INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Epstein-Barr virus (EBV) is a human herpesvirus that is estimated to infect up to 90% of the world's population (28, 29). EBV infection is associated with several human diseases, including infectious mononucleosis, nasopharyngeal carcinoma, Burkitt's lymphoma, and, in immunosuppressed patients, B-cell and T-cell lymphomas (28, 29). Thus, EBV is a serious pathogen and poses a significant threat to human health.

Like other herpesviruses, primary infection with EBV is followed by a persistent infection of the human host. Oropharyngeal epithelium is thought to be the primary site of EBV infection and replication and of viral spread (28-30, 51), while B cells are the major site of persistent latency (28-31). Eleven of EBV's approximately 100 viral genes are expressed during latency. These include ones encoding the EBV-encoded nuclear antigens (EBNAs 1 to 6), the latent membrane proteins (LMPs 1 and 2), two EBV-encoded small nuclear RNAs, and the BamHIA transcripts (28, 29). Expression of a subset of the latter genes is sufficient to immortalize B cells and to maintain steady-state levels of the viral genome (31). Treatment of certain latently infected B-cell populations with reagents such as phorbol esters (5, 17, 64), Ca²⁺ ionophores (15), sodium butyrate (26, 38), and serum factors (2) or cross-linking of surface anti-immunoglobulin (12, 21, 50, 53) leads to cellular differentiation and the concomitant induction of the rest of EBV's genes, followed by viral genome replication to higher copy number and production of infectious virus particles (28-30). Two immediate-early genes, BZLF1 and BRLF1, are the first to be expressed during induction of EBV out of latency (11, 23, 32, 52). The protein products of both of these genes are strong transcriptional transactivators (9, 16). The product of the BZLF1 gene, referred to as Zta, ZEBRA, or EB1, plays a crucial role in the disruption of latency and initiation of the viral infectious cycle (10, 11, 41, 46, 52). Thus, regulation of Zta expression is critical to the state of EBV in cells.

The transcriptional regulation of the BZLF1 promoter (also referred to as Zp) has been studied extensively. Zp exhibits very low basal activity and is readily activated by inducers of the viral infectious cycle. The cis-acting elements necessary both for basal activity and for response to exogenous inducers lie within the nucleotide (nt) 221 to +12 region of the promoter relative to the transcriptional initiation site (12, 17, 18). These elements have been divided into three classes (Fig. 1; also, see reference 36 and references therein). Four AT-rich elements, termed ZIA to ZID, are dispersed throughout the promoter. They can bind the transcription factors Sp1 and Sp3 (34) and myocyte enhancer factor 2D (37, 44). A second type of element, ZII, shares significant homology with the consensus CRE/AP-1 binding site (1, 13, 25, 54). Finally, a region called ZIII contains multiple binding sites for the Zta protein itself (18). It has been proposed that activation of the BZLF1 gene occurs via a two-step process involving induction by exogenous factors mediated through the ZI and ZII domains, followed by autoactivation by Zta binding to the ZIII elements (18).

View larger version (24K): [in this window] [in a new window]   FIG. 1.   BZLF1 promoter. (A) Schematic representation of the cis-acting elements and their binding factors present within the 221 to +20 region of the BZLF1 promoter. The rectangles denote the approximate locations of the cis-acting sites. Binding factors are indicated above the sites. ZV, previously unknown cis-acting element identified here along with its binding factor, ZVR. Solid bars, cis-acting sites of previously identified negative regulatory elements for which trans-acting factors are not yet known. Numbering is relative to the transcriptional initiation site at +1. (B) Nucleotide sequence of the 30 to +20 region of the BZLF1 promoter used as the probe to identify the regulatory element ZV. The ZV site sequence is boxed.

During the latent state of infection, expression of the BZLF1 gene remains quiescent, suggesting the presence of silencing elements within the promoter. Recently, Liu et al. (36) identified a negative, cis-acting element located between nt 77 and 70, immediately upstream of the consensus CRE/AP-1 binding site. Mutations within this region, termed ZIIR, relieve repression of both basal and activated transcription. Unfortunately, the authors failed to detect a specific protein complex that recognized this sequence. Recently, Zhang et al. (60) showed that the ubiquitous factor Sµbp-2 represses transcription of the BZLF1 promoter, with this repression being significantly affected by an element located between nt 93 and 79. Whether this or the ZIIR element is an Sµbp-2 DNA-binding site remains unclear.

