c-sis/PDGF-B promoter transactivation by the Yax protein of human T-cell leukemia virus type 1.

The human c-sis proto-oncogene promoter is transactivated by the human T-cell leukemia virus type 1 Tax protein in human Jurkat T-cells. Transactivation was >7-fold in Jurkat cells stably expressing the Tax protein (Jurkat-Tax) than in Jurkat E6.1 cells and was further enhanced in Jurkat-Tax cells stimulated with 12-O-tetradecanoylphorbol-13-acetate and the calcium ionophore, ionomycin. Deletion analysis showed that a 167-base pair promoter fragment retained full Tax responsiveness. Insertion of this minimal Tax-responsive region into a heterologous, minimal promoter resulted in approximately a 7-fold increase of transcriptional activation in the presence of Tax. Linker-scanning insertion analysis of this region identified Tax-responsive elements at nucleotides -64 to -45 (TRE1) and -34 to -15 (TATA box region). TRE1 contains a consensus binding site for the Sp family of transcription factors. The TATA box region corresponds to the TATA box and its 3'-neighboring sequence. Gel-shift and antibody supershift analysis of TRE1-binding proteins in unstimulated Jurkat E6.1 and Jurkat-Tax nuclear extracts identified Sp1 and Sp3 as the main TRE1 binding factors. Nuclear extracts from stimulated Jurkat E6.1 and Jurkat-Tax cells identified an additional TRE1 binding factor, Egr-1. These studies define a novel mechanism whereby Tax transactivates the c-sis promoter.


The human c-sis proto-oncogene promoter is transactivated by the human T-cell leukemia virus type 1 Tax protein in human Jurkat T-cells. Transactivation was
Human T-cell leukemia virus type 1 (HTLV-1) 1 is the etiologic agent of adult T-cell lymphoma/leukemia (1) and transforms normal human T lymphocytes in vitro (2,3). The cellular events whereby HTLV-1 infection leads to T-cell transformation are not clearly understood. Unlike many other acutely transforming retroviruses, the HTLV-1 genome does not encode an oncogene homologous to a cellular sequence (4). This suggests that HTLV-1 transformation occurs through another mechanism mediated by a virus-produced regulatory protein(s) which may transactivate some cellular gene(s) involved in cell proliferation. Indeed, the HTLV-1 regulatory protein Tax is a potent transactivator of the HTLV-1 long terminal repeat (LTR) and numerous cellular genes, including IL-2 (5), IL-2R-␣ (5), granulocyte macrophage/colony-stimulating factor (5), transforming growth factor ␤ (5), c-fos (5), and c-sis (6). Tax does not appear to bind DNA directly (7,8). Rather, recent evidence indicates that Tax activates transcription by inducing or modifying the activity of certain host transcription factors, including members of the activating transcription factor/cAMP response element binding protein (ATF/CREB) family of proteins, serum response factor, Fos-Jun, and NF-B (9 -16).
T-cells infected with HTLV-1 express high levels of transcripts for the c-sis proto-oncogene (17,18), which encodes the B-chain of platelet-derived growth factor (PDGF) (19,20). Expression of the c-sis gene is restricted to selected cell types, which include activated monocytes (21) and megakaryocytes (22), developing placenta (23), and vascular smooth muscle cells (24). It is not normally expressed, however, in lymphocytes. PDGF is a potent mitogen for cells of mesenchymal origin (25). Biologically active PDGF is a dimeric protein consisting of homo-and heterodimeric combinations of two polypeptide chains, A and B (26). The major function of PDGF is to induce mitosis in quiescent target cells. PDGF exerts its effects through binding to two types of receptors, the ␣ receptor, which binds both A and B chains with high affinity, and the ␤ receptor, which binds only the B chain (27). PDGF was first implicated in the process of transformation when one of its peptide chains, the B-chain/c-sis, was found to be homologous to the viral sis oncogene (v-sis) (19,20). Expression of a recombinant, wild-type human c-sis/PDGF-B gene in mouse 3T3 cells, which express both the ␣ and ␤ PDGF receptors, resulted in the transformation of these cells (28). Interestingly, it has also been demonstrated that HTLV-1-infected T-cells express high levels of PDGF-␤ receptor transcripts and synthesize protein that can be immunoprecipitated with antibodies specific for the PDGF receptor that binds the PDGF-B homodimer and the PDGF-AB heterodimer (17). These findings raise the possibility that cells which constitutively synthesize both a mitogenic growth factor and its receptor might acquire an autostimulatory mechanism that does not necessarily require secretion of the mitogenic ligand (29,30).
