The Tax Protein of Human T-cell Leukemia Virus Type 1 Mediates the Transactivation of the c-sis/Platelet-derived Growth Factor-B Promoter through Interactions with the Zinc Finger Transcription Factors Sp1 and NGFI-A/Egr-1*

Transcriptional up-regulation of the c-sis/platelet-derived growth factor-B (PDGF-B) proto-oncogene by the Tax protein of human T-cell leukemia virus type 1 has been implicated as one possible mechanism of cellular transformation by human T-cell leukemia virus type 1. In previous work, we identified an essential site in the c-sis/PDGF-B promoter, Tax-responsive element 1 (TRE1), necessary for transactivation by Tax. We also identified Sp1, Sp3, and NGFI-A/Egr-1 as the primary nuclear transcription factors binding to TRE1 which mediate Tax responsiveness. In the present work, we have investigated the mechanism(s) whereby Tax transactivates the c-sis/PDGF-B proto-oncogene. In vitrotranscription assays showed that Tax was able to significantly increase the transcriptional activity of a template containing the −257 to +74 region of the c-sis/PDGF-B promoter. Electrophoretic mobility shift assay analysis showed that Tax increased the DNA binding activity of both Sp1 and NGFI-A/Egr-1 using a TRE1 probe. Analysis of Tax mutants showed that two mutants, IEXC29S and IEXL320G, were unable to significantly transactivate the c-sis/PDGF-B promoter. Finally, co-immunoprecipitation analysis revealed that Tax is able to stably bind to both Sp1 and NGFI-A/Egr-1. Interestingly, co-immunoprecipitation analysis also revealed that Tax mutant IEXC29S is unable to interact with NGFI-A/Egr-1, whereas Tax mutant IEXL320G is able to interact with NGFI-A/Egr-1.

The c-sis proto-oncogene, which encodes the B-chain of platelet-derived growth factor (PDGF) (23,24), is actively transcribed in T-cells infected with HTLV-1. The transcript has been cloned and sequenced, and is apparently normal (25)(26)(27). Expression of c-sis is normally tightly regulated in a cell typeand developmental stage-specific manner and is thought to play a role in wound healing and early development (28). It is not normally expressed, however, in lymphocytes. PDGF is a potent mitogen and chemoattractant for cells of mesenchymal origin (29). Biologically active PDGF is a dimeric protein consisting of homo-and heterodimeric combinations of two polypeptide chains, A and B (30). 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 (31). PDGF was first implicated in the process of transformation when one of its polypeptide chains, the B-chain/c-sis, was found to be homologous to the viral sis oncogene (v-sis) (23,24). Indeed, 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 (32).
Because lymphocytes in general were thought to have no receptors for PDGF, it was previously postulated that the c-sis/ PDGF-B expression in HTLV-1 infected T-cells could possibly provide a paracrine function, perhaps stimulating stromal cells, or other nonlymphocytic cells known to have PDGF receptors. More recently, however, it has been demonstrated that HTLV-1 infected T-cells also 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 het-erodimer (33). These findings suggest the possibility that HTLV-1 infected T-cells might acquire an auotcrine mechanism of cell proliferation that involves PDGF.
With regard to the regulatory mechanism(s) that underlies c-sis/PDGF-B expression in HTLV-1-infected T-cells, previous work from our laboratory (34), and others (35)(36)(37)(38)(39), identified a regulatory site at Ϫ64 to Ϫ45 within the c-sis/PDGF-B promoter, that we named Tax-responsive element 1 (TRE1). This regulatory site was shown to be essential for transactivation by Tax. In addition, electrophoretic mobility shift assay (EMSA) analysis and antibody supershift analysis of TRE1-binding proteins in Jurkat-E6.1 and Jurkat-Tax cell nuclear extracts, along with nuclear extracts prepared from the HTLV-1 infected T-cell line, HUT102, identified the Sp family members, Sp1 and Sp3, along with a member of the immediate early response gene family, NGFI-A/Egr-1 (for the sake of simplicity, NGFI-A/Egr-1 from this point on will be referred to as Egr-1), as the main TRE1-binding factors mediating Tax-responsiveness.
In the current study, we have investigated the mechanism(s) whereby Tax transactivates the c-sis/PDGF-B proto-oncogene promoter. In vitro transcription analysis showed that Tax was able to markedly increase RNA synthesis from a template containing the Ϫ257 to ϩ74 region of the c-sis/PDGF-B promoter. Site-directed mutagenesis of the TRE1 region was used to identify a CCACCC and GNGNGGGNG motif essential for transactivation by Tax. EMSA analysis hinted at a possible mechanism of Tax transactivation in that Tax was able to substantially increase the DNA binding activity of both Sp1 and Egr-1 to their DNA recognition sites contained within TRE1. In addition, EMSA analysis also indicated the possibility of ternary complex formation consisting of Sp1 or Egr-1, Tax, and DNA, suggesting that Tax might stably interact with both Sp1 and Egr-1 through protein-protein contacts. We show by co-immunoprecipitation analysis that Tax does indeed stably interact with both Sp1 and Egr-1 in both the absence of DNA and from intact cells. Finally, we identified two Tax mutants, C29S and L320G, that were unable to transactivate the c-sis/PDGF-B promoter by more than 2-fold. Interestingly, co-immunoprecipitation analysis also revealed that Tax mutant C29S was unable to interact with Egr-1, whereas, Tax mutant L320G was able to interact with Egr-1. The results presented in this article support the possibility of the existence of an additional, as yet uncharacterized, pathway of transactivation by Tax involving members of the zinc finger family of transcription factors.
Transfections and Luciferase Assays-5 ϫ 10 6 Jurkat-Tax cells were transfected with 5 g of either the pRALuc luciferase reporter plasmid, the pRALuc linker-scanning mutants, Ϫ64/Ϫ55 and Ϫ54/Ϫ45, or the pRALuc site-directed mutant constructs, YL0-YL7, using the Lipofectin reagent method protocol (Life Technologies, Inc.). When stimulated, the cells were divided equally 36 h post-transfection 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 (PBS) 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 for the luciferase assays as described previously (34). For the analysis of the Tax mutants, 5 ϫ 10 6 Jurkat E6.1 cells were transfected with 3 g of the pRALuc luciferase reporter plasmid alone or, plus, where indicated, 3 g of the wild-type Tax expression plasmid, IEX, or the Tax mutant expression plasmids, IEXS10A, IEXC29S, IEXH43Q, IEXS258A, and L320G, using the Lipofectin reagent method protocol (Life Technologies, Inc.). Twenty-four hours post-transfection, the cells were stimulated with TPA (10 ng/ml) and ionomycin (0.4 g/ml). Twenty-four hours post-stimulation, cell lysates were prepared and luciferase assays were carried out as described above.
