Role of the Transcription Start Site Core Region and Transcription Factor YY1 in Rous Sarcoma Virus Long Terminal Repeat Promoter Activity

Abstract

The Rous sarcoma virus (RSV) long terminal repeat (LTR) contains a transcriptionally potent enhancer and promoter that functions in a variety of cell types. Previous studies have identified the viral sequences required for enhancer activity, and characterization of these elements has provided insight into the mechanism of RSV transcriptional activity. The objective of this study was to better define the RSV LTR promoter by examining the transcription start site core (TSSC) region. Deletion of the TSSC resulted in complete loss of transcriptional activity despite the presence of a functional TATA box, suggesting that the TSSC is required for viral expression. Homologies within the TSSC to the DNA binding motif of YY1 suggested that it might regulate promoter activity. YY1 has been shown to regulate transcription in some cellular genes and viral promoters by binding to sites overlapping the transcription start site. Gel shift assays using YY1 antibody identified YY1 as one of three complexes that bound to the TSSC. Mutation of the YY1 binding site reduced RSV transcriptional activity by more than 50%, suggesting that YY1, in addition to other TSSC-binding factors, regulates RSV transcription. Furthermore, in vitro transcription assays performed with Drosophila embryo extract (devoid of YY1 activity) showed decreased levels of RSV transcription, while transient transfection experiments overexpressing YY1 demonstrated that YY1 could transactivate the RSV LTR ∼6- to 7-fold. We propose that the TSSC plays a vital role in RSV transcription and that this function is partially carried out by the transcription factor YY1.

The Rous sarcoma virus (RSV) long terminal repeat (LTR) contains enhancer and promoter sequences that are active in a wide variety of eukaryotic cells (19), providing an excellent model system in which to study the molecular events involved in eukaryotic gene expression. Cullen et al. (15) demonstrated that the RSV LTR contains all of the functional elements required for efficient transcription (19). Deletion analysis and enhancer trap experiments localized the viral sequences necessary for enhancer activity to sequences within the U3 region of the LTR consisting of the nucleotides spanning from −229 to −54 (relative to the transcription start site) (15, 20, 38, 43, 76). We have previously defined three types of enhancer factor complexes (designated EFI, EFII, and EFIII) that bind to different sites in the RSV LTR enhancer (6, 17, 22, 63, 64) (Fig. 1). Other investigators have also characterized protein factors that bind to these sites within the RSV LTR enhancer (8, 18, 33, 52, 59–61, 64). The RSV LTR core promoter contains a well-conserved TATA box appropriately positioned at ∼30 bp upstream of the transcription initiation site (15, 20). Deletion of the enhancer and/or promoter elements results in significant loss of transcriptional activity.

FIG. 1.

Schematic diagram depicting the RSV LTR regulatory sequences. The RSV enhancer spans from nt −229 to −54. Shown within the enhancer are the DNA cis-acting elements EFI, EFII, and EFIII. Each of these enhancer factor binding sites has been previously characterized and shown to be crucial for viral expression (6, 17, 22, 63, 64). Broadly, each EFI element contains an inverted CCAAT motif (Y box). The EFII site is composed of two C/EBPβ consensus sequences, and the EFIII elements contain CArG motifs. The RSV promoter contains a canonical TATA box at approximately −30 bp. The sequence of the TSSC element is shown. The TSSC consists of nt −5 to +26 and overlaps the transcription start site. The initiating nucleotide is marked with an arrow.

The initiation of mRNA synthesis is a fundamental regulatory point in gene expression. Two control elements commonly found in the core promoter of protein-coding genes are the TATA box, located 25 to 30 bp upstream of the transcription initiation site, and the initiator (Inr), which overlaps the transcription start site (9, 24, 30, 34, 70–72). Core promoters can contain a TATA box, an Inr, both, or neither of these control elements. During transcription initiation, the TATA box is first recognized by the general transcription factor TFIID. TFIID is a multisubunit complex consisting of TATA-binding protein (TBP) and several TBP-associated factors (TAFs) (27). The TBP subunit is responsible for recognition of the A/T-rich TATA box sequence. Following template recognition, the formation of a preinitiation complex (PIC) competent for mRNA synthesis is completed through a multistep process in which the other general transcription factors (TFIIA, TFIIB, TFIIF-polymerase II, TFIIE, and TFIIH) are recruited to the DNA template (reviewed in references 14, 50, and 51).

The pyrimidine-rich Inr element (core consensus sequence Py₂CAPy₄) functions similarly to the TATA box in that it is sufficient for directing accurate basal transcription by RNA polymerase II (30, 72). The Inr can function independently or synergistically with the TATA box to direct formation of the PIC, and many core promoters that contain one or both of these elements have been characterized (12). Although many genes that contain a functional Inr have been characterized, the mechanism by which the Inr directs formation of the PIC and the proteins that specifically interact with the Inr element to affect TFIID recruitment and/or function remain unclear. However, several transcription factors, including TFIID (5, 11, 32, 72), TFII-I (31, 46, 53–56), USF (16, 55), E2F (34), specific TAFs, RNA polymerase II (12, 51), and the multifunctional transcription factor YY1 (65, 73), have been reported as Inr-binding proteins.

YY1 (also called δ, UCRBP, and NF-E1) is a 414-amino-acid zinc finger protein with an apparent molecular mass of 65 to 68 kDa in an sodium dodecyl sulfate-polyacrylamide gel (4, 26, 49, 68). YY1 is ubiquitously expressed, belongs to the GLI-Krüppel family of proteins, and is highly conserved between mouse and human (98.6% amino acid identity) (26). The YY1 consensus binding site contains a conserved 5′-CAT-3′ core flanked by variable regions (28, 35, 77). YY1 regulates the expression of a variety of cellular and viral genes by functioning as a repressor or an activator of transcription (reviewed in references 67 and 69). In particular, YY1 has been shown to regulate the expression of the adeno-associated virus P5 promoter (AAV P5) (65, 73) and the cytochrome oxidase Vβ subunit promoter (COX Vβ) (3) by binding to their transcription start sites. Usheva and Shenk have demonstrated that a complex containing only YY1, TFIIB, and polymerase II is able to direct specific basal transcription from the AAV P5 promoter (73) and recently proposed a mechanistic basis for the TBP-like function of YY1 (74). Indeed, YY1 has been shown to be a component of the RNA polymerase II holoenzyme as well as to interact with several proteins involved in polymerase II transcription, including TBP, TFIIB, TAF_II55, and p300 (1, 45, 74).

In this report, we characterize the sequences that comprise the RSV LTR transcription start site core (TSSC), which is positioned downstream of the TATA box and overlaps the transcription initiation site (Fig. 1). Our results demonstrate that deletion of the TSSC results in a complete loss of transcriptional activity from the LTR, despite the presence of an intact TATA box. We determined that three protein-DNA complexes, TSSC(A), TSSC(B), and TSSC(C), bind specifically to the TSSC region and identified TSSC(C) as the multifunctional transcription factor YY1. We demonstrate that the YY1 protein and its binding site are necessary for full enhancer and promoter activity from the RSV LTR and that YY1 activates transcription from the RSV LTR. Our data suggest that the RSV promoter contains a basal control element that is regulated by YY1, overlaps the transcription start site, and is crucial for viral expression.

MATERIALS AND METHODS

Cell culture, transfections, and enzymatic assays.

HeLa (cervix carcinoma) cells and QT6 (quail fibrosarcoma) cells were obtained from the American Type Culture Collection. HeLa cells were maintained in a 1:1 mixture of Dulbecco’s modified Eagle’s medium and Ham’s F-12 nutrient mixture (DMEM/F-12) enriched with 10% calf serum with ferric sulfate complex at 6 mg/ml. QT6 cells were maintained in medium 199 supplemented with 10% tryptose phosphate broth, 5% fetal bovine serum, and 1% chicken serum. In addition, the following were added to each medium: penicillin G sodium (25 U/ml), streptomycin sulfate in 0.85% saline (25 mg/ml), sodium bicarbonate (HeLa) (2.44 g/liter), and sodium bicarbonate (QT6) (1.7 g/liter).

Transfections were performed by the calcium phosphate coprecipitation technique as described by Graham and van der Eb (21). All cells were plated in a 60-mm-diameter dish 1 day prior to transfection at a density of 5 × 10⁵ cells/dish. HeLa cell transfections were performed with 10 μg of the plasmid DNA indicated in the relevant figure legend per 5 ml of DMEM/F-12. QT6 transfections were performed with either 2.5 μg or 50 ng of the plasmid DNA indicated in the relevant figure legend per 5 ml of medium 199.

Transfections overexpressing YY1 were performed with a total of 22.5 μg of DNA, consisting of 2.5 μg of reporter plus variable amounts of expression vector supplemented with parental plasmid to adjust for equal amounts of input DNA. Cells were exposed to the CaPO₄-DNA precipitate for 24 h (HeLa) or 6 to 8 h (QT6), at which time the medium was removed and replaced with fresh supplemented medium; cells were then allowed to grow for 36 to 48 hours. Cells were harvested by being washed once in cold (4°C) phosphate-buffered saline, incubated for 15 min at room temperature in phosphate-buffered saline containing 5 mM EDTA and 5 mM EGTA, and scraped from the plates. The cells were collected at 4°C by centrifugation for 5 min at 630 × g.

For chloramphenicol acetyltransferase (CAT) assays, the cell pellet was resuspended in 250 μl of 0.25 M Tris (pH 8.0) containing 1 mM phenylmethylsulfonyl fluoride, lysed by sonication, and clarified by centrifugation for 15 min at 10,000 × g as described by Boulden and Sealy (6). CAT assays were performed by the method of Nordeen et al. (48), and CAT activity was quantitated by liquid scintillation counting. We used the Promega luciferase assay system with reporter lysis buffer for luciferase assays carried out by the manufacturer’s protocol. Luciferase activity was assayed by using an Analytical Luminescence Laboratory Monolight 2010 luminometer with either 10 μl of extract or 10 μl of a 1:100 dilution of cell lysate. The total protein in each sample was determined by the Bradford assay, and reporter activity was normalized to the total protein in each sample.

Gel shift assays.

HeLa cell nuclear extracts were previously prepared by the method of Shapiro et al. (66). Gel shift assays were performed with a 1:10 dilution of HeLa cell nuclear extract, 1.25 μg of poly(dI-dC) · poly(dI-dC), and 0.5 ng of either TSSC wild-type, TSSC_5′mYY1, or TSSC_3′mYY1 ³²P-labeled DNA in a final volume of 20 μl containing 10 mM HEPES (pH 8), 5 mM Tris (pH 7.9), 2 mM dithiothreitol, 1 mM EDTA, 50 mM NaCl, and 20% glycerol. The probe used in each assay is indicated in the relevant figure legend. Nonradiolabeled competitor DNAs were added to the reaction before the addition of the probe, and the HeLa nuclear extract was always added to the reactions last. All antibodies were purchased from Santa Cruz Biotechnology, and supershift assays were performed by adding the antibody indicated in the relevant figure legend just prior to the addition of the HeLa nuclear extract. All binding reactions were carried out for 30 min at room temperature and then electrophoresed on a native 6% polyacrylamide gel containing 25 mM Tris base, 190 mM glycine, and 1 mM EDTA (pH 8) (TGE). Gels were subsequently dried for autoradiography.

Radiolabeled and competitor DNAs.

Oligonucleotides for radiolabeled probes and competitor DNAs were prepared by automated DNA synthesis in the Diabetes Research and Training Center DNA Core (Vanderbilt University) and purified over desalting columns followed by n-butanol extraction. The amount of DNA was quantitated by absorbance measurements at 260 nm. Oligonucleotides were labeled by using T4 polynucleotide kinase, electrophoresed on a 12% native polyacrylamide gel in Tris-borate-EDTA, and purified by electroelution as previously described (7). Nonradiolabeled competitor DNAs were purified and quantitated as described above. The sequences for the oligonucleotides used in this study are shown in Table 1.

TABLE 1.

Oligonucleotides used in this study

Oligonucleotide	Sequencea
TSSC	TAAACGCCATTTTACCATTCACCACATTGGT
	ATTTGCGGTAAAATGGTAAGTGGTGTAACCA
TSSCm5′YY1	TAAACGttgTTTTACCATTCACCACATTGGT
	ATTTGCaacAAAATGGTAAGTGGTGTAACCA
TSSCm3′YY1	TAAACGCCATTTTAttgTTCACCACATTGGT
	ATTTGCGGTAAAATaacAAGTGGTGTAACCA
TSSCm5′3′YY1	TAAACGttgTTTTAttgTTCACCACATTGGT
	ATTTGCaacAAAATaacAAGTGGTGTAACCA
YY1 consensus	CGCTCCGCGGCCATCTTGGCGGCTGGT
	GCGAGGCGCCGGTAGAACCGCCGACCA
XDH/XO Inr2	CCGGGAGGCGTATCTTTCAAGTTGCAGGGCAGT
	GGCCCTCCGCATAGAAAGTTCAACGTCCCGTCA

^aNon-wild-type sequences are in lowercase. 

Site-directed mutagenesis.

Single-stranded DNA to introduce specific mutations within the RSV LTR promoter was prepared by standard mutagenesis techniques (36, 37, 75) using Escherichia coli CJ236 (F′ dut ung mutant) (Invitrogen), and M13K07 phage (Promega). Oligonucleotides containing specific mutations flanked on either side by wild-type sequences were prepared and quantitated as described above. The oligonucleotides were phosphorylated with T4 polynucleotide kinase as described by Boulden and Sealy (7) except that nonradioactive ATP was used. Double-stranded DNA was synthesized by elongation of a primer that was annealed to the single-stranded DNA as described elsewhere (37). The presence of the specific mutation was confirmed by chain termination DNA sequencing of the plasmids as described by Sanger et al. (62). The oligonucleotides used to generate the ClaI restriction site (SRA ClaI), TSSC deletion (ΔTSSC), and 5′YY1 mutation in the RSV LTR (5′mYY1) are 5′-TGCAATGCGGAATTCATCGATTCGTCCAATCCA-3′, 5′-TCAACCCAGGTGCACTTGTATCGAGCTAGG-3′, and 5′-GGTGAATGGTCAAACAACGTTTATTGTATC-3′, respectively.

Plasmid construction of reporter genes.

Plasmid p(B)SRA was created from the reporter gene plasmid p(B)SRA CAT (6). Plasmid p(B)SRA CAT was digested with the restriction enzyme AccI, blunt ended with Klenow fragment of DNA polymerase, and then digested with BamHI. The digested products were size fractionated on a 1% agarose gel, and the appropriate DNA fragment was excised and purified on a Spin-X column (Costar no. 8169), followed by phenol-chloroform (1:1) extraction, chloroform-isoamyl alcohol (24:1) extraction, and precipitation in 2 volumes of ethanol and 1/10 volume of 3 M sodium acetate (pH 5) at −20°C overnight. This fragment, which contains RSV LTR sequences from −489 to +103 of the provirus linked to the CAT gene, was subcloned into the EcoRV and BamHI sites of plasmid pBluescript KS II(+) (Stratagene). The oligonucleotide 5′-TGCAATGCGGAATTCATCGATTCGTCCAATCCA-3′ was used for site-directed mutagenesis to introduce a ClaI site (shown in boldface) into this plasmid at the beginning of the RSV LTR promoter sequence (−54 relative to the transcription start site). The resulting plasmid, designated p(B)SRA CAT, was used to generate the p(B)e⁻CAT vector by digestion with ClaI to remove a 624-bp fragment (containing the RSV enhancer and vector sequence), and the remaining DNA was recircularized by ligation. The reporter constructs containing the mutation in the TSSC 5′YY1 site were created by site-directed mutagenesis from their wild-type counterpart and resulted in the following mutation (indicated in lowercase) of the TSSC 5′YY1 site: CGCCATTTTA→CGttgTTTTA. Plasmids pSRA-Luc and pSRA_5′mYY1-Luc were derived from plasmids p(B)SRA CAT and p(B)SRA_5′mYY1 CAT, respectively. The HindIII restriction fragments of plasmids p(B)SRA CAT and p(B)SRA_5′mYY1 CAT (containing the RSV enhancer and minimal promoter) were subcloned into the HindIII site in the pGL3-basic luciferase vector (Promega). Plasmids pe⁻-Luc and pe⁻_5′mYY1-Luc were prepared by inserting the XhoI-HindIII restriction fragments from p(B)e⁻CAT and p(B)e⁻_5′mYY1CAT, respectively, into the XhoI-HindIII restriction sites of the pGL3-basic luciferase vector. The expression vector pSVK3/YY1 has been previously described (42).

In vitro transcription and primer extension experiments.

Helascribe HeLa cell transcription extract and Drosophila embryo extract were purchased from Promega and used in all transcription assays. Transcription reactions were performed by preincubating 6 μl of HeLa extract or 2 μl of Drosophila extract with 400 ng of supercoiled template DNA on ice for 30 min in a buffer containing ∼20 U of RNAsin 7.5 mM MgCl₂, 70 mM KCl, 20 mM HEPES (pH 7.9), 0.2 mM EDTA (pH 8), 0.2 mM EGTA, and 2 mM dithiothreitol. After addition of ribonucleoside triphosphates to a final concentration of 500 μM, the reactions were incubated for 1 h at 30°C. The RNA was purified by digesting protein present in the reactions with 0.2% sodium dodecyl sulfate and 125 μg of proteinase K per ml at 37°C for 1 h and removing the contaminating protein by phenol-chloroform and chloroform-isoamyl alcohol extraction. The RNA was then precipitated in 2 volumes of ethanol and 1/10 volume of 3 M sodium acetate (pH 5).

Primer extension analysis was carried out with a synthetic oligonucleotide designated 3′CAT (5′-CTCCATTTTAGCTTCCTTA-3′), which is complementary to the CAT gene sequence present downstream of the promoter elements. Reactions containing 5 × 10⁵ cpm of the 3′CAT primer and the purified RNA were annealed in 20 mM Tris (pH 7.5)–250 mM NaCl–1 mM EDTA (pH 8) at 55°C for 1 h. Elongation of the transcripts was performed by adding reverse transcriptase buffer (Promega), 500 μM deoxynucleoside triphosphates, and ∼10 U of avian myeloblastosis virus reverse transcriptase and incubating the reactions for 1 h at 42°C. The extended products were separated on a 8% denaturing polyacrylamide gel and dried for autoradiography.

RESULTS

The activity of the RSV LTR promoter is dependent on specific sequences around the transcription start site.

