Skip to main content
NCBI home page
Journal of Virology logoLink to Journal of Virology
. 1999 Apr;73(4):3258–3263. doi: 10.1128/jvi.73.4.3258-3263.1999

Interaction of Gli2 with CREB Protein on DNA Elements in the Long Terminal Repeat of Human T-Cell Leukemia Virus Type 1 Is Responsible for Transcriptional Activation by Tax Protein

Shingo Dan 1, Akira Tanimura 1, Mitsuaki Yoshida 1,*
PMCID: PMC104089  PMID: 10074179

Abstract

The long terminal repeat (LTR) of human T-cell leukemia virus type 1 (HTLV-1) has two distinct DNA elements, one copy of TRE2S and three copies of a 21-bp sequence that respond to the viral trans-activator protein, Tax. Either multiple copies of the 21-bp sequence or a combination of one copy each of TRE2S and 21-bp sequence is required for efficient trans activation by Tax. In the trans activation of multiple copies of 21-bp sequence, CREB/ATF protein plays an essential role in forming a complex with Tax. To understand the role of TRE2S in trans activation of one copy of 21-bp sequence, we examined protein binding to the DNA elements by DNA affinity precipitation assay including Gli2 protein binding to TRE2S and CREB protein binding to 21-bp sequence. Binding of CREB to a DNA probe containing both elements, TRE2S-21bp probe, was dependent on Gli2 protein under restricted conditions and was enhanced in a dose-dependent fashion by the binding of Gli2 protein to the same probe. Mutation in either element abolished the efficient binding of CREB. A glutathione S-transferase fusion protein of a fragment of Gli2 was able to bind to CREB. Therefore, Gli2-CREB interaction on the DNA probe is proposed to stabilize CREB binding to DNA. Tax can bind to CREB protein on the DNA; therefore, stabilization of DNA binding of CREB results in more recruitment of Tax onto DNA. Conversely, Tax increased the DNA binding of CREB, although it had almost no effect on the binding of Gli2. These results suggest that Gli2 binds to the DNA element and interacts with CREB, resulting in more recruitment of Tax, which in turn stabilizes DNA binding of CREB. Similar cooperation of the protein binding to TRE2S-21bp probe was also observed in nuclear extract of an HTLV-1-infected T-cell line. Consistent with the Gli2-CREB interaction on the DNA elements, Tax-mediated trans activation was dependent on the size of the spacer between TRE2S and 21-bp sequence. The effective sizes of the spacer suggest that TRE2S in the LTR would cooperate with the second and third copies of the 21-bp sequence and contribute to trans activation of the viral gene transcription.

Infection with human T-cell leukemia virus type 1 (HTLV-1) (21, 33) is etiologically associated with adult T-cell leukemia (34) and tropical spastic paraparesis (10, 19). Patients with these diseases have serum antibodies against HTLV-1 proteins, indicating persistent expression of HTLV-1 in individuals. The mechanisms of the viral gene expression and replication are thus important issues in understanding its pathogenesis and possibly controlling the diseases.

Gene expression of retroviruses is regulated by cis elements in the long terminal repeats (LTRs) at both termini of the proviral genome. HTLV-1, however, has a unique regulatory system to enhance its own gene expression: HTLV-1 transcription is trans activated by its own product, Tax (4, 6, 25, 26), responding to three repeats of 21-bp sequence in the LTR (7, 24), and TRE2S (18, 30, 31). Reconstitution experiments have revealed that at least two direct repeats of the 21-bp sequence alone are sufficient for efficient activation by Tax (7, 20, 24); however, one copy of the 21-bp sequence is activated only weakly. In the activation of multiple copies of the 21-bp sequence, Tax protein does not bind to DNA directly but binds to CREB (cyclic AMP response element binding protein), which specifically binds to the 21-bp sequence (27, 35). Tax, on the other hand, binds to CBP (CREB binding protein) and forms a transcriptionally active complex, 21bp-CREB-Tax-CBP, without phosphorylation of CREB at the specific site (16, 17, 27, 35).

On the other hand, another element termed TRE2S was proposed to contribute to trans activation induced by Tax (1). However, the TRE2S element requires the 21-bp sequence to respond to Tax protein (18, 30, 31); furthermore, TRE2S alone was totally inactive even in its multiple form (30, 31). TRE2S, therefore, seems to have a unique property as an enhancer, although enhancers are generally active in a multiple form. As binding proteins for TRE2S sequence, the Gli2 family of Gli-Krüppel family proteins with zinc finger motifs, including four isoforms (α, β, γ, and δ), was isolated (30). The Gli2 proteins show high homology with Gli1 and Gli3 in their zinc finger motifs (14, 22). The Gal4 fusion protein of Gli2 isoforms enhanced gene expression from a reporter carrying the Gal4-binding site and the 21-bp sequence in the presence of Tax, but not in the absence of Tax. These previous observations suggest that binding of Gli2 to TRE2S is involved in Tax-mediated trans activation cooperating with 21-bp sequence (31), implying that another cellular signal may control HTLV-1 gene expression in addition to that through 21-bp sequence.

