Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

The enhancer factor R of Epstein-Barr virus (EBV) Is a sequence-specific DNA binding protein

The enhancer factor R of Epstein-Barr virus (EBV) Is a sequence-specific DNA binding protein Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6835 The enhancer factor R of Epstein-Barr virus (EBV) Is a sequence-specific DNA binding protein Henri Gruffat, Evelyne Manet, Agnes Rigolet and Alain Sergeant* Laboratoire de Virologie Moleculaire, Ecole Normale Superieure de Lyon, UMR 49 CNRS-ENS, 46 Allee d'ltalie, 69364 Lyon Cedex 07, France Received September 20, 1990; Revised and Accepted October 26, 1990 ABSTRACT In cells latently infected with EBV, the switch from whose coding sequences are created by joining the ends of the latency to productive infection is linked to the linear virus (6; 7). expression of two EBV transcription factors called EB1 The latent EBV genome is spontaneously activated in particular (or Z) and R. EB1 is an upstream element factor which cell lines, where between 0.5% and 5% of the cells produce has partial homology to the AP1/ATF family, whereas viruses. It can also be activated by various chemical agents R is an enhancer factor. In the R-responslve enhancer including the tumor promoter 12-O-tetradecanoyl-phorbol of the replication origin only active during the EBV lytlc 13-acetate (TPA) (8). In both cases, the activation seems to be cycle (ORIIyt), R-responsive elements are located in a linked to the expression of two EBV-encoded transactivators of region of about 70 bp (RRE-DR). Here we show that R, early gene promoters, EB1 (also called Z) and R (9; 10,11; 12; produced either by In vitro translation, or present in 13; 14; 15). EB1 is encoded by the open reading frame (ORF) nuclear extracts from HeLa cells constitutively BZLF1 and is expressed from two promoters, PZ and PR, either producing R, binds directly to and protects against as a lkb monocistronic mRNA or as 3 and 4 kb mRNAs DNAase I digestion, two regions in RRE-DR. Using generated by alternative splicing and expressing both EB1 and mobility shift assay and DMS interference, we have R, the BRLF1 ORF encoded factor (Figure 1A) (16). characterized the contact-points between R and the EB1 seems to have a key role in the induction of the lytic cycle DNA. Two binding sites, RRE-DR1 and RRE-DR2, were (10; 11; 12). It is a DNA binding protein (17; 18; 19) that characterized and are contiguous in RRE-DR. R binds positively autoregulates its own promoters but also activates to these two sites probably by simultaneously transcription from quite different responsive elements including contacting two sequences within the sites, which are AP-1 binding sites (20; 21). EB1 does not seem to be a factor separated by 7 bp in RRE-DR1, cctGTGCCttgtcccGT- that can act at distances more than 100 to 200 bp from the TATA GGACaatgtccc, and by 6bp in RRE-DR2, caatGTCCC- box (20). R, however, seems to be a factor that can act at tccagcGTGGTGgctg. Direct Interaction of R with its distances over thousands of base pairs and several R targets have cognate sequences is conferred by its N-terminal 355 been identified (22; 23a; 23b; 24; 25). One is part of the amino-acids. Directed mutagenesis in RRE-DR, of either duplicated promoter DR/DL (26; 22), and overlaps with the R-binding site, impaired binding of R in vitro and, as enhancer of the EBV origins of replication active only during assayed by transient expression in HeLa cells, impaired the lytic cycle and called ORIIyt (27). R-activation by a factor of two. This suggests that RRE- ORIIyt activity is dependent on the EBV-encoded DNA DR1 and RRE-DR2 do not respond cooperatively to R. polymerase, on the BZLF1 encoded transcription factor EB1, and on the presence of the enhancer located upstream from the DR/DL TATA boxes (27) (Figure 1A). This enhancer has two INTRODUCTION functionally distinct regions, A and B. Region A is constitutively The human herpes virus EBV (Epstein-Barr virus) infects and active in all cell lines tested so far except lymphoid B cells, whereas region B was transactivated by R in all cell lines tested immortalizes peripheral B lymphocytes, resulting in the (22). One R target in the DR enhancer B region has been reduced establishment of a latent infection. In such latently infected B to 28 bp and contains the double palindromic sequence B0, CC- cells, the entire EBV genome is maintained mainly as a plasmid, TGTGCC7TG7TCCGTGG4CA4TGTCC (22). When this 28bp and its expression reduced to a few genes: those encoding two DNA fragment was placed upstream from the rabbit /3-globin small RNAs (EBERS) (1), the six Epstein-Barr Nuclear Antigens promoter, it mediated an 8 fold R-induction (28). However, the (EBNA-1, -2, -3A, -3B, 3C and LP) (For a review see ref 2), double palindrome TTGTCCCGTGGACAATGTCC either alone the BHRF1 encoded protein (3; 4) , the latent membrane protein or duplicated did not confer R-responsiveness to the /3-globin (LMP) (5) and the terminal membrane proteins (TP1 and TP2), * To whom correspondence should be addressed Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6836 Nucleic Acids Research, Vol. 18, No. 23 promoter (28). The contiguous and partially overlapping sequence The RNA obtained were used to program protein synthesis in GTCCCTCCAGCGTGGTGGCTGCC, called Bl, mediated a 3 messenger dependent rabbit reticulocyte lysates (Promega) using fold R-activation when placed upstream of the /3-globin promoter 14C-L-Leucine. (28). However, when linked to each other, BO and Bl (RRE- DNAase I footprints DR), mediated a 25 fold R-activation (28). We show in this report that R binds in the B enhancer region, Footprints experiments were made using an EBV ORDyt enhancer to sequences which fail within the RRE-DR. Moreover, we show probe derived from plasmid pG2-899/741, 5'-labelled by 32P by mobility shift assay and DMS interference, that R binds at either position 741 (probe pG2EC) or at position 899 (plasmid independently in vitro to two sites within the RRE-DR (the RRE- pG2ENC) (Figure 3). The experimental procedure is described DR 1 and the RRE-DR2). Each binding site covers about 18 bp, in ref 33. where R probably simultaneously contacts two core sequences Electrophretic Mobility Shift Assay (EMSA) separated by 6 or 7 bp. In the RRE-DR, mutations that impaired binding of R to one site or the other, only reduced the R- 2/xl of in vitro translation extract, were incubated with 2 x 104 transactivation by a factor of 2, suggesting that the RRE-DR 1 cpm of ^P-labeled-RREs (figure 1A). Incubations were carried out in 0,5 mM MgCl , 10 mM HEPES-KOH (pH 7.9), 0.5 mM and the RRE-DR2 do not respond cooperatively to R. DTT, 0.5 mM PMSF, 150 mM KC1, 10% glycerol, at 25°C for 30 minutes. The mixture was loaded onto a 4.5% MATERIALS AND METHODS polyacrylamide gel (29 to 1 crosslinked), 0.2XTBE. The R-RRE Cloned DNA templates complexes (B) were separated from the non-complexed DNA (F) by migration at 10 V/cm and visualized by autoradiography. The EB1 and R expression vectors have been described extensively elsewhere (16). Briefly, they are pUC18 derivatives DMS interferences containing the ORFs BZLF1 and BRLF1 that code respectively for the EBV transactivating factors EB1 and R, placed under the 5 x 105 cpm of the DNA probes was methylated using 1 yX of control of the SV40 early promoter-enhancer. Plasmid pG2 DMS during 3 mn. at 18°C. The methylated probe was then contains the rabbit /3-globin gene with the M13mpl2 polylinker incubated with 2/il of in vitro translation extract. After EMSA cloned 5 ' to the /3-globin promoter (29). Plasmids pG2.899/741, assay, the retarded DNA probe (B) and the non-retarded DNA pG2.899/752, pG2.899/775, pG2.899/812, pG2.737/805, were probe (F) were electroeluted and incubated in 100 y.