With regard to the regulatory mechanism(s) that underlies c-sis/PDGF-B expression in HTLV-1-infected T-cells, a previous study provided preliminary deletion analysis of the 5Јflanking region and demonstrated that activation was due to the HTLV-1 regulatory protein Tax (6). In the current study, we have analyzed the Tax-mediated transactivation of the c-sis/ PDGF-B promoter in human Jurkat T-cells and a Jurkat T-cell line stably expressing the Tax protein (Jurkat-Tax) (31). We prepared a series of 5Ј-promoter deletion mutants and a series of linker-scanning insertion mutants and used them to identify two sites, Tax-responsive element 1 (TRE1) and the TATA box region, essential for Tax-mediated transactivation. Gel-shift and antibody supershift analysis of Jurkat E6.1 and Jurkat-Tax cells, along with the HTLV-1-infected T-cell line, HUT102, showed preferential binding of three nuclear proteins to a site within the TRE1 element.
Plasmids-pSISCAT has been described previously (32). 5Ј-pSISCAT promoter-deletion mutants were constructed by polymerase chain reaction amplification of the designated promoter regions using specific 5Ј (forward) XbaI-linkered oligonucleotides and a common 3Ј (reverse) PstI-linkered oligonucleotide as primers. The resulting promoter fragments were digested with both XbaI and PstI and cloned into the XbaI and PstI sites of pSISCAT. ptSmFNCAT minimal promoter constructs were made by polymerase chain reaction amplification of the designated promoter regions using specific 5Ј (forward) XbaI-linkered oligonucleotides and a common 3Ј (reverse) XbaI-linkered oligonucleotide as primers. The resulting promoter fragments were digested with XbaI and cloned in the forward and reverse orientations into the XbaI site of pTA-FN-CAT (a generous gift of Douglas C. Dean) (33). phSmFNCAT minimal promoter constructs were made by cloning annealed oligonucleotides containing the minimal fibronectin promoter region Ϫ28 to ϩ8 (5Ј-atcCCCATATAAGCCCGGCTCCCGCGCTCCGACGCCCGCtgca-3Ј (sense strand), 5Ј-GCGGGCGTCGGAGCGCGGGAGCCGGGCTTATA-TGGGgat-3Ј (antisense strand)) into the EcoRV and PstI sites of the ptSmFNCAT minimal promoter constructs (terminal nucleotides shown in lowercase type were added to create EcoRV and PstI sites). pRALuc and pRALuc linker-scanning mutants have been described previously (34,35). The Tax expression plasmid, pcTax, was a generous gift from Warner C. Greene (36). pHTLV1CAT (6) and pCAT (37) have been described previously. All mutants were sequenced by the dideoxy chain termination method (U.S. Biochemical).
Transfections, CAT Assays, and Luciferase Assays-5 ϫ 10 6 Jurkat E6.1 or 5 ϫ 10 6 Jurkat-Tax cells were transfected with either 20 g of chloramphenicol acetyltransferase (CAT) reporter plasmid or 20 g of luciferase reporter plasmid, plus, where indicated, 20 g of a Tax expression plasmid, using the DEAE-dextran method as described previously (38). When stimulated, 36 h post-transfection the cells were divided equally into two flasks; one flask was supplemented with TPA (10 ng/ml) and ionomycin (0.4 g/ml), and the other flask received the same volume of solvent. Thirty-six hours after stimulation, the cells were harvested by centrifugation, washed with phosphate-buffered saline and cell lysates were prepared by three cycles of freeze-thawing in an ethanol/dry ice bath and 37°C water bath. Cell extracts were normalized for protein content by a commercially available kit (Bio-Rad). Equal amounts of protein were used in CAT assays (32) and luciferase assays (39, 40) as described previously.

Identification of the Minimal Sequence Necessary for Tax
Responsiveness in the c-sis/PDGF-B Promoter-A previous study provided preliminary deletion analysis of the c-sis/ PDGF-B promoter in human Jurkat T-cells and showed that a 406-bp fragment (containing 386 bp 5Ј to the mRNA initiation site, as well as 16 bp 3Ј to the mRNA initiation site) fused to the chloramphenicol acetyltransferase (CAT) reporter gene (pSIS-CAT) retained full Tax responsiveness (6). As shown in Fig. 1, when pSISCAT was transiently transfected into Jurkat E6.1 cells either alone or cotransfected with a Tax expression plasmid, CAT activity was increased Ͼ7-fold in the presence of Tax. A Ͼ7-fold increase in CAT activity was also observed when pSISCAT was transiently transfected into Jurkat-Tax cells compared with that observed with Jurkat E6.1 cells in the absence of Tax. Similar results were obtained with the positive control vector pHTLV1CAT. (Fig. 1). As a result, instead of cotransfecting Jurkat E6.1 cells with a Tax-expression plasmid, we decided to use Jurkat-Tax cells to further investigate the Tax-mediated transactivation of the c-sis/PDGF-B promoter.
To define more precisely the minimal sequence required for Tax-mediated transactivation, a series of 5Ј-promoter deletion mutants, generated from pSISCAT, was constructed (Fig. 2). These constructs were transfected into either Jurkat E6.1 or Jurkat-Tax cells and CAT activity was measured. Constructs d-242 and d-151 retained full Tax-responsiveness. Deletions downstream of d-151, however, markedly attenuated Tax responsiveness (Fig. 2). These data identify the minimal Taxresponsive region as that spanned by d-151.
Enhancement of pTA-FN-CAT Expression by the Minimal Tax-responsive Region of the c-sis/PDGF-B Promoter-To investigate whether the region we identified as the minimal Tax-responsive region was capable of conferring Tax-responsiveness onto a heterologous, minimal promoter, the region spanning nucleotides Ϫ151 to Ϫ46 (just upstream of the c-sis/ PDGF-B promoter TATA box) was inserted into the CAT expression plasmid pTA-FN-CAT, in both the forward and reverse orientations (Fig. 3A). In its minimal promoter, the pTA-FN-CAT expression plasmid contains only the fibronectin gene promoter TATA box and RNA initiator sequence. The distance between nucleotide Ϫ46 of the c-sis/PDGF-B promoter and the fibronectin gene promoter TATA box is identical with the distance between nucleotide Ϫ46 and the TATA box of the wildtype c-sis/PDGF-B promoter. As shown in Fig. 3B, when inserted in the forward orientation (phSmFNCAT-151/-46F), the wild-type c-sis/PDGF-B promoter oligonucleotide Ϫ151 to Ϫ46 increased the level of activity approximately 7-fold in the presence of Tax. When inserted in the reverse orientation (phSmFNCAT-151/-46R), the level of activity was increased approximately 6-fold in the presence of Tax. Activity was further enhanced in the presence of Tax when the cells were stimulated with TPA and the calcium ionophore, ionomycin. In contrast, when the distance between nucleotide Ϫ46 of the c-sis/PDGF-B promoter and the fibronectin gene promoter TATA box was increased by 41 nucleotides compared with that of nucleotide Ϫ46 and the TATA box of the wild-type c-sis/ PDGF-B promoter, Tax responsiveness was completely abolished (data not shown). Thus, it appears that the observed Tax responsiveness is distance/spacing-dependent. These results indicated that the Ϫ151 to Ϫ46 region of the c-sis/PDGF-B promoter served the function of a Tax-responsive element.