To understand the mechanism of the cooperation between TRE2S and the 21-bp sequence in the LTR during trans activation induced by Tax, we analyzed the binding of Gli2, CREB, and Tax to DNA elements by DNA affinity precipitation (DNAP) assay. We demonstrated that Gli2 binding enhances the binding of CREB proteins to DNA elements, which enables recruitment of more Tax on the DNA. Interaction between Gli2 and CREB was also directly demonstrated without DNA. Consistent with the Gli2-CREB interaction on the DNA elements, the distance between TRE2S and 21-bp sequence affected the trans activation induced by Tax protein. From these results, we proposed that the binding of Gli2 to CREB bound to the respective DNA element stabilizes the complex and enhances recruitment of Tax onto the complex forming Gli2-CREB-Tax, and Tax in turn stabilizes the DNA binding of CREB protein. These protein interactions of Gli2, CREB, and Tax would be the mechanism of the trans activation of transcription in the presence of Tax.

MATERIALS AND METHODS

Cells and plasmids.

FL cells, a human amnion cell line, and 293T cells, an adenovirus-transformed human embryonic kidney cell line carrying simian virus 40 large T antigen, were maintained in Dulbecco’s modified Eagle medium supplemented with 5% fetal calf serum. Hut102 cells, a T-cell line infected with HTLV-1, were maintained in RPMI 1640 with 10% fetal calf serum. A reporter plasmid, pTK-Luc, containing a basic promoter of thymidine kinase which is linked to the luciferase gene, and pUCdN55-CAT, carrying the promoter region of HTLV-1 LTR, from which enhancers were deleted, were previously described (31). Into these basic constructs, TRE2S, consisting of 25 bp (31) and the 21-bp sequence, was inserted, constructing pTRE2S-(n)-21bp-TK-Luc. pTRE2S-TK-Luc and p21bp-TK-Luc were also constructed as controls. Another series with the HTLV-1 promoter and chloramphenicol acetyltransferase (CAT), dN55-CAT, was similarly constructed. Between the TRE2S sequence and the 21-bp sequence of these reporters, spacer sequences of specified length (see Fig. 5) were inserted. Expression vectors, pCG-Tax (8), pCG-d3 (27), pCG-Gli2β (30), and pHisT-pET (13), for bacterial expression were previously described. Mutations in TRE2S and 21-bp sequence were as follows: TRE2S, CCGGGAAGCCACCGGGAACCACCCA; TRE2M, CCGGGAAGCCACCGGGAACAAATTA; 21-bp sequence, AGGCGTTGACGACAACCCCTG; 21-M, AGGCGTACACGACAACCCCTG.

FIG. 5.

FIG. 5

Open in a new tab

(A) Alignment of Tax-responsive elements, 21-bp sequence and TRE2S, in the LTR of HTLV-1. Hatched and open boxes with arrows are 21-bp sequence and TRE2S, respectively. (B) Size and sequence of the spacer used for the assay described for panel C. (C) Effect of size of spacer on trans activation induced by Tax. □ and ⊡, luciferase activity (104 units per microgram of protein) in the absence or presence of Tax, respectively; ●, the ratio of □ to ⊡. ⦿ and ○, fold activation by Tax with a reporter carrying either 21-bp sequence or TRE2S alone, respectively. (D) Effect of relative positions of, and spacers between, TRE2S and 21-bp sequence.

Transfection and assays for CAT and luciferase.

The reporter and effector plasmids were transfected into 5 × 105 FL cells according to the calcium phosphate procedure, adjusting the total DNA to 10 μg with salmon sperm DNA as described previously (6). For the expression of Tax, 0.05 μg of pCG-Tax was cotransfected. After 40 h, cells were harvested and subjected to assay for CAT or luciferase activity as described previously (6). Under the conditions used, the activity was linearly proportional to the incubation time and the protein concentration. The assay was repeated at least twice to confirm reproducibility. CAT activity was defined as percent acetylation of chloramphenicol per 100 μg of protein in 30 min at 37°C, and luciferase activity was expressed as arbitrary units as previously described (6).

DNAP assay.

CREB and Tax used in the assays were histidine-tagged proteins produced in Escherichia coli and purified as described previously (13). Nuclear extracts (NEs) of 293T cells and Hut102 cells were prepared according to Dignam’s method (3). The expression plasmid for Gli2β was transfected into 293T cells by the calcium phosphate procedure (7). After 48 h, cells were suspended in hypotonic buffer (20 mM HEPES-NaOH [pH 7.9], 5 mM KCl, 0.5 mM MgCl2, 0.5 mM dithiothreitol [DTT], 1 mM phenylmethylsulfonyl fluoride, 1 μg of aprotinin per ml, 1 μg of pepstatin per ml, 1 μg of leupeptin per ml) and homogenized in a Dounce homogenizer. Nuclei were pelleted and extracted with a high-salt buffer (20 mM HEPES-NaOH [pH 7.9], 25% [vol/vol] glycerol, 500 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM DTT, 1 mM phenylmethylsulfonyl fluoride, 1 μg of aprotinin per ml, 1 μg of pepstatin per ml, 1 μg of leupeptin per ml). The extract was dialyzed against a buffer (20 mM HEPES-NaOH [pH 7.9], 80 mM KCl, 0.5 mM MgCl2, 10 mM ZnSO4, 0.5 mM DTT). DNA probes carrying TRE2S and/or the 21-bp sequence were generated by PCR with biotinylated primers as described previously (27).