\ of 1M made by ligating subregions of the B region of the DR enhancer piperidine for 30 mn. at 90°C. An equal amount of the radioactive 425 bp 5 ' to the /3-globin promoter (see figures 2B and 3B). All B and F probes was analysed on 8% polyacrylamide sequencing clones were verified by sequencing. Numbers after pG2 describe gels and visualized by autoradiography. the position and the orientation of the border of the inserted Cell culture and transfections enhancer subregions and they refer to the map coordinates described in Figures 1A and 2A. Plasmid pSV2/3 (28) expresses HeLa cells were grown in DMEM (Gibco) supplemented with a chimeric SV40-/3-globin RNA and was cotransfected as an 10% (v/v) fetal calf serum. The plasmids used for transfection internal control for transient expression experiments. Plasmid were prepared by the alkaline lysis method and purified through pSVO contains the SV40 Hpa II (map position 346) to Hind in two CsCl gradients. HeLa cells were seeded at 106 cells per 100 (map position 5171) fragment cloned in pUC19 digested with mm Petri dish 8 h prior to transfection. Transfections were Hind HI and BamH I. This plasmid was included in transfections performed by the calcium precipitate method (30). Cells were to keep the amount of SV40 early promoter sequences constant, mixed with the appropriate DNA(s), and the DNAs were in the since EB1 and R are expressed under the control of the SV40 same topological state as assayed by agarose gel electrophoresis. early promoter. Usually 15/tg of DNA were used per 100 mm dish including: 0,5/tg of R-expressing vector, 5/tg of reporter promoters, pSVO Production of HeLaN and HeLaR cells when required, 0,5/tg of pSV2/3 as internal control and pUC19 up to 15/tg. HeLa cells were transfected with plasmids p36/7polyA or p36/7polyAR (Figure 2A), by the calcium precipitate method RNA extraction and SI nuclease mapping (30). 20 hrs after transfection, the cells were exposed to G418 at a concentration of 800mg/L. Every three days, the cells were The cells transfected were lysed by NP40 as described elsewhere washed and placed in fresh medium plus G418. Several G418 (29). Nuclei were pelleted and RNA phenol extracted from the resistant clones were isolated, amplified and frozen. Two of them cytoplasmic fraction. 10 to 40 /tg of total cytoplasmic RNA was were selected and called HeLaR (for cells transfected with hybridized overnight at 30°C in 50% formamide, 0.3M NaCl, p36/7polyA R) and HeLaN (for cells transfected with 0.01M Tris-HCl pH 7.4 to 5'- 32P-labelled synthetic single- p36/7polyA). stranded DNA probes (Figure IB). The hybrids were digested 2 hours at 20°C with 5U of SI nuclease per 10 /xg of RNA. The HeLa cells nuclear extracts size of the SI protected DNA fragment was analysed on 8% (w/v) acrylamide/8.3 M urea gels. Quantification was made by cutting Nuclear extracts were prepared from HeLaN and HeLaR cells, the specific SI-protected bands out of the gel and counting the by the method of Dignam (31), with the modifications introduced radioactivity. The results were corrected as follows: (i) according by Wildeman et al. (32). to the efficiency of transfection as evaluated by counting the Production of R and mutant proteins in vitro radioactivity present in SI-protected probes corresponding to specific SV40 early RNA expressed from plasmid pSV2/3 and The BRLF1 ORF was cloned in plasmids pSPT18 or pSPT19 (ii) according to the activities of the different constructions in (Boehringer Mannheim). Cloned inserts within the polylinker the absence of R. region were transcribed from either the SP6 or the T7 promoters. Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6837 RESULTS |—^ft^n#*tft)pzi MH.FI The ORDyt RRE contains two regions responding to R EBI ** , „ „. ,„ U5 t. Tl Ui III U2IM US We have previously shown that the ORDyt enhancer sequences located between positions -741 to -899 (region B) upstream from the IR4 gene CAP site (Figure 1A), responded to R in a non cell-specific manner when linked to an heterologous promoter (22; 28). We show here, that 69bp within the R-responsive enhancer contains the R-responsive element. Enhancer deletion mutants placed 425 bp upstream of the rabbit /3-globin promoter in plasmid pG2 (Figure IB), were transfected in HeLa cells, and their response to increasing amounts of R-expressing vector was evaluated by quantitative SI analysis. We do not show that increasing amounts of R-expressing vector result in increasing amounts of R protein. However, since specific R-induced transcription increased in response to increasing amounts of R- expressing vector (see figure IB), we assume that this is rather due to increasing amounts of R-protein. The enhancer fragments inserted increased 2 fold the basal activity of the /3-globin promoter in plasmid pG2 (FigureIB, compare lane 1 with lanes 2 and 3), suggesting that cellular factors participate in the function of this enhancer. R highly stimulated the expression of specific /3-globin transcripts when region B (plasmid pG2-899/741), was inserted 5' to the /3-globin promoter (Fig. IB, lanes 7, 11 and I 0 tl 12 I ) T4 16 « T7 15). Deletion of sequences between bp -741 and -75 2 (plasmid pG2-899/752), decreased the R-activation effect by a factor 2, and this was seen at low R concentrations (Figure IB, lanes 8 and 12), but not at high R concentrations (Figure IB, lane 16). Therefore, the effect of the deletion can be compensated by the amount of R protein expressed. Deletion of sequences between bp -752 and -775 (plasmid pG2-899/775), impaired the induction of /3-globin transcripts by R, and this effect could not be compensated by the amount of R expressed (FigureIB, lanes Figure 1. An R-rcsponsive enhancer overlaps with ORDyt. (A) Two EBV ORFs 9, 13 and 17). However, the remaining sequences retained a low have been shown to code for transcription factors: BZLF1 codes for EBI and BRLF1 codes for R. ORTlyt overlaps with