In earlier reports, Montalvo et al. (42) identified a negative, cis-acting region they called ZIV, between nt 551 and 386 relative to the transcriptional start site of the BZLF1 promoter. They went on to show that the transcription factor YY1 recognized a sequence within nt 433 to 386 (43). Similarly, Schwarzmann et al. (49) identified a silencing element, termed HI, repeated throughout the BZLF1 promoter region five times. This element contains the consensus sequence 5'- ACAGA(T/G)G(A/G)-3'. Four of the five HI elements are located between nt 551 and 227. The fifth HI element, termed HI, is located between nt 60 and 53.

We report here the identification of a previously unknown regulatory element within the nt 17 to 12 region of the BZLF1 promoter that plays a significant role in maintaining basal activity at low levels. In keeping with previous nomenclature identifying cis-acting regions of the BZLF1 promoter, we named this element ZV. We also show that ZV sequence-specifically binds a cellular factor(s) we call ZVR for "ZV regulator." ZVR activity is present in several cell lines, including EBV-positive and -negative B-lymphocytic cell lines. ZVR fails to recognize previously identified negative regulatory elements in the BZLF1 promoter. Thus, ZVR is a novel regulator of the BZLF1 promoter. It probably plays a significant role in regulation of the life cycle of EBV as well.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cells. All B-lymphocytic cell lines were grown in 100-mm tissue culture dishes and maintained at 37°C in a 5% CO₂ atmosphere. The EBV-negative Burkitt's lymphoma cell line DG75 and the EBV-positive B lymphocytic cell lines 721 (27) and GG68, a clone of EBV strain P3/HR1-infected B lymphocytes (55), were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum and 100 U of penicillin and streptomycin per ml. The human papillomavirus (HPV)-negative human cervical cell line C33A and the monkey kidney cell line CV-1 were grown in Dulbecco's modified Eagle's medium supplemented with 10 or 5% fetal bovine serum, respectively, and 100 U of penicillin and streptomycin per ml. The human breast cancer cell line MCF-7 was grown in RPMI 1640 medium supplemented with 10% fetal bovine serum, 6 ng of insulin per ml, 3 µg of glutamine per ml, and 100 U of penicillin and streptomycin per ml.

Plasmids. Plasmid DNAs were constructed by standard recombinant DNA techniques (48). Plasmid 221ZpCAT (reference 36 and references therein), a generous gift from Sam Speck, contains the nt 221 to +12 region relative to the transcription initiation site of the BZLF1 promoter driving the expression of the chloramphenicol acetyltransferase gene. We transferred this promoter sequence and variants of it into luciferase reporter plasmids by insertion of appropriate PCR-generated fragments into the KpnI and HindIII restriction sites of the pGL3 basic luciferase vector (Promega Corp., Madison, Wis.). The PCR fragments were obtained using 221ZpCAT as a template and the following oligonucleotides as primers. The forward primer, 5'-GAGGTACCCCATGCATATTTCAACTGGGCTGTCTATTTTTGACACCAGCTT-3', annealed to nt 221 to 178 of the BZLF1 promoter. The primers containing the wild-type sequence or mutations (underlined) within or adjacent to the ZVR DNA-binding site annealed to the complementary strand corresponding to the +10 to 40 region of the BZLF1 promoter. They were 5'- GTGTAAGCTTGCAAGGTGCAATGTTTAGTGAGTTACCTGTCTAACATCTCCC-3' for WTZpLUC, 5'-GTGTA AGCTTGCAAGGTGCAATGTTTAGTGAGTTAgCTGTCTAACATCTCC C-3' for 23CZpLUC, 5'-GTGTAAGCTTGCAAGGTGCAATGTTTAGTGAGTTAgCTGTCTAAgATCTCCC-3' for 23C/14CZpLUC, 5'- GTGTAAGCTTGCAAGGTGCAATGTTTAGTGAGTTACCTGTCctgCATCTCCC-3' for 22/20CAGZpLUC, 5'-GTGTAAGCTTGCAAGGTGCAATGTTTAGTGAGTTACCTactTAACATCTCCC-3' for 19/17AGTZpLUC, 5'- GTGTAAG CTTGCAAGGTGCAATGTTTAGTagaTTACCTGTCTAACATCTCCC-3' for 10/8TCTZpLUC, and5'-GTGTAAGCTTGCAAGGTGCAATGTTTAGTGAGTgACCTGTCTAACATCTCCC-3' for 12CZpLUC. The sequences of the promoter regions of all of the luciferase reporter plasmids were confirmed by DNA sequence analysis. All oligonucleotides were purchased from Integrated DNA Technologies Inc. (Coralville, Iowa).