Typical DNAP assays were carried out in a total volume of 400 μl basically as follows unless otherwise specified. DNA probe (10 ng) was incubated at 25°C for 10 min with Gli2-containing NE (100 μg of protein unless otherwise specified), His-CREB (20 ng), and/or His-Tax (10 ng). DNA-protein complex was isolated by addition of Dynabeads M-280–streptavidin (Dynal), and half or one-third of the isolated complexes was subjected to each immunoblot analysis with rabbit anti-Gli2 (31), anti-CREB (27), or anti-Tax antibodies (15). For the analysis with the HTLV-1-infected T-cell line (Fig. 5), increased doses of the components were used: 200 μg of protein of NE and 20 ng of DNA probe; furthermore, total complexes isolated were applied to a single gel for immunoblot analysis. Therefore, the sensitivity of the experiments with infected cell lines was nearly 10-fold higher than that of those with transfected cells The bands were visualized with protein A-conjugated horseradish peroxidase and an enhanced chemiluminescence detection system (Amersham Co. Ltd.). The density of each band was measured by scanning the film with BASTATION software (Fuji Film, Tokyo, Japan). The binding of proteins to DNA-protein complexes was dependent on the dose of Gli2-containing NE and also of CREB, and thus the system reflects the binding capacities of Gli2 and CREB proteins.

In vitro binding of glutathione S-transferase (GST)—Gli2 to CREB.

An expression plasmid of the fusion protein, GST-Gli2β(1–521), was constructed by inserting an N-terminal fragment of Gli2β cDNA that codes for amino acids 1 to 521 into a vector, pGEX. pGEX-Gli2β(1–521) was introduced into E. coli, and the production of GST-Gli2β(1–521) was induced by addition of 1 mM isopropyl-β-d-thiogalactoside (IPTG). The cells were collected and lysed by sonication in phosphate-buffered saline containing Triton X-100. The lysates were mixed with glutathione-conjugated Sepharose beads (Pharmacia, Piscataway, N.J.), and the complexes were isolated. The GST-Gli2β(1–521) conjugated with the beads was then incubated with purified CREB protein as previously described (12). The GST fusion proteins on the beads were collected and subjected to immunoblot analysis with anti-CREB antibody (27).

RESULTS

Cooperative DNA binding of Gli2 and CREB.

We have previously reported that a combination of TRE2S and 21-bp sequence was trans activated by Tax protein but either element alone was not (30). The proteins that bind to TRE2S are Gli2 proteins (30), new members of the Gli-Krüppel protein family related to the previously reported THP (31), and the proteins that bind to 21-bp sequence are CREB/ATF (27). The requirement for these two DNA elements in trans activation induced by Tax protein suggests specific interaction of these proteins bound to each element. To examine this possibility, DNA binding capacities of Gli2 and CREB proteins were analyzed. As a protein source of Gli2β, a long isoform of Gli2, an NE of 293T cells transfected with Gli2β expression vector was used, since production of Gli2β in E. coli was unsuccessful. A biotinylated DNA probe was incubated with increasing doses of Gli2β and CREB, and then the DNA-protein complexes were isolated by means of streptavidin beads. The isolated complexes were analyzed by immunoblotting with antibodies against Gli2 or CREB protein (Fig. 1).

FIG. 1.

FIG. 1

Open in a new tab

Cooperative binding of Gli2β and CREB proteins to a DNA probe carrying TRE2S and 21-bp sequence. (A) Effect of increasing doses of Gli2β and CREB on their binding to TRE2S-21bp DNA. (B) Mutational effect of DNA probes on the binding of Gli2β and CREB to DNA elements. Biotinylated DNA probe was incubated with purified His-CREB and an NE of 293T cells transfected with a Gli2β expression vector (NE-Gli2) or untransfected cells (NE). DNA-protein complexes were then isolated, the complexes were divided into two portions, and each was separately subjected to immunoblot analysis with rabbit anti-Gli2 or anti-CREB antibodies. The numbers 10 to 40 for NE-Gli2 or 60 to 100 for NE indicate total protein (micrograms) of each NE, and +1 and +3 for CREB roughly correspond to 20 and 50 ng of purified His-CREB, respectively. Open and hatched boxes with arrows represent TRE2S and 21-bp sequence, respectively, and those with asterisks represent their mutants. Numbers at right show molecular mass in kilodaltons.