the DR promoter where four EBI but detectable R-inducibility. In conclusion, it seems that there binding sites have been mapped (ZRE1 to ZRE4). ORIlyt is composed of three are at least two R-responsive elements within or overlapping with regions, indicated by open rectangles (27). Numbers over the rectangles indicate the sequences located between positions -742 and —775. This the end-points of the regions with respect to position 52787 as + 1 on the EBV is likely to be the case, since a DNA fragment overlapping this B95-8 sequence (34). Horizontal arrows indicate palindromic se<juences. ori region and placed upstream of the /3-globin gene (plasmid indicates the origin of replication. Oriryt is overlapping with a promoter controlling the expression of a short repeted sequence called IR4. Proximal to the CAP site, pG2-737/805), rendered this promoter responsive to R (Figure there are four EB1 binding sites called ZRE1 to 4 . Distal to the CAP site, region IB, lanes 6, 10 and 14), and the level of R-activation was -63 9 to -79 4 is the ORIlyt enhancer. The sequence of the R-rcsponsive region comparable with that obtained with region B at all R in the ORIlyt enhancer and denoted domain B is shown, and has been mutated concentrations tested (compare lanes 7,11 and 15 with lanes 6, by progressive deletion. The structures of the mutants are shown (thick lines). 10 and 14). However, there could be only one R-responsive Numbers at both ends of the thick lines indicate the end points of the deletions in domain B, with respect to position 52787 as +1 on the EBV B95-8 sequence. element, and the stimulatory effect might be due to a stabilizing These mutants have been linked to the 0-globin promoter (plasmid pG2). The effect of flanking sequences (-74 1 to -752). Plasmid pSV2/3, schematic structure of SI nuclease DNA probes and the size of the SI protected expressing a chimeric SV40-/3-globin RNA under the control of DNA fragments are also presented. (B) HeLa cells were transfected with the the SV40 early promoter-enhancer sequences, was cotransfected different constructs in the conditions indicated in the top panel. Transcnptional activation was determined by quantitative SI analysis of total cellular RNA isolated as a control for transfection efficiency (see figure IB, pSV2/S from transfected cells . Plasmid pSV2/3 expresses an SV4O-/3-globin hybrid RNA internal control). An equal amount of specifically initiated RNA under the control of the SV40 early promoter-enhancer was cotransfected as an was found in every transfection, indicating that the results internal control (28). The specific start sites of 0-globin and the early-early start described above could be compared. sites of SV40 are indicated by 0 and SV respectively. HeLaR cells constitutively produce a functional R protein polypeptide recognized by a rabbit anti-R polyclonal antibody Having shown that at least two R-responsive elements are located (Figure 2B, lane 2), not detected in HeLaN cells containing only between positions -737 and -805 , we wanted to determine if the plasmid p36/7polyA (Figure 2B, lane 1). This polypeptide R binds directly to this region and where. First, we made a HeLa migrated with the same apparent molecular weight (90 Kda), as cell line (HeLaR) constitutively producing the R protein. HeLaR the R protein expressed in Raji cells treated by TPA (Figure 2B, cells contain an integrated plasmid (p36/7polyAR, Figure 2A) compare lanes 2 and 4). We transfected the rabbit /3-globin gene carrying the neomycin gene as a selection marker, and the R gene into HeLaR cells, either enhancerless (plasmid pG2), or linked (Figure 2A, ORF BRLF1), fused to the metal-inducible human to the R-responsive enhancer (plasmid pG2-737/805) (Figure 2C). metallothionein HA promoter. These HeLaR cells produce a Treatment of the HeLaR cells by ZnCl , did not lead to 2 Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6838 Nucleic Acids Research, Vol. 18, No. 23 2 and 4). An equal amount of RNA was expressed from plasmid pSV2/J cotransfected as an internal control, suggesting that the p36/7 results described above could be compared (Figure 2C, lanes 5 to 8). In conclusion, HeLaR cells produce a functional R polypeptide, and the amount of R produced can be increased by ZnCl . Wild type R protein and a C-terminal deleted mutant of R (Manet et al, unpublished data) were also produced by in vitro translation. RNAs transcribed from the SP6 promoter located upstream from the R gene cloned in plasmid pSPT19 (Figure 2A, plasmid pSPT19R) or of the deleted mutant gene cloned in the same plasmid, were used to programm protein synthesis in messenger-dependent rabbit reticulocyte lysates. In these lysates, a 90Kda 14C-Leucine-labelled polypeptide corresponding to the wild type R was detected after SDS-PAGE and autoradiogaphy (Figure 2B, lane 7) whereas a 50 Kda 14C-Leucine-labelled polypeptide was detected for the C-terminal deleted mutant. R binds to the RRE-DR in vitro Four sources of proteins, HeLaR and HeLaN cells nuclear extracts, and reticulocyte lysates containing R or BMV (Bromomosaic virus) proteins translated in vitro were used in conjunction with 5'-end-labeled double stranded DNA probes covering the RRE-DR to determine the sequence-specific DNA P62-B05/737 properties of R by DNAase I footprinting. One probe was 5'-end- labeled close to position -741 and called pG2ENC (Figure 3A), and the other probe was 5'-end-labeled on the opposite strand close to position -899 and called pG2EC (Figure 3C). Several pG2 regions of probe pG2ENC, called I, II, HI, IV and V, were tokttt e-Sl.H« protected against DNAase I digestion by a HeLaN nuclear extract Inductd tin (Figure 3B, lanes 5 and 6), as compared to the DNAase I Non Induced ISOun ZnCl digestion pattern of the probe in absence of protein extract (Figure 3B, lanes 2 to 4). Moreover, two DNAase I hypersensitive sites pSVjB called b and c, were also induced. However, the HeLaR nuclear (Intaraal control) extract strongly increased the protection of regions I ( — 741 to 9 6 7 8 -760 ) and II (-76 0 to -800) (Figure 3D, lanes 7 and 8), and induced the DNAase I hypersensitive site a. These results suggest that R is binding directly to regions I and II. They are also compatible with the idea that the HeLaR extracts contain cellular factors induced by R that bind to region I and II. The same factor sv*- might be present in HeLaN cells at lower concentrations. To ff determine further if R directly binds to these regions, DNAase I footprints were performed with extracts containing in vitro Figure 2. HeLaR cells produce a functional R polypeptide (A) Schematic translated R. In this case, only region I and a shorter segment representation of the plasmids used to generate HeLaR and HeLaN cells, and of region II were protected against DNAase I digestion (Figure to produce R by in vitro