Enhanced Activation of Tax-dependent Transcription of Human T-cell Leukemia Virus Type I (HTLV-I) Long Terminal Repeat by TORC3*

Tax, a protein encoded by the env-pX gene of human T-cell leukemia virus type I (HTLV-I), interacts with various host cell transcription factors. Tax activates transcription from the long terminal repeat (LTR) of HTLV-I through association with cyclic AMP-responsive element-binding protein (CREB). Here, we present evidence that transducer of regulated cyclic AMP-response element-binding protein 3 (TORC3), a co-activator of CREB, is involved in Tax-induced transcriptional activation from the HTLV-I LTR. By using a luciferase assay system, we show that TORC3 alone can enhance transcription from the HTLV-I LTR, as well as from a cellular cyclic AMP-response element (CRE). Interestingly, we find that co-expression of TORC3 and Tax dramatically increased transcriptional activation at the HTLV-I LTR. We also show by glutathione S-transferase pull-down and co-immunoprecipitation experiments that TORC3 interacts with Tax. Using deletion mutant analysis, we identify the Tax interaction domain of TORC3 as a region spanning from amino acid 1 to 103, which contains a coiled-coil domain. These results provide important clues toward understanding the molecular mechanism of Tax-dependent transcriptional activation of the HTLV-I LTR.

The viral oncoprotein Tax is encoded by the env-pX gene (8) and plays a principal role in the regulation of the proliferation and transformation of HTLV-I-infected T cells. Tax is a pleiotropic factor that interacts with many cellular proteins to regulate gene expression. For example, Tax regulates the NF-B pathway using a variety of mechanisms. Tax is not only able to activate NF-B signaling by binding to NF-B, IB, and IKK␦ (9 -11) but also activates transcription of cellular genes with NF-B-binding sites in their promoter regions (12,13), such as Bcl-xL, interleukin-2, and interleukin-2 receptor, to regulate proliferation and transformation of HTLV-I-infected cells (14,15). Furthermore, Tax also directly binds to p16 ink4a , a member of the INK4 family of cyclin-dependent kinase inhibitors, and suppress its function (16 -18), which leads to aberrant cell cycle progression.
In addition to regulating expression of host cell genes, Tax also regulates transcription of viral genes by interacting directly with the cellular transcriptional factor CREB to activate the HTLV-I-LTR (19). The Tax/CREB complex binds to the Tax-responsive element (TxRE) in the HTLV-I LTR, promoting viral transcription through the recruitment of co-activators such as CBP/p300 and p300/CREB-binding protein-associated factor, a process that does not require phosphorylation of CREB (20 -22). The HTLV-I LTR is composed of three tandem repeats of TxRE, which contains a cyclic AMP-responsive element (CRE) flanked by GC-rich sequences. Although the Tax/CREB complex appears to be recruited predominantly to the TxRE, it is not recruited to the cellular CRE as to a functional complex, suggesting that the CRE-flanking sequence provides specificity for binding of various activation complexes (23). Indeed, protease sensitivity analysis demonstrated that the conformation of CREB differs according to the specific context of the CRE to which it is bound, indicating that the conformation of CREB is regulated by CRE-flanking sequences (24). Specifically, the weak association of CREB with the CRE sequence of TxRE appears to expose its Tax-binding domain, allowing easy access of Tax to CREB. Binding of Tax may then tighten the ternary complex composed of CREB, Tax, and TxRE. Analysis of this ternary complex suggested that the GC-rich sequence flanking the CRE element in TxRE is crucial for interaction with Tax since this sequence was protected by Tax in a manner that depended on the dose of CREB (25). Moreover, a drug that binds the minor groove of DNA was able to disrupt Tax binding to the CRE-flanking sequence, suggesting that the interaction of Tax with nucleotides in the minor groove of the flanking GC-rich sequence plays an important role in establishing the stable ternary complex. Interaction of protein factors with the flanking sequences in TxRE in HTLV-I-LTR in vivo has been also demonstrated by DNA protection assays (26).