When the DNA probes contained both TRE2S and 21-bp sequence, increasing doses of Gli2β and CREB resulted in greater binding of the respective proteins to the DNA elements (Fig. 1A). The results revealed that the assay system is dependent on the dose of each protein and thus reflects total binding capacities of the proteins in the system. NE of untransfected cells contained Gli2 and CREB proteins; however, no significant bindings of these endogenous proteins were observed (lane 1), probably due to their low levels in the reaction mixture. The binding of Gli2β was not significantly affected by increasing doses of CREB (lanes 4 to 6, 7 to 9, and 10 to 12), but the binding of CREB was enhanced by increasing doses of Gli2β (lanes 5, 8, and 11 and 6, 9, and 12). Furthermore, with a low dose of Gli2β (Fig. 1A, lanes 1 to 3), CREB did not bind to DNA at significant levels under the conditions used. Total protein in each reaction was adjusted to 100 μg by adding an NE of untransfected cells; therefore, these observations exclude any possible contribution of other endogenous proteins to the observed binding of CREB. We further suspected a possibility that Gli2β-induced proteins might be involved in the CREB binding through the TRE2S element. However, any other specific binding protein was detected in addition to Gli2β by gel electrophoresis of DNA-protein complexes followed by silver staining (data not shown). Therefore, it is highly unlikely that any Gli2β-induced protein affected the DNA binding of CREB. We thus concluded that there was cooperative binding of Gli2β and CREB to the DNA probe.

To substantiate the cooperative binding of Gli2β and CREB, a mutation was introduced into either TRE2S or the 21-bp sequence (Fig. 1B). When a mutation was introduced into TRE2S, both Gli2β and CREB no longer bound to the DNA (lanes 5 to 8). On the other hand, when the 21-bp sequence was mutated, the binding of Gli2β was not affected significantly, but the binding of CREB was markedly reduced (lanes 9 to 12). Two DNA elements, therefore, are both indispensable for the efficient binding of CREB to the DNA probe in cooperation with Gli2β.

The possible interaction of Gli2β with CREB without DNA was examined with GST fusion protein of Gli2. A fragment of Gli2β (amino acids 1 to 521) fused to the GST domain was produced in E. coli, purified, and used for the binding to CREB protein. As shown in Fig. 2, CREB protein bound to GST-Gli2β but not to the GST domain alone. These observations clearly indicate that Gli2β can bind to CREB without DNA motifs.

FIG. 2.

FIG. 2

Open in a new tab

Binding of Gli2 fragment to CREB protein without DNA. A GST fusion protein of a fragment of Gli2 containing amino acids 1 to 521 (asterisk) was incubated with CREB, and the complexes formed were isolated with glutathione-Sepharose beads. The isolated proteins were analyzed by immunoblotting with anti-CREB antibody.

Effects of Gli2 and CREB on association of Tax with DNA.

As described above, Gli2β binding to TRE2S-21bp enhanced the binding of CREB. Previously, CREB protein has been shown to recruit Tax protein onto the repeated 21-bp elements, and Tax in turn stabilizes the CREB-DNA complex (27, 35). Therefore, we examined the effect of Gli2 on recruitment of Tax onto TRE2S-21bp probe. As shown in Fig. 3, the presence of both Gli2 and CREB induced greater binding of Tax (Fig. 3A, subpanel c, lane 8) than did the presence of either protein alone (Fig. 3, lanes 6 and 7). Since CREB binding was enhanced by Gli2 (Fig. 3A, subpanel b, lanes 2 and 4 and 6 and 8), it is conceivable that Gli2 bound to TRE2S interacts with CREB, enhances CREB binding to DNA, and results in greater recruitment of Tax. Furthermore, addition of Tax to the binding system enhanced the binding of CREB to DNA (Fig. 3, lanes 2 and 4 and 6 and 8). The result was similar to the previous findings on increased binding of CREB to multiple copies of 21-bp sequence in the presence of Tax (2, 32, 35), therefore suggesting stabilization of the DNA binding of CREB by Tax protein. The proposed mechanism is further supported by examining a mutant of Tax. A mutant of Tax that cannot bind to CREB (27), d3, was not bound significantly to the DNA-protein complex even in the presence of CREB (Fig. 3A, subpanel d, lanes 11 and 12). Consistent with these observations, mutant d3 was completely inactive in trans activation of gene expression directed by TRE2S-21bp (Fig. 3B). These results suggest that recruitment of CREB and Tax onto DNA elements by Gli2 binding and stabilization of the complex by Tax is the mechanism of Tax-mediated trans activation of one copy of the 21-bp element.

FIG. 3.

FIG. 3

Open in a new tab

Binding of Gli2β, CREB, and Tax to DNA probe and trans activation of gene expression. (A) Cooperative binding of Gli2β and CREB in the absence and presence of wild-type Tax (a to c) or mutant d3 of Tax (d). Assays for DNA binding of proteins were carried out as described for Fig. 1. Plus signs for Gli2, CREB, and Tax or d3 represent roughly 100 μg of total NE and 20 and 10 ng of purified protein, respectively. One-third each of the isolated complexes was separately subjected to immunoblot analysis with anti-Gli2, anti-CREB, or anti-Tax antibody. Numbers at right show molecular mass in kilodaltons. (B) Tax-mediated trans activation of gene expression directed by TRE2S and 21-bp sequence. CAT activity is expressed as the ratio to that with reporter alone without Tax or mutant d3 of Tax. Wt, wild type.

A noteworthy result is that Tax was detected in the DNA-protein complex even without detectable binding of CREB (Fig. 3C, lane 7). This result suggests that Gli2 may be able to bind to Tax. However, the significance of the putative interaction is not convincing because the binding was variable from one assay to another, suggesting a rather weak interaction.