translation. (B) HeLaR cells but not HeLaN cells induced 3B, lanes 16 and 17). As such a footprint is not observed with by ZnCl , produce a 90 Kda polypeptide immuno-related to R produced in Raji cells where the EBV early gene products EBl and R have been induced by TPA extracts containing proteins translated in vitro from BMV RNA (lanes 1 to 4). The proteins were revealed b y western Wot using two rabbits antisera, in the same conditions (Figure 3B, lanes 13 to 15), the DNAase one ami R (35) and one ami EBl, mixed together. R and a truncated version I footprint obtained with R in vitro translated can be considered of R (RD97st) were also produced by in vitro translation and separated on PA- as specific. As shown in Figure 3A, regions I and II overlap the GE (lanes 5 to 8). (C) To determine if the 90 Kda polypeptide produced by HeLaR sequences located between positions -737 and -775, which cells was transactivating the RRE-DR linked to the 0-globin promoter, HeLaR cells were transfected with plasmid pG2-805/737 as indicated on the top panel. are those that mediate R-activation of the ORIlyt enhancer (see The level of correctly initiated /3-globin RNA (fi) was determined by SI nuclease Figure 1). analysis (lanes 1 to 4). The amount of RNA expressed from pSV2)3 (SV) cotransfected as an internal control, is presented in lanes 5 to 8, and corresponds DNAase I footprints were also performed on probe pG2EC to the RNA samples presented in lanes 1 to 4. (Figure 3C). No clear footprint was detected on this strand with either the HeLaN or the HeLaR nuclear extract (Figure 3D, lanes 1 to 9). However, a strong DNAase I hypersensitive site called an increased expression of specifically initiated |3-globin mRNA a, was detected with the HeLaR nuclear extract (Figure 3D, lanes from plasmid pG2 (Figure 2C, lanes 1 and 3). However, in 7 to 9). This strong hypersensitive site was also induced by the in vitro translated R protein, but this time there was a clear HeLaR cells, /3-globin RNA levels were increased when the R- footprint overlapping the sequences located between positions responsive enhancer was inserted in front of the /3-globin promoter (Figure 2C, compare lanes 1 and 2). Moreover, this -73 7 and -77 5 (Figure 3D, lanes 14 and 15). As previously, R-enhanced /3-globin RNA transcription was further increased such a footprint is not seen with extracts containing proteins in by treatment of the cells by ZnC12 (Figure 2C, compare lanes vitro translated from BMV RNA (Figure 3D, lanes 12 and 13). Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6839 la vttr* tranalitM R HaiaB Htlii r aitrKt HH M MClaar <itr*cl ia vitro truslatt d R 31ca I I -COAUTCGACBSAACASeSCACCTIlTTACWEMGCTCECACCACCBAC -i- -741 -737 -775 S' ICO •l*-GCTC6*GCTGCCXrGTCCC6T66ACAA|6TCCCTCCASCGTG^TCeCTQ -74 1 -751 -775 CCTTTGGGATGCATC ACTTT 6 A DCC AC T AAGC C C CC 6T T GCT C G CCT T GC CT 6C C T G6AAACCCTACSTA8TBAAACTCG6TBATTCOSSe6CAACGAGCGGAACG6ACBBA -ooa -a'u iy -OO9 c -012 V CACCAT6ACACACTAAGCCCCT6CTAATCCAT6AGCCCCGCCTTTAG6AAGCACC STHTMTSTeTMTTCtG6«ACSArrAGCTACTCE£ECCECAAATCCTTCSTGCTB CASSfCCCTWCAWTCT - HIMIII* ACGTCCCGGGATCCTCTAGA-Hlndlll -m In vttr* trwulattos ff»cl««r Cntr*ct !• vitro trHnUtion Htt«JI HtiiR 1 2 3 4 S 6 7 6 •• * «!;:; • Figure 3. R binds to the RRE-DR. (A) and (C) Schematic representation of the sequences protected against DNAase I digestion by the different protein extracts used in the footprint experiment. Horizontal braids indicate footprints induced by in vitro translated R. Horizontal closed bars indicate footprints induced by HeLaR extracts. Open bars indicate footprints induced by HeLaN extracts. Vertical arrows and letters indicate DNAase I hypersensitive sites. (B) and (D) DNAase I footprints with decreasing amounts of DNAase I, on both strands of a DNA probe located between positions -741 to -899 . The protein extracts used are indicated in the top panel. Vertical bars and numbers indicate footprints. Horizontal arrows and letters indicate DNAase I hypersensitive sites. Closed triangles indicate decreasing amounts of DNAase I. Thus, in conclusion, R seems to bind directly to the sequences protein in conjunction with three synthetic double-stranded (RRE-DR) mediating the R-activation in a transient expression oligonucleotides overlapping regions I and II, and called BO, Bl and BR (Figure 4A). We also made use of a fourth double assay, as defined by deletion mutagenesis (Figure IB). stranded oligonucleotide called B2 (Figure 4A), which The R DNA-binding domain is located in the N-terminal 355 corresponds to a region protected by HeLaR extracts but not by amino-acids in vitro translated R (-78 7 to -807). Probe B2 should not therefore bind in vitro translated R. As judged by mobility shift As discussed above, R seems to bind in vitro to a region assays and competitions, specific binding of R was observed to overlapping positions -741 and —780 in the RRE-DR, and probes BO, Bl and BR, but not to probe B2. The in vitro possibly to two sites within region I and n . Interestingly, deletion mutagenesis in this region also showed a two-step decrease in translated R bound with a higher affinity to BO (Figure 4B, lane the R-responsiveness, corresponding approximately to regions 2) and to BR (Figure 4B, lane 14), than to Bl (Figure 4B, lane I and II. To localize further where R bound in the RRE-DR, 8). This was also seen when a 400 fold molar excess of unlabeled we made mobility shift assays using the in vitro translated R B0, Bl and BR probes were used as competitors. Probe B0 Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6840 Nucleic Acids Research, Vol. 18, No. 23 80 S I I '• -Induced footprint I RRE CCTGTGCCT1GTCCCGTGGACAATGTCCCTCCAGCGTGGTGGCTGCCTTT66GATGCATCACTTT6A6CC BO CCTGTGCCTTGTCCC6T66ACAATGTCCC B l GTCCCTCCAGCGTG6T66CT6CC B2 GGGATGCATCACTTT6A6CC BR TTGTCCCGTGGACAATGTCCCTCCA6CGT6GT6GCT6 probes -80- \- -B0- H h H proteins / L | R- ' L r- _|L | KD97 |L -R- competitors / / / BO Bl B2 / / / BR BO B2 / BO B2 / / BO B2 / 12 13 M E » 17 6 7 8 9 10 II 1 2 3 4 5 6 7 8 Figure 4. Mapping of R binding sites by gel retardation assay and DNA competition (A) Schematic representation of DNA probes used in the binding assays and competitions. (B) Probes BO, Bl and BR were incubated with protein extracts and with DNA competitors as indicated in the top panel. (C) In vitro translated R and RD97st proteins were incubated with probe BO and DNA competitors as indicated in the top panel. In (B) and (C), /, indicates no protein added; L, indicates lysate translated without exogenous RNA; DNA competitors were added at a mass ratio of 100; Free (F) and bound (B) DNA probe are indicated on the left of the autoradiographies. competed efficiently for the binding of R on probes BO (Figure 5'-GTGGAC-3' (Figure 5B). Such strong interferences suggest 4B, lane 3), Bl (Figue 4B, lane 9) and BR (Figure 4B, lane 12). that