Nuclear extracts. Nuclear extracts were prepared essentially as described by Dignam et al. (14) as modified by Zuo (63). B-lymphocytic cells (5 × 10⁸), grown to a density of approximately 1 × 10⁶/ml, were harvested by centrifugation at 2,000 × g for 10 min at 4°C. The cells were washed twice with cold phosphate-buffered saline (PBS), and the packed-cell volume (PCV) was determined. The cells were resuspended in 2 PCVs of buffer A (10 mM HEPES [pH 7.9], 1.5 mM MgCl₂, 10 mM KCl, 1 mM dithiothreitol [DTT]), incubated for 10 min on ice, and lysed by 10 strokes in a Dounce homogenizer using a B pestle. The nuclei were recovered by centrifugation at 17,000 × g for 30 min at 4°C and resuspended in 3 ml of buffer C (20 mM HEPES [pH 7.9], 1.5 mM MgCl₂, 0.42 M NaCl, 0.2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride [PMSF], 1 mM DTT, 25% glycerol) per 10⁹ cells. The resuspended nuclei were extracted by 5 strokes in a Dounce homogenizer using a B pestle. Extraction was continued for an additional 30 min at 4°C. The nuclear debris was removed by centrifugation at 17,000 × g for 30 min at 4°C. The supernatant containing the nuclear extract was dialyzed against 50 volumes of buffer D (20 mM HEPES [pH 7.9], 6 mM MgCl₂, 100 mM KCl, 0.2 mM EDTA, 1 mM PMSF, 1 mM DTT, 20% glycerol) overnight at 4°C. Aliquots of nuclear extract were stored at 70°C until use.

EMSAs. Electrophoretic mobility shift assays (EMSAs) were performed as follows. The probes consisted of gel-purified, double-stranded synthetic oligonucleotides that had been 5'-end labeled with T4 polynucleotide kinase and 50 µCi of [-³²P]ATP. The binding reaction mixtures typically contained 2 to 12 µg of nuclear extract incubated in 20 M HEPES (pH 7.9)-0.1 M KCl-6 mM MgCl₂-4 µg of poly(dI-dC)·(dI-dC)-0.5 mM PMSF-0.5 mM DTT-8% Ficoll. Following incubation for 20 min at 4^oC, 0.5 to 1.0 ng (25,000 to 50,000 cpm) of the desired probe was added and the mixture was incubated at 25°C for 15 min. For those experiments in which competition EMSAs were performed, the desired unlabeled competitor oligonucleotides were added to the reaction mixture, which was incubated for 20 min at 4°C, after which the desired probe was added and the incubation was continued. The protein-DNA complexes were separated from the free probe by electrophoresis at 200 V for 2 h at 4°C in a nondenaturing 4% polyacrylamide gel with 0.5× Tris-borate-EDTA as the running buffer. The gels were dried and exposed to X-ray films.