Bacterial Expression and Purification of Recombinant TaxH 6 and Egr-1H 6 -Purification of TaxH 6 proceeded as follows: Escherichia coli BL21(DE3) cells transformed with pTaxH 6 were grown at 37°C in 1 liter of Terrific Broth medium containing 100 g/ml ampicillin until A 600 ϭ 1.0 -1.5, then induced for TaxH 6 expression with 40 M isopropyl-1-thio-␤-D-galactopyranoside at room temperature overnight. Cells were harvested and resuspended in 20 ml of ice-cold PBS containing 300 mM NaCl, 0.25 mM PMSF, 0.5 mM 2-mercaptoethanol, and 10 mM imidazole. The cells were ruptured by sonication carried out at a 60% duty cycle for 4 ϫ 1 min duty cycle. After centrifugation at 16,000 ϫ g for 30 min at 4°C, the supernatant was mixed with 2 ml of Ni 2ϩ -NTAagarose (Qiagen) at 4°C for at least 2 h. The protein bound gel matrix was then packed into a column (1.5 cm ϫ 10 cm) and washed with 40 ml of the same buffer containing 40 mM imidazole. TaxH 6 protein was then eluted with a 50-ml gradient of 50 -500 mM imidazole. A negative control bacterial extract (prepared in the same manner from bacteria that were transformed with a vector containing no insert DNA) was also prepared. Purification of NGFI-A/Egr-1H 6 proceeded as follows: E. coli BL21(DE3)pLysS cells transformed with pJDM1731 were grown at 37°C in 1 liter of Terrific Broth medium containing 100 g/ml ampicillin and 60 g/liter chloramphenicol until A 600 ϭ 0.7-1.0, then induced for NGFI-A/Egr-1H 6 expression with 800 M isopropyl-1-thio-␤-Dgalactopyranoside at 37°C for 5 h. Cells were harvested and resuspended in 20 ml of ice-cold PBS containing 300 mM NaCl, 0.25 mM PMSF, 0.5 mM 2-mercaptoethanol, and 10 mM imidazole. The cells were ruptured by sonication carried out at a 60% duty cycle for 4 ϫ 1 min duty cycle. After centrifugation at 16,000 ϫ g for 30 min at 4°C, the supernatant was mixed with 2 ml of Ni 2ϩ -NTA-agarose (Qiagen) at 4°C for at least 2 h. The protein bound gel matrix was then packed into a column (1.5 cm ϫ 10 cm) and washed with 40 ml of the same buffer containing 40 mM imidazole. NGFI-A/Egr-1H 6 protein was then eluted with a 50-ml gradient of 50 -500 mM imidazole. Fractions collected from both the TaxH 6 purification and the NGFI-A/Egr-1H 6 purification were analyzed by SDS-7.5% polyacrylamide gel electrophoresis, followed by Coomassie Brilliant Blue staining. Purity of each was determined to be approxi-mately Ͼ95%. Fractions containing purified TaxH 6 and NGFI-A/Egr-1H 6 were further analyzed by Western blotting using antisera against Tax (␣TaxC; a generous gift from W. C. Greene), and NGFI-A/Egr-1 (Santa Cruz Biotechnology). The fractions containing either purified TaxH 6 or NGFI-A/Egr-1H 6 were dialyzed for 6 -24 h at 4°C against buffer containing 20 mM HEPES (pH 7.9), 0.1 M KCl, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM PMSF, and 20% (v/v) glycerol and frozen in aliquots at Ϫ70°C. Protein concentrations were determined with a commercially available kit (Bio-Rad).
EMSAs-The following double-stranded oligonucleotide probes, corresponding to nucleotides Ϫ83 to Ϫ45 of the c-sis/PDGF-B promoter and containing the TRE1 region, or a mutated CCACCC motif of the TRE1 region, respectively, were prepared by annealing complementary, single-stranded oligonucleotides (Life Technologies, Inc.) in a thermocycler (95°C for 5 min, cool to 25°C at 1°C/min): TRE1, 5Ј-GCCAGAAGAG-GAAAGGCTGTCTCCACCCACCTCTCGCAC-3Ј; TRE1-mSp1, 5Ј-GC-CAGAAGAGGAAAGGCTGTCTGATCGAACCTCTCGCAC-3Ј. Each probe was end-labeled with [␥-32 P]ATP (ICN Biomedicals, Inc.) and T4 polynucleotide kinase (New England Biolabs). Typical in vitro binding reactions (20 l) contained approximately 1 ng of either purified Sp1 (Promega) or purified NGFI-A/Egr-1H 6 , 1 g of poly(dI-dC) (Amersham), 1 ϫ Superdex buffer (13) (25 mM HEPES (pH 7.9), 12.5 mM MgCl 2 , 10 M ZnSO 4 , 150 mM KCl, 4 mM 2-mercaptoethanol, 20% (v/v) glycerol, 0.1% Nonidet P-40), and 10 -20 fmol of 32 P-labeled probe (15-30 ϫ 10 3 cpm). In samples containing TaxH 6 , approximately 400 ng of purified TaxH 6 was used in each reaction. EMSA analysis with dilutions of TaxH 6 down to 25 ng was also performed, with similar results observed (data not shown). Once all of the appropriate factors had been added to each sample, the samples were allowed to incubate 5 min at room temperature in 1 ϫ Superdex buffer prior to the addition of labeled probe. After the addition of labeled probe, the mixture was incubated a further 10 min at room temperature. The DNA-protein complexes were then resolved by electrophoresis on a 5% nondenaturing, polyacrylamide gel (acrylamide/N,NЈ-methylenebisacrylamide weight ratio, 49:1) at 165 V for 4 h at 4°C in 1 ϫ TGE buffer (25 mM Tris-HCl (pH 8.5), 190 mM glycine, 1 mM EDTA). In reactions that included cold (unlabeled) oligonucleotide competitors, the samples were allowed to incubate with the cold oligonucleotide probes, along with poly(dI-dC) in 1 ϫ Superdex buffer, for 10 min at room temperature prior to the addition of the labeled DNA probe. In reactions containing ␣TaxC Ab, 2 l of ␣TaxC antiserum was added to the appropriate samples and incubated for 1 h at 4°C, along with poly(dI-dC) in 1 ϫ Superdex buffer, for 1 h at 4°C prior to the addition of the labeled DNA probe. The gels were then dried under vacuum at 80°C for 1 h and exposed to XAR-5 film.
Immunoprecipitation-Co-immunoprecipitation reactions (50 l), using purified proteins, contained approximately 200 ng of either purified Sp1 (Promega) or purified NGFI-A/Egr-1H 6 , and approximately 400 ng of purified TaxH 6 in 1 ϫ Superdex buffer. The samples, containing the various purified proteins, were allowed to incubate for 15 min at room temperature prior to the addition of 5 l of pooled Tax monoclonal antibodies (␣Tax mAbs) (45), 100 l of 5 ϫ Superdex buffer, and 345 l of dH 2 O. For co-immunoprecipitation reactions involving HTLV-1-infected cells, whole cell extracts were prepared from 100 ϫ 10 6 MT2 cells. The cells were washed 1 time with 5 ml of PBS. The cells were then lysed with 2 ml of RIPA buffer (0.5% deoxycholic acid, 0.1% SDS, 1% Nonidet P-40) for 5 min at room temperature and clarified by spinning at 14,000 ϫ g for 5 min at 4°C. The supernatant was divided equally into two samples; to one sample was added 7 l of pooled ␣Tax mAbs, the other sample did not receive any pooled ␣Tax mAbs as a control. The samples from both the purified protein co-immunoprecipitation reactions, and the co-immunoprecipitation reactions involving whole cell extracts prepared from HTLV-1-infected cells, were then incubated for 1 h on a rotator at 4°C. 50 l of a 50% slurry solution of immobilized rProtein A-Sepharose (Repligen Corp.) was then added to each sample. The reaction mixtures were incubated for 18 -24 h on a rotator at 4°C. The beads were then gently washed 5-8 times with 500 l of RIPA buffer. The bound proteins were eluted in SDS sample loading buffer, subjected to electrophoresis on a 7.5% denaturing polyacrylamide gel, transferred to nitrocellulose, and Western blotted for Tax (␣TaxC), Sp1 (Santa Cruz Biotechnology), and NGFI-A/Egr-1 (Santa Cruz Biotechnology) by enhanced chemiluminescence (ECL) (Amersham). After each individual Western blot, the blot was stripped by submerging the membrane in stripping buffer (100 mM 2-mercaptoethanol, 2% SDS, 62.5 mM Tris-HCl, pH 6.7) and incubating at 50°C for 30 min. The same blot (membrane), for each, separate, co-immunoprecipitation was used for all Western blots pertaining to that co-immunoprecipitation. Bands were quantitated with a Densitometer (Pharmacia Biotech Inc.).
Immunoprecipitation Analysis of Tax Mutants in COS-7 Cells-5 g of either the wild type Tax expression plasmid, IEX, or the Tax mutant expression plasmids, IEXC29S and IEXL320G, were transfected into 40 -60% confluent COS-7 cells in a 100-mm tissue culture dish, either alone or with 5 g of the Egr-1 expression plasmid, pJDM948, using the Lipofectin reagent method protocol (Life Technologies, Inc.). 5 g of pJDM948 was also transfected alone into 40 -60% confluent COS-7 cells in a 100-mm tissue culture dish as a control for Egr-1 expression. Each of the transfections was performed in triplicate. Forty-eight hours post-transfection, the cells were washed with PBS, harvested by trypsinization, and pelleted by centrifugation at 14,000 ϫ g for 5 min at 4°C. The cells from all samples, except the Egr-1 alone sample, were then lysed by resuspending the cell pellet in 0.5 ml of RIPA buffer for 5 min at room temperature and clarified by spinning at 14,000 ϫ g for 5 min at 4°C. For the Egr-1 alone sample, a whole cell extract was prepared by three cycles of freeze-thawing in an ethanol/dry ice bath and 37°C water bath. 5 l of pooled ␣Tax mAbs was added to each sample, except the Egr-1 alone sample, and the samples were then incubated for 1 h on a rotator at 4°C. 50 l of a 50% slurry solution of immobilized rProtein A-Sepharose (Repligen Corp.) was then added to each sample. The reaction mixtures were incubated for 18 -24 h on a rotator at 4°C. The beads were then gently washed 5-8 times with 500 l of RIPA buffer. The bound proteins were eluted in SDS sample loading buffer, subjected to electrophoresis on a 7.5% denaturing polyacrylamide gel, transferred to nitrocellulose, and Western blotted for Tax (␣TaxC) and Egr-1 (Santa Cruz Biotechnology) by ECL (Amersham). After each Western blot, the blot was stripped by submerging the membrane in stripping buffer and incubating at 50°C for 30 min. The same blot (membrane) was used for both Western blots.
In Vitro Transcription Reactions-Typical in vitro transcription reactions (25 l) contained 50 g of stimulated Jurkat E6.1 nuclear extract, 500 ng of glass-bead purified template DNA (the 954-bp NcoI-EcoRI fragments of either pRALuc or the pRALuc linker-insertion mutants, Ϫ64/Ϫ55, Ϫ54/Ϫ45, and Ϫ34/Ϫ25), 5 mM MgCl 2 , 400 M each ATP, CTP, and GTP, 40 M UTP, and 10 Ci of [␣-32 P]UTP (3000 Ci/mmol) in buffer A. Reactions involving Tax contained approximately 400 ng of TaxH 6 . The reactions were preincubated for 15 min at 30°C prior to the addition of the NTPs and then were allowed to proceed for 1 h at 30°C after the addition of the NTPs. When indicated, ␣-amanitin was added to the in vitro transcription reactions to a final concentration of 5 g/ml. Reactions were terminated by adding 175 l of stop solution (300 mM Tris-HCl, pH 7.4, 300 mM sodium acetate, 0.5% SDS, 2 mM EDTA, 3 g/ml tRNA), and extracted with 200 l phenol:chloroform: isoamyl alcohol (25:24:1). Each reaction was then passed over a Micro-Spin S-300 HR column (Pharmacia Biotech). The reactions were then precipitated with 500 l of 100% ethanol, placed on dry ice for 15 min, pelleted, washed with 70% ethanol, pelleted, and dried in a vacuum dessicator for 10 min. The pellets were resuspended in loading dye, heated at 90°C for 10 min, and analyzed by gel electrophoresis on a 5% denaturing polyacrylamide gel containing 7 M urea. The gels were then dried under vacuum at 80°C for 1 h, exposed to XAR-5 film, and bands corresponding to full-length transcripts were quantitated with a densitometer (Pharmacia Biotech Inc.).