R contacts simultaneously both sequences rather than binding to them separately. Similar results were observed with probe BR, However, probe Bl competed less efficiently than probe BO for where strong interferences were detected on guanine residues in the binding of R on probe BO (Figure 4B, lane 4). As expected no R binding to the non-specific probe B2 could be detected (not the sequences 5'-GTCCC-3' and 5'-CGTGGGTG-3' (Figure shown), and unlabeled B2 probe did not detectably compete for 5C). Again, such strong interferences suggest that R contacts the the binding of R on labeled probes BO (Figure 4B, lane 5), Bl two sequences simultaneously. This was strengthened by the fact that there was no interference on the sequence 5'-GTGGAC-3' (Figure 4B, lane 11), and BR (Figure 4B, lane 17). The truncated present alone in probe BR, whereas interferences on this sequence R protein (RD97st) generated by in vitro-transcription/translation of the C-terminal deleted BRLF1 mutant construct (Figure 2B, were observed in probe BO (see Figure 5B). Similarly, there was lane 8), still bound to probe BO. RD97st contains the N-terminal no interference on the sequence 5'-GTCCC-3' located between 355 amino-acids of R, and has a higher affinity for probe BO positions —760 and -76 5 and present alone in probe BO (Figure 5B), whereas interferences were found on this sequence in probe than the wild type R protein (Figure 4C, compare lanes 1 and BR (Figure 5C). Moreover, on both the coding and the noncoding 5). Again, the interaction was specific since binding of RD97st to probe BO was competed out by an excess of unlabelled probe strands of a double stranded oligonucleotide covering the entire BO (Figure 4C, lane 6), but not by an excess of probe B2 (Figure RRE-DR, interferences were observed on the four sequences 4C, lane 7). Similar results were obtained with probes Bl and described above but with a much weaker intensity (not shown). The results of the methylation interferences are summarized in BR (not shown). These results suggest that there are at least two Figure 5A, and suggest that R binds to two contiguous sites in R binding sites in the RRE-DR. Furthermore, specific binding to these sites is still observed when the R protein is reduced to the RRE-DR. Each site probably consists of two regions directly the N-terminal 355 amino-acids. contacted by R and separated by 6 or 7 bp, 5'-GTGCCTTGT- CCCGTGGAC-3' and 5-GTCCCTCCAGCGTGGTG-3', called RRE-DR 1 and RRE-DR2 respectively. Similar results have R binds to two contiguous sites in the RRE-DR been obtained with the R deletion mutant RD97st (not shown). Since in vitro translated R bound to probes BO and BR, the contact points of R with these double-stranded oligonucleotides were The RRE-DR1 and the RRE-DR2 do not respond probed chemically by modification of the DNA with dimethyl cooperatively to R sulfate (methylation interference analysis). After partial methylation at purine residues, the modified probes were used In order to further evaluate the contribution of each R-binding in the gel retardation assay. Free (F) and bound (B) probe DNA site in the R-induced activity of the RRE-DR, we mutated the was recovered, subjected to chemical cleavage at methylation RRE-DR1 (Figure 6A, plasmid pG2M2), the RRE-DR2 (Figure sites, and the products were resolved by gel electrophoresis under 6A, plasmid pG2M3) or both (Figure 6A, plasmid pG2M4). denaturing conditions. On both the coding (C) and the non-coding These mutations impaired totally the binding of R on the isolated (N Q strands of probe BO, strong interferences were detected on RRE-DR 1 or RRE-DR2 in vitro (not shown). Each mutant and guanine residues in the sequences 5'-GTGCC-3' and the wild type RRE were placed upstream of the /3-globin promoter Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6841 -737 -74S -7«« -Tit TCTCCCTCCMCtTWTCSCTaXTTTCSUTCCATUCTTTSMCC »62RRE CCrCTGCCTTGTCCCGTCCACAItrcTCCtrCCMCtTeSTBSCTBCCTTTCCCATGCATCACTTTCMCC i R-lnducad footprint prob t puC I B 0 0 *•« -7?7.. ._ - r C 5-<Mcrc<iCTKCCCCTCTCCCTTCTCCCSTSS*CMTCTCCCCaG«TCCTCTMMCTCGU »C6AATTCCC666TTCTA6A6CTC 6 prob« »62 BR , EcoRi Sm«l xtai SK I Xhoi 45 7i0 Xh» I I , E» ftl rabbit B-globin C S'-CTCSMTTSTCCCCTCC*CMT{TCCtTCCACcSTSTGSCTSWn 1*fTTr HIWntTtTI>fTTr r Tnrny|*j|j'"'g''"' r "—**""' • f «C e 50»« B C lp62RREn2 t13 rul l pG2 RRE H2 113 rwl 1*62 RRE M2 H3 tw l prooa pUCIBBO pro»« p62BR fold induction I 17 8 7 13 2 32 19 21 6 C MC C He I 2 3 4 5 6 7 B 9 10 II 12 13 14 IS G*A F B S.A F B 6»A F B S'A F B PSV2B (liit«m«l cmtr*l> I 2 3 4 5 6 7 B 9 10 II 12 13 14 IS Figure 6. The paired non-homologous RRE sites in the RRE-DR do not respond G • • <•. synergkally to R. (A) Schematic representation of the /3-globin gene constructions in which the RRE-DR (pG2RRE), or mutants in the RRE-DR1 (pG2M2) in the RRE-DR2 (pG2M3), and in both RREs (pG2M4), have been inserted. (B) The various constructs were transfected in HeLa cells as indicated in the top panel, and the level of correctly initiated /3-globin transcripts(/3) was evaluated by quantitative SI nuclease assay. Fold induction represents the ratio of the radioactivity counted in each SI-protected band obtained in presence of R on the radioactivity counted in the corresponding SI-protected band obtained in absence of R. This has been done for the two R concentrations. ( Q SI nuclease analysis of RNA initiated from plasmid pSV2j3 (SV) cotransfected as an internal control. RNA samples 1 to 15 are the same as in (B). saturating (lanes 8 and 9) than saturating amounts of R Qanes 13 and 14). Finally, when both the RRE-DR1 and the RRE-DR2 were mutated (mutant M4), the R-induced activation was strongly Figure 5. Fine mapping of R binding sites by methylation interference (A) impaired, and this was seen at 500 ng as well as 50 ng of R- Sequences of the R-Responsive Enhancer (RRE) and of double stranded DNA expression vector (lanes 10 and 15). These results suggest that probes used for formation of protein-DNA complexes (pUC18BO and pG2BR) the RRE-DR1 and the RRE-DR2 do not respond cooperatively The R binding site is indicated by an open bar under RRE. The G residues involved to R. An equal amount of RNA was expressed from plasmid in the formation of specific nucleoprotein complexes on the probes are shown by black squares. DNA binding assays were performed under standard conditions pSV2/S (Figure 6C, lanes 1 to 15), indicating that the results except that probes pUC 18 BO and pG2 BR (5'-end-labeled either on the coding presented above could be compared. However, it should be noted (C) or the noncoding (NC) strand) were treated with dimethyl sulfate before that the /3-globin promoter, which lacks R-binding sites, is weakly incubation with the in vitro translated protein R. (B) and (C) G+ A is a Maxam stimulated by R at high concentrations of the factor, suggesting and Gilbert