Transient transfections and luciferase assays. DG75 cells were transfected by the DEAE-dextrandimethyl sulfoxide (DMSO) shock method described by Liu et al. (36) with the minor modification that the 20% DMSO stock solution was prepared in RPMI 1640 medium. Transfected cells were harvested 72 h posttransfection, washed twice with PBS, and suspended in 300 µl of luciferase assay cell lysis buffer provided by the manufacturer (Promega Corp., Madison, Wis.). The suspension was cleared of debris by microcentrifugation. Luciferase activity present in the cell extracts was assayed according to the manufacturer's protocol (Promega Corp.). Luciferase activities were normalized to protein concentrations determined by a modified Bradford assay that utilizes a protein assay dye (Bio-Rad Laboratories, Hercules, Calif.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Identification of a novel sequence-specific binding site in the BZLF1 promoter. To look for putative binding sites for cellular factors that might regulate the BZLF1 promoter, we performed EMSAs using nuclear extracts prepared from DG75 cells and radiolabeled double-stranded oligonucleotides corresponding to various regions of the BZLF1 promoter as probes. We observed the binding of a factor(s) to the 30 to +20 region of the BZLF1 promoter (Fig. 2A, lane 2), a region not previously reported to bind regulatory factors. Data from competition EMSAs indicated that the cellular factor(s) present in the DG75 nuclear extract bound sequence specifically to this region of the BZLF1 promoter (Fig. 2A, lanes 3 to 5 versus lanes 6 to 8). Four regulatory regions of the BZLF1 promoter have been defined previously. Therefore, we chose to term this new element ZV.

View larger version (56K): [in this window] [in a new window]   FIG. 2.   Identification of a novel binding site within the BZLF1 promoter. (A) A cellular factor(s) binds sequence specifically to the 30 to +20 region of the BZLF1 promoter. Six micrograms of protein from a nuclear extract prepared from DG75 cells was incubated with the indicated amounts of unlabeled double-stranded DNA corresponding to the 30 to +20 (lanes 3 to 5) and 165 to 115 (lanes 6 to 8) regions of the BZLF1 promoter as competitors, respectively, and then with a radiolabeled double-stranded oligonucleotide corresponding to the 30 to +20 region of the BZLF1 promoter as probe. The protein-DNA complexes were separated from free probe by electrophoresis in a native 4% polyacrylamide gel. The location of the sequence-specific binding complex (ZVR) identified here is indicated. (B) ZVR binding localizes to the 25 to 11 region of the BZLF1 promoter. Competition EMSAs were performed as described for panel A. Unlabeled double-stranded DNAs corresponding to the 30 to +20 (lanes 3 to 5) and 165 to 115 (lanes 6 to 8) regions of the BZLF1 promoter and unlabeled double-stranded oligonucleotides containing the sequences 5'-ATGTTAGACAGGTAACTCACTAAACATTGCC-3' (25/+5, lanes 9 to 11) and 5'- CTCACTAAACATTGCACCTTGCCGGCCACC-3' (10/+20, lanes 12 to 14) were used as competitors.

View larger version (72K): [in this window] [in a new window]   FIG. 3.   The cellular factor ZVR recognizes bases within the 19 to 11 region of the BZLF1 promoter. (A) Competition EMSAs performed by incubation of 6 µg of protein obtained from a DG75 nuclear extract with the indicated amounts of the 25-bp oligonucleotides shown in panel B as competitors and approximately 1 ng of the 30 to +20 radiolabeled probe. The protein-DNA complexes were separated from free probe by electrophoresis in a native 4% polyacrylamide gel. No comp, no competitor. (B) Nucleotide sequences of the oligonucleotides used as competitors in the experiment in panel A and binding affinities for ZVR as determined from these data.

View larger version (74K): [in this window] [in a new window]   FIG. 4.   Identification of specific bases within the BZLF1 promoter necessary for binding of ZVR. (A) Competition EMSAs performed by incubation of 6 µg of protein obtained from a DG75 nuclear extract with the indicated amounts of the 25-bp competitor oligonucleotides shown in panel B and approximately 1 ng of the 30 to +20 radiolabeled probe. The protein-DNA complexes were separated from free probe by electrophoresis in a native 4% polyacrylamide gel. No comp, no competitor. (B) Nucleotide sequences of the oligonucleotides used as competitors and binding affinities for ZVR as determined from the data shown in panel A.

Mutations within ZV relieve transcriptional repression of the BZLF1 promoter. Having identified a previously unknown sequence-specific binding site in the BZLF1 promoter, we wished to determine the role this element plays in regulating expression of the BZLF1 promoter. To achieve this end, we cloned the wild-type and mutant versions of the nt 221 to +10 region of the BZLF1 promoter into a luciferase reporter plasmid. DG75 cells were transiently transfected in parallel with these plasmids and incubated at 37°C for 72 h in the presence or absence of TPA plus ionomycin as inducers. Mutations in sequences flanking ZV that do not affect protein binding had little or no effect on the transcriptional activity of the BZLF1 promoter in either the uninduced or the induced state (Fig. 5, mutants 23C, 22/20CAG, and 10/8TCT versus the wild type). On the other hand, mutations that abrogate protein binding increased uninduced transcriptional activity 2- to 4-fold and induced activity 40- to 180-fold (Fig. 5, mutants 19/17 AGT, 23C/14C, and 12C versus the wild type). Interestingly, the mutations that partially relieve repression of the BZLF1 promoter permit superactivation of transcription by the inducers: whereas the wild-type promoter was activated 6- to 7-fold by the inducers, the mutant promoters were activated 20- to 45-fold above their uninduced levels. Thus, we conclude that the 17 to 12 region of the BZLF1 promoter sequence-specifically binds a factor(s) that can function as a potent repressor of the BZLF1 promoter when inducers are not present. We named this factor ZVR, for "ZV regulator."

View larger version (20K): [in this window] [in a new window]   FIG. 5.   Correlation between binding of ZVR and repression from the BZLF1 promoter. (A) Structure of the wild-type reporter plasmid WTZpLUC constructs. The cloning vector was the pGL3 basic vector, whose sequences are not shown. Arrow, transcription initiation site. The location of the ZV element is indicated, as are sequences encoding luciferase (LUC). Nucleotides are numbered relative to the transcription initiation site. (B) Effects of ZV mutations on transcription in DG-75 cells. Luciferase reporter plasmids (2 µg/100-mm dish) containing the wild-type sequence or the indicated mutations in the BZLF1 promoter were transfected in parallel into DG75 cells, incubated for 72 h, and then harvested. Luciferase activities were determined and normalized to the protein concentration of each extract. The data are presented relative to the activity observed for the wild-type promoter not treated with inducers (rel. to WT). They are the means with standard errors of the means for three sets of transfections performed on different days. (C) Effects of ZV mutations on transcription in the presence of inducers. Cells were treated in parallel with the ones in panel B, except for incubation after transfection with TPA (20 ng/ml) plus ionomycin (1 µM) until harvesting.

Repression of the BZLF1 promoter through the ZV element requires the presence of ZVR DNA-binding activity. If the ZV element does, indeed, function by binding ZVR, mutations in ZV may have little effect on transcriptional activity in cell lines lacking ZVR activity. Thus, we tested several additional cell lines for ZVR activity: the HPV-negative human cervical cell line C33A, the human breast cancer cell line MCF-7, and the monkey kidney cell line CV-1. All three are epithelium derived. Nuclear extracts were prepared from each of these cell lines and assayed for ZVR DNA-binding activity as described above. The C33A and CV-1 cell lines were found to contain as much ZVR DNA-binding activity as do DG75 cells, if not slightly more (Fig. 6, lanes 5 to 7 and 8 to 10, respectively, versus lanes 2 to 4). MCF-7 cells appeared to lack detectable ZVR DNA-binding activity (Fig. 6, lanes 11 to 13), at least with the electrophoretic mobility observed with the other cell lines. Therefore, MCF-7 cells likely represent a cell line one can use to examine regulation of the BZLF1 promoter in the absence of significant levels of ZVR.