However, these results do not fully explain the mechanism of Tax-dependent transcription of HTLV-I LTR. First of all, the physiological relevance of the ternary complex reconstituted by in vitro synthesized CREB and Tax may be questioned since no evidence has been presented that this complex has any biological function. Furthermore, protein complexes interacting with the flanking sequence in vivo may contain additional factors besides CREB and Tax. For example, we previously identified a cellular protein, TAXREB803/SRL300, which interacts with the 3Ј-flanking region of TxRE (27). This protein enhances Tax-dependent transcription and CREB binding to TxRE in cooperation with Tax. Thus, the precise mechanism of how Tax activation of CREB leads to transcriptional activation of the LTR remains to be elucidated.
Recently, TORCs were identified as a family of CREB coactivators that bind to CREB and enhance CRE-mediated transcription in a phosphorylation-independent manner (28,29). This transcriptional activation is facilitated by recruitment of the TATA-binding protein-associated factor 130 (TAF130) component of transcription factors IID (TFIID) to CREB. One member of the family, TORC3, is expressed at high levels in B and T lymphocytes (29). Based on these studies, we hypothesized that TORC3 might be involved in the transcriptional activation of the HTLV-I LTR by Tax. In this study, we examine the effect of TORC3 on Tax-mediated transcription regulated by the HTLV-I LTR. We find that TORC3 dramatically enhances Taxmediated viral promoter activity. Furthermore, we show that TORC3 interacts with Tax directly via the N-terminal region of TORC3, which contains a coiled-coil domain.
Cell Culture-HEK-293T cells were maintained in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin and were cultured at 37°C in a 5% CO 2 humidified chamber.
Plasmid Transfection-HEK-293T cells were grown to 50% confluence in 60-mm dishes or 24-well dishes in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. Cells were transfected with expression vectors or corresponding empty vector constructs using FuGENE 6 (Roche Applied Science) according to the manufacturer's instructions.
Antibodies-Rat anti-HA (3F10; Roche Applied Science) and mouse anti-Myc (9E10; Santa Cruz Biotechnology) antibodies were purchased commercially. A rabbit polyclonal antibody specific for Tax was raised against the bacterially expressed GST-Tax (233-353 aa) fusion protein. Horseradish peroxidase-linked goat antibodies to rat IgG were from Jackson ImmunoReseach Laboratories. Horseradish peroxidase-linked goat antibodies to mouse or rabbit IgG were from Amersham Biosciences.
GST Pull-down Assay-GST fusion proteins were expressed in Escherichia coli and isolated by an affinity column with glutathione-Sepharose-4B beads (Amersham Biosciences). GST fusion proteins bound to the resin were then incubated with [ 35 S]methionine-labeled Tax, which was synthesized by in vitro translation using the TNT-coupled transcription-translation system (Promega). The binding reactions were carried out in 600 ml of GST binding buffer (20 mM Tris-HCl (pH 8.0), 150 mM NaCl, 5% glycerol, 1% Nonidet P-40, 0.5 mM EDTA, and 1 mM dithiothreitol, supplemented with a Complete protease inhibitor mixture tablet (Roche Applied Science)) for 2 h at 4°C. The beads were washed three times with 1 ml of GST binding buffer. The bound proteins were eluted by adding 20 ml of sample buffer and resolved by SDS-PAGE and then analyzed by autoradiography.
Co-immunoprecipitation Assay-Co-immunoprecipitation assays were performed using HEK-293T cells. The cells were lysed with radioimmune precipitation buffer (50 mM Tris-HCl (pH 8.0), 50 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40, and 1 mM dithiothreitol, supplemented with a Complete protease inhibitor mixture tablet). Cell debris was removed by centrifugation for 5 min. The lysates were first cleared with protein G beads for 15 min and then incubated with 2 g of anti-Tax for 2 h at 4°C and incubated for another 1 h following the addition of 20 ml of protein G bead slurry. The beads were then washed three times with radioimmune precipitation buffer, and immunoprecipitates were resolved by SDS-PAGE and analyzed by Western blot.
Luciferase Reporter Assay-Luciferase assays were performed with the dual luciferase assay system (Promega) according to the manufacturer's instructions. Luciferase activities in cell lysates were normalized relative to the Renilla luciferase activity derived from co-transfected pRL-RSV-luc (31). All reporter assays were performed in triplicate, and standard errors (S.E.) were denoted by the bars in the figures.