Tax-mediated Transactivation of the c-sis/PDGF-B Promoter Is Further Enhanced in Stimulated Cells-The observed increase in Tax-mediated transactivation of the minimal Taxresponsive region constructs upon stimulation prompted us to investigate the effect of stimulation on Tax-mediated transactivation of the wild-type c-sis/PDGF-B promoter. pRALuc (a vector in which the c-sis/PDGF-B promoter had been previously fused to the luciferase reporter gene) was transiently transfected into either Jurkat E6.1 or Jurkat-Tax cells, in the presence or absence of stimulation with TPA and ionomycin. Luciferase activity was then measured. As shown in Fig. 4, stimulation of Jurkat E6.1 cells resulted in approximately a 2-3-fold increase in luciferase activity, whereas stimulation of Jurkat-Tax cells resulted in a Ͼ30-fold increase in luciferase activity.
Localization of the Position(s) Contained within the Taxresponsive Region Necessary for Tax-mediated Transactivation-To identify the exact position(s) within the Tax-responsive region of the c-sis/PDGF-B promoter necessary for Taxmediated transactivation, we conducted saturation mutation of this region by systematic, linker-scanning substitution. We created a series of systematic linker-scanning mutants generated from pRALuc. In these mutants, an invariant 10-bp oligonucleotide sequence (AACGATCGAT) containing a PvuI restriction site was used to substitute consecutive segments of the wild-type c-sis/PDGF-B promoter from nucleotides Ϫ254 to Ϫ15. Linker-scanning mutants corresponding to the region Ϫ154 to Ϫ15 were transiently transfected into either Jurkat E6.1 or Jurkat-Tax cells, and the levels of luciferase activity were measured. As shown in Fig. 5, we observed two regions that were essential for Tax-mediated transactivation. The first region, ablated by mutants Ϫ64/Ϫ55 and Ϫ54/Ϫ45, was designated Tax-responsive element 1 (TRE1). The second region, ablated by mutants Ϫ34/Ϫ25 and Ϫ24/Ϫ15, included the TATA box and its 3Ј-neighboring sequence. When compared with wild-type level of activity, mutation of either TRE1 or the TATA box region resulted in substantial reductions in Taxmediated luciferase expression (Fig. 5). Similar results were obtained using stimulated cells (data not shown). The level of Tax-mediated luciferase expression did not differ significantly from wild-type when linker-scanning mutants corresponding to the region located between nucleotides Ϫ154 and Ϫ94 were tested (data not shown).

TRE1 Binds Nuclear Factors in Human
Leukemic T-cell Lines-Identification of the TRE1 site in the c-sis/PDGF-B promoter prompted us to investigate whether this element bound nuclear factors. A set of double-stranded oligonucleotides (corresponding to nucleotides Ϫ83 to Ϫ45) containing the TRE1 site was prepared. The TRE1 oligonucleotide was labeled and used as a probe for in vitro electrophoretic mobility shift assays (EMSAs). Four major EMSA complexes were observed after the TRE1 probe was incubated with nuclear extracts prepared from Jurkat E6.1 (Fig. 6, lanes 2 and 3) and Jurkat-Tax cells (lanes 4 and 5). Three of the major EMSA complexes (C1, C2, and C4) were seen with nuclear extracts prepared from both unstimulated Jurkat E6.1 (lane 2) and Jurkat-Tax cells (lane 4) and with extracts prepared from TPA/ionomycintreated Jurkat E6.1 (lane 3) and Jurkat-Tax cells (lane 5). However, the major remaining EMSA complex observed (C3) with extracts from Jurkat E6.1 and Jurkat-Tax cells (lanes 3 and 5), appeared only upon stimulation with TPA and ionomycin. Interestingly, when nuclear extracts prepared from unstimulated HUT102 cells were incubated with the TRE1 probe, the same four major EMSA complexes were observed (C1-C4) as with unstimulated and stimulated extracts from Jurkat E6.1 and Jurkat-Tax cells (lane 6). In addition, a fifth major EMSA complex (C5), which migrated between C3 and C4, was observed only with extract prepared from HUT102 cells (lane 6). These EMSA complexes were specific for the probe, since their formation was blocked by unlabeled TRE1 probe using extract prepared from stimulated Jurkat-Tax cells (Fig. 7, compare lane 3 versus lane 2). Similar results were obtained using extracts prepared from unstimulated Jurkat E6.1, Jurkat-Tax, and HUT102 cells, as well as stimulated Jurkat E6.1 cells (data not shown).