In Vitro Transcription from the c-sis/PDGF-B Promoter Is
Enhanced by HTLV-1 Tax-The effect(s) of HTLV-1 Tax on transcription of the c-sis/PDGF-B promoter was investigated using an in vitro run-off transcription assay supported by nuclear extracts prepared from stimulated Jurkat E6.1 cells. Transcription from the template DNA fragment, containing the c-sis/PDGF-B promoter, was designed to generate an RNA transcript of 697 nucleotides (represented schematically in Fig.  1A). The results presented in lane 2 of Fig. 1B demonstrate that the c-sis/PDGF-B promoter is efficiently transcribed using stimulated Jurkat E6.1 nuclear extracts in the in vitro run-off transcription assays. The addition of Tax to the in vitro run-off transcription assays resulted in an 11-fold increase in transcription above that observed in the absence of Tax (Fig. 1B, compare lane 3 versus lane 2). To demonstrate that the transactivation activity was specific to Tax, a control extract containing no Tax protein (see "Materials and Methods") was added to the reactions. As shown in Fig. 1B, lane 4, the addition of the control extract did not result in an increase of transcrip-tion. To demonstrate that the transcription observed in our in vitro run-off transcription assays was mediated by RNA polymerase II (pol II), the pol II inhibitor, ␣-amanitin (a bicyclic octapeptide from the mushroom Amanita phalloides), was added to the reactions. At low concentrations, ␣-amanitin selectively inhibits pol II-dependent transcription through binding to the largest subunit of pol II and blocking transcription elongation (46 -49). Transcription from the c-sis/PDGF-B promoter was completely abolished upon addition of ␣-amanitin to 5 g/ml (Fig. 1B, lane 5), as well as transactivation by Tax (Fig. 1B, lane 6).
To demonstrate the necessity of the TRE1 region for both basal transcription and transactivation by Tax, in vitro run-off transcription reactions were performed using the pRALuc linker-insertion mutants Ϫ64/Ϫ55, Ϫ54/Ϫ45, and Ϫ34/Ϫ25 as templates. Each mutant contains a 10-bp substitution linker sequence at the designated nucleotides. Mutants Ϫ64/Ϫ55 and Ϫ54/Ϫ45 disrupt TRE1 while mutant Ϫ34/Ϫ25 disrupts the TATA box/initiator region. As shown in Fig. 1B, transcription from the mutant templates Ϫ64/Ϫ55 and Ϫ54/Ϫ45 is significantly reduced both in the absence and presence of Tax (compare lane 7 versus lanes 1 and 8, and lane 9 versus lanes 1 and 10, respectively). When the TATA box/initiator region is disrupted by mutant Ϫ34/Ϫ25, transcription from this template is completely abolished in both the absence and presence of Tax (Fig. 1B, lanes 11 and 12). These data demonstrate Tax's ability to enhance pol II-dependent transcription from the c-sis/ PDGF-B promoter in an in vitro run-off transcription assay along with the relative importance of the TRE1 region for both basal transcription and Tax-mediated transactivation.
Site-directed Mutagenesis of TRE1-To further investigate the regulatory region contained within TRE1, a series of c-sis/ PDGF-B promoter mutants containing high-resolution mutations (Fig. 2) within TRE1, driving expression of the firefly luciferase reporter gene, was constructed. These site-directed mutants were transiently transfected into Jurkat-Tax cells, which were then stimulated with TPA and ionomycin to transiently activate the expression of Egr-1. Luciferase activity was then measured. As shown in Fig. 2, mutants YL0-YL3 each resulted in a decrease in activity comparable to that observed with either of the two 10-bp linker insertion mutants, Ϫ64/Ϫ55 and Ϫ54/Ϫ45. In contrast, mutants YL4-YL7 were each significantly lower in activity than the wild-type, pRALuc, but no single mutant caused a reduction equal to that observed with either Ϫ64/Ϫ55 or Ϫ54/Ϫ45. This analysis indicated the importance of a CCACCC motif (ablated by mutants YL1-YL3), along with the importance of a GNGNGGGNG motif (ablated by mutants YL1-YL5), for Tax-responsiveness.
The CCACCC motif has been previously identified as a cisacting regulatory element, capable of binding members of the Sp family of transcription factors, in the promoters of several genes, some of which include the human embryonic ⑀-globin gene (50), the parathyroid hormone-related protein (PTHrP) gene (51), the human IL-2 gene (52), and the human leukosialin (CD43) gene (53). Likewise, the GNGNGGGNG motif (54) has been previously identified as a cis-acting regulatory element, capable of binding members of the immediate-early transcription factor gene family, in the promoters of several genes, some of which include the phenylethanolamine N-methyltransferase gene (55), the human tumor necrosis factor gene (56), and the human IL-2 gene (52). Shown in Fig. 3A are the consensus DNA binding sequences for the Sp transcription factor family member, Sp1 (57), containing the CCACCC motif in its core, and the immediate-early transcription factor gene family member, Egr-1 (54). It is interesting to note that TRE1 contains both the Sp1 core recognition sequence and the Egr-1 consensus DNA recognition sequence (Fig. 3A). Initial EMSA analyses of T-cell line nuclear proteins that bound to TRE1 demonstrated that both Sp1 and Egr-1 did indeed bind to TRE1, and that this binding was essential for Tax responsiveness (34). Similar regions found in the PTHrP gene promoter and the IL-2 gene promoter, two genes that have also been demonstrated to be transactivated by Tax (58,59), are shown in Fig. 3B. The PTHrP promoter contains the core of the Sp1 consensus DNA recognition sequence (the CCACCC motif), whereas, the IL-2 promoter contains both the core of the Sp1 consensus DNA recognition sequence (the CCACCC motif), and the Egr-1 consensus DNA recognition sequence (the GNGNG-GGNG motif).