RNA Interference-A duplex of 21-nucleotide small interfering RNA (siRNA) specific to mRNA for TORC3, including two nucleotides of deoxythymidine at the 3Ј end, was synthesized (Qiagen) and annealed according to the manufacturer's protocol. The sense sequence of the duplex was 5Ј-CACCAUCCUGCCAGAAGACdTdT, and the antisense sequence was 3Ј-GUCUUCUGGCAGGAUGGUGdTdT. HEK-293T cells were transfected with 25 l of 20 M duplex RNA per 6-cm dish with Oligofectamine (Invitrogen) according to the manufacturer's protocol. Twenty-four hours after transfection, the medium was replaced with the normal culture medium. For the RNA analysis, total RNA was isolated from cells 24 h after siRNA transfection and used as a template for RT-PCR. Primers used were 5Ј-AAAGAATTCGCCTTCGAGCAGCT-CATGACC-3Ј for the forward orientation and 5Ј-TTTCTCGAGGAA-GCGGTTCCTGGATGGCCT-3Ј for the reverse orientation.
RNA Isolation, RT-PCR, and Reporter Assay-Total RNA was extracted by ISOGEN (Nippon Gene) according to the manufacturer's protocol. To obtain TORC3 cDNA from the isolated RNA, reverse transcription was carried out using oligo(dT) 12-18 as a primer in a reaction volume of 20 l, in which 1 g of total RNA was present. The cDNA was subjected to PCR with Taq polymerase (Promega) using 5Ј-AAAGAAT-TCGCCTTCGAGCAGCTCATGACC-3Ј as a forward primer and 5Ј-TT-TCTCGAGGAAGCGGTTCCTGGATGGCCT-3Ј as a reverse primer. PCR was run for 25 cycles of denaturation (94°C for 30 s), annealing (52°C for 30 s), and extension (72°C for 1 min). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA was used as an internal control for the RT-PCR. The primers used to amplify GAPDH mRNA were 5Ј-ATGGGGAAGGTGAAGGTCGG-3Ј and 5Ј-TGGAGGGATCTCGCTC-CTGG-3Ј. PCR for GAPDH was conducted for 20 cycles.
For the reporter assay, at 24 h after the transfection with siRNA, the reporter plasmids were transfected together with pcDNA3-Tax by using FuGENE 6. At 24 h after the transfection of the reporter plasmids, expression of firefly luciferase from pGL3-LTR-luc and Renilla luciferase from pRL-RSV-luc was evaluated with the dual luciferase reporter assay system.

TORC3 Activates Transcription at the HTLV-I LTR as
Well as Cellular CRE-containing Promoters-Since TORC3 is a coactivator of the transcription factor CREB that enhances CREmediated transcription of target genes, we explored the effect of TORC3 on Tax-dependent transcription from the CRE-contain-ing HTLV-I LTR (Fig. 1A). We first performed a luciferase assay using a pLTR-luc reporter plasmid. We found that luciferase activity was enhanced by expression of either Tax or TORC3 (Fig. 1B) and was further enhanced by co-expression of Tax and TORC3. To determine whether TORC3 can activate CREB-dependent transcription of genes whose promoters contain a cellular CRE sequence, we transfected TORC3 into HEK-293T cells carrying the luciferase reporter plasmid, pCRE-luc, which has four repeats of the CRE consensus sequence (Fig.  1C). TORC3 alone activates pCRE-luc ϳ200-fold above untransfected control cells, whereas Tax was unable to activate transcription from pCRE-luc (Fig. 1D). However, co-expression of Tax and TORC3 in HEK-293T cells greatly enhanced the expression of luciferase from pCRE-luc (Fig. 1D), suggesting a synergistic interaction between these two factors that may promote high levels of transcription in a cellular context.
Since Tax potentiates the binding of CREB to the TxRE in the HTLV-I 5Ј-LTR to initiate viral gene expression, we next performed a reporter assay using pTxRE-luc. This reporter plasmid has in its promoter region five copies of the Taxresponsive element, which contains a core CRE element flanked by GC-rich sequences (Fig. 1C). In HEK293T cells transiently expressing TORC3 alone, transactivation of the HTLV-I LTR reached levels ϳ700-fold above background (Fig.  1B). Intriguingly, co-expression of Tax and TORC3 dramatically enhanced pTxRE-luc expression to over 5,000-fold above background (Fig. 1D). These results indicate that although TORC3 alone can enhance transcription from the HTLV-I LTR via the TxRE, Tax is able to synergize with TORC3 to further increase transcription levels.