Due to the fact that the TRE1 probe contained 19 nucleotides upstream of the Ϫ64 to Ϫ45 TRE1 region itself, it was possible that one or more of the EMSA complexes were binding to this region. To investigate this possibility, we conducted a competition experiment using an unlabeled probe containing only the TRE1 region itself (Ϫ67 to Ϫ42). As shown in Fig. 7, formation of the EMSA complexes was blocked when extract prepared from stimulated Jurkat-Tax cells was incubated with unlabeled Ϫ67/Ϫ42 probe (compare lane 4 versus lane 2). Similar results were obtained using extracts prepared from unstimulated Jurkat E6.1, Jurkat-Tax, and HUT102 cells, along with stimulated Jurkat E6.1 cells (data not shown). These results indicated that the major EMSA complexes were indeed binding Tax Activation of c-sis to the TRE1 region itself.
Identification of Sp1, Sp3, and Egr-1 as the Main TRE1 Binding Factors in Human Leukemic T-cell Line Nuclear Extracts-Analysis of the nucleotide sequence within TRE1 revealed the presence of a CACCC regulatory motif. Previous work has demonstrated that this motif represents a cis-acting element for the Sp family of transcription factors (42)(43)(44)(45). To investigate whether members of the Sp family of transcription factors were present in any of the EMSA complexes obtained with the TRE1 probe mentioned above, we performed a competition experiment using an unlabeled consensus Sp family probe. As shown in Fig. 7, formation of complexes C1, C2, and C4, but not C3, was blocked when extract prepared from stimulated Jurkat-Tax cells was incubated with unlabeled consensus Sp family probe (compare lane 5 versus lane 2). Similar results were obtained using extracts prepared from unstimulated Jurkat E6.1, Jurkat-Tax, and HUT102 cells, along with stimulated Jurkat E6.1 cells (data not shown). We next used antibodies raised against the Sp family members, Sp1 and Sp3, in EMSA-antibody supershift assays. A monoclonal antibody specific for human Sp1 was able to supershift complex C1 when extract prepared from stimulated Jurkat-Tax cells was pre-treated with the Sp1 antiserum (lane 6). In addition, a polyclonal antibody specific for human Sp3 was able to supershift both complex C2 and complex C4 (lane 7). Similar results were obtained using extracts prepared from unstimulated Jurkat E6.1, Jurkat-Tax, and HUT102 cells, along with stimulated Jurkat E6.1 cells (data not shown).