Cooperative binding of Gli2β, CREB, and Tax in vivo.

To confirm the cooperative binding of Gli2β, CREB, and Tax to DNA in HTLV-1-infected cells, DNAP assay was carried out with an extract from an HTLV-1-infected cell line, Hut102. The DNA probe with the 21-bp sequence alone precipitated only low levels of CREB and Tax (Fig. 4, lane 1), as expected from the previous findings. The TRE2S probe alone precipitated Gli2β and a low level of Tax but not CREB (lane 2). One might expect binding of CREB to the TRE2S probe since CREB can bind to Gli2 without 21-bp sequence, as shown in Fig. 2; however, CREB binding was not detectable in this assay (lane 2), probably due to the low level of CREB protein in the NE used. Mixed probes of the 21-bp sequence and TRE2S gave a pattern similar to the sum of lanes 1 and 2 as expected (lane 3), indicating no cooperation in the protein binding. However, when the DNA probe carried both the 21-bp sequence and TRE2S, CREB and Tax proteins were precipitated at much higher levels than with the mixed probes (4.1- and 6.0-fold, respectively) (lane 4). These increased levels of binding of CREB and Tax on the DNA probe can be explained by the cooperation of Gli2, CREB, and Tax on the DNA elements in HTLV-1-infected cells. Although possible involvement of proteins other than Gli2 cannot be ruled out rigorously in this type of experiment, the cooperative binding of Gli2, CREB, and Tax that is observed in the reconstituted system was strongly suggested also for naturally infected T cells.

FIG. 4.

FIG. 4

Open in a new tab

Binding of Gli2, CREB, and Tax proteins to TRE2S-21bp DNA probe in NE of HTLV-1-infected cell line. The NE (200 μg of protein) of Hut102 cells, an HTLV-1-infected T-cell line, was used for a DNAP assay as described above for Fig. 1, but the total of the DNA-protein complexes isolated was subjected to an immunoblot analysis with a mixture of antibodies against Gli2, CREB, and Tax proteins. The intensity of each band was estimated with BASTATION software (Fuji Film), and the ratios to those with mixed probes of 21-bp sequence and TRE2S are also indicated. (−), too low for significant estimation. Numbers at right show molecular mass in kilodaltons.

Spacing effect of TRE2S and 21-bp sequence on trans activation.

TRE2S and three copies of the 21-bp sequence are aligned in the LTR of HTLV-1 in the order 21bp-21bp-TRE2S-21bp with spacers of 27, 25, and 29 bases (20, 23) (Fig. 5A). The distances between these elements in the LTR are rather different from the 11-bp spacer in the constructs used in these assays. Therefore, we examined the effect of the distance between these two elements on the expression of the luciferase gene from the reporter, pTRE2S-(n)-21bp-TK-Luc. Series of reporters with various lengths of spacer were constructed (Fig. 5B) and transfected into FL cells with or without Tax expression vector. The luciferase activity expressed in the absence of Tax was relatively constant among the reporters, but the activity in the presence of Tax varied depending on the length of the spacer (Fig. 5C). Similar results were also obtained with another promoter with CAT, pTRE2S-(n)-21bp-dN55-CAT, which contained the HTLV-1 promoter region instead of thymidine kinase promoter and luciferase (data not shown). Therefore, the periodical effect of the spacer length was not unique to the TK-Luc reporter. These results suggest that both TRE2S and the 21-bp sequence are required in a stereospecific arrangement on the double-stranded DNA to allow the interaction of two binding proteins, Gli2 and CREB.

The results in Fig. 5C, on the other hand, show that the 29-bp spacer between TRE2S and 21-bp sequence is active in responding to Tax. This clearly suggests that TRE2S and the third copy of 21-bp sequence in the LTR are an active form in the trans activation, since they are in the same order with 29-bp spacer as in the reporter in Fig. 5B. However, the 25-bp sequence between the second copy of 21-bp sequence and TRE2S in the LTR does not fit the active form in Fig. 5C. The alignment of the elements in the reporter was different from that in the LTR; therefore, we next altered the order of TRE2S and 21-bp sequence in the reporter construct. The alteration with a 20-bp spacer reduced the trans activation markedly. However, extension of the spacer from 20 to 24 bp, which is close to the size of the 25-bp sequence in the LTR, almost restored the activity (Fig. 5D). These results support the idea that the second copy of the 21-bp sequence in the LTR also cooperates with TRE2S in trans activation.

DISCUSSION

The LTR of HTLV-1 has two types of cis-acting transcriptional elements, TRE2S and the 21-bp sequence, that respond to Tax. These elements are activated by Tax protein in either a combination of multiple copies of the 21-bp sequence (7, 20, 24) or one copy each of TRE2S and 21-bp sequence (18, 30, 31). Analyzing the second system with TRE2S-21bp, we found that Gli2 protein binds to TRE2S, enhances the binding of CREB to DNA probe, and finally results in enhanced recruitment of Tax on the DNA elements. The recruited Tax in turn stabilizes the DNA binding of CREB. We propose that this would be a possible mechanism for the Tax-mediated trans activation of one copy of the 21-bp element.