View larger version (104K): [in this window] [in a new window]   FIG. 6.   ZVR DNA-binding activity is present in C33A and CV-1 cells but not MCF-7 cells. The indicated amounts of protein from nuclear extracts obtained from DG75 cells (lanes 2 to 4), C33A cells (lanes 5 to 7), CV-1 cells (lanes 8 to 10), and MCF-7 cells (lanes 11 to 13) were incubated with approximately 1 ng of the 30 to +20 radiolabeled probe. The DNA-protein complexes were separated from free probe by electrophoresis in a native 4% polyacrylamide gel.

View larger version (17K): [in this window] [in a new window]   FIG. 7.   The presence of ZVR correlates with regulation of expression of the BZLF1 promoter via the ZV element. Luciferase reporter plasmids (2 µg/100-mm dish) containing the wild-type sequence or the 23C/14C mutation in the BZLF1 promoter were transfected in parallel into C33A (A), CV-1 (B), and MCF-7 (C) cells. Cells were incubated for 48 h with or without TPA (20 ng/ml) plus ionomycin (1 µM) and then harvested. Luciferase activities (act.) were determined and normalized to the protein concentration of each extract. The data are presented relative to the activity observed for the wild-type promoter not treated with inducers (rel. to WT). They are the means with standard errors of the means for three experiments performed on different days.

ZVR does not bind previously identified repressor elements within the BZLF1 promoter. Three other negative cis-acting elements have previously been identified within the nt 221 to +20 region of the BZLF1 promoter. No sequence-specific factors have been identified to date that bind these three negative elements. One, called ZIIR, is centered around nt 70 (36). Another, HI, is located between nt 59 and 52 (49). The third region maps to nt 93 to 79 (60). To examine whether ZVR recognizes any of these sequences, we performed EMSAs utilizing as competitors unlabeled double-stranded oligonucleotides containing the sequences of the ZIIR, HI, and 93/79 elements (Fig. 8B). None of these oligonucleotides competed effectively for ZVR DNA binding (Fig. 8A). Thus, the factor(s) that binds to the 17 to 12 region of the BZLF1 promoter is distinct from the ones yet to be identified that interact with the previously known negative regulatory elements of the BZLF1 promoter.

View larger version (74K): [in this window] [in a new window]   FIG. 8.   ZVR is a previously unidentified DNA-binding factor. (A) Competition EMSAs performed by incubation of 6 µg of nuclear extract obtained from DG75 cells with the radiolabeled 30 to +20 region oligonucleotide as probe and the indicated amounts of unlabeled 30 to +20 wild-type DNA (lanes 3 to 5), 23/14 mutant DNA (lanes 6 to 8), or the double-stranded oligonucleotides shown in panel B containing the ZIIR (lanes 9 to 11), the HI (lanes 12 to 14), or the 93/79 (lanes 15 to 17) element as a competitor. (B) Sequences of the oligonucleotides used as competitors in the experiment shown in panel A. Bases shown by others to play roles in repression of the BZLF1 promoter are boxed.

Effect of EBV latent products on ZVR DNA-binding activity. To determine the effects of EBV latent proteins on the DNA-binding activity of ZVR, we performed EMSAs with nuclear extracts obtained from two EBV-positive B-lymphocytic cell lines, GG68 and 721. ZVR DNA-binding activity was readily observed in both of these cell lines (Fig. 9). Competition EMSAs with wild-type and mutant oligonucleotides as competitors confirmed that these DNA-protein complexes contained ZVR (data not shown). Thus, the amount of DNA-binding activity present in B-lymphocytic cells is probably not appreciably affected by the presence of the EBV latent gene products responsible for cellular immortalization. However, the effects EBV latent gene products have on other activities of ZVR, e.g., transcriptional activities, remain unknown.

View larger version (80K): [in this window] [in a new window]   FIG. 9.   Presence of EBV's products expressed during latency does not affect the DNA-binding activity of ZVR. Nuclear extracts were prepared from the EBV-negative cell line DG75 (lanes 2 to 4) and the EBV-positive cell lines 721 (lanes 5 to 7) and GG68 (lanes 8 to 10). The indicated amounts of protein from these extracts were incubated with approximately 1 ng of the 30 to +20 region radiolabeled probe. The DNA-protein complexes were separated from free probe by electrophoresis in a native 4% polyacrylamide gel.