Localization of the DNA sequences essential for RSV enhancer activity has previously been shown to reside within the U3 region of the LTR (15, 20, 38, 43, 76). However, previous studies to delineate the core promoter sequences that are necessary for efficient viral expression have demonstrated variable requirements for sequences downstream of the TATA box (15, 47). Therefore, to characterize the potential of the RSV transcription start site region to regulate transcription, we initially analyzed promoter constructs containing a deletion of the TSSC region. The plasmid, p(B)SRA, which contains the LTR enhancer and promoter and a similar construct, p(B)e⁻, that lacks the enhancer were modified by site-directed mutagenesis techniques to create the deletion mutants p(B)SRA_ΔTSSC and p(B)e⁻_ΔTSSC, respectively (Fig. 2A). The TSSC deletion removed the sequences from −5 to +26 with respect to the transcription start site in the wild-type LTR. In vitro transcription assays were performed to identify mRNA transcripts initiating from within the LTR constructs. Transcription from the deletion mutants would be directed by the intact TATA box or the TATA box in conjunction with the upstream enhancer elements. The results of these in vitro transcription experiments are shown in Fig. 2B. Correctly initiated transcripts were detected with the wild-type DNA templates containing both the TATA box and the TSSC (lanes 1 and 3). However, with both of the ΔTSSC mutant plasmids, no transcripts of 150 nucleotides (nt) were detected as would be expected from mRNA synthesis directed by the TATA box [repeated with two independent preparations of p(B)SRA_ΔTSSC, and p(B)e⁻_ΔTSSC]. An increase in start site heterogeneity can occur with alterations of basal control elements, but this was not observed with the deletion mutant plasmids. These results suggest the presence of a critical basal control element within the TSSC that is absolutely required for promoter activity. Thus, the RSV LTR TATA box, although well conserved, is not capable of directing transcription initiation alone or in conjunction with the upstream enhancer elements; rather, its activity is dependent on specific downstream sequences contained within the TSSC to function.

FIG. 2.

Identification of the TSSC as a critical transcriptional regulatory element. (A) Schematic diagram of the reporter constructs used in the in vitro transcription and transfection experiments described below. Plasmids p(B)SRA and p(B)SRA_ΔTSSC both contain the RSV enhancer, whereas p(B)e⁻ and p(B)e⁻_ΔTSSC do not contain the RSV enhancer. p(B)SRA_ΔTSSC and p(B)e⁻_ΔTSSC contain identical deletions of the TSSC element (nt −5 to +26). Each reporter plasmid drives the expression of the CAT gene. (B) In vitro transcription assays were carried out with 400 ng of template DNA and 8 to 12 μg of HeLa cell nuclear extract. The RNA transcripts were analyzed by primer extension using the 3′CAT gene primer (5′-CTCCATTTTAGCTTCCTTA-3′). The arrow indicates the position of the 150-nt band which resulted from the cDNA products of correctly initiated transcripts from within the TSSC element. The autoradiograph shown is representative of at least three experiments. Similar results were obtained with different preparations of both HeLa extract and p(B)SRA_ΔTSSC and p(B)e⁻_ΔTSSC DNA templates. (C) Transient transfection assays were performed by introducing 10 μg of the CAT reporter construct p(B)SRA or p(B)e⁻ or the corresponding transcription start site core deletion mutant construct p(B)SRA_ΔTSSC or p(B)e⁻_ΔTSSC into HeLa cells. Protein extracts were prepared from cells harvested 36 to 48 h posttransfection as described in Materials and Methods. The CAT activity for each transfection was calculated and normalized to total protein as described in Materials and Methods. The graph represents data from at least three separate experiments and two different preparations of plasmids p(B)SRA_ΔTSSC and p(B)e⁻_ΔTSSC. The error bars show standard errors.

Previous studies analyzing transcription start site sequence requirements have detected differences in promoter activity when assayed in vitro and in vivo (13, 30, 78). For this reason, we extended the analysis of the RSV LTR by performing transient transfection assays to analyze the contribution of the TSSC region to promoter activity in vivo. For these experiments, the reporter constructs p(B)SRA, p(B)e⁻, p(B)SRA_ΔTSSC, and p(B)e⁻_ΔTSSC were transfected into HeLa cells and the transcriptional activity from each promoter construct, as assayed by CAT activity, was compared to the activity of extracts from mock-transfected cells. Figure 2C shows the results obtained in vivo. Deletion of the TSSC reduced the transcriptional activity to background levels, similar to the in vitro results described above. Transcriptional activation by the RSV enhancer was not detected in vitro, but in vivo an ∼300-fold activation was seen [compare p(B)SRA to p(B)e⁻]; however, the loss of transcription upon deletion of the TSSC sequences was independent of the presence of the RSV enhancer, suggesting that the TSSC is indeed a basal control element and is not solely required for enhancer function. Taken together, these results demonstrate that the RSV LTR TSSC is a core promoter element required for both enhanced and basal transcriptional activity.

The TSSC sequence binds three protein complexes and contains YY1 binding sites.

We showed that the TSSC was essential for RSV LTR transcriptional activity. TSSC function is most likely mediated by sequence-specific DNA-binding factors. Therefore, to begin the analysis of the TSSC, we performed gel shift assays using HeLa cell nuclear extract to detect any transcription factors binding to this region of the DNA. We observed three protein-DNA complexes, designated TSSC(A), TSSC(B), and TSSC(C), binding to the element (Fig. 3A, lane 1). To determine the specificity of these protein complexes for the TSSC element, we performed competition gel shift assays. Wild-type nonradiolabeled TSSC DNA competed all three complexes (lane 2), whereas a 33-bp nonspecific competitor oligonucleotide containing sequences of the xanthine dehydrogenase/xanthine oxidase promoter (XDH/XO) did not affect the binding of any of the three complexes (lane 3). A serendipitous inspection of the TSSC sequence identified two potential binding sites for the multifunctional transcription factor YY1 (Fig. 3B) (26). YY1 has been shown to bind to sequences overlapping transcription start sites, where it functions to activate transcription (2, 3, 65). As shown in Fig. 3B, the TSSC contains two copies of the core sequence 5′-CCAT-3′ that is necessary for YY1 binding (28, 30). Based on the sequence similarity between the TSSC and YY1 binding sites, we explored the possibility that YY1 is a component of these complexes.

FIG. 3.