One copy itself of 21-bp sequence is not an efficient target for the binding of CREB protein, which is responsible for the trans activation (27, 35), and is not efficiently trans activated by Tax (7, 20, 24). However, when the probe has 21-bp sequence together with TRE2S, efficient binding of CREB to one copy of the 21-bp sequence is achieved by the binding of Gli2 protein to TRE2S adjacent to the 21-bp sequence. Although Gli2 is able to bind to CREB without DNA elements, Gli2 was unable to affect the CREB binding to 21-bp sequence unless TRE2S was present in cis. Therefore, interaction of Gli2 protein with CREB on the DNA elements seems to play a role in increasing the binding of CREB. Because CREB did not enhance Gli2 binding to DNA effectively, it is conceivable that Gli2 binds stably to TRE2S and then interacts with CREB that is bound unstably to the 21-bp sequence in cis.

The proposed interaction on the DNA between two proteins should be specific, since TRE2S was unable to trans activate an NF-κB binding site in a TRE2S–NF-κB construct (31a). The importance of the proposed interaction of Gli2 with CREB is supported by the finding that Tax-mediated trans activation depends on the distance between TRE2S and 21-bp sequence. Probably, a certain distance and configuration of these two elements are required for the efficient physical interaction of the respective binding proteins.

One of the principles of the Tax-mediated trans activation of 21-bp sequence, NF-κB site, and SRE (serum responsive element) is the indirect association of Tax with the DNA element through binding to CREB/ATF (27, 35), NF-κB family proteins (28, 29), and serum response factor (5, 27). In our system with TRE2S-21bp, augmentation of CREB binding to DNA by Gli2 resulted in greater recruitment of Tax on the specific DNA. It was reported previously that Tax increased the binding of CREB to multiple copies of 21-bp sequence (2, 32); therefore, it is expected that CREB binding to DNA is enhanced by Tax also in the system with TRE2S-21bp. In fact, CREB binding was enhanced by Tax as well as Gli2 in the TRE2S-21bp system. The importance of CREB-Tax interaction in the trans activation of TRE2S-21bp was also supported by examination of a mutant of Tax; a mutant of Tax, d3, which cannot bind to CREB, was unable to bind to TRE2S-21bp DNA and also unable to trans activate the transcription. Therefore, the principle of the trans activation of TRE2S-21bp would be the same as that for the 21bp-21bp system. That is, Tax protein on CREB has been proposed to recruit another transcription coactivator, CBP/p300, independently from phosphorylation of CREB protein (16). The cooperative binding of Gli2, CREB, and Tax proteins was also detected in a nuclear extract of HTLV-1-infected T cells, suggesting the cooperation of these proteins in vivo also.

It is reported that a protein in the same family as Gli2 interacts with CREB/ATF on a specific DNA and regulates transcription: YY1, a protein in the Gli-Krüppel protein family, interacts with CREB/ATF on the DNA elements and inhibits the expression of the c-fos gene (9, 11). This interaction was mediated through the zinc finger motifs (amino acids 347 to 392) of YY1 (37). The interaction domain has 68% homology with the corresponding region in the zinc finger motif (amino acids 152 to 197) of Gli2. Therefore, the corresponding domain of Gli2 might interact with CREB protein on the DNA. The inhibitory complex of YY1-CREB/ATF on the DNA is disrupted by an adenovirus-transforming protein, E1A (36), which is an activator of transcription of specific genes. In our system, we found the stabilization of the CREB-Tax complex to occur by adjacent binding of Gli2 to the DNA elements and activation of the transcription. Apparently, the basic strategies of the viral trans activation in these systems are unrelated even though similar proteins are involved in the mechanisms.

It is established that multiple copies of the 21-bp sequence are sufficient for the maximum trans activation by Tax protein. Therefore, three repeats of 21-bp sequence in the LTR could be responsible for trans activation. TRE2S in the LTR is separated by 25 bp upstream from the second 21-bp sequence and by 29 bp downstream from the third 21-bp sequence. Our data indicate that both spacers in the respective orientation are suitable for efficient trans activation and, therefore, suggest an active contribution of the TRE2S element in Tax-induced activation of viral gene expression. Therefore, Tax-mediated trans activation described here is an alternative mechanism for the activation of viral gene expression involving the binding of Gli2 and CREB.

