An ATF/CRE element mediates both EBNA2-dependent and EBNA2-independent activation of the Epstein-Barr virus LMP1 gene promoter

Abstract

The Epstein-Barr virus (EBV) latent membrane protein 1 (LMP1) is a viral oncogene whose expression is regulated by both viral and cellular factors. EBV nuclear antigen 2 (EBNA2) is a potent transactivator of LMP1 expression in human B cells, and several EBNA2 response elements have been identified in the promoter regulatory sequence (LRS). We have previously shown that an activating transcription factor/cyclic AMP response element (ATF/CRE) site in LRS is involved in EBNA2 responsiveness. We now establish the importance of the ATF/CRE element by mutational analysis and show that both EBNA2-dependent activation and EBNA2-independent activation of the promoter occur via this site but are mediated by separate sets of factors. An electrophoretic mobility shift assay (EMSA) with specific antibodies showed that the ATF-1, CREB-1, ATF-2 and c-Jun factors bind to the site as ATF-1/CREB-1 and ATF-2/c-Jun heterodimers whereas the Sp1 and Sp3 factors bind to an adjacent Sp site. Overexpression of ATF-1 and CREB-1 in the cells by expression vectors demonstrated that homodimeric as well as heterodimeric forms of the factors transactivate the LMP1 promoter in an EBNA2-independent manner. The homodimers of ATF-2 and c-Jun did not significantly stimulate promoter activity. In contrast, the ATF-2/c-Jun heterodimer had only a minor stimulatory effect in the absence of EBNA2 but induced a strong transactivation of the LMP1 promoter when coexpressed with this protein. Evidence for a direct interaction between the ATF-2/c-Jun heterodimeric complex and EBNA2 was obtained by EMSA and coimmunoprecipitation experiments. Thus, our results suggest that EBNA2-induced transactivation via the ATF/CRE site occurs through a direct contact between EBNA2 and an ATF-2/c-Jun heterodimer. EBNA2-independent promoter activation via this site, on the other hand, is mediated by a heterodimeric complex between the ATF-1 and CREB-1 factors.

FIG. 1

Schematic presentation of the LRS in the B95-8 EBV genome. The scale refers to the position relative to the transcription initiation site from the EDL1 promoter. Transcription factor-binding sites previously identified as involved in regulation of the LMP1 promoter are indicated by open boxes. The Sp and the ATF/CRE sites are defined in the present investigation.

FIG. 2

EBNA2-induced transactivation of LRS depends on intact Sp and ATF/CRE motifs in LRS. Mutations were introduced in the pgLRS(−106)CAT and pgLRS(−634)CAT plasmids, respectively, as indicated in Materials and Methods. The reporter plasmids were cotransfected with pEΔA6 (+EBNA2) or with an equivalent amount of pSV2gpt (−EBNA2) in the EBV-negative B-cell line DG75. The CAT activity is given as relative chloramphenicol acetylation expressed as a percentage of the activity obtained with pgLRS(−634)CAT in the presence of EBNA2. The 100% value corresponded to acetylation of 97% of the substrate in the assay. The values are the mean of three independent transfections. The standard errors are indicated by error bars.

FIG. 3

EBNA2-induced transcription initiates at the correct LMP1 promoter site in reporter plasmids. RNA was prepared from DG75 cells transfected with the EBNA2 expression vector pEΔA6 or the pSV2gpt control vector and the indicated LRS CAT reporter plasmids and subjected to RNase protection analysis with a ³²P-labelled probe corresponding to positions −106 to +40 of LRS and the first part of the CAT gene. Lanes: 1, probe only; 2, pgLRS(−54)CAT and pSV2gpt; 3, pgLRS(−54)CAT and pEΔA6; 4, pgLRS(−106)CAT and pSV2gpt; 5, pgLRS(−106)CAT and pEΔA6; 6, pgLRS(−634)CAT and pSV2gpt; 7, pgLRS(−634)CAT and pEΔA6; 8, pgCAT and pSV2gpt; 9, pgCAT and pEΔA6; 10, DNA size markers. The band corresponding to the common initiation site in the EDL1 promoter is indicated by the solid arrow. Bands corresponding to nonspecific initiation upstream of LRS in the vector part of the reporter plasmids are indicated by dotted arrows. The lengths of these protected fragments differ depending on the plasmid. The solid arrowhead indicates a band present in all samples, probably due to incomplete RNase cleavage.

FIG. 4

Sp and ATF/CRE transcription factors in B-lymphoid cells bind to LRS. A ³²P-labelled double-stranded synthetic oligonucleotide corresponding to the −50 to −19 LRS region was incubated with nuclear extracts from DG75 cells and subjected to EMSA. Lane 1 shows the binding pattern obtained with the nuclear extract. Competition reactions was carried out as indicated below the autoradiogram and described in Materials and Methods. In lanes 2 to 5, the binding mixtures contain a 300-fold excess of unlabelled competitor over probe; in lanes 6, 8, and 10, they contain a 200-fold excess; and in lanes 7, 9, and 11, they contain a 300-fold excess. Some of the competitors were mutated at the Sp and/or the ATF/CRE sites as specified in Materials and Methods. Five complexes indicated by solid arrows are considered specific and designated Sp (bands remaining after competition with an LRS fragment that contained a mutated Sp site) and ATF/CRE (bands remaining after competition with an LRS fragment that contained a mutated ATF/CRE site), respectively. Three nonspecific bands that were not abolished by competition with unlabelled probe are indicated by dotted arrows.

FIG. 5

Identification of the transcription factors interacting with the Sp and ATF/CRE motifs in LRS. (A) Nuclear extract of DG75 cells was incubated under binding conditions with a ³²P-labelled double-stranded oligonucleotide corresponding to the −50 to −19 LRS region in the presence of a 300-fold molar excess of the competitor LRS −50/−19 with a mutated Sp site. Antibody supershifts were carried out by incubation with a goat polyclonal antibody against Sp1, a rabbit polyclonal antibody against Sp3, and a mixture of the antibodies, as indicated below the autoradiogram. The reaction mixtures were analyzed by EMSA. Three specific complexes are indicated by solid arrows, one designated Sp1 and two designated Sp3. Two bands that were not abolished by competition are indicated by the dotted arrows. The position of the anti-Sp1 antibody-shifted complex is shown by the solid arrowhead. (B and C) EMSA and antibody supershift analysis were performed by incubating nuclear extract of DG75 cells under binding conditions with a ³²P-labelled double-stranded oligonucleotide corresponding to the LRS −50 to −19 region with a mutated Sp site and with antibodies as indicated below the autoradiogram. Two bands that were not abolished by competition are indicated by the dotted arrows. (B) Three specific complexes are indicated by solid arrows, two of which are designated ATF-1, CREB-1 since they contain both factors. The positions of the antibody complexes are indicated by an open arrowhead for the anti-CREB-1 shift and solid arrowheads for the anti-ATF-1 shifts. (C) The third of the three specific complexes indicated by solid arrows is identified as ATF-2, c-Jun. The positions of the immunologically shifted complexes are shown by the solid arrowheads for the anti-ATF-1 shifts and the open arrowhead for the anti-c-Jun shift.

FIG. 6

Sp1 transactivates the LMP1 promoter independently of EBNA2. The pCMV-Sp1 expression vector or an equivalent amount of the empty pCMV control plasmid was cotransfected with the EBNA2 expression vector pEΔA6 or the pSV2gpt plasmid with the pgLRS(−106)CAT reporter plasmid or the mutated derivative pgLRS(−106)(Spmut)CAT in DG75 cells. The CAT activity is expressed as percent chloramphenicol acetylation, with the value obtained with the pCMV plasmid together with pEΔA6 and pgLRS(−106)CAT as 100%. The standard errors are indicated by error bars. The 100% value corresponded to 22% conversion of substrate to product in the CAT assay. The values presented are the mean of four independent transfections.

FIG. 7

The LMP1 promoter can be transactivated by CREB-1 and ATF-1 homo- and heterodimers independently of EBNA2 and by a c-Jun/ATF-2 heterodimer in an EBNA2-dependent manner. (A) The pc(ATF-1), and pc(CREB-1) expression vectors, separately or mixed, or the pc control plasmid was cotransfected with the EBNA2 expression vector pEΔA6 or an equivalent amount of pSV2gpt and the reporter plasmids pgLRS(−106)CAT or pgLRS(−106)(ATF/CREmut)CAT into DG75 cells, as detailed in Materials and Methods. The CAT activity is expressed as the percent chloramphenicol acetylation relative to the value obtained in transfections with the pc plasmid together with pEΔA6 and pgLRS(−106)CAT. The 100% value corresponded to 21% conversion of substrate to product in the CAT assay. The standard errors are indicated with error bars. The values shown are the mean of three independent transfections. (B) The pc(ATF-2) and pc(c-Jun) expression vectors, separately or in combination, or the pc control plasmid was cotransfected with the EBNA2 expression vector pc(BYRF) or an equivalent amount of the pc plasmid and the reporter plasmids pgLRS(−106)CAT or pgLRS(−106)(ATF/CREmut)CAT into DG75 cells, as detailed in Materials and Methods. The CAT activity is expressed as the percent chloramphenicol acetylation relative to the value obtained in transfections with the pc plasmid together with pc(BYRF) and pLRS(−106)CAT. The 100% value corresponded to 12% conversion of substrate to product in the CAT assay. The standard errors are indicated by error bars. The values shown are the mean of three independent transfections.