Characterization of the RSV TSSC element by gel shift analysis. (A) Three specific protein complexes were found to bind to the TSSC element by performing a gel shift assay using HeLa nuclear extract and 0.5 ng of ³²P-labeled TSSC DNA in the presence of 1.25 μg of poly(dI-dC) · poly(dI-dC) as described in Materials and Methods. Either no competitor (lane 1) or a 100-fold molar excess of competitor oligonucleotide (lane 2 and 3) was added to the binding reactions. Each complex was competed for binding by addition of TSSC wild-type competitor (lane 2), but a nonspecific competitor, XDH/XO, did not affect the binding of the TSSC complexes (lane 3). (B) The consensus sequence for YY1 is shown. The YY1 binding site contains a conserved core 5′-CCAT-3′, which is essential for efficient binding, and is flanked on either side by variable regions. Two potential binding sites for YY1 were identified within the TSSC element. Each putative binding site for YY1 is marked with a black bar, and the conserved core is highlighted with a black box.

Identification of TSSC(C) as YY1.

The binding specificity of the TSSC complexes was addressed by performing gel shift assays in the presence of a YY1 competitor DNA that contained the YY1 consensus site centered in the oligonucleotide sequence. Figure 4A shows that the TSSC(C) complex was eliminated in the presence of competitor, and the TSSC(A) and -(B) protein-DNA complexes were not competed with the YY1 consensus oligonucleotide. Therefore, only the TSSC(C) complex required a YY1 sequence element for complex formation.

FIG. 4.

YY1 binds to the TSSC element. (A) Competition gel shift assays were performed with a 27-bp oligonucleotide containing the consensus sequence for YY1 (see Materials and Methods). Increasing amounts of YY1 consensus DNA were added to gel shift reactions (lanes 2 to 5) to demonstrate that the TSSC(C) band was specifically competed, while TSSC(A) and TSSC(B) shifts remained unaffected. NE, nuclear extract. (B) Gel shift assays were performed with HeLa nuclear extract and 0.5 ng of ³²P-labeled TSSC DNA in the presence of 1.25 μg of poly(dI-dC) · poly(dI-dC) as described in Materials and Methods. Either no antibody (lane 1) or 1 μl of polyclonal antibody to YY1 (lane 2), SRF (lane 3), or the Gal4 DBD (lane 4) was added to the reactions prior to the addition of the HeLa extract. Arrow 1 indicates the supershifted complex observed when anti-YY1 antibody is added. Antibodies to SRF or Gal4DBD do not affect the binding activity of any of the TSSC complexes.

To determine if any of the TSSC complexes are immunologically related to YY1, we performed gel shift assays using a polyclonal antibody directed toward YY1. As seen in Fig. 4B, the TSSC(C) complex is shifted to a position comigrating with the TSSC(A) complex upon addition of YY1 antibody. In gel shift assays using a different preparation of YY1 antibody, we alternatively observed an elimination of TSSC(C) binding whereas TSSC(A) remained visibly unaffected by the addition of anti-YY1 (data not shown). TSSC(C) was unaffected by addition of antibody directed against either serum response factor (SRF) or the DNA binding domain of the yeast activator Gal4 (Gal4 DBD). The two upper complexes, TSSC(A) and -(B), were not affected by addition of YY1, SRF, or Gal4 DBD antibodies. In addition, gel shift assays performed with the TSSC element demonstrated that the electrophorectic mobilities of TSSC(C) and in vitro translated YY1 protein were identical (data not shown). Taken together these data confirm that the TSSC(C) complex is composed of YY1.

YY1 binds exclusively to the 5′YY1 TSSC consensus site.

We have demonstrated that TSSC(C) is composed of YY1. The TSSC element contains two sequences with homology to the YY1 consensus binding site. Table 2 compares the YY1 sites present in the RSV LTR TSSC element with YY1 sites in a variety of gene promoters and enhancers. The TSSC 5′YY1 site is identical to the YY1 consensus motif (28), compared to 55% sequence similarity with the TSSC 3′YY1 site (Table 2). To determine the site specificity of YY1 within the TSSC element, we performed gel shift assays using competitors with mutations in the YY1 binding sites.

TABLE 2.

Sequence comparison of the YY1 binding sites within the RSV LTR with YY1 sites in other genes and enhancers

Gene or enhancer	Sequence	Reference
RSV LTR (5′YY1)	CGCCATTTT	This study
RSV LTR (3′YY1)	TACCATTCA	This study
Moloney murine leukemia virus LTR	CGCCATTTT	28
AAV P5 (−60)	CGACATTTT	28
AAV P5 (+1)	CTCCATTTT	28
rpL32	TGCCATCTG	26
rpL30	GCCCATCTT	26
Cox Vβ	GCCCATCTT	3
Serum amyloid A1	CACCATGTC	42
δ-Globin	TGACATATT	28
Igκ3′ enhancer	CTCCATCTT	49
GRP78/BiP	GGCCAGCTT	39
YY1 consensus	CGCCATTTT

Oligonucleotides containing a mutation of the core consensus sequence in either the 5′YY1 site (m5′YY1), 3′YY1 site (m3′YY1), or both sites (m5′3′YY1) of the TSSC element were prepared. The mutations changed the YY1 core consensus from CCAT to ttgT. Javahery et al. demonstrated that these mutations eliminate YY1 binding (30). The results of these experiments are shown in Fig. 5A. Addition of increasing amounts of either wild-type or m3′YY1 oligonucleotide competed all three complexes (lanes 2 to 9). The competitors m5′YY1 and m5′3′YY1 competed effectively for TSSC(A) and -(B) complex formation but not TSSC(C)/YY1 formation. This suggested that the TSSC(C)/YY1 complex was unable to bind the TSSC when a mutation in the 5′YY1 site was present but could actively bind the TSSC element in the presence of a 3′YY1 mutation (lanes 10 to 17).

FIG. 5.

Specificity of YY1 binding to the TSSC element. (A) Competition gel shift assays were performed to define the binding specificity of YY1 to the TSSC element. Assays were performed with HeLa nuclear extract and labeled TSSC DNA as the probe as described in Materials and Methods. Each competitor used was added in 5, 25, 50, and 100-fold molar excess. Lane 1 contained no competitor; lanes 2 to 5 contained the wild-type competitor TSSC element; lanes 6 to 17 contained competitor with mutations in the YY1 core consensus site. The YY1 core sequence 5′-CCAT-3′ was changed to 5′-ttgT-3′ in the 5′ and 3′ sites of TSSC_m5′YY1 and TSSC_m3′YY1 oligonucleotides, respectively. A YY1 TSSC double mutant that contained the same mutations as described above in each site was used as competitor in lanes 14 to 17. (B) The ability of the TSSC YY1 mutant oligonucleotides to form a YY1 complex was examined by gel shift assay. Wild-type TSSC and 5′ and 3′YY1 mutant oligonucleotides were labeled and used in gel shift assays. When TSSC_m5′YY1 was used as a probe TSSC(A) and -(B) complexes were detected, but TSSC(C)-YY1 binding activity was not observed (lane 2). Each of TSSC(A), -(B), and -(C) was able to bind the TSSC_m3′YY1 oligonucleotide (lane 3).