REFERENCES

  • 1.Brady J, Jeang K T, Duvall J, Khoury G. Identification of p40x-responsive regulatory sequences within the human T-cell leukemia virus type I long terminal repeat. J Virol. 1987;61:2175–2181. doi: 10.1128/jvi.61.7.2175-2181.1987. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Brauweiler A, Garl P, Franklin A A, Giebler H A, Nyborg J K. A molecular mechanism for human T-cell leukemia virus latency and Tax transactivation. J Biol Chem. 1995;270:12814–12822. doi: 10.1074/jbc.270.21.12814. [DOI] [PubMed] [Google Scholar]
  • 3.Dignam J D, Lebovitz R M, Roeder R G. Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei. Nucleic Acids Res. 1983;11:1475–1489. doi: 10.1093/nar/11.5.1475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Felber B K, Paskalis H, Kleinman-Ewing C, Wong-Staal F, Pavlakis G N. The pX protein of HTLV-I is a transcriptional activator of its long terminal repeats. Science. 1985;229:675–679. doi: 10.1126/science.2992082. [DOI] [PubMed] [Google Scholar]
  • 5.Fujii M, Tsuchiya H, Chuhjo T, Akizawa T, Seiki M. Interaction of HTLV-I Tax1 with p67SRF causes the aberrant induction of cellular immediate early genes through CArG boxes. Genes Dev. 1992;6:2066–2076. doi: 10.1101/gad.6.11.2066. [DOI] [PubMed] [Google Scholar]
  • 6.Fujisawa J, Seiki M, Kiyokawa T, Yoshida M. Functional activation of the long terminal repeat of human T-cell leukemia virus type I by a trans-acting factor. Proc Natl Acad Sci USA. 1985;82:2277–2281. doi: 10.1073/pnas.82.8.2277. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Fujisawa J, Seiki M, Sato M, Yoshida M. A transcriptional enhancer sequence of HTLV-I is responsible for trans-activation mediated by p40 chi HTLV-I. EMBO J. 1986;5:713–718. doi: 10.1002/j.1460-2075.1986.tb04272.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Fujisawa J, Toita M, Yoshimura T, Yoshida M. The indirect association of human T-cell leukemia virus tax protein with DNA results in transcriptional activation. J Virol. 1991;65:4525–4528. doi: 10.1128/jvi.65.8.4525-4528.1991. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Gedrich R W, Engel D A. Identification of a novel E1A response element in the mouse c-fos promoter. J Virol. 1995;69:2333–2340. doi: 10.1128/jvi.69.4.2333-2340.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Gessain A, Barin F, Vernant J C, Gout O, Maurs L, Calender A, de The G. Antibodies to human T-lymphotropic virus type-I in patients with tropical spastic paraparesis. Lancet. 1985;ii:407–410. doi: 10.1016/s0140-6736(85)92734-5. [DOI] [PubMed] [Google Scholar]
  • 11.Gualberto A, LePage D, Pons G, Mader S L, Park K, Atchison M L, Walsh K. Functional antagonism between YY1 and the serum response factor. Mol Cell Biol. 1992;12:4209–4214. doi: 10.1128/mcb.12.9.4209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hirai H, Fujisawa J, Suzuki T, Ueda K, Muramatsu M, Tsuboi A, Arai N, Yoshida M. Transcriptional activator Tax of HTLV-1 binds to the NF-kappa B precursor p105. Oncogene. 1992;7:1737–1742. [PubMed] [Google Scholar]
  • 13.Hoffmann A, Roeder R G. Purification of his-tagged proteins in non-denaturing conditions suggests a convenient method for protein interaction studies. Nucleic Acids Res. 1991;19:6337–6338. doi: 10.1093/nar/19.22.6337. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kinzler K W, Vogelstein B. The GLI gene encodes a nuclear protein which binds specific sequences in the human genome. Mol Cell Biol. 1990;10:634–642. doi: 10.1128/mcb.10.2.634. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Kiyokawa T, Kawaguchi T, Seiki M, Yoshida M. Association of the pX gene product of human T-cell leukemia virus type-I with nucleus. Virology. 1985;147:462–465. doi: 10.1016/0042-6822(85)90149-7. [DOI] [PubMed] [Google Scholar]
  • 16.Kwok R P, Laurance M E, Lundblad J R, Goldman P S, Shih H, Connor L M, Marriott S J, Goodman R H. Control of cAMP-regulated enhancers by the viral transactivator Tax through CREB and the co-activator CBP. Nature. 1996;380:642–646. doi: 10.1038/380642a0. [DOI] [PubMed] [Google Scholar]
  • 17.Laurance M E, Kwok R P, Huang M S, Richards J P, Lundblad J R, Goodman R H. Differential activation of viral and cellular promoters by human T-cell lymphotropic virus-1 tax and cAMP-responsive element modulator isoforms. J Biol Chem. 1997;272:2646–2651. doi: 10.1074/jbc.272.5.2646. [DOI] [PubMed] [Google Scholar]
  • 18.Marriott S J, Boros I, Duvall J F, Brady J N. Indirect binding of human T-cell leukemia virus type I tax1 to a responsive element in the viral long terminal repeat. Mol Cell Biol. 1989;9:4152–4160. doi: 10.1128/mcb.9.10.4152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Osame M, Usuku K, Izumo S, Ijichi N, Amitani H, Igata A, Matsumoto M, Tara M. HTLV-1 associated myelopathy, a new clinical entity. Lancet. 1986;i:1031–1032. doi: 10.1016/s0140-6736(86)91298-5. . (Letter.) [DOI] [PubMed] [Google Scholar]