HTLV-1 Tax Enhances the DNA Binding Activity of Sp1 and Egr-1-With the knowledge that two of the main TRE1-binding proteins consisted of Sp1 and Egr-1, and that this binding was essential for Tax responsiveness, we next sought to investigate whether Tax could enhance the binding of either Sp1 or Egr-1, or both, to TRE1 by EMSA analysis. A previous report on the effects of Tax on the DNA binding activity of eukaryotic transcription factors demonstrated that Tax enhanced the sitespecific DNA binding activity of Sp1 to a GC-rich recognition sequence, containing a CCACCC motif, found within the HTLV-1 promoter (13). This Sp1-binding site in the HTLV-1 promoter is located in a region previously shown to be Taxresponsive in vivo (60). As shown in Fig. 4, Tax  Egr-1H 6 C1). The binding of both Sp1 and Egr-1 was specific for the TRE1 probe since an excess of unlabeled TRE1 probe competed for their binding (lane 4 and lane 8, respectively). Interestingly, the addition of Tax to the binding reaction mixtures of both Sp1 and Egr-1 also resulted in the formation of an additional shifted complex (Sp1 C2 and Egr-1H 6 C2, respectively) migrating slower than either Sp1 C1 or Egr-1H 6 C1 (lanes 3 and 7, respectively). We next investigated the importance of the CCACCC and GNGNGGGNG motifs, contained within TRE1, for Sp1 and Egr-1 binding, in both the presence and absence of Tax. As shown in Fig. 4, the binding of both Sp1 and Egr-1, in the presence and absence of Tax, was completely abolished when a probe containing both a mutated CCACCC motif, and GNGNGGGNG motif, was used in the EMSA analysis (lanes 9 -13). Thus, these results demonstrate Tax's ability to increase the DNA binding activity of the zinc finger transcription factors, Sp1 and Egr-1, to TRE1, along with the relative importance of the CCACCC and GNGNGGGNG motifs contained within TRE1.
Ternary Complex Formation between Either Sp1 or Egr-1, Tax, and DNA-The observation that the addition of Tax to the binding reaction mixtures resulted in the formation of additional, slower migrating, shifted complexes ( Fig. 4 and Fig. 5, Sp1 C2 and Egr-1H 6 C2) prompted us to investigate the possibility that Tax, itself, was stably incorporated into the complex. We addressed this possibility by antibody supershift analysis utilizing the TRE1 probe mentioned above. As shown in Fig. 5, when an antibody directed against Tax was included in either the Sp1 (compare lane 4 versus lanes 2 and 3) or Egr-1 (compare lane 7 versus lanes 5 and 6) samples containing Tax, not only was Tax-mediated enhancement of binding abolished, but also, basal binding of Sp1 and Egr-1 was completely abolished. Surprisingly, the addition of anti-Tax antibody did not give rise to the formation of antibody supershifted complexes, but, rather, completely blocked the formation of all shifted complexes. This effect was not due to the antibody directly interfering with the binding activity of Sp1 or Egr-1, since their binding was not inhibited by the antibody in the absence of Tax (lane 8 and lane 9, respectively). In addition, this effect of the antibody was specific for Tax, since preimmune serum did not affect Taxmediated enhancement of either Sp1 or Egr-1 binding (lane 10

FIG. 3. TRE1 of the c-sis/PDGF-B promoter contains Sp1 and Egr-1 consensus DNA recognition sequences similar to those found in the Tax-responsive promoters of the PTHrP and IL-2 genes.
A, consensus DNA recognition sequences of the zinc finger transcription factors Sp1 and Egr-1. The "CCACCC" motif, which comprises the core of the Sp1 DNA recognition sequence, is indicated. The CCACCC core motif of the Sp1 DNA recognition sequence and the Egr-1 consensus DNA recognition sequence contained within TRE1 (nucleotides Ϫ64 to Ϫ45) are indicated. B, representation of the Sp1 CCACCC core motif found in the Ϫ64 to Ϫ45 region of the PTHrP gene promoter, and, the Sp1 CCACCC core motif and Egr-1 consensus DNA recognition sequence found in the Ϫ303 to Ϫ280 region of the IL-2 gene promoter. q indicates a nucleotide which does not conform to the Egr-1 consensus DNA recognition sequence. and lane 11, respectively). Thus, these results support the possibility that Tax is incorporated into the shifted complexes, resulting in the formation of ternary complexes, possibly through protein-protein contacts with Sp1 and Egr-1.
HTLV-1 Tax Physically Interacts with Both Sp1 and Egr-1-To investigate the potential interaction of Tax with either Sp1 or Egr-1, in vitro co-immunoprecipitation experiments were performed using purified proteins. Samples containing either Sp1, plus and minus TaxH 6 , or Egr-1H 6 , plus and minus TaxH 6 , were incubated in Superdex buffer, immunoprecipitated with ␣-Tax monoclonal antibodies and the immune com-plexes were analyzed by Western blotting. As shown in Fig. 6A, for co-immunoprecipitation reactions involving Sp1, a band of approximately 95 kDa (corresponding to Sp1) was detected in the sample incubated in the presence of Tax (lane 3), whereas it was not detected in the sample incubated in the absence of Tax (lane 5). Similarly, for co-immunoprecipitation reactions involving Egr-1H 6 , a band of approximately 85 kDa (corresponding to Egr-1H 6 ) was detected in the sample incubated in the presence of Tax (Fig. 6B, lane 4), whereas it was not detected in the sample incubated in the absence of Tax (Fig. 6B,  lane 5). Similar amounts of TaxH 6 were immunoprecipitated in  2-4), or purified Egr-1H 6 (ϳ1 ng) (lanes 6 -8) were incubated with a probe containing the TRE1 site of the c-sis/PDGF-B promoter. Binding reactions were carried out in the absence (lanes 2 and 6, respectively) or presence of TaxH 6 (lanes 3 and 7, respectively). To demonstrate specificity of binding, Sp1 (lane 4) and Egr-1H 6 (lane 8) were incubated in the presence of a 1000-fold molar excess of unlabeled TRE1 probe. As a control, the TRE1 probe was incubated by itself (lane 1) or with TaxH 6 only (lane 5). To demonstrate the importance of TRE1 for Sp1 and Egr-1H 6 binding, Sp1 and Egr-1H 6 were incubated with a probe containing a mutated TRE1 site, TRE1-mSp1 (see "Materials and Methods"), in the absence (lanes 10 and 12, respectively) or presence of TaxH 6 (lanes 11 and 13, respectively). Binding reactions were carried out as described under "Materials and Methods." Similar results were observed using dilutions of TaxH 6 down to 25 ng. Sp1 shifted complexes, Sp1 C1 and Sp1 C2, are indicated on the left side of the gel by arrows. Egr-1H 6 shifted complexes, Egr-1H 6 C1 and Egr-1H 6 C2, are indicated on the right side of the gel by arrows. Lanes containing TRE1 probe, or the mutated TRE1 probe, TRE1-mSp1, are indicated. Binding reactions were carried out as described under "Materials and Methods." Sp1 shifted complexes, Sp1 C1 and Sp1 C2, are indicated on the left side of the gel by arrows. Egr-1H 6 shifted complexes, Egr-1H 6 C1 and Egr-1H 6 C2, are indicated on the right side of the gel by arrows. *nsp indicates a nonspecific shifted complex. each sample (Fig. 6C, lanes 2-4).