To directly test the ability of TSSC(C)/YY1 to bind to the 5′ and/or 3′ site within the TSSC, we performed gel shift assays with the radiolabeled mutant YY1 oligonucleotides described above. The results of these experiments are shown in Fig. 5B. While TSSC(B) binding activity was not affected by the YY1 mutations, the formation of TSSC(C)/YY1 was dependent on an intact 5′YY1 binding site (compare lanes 1 and 2). Also, an oligonucleotide with a deletion of the 5′YY1 site formed only TSSC(A) and -(B) (data not shown). Mutation of the 3′YY1 site did not affect TSSC(C)/YY1 complex formation [TSSC(C) in lane 3]. No TSSC(A) binding activity was detected in the presence of the 3′YY1 mutation (lane 3). The mutation of the 3′YY1 site may have produced a lower affinity binding site for the TSSC(A) factor that prohibited its binding to TSSC_m3′YY1 DNA but still competed when present at high concentrations. This is likely since at lower levels of competitor, the oligonucleotides with the 3′YY1 mutation do not compete analogously to competitors with this site intact (Fig. 5A; compare lanes 2 and 10 to lanes 6 and 14). These results show that YY1 binds exclusively to the 5′YY1 site in the TSSC element and that the 3′YY1 site may bind the TSSC(A) factor, whose binding site is likely to include the CCA nucleotides located at positions +10 to +12.

The RSV LTR 5′YY1 site is required for normal viral enhancer and promoter activity.

We wanted to test the functional significance of the YY1 binding site in RSV LTR transcription. To do this, the CAT reporter constructs p(B)SRA_m5′YY1 and p(B)e⁻_m5′YY1 (Fig. 6A), which contain mutations identical to those which prohibited TSSC(C)/YY1 complex formation, were used in transcription assays in vitro. As shown in Fig. 6B, specific initiation of transcription from the YY1 mutant DNA templates, which would produce a 150-nt transcript, is significantly less than wild-type activity. In addition, we detected a transcript initiating slightly further upstream in the mYY1 templates. The increased heterogeneity in start site selection from the mYY1 templates indicates that YY1 may play a role in specifying the initiating nucleotide. In addition, the reduced levels of specific transcripts suggests a role for YY1 in basal transcription.

FIG. 6.

Effect of YY1 mutation RSV transcriptional activity. (A) Schematic diagram of the reporter constructs used in the in vitro transcription and transfection experiments described below. The reporter constructs p(B)SRA and p(B)e⁻ are depicted in Fig. 2A. The mutation of the 5′YY1 site was identical to the mutation used in the gel shift analyses in Fig. 4. The reporter plasmids pSRA-Luc, pSRA-Luc_m5′YY1, pe⁻-Luc, and pe⁻-Luc_m5′YY1 contain enhancer and promoter sequences identical to those in the corresponding CAT constructs but drive the expression of the luciferase gene. (B) In vitro transcription assays were carried out with 400 ng of template DNA and 6 μl of HeLa cell nuclear extract. The RNA transcripts were analyzed by primer extension using a CAT gene primer. The arrow indicates the position of the 150-nt band which results from the cDNA products of correctly initiated transcripts from within the TSSC element. (C) HeLa cells were transfected with 10 μg of either p(B)SRA, p(B)SRA_m5′YY1, p(B)e⁻, or p(B)e⁻_m5′YY1 or no DNA (Mock) as described in Materials and Methods. QT6 cells were transfected with 50 ng of the enhancer-containing construct pSRA-Luc or pSRA-Luc_m5′YY1 or 2.5 μg of the enhancerless construct pe⁻-Luc or pe⁻-Luc_m5′YY1 as described in Materials and Methods. The CAT or luciferase activity for each transfection was calculated and normalized to total protein as described in Materials and Methods. The reporter activity was calculated and graphed relative to the corresponding wild-type construct, which was set at 100%. The graph represents data from at least three separate experiments. The error bars show standard errors; each star indicates the construct used for normalization and therefore does not contain standard error.

To test the functional significance of the YY1 binding site within the TSSC in vivo, we assessed the effect of the 5′YY1 mutation on reporter activity in HeLa cells and the Japanese quail fibrosarcoma cell line QT6 (a natural avian host cell line susceptible to infection by the RSV). The activities of the 5′YY1 mutant DNA templates were compared to the activity of their wild-type counterpart (set at 100%), and the results of these assays are shown in Fig. 6C. In HeLa cells, mutation of the 5′YY1 site reduced transcriptional activity of the enhancer-containing promoter to 43.8% ± 4.0% and that of the enhancerless promoter to 52.9% ± 5.9%. From these results, we conclude that it is unlikely that the 3′YY1 site is able to compensate for the 5′ mutation by binding YY1, since neither the TSSC(C)/YY1 complex nor the presence of a newly shifted complex was detected when the TSSC_5′mYY1 oligonucleotide, which leaves the 3′YY1 site unaffected, was used as a probe in gel shift assays (Fig. 5B). However, the effect of the 5′YY1 mutation on transcriptional activity was not analogous to the effect that we observed when the TSSC region was deleted (Fig. 2). Since binding of the TSSC(A) and -(B) complexes remained when the 5′YY1 site was altered, these factors would most likely contribute to the functional activity in the 5′YY1 mutant templates but would not be present in the ΔTSSC constructs to augment RSV activity. In any case, we observed that the transcriptional activity of p(B)SRA_m5′YY1 remained ∼260-fold higher than the activity of p(B)e⁻_m5′YY1 (data not shown); this indicates that in HeLa cells, enhancer function is not coupled to the YY1 binding motif and suggests a role for YY1 in basal promoter function. In QT6 cells, mutation of the 5′YY1 site reduced transcriptional activity of the enhancer containing promoter to 54.1% ± 7.9%, similar to the result for HeLa cells. Interestingly, mutation of the 5′YY1 site in the enhancerless construct had a much more profound effect on transcriptional activity from the RSV LTR, reducing activity to 12.7% ± 0.9%. These results were similar to the effect observed when the TSSC region was deleted (Fig. 2). Apparently in a natural host cell, the RSV minimal promoter is more dependent on the YY1 binding site for transcriptional activity. Whether this reflects cell-type-specific differences in the ability of TSSC(A) or TSSC(B) to compensate for YY1 loss remains to be determined.