  • 20.Paskalis H, Felber B K, Pavlakis G N. Cis-acting sequences responsible for the transcriptional activation of human T-cell leukemia virus type I constitute a conditional enhancer. Proc Natl Acad Sci USA. 1986;83:6558–6562. doi: 10.1073/pnas.83.17.6558. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Poiesz B J, Ruscetti F W, Gazdar A F, Bunn P A, Minna J D, Gallo R C. Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T-cell lymphoma. Proc Natl Acad Sci USA. 1980;77:7415–7419. doi: 10.1073/pnas.77.12.7415. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Ruppert J M, Vogelstein B, Arheden K, Kinzler K W. GLI3 encodes a 190-kilodalton protein with multiple regions of GLI similarity. Mol Cell Biol. 1990;10:5408–5415. doi: 10.1128/mcb.10.10.5408. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Seiki M, Hattori S, Hirayama Y, Yoshida M. Human adult T-cell leukemia virus: complete nucleotide sequence of the provirus genome integrated in leukemia cell DNA. Proc Natl Acad Sci USA. 1983;80:3618–3622. doi: 10.1073/pnas.80.12.3618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Shimotohno K, Takano M, Teruuchi T, Miwa M. Requirement of multiple copies of a 21-nucleotide sequence in the U3 regions of human T-cell leukemia virus type I and type II long terminal repeats for trans-acting activation of transcription. Proc Natl Acad Sci USA. 1986;83:8112–8116. doi: 10.1073/pnas.83.21.8112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Sodroski J, Rosen C, Goh W C, Haseltine W. A transcriptional activator protein encoded by the x-lor region of the human T-cell leukemia virus. Science. 1985;228:1430–1434. doi: 10.1126/science.2990028. [DOI] [PubMed] [Google Scholar]
  • 26.Sodroski J G, Rosen C A, Haseltine W A. Trans-acting transcriptional activation of the long terminal repeat of human T lymphotropic viruses in infected cells. Science. 1984;225:381–385. doi: 10.1126/science.6330891. [DOI] [PubMed] [Google Scholar]
  • 27.Suzuki T, Fujisawa J I, Toita M, Yoshida M. The trans-activator tax of human T-cell leukemia virus type 1 (HTLV-1) interacts with cAMP-responsive element (CRE) binding and CRE modulator proteins that bind to the 21-base-pair enhancer of HTLV-1. Proc Natl Acad Sci USA. 1993;90:610–614. doi: 10.1073/pnas.90.2.610. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Suzuki T, Hirai H, Fujisawa J, Fujita T, Yoshida M. A trans-activator Tax of human T-cell leukemia virus type 1 binds to NF-kappa B p50 and serum response factor (SRF) and associates with enhancer DNAs of the NF-kappa B site and CArG box. Oncogene. 1993;8:2391–2397. [PubMed] [Google Scholar]
  • 29.Suzuki T, Hirai H, Yoshida M. Tax protein of HTLV-1 interacts with the Rel homology domain of NF-kappa B p65 and c-Rel proteins bound to the NF-kappa B binding site and activates transcription. Oncogene. 1994;9:3099–3105. [PubMed] [Google Scholar]
  • 30.Tanimura A, Dan S, Yoshida M. Cloning of novel isoforms of the human Gli2 oncogene and their activities to enhance tax-dependent transcription of the human T-cell leukemia virus type 1 genome. J Virol. 1998;72:3958–3964. doi: 10.1128/jvi.72.5.3958-3964.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Tanimura A, Teshima H, Fujisawa J, Yoshida M. A new regulatory element that augments the Tax-dependent enhancer of human T-cell leukemia virus type 1 and cloning of cDNAs encoding its binding proteins. J Virol. 1993;67:5375–5382. doi: 10.1128/jvi.67.9.5375-5382.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31a.Tanimura, A., and M. Yoshida. Unpublished observations.
  • 32.Yin M J, Gaynor R B. Complex formation between CREB and Tax enhances the binding affinity of CREB for the human T-cell leukemia virus type 1 21-base-pair repeats. Mol Cell Biol. 1996;16:3156–3168. doi: 10.1128/mcb.16.6.3156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Yoshida M, Miyoshi I, Hinuma Y. Isolation and characterization of retrovirus from cell lines of human adult T-cell leukemia and its implication in the disease. Proc Natl Acad Sci USA. 1982;79:2031–2035. doi: 10.1073/pnas.79.6.2031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Yoshida M, Seiki M, Yamaguchi K, Takatsuki K. Monoclonal integration of human T-cell leukemia provirus in all primary tumors of adult T-cell leukemia suggests causative role of human T-cell leukemia virus in the disease. Proc Natl Acad Sci USA. 1984;81:2534–2537. doi: 10.1073/pnas.81.8.2534. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zhao L J, Giam C Z. Human T-cell lymphotropic virus type I (HTLV-I) transcriptional activator, Tax, enhances CREB binding to HTLV-I 21-base-pair repeats by protein-protein interaction. Proc Natl Acad Sci USA. 1992;89:7070–7074. doi: 10.1073/pnas.89.15.7070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Zhou Q, Engel D A. Adenovirus E1A243 disrupts the ATF/CREB-YY1 complex at the mouse c-fos promoter. J Virol. 1995;69:7402–7409. doi: 10.1128/jvi.69.12.7402-7409.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Zhou Q, Gedrich R W, Engel D A. Transcriptional repression of the c-fos gene by YY1 is mediated by a direct interaction with ATF/CREB. J Virol. 1995;69:4323–4330. doi: 10.1128/jvi.69.7.4323-4330.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Journal of Virology are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES