The BZLF1 promoter of Epstein-Barr virus is controlled by E box-/HI-motif-binding factors during virus latency

Abstract

The BZLF1 open reading frame of EpsteinBarr virus (EBV) encodes an important transactivator of replication. During latency, transcription of this gene is switched off. HI motifs have been shown to cause negative regulation of the promoter. Using yeast one-hybrid assays, we isolated the E box-binding protein, E2-2, interacting with these motifs. Electrophoretic mobility shift assays demonstrated that E2-2 binds to HIα, HIβ and HIγ, which contain E box consensus binding sites. Deletion of the HI-associated E boxes and overexpression of E2-2 in transfection assays revealed that these elements act as repressors in lymphoid cells. In contrast, in epithelial cells they contribute to the increased responsiveness of the promoter to transactivation by the BZLF1 protein. The data presented are in accord with an alternative and exclusive binding of different cell type- and differentiation-specific factors, such as E2-2, to the HI-associated E boxes in lymphoid and epithelial cells. This implies a role in cell type-specific virus replication.

The authors wish to thank Dr Hans Helmut Niller and Dr Nadja Prang for stimulating discussions. This work was supported by the Deutsche Forschungsgemeinschaft DFG (Wo227/6 and Schw566/7).

Footnotes

†Present address: MRC, Laboratory of Molecular Biology, Division of Protein and Nucleic Acid Chemistry, Cambridge, UK.

EpsteinBarr virus (EBV) establishes life-long persistence in humans. Replication of the virus depends on the type and state of differentiation of the infected cell. Whereas lymphoid cells are predominantly latently infected, differentiated epithelial cells are permissive for virus replication (Becker et al., 1991; Young et al., 1991). It is thought that a subpopulation of resting B lymphocytes serves as a latent reservoir of EBV (Tierney et al., 1994; Miyashita et al., 1995; Decker et al., 1996; Babcock et al., 1998). Transcription of the EBV open reading frame BZLF1, which leads to virus replication, is controlled by both down-regulation and transactivation (Flemington & Speck, 1990; Montalvo et al., 1991, 1995; Shimizu & Takada, 1994; Walling et al., 1994; Ruf & Rawlins, 1995; Wang et al., 1997; Kraus et al., 2001). Furthermore, a post-transcriptional mechanism interferes with processing of the primary transcript (Prang et al., 1995, 1999), and the activity of the gene product is regulated at the post-translational level, (Daibata et al., 1992; Katz et al., 1992; Kenney et al., 1992; Sinclair et al., 1992; Gutsch et al., 1993; Zhang et al., 1994). The objective of this work was to identify cellular proteins that interact with sequence motifs of the HI type in the promoter of BZLF1 (Schwarzmann et al., 1994) and thereby down-regulate transcriptional activity in latently infected lymphoid cells. Five copies of HI motifs are localized within 500 nucleotides upstream of the start site of transcription.

To identify HI binding factors we performed yeast one-hybrid assays (Matchmaker One Hybrid System; Clontech). As molecular baits for fishing and screening, we used the central regions of the four HI motifs HIα, HIβ, HIγ and HIδ (Schwarzmann et al., 1994). Five copies of each individual HI sequence were fused in a head-to-tail orientation (HIα, 5'-ACAGATGAACAGATGAACAGATGAACAGATGAACAGATGA-3'; HIβ, 5'-ACATATGGACATATGGACATATGGACATATGGACATATG-3'; HIγ, 5'- ACAGATGGACAGATGGACAGATGGACAGATGGACAGATGG-3'; HIδ, 5'-ACAGAGGAACAGAGGAACAGAGGAACAGAGGAACAGAGGA-3'). These sequences were cloned in both orientations into the reporter plasmids. One of the cDNA clones, which we isolated with an HIγ-specific probe, encoded the transcription factor E2-2 (Bain et al., 1993). E2-2 belongs to the family of E box-binding proteins. Computer analysis detected putative binding sites (5'-CANNTG-3') in the HI motifs HIα, HIβ and HIγ. No consensus sequence was found in HIδ. In vitro translation of the isolated E2-2 cDNA yielded a protein with the expected molecular mass (approximately 75 kDa; Fig. 1A). However, the recombinant protein did not bind to the published consensus sequences CT1 (Pscherer et al., 1996) and µE5 (Shen & Kadesch, 1995) (data not shown). This was thought to be due to incorrect folding or missing modifications, since a monoclonal antibody specific for E2-2 (BD Pharmingen) recognized the in vitro-translated product very inefficiently (Fig. 1B, lane 4). Therefore, further experiments were performed with E2-2-positive nuclear extracts. Native E2-2 was present in nuclear extracts of the lymphoid cell line BJAB (Fig. 1B, lane 1) (Ben Bassat et al., 1977) and in the neuroblastoma cell line NGP (Fig. 1B, lane 3) but was not detected in the teratocarcinoma cell line Tera 9117 (Fig. 1B, lane 2).

(51K):

Fig. 1. (A) In vitro translation of E2-2 in the presence of [³⁵S]methionine. The translation products were separated by PAGE and visualized by autoradiography. (B) Presence of E2-2 in different cell lines. Seventy-five µg of nuclear extract of the lymphoid cell line BJAB (lane 1), the teratocarcinoma line Tera 9117 (lane 2), the neuroblastoma cell line NGP (lane 3) and 10 µl of in vitro-translated E2-2 (lane 4) were separated by PAGE. Proteins were transferred to a nitrocellulose membrane by Western blotting and were analysed with a monoclonal antibody specific for E2-2 (BD Pharmingen).

To confirm that cellular proteins interact with the HI motifs, we performed electrophoretic mobility shift assays (EMSAs) (Fig. 2). DNA probes with nucleotide sequences identical to those used for fishing and screening in the yeast one-hybrid assay were used. Competitor oligonucleotides were added at up to 60-fold excess. In EMSAs with BJAB extract, a number of specific protein complexes (Fig. 2A, indicated by arrows) were formed with the four most distantly located HI sites. In lanes 2, 7, 12 and 17, the complexes are visible without competition. These HI-specific complexes were also formed in the presence of non-specific competitor oligonucleotides (Oct1: 5'-TGTCGAATGCAAATCACTAGAA-3', Fig. 2A, lanes 4, 9, 14 and 19; YY1: 5'-CGCTCCGCGGCCATCTTGGCGGCTGGT-3', Fig. 2A, lanes 5, 10, 15 and 20) but disappeared when the corresponding unlabelled probes were used for competition (Fig. 2A, lanes 3, 8, 13 and 18).

(85K):

Fig. 2. E box-binding factors specifically interact with the HI motifs. (A) The HI motifs bind several factors. Shown are EMSAs with 7·5 µg of nuclear extract of the EBV-negative BJAB cells. ProteinDNA-complexes were separated by non-denaturing PAGE and analysed by autoradiography. Different HI binding sites were used as radiolabelled probes: HIα (lanes 15), HIβ (lanes 610), HIγ (lanes 1115) and HIδ (lanes 1620). Negative controls without nuclear extracts are shown in lanes 1, 6, 11 and 16, and results with no competitors in lanes 2, 7, 12 and 17. Non-labelled oligonucleotides were added as competitors in 60-fold excess compared with the labelled probes (lanes 35, 810, 1315 and 1820). The type of competitor (HIγ, Oct-1 or YY1) is indicated above the lanes. (B) Transcription factors interact with the HI-associated E boxes. An HIγ-specific oligonucleotide probe was used in all lanes and complex formation was competed with unlabelled oligonucleotides containing binding sites for HIγ (lane 3), CT1 (lanes 4 and 9), µE5 (lanes 5 and 10), µE5* (lane 6), ATF (lane 7), HIα* (lane 11), HIβ* (lane 12) and HIγ* (lane 13). Negative control without protein is shown in lane 1 and positive controls without competitor in lanes 2 and 8. The E box-specific complexes are indicated by an arrow; other HI-specific complexes are marked with asterisks. (C) The HI-associated E boxes bind E2-2. The cell line H1299 was transfected with an expression plasmid for either E2-2 or pUC18. Crude nuclear extracts were prepared and used for gel retardation assays using the radiolabelled probes HIα (lanes 13), HIβ (lanes 46) and HIγ (lanes 79). Negative controls without nuclear extracts are shown in lanes 1, 4 and 7. E2-2-specific complexes that increased in intensity with extract from E2-2-transfected cells are indicated by arrows. In lanes 1023, the E2-2-specific complexes that formed with the E2-2 consensus sequence µE5 (labelled probe) were competed with a 20- and 40-fold excess of the non-specific competitor CT2 (lanes 13 and 14), the specific competitor CT1 (lanes 15 and 16) and the HI sequences HIα (lanes 17 and 18), HIβ (lanes 19 and 20) and HIγ (lanes 21 and 22). Lane 11 is a negative control with H1299 cells transfected with pUC18, lanes 12 and 23 are without competitor.

To identify those complexes containing E box-binding factors, we used the HIγ sequence as a probe and challenged with the published E2-2-specific binding sites CT1 and µE5 (Fig. 2B). It was clearly visible that both the published binding sites (Fig. 2B, lanes 4 and 5) and the unlabelled cold probe (Fig. 2B, lane 3) interfered with the formation of the slowest-migrating complex and thus demonstrated the specific binding of factors to the E boxes. In contrast, there was no competition with the mutated binding site µE5* (Fig. 2B, lane 6). The binding site for ATF (5'-AGAGATTGCCTGACGTCAGAGAGCTAG-3') was also unable to compete with the E box-specific complex (Fig. 2B, lane 7). However, formation of a faster complex was significantly reduced suggesting that ATF (or a related factor) also interacted with the HI motif HIγ.

To confirm that the slowest-migrating complex formed at HI-associated E boxes, the complex was challenged with binding sites HIα, HIβ, HIγ that had been mutated in their E boxes (HIα*: 5'-CCTCCAACATGCAACTTGCCTCG-3'; HIβ*: 5'-CTGTCCACAATACGCTGCTTCCTCC-3'; HIγ*: 5'-CCATATGTGGACACTACGACCTGAGC-3') (Fig. 2B). As shown before, the CT1 site and the µE5 site abolished formation of the complex (Fig. 2B, lane 9 and 10, respectively). In contrast, the mutated HI sequences did not interfere (Fig. 2B, lanes 1113) although they still competed with other HI-specific complexes indicated by asterisks. This demonstrated again that the E boxes localized within the HI motifs are specifically bound by proteins.

To demonstrate that the E2-2 protein interacts with the HI-associated E boxes, we performed supershift assays with an E2-2-specific monoclonal antibody. However, the antibody did not work sufficiently well in this type of assay (data not shown). In an alternative experimental approach, H1299 cells, which synthesize very low amounts of E2-2, were transiently transfected with an expression plasmid encoding E2-2 or pUC18 (as a negative control). By creating such an E2-2-positive/negative test system, we wanted to see whether we could unequivocally identify E2-2-specific complexes with HI motifs (Fig. 2C). Using the E2-2-positive extract with the HI motifs HIα, HIβ and HIγ (Fig. 2C, lanes 3, 6 and 9, respectively), E2-2-specific complexes were clearly visible (indicated by arrows). These did not form or were significantly less intense with control extract of H1299 cells transfected with pUC18 (Fig. 2C, lanes 2, 5 and 8). In another shift experiment, the E2-2-specific complex, which formed with the µE5 consensus probe (Fig. 2C, lanes 12 and 23), could be competed with increasing amounts of HIα, HIβ and HIγ oligonucleotides (Fig. 2C, lanes 1722), as well as with a positive control, CT1 (Fig. 2C, lanes 15 and 16). The EMSA with E2-2-positive/negative extracts clearly demonstrated that E2-2 does bind to the HI-associated E boxes.

Based on EMSA data, we finally performed functional transfection assays with BZLF1 promoter-derived reporter plasmids (Fig. 3). Transfection of the Burkitt's lymphoma cell line DG75 was carried out by electroporation using double pulses: 750 V, 2350 Ω, 25 µF and 100 V, 2310 Ω, 3000 µF (Gene Pulser; Bio-Rad). The nasopharyngeal carcinoma (NPC) epithelial line CNE-L was transfected by calcium phosphate precipitation. To construct reporter plasmids, the regulatory region of BZLF1 was isolated by PCR (primers: 5'-CGGCAAGGAGATCTGTTTAGTG-3' and 5'-GGATCCCTAACGGTACCCCCGG-3'), inserted into the reporter plasmid pGL2basic (Promega) and the E boxes were mutated by site-directed mutagenesis (QuikChange site-directed mutagenesis kit; Stratagene) using the following primers: HIα*: 5'-CCTCCAACATGCAACTTGCCTCG-3' and 5'-CGAGGCAAGTTGCATGTTGGAGG-3'; HIβ*: 5'-CTGTCCACAATACGCTGCTTCCTCC-3' and 5'-GGAGGAAGCAGCGTATTGTGGACAG-3'; HIγ*: 5'-CCATATGTGGACACTACGACCTGAGC-3' and 5'-GCTCAGGTCGTAGTGTCCACATATGG-3'. The reporter constructs were co-transfected with expression plasmids for BZLF1 protein (Zta) to measure the impact of E box sequences on Zta-mediated transactivation. In the lymphoid cell line (DG75), transactivation ranged from 66- to 290-fold (Fig. 3B, lanes 610). Mutation of the E boxes within the HI motifs HIα and HIγ increased transactivation 2·4-fold compared with the original promoter (Fig. 3B, lanes 7 and 9). A triple mutation in the three HI motif-associated E boxes showed a cumulative effect and increased transactivation 4·4-fold compared with the original promoter (Fig. 3B, lane 10), whereas mutation of the HI motif HIβ had no stimulating effect (Fig. 3B, lane 8). In the NPC epithelial line (CNE-L) transactivation rates were quite high (up to 3000-fold) (Fig. 3A, lane 6). The relatively strong transactivation of the original promoter in the epithelial line and the enhancement of transactivation in the lymphoid cell line after mutation of the E boxes were in accordance with the absence and presence, respectively, of E2-2 in both cell lines. (CNE-L tested negative with EMSA; data not shown.) In contrast to the lymphoid cell line, mutation of the HI-associated E boxes significantly reduced Zta-mediated transactivation in the epithelial line (Fig. 3A, lanes 710). This indicated that in the latter case the E boxes were not only free of inhibitory factors but were also involved in positive regulation increasing responsiveness to Zta.

(25K):

Fig. 3. The HI-associated E boxes are functional regulatory motifs in lymphoid and epithelial cells. (A) Nasopharyngeal carcinoma (NPC) epithelial cell line CNE-L and (B) Burkitt's lymphoma (BL) cell line DG75. The reporter plasmids with the original BZLF1 promoter (lanes 1, 6, 11 and 16), with mutations in the E box in HIα (lanes 2, 7, 12 and 17), with mutations in the E box in HIβ (lanes 3, 8, 13 and 18), with mutations in the E box in HIγ (lanes 4, 9, 14 and 19) and with mutations in all three E boxes (lanes 5, 10, 15 and 20) were co-transfected with expression plasmids for Zta/BZLF1 (lanes 610), E2-2 (lanes 1115) or Zta and E2-2 (lanes 1620). Promoter activities were calculated by measuring the concentration of the reporter protein luciferase. (C) E2-2 does not interfere with the expression of Zta from the co-transfected plasmid. DG75 cells were co-transfected with a constant amount (5 µg) of expression plasmid for Zta and increasing amounts of expression plasmids for E2-2 (0 µg, lanes 13; 5 µg, lanes 46; 10 µg, lanes 79 and 20 µg, lanes 1012). Cells (2x10⁵) from each transfection were analysed by PAGE and Western blotting with a Zta-specific polyclonal rabbit serum. In lane 13, 2x10⁵ EBV-negative and Zta-negative BJAB cells were applied as a negative control. In lanes 1418, a series of twofold dilutions of purified Zta protein were applied as a marker for quantification, starting with 5 ng protein in lane 14.

To demonstrate that E2-2 was responsible for the inhibitory effect on transactivation, both cell lines were simultaneously co-transfected with expression plasmids for E2-2 and Zta and with the reporter plasmids (Fig. 3A, B). Both in the lymphoid and in the epithelial cell line, overexpression of E2-2 reduced transactivation. This effect was strongest in the epithelial line CNE-L with the original promoter that contained all E boxes responding to E2-2 (approx. 60 % reduction; Fig. 3A, lane 16) and decreased when the E boxes were mutated (Fig. 3A, lanes 1720).

In the lymphoid cell line, the E2-2 effect was less significant with the original promoter due to the high background expression of E2-2 (Fig. 3B, lane 16). However, overexpression also significantly reduced transactivation (to 62 %) in the case of mutations in HIα (Fig. 3B, lane 17) and in HIγ (Fig. 3B, lane 19). In the case of the triple E box mutation, E2-2 reduced Zta-mediated transactivation to 43 % (Fig. 3B, lane 20). A Western blot control of a co-transfection experiment with a Zta-specific antibody ensured that the observed transcriptional repression was not an artefact due to the inhibition of expression of Zta from the transfected vector by E2-2 (Fig. 3C).

Since the triple mutants still responded to overexpressed E2-2, it seems likely that other E boxes outside the HI motifs confer additional repression. Indeed, three additional E boxes were localized outside the HI sequences at positions -520 to -515 (5'-CACTTG-3'), -344 to -339 (5'-CATTTG-3') and -208 to -203 (5'-CAACTG-3'). Recently, the E box nearest to the start site of transcription was reported to be part of a silencer element (Kraus et al., 2001).

Thus, EMSA experiments and transfection assays have demonstrated that E box-binding proteins, in particular E2-2, interact with the core region of the HI motifs HIα, HIβ and HIγ in the BZLF1 promoter. Mutations of the E boxes within the HI motifs HIα and HIγ released transcriptional repression in the lymphoid cell line, demonstrating their role as cell-type-specific repressors. A mutation of the E box in HIβ had no influence on transactivation. In contrast to the lymphoid cells, in the epithelial cell line mutation of all three HI-associated E boxes significantly reduced Zta-mediated transactivation, indicating a positive regulatory property in this type of cell.

These data suggest a model where the HI motifs HIα and HIγ are bi-functional elements. Both inhibitory and stimulatory factors can bind in an exclusive manner to the same HI-associated E boxes. Inhibitory E box-binding factors such as E2-2 are expressed in a cell-type-specific manner in lymphoid cells, since mutations of HI-associated E boxes led to increased transactivation. In this context, the role of the E box in the HIβ motif needs to be analysed further. In epithelial cells, such as NPC, the E boxes contribute to enhanced responsiveness to Zta-mediated transactivation by binding (epithelial) cell-type- and differentiation-dependent stimulatory factors.

An alternative model is that the E boxes are situated close to positive regulatory elements with which they interfere. The inhibitory effects result from sterical hindrance or binding of histone deacetylases.

Taking into account the fact that E2-2 and other E box-binding proteins are expressed in a differentiation-dependent manner in lymphoid cells (Church et al., 1985; Ephrussi et al., 1985; Bain et al., 1993; O'Riordan & Grosschedl, 1999) and that virus replication depends on cell differentiation, the HI motifs and the associated E boxes appear to contribute to the stabilization of latency and the control of virus replication in B lymphoid cells via binding of E box-binding proteins.