The YY1 protein positively regulates the RSV LTR.

To directly confirm that YY1 exerts a positive effect on the RSV LTR, we tested whether overexpression of YY1 could transactivate the RSV LTR in cotransfection assays. Increasing amounts of the YY1 expression plasmid pSVK3/YY1 were cotransfected into HeLa cells with either the pSRA-Luc or pe⁻-Luc plasmid. As shown in Fig. 7A, compared to a control containing the reporter plus the parental expression vector, a dose-dependent response was observed for both the enhancer-containing and enhancerless constructs, with maximal stimulation of YY1 expression plasmid resulting in a ∼6- to 7-fold activation of pSRA-Luc or pe⁻-Luc. By immunoblot analysis, we detected increasing amounts of YY1 protein expressed with increasing amounts of the pSVK3/YY1 expression vector following transfection into HeLa cells (data not shown). As an additional control, the empty reporter vector plus the maximal amount of YY1 expression plasmid was included in the cotransfection assays. The low level of expression observed with the empty reporter vector was not altered by YY1 overexpression. Taken together, these data suggest that the YY1 protein plays a positive regulatory role in RSV LTR transcription.

FIG. 7.

YY1 activates transcription from the RSV LTR. (A) YY1 was tested for its ability to transactivate the RSV LTR. The indicated amounts of the reporter construct pSRA-Luc or pe⁻-Luc or the empty vector pGL3-basic were cotransfected into HeLa cells with the indicated amounts of the YY1 expression plasmid pSVK3/YY1 as described in Materials and Methods. The total amount of DNA was kept constant by addition of the empty pSVK3 expression vector. Cells were harvested 36 to 48 h posttransfection. Luciferase activity was calculated from either 10 μl (pe⁻-Luc) or 10 μl of a 1:100 dilution (pSRA-Luc) of extract and normalized to total protein present in the cell lysate (see Materials and Methods). The graph represents the fold activation of at least three independent experiments, with the specific activation and standard error for each experiment indicated below. RLU, relative luciferase units. (B) In vitro transcription experiments were carried out with HeLa nuclear extract or Drosophila (Dros.) embryo extract with 400 ng of the DNA template, p(B)SRA (lanes 1 and 2), p(B)e⁻ (lanes 3 and 4), or the YY1-independent promoter plasmid phosphoenolpyruvate carboxykinase (PEPCK) (lanes 5 and 6). Transcripts were analyzed by primer extension as described in Materials and Methods. The arrows indicate the positions of correctly initiated transcripts. Autoradiographs are representative of results obtained from at least three separate experiments.

Much of the general transcriptional machinery is largely conserved between organisms as divergent as yeast, Drosophila, and human (25). Although Drosophila and HeLa cells contain functionally interchangeable RNA polymerase II basal transcription factors, many sequence-specific DNA-binding factors are not conserved. Since YY1 activity is absent in Drosophila embryo extracts (65), we tested the ability of this extract to support RSV LTR transcription. Transcription experiments performed in vitro with either p(B)SRA or p(B)e⁻ and Drosophila embryo extract showed that specific initiation was reduced compared to identical experiments using HeLa nuclear extract (Fig. 7B; compare lanes 1 and 2 and lanes 3 and 4). An alternative RNA product slightly larger than the expected 150-nt RNA transcript was detected from both promoter plasmids (lanes 2 and 4), similar to the effect of the 5′YY1 mutation. To account for possible differences in activity between the Drosophila embryo extract and HeLa nuclear extract, we tested the ability of these extracts to support accurate basal level transcription from a YY1-independent DNA template, phosphoenolpyruvate carboxykinase promoter, that contains the minimal sequence requirements for basal-level expression (29, 44). Correctly initiated transcripts corresponding to 110 nt were detected in each extract, and the level of transcription observed in the Drosophila embryo extract was comparable with that seen in the HeLa cell nuclear extract (lanes 5 and 6). This finding suggests that the removal of a sequence-specific DNA-binding protein that interacts with the viral promoter, such as YY1, functions to activate transcription, as well as contribute to start site selection.

DISCUSSION

Although the critical cis-acting DNA elements required for RSV enhancer function have largely been determined, important core promoter elements outside the TATA box have not been described in detail. Clearly, elucidation of transcriptional regulatory mechanisms relies on thorough characterization of promoter sequence. To understand the potent transactivation potential of enhancer-bound activators, we have further characterized the RSV LTR promoter to identify possible targets for regulation within the basal transcriptional machinery. In this study, we have identified a core region, the TSSC, which encompasses the transcription start site and is absolutely required for enhancer and promoter activity both in vitro and in vivo. A consensus binding site for the transcription factor YY1 contained within the TSSC element was identified, and immunological techniques were able to confirm that YY1 is a component of one of the TSSC complexes formed on the viral promoter. We demonstrated that viral promoter activity relies on maintenance of an intact YY1 binding site, since specific mutations within the YY1 core consensus sequence that prohibit YY1 binding activity also impair the transcriptional response of the RSV promoter in vitro and in vivo. Mutational analysis of the YY1 site in avian cells, which are naturally susceptible to infection by RSV, revealed that the YY1 site is required for full promoter and enhancer activity in this host cell. The YY1 protein positively regulates the RSV promoter, since overexpression of YY1 enhanced transcriptional activity. In addition to this, in vitro transcription assays using Drosophila embryo extracts devoid of YY1 activity failed to support wild-type levels of transcription. Although activity of the RSV enhancer is dependent on the intact TSSC sequence, the function of the enhancer region is not directly linked to the activity of YY1, since in HeLa cells, mutation of this site does not diminish enhancer function. Instead, the TSSC and the YY1 binding motif therein appear to contribute to basal promoter activity.