To further demonstrate the relevance of these interactions, co-immunoprecipitation reactions were performed using whole cell extracts prepared from the HTLV-1 infected T-cell line, MT2. As shown in Fig. 7A, for co-immunoprecipitation reactions involving Sp1, a band of approximately 95 kDa (corresponding to Sp1) was detected in the sample incubated in the presence of ␣-Tax monoclonal antibodies (lane 1), whereas it was not detected in the sample incubated in the absence of ␣-Tax monoclonal antibodies (lane 2). This band also comigrated with the bands detected in the positive control samples using nuclear extract prepared from mitogen-stimulated Jurkat E6.1 cells and purified Sp1 (Fig. 7A, lanes 4 and 5, respectively). Similarly, for co-immunoprecipitation reactions involving Egr-1, a band of approximately 85 kDa (corresponding to Egr-1) was detected in the sample incubated in the presence of ␣-Tax monoclonal antibodies (Fig. 7B, lane 1), whereas it was not detected in the sample incubated in the absence of ␣-Tax monoclonal antibodies (Fig. 7B, lane 2). This band also comigrated with the band detected in the positive control sample FIG. 6. Co-immunoprecipitation of purified Sp1 and Egr-1 with purified Tax. Either ϳ200 ng of purified Sp1 and ϳ400 ng of purified TaxH 6 or ϳ200 ng of purified Egr-1H 6 and ϳ400 ng of purified TaxH 6 were preincubated in 1 ϫ Superdex buffer for 15 min at room temperature and co-immunoprecipitation (Co-IP) reactions were carried out as described under "Materials and Methods." A, immunoblot against Sp1 demonstrating co-immunoprecipitation (Co-IP) with TaxH 6 using ␣Tax mAbs (lane 3). The Sp1 band is indicated by an arrow to the right of the gel. B, immunoblot against Egr-1H 6 demonstrating co-immunoprecipitation with TaxH 6 using ␣Tax mAbs (lane 4). A mouse secondary antibody, recognizing the heavy chain component of the ␣Tax mAbs, was included in the co-immunoprecipitation reaction to demonstrate equivalent amounts of protein in each sample (lanes 1-6). * indicates a nonspecific band. The Egr-1H 6 , heavy chain, and * nonspecific bands are indicated by arrows to the right of the gel. C, immunoblot against TaxH 6 demonstrating equivalent amounts of TaxH 6 protein in each co-immunoprecipitation reaction (lanes 2-4). The TaxH  using nuclear extract prepared from mitogen-stimulated Jurkat E6.1 cells (Fig. 7B, lane 4). Taken together, these findings demonstrate that Tax physically interacts with both Sp1 and Egr-1, through protein-protein contacts.
Ability of HTLV-1 Tax Mutants to Transactivate the c-sis/ PDGF-B Promoter-To examine the transactivation activities of Tax mutants on the c-sis/PDGF-B promoter, Jurkat E6.1 cells were transiently transfected with the pRALuc luciferase reporter plasmid either alone or together with the wild-type or mutant Tax expression plasmids, IEX, IEXS10A, IEXC29S, IEXH43Q, IEXS258A, and IEXL320G. The cells were then stimulated with TPA and ionomycin to transiently activate the expression of Egr-1. Luciferase activity was then measured. As shown in Fig. 8, wild-type Tax, IEX, resulted in a greater than 14-fold induction of the c-sis/PDGF-B promoter and Tax mutants IEXS10A, IEXH43Q, and IEXS258A all resulted in a 4-fold or greater induction. In contrast, Tax mutants IEXC29S (Ͻ2-fold) and IEXL320G (2-fold) were significantly impaired in their ability to transactivate the c-sis/PDGF-B promoter.
Analysis of HTLV-1 Tax Mutants IEXC29S and IEXL320G to Interact with Egr-1-With the knowledge that the HTLV-1 Tax mutants IEXC29S and IEXL320G were unable to significantly transactivate the c-sis/PDGF-B promoter, we reasoned that this might be a result of their inability to interact with Egr-1. The idea of focusing on the interaction between Tax and Egr-1, and not Tax and Sp1, is based on the observations that Sp1 and Egr-1-binding sites in the c-sis/PDGF-B promoter overlap and Sp1 is displaced by increasing amounts of Egr-1 that arise upon stimulation (37). To investigate Tax-Egr-1 interactions, we performed a series of co-immunoprecipitation reactions in which the wild-type Tax expression plasmid, IEX, or the Tax mutant expression plasmids, IEXC29S and IEXL320G, were transiently cotransfected into COS-7 cells along with the Egr-1 expression plasmid, pJDM948, and immunoprecipitated with ␣-Tax monoclonal antibodies. The immune complexes were then analyzed by Western blotting. As shown in Fig. 9, a band of approximately 85 kDa, corresponding to Egr-1, was detected in the sample containing wild-type Tax, IEX (lane 5), and interestingly, also in the sample containing the Tax mutant IEXL320G (lane 7). This band also comigrated with the band detected in the positive control sample using a whole cell extract prepared from COS-7 cells transfected with the Egr-1 expression plasmid alone (Fig. 9, lane 8). In contrast, this same band, corresponding to Egr-1, was not detected in the sample containing the Tax mutant IEXC29S (Fig. 9, lane 6). Thus, these results suggest that protein-protein interactions between Tax and Egr-1 are necessary, but not sufficient, to support Tax-mediated transactivation of the c-sis/PDGF-B promoter.