Treatment with TPA plus ionomycin enhances the DNA-binding activity of ZVR. The BZLF1 promoter contains a recognizable CRE/AP-1 motif previously shown to be essential for transcriptional activation by phorbol esters (5, 17). Treatment with inducing agents enhances the binding activity of the trans-acting factors that recognize these sequences (33). Likewise, incubation of cells with TPA has been shown to relieve binding activities of transcriptional repressors that recognize elements lying 300 to 400 bp upstream of the transcriptional start site of the BZLF1 promoter (43, 49). To determine whether inducers also affect ZVR DNA-binding activity in ways that might directly contribute toward activation of the BZLF1 promoter, we prepared nuclear extracts from DG75 cells treated with TPA plus ionomycin for 72 h prior to harvesting the cells. In contrast to the response observed with previously identified transcriptional repressors of the BZLF1 promoter, treatment with TPA plus ionomycin led to a threefold enhancement of ZVR DNA-binding activity (Fig. 10A, lanes 2 to 4 versus lanes 5 to 7). Competition EMSAs confirmed that this binding activity was, indeed, ZVR (Fig. 10B). Thus, we conclude that induction of the BZLF1 promoter by TPA plus ionomycin does not occur in part via inactivation of binding of ZVR to the BZLF1 promoter. Other models by which it might act are discussed below.

View larger version (85K): [in this window] [in a new window]   FIG. 10.   ZVR DNA-binding activity is increased by treatment with inducers of the BZLF1 promoter. (A) EMSAs performed with the indicated amounts of protein obtained from a nuclear extract prepared from DG75 cells either untreated (lanes 2 to 4) () or treated (lanes 5 to 7) (+) with TPA (20 ng/ml) plus ionomycin (1 µM) for 72 h prior to harvesting. Each reaction mixture also contained approximately 1 ng of radiolabeled probe corresponding to the 30 to +20 region of the BZLF1 promoter. The protein-DNA complexes were separated from free probe by electrophoresis in a native 4% polyacrylamide gel. The position of the ZVR-DNA complex is indicated. (B) Competition EMSAs performed with nuclear extract prepared from DG75 cells treated with TPA (20 ng/ml) plus ionomycin (1 µM). The reactions were performed with 6 µg of protein, the indicated amounts of the unlabeled wild-type (lanes 3 to 5) or mutant (lanes 6 to 8) double-stranded oligonucleotides shown in Fig. 3B as competitors, and the radiolabeled probe corresponding to the 30 to +20 region of the BZLF1 promoter. No comp, no competitor.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We report here the identification within the BZLF1 promoter of a previously unknown regulatory element called ZV and the cellular factor(s), named ZVR, which sequence-specifically binds to it. The ZV element maps to nt 17 to 12 of the BZLF1 promoter (Fig. 3 and 4), although flanking bases may contribute to its specificity. We demonstrated a correlation between binding of ZVR to the ZV element and regulation of the BZLF1 promoter both with mutations in the ZV element (Fig. 5) and with cell lines lacking ZVR DNA-binding activity (Fig. 6 and 7). We further showed that ZVR does not bind to previously identified negative regulatory elements of the BZLF1 promoter (Fig. 8). Thus, we conclude that ZVR is a previously unknown regulator of the BZLF1 promoter. Finally, we found that ZVR DNA-binding activity remains abundant in lymphocytes that had been immortalized by EBV and express EBV's latent gene products (Fig. 9), and it is not inactivated by treatment of cells with the inducers TPA and ionomycin (Fig. 10).

ZV element. Using EMSAs and nuclear extracts from DG75 cells as a protein source, we identified a novel factor-binding site located at nt 17 to 12 of the BZLF1 promoter (Fig. 2 to 4). Our data showed that these 6 bp are required for binding activity. Mutations outside this region did not affect binding. However, we have not ruled out the possibility that additional bases also contribute to binding specificity.