The absolute requirement for the TSSC demonstrated here (Fig. 2) indicates that TSSC is an essential core element of the RSV promoter. This is in contrast to other promoters whose transcription efficiency appears to be unaffected or only slightly reduced by deletion of sequences downstream of the TATA box (47). The TSSC likely plays a central role in the activation of RSV through cooperation with other transactivating factors. One possible hypothesis for TSSC regulation of RSV is that it functions similarly to an Inr. Inr elements have been identified in several cellular and viral promoters (69), including the AAV P5 promoter (30, 40, 65) and the COX Vβ promoter (3). The sequence requirements for an Inr are defined as Py Py A₊₁ N T/A Py Py (10, 30, 35). Comparison of the Inr consensus sequence to those of the RSV LTR TSSC reveals little homology. However, the consensus Inr sequence is not absolutely conserved for every defined Inr element, and like all previously defined Inr elements, the TSSC encompasses the sequence which contains the major transcription initiation site. Similar to the RSV TSSC, the human immunodeficiency virus type 1 (HIV-1) core promoter contains an element termed the SSR (start site region) that overlaps the transcription start site and strongly influences promoter strength but exhibits only partial sequence similarity to the consensus Inr sequence (79). The presence of a prototypical Inr element in the RSV and HIV-1 promoters, like those defined for the terminal deoxynucleotidyltransferase and adenovirus major late promoters, remains unclear.

In TATA-less promoters the Inr functions analogously to a TATA box, and in TATA-containing promoters the Inr can augment the strength of the TATA box. We have found that in the absence of the TSSC, the RSV TATA box is unable to support activated or basal levels of transcription, suggesting that the TSSC and the TATA box are codependent, since transcription is also severely decreased in the absence of the TATA box (reference 47 and unpublished data). In this regard, Jahavery et al. described several synthetic Inr elements whose activity is dependent on an A+T-rich sequence 30 nt upstream of the start site (30). In addition, HIV-1 SSR activity is dependent on the presence of at least a weak TATA box appropriately positioned upstream of the start site (79). The RSV TSSC and HIV-1 SSR may represent a different class or subset of Inr elements whose activity is required for transcription but dependent on the TATA box. The mechanism by which these transcription start site elements function either alone or in conjunction with a TATA box is not known, nor is it clear why some promoters contain both a TATA box and an Inr. The requirement for both elements within the viral promoter could serve to increase the proficiency at which the virus can integrate activating signals, to allow for effective viral transcriptional activity in a variety of cellular environments, to increase the recruitment of the general transcription machinery to its promoter, or to stabilize the PIC.

Several candidate proteins that recognize the transcription initiation site, including TFIID (5, 11, 32, 72), TFII-I (31, 46, 53–56), USF (16, 55), E2F (34), specific TAFs, RNA polymerase II (12, 51), and YY1 (65, 73), have been suggested. YY1 possesses the unique property of regulating transcription by functioning as an activator or a repressor of transcription, depending on the gene context. However, little is known about the mechanism(s) by which YY1 activates transcription. YY1 contains a bipartite transactivation domain composed of two acidic regions at the N terminus and two domains (DNA binding and Gly/Ala rich) important for protein-protein interactions (1). YY1 is known to associate with several factors, many of which are components of the basal transcription machinery (1, 45, 69). Presumably, it is these interactions that modulate much of the activity of YY1. In our analysis of the RSV LTR in vivo, we observed a consistent twofold decrease in promoter strength upon mutation of the YY1 binding site. This decrease in activity is due to the loss of YY1 binding, and not a fortuitous mutation in the putative Inr sequence, since the mutation of the 5′YY1 site produces a closer match to the consensus Inr sequence. A YY1 binding site coincident with an Inr was first described for the AAV p5 (65) and COX Vβ (3) promoters. The YY1 binding site in each of these promoters flanks the +1 site, and transactivation of the Inr sequence in the COX Vβ promoter was dependent on YY1 binding (3). By mutational analysis of the AAV P5 promoter, Lo and Smale showed that mutation of the YY1 consensus sequence resulted in a twofold reduction in promoter strength (40), consistent with our observations in the mutational analysis of the RSV LTR. We also observed that overexpression of the YY1 protein produced a dose-dependent response on both the enhancer-containing and enhancerless RSV LTR reporter constructs, with maximal stimulation resulting in a six- to sevenfold increase in transcriptional activity. The result of YY1 overexpression was more profound than the effect observed by simply prohibiting binding of YY1 to its cognate site. Mutation of the YY1 binding site would prohibit only those interactions dependent on DNA binding; however, recruitment of YY1 to the promoter via protein-protein interactions between YY1 and other factors would be maintained. The increased effect of overexpression of YY1 may simply reflect its ability to physically interact with components of the basal transcription machinery to effect viral promoter activity. Indeed, YY1 has been shown to be a component of the RNA polymerase II holoenzyme. Taken together, these data suggest an auxiliary function for YY1 in Inr activity, perhaps to aid in specifying the precise position of the initiation site, rather than a critical role for YY1 participating in the recruitment of RNA polymerase II as suggested by earlier experiments (73).

Viral promoters are often used experimentally to investigate the mechanisms of transcriptional activation. For this reason, a clear understanding of viral promoter structure is of great interest. The TSSC element located within the RSV LTR is a key regulator of viral enhancer and promoter activity. Although experimental evidence for functional Inr elements is present for some viral promoters (41), it is uncertain whether the TSSC functions as a bona fide Inr or represents a subset of transcription start site elements that contribute to basal transcriptional activity in a different manner. The transcriptional activity of the TSSC can most likely be attributed to the binding of YY1 and other sequence-specific factors [TSSC(A) and TSSC(B)], since YY1 is not singularly responsible for the effect of TSSC function, although it certainly contributes to promoter activity. The mechanism by which the YY1 protein exerts its action is as yet undetermined. However, the finding that YY1 interacts functionally with the RSV LTR may lead to a clearer understanding of RSV enhancer function, since YY1 has been shown to interact with proteins that bind to Y-box sequences present in the LTR enhancer (39, 57, 58). In addition to YY1, TFII-I, which binds to the terminal deoxynucleotidyltransferase and adenovirus major promoter Inrs (55), also interacts with the TSSC element (47a). Experiments to determine the functional significance of this binding are currently under investigation. Interestingly, TFII-I has recently been shown to promote the formation of a stable SRF complex on its DNA-binding element through direct interaction with SRF (23). This is of potential significance since two SRF binding sites are present within the RSV LTR enhancer. Binding of YY1 and TFII-I near the transcription start site could influence the formation of a specific DNA topology that allows the enhancer to effectively communicate with the basal transcription machinery and increase the efficiency of mRNA synthesis. Future studies to evaluate the specific sequences within the TSSC that contribute to its function and to evaluate possible protein-protein interactions between enhancer-bound and basal factors are required to define the powerful transcriptional response of the RSV LTR.