FIG. 8

EBNA2 does not affect the level or phosphorylation state of c-Jun or ATF-2 in DG75 cells. The EBNA2 expression vector pEΔA6 or equivalent amounts of the pSV2gpt control plasmid or the E1A 13S expression vector were transfected together with the CD2 expression vector pE300CY6 in DG75 cells. The transfected cells were selected for their CD2 expression with magnetic beads. The cells were lysed and equal amounts of protein extract were analyzed by SDS-PAGE and immunoblotting. The antibodies used in panel A were anti-c-Jun, anti-phospho-c-Jun(Ser63), and anti-phospho-c-Jun(Ser73), and those used in panel B were anti-ATF-2 and anti-phospho-ATF-2(Thr71). NIH 3T3 cell extracts containing nonphosphorylated or phosphorylated forms of c-Jun (panel A, lanes 1 and 2) and ATF-2 (panel B, lanes 1 and 2), respectively, were used as controls of antibody activity. The anti-c-Jun(Ser73) antibody also detected JunD phosphorylated at the Ser100 residue.

FIG. 9

In vitro-translated EBNA2 abrogates the binding of the in vitro-translated heterodimer c-Jun/ATF-2, but not the respective homodimeric forms, to the ATF/CRE site. The ³²P-labelled oligonucleotide probes indicated in the figure were incubated with DG75 nuclear extract and/or in vitro-translated proteins and analyzed by EMSA. The three specific complexes obtained with DG75 nuclear extract and identified above are indicated by solid arrows and designated ATF-1, CREB-1 and ATF-2, c-Jun. In vitro-translated proteins are denoted by the prefix IVT, and the positions of the corresponding complexes are shown by solid arrows. Nonspecific bands that were not abolished by competition are indicated by the dotted arrows. (A) The binding-reaction mixtures contained a ³²P-labelled −50 to −19 LRS oligonucleotide probe with a mutated Sp site and DG75 nuclear extract (lane 1) or the same probe with in vitro-translated ATF-2 (lanes 2 to 6). Antibody supershifts were performed by incubation with a rabbit polyclonal antibody against ATF-2 (lane 3) or a rabbit polyclonal antibody against CREB-2 (lane 4). A supershifted band is indicated by a solid arrowhead. In lanes 5 and 6, aliquots of reticulocyte in vitro translation reactions with EBNA2 DNA or control DNA were added to the binding-reaction mixtures. (B) The binding-reaction mixtures contained a ³²P-labelled AP-1 consensus oligonucleotide probe and in vitro-translated c-Jun protein (lane 1). Antibody supershifts were performed by incubation with a mouse monoclonal antibody against c-Jun (lane 2) or a goat polyclonal antibody against Jun B (lane 3). A supershifted band is indicated by a solid arrowhead. In lanes 4 and 5, aliquots of reticulocyte in vitro translation reaction mixtures with EBNA2 DNA or control DNA were added to the binding-reaction mixtures. (C) The binding-reaction mixtures contained ³²P-labelled −50 to −19 LRS oligonucleotide probe with a mutated Sp site and DG75 nuclear extract (lane 1) or in vitro-translated ATF-2 and c-Jun protein (lanes 2 to 7). Antibody supershifts were performed by incubation with a rabbit polyclonal antibody against ATF-2 (lane 3), a mouse monoclonal antibody against c-Jun (lane 4), or a goat polyclonal antibody against Jun B (lane 5). Supershifted bands are indicated by solid and open arrowheads. In lanes 6 and 7, aliquots of reticulocyte in vitro translation reaction mixtures with EBNA2 DNA or control DNA were added to the binding-reaction mixtures.

FIG. 10

EBNA2 coimmunoprecipitates with the c-Jun/ATF-2 heterodimer. The EBNA2 expression vector pc(BYRF) or equivalent amounts of the pc control plasmid were transfected together with pc(c-Jun) and pc(ATF-2) and the CD2 expression vector pE300CY6 in DG75 cells. The transfected cells were selected for CD2 expression with magnetic beads. After lysis of the cells, the proteins were immunoprecipitated with specific antibodies and adsorption to protein A/G agarose, eluted, and analyzed by SDS-PAGE and immunoblotting. Cell extracts that had not been subjected to immunoprecipitation were analyzed in lanes 1 to 3, and control immunoprecipitations without the specific antibody but including the adsorption step with protein A/G were analyzed in lanes 10 to 12. Lanes: 1, DG75 cells transfected with pc(BYRF); 2, DG75 cells transfected with the pc plasmid; 3, B95-8 cells; 4, anti-ATF-2 precipitate from DG75 cells transfected with pc(BYRF); 5, anti-ATF-2 precipitate from DG75 cells transfected with the pc plasmid; 6, anti-ATF-2 precipitate from B95-8 cells; 7, anti-c-Jun precipitate from DG75 cells transfected with pc(BYRF); 8, anti-c-Jun precipitate from DG75 cells transfected with the pc plasmid; 9, anti-c-Jun precipitate from B95-8 cells; 10, protein A/G agarose eluate from DG75 cells transfected with pc(BYRF); 11, protein A/G agarose eluate from DG75 cells transfected with the pc plasmid; 12, protein A/G agarose eluate from B95-8 cells. Antibodies used for visualizing the proteins on the immunoblots were anti-ATF-2 (A), anti-c-Jun (B), and a human serum containing anti-EBNA2 antibodies (C). The positions of ATF-2, c-Jun, EBNA2, and immunoglobulin heavy chains (Ig H) are indicated by the solid arrowheads.