References

Babcock, G. J., Decker, L. L., Volk, M. & Thorley-Lawson, D. A. (1998). EBV persistence in memory B cells in vivo. Immunity 9, 395404.[CrossRef][Medline]

Bain, G., Gruenwald, S. & Murre, C. (1993). E2A and E2-2 are subunits of B-cell-specific E2-box DNA-binding proteins. Mol Cell Biol 13, 35223529.[Abstract/Free Full Text]

Becker, J., Leser, U., Marschall, M., Langford, A., Jilg, W., Gelderblom, H., Reichart, P. & Wolf, H. (1991). Expression of proteins encoded by EpsteinBarr virus trans-activator genes depends on the differentiation of epithelial cells in oral hairy leukoplakia. Proc Natl Acad Sci U S A 88, 83328336.[Abstract/Free Full Text]

Ben Bassat, H., Goldblum, N., Mitrani, S. & 7 other authors (1977). Establishment in continuous culture of a new type of lymphocyte from a "Burkitt like" malignant lymphoma (line D.G.-75). Int J Cancer 19, 2733.[Medline]

Church, G. M., Ephrussi, A., Gilbert, W. & Tonegawa, S. (1985). Cell-type-specific contacts to immunoglobulin enhancers in nuclei. Nature 313, 798801.[CrossRef][Medline]

Daibata, M., Humphreys, R. E. & Sairenji, T. (1992). Phosphorylation of the EpsteinBarr virus BZLF1 immediate-early gene product ZEBRA. Virology 188, 916920.[CrossRef][Medline]

Decker, L. L., Klamen, L. D. & Thorley-Lawson, D. A. (1996). Detection of the latent form of EpsteinBarr virus DNA in the peripheral blood of healthy individuals. J Virol 70, 32863289.[Abstract]

Ephrussi, A., Church, G. M., Tonegawa, S. & Gilbert, W. (1985). B lineage-specific interactions of an immunoglobulin enhancer with cellular factors in vivo. Science 227, 134140.[Abstract/Free Full Text]

Flemington, E. & Speck, S. (1990). Identification of phorbol ester response elements in the promoter of EpsteinBarr virus putative lytic switch gene BZLF1. J Virol 64, 12171226.[Abstract/Free Full Text]

Gutsch, D., Holley Guthrie, E., Zhang, Q. & Kenney, S. (1993). NF-kB can physically interact with the EpsteinBarr virus Z protein and inhibit Z function. In XIII International Herpesvirus Workshop, July 2530. 30.

Katz, D. A., Baumann, R. P., Sun, R., Kolman, J. L., Taylor, N. & Miller, G. (1992).Viral proteins associated with the EpsteinBarr virus transactivator, ZEBRA. Proc Natl Acad Sci U S A 89, 378382.[Abstract/Free Full Text]

Kenney, S. C., Holley Guthrie, E., Quinlivan, E. B., Gutsch, D., Zhang, Q., Bender, T., Giot, J. F. & Sergeant, A. (1992). The cellular oncogene c-myb can interact synergistically with the EpsteinBarr virus BZLF1 transactivator in lymphoid cells. Mol Cell Biol 12, 136146.[Abstract/Free Full Text]

Kraus, R. J., Mirocha, S. J., Stephany, H. M., Puchalski, J. R. & Mertz, J. E. (2001). Identification of a novel element involved in regulation of the lytic switch BZLF1 gene promoter of EpsteinBarr virus. J Virol 75, 867877.[Abstract/Free Full Text]

Miyashita, E. M., Yang, B., Lam, K. M., Crawford, D. H. & Thorley Lawson, D. A. (1995). A novel form of EpsteinBarr virus latency in normal B cells in vivo. Cell 80, 593601.[CrossRef][Medline]

Montalvo, E. A., Shi, Y., Shenk, T. E. & Levine, A. J. (1991). Negative regulation of the BZLF1 promoter of EpsteinBarr virus. J Virol 65, 36473655.[Abstract/Free Full Text]