DISCUSSION
In this study, we have investigated the mechanism of HTLV-1 Tax-mediated transactivation of the c-sis/PDGF-B promoter by in vitro transcription assay and site-directed mutation analysis, along with EMSA and co-immunoprecipitation analysis. We have shown by in vitro transcription assay analysis that addition of Tax to the in vitro transcription reaction led to an 11-fold increase in RNA synthesis (Fig. 1B, compare  lane 3 versus lane 2). It has previously been reported that Tax-mediated in vitro transcription from the HTLV-1 promoter is resistant to low levels of ␣-amanitin (at low concentrations, ␣-amanitin selectively inhibits RNA pol II) in reactions containing either HeLa or CEM T-cell extract. This is surprising in light of the fact that the HTLV-1 promoter contains all of the structural features of a typical RNA pol II transcription template: a TATA box ϳ30 bp upstream of the transcription initiation site, binding sites for several pol II transcription factors, and long poly(A) ϩ RNA is synthesized from the integrated HTLV-1 proviral DNA in vivo. In addition, in vitro transcription experiments, supported by either HeLa or CEM T-cell extracts, showed that the HTLV-1 promoter contains overlapping transcription units which utilize the same transcription initiation site as that of pol II-dependent transcription. Fur- thermore, neutralization and depletion experiments with pol II antibodies strongly suggested that classical pol II is not involved in the HTLV-1 overlapping transcription unit (61,62).
Similarly, the c-sis/PDGF-B promoter contains all of the structural features of a typical RNA pol II transcription template: a TATA box ϳ30 bp upstream of the transcription initiation site, binding sites for several pol II transcription factors, and poly(A) ϩ RNA transcripts are synthesized from the promoter. Given the above mentioned data and the fact that our in vitro run-off transcription assays were supported by nuclear extracts prepared from stimulated Jurkat E6.1 T-cells, we tested our in vitro transcription assays for pol II-dependent transcription. We showed that both transcription in the absence of Tax, and Tax-mediated transactivation, was completely inhibited at a low concentration of ␣-amanitin (Fig. 1B,  compare lanes 5 and 6 versus lanes 2 and 3). In addition, we also demonstrated the relevant importance of the TRE1 region for Tax-mediated transactivation by using DNA templates containing linker-insertion substitution mutations within TRE1 (Fig. 1B, lanes 7-10). This would suggest that the binding of Sp1 or Egr-1 to TRE1 is essential for transactivation mediated by Tax. Together, these results indicated that Tax was able to greatly stimulate RNA synthesis in vitro from the c-sis/ PDGF-B promoter, and that this Tax-mediated transactivation was dependent not only on RNA pol II, but also on the binding of either Sp1 or Egr-1 to the TRE1 region.
When the SIS-Luc TRE1 site-directed mutants were analyzed for Tax-responsiveness, a CCACCC motif (mutants YL1-YL3) and a GNGNGGGNG motif (mutants YL1-YL5) were clearly identified as being essential for conferring Tax responsiveness in stimulated Jurkat-Tax cells (Fig. 2). That the results observed in this experiment were indeed due to Tax and not a result of the effects of TPA and ionomycin stimulation is evidenced by the fact that we have previously (34) analyzed a series of linker-scanning substitution mutants within the TRE1 region using unstimulated Jurkat E6.1 cells versus unstimulated Jurkat-Tax cells and observed the same results as those presented in this article. It should also be noted that stimulation in the presence of Tax results in a synergistic activation of the c-sis/PDGF-B promoter. Stimulation is necessary to transiently activate the expression of Egr-1.
The CCACCC motif has been previously reported to be a positive regulatory element in the promoters of several genes, capable of binding members of the Sp family of zinc finger transcription factors (50 -53). Indeed, the nucleotides CCACCC also comprise the core nucleotide sequence of the DNA recognition sequence for the Sp family member, Sp1 (Fig. 3A) (57). This same CCACCC motif has also been demonstrated to be critical for transcription of the c-sis/PDGF-B gene in the human osteosarcoma cell line, U2-OS, involving the Sp family members, Sp1 and Sp3 (35,38). Likewise, the GNGNGGGNG motif has also been previously reported to be a positive cis-acting regulatory element found in the promoters of several genes, capable of binding the zinc finger transcription factor Egr-1, which is a member of the immediate-early transcription factor gene family (52,55,56). Previous work from our laboratory (34) demonstrated that the main nuclear factors binding to TRE1 from the stimulated T-cell lines, Jurkat E6.1 and Jurkat-Tax, along with the HTLV-1 infected T-cell line, HUT102, included the Sp family members Sp1 and Sp3, along with the immediate-early transcription factor gene family member, Egr-1. Furthermore, it should be noted that since the TRE1 region of the c-sis/PDGF-B promoter contains overlapping binding sites for the Sp family members, Sp1 and Sp3, and the immediate-early transcription factor gene family member, Egr-1, these factors cannot associate together and simultaneously bind to TRE1.
Indeed, it has recently been shown that in unstimulated cells, Sp1 occupies this element and is displaced by increasing amounts of Egr-1 upon stimulation (37).
Both the CCACCC motif, and the GNGNGGGNG motif, are also found in the promoters of two genes previously shown to be up-regulated and transactivated by Tax. These include the PTHrP gene and the IL-2 gene (Fig. 3B) (58,59). These motifs represent binding sites for members of the zinc finger family of transcription factors. This proves to be interesting in light of previous experiments carried out to investigate the mechanism(s) of Tax-mediated transactivation. Site-directed mutational analysis of Tax has allowed for an approximate demarcation of regions within the Tax protein itself necessary for transactivation (41,63). These analyses indicated that Tax transactivation of viral and cellular promoters occurs through at least two distinct cellular transcription pathways: 1) the CREB/ATF family and 2) the NFB/Rel protein family of transcription factors. It now appears that an additional Tax-mediated transactivation pathway exists involving members of the zinc finger family of transcription factors.