sequence ladder. F, is the non-retarded probe, and B is the retarded that R present at high concentrations can activate by binding non complex. Black squares indicate the G residues whose methylation strongly prevents the formation of complexes, open squares those which are weakly involved specifically to the /3-globin promoter. Moreover, /3-globin specific transcription from mutant M4 was still weakly activated by high concentrations of R (Figure 6B, compare lanes 2 and 6), suggesting that the mutations did not impair totally R binding in plasmid pG2 (Figure 6A). These reporter plasmids were in vivo. transfected into HeLa cells either alone or along with various amounts of R-expression vector, and their transcriptional activity was evaluated by quantitative SI analysis. Representative results DISCUSSION are shown in Figure 6B. As expected, the RRE-DR located upstream of the /3-globin promoter resulted in a strong In this report, we provide the first evidence that the N-terminal transcriptional induction dependent on the presence of R (lane 355 amino-acids of the EBV-encoded transcription factor R, 7), and this induction reached a plateau at 500ng of R-expression confer sequence-specific DNA binding properties to this protein vector (lane 12). The three base-pairs change in the RRE-DR1 in vitro. (mutant M2, lane 8), or in the RRE-DR2 (mutant M3, lane 9), Furthermore, using DNAase I footprints we have also shown resulted in a reduction of about 50% of the R-induced activation that HeLa cell proteins bound to regions I and II in the RRE- of the /3-globin promoter, and this was more obvious at non- DR, where two R-binding sites have also been identified. The Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6842 Nucleic Acids Research, Vol. 18, No. 23 protection against DNAase I digestion of regions I and II was early promoter PM, controlling the expression of a factor acting at the post-transcriptional level (23b; 24). Bidirectional deletion much more pronounced with HeLa cell extracts containing R than mutagenesis and DNAase I footprints indicate that there could with HeLa cell protein extracts. At present, we do not know if be four R-binding sites, with sequences only partially related to this reflects the induction by R of such cellular factor, or if there the one described in this publication (H. Gruffat and M. Buisson, is cooperativity between R and these cellular factors, or if R competes with these cellular factors for binding on the RRE-DR. unpublished results). Therefore, as in the case of the EBV- encoded upstream element factor EB1 (18), the EBV enhancer Whether these cellular factors are contacting DNA at the same factor R could bind to many sites which are imperfectly sites as R, and if such proteins contribute to the R-activation of homologous and rather degenerated. This possibility is interesting the enhancer is also at present unknown. However, mutations if one considers the regulation of EBV early genes expression. which affect the purine residues contacted by R, impaired R- binding in vitro and also impaired the R-induced transcriptional Two transcription factors, EB1 and R, seem to be involved in the activation of EBV early genes expression. EB1 expressed in activation in a transient expression assay (Figure 6B). This EBV latently infected B cells, activates the expression of R and suggests that R-mediated transcriptional activation requires direct of many EBV early promoters (12; 22). Many EB1 binding sites interaction between R and its cognate binding sites. Nevertheless, but fewer EB1 responsive elements (ZRE) have been identified. one cannot rule out the possibility that although R can directly bind in vitro to the RRE-DR, in vivo, the function of R would These ZRE are different but partially conserved (18). R expressed in EBV latently infected cells activates as many promoters as be to stabilize the binding of cellular factors to the RRE-DR, does EB1, but R does not induce the expression of EB1 (A. or to induce the expression of such cellular factors. In that case, Chevallier-Greco, personal communication). This suggests, that transcriptional activation mediated by R in vivo would be due there must be many R-responsive elements in the EBV genome, solely to the cellular factors stably bound to the RRE-DR. Therefore, it could be that R is not a transcription factor per se, and the multiplicity of these elements calls for imperfect conservation of the binding sites. and in addition to the domain required for specific DNA binding and/or stabilization of cellular protein interaction with DNA, it could be that R does not have a domain responsible for ACKNOWLEDGEMENTS transcriptional activation. We have shown here that the R N- terminal 355 amino-acids are sufficient for specific DNA binding. We wish to thank Conrad B. Bluink for editorial assistance, However, this part of the R protein did not permit transcriptional P. Jalinot for plasmid pG2, B. Verrier for plasmid p36/7polyA, activation from the RRE-DR (E. Manet, unpublished data). It H. Wolf and H. Miller for the gift of the R and Z antisera. This remains to be established if the C-terminal part of R linked to work was financially supported by the Federation Nationale des an heterologous DNA binding domain will allow transcriptional Centres de Lutte centre le Cancer (F.N.C.L.C.C), by activation from the appropriate binding sites, demonstrating that I.N.S.E.R.M. (contrat n° 871015), and by the Association pour R is a transcription factor per se. la Recherche sur le Cancer (A.R.C., n° 6810). H.G. is a recipient of an MRT fellowship. In this report we also show that the contiguous site RRE-DR 1 and RRE-DR2 act additively but not synergically at low R concentrations (Figure 6B). However we have previously REFERENCES reported that the probe BO containing the RRE-DR 1 mediated 1. Lerner, M. , Andrews, N., Miller, G. and Steitz, J. (1981) Proc. Natl. R-activation of an heterologous promoter and that probe Bl Acad. Sci. USA, 78, 805-809. containing the RRE-DR2 did not, whereas when linked to each 2. Miller, G. (1985) In B. Fields (ed.), Virology. Raven Press, Publishers, other the paired RRE-DR sites responded synergically to R (28). New York, 563-590. These previous results can now be explained by the fact that R 3. Bodescot, M., and Perricaudet, M. (1986) Nucleic Acids Res., 14, 7103-7114. binds poorly to the RRE-DR2 in probe Bl (Figure 4B) due to 4. Pfitzner, A. J., Strominger, J. L. and Speck, S. H. (1978) J. Virol., 61 , the absence of flanking sequences which probably stabilize the 2943-2946. binding. In effect, in probe BR, which contains these flanking 5. Hermessy, K. , Fennewald, S., Hummel, M., Cole, T. and Kieff., E. (1984) sequences, R binds as efficiently to the RRE-DR2 as it does to Proc Nad. Acad. Sci. USA, 81, 7207-7211 . the RRE-DR1 in probe BO (Figure 4B). Therefore, in our original 6. Laux, G. , Perricaudet, M. and Farrell, P. J. (1988) EMBO J., 7, 769-744. 