Since formation of complex C3 occurred only upon stimulation with TPA and ionomycin and could not be blocked with unlabeled consensus Sp family probe, we reasoned that it might contain a member of the immediate early response genes that are transiently activated in stimulated cells. Previous studies have demonstrated that binding sites for the transcription factor Sp1 often also contain overlapping, cryptic binding sites for the early growth response factor Egr-1 (46 -50). A polyclonal antibody specific for human Egr-1 was able to supershift complex C3 when extract prepared from stimulated Jurkat-Tax cells was pretreated with the Egr-1 antiserum (Fig.   FIG. 5. Relative luciferase activity of the c-sis/PDGF-B promoter linkerscanning insertion constructs in Jurkat E6.1 cells versus Jurkat-Tax cells. 20 g of each reporter construct was transiently transfected into either Jurkat E6.1 or Jurkat-Tax cells. The cells were subsequently lysed and assayed for luciferase activity as described under "Materials and Methods." After subtraction of background p0Luc activity from all of the reporter constructs, pRALuc activity in Jurkat-Tax cells was arbitrarily given a value of 1, and the activities of the other transfections were adjusted relative to this activity. The region of linker-scanning insertion within the promoter is represented by a black shaded box within each construct. The precise location of each linker-scanning insertion is indicated as the number of base pairs upstream of the mRNA initiation site. Error bars represent 1 S.D. calculated from at least two independent experiments. 7, lane 8). Similar results were obtained using extracts prepared from unstimulated HUT102 cells and from stimulated Jurkat E6.1 cells (data not shown). These results support the conclusion that Sp1, Sp3, and Egr-1 are the main TRE1 binding factors of the c-sis/PDGF-B promoter in Jurkat and HUT102 human leukemic T-cell lines. DISCUSSION In this study, we have examined the HTLV-1 Tax-mediated transactivation of the c-sis/PDGF-B promoter in human Jurkat T-cells by deletion, linker-scanning substitution, and gel shift analysis. By transient transfection analysis of 5Ј-promoter deletion mutants, a sequence consisting of 151 bp upstream and 16 bp downstream of the mRNA initiation site was found to retain full Tax responsiveness (Fig. 2). Insertion of this minimal Tax-responsive region, both in the forward and reverse orientations, into a heterologous, minimal expression vector resulted in a Ͼ6-fold increase in transcription in the presence of Tax. Tax responsiveness was further enhanced upon stimulation with TPA and ionomycin (Figs. 3 and 4).
To further examine the minimal Tax-responsive region of the c-sis/PDGF-B promoter, we constructed a series of linker-scanning mutants in which each plasmid received a 10-bp substitution linker sequence (containing a PvuI restriction site) at a single site within this region (Fig. 5). There is a 2-fold advantage in using linker-scanning mutants to identify regulatory elements. First, wild-type DNA sequence and promoter architecture are preserved throughout the entire promoter region being analyzed except for the 10-bp substitution site, leaving 5Ј regions intact. Second, since the substitution linker was the same in each mutant, this served to minimize any extrinsic effect(s) attributable to the substitution linker sequence.
When we analyzed the linker-scanning mutants for Tax responsiveness, we identified two cis-acting elements necessary for Tax-mediated transactivation. The first element, which we named TRE1, was located between nucleotides Ϫ64 and Ϫ45. TRE1 was shown to contain a consensus binding sequence for the Sp family of transcription factors, CACCC. This same 20nucleotide sequence was also identified (34) as essential for the activation of the c-sis/PDGF-B promoter that is observed in TPA-treated K562 cells as they differentiate into megakaryocytes. The second element, located between nucleotides Ϫ34 and Ϫ15, corresponded to the TATA box and its 3Ј-neighboring sequence of the c-sis/PDGF-B promoter. This result indicated that the Tax-mediated increase of transcription in Jurkat cells was initiated within the initiator region of the c-sis/PDGF-B promoter and was dependent upon the TATA-binding-RNA initiation complex.
It is interesting to note that when promoter elements previously reported to be critical for c-sis/PDGF-B expression in bovine aortic endothelial cells and human umbilical vein endothelial cells (51) were mutated by linker-scanning substitution mutants Ϫ93 to Ϫ84, Ϫ84 to Ϫ75, and Ϫ74 to Ϫ65, Taxmediated transactivation in Jurkat cells was not substantially decreased (Fig. 5). The Ϫ93 to Ϫ84 region contains an AP-1-like consensus binding sequence (Ϫ92 to Ϫ86), while the Ϫ84 to Ϫ65 region contains a consensus binding site for the Ets family of transcription factors (Ϫ78 to Ϫ68). The results observed with the Ϫ84 to Ϫ65 region were surprising in light of the fact that several reports have demonstrated the importance of Ets consensus binding sites for the Tax-mediated transactivation of the HTLV-1 LTR (11,52,53). In addition, it has been shown that multiple Ets family members are present in resting and activated human T-cells (54 -57).