Role of ZVR in regulation of the BZLF1 promoter. Our experiments demonstrated a correlation between binding of ZVR to the ZV element and repression of the BZLF1 promoter. First, we found that promoters containing mutations in the ZV element that abrogated ZVR binding exhibited higher levels of both basal and induced transcription than the wild-type promoter in ZVR-positive cells (Fig. 3 to 7). Second, statistically significant differences were not observed between the wild-type and mutant promoters in the cell line MCF-7, which lacks ZVR DNA-binding activity (Fig. 7C). Thus, the ZV element is a regulatory element of the BZLF1 promoter that functions as a transcriptional silencing element in the absence of inducers.

View larger version (33K): [in this window] [in a new window]   FIG. 11.   Model for regulation of the BZLF1 promoter by ZVR. See the text for details. Relative levels of transcription are indicated by the heights and thicknesses of the arrows. Rectangles, ZV elements; X, mutations within the ZV element; R, cellular repressors inactivated by inducers such as TPA; A, sequence-specific positive factors activated by inducers in B lymphocytes (triangles) or positive factors present in epithelial cells (rhomboids).

Effect of EBV-encoded factors on ZVR activities. Negative regulatory elements of the BZLF1 promoter lie both within (36, 43, 60) and outside (42, 43, 49) the 221 to + 20 region. We showed here that ZVR does not bind any of the known proximal negative elements (Fig. 8). Specific factors that recognize the distal negative elements have been found only in nuclear extracts obtained from cells latently infected with EBV (49). We found that ZVR is abundant not only in B lymphocytes that express EBV's latent gene products (Fig. 9) but also in B lymphocytes that do not (Fig. 2 to 4). Thus, ZVR DNA-binding activity is independent of EBV status. Therefore, we conclude that ZVR is, indeed, distinct from any of the previously identified repressors of the BZLF1 promoter.

Effect of inducers on ZVR activities. We found that treatment of cells with the inducers TPA and ionomycin did not eliminate the DNA-binding activity of ZVR; if anything, it enhanced it (Fig. 10). Likewise, Grove and Mastro (24) found enhanced binding activity in extracts obtained from TPA-treated bovine lymphocytes of a transcription factor that interacts with the negative regulatory element (NRE-A) of the interleukin-2 (IL-2) promoter. Since the sequences of these two elements are identical, it is quite likely that we have identified a similar binding activity. Yet to be determined is the effect of inducers on ZVR's transcriptional activities. Possibly, inducers change ZVR from a repressor to an activator by modifying ZVR directly or altering factors with which it interacts, e.g., corepressor and coactivator complexes, some of which may be lymphoid specific. The inducers may also be directly or indirectly increasing the specific DNA-binding activity or half-life of ZVR. Alternatively, although less likely, the inducers may lead indirectly to another, similar-mobility factor binding to the ZV site. On the other hand, treatment of EBV-positive cells with TPA relieves binding to the distal HI and YY1 elements (43, 49). Thus, relief of repression and activation of transcription represent non-mutually exclusive mechanisms by which TPA may induce the transcriptional activity of the BZLF1 promoter.

Possible identity of ZVR. We showed here that ZVR recognizes the sequence 5'-CAGGTA-3'. This sequence is identical to the NRE-A within the IL-2 promoter (57). Originally, a T-cell-specific zinc finger binding protein, termed Nil-2-a, was identified as the transcription factor that mediates its effects through NRE-A (57). Subsequently, a similar zinc finger/homeodomain protein, termed ZEB for "zinc finger E-box binding protein," was isolated from B cells (19, 20, 58). It is now known that Nil-2-a is a partial cDNA clone of ZEB. ZEB represents the full-length protein and is present in a number of cell types including T cells.

Role of ZVR in the life cycle of EBV. The fundamental question yet to be answered is the role played by ZVR in EBV latency and the induction out of latency. We hypothesize that ZVR repressor activity may predispose a cell toward establishing a latent state of infection. Target cells of EBV lacking ZVR repressor activity may overproduce Zta, producing infectious virus rather than entering a latent immortalized state. Generating strains of EBV containing mutations within the ZV element and cell lines with levels of ZVR that can be regulated experimentally should enable one to determine the role of this regulatory element and its trans-acting factors in the life cycle of EBV.