Montalvo, E. A., Cottam, M., Hill, S. & Wang, Y.-C. J. (1995). YY1 binds to and regulates cis-acting negative elements in the EpsteinBarr virus BZLF1 promoter. J Virol 69, 41584165.[Abstract]

O'Riordan, M. & Grosschedl, R. (1999). Coordinate regulation of B cell differentiation by the transcription factors EBF and E2A. Immunity 11, 2131.[CrossRef][Medline]

Prang, N., Wolf, H. & Schwarzmann, F. (1995). EpsteinBarr virus lytic replication is controlled by a posttranscriptional mechanism of BZLF1. J Virol 69, 26442648.[Abstract]

Prang, N., Wolf, H. & Schwarzmann, F. (1999). Latency of EpsteinBarr virus is stabilized by antisense-mediated control of the viral immediate-early gene BZLF-1. J Med Virol 59, 512519.[CrossRef][Medline]

Pscherer, A., Dorflinger, U., Kirfel, J., Gawlas, K., Ruschoff, J., Buettner, R. & Schule, R. (1996). The helix-loop-helix transcription factor SEF-2 regulates the activity of a novel initiator element in the promoter of the human somatostatin receptor II gene. EMBO J 15, 66806690.[Medline]

Ruf, I. K. & Rawlins, D. R. (1995). Identification and characterization of ZIIBC, a complex formed by cellular factors and the ZII site of the EpsteinBarr virus BZLF1 promoter. J Virol 69, 76487657.[Abstract]

Schwarzmann, F., Prang, N., Reichelt, B., Rinkes, B., Haist, S., Marschall, M. & Wolf, H. (1994). Negative cis-acting elements in the distal part of the promoter of EpsteinBarr virus trans-activator gene BZLF1. J Gen Virol 75, 19992006.[Abstract/Free Full Text]

Shen, C. P. & Kadesch, T. (1995). B-cell-specific DNA binding by an E47 homodimer. Mol Cell Biol 15, 45184524.[Abstract]

Shimizu, N. & Takada, K. (1994). Analysis of the BZLF1 promoter of EpsteinBarr virus: identification of an anti-immunoglobulin response sequence. J Virol 67, 32403245.

Sinclair, A. J., Brimmell, M. & Farrell, P. J. (1992). Reciprocal antagonism of steroid hormones and BZLF1 in switch between EpsteinBarr virus latent and productive cycle gene expression. J Virol 66, 7077.[Abstract/Free Full Text]

Tierney, R. J., Steven, N., Young, L. S. & Rickinson, A. B. (1994). EpsteinBarr virus latency in blood mononuclear cells: analysis of viral gene transcription during primary infection and in the carrier state. J Virol 68, 73747385.[Abstract/Free Full Text]

Walling, D. M., Perkins, A. G., Webster-Cyriaque, J., Resnick, L. & Raab-Traub, N. (1994). The EpsteinBarr virus EBNA-2 gene in oral hairy leukoplakia: strain variation, genetic recombination, and transcriptional expression. J Virol 68, 79187926.[Abstract/Free Full Text]

Wang, Y. C., Huang, J. M. & Montalvo, E. A. (1997). Characterization of proteins binding to the ZII element in the EpsteinBarr virus BZLF1 promoter: transactivation by ATF1. Virology 227, 323330.[CrossRef][Medline]

Young, L. S., Lau, R., Rowe, M. & 7 other authors (1991). Differentiation-associated expression of the EpsteinBarr virus BZLF1 transactivator protein in oral hairy leukoplakia. J Virol 65, 28682874.[Abstract/Free Full Text]

Zhang, Q., Gutsch, D. & Kenney, S. (1994). Functional and physical interaction between p53 and BZLF1: implications for EpsteinBarr virus latency. Mol Cell Biol 14, 19291938.[Abstract/Free Full Text]

Received 31 July 2002; accepted 3 December 2002.