Determination of the Region of TORC3 Important for Enhancement of Tax-mediated Transcription-To identify the region required for Tax-mediated transcription, we generated several deletion mutants of TORC3 ( Fig. 2A) and examined the effect of these mutant proteins on Tax-mediated transcription. When HEK-293T cells were co-transfected with TORC3 (1-298 aa) and Tax, the level of the reporter activity remained at two-thirds of that of cells expressing Tax and full-length TORC3 (Fig. 2B). However, when TORC3 (298 -620 aa) or TORC3 (95-620 aa), which lack the N-terminal region of TORC3, were co-transfected with Tax, the reporter activity was at levels comparable with those observed in cells expressing Tax alone (Fig. 2B). These results suggest that the N-terminal region of TORC3 is involved in Tax-mediated transactivation via the TxRE element. In fact, TORC3 (1-103 aa), which contains only the N-terminal region, showed significant activity.
TORC3 Interacts with Tax to Promote the Transactivation via the CRE Sequence-The N-terminal region of TORC3 contains a coiled-coil domain (ϳ70 aa), a motif that has often been implicated in protein-protein interactions. Since TORC3 is a CREB-interacting protein, we hypothesized that TORC3 enhancement of transcription is mediated through its association with Tax. To test this possibility, we performed a GST pulldown assay using bacterially generated GST-TORC3 fusion proteins and in vitro-translated Tax protein. As shown in Fig.  2C, GST-TORC3, but not GST alone, was able to associate with [ 35 S]methionine-labeled Tax (Fig. 2C, lanes 2 and 3). We found that Tax was able to bind to the region of TORC3 containing the coiled-coil domain (Fig. 2C, lanes 4 and 6) but was not able to interact with GST-TORC3 (95-620 aa) or GST-TORC3 (298 -620 aa), both of which lacked the coiled-coil domain (Fig. 2C,  lanes 5 and 7).
Next, to analyze the interaction of TORC3 with Tax in cells, co-immunoprecipitation experiments were performed. HEK-293T cells were co-transfected with expression plasmids encoding Tax and Myc-tagged TORC3. The immunocomplex precipitated with the anti-Tax antibody was resolved by SDS-PAGE and analyzed by Western blot using the anti-Myc antibody. A 78-kDa band corresponding to the expected size of TORC3 was detected (Fig. 2D, lane 2), suggesting that TORC3 does in fact interact with Tax in cells. This result, together with the data shown above, suggests that the interaction of TORC3 with Tax is required to enhance transcription synergistically via CRE promoter sequences.
TORC3 Functions as a Co-activator for Tax-mediated Transcriptional Activation-To examine whether TORC3 has effects on Tax-dependent transcription activity, we constructed an expression plasmid encoding a GAL4 DNA-binding domain-Tax fusion protein, GAL4-Tax, as well as a luciferase reporter plasmid, pGL3-luc, which contains five GAL4-binding sites and an SV40 minimal promoter. These plasmids were introduced into HEK-293T cells together with plasmids expressing fulllength TORC3 or TORC3 deletion mutants (Fig. 3). Expression of GAL4-Tax alone is able to activate pGL3-luc. However, coexpression of GAL4-Tax and full-length TORC3 induced a 3-fold increase in pGL3-luc activation as compared with GAL4-Tax alone. Co-expression of GAL4-Tax and TORC3 (1-298 aa) also resulted in enhanced luciferase activity, whereas we observed no synergy in cells co-expressing GAL4-Tax and TORC3 (298 -620 aa). These results indicate that the interaction between Tax and TORC3 allows TORC3 to function as a coactivator for Tax.
Gene Silencing of TORC3 by siRNA Reduces Tax-mediated Transcriptional Activation-To examine the function of TORC3 under physiological conditions, we employed siRNA technology, using a synthetic oligonucleotide that forms a duplex RNA encoding nucleotides 1638 -1646 of TORC3. Treatment with TORC3 siRNA efficiently reduced the level of TORC3 mRNA in HEK-293T cells to 10% of the control level, whereas the level of GAPDH RNA remained unchanged (Fig.  4A). TORC3 siRNA treatment also resulted in an approximately two-thirds decrease in Tax-mediated transactivation, as measured by pTxRE-luc and pLTR-luc expression (Fig. 4, B and C). The values of luciferase activity after 48 h of transfection were almost same as those observed at 24 h (data not shown). Taken together, these data suggest that TORC3 plays an important role in Tax-mediated transactivation of the HTLV-I LTR.

