathological alteration in several cancer cells including ATLL cells. We reported previously that IKK phosphorylates p65 at Ser-536 (22). In the present study, enhanced phosphorylation of p65 was observed in Tax-positive cells, suggesting that p65 phosphorylation will be a useful marker for detecting the constitutive activation of NF-B in cancer cells. We also found that TAK1 is constitutively activated through Tax-mediated over-expression of TAB2 ( Fig. 2A). TAK1 has been shown to play a critical role in NF-B activation in response to various extracellular stimuli including T cell receptor activation (5, 7, 9 -11, 15, 23). Therefore, we hypothesized that Tax-dependent activation of TAK1 might induce IKK activation; however, our data did not support this hypothesis (Fig. 2C). Wu and Sun (24) recently reported that TAK1 is involved in Tax-induced activation of NF-B; however, no direct evidence of the role of TAK1 in Tax induction of NF-B in HTLV-1-transformed cells was shown. In contrast, this led to the conclusion of direct blockade of TAK1 kinase activity using both chemical inhibitor and siRNA in HTLV-1-transformed cells. They also showed that TAK1 associates with Tax-IKK complex in SLB-1 cells (Tax-positive HTLV-1-transformed cells). We also observed constitutive activation of TAK1 in the cell line (data not shown); however, the association of TAK1 with the Tax-IKK complex could not be detected in HuT-102 cells (Fig. 3C). Consistent with our data, Ghoda et al. (25) reported that Tax is not able to induce IKK phosphorylation in TAK1 Ϫ/Ϫ MEFs. Further study will be necessary to settle this issue; however, we currently believe that Tax-dependent IKK activation is regulated by a factor other than TAK1. Recently, protein phosphatase 2A has been reported as a candidate for this factor (26).
The JNK pathway has been widely accepted as another main signal triggered by activated TAK1. It should be noted that TAK1 is essential for activation of the JNK-ATF2 pathway in HTLV-1-infected cells. RNAi experiment also revealed that JNK1, but not JNK2, is essential for activation of ATF2. In addition, persistent activation of p38␣ negatively regulated the TAK1-JNK1-ATF2 pathway by phosphorylating TAB1. Moreover, the Wnt/␤-catenin pathway is known as a target for TAK1 (27,28); however, TAK1 inhibitor had no inhibitory effect on the stable expression of ␤-catenin in HTLV-1-infected cells (data not shown). Collectively, we claim that the constitutive activation of TAK1 led to the activation of stress-activated MAPKs but not NF-B.
It has been reported that the expressions of various genes, including cIAP 2 , CCR6, CCR4, CXCR4, and STAT5, are up-regulated in Tax-positive ATLL cells compared with Tax-negative T cells. We therefore assessed the effects of TAK1 knockdown on the expression of these genes; however, no inhibitory effect was observed (data not shown). In contrast, knockdown of NF-B p65 resulted in the down-regulation of cIAP 2 gene expression (data not shown). This was correlated with the fact that IKK inhibitor, but not TAK1 inhibitor, induced cell death (Fig. 3E). Microarray analysis will help our understanding of the physiological function of activated TAK1 in HTLV-1-transformed cells.
Another important finding in the present study is that TAB2, but not TAB3, is overexpressed in a Tax-dependent fashion. TAB2 has been characterized as an adaptor protein that regulates the activation of TAK1. Interestingly, it has recently been reported that TAB2 localizes in the nucleus and is able to associate with a transcriptional co-repressor NcoR, an epidermal growth factor receptor family (ErbB4), and nuclear receptors such as androgen receptor (29,30). These reports suggest that TAB2 has several potential physiological functions apart from TAK1 activation. Therefore, further study from the viewpoint of TAB2 overexpression, including analysis of the transcriptional mechanism of TAB2 expression and a search for its associate proteins in HTLV-1-infected cells, will shed light on the pathophysiological importance of our findings. In addition, it is interesting to speculate whether TAB2 is overexpressed in other cancer cells.
In summary, we have provided a new insight into the pathophysiological function of TAK1 in HTLV-1-infected T cells. TAK1 has also been shown to be an oncoprotein component of the Epstein-Barr virus latent membrane protein 1 complex (31,32). These findings altogether raise the possibility that TAK1 is a key molecule in virus-related oncogenesis; however, virus infection and oncoproteins dramatically change the intracellular environment. Further basic and clinical investigations will be essential in exploring this possibility.