EMSA analysis showed that Tax dramatically enhanced the site-specific DNA binding activity of the zinc finger transcription factors Sp1 (Fig. 4, compare lane 2 versus lane 3) and Egr-1 (Fig. 4, compare lane 6 versus lane 7) to a probe containing the TRE1 site. This is consistent with previous reports which suggested that Tax may deregulate target genes by enhancing the DNA binding activity of the cellular transcription factors that recognize the Tax-responsive promoter elements (19,20,64). Indeed, Tax has been shown to increase the DNA binding activity of a wide variety of eukaryotic transcription factors, including, members of the CREB/ATF family of proteins, serum response factor, Fos-Jun, and the NFB subunits, p50 and p65 (13,14). Mutation of the CCACCC and GNGNGGGNG motifs prevented the binding of Sp1 and Egr-1, respectively, to the TRE1 probe, in both the presence and absence of Tax (Fig. 4,  lanes 9 -13). This further demonstrated the importance of each of these motifs for conferring Tax responsiveness.
EMSA analysis also indicated that Tax altered the mobility of both the Sp1-DNA complex (Fig. 4, lane 3, Sp1 C2) and the Egr-1-DNA complex (Fig. 4, lane 7, Egr-1H 6 C2), suggesting that the observed increase in DNA binding activity for Sp1 and Egr-1 (mediated by Tax) involves Tax entering into the complex, leading to the formation of ternary complexes. It is possible that the Sp1 C2 and Egr-1H 6 C2 complexes are the result of Tax increasing the affinity of a less abundant complex that was previously not visible in the Tax-minus samples (Fig. 4,  lanes 2 and 6, respectively). However, this was not the case since these C2 complexes were not observed upon overexposure of the autoradiogram (data not shown). This was further substantiated by the finding that an antibody directed against the C-terminal 13 amino acids of Tax was able to completely abolish the formation of the Sp1-DNA complexes Sp1 C1 and Sp1 C2 (Fig. 5, lane 4), and the Egr-1H 6 complexes, Egr-1H 6 C1 and Egr-1H 6 C2 (Fig. 5, lane 7), when added to the samples containing Tax. This is indicative of a potentially strong interaction between Tax and both Sp1 and Egr-1, in which binding of the antibody to Tax sterically interferes with their ability to bind DNA. As a result, if all of the Sp1 and Egr-1 in the sample is complexed with Tax, then, no binding whatsoever should occur.
Further investigation of the physical interaction between Tax and both Sp1 and Egr-1, by co-immunoprecipitation analysis, showed that Tax did, indeed, stably interact with, and bind to, both purified Sp1 (Fig. 6A, lane 3) and purified Egr-1 (Fig. 6B, lane 4). In addition, it was also demonstrated by co-immunoprecipitation analysis that Tax interacts with both Sp1 (Fig. 7A, lane 1) and Egr-1 (Fig. 7B, lane 1) in intact cells. These results represent the first demonstration that Tax physically interacts with members of the zinc finger family of transcription factors.
Analysis of the ability of Tax mutants to transactivate the c-sis/PDGF-B promoter identified two Tax mutants, IEXC29S and IEXL320G, that were significantly impaired in their ability to transactivate the c-sis/PDGF-B promoter (Fig. 8). Coimmunoprecipitation analysis also revealed that Tax mutant IEXC29S was unable to interact with Egr-1 (Fig. 9, compare lane 5 versus lane 6), whereas, in contrast, Tax mutant IEXL320G was able to interact with Egr-1 (Fig. 9, compare lane 5 versus lane 7). This would suggest that an interaction between Tax and Egr-1 is necessary, but not sufficient, to support Tax-mediated transactivation of the c-sis/PDGF-B promoter. The fact that Tax mutant IEXL320G still binds to Egr-1, but fails to transactivate the c-sis/PDGF-B promoter, is not unexpected. The defect in IEXL320G most likely results from a mutation in the transactivation domain of Tax. Indeed, the presence of such a domain in Tax has been previously demonstrated (65). It was shown that amino acid residues 284 to 322 in the carboxyl-terminal region of Tax constituted a functional transactivation domain. Similar mutations of the same amino acid residues in Tax have been reported to be defective in transactivation of the HTLV-1 long terminal repeat by Tax, but have been shown to still be able to interact with CREB and the HTLV-1 21-bp repeats to assemble ternary Tax-CREB-DNA complexes (66).
As mentioned above, Tax is able to increase the DNA binding activity of a wide variety of eukaryotic transcription factors. Perhaps the best characterized of these is the basic regionleucine zipper (bZIP) DNA-binding domain family, which includes members of the CREB/ATF protein family. Dimerization is essential for bZIP proteins to bind DNA. Consequently, it was demonstrated that Tax stimulated the DNA binding activities of CREB/ATF proteins both quantitatively and qualitatively (67-69). First, it was shown that Tax enhanced dimerization, thereby increasing the overall level of DNA binding. Second, Tax altered DNA-binding site selectivity. Several reports then demonstrated that Tax was able to physically interact with several members of the CREB/ATF protein family (66,70,71). A more recent report demonstrated that the physical interaction between the bZIP protein, CREB, and Tax to form a Tax-CREB-DNA ternary complex involves the incorporation of Tax into the ternary complex as a dimer. In addition, the ability of Tax to form a dimer was shown to be necessary for its interaction with CREB (72). It has not escaped our attention that Tax's interactions with bZIP proteins deals exclusively with the ability of both Tax, and the bZIP protein, to dimerize. When one considers the situation involving the interaction(s) between Tax and members of the zinc finger family of transcription factors, the fact that zinc finger transcription factors do not normally dimerize prior to DNA binding, or, that dimerization is not essential for their DNA binding activity, immediately becomes apparent.
The results presented in this article support the possibility of the existence of an additional, as yet uncharacterized, pathway of transactivation by Tax involving members of the zinc finger family of transcription factors. Several questions as to the exact nature of this interaction between Tax and zinc finger family transcription factors still remain to be answered. First, and foremost, the issue of whether Tax dimerization is necessary for engaging in interactions with zinc finger transcription factors will have to be addressed. In addition, a more detailed analysis of the domains within Tax that are responsible for interacting with the various zinc finger family transcription factor members will need to be undertaken. Experiments are currently underway to address these and other aspects of the Tax-mediated transactivation of the c-sis/PDGF-B proto-oncogene.