EMSA and antibody supershift analysis showed preferential binding of three nuclear proteins, Sp1, Sp3, and Egr-1, to the TRE1 element (Fig. 7, lanes 6 -8). Since TRE1 was shown to contain a consensus binding sequence for the Sp family of transcription factors, it was not a surprise that the Sp family members, Sp1 and Sp3, were identified as two of the three main TRE1 binding factors. This same CACCC regulatory motif has also been identified in a Tax-responsive element of the HTLV-1 LTR (52). The binding of Egr-1 to TRE1, upon stimulation, was also not unexpected due to the fact that, in many cases, binding sites for the transcription factor Sp1 also contain overlapping, cryptic binding sites for Egr-1 (46 -50). In addition, the fact that binding of Egr-1 was shown to occur only upon stimulation, is consistent with induction of Egr-1 synthesis in lymphocytes in response to mitogenic stimulation (58). Furthermore, the binding of Egr-1 to TRE1 using extracts prepared from unstimulated HUT102 cells (data not shown) was consistent with the finding that HTLV-1 Tax induces the expression of various immediate early serum response genes, including Egr-1 (59). Thus, Tax may replace/bypass growth signals, at least in part, in this manner.
These results represent the first demonstration of an Egr-1/Sp family cis-acting element mediating Tax responsiveness of a cellular gene. The cis elements mediating Tax responsiveness of various other cellular genes are diverse. The CArG box in the 5Ј-flanking sequence of c-fos is a Tax-responsive element (60). p67 SRF , a CArG binding factor which is constitutively localized in the nucleus, mediates transcriptional activation by Tax through direct interaction with Tax. Tax transactivation of the IL-2R␣ gene is mediated through NF-B, which is induced by Tax and subsequently interacts with the NF-B motif (GGG-GAATCTCCC) in the IL-2R␣ promoter (61,62). Tax responsiveness of the human granulocyte macrophage/colony-stimulating factor promoter is mediated through a 22-bp sequence, containing CATT(A/T) repeats, which is also required for mitogen inducibility of the same promoter (63,64). The critical region in the transforming growth factor ␤ promoter which is required for Tax-mediated induction contains a sequence which is similar to an AP-1 binding site (65).
It remains to be seen how Tax mediates the transactivation of the c-sis/PDGF-B promoter. One clue stems from the recent findings that Tax transactivation may involve enhancement in the DNA binding activity of target transcriptional regulatory proteins (9). These findings demonstrated that Tax was able to enhance the site-specific DNA binding activity of serum response factor and Fos-Jun and modestly enhanced the binding of the NF-B subunits, p50 and p65. In addition, they also showed that Tax was able to increase the DNA binding activity of the eukaryotic transcription factors ATF-1, Sp1, and GAL4. In accordance, recent evidence suggests that Tax may stimulate HTLV-1 transcription, at least in part, through enhanced binding of ATF/CREB proteins to their recognition elements within the Tax-responsive 21-bp repeats of the viral LTR (13). It may be that one possible mechanism of Tax-mediated c-sis/ PDGF-B promoter transactivation involves enhancing the DNA binding activity of Sp1 by Tax. With regard to Egr-1, in addition to up-regulating its expression, Tax may also have an effect on its DNA binding activity. Cooperative versus antagonistic binding between Sp1, Sp3, and Egr-1 will also need to be addressed, since all three are present in unstimulated nuclear extracts prepared from HUT102 cells (Fig. 6, lane 6). In addition, further insight into the mechanism of c-sis/PDGF-B promoter transactivation by Tax will be gained by identification of the protein(s) comprising the major EMSA complex C5, observed with nuclear extracts prepared from HUT102 cells (Fig.  6, lane 6). Future efforts should determine what functional role(s) each of the TRE1 binding proteins play in supporting Tax-mediated c-sis/PDGF-B promoter transactivation.