7. Sample, J., Liebowitz, D. and Kiefl", E. (1989) J. Virol., 63, 933-937. report (28), the apparent synergy between the paired RRE sites 8. Zur Hauscn, H. , J. O'Neil, E. , Freese, U. K. and Hecher, E. (1978) Nature probably reflected stabilization of R binding to the RRE-DR2 (London), 272, 373-375. by flanking sequences. 9. Miller, G., Rabson, M. and Heston, L. (1984) J. Virol., 50, 174-182. 10. Countryman, J. and Miller, G. (1985) Proc. Natl. Acad. Sci. U. S. A., 82, The additive effect described above between the RRE-DR 1 and 4085-4089. the RRE-DR2 is not seen anymore at high R concentrations. This 11. Countryman, J. K., Jenson, H., Grogan, E. and Miller, G. (1986) Cancer could be due to the fact that binding of R to a single RRE, or Cells, 4, 517-523. R-induced binding of a cellular factor to a single RRE, results 12. Chevallier-Greco, A. , Manet, E., Chavrier, P., Mosnier, C , Daillie, J. in maximal stimulation of transcription. and Sergeant, A. (1986) EMBO J., 5, 3243-3249. 13. Takada, K., Shimizu, N., Sakuma, S. and Ono, Y. (1986) J. Virol., 57, The two R-binding sites characterized in this report are partially 1016-1022. homologous. Each seems to be composed of two regions 14. Chevallier-Greco, A., Manet, E., Chavrier, P., Urier, G., Buisson, M. contacted directly by the protein, and separated by 7 bp in the Daillie, J. and Sergeant, A. (1987) In D. L. Ablashi (ed), Epstein-Barr virus case of the RRE-DR 1, CCTGTGCCTTGTCCCGTGGACA- and Human diseases. The Humana Press, Clifton, N. J., 157-161. 15. Hardwick, J. M., Ueberman, P. M. and Hayward, S. D. (1988) J. Virol., ATGTCCC, and by 6bp in the case of the RRE-DR2, CAATGT- 62, 2274-2284. CCCTCCAGCGTGGTGGCTG. The binding sites are not 16. Manet, E. , Gruffat, H., Trescol-Biemont, M-C., Moreno, N., Chambard, palindromic and whether R interacts with the RRE sites as a P., Giot, J. F. and Sergeant, A. (1989) EMBO J., 8, 1819-1826. monomer or as a dimer is not as yet known. We, and others, 17. Farrell, P. J., Rowe, D. T., Rooney, M. C. and Kouzarides, T. (1989) EMBO have characterized another R-responsive enhancer in the EBV J., 8, 127-132. Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6843 18. Lieberman, P. M. and Berk, A. J. (1990) J. Virol., 64, 2560-2568. 19. Chang, Y. N., Dong, D. L. Y., Hayward, G. S. and Hayward, S. D. (1990) J. Virol., 64, 3358-3369. 20. Urier, G. , Buisson, M., Chambard, P. and Sergeant, A. (1989) EMBO J., 8, 1447-1453. 21. Remington, E. and Speck, S. H. (1990) J. Virol., 64, 1227-1232. 22. Chevallier-Greco, A. , GrufTat, H., Manet, E., Calender, A. and Sergeant, A. (1989) J. Virol., 63, 615-623. 23a. Kenney, S., Holley-Guthrie, E., Mar, E. C. and Smith M. (1989) J. Virol., 63, 3878-3883. 23b. Kenney, S., Kamine, J., Holley-Guthrie, E., Mar, E. C , Lin, J. C , Markovitz, D. and Pagano, J. (1989). J. Virol., 63: 3870-3877. 24. Buisson, M., Manet, E., Biemont, M. C , GrufTat, H., Durand, B. and Sergeant, A. (1989) J. Virol., 63, 5276-5284. 25. Cox, M. Leahy, J. and Hardwick, M. (1990) J. Virol., 64, 313-321. 26. Laux, G., Freese, U. K. and Bornkamm, G. (1985) J. Virol., 56, 987-995 . 27. Hammerschmidt, W., and Sugden, B. (1988) Cell, 55, 427-433 . 28. GrurTat, H., Moreno, N. and Sergeant, A. (1990) J. Virol., 64, 2810-2818. 29. Jalinot, P. and Kedinger, C. (1986) Nuc. Acids Res., 14, 2651-2669. 30. Graham, F. L. and Van Der Eb, A. J. (1973) Virology, 52, 456-467. 31 Dignam, J. D., Lebowitz, R. M. and Roeder, R. G. (1983) Nucl. Acids Res., 11, 1475-1489. 32 Wildeman, A. G., Sassone-Corsi, P., Grudstrom, T., Zenke, M. and Chambon, P. (1984) EMBO J., 3, 3129-3133. 33. Galas, D. and Schmitz, A. (1978) Nucl. Acids Res., 5, 3157-3170. 34. Baer, R , Bankier, A. T., Biggin, M. D., Deininger, P. L., Farrell, P. J., Gibson, T. J., Hatfull, G., Hudson, G. S., Satchwell, S.C., Seguin, C , Tuffnell, P. S. and Barrell, B. G. (1984) Nature , 310, 207-211. 35. Marshall, M., Leser, U , Seibl, R. and Wolf, H. (1989) J. Virol. 63, 938-942 . http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Nucleic Acids Research Oxford University Press

The enhancer factor R of Epstein-Barr virus (EBV) Is a sequence-specific DNA binding protein

Download PDF
9 pages

Loading next page...
 
/lp/oxford-university-press/the-enhancer-factor-r-of-epstein-barr-virus-ebv-is-a-sequence-specific-PuzyEVSQt0

References (8)

Publisher
Oxford University Press
Copyright
© 1990 Oxford University Press
ISSN
0305-1048
eISSN
1362-4962
DOI
10.1093/nar/18.23.6835
Publisher site
See Article on Publisher Site

Abstract

Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6835 The enhancer factor R of Epstein-Barr virus (EBV) Is a sequence-specific DNA binding protein Henri Gruffat, Evelyne Manet, Agnes Rigolet and Alain Sergeant* Laboratoire de Virologie Moleculaire, Ecole Normale Superieure de Lyon, UMR 49 CNRS-ENS, 46 Allee d'ltalie, 69364 Lyon Cedex 07, France Received September 20, 1990; Revised and Accepted October 26, 1990 ABSTRACT In cells latently infected with EBV, the switch from whose coding sequences are created by joining the ends of the latency to productive infection is linked to the linear virus (6; 7). expression of two EBV transcription factors called EB1 The latent EBV genome is spontaneously activated in particular (or Z) and R. EB1 is an upstream element factor which cell lines, where between 0.5% and 5% of the cells produce has partial homology to the AP1/ATF family, whereas viruses. It can also be activated by various chemical agents R is an enhancer factor. In the R-responslve enhancer including the tumor promoter 12-O-tetradecanoyl-phorbol of the replication origin only active during the EBV lytlc 13-acetate (TPA) (8). In both cases, the activation seems to be cycle (ORIIyt), R-responsive elements are located in a linked to the expression of two EBV-encoded transactivators of region of about 70 bp (RRE-DR). Here we show that R, early gene promoters, EB1 (also called Z) and R (9; 10,11; 12; produced either by In vitro translation, or present in 13; 14; 15). EB1 is encoded by the open reading frame (ORF) nuclear extracts from HeLa cells constitutively BZLF1 and is expressed from two promoters, PZ and PR, either producing R, binds directly to and protects against as a lkb monocistronic mRNA or as 3 and 4 kb mRNAs DNAase I digestion, two regions in RRE-DR. Using generated by alternative splicing and expressing both EB1 and mobility shift assay and DMS interference, we have R, the BRLF1 ORF encoded factor (Figure 1A) (16). characterized the contact-points between R and the EB1 seems to have a key role in the induction of the lytic cycle DNA. Two binding sites, RRE-DR1 and RRE-DR2, were (10; 11; 12). It is a DNA binding protein (17; 18; 19) that characterized and are contiguous in RRE-DR. R binds positively autoregulates its own promoters