DISCUSSION
The identification of TORCs as CREB modulators (28) prompted us to test the effects of TORC3, which is expressed primarily in B and T lymphocytes (29), on CRE-dependent HTLV-I gene expression induced by Tax. Since the TxRE of the HTLV-I LTR contains a CRE sequence, we examined the effect of TORC3 on transcription of a reporter plasmid containing TxRE in its promoter. We found that although TORC3 was able to enhance transcriptional activation (Fig. 1D), co-expression of Tax with TORC3 resulted in a dramatic increase in expression level. Furthermore, HEK-293T cells co-transfected with Tax and a truncated version of TORC3 (1-298 aa) had a similar level of pTxRE-luc transcription as cells co-expressing fulllength TORC3 and Tax. However, co-transfection of Tax with either TORC3 (298 -620 aa) or TORC3 (95-620 aa), truncated proteins that lack the N-terminal region, yielded luciferase expression levels similar to those of cells expressing Tax alone (Fig. 2B). We further demonstrated that Tax interacts with the N-terminal region of TORC3, which contains a coiled-coil domain (Fig. 2, C and D). Together, these results suggest that TORC3 enhances CREB activity via CRE sequences through interaction with Tax.
Based on these results, we suggest that TORC3 is included in a complex composed of Tax, CREB, and the viral promoter sequence TxRE. Furthermore, the fact that TORC3 enhanced transcription of a cellular gene containing a CRE element in the promoter and that Tax was able to further enhance this . At 24 h after transfection, 50 ng of reporter plasmid (pTxRE-luc or pLTR-luc) and 100 ng of Tax plasmid with (ϩ) or without (Ϫ) were transfected into HEK-293T cells. As a control experiment, a reporter assay using pTK-luc was also conducted in the same conditions used for pTxRE-luc. At 24 h after transfection, luciferase activity was measured. Relative luciferase activities normalized to those of control cell lysates without ectopic Tax and TORC3 expression are shown. Average values of three independent experiments are shown, and standard error (S.E.) is denoted by the error bars. activation (Fig. 1D) also suggest that a Tax/CREB/TORC3 complex associates with cellular CRE sequences to induce transcription. Interestingly, previous work showed that Tax interacts weakly with CREB-2 bound to a cellular CRE motif in the somatostatin promoter (32). It is thus possible that TORC3 has the general function of enhancing transcription of genes with CRE sequences in their promoters and that the level of enhancement depends at least in part on the level of TORC3 in cells. Since TORC2 and TORC3 are predominantly expressed in lymphocytes (29), it is likely that HTLV-I-infected T lymphocytes have the potential to activate genes with CRE-containing promoters.
Furthermore, our RNA interference experiments indicated that in cells treated with TORC3 siRNA, Tax-dependent transcription was reduced to less than a third of that in cells treated by control siRNA. This result suggests that Tax requires TORC3 for transcriptional activation of genes carrying TxRE in the promoter, including the HTLV-I LTR, and that TORC3 is an essential component of the Tax/CREB/TxRE ternary complex.
The mechanism by which TORC3 and Tax act synergistically to activate CREB-mediated expression at the HTLV-I LTR remains unclear. However, it has been reported that CBP/p300 and p300/CREB-binding protein-associated factor can be recruited to the HTLV-I TxRE through direct interaction with Tax (7,21,22,33). Thus, it is possible that in vivo, a tetrameric complex of Tax/CREB/TORC3/TxRE may recruit additional coactivator complexes to regulate transcription.
Interestingly, although we identified TORC3 (1-103 aa) as the minimal Tax-binding region, we could detect only a slight enhancement in the level of Tax-mediated transcription as compared with that of TORC3 (1-298 aa) and full-length TORC3 (Fig. 2B). This observation raises questions about how TORC3 induces such a dramatic enhancement in Tax-mediated transcription. In our reporter assay using GAL4-Tax fusion protein, TORC3 (1-298 aa) caused slightly elevated activation levels (Fig. 3). Thus, it is likely that TORC3 (104 -298 aa) recruits cellular factors required for full transcriptional activation. Further clarification of the mechanism of Tax-mediated transcription from the LTR is necessary and may contribute to the development of an agent that inhibits HTLV-I replication.