but also activates to these two sites probably by simultaneously transcription from quite different responsive elements including contacting two sequences within the sites, which are AP-1 binding sites (20; 21). EB1 does not seem to be a factor separated by 7 bp in RRE-DR1, cctGTGCCttgtcccGT- that can act at distances more than 100 to 200 bp from the TATA GGACaatgtccc, and by 6bp in RRE-DR2, caatGTCCC- box (20). R, however, seems to be a factor that can act at tccagcGTGGTGgctg. Direct Interaction of R with its distances over thousands of base pairs and several R targets have cognate sequences is conferred by its N-terminal 355 been identified (22; 23a; 23b; 24; 25). One is part of the amino-acids. Directed mutagenesis in RRE-DR, of either duplicated promoter DR/DL (26; 22), and overlaps with the R-binding site, impaired binding of R in vitro and, as enhancer of the EBV origins of replication active only during assayed by transient expression in HeLa cells, impaired the lytic cycle and called ORIIyt (27). R-activation by a factor of two. This suggests that RRE- ORIIyt activity is dependent on the EBV-encoded DNA DR1 and RRE-DR2 do not respond cooperatively to R. polymerase, on the BZLF1 encoded transcription factor EB1, and on the presence of the enhancer located upstream from the DR/DL TATA boxes (27) (Figure 1A). This enhancer has two INTRODUCTION functionally distinct regions, A and B. Region A is constitutively The human herpes virus EBV (Epstein-Barr virus) infects and active in all cell lines tested so far except lymphoid B cells, whereas region B was transactivated by R in all cell lines tested immortalizes peripheral B lymphocytes, resulting in the (22). One R target in the DR enhancer B region has been reduced establishment of a latent infection. In such latently infected B to 28 bp and contains the double palindromic sequence B0, CC- cells, the entire EBV genome is maintained mainly as a plasmid, TGTGCC7TG7TCCGTGG4CA4TGTCC (22). When this 28bp and its expression reduced to a few genes: those encoding two DNA fragment was placed upstream from the rabbit /3-globin small RNAs (EBERS) (1), the six Epstein-Barr Nuclear Antigens promoter, it mediated an 8 fold R-induction (28). However, the (EBNA-1, -2, -3A, -3B, 3C and LP) (For a review see ref 2), double palindrome TTGTCCCGTGGACAATGTCC either alone the BHRF1 encoded protein (3; 4) , the latent membrane protein or duplicated did not confer R-responsiveness to the /3-globin (LMP) (5) and the terminal membrane proteins (TP1 and TP2), * To whom correspondence should be addressed Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6836 Nucleic Acids Research, Vol. 18, No. 23 promoter (28). The contiguous and partially overlapping sequence The RNA obtained were used to program protein synthesis in GTCCCTCCAGCGTGGTGGCTGCC, called Bl, mediated a 3 messenger dependent rabbit reticulocyte lysates (Promega) using fold R-activation when placed upstream of the /3-globin promoter 14C-L-Leucine. (28). However, when linked to each other, BO and Bl (RRE- DNAase I footprints DR), mediated a 25 fold R-activation (28). We show in this report that R binds in the B enhancer region, Footprints experiments were made using an EBV ORDyt enhancer to sequences which fail within the RRE-DR. Moreover, we show probe derived from plasmid pG2-899/741, 5'-labelled by 32P by mobility shift assay and DMS interference, that R binds at either position 741 (probe pG2EC) or at position 899 (plasmid independently in vitro to two sites within the RRE-DR (the RRE- pG2ENC) (Figure 3). The experimental procedure is described DR 1 and the RRE-DR2). Each binding site covers about 18 bp, in ref 33. where R probably simultaneously contacts two core sequences Electrophretic Mobility Shift Assay (EMSA) separated by 6 or 7 bp. In the RRE-DR, mutations that impaired binding of R to one site or the other, only reduced the R- 2/xl of in vitro translation extract, were incubated with 2 x 104 transactivation by a factor of 2, suggesting that the RRE-DR 1 cpm of ^P-labeled-RREs (figure 1A). Incubations were carried out in 0,5 mM MgCl , 10 mM HEPES-KOH (pH 7.9), 0.5 mM and the RRE-DR2 do not respond cooperatively to R. DTT, 0.5 mM PMSF, 150 mM KC1, 10% glycerol, at 25°C for 30 minutes. The mixture was loaded onto a 4.5% MATERIALS AND METHODS polyacrylamide gel (29 to 1 crosslinked), 0.2XTBE. The R-RRE Cloned DNA templates complexes (B) were separated from the non-complexed DNA (F) by migration at 10 V/cm and visualized by autoradiography. The EB1 and R expression vectors have been described extensively elsewhere (16). Briefly, they are pUC18 derivatives DMS interferences containing the ORFs BZLF1 and BRLF1 that code respectively for the EBV transactivating factors EB1 and R, placed under the 5 x 105 cpm of the DNA probes was methylated using 1 yX of control of the SV40 early promoter-enhancer. Plasmid pG2 DMS during 3 mn. at 18°C. The methylated probe was then contains the rabbit /3-globin gene with the M13mpl2 polylinker incubated with 2/il of in vitro translation extract. After EMSA cloned 5 ' to the /3-globin promoter (29). Plasmids pG2.899/741, assay, the retarded DNA probe (B) and the non-retarded DNA pG2.899/752, pG2.899/775, pG2.899/812, pG2.737/805, were probe (F) were electroeluted and incubated in 100 y.\ of 1M made by ligating subregions of the B region of the DR enhancer piperidine for 30 mn. at 90°C. An equal amount of the radioactive 425 bp 5 ' to the /3-globin promoter (see figures 2B and 3B). All B and F probes was analysed on 8% polyacrylamide sequencing clones were verified by sequencing. Numbers after pG2 describe gels and visualized by autoradiography. the position and the orientation of the border of the inserted Cell culture and transfections enhancer subregions and they refer to the map coordinates described in Figures 1A and 2A. Plasmid pSV2/3 (28) expresses HeLa cells were grown in DMEM (Gibco) supplemented with a chimeric SV40-/3-globin RNA and was cotransfected as an 10% (v/v) fetal calf serum. The plasmids used for transfection internal control for transient expression experiments. Plasmid were prepared by the alkaline lysis method and purified through pSVO contains the SV40 Hpa II (map position 346) to Hind in two CsCl gradients. HeLa cells were seeded at 106 cells per 100 (map position 5171) fragment cloned in pUC19 digested with mm Petri dish 8 h prior to transfection. Transfections were Hind HI and BamH I. This plasmid was included in transfections performed by the calcium precipitate method (30). Cells were to keep the amount of SV40 early promoter sequences constant, mixed with the appropriate DNA(s), and the DNAs were in the since EB1 and R are expressed under the control of the SV40 same topological state as assayed by agarose gel electrophoresis. early promoter. Usually 15/tg of DNA were used per 100 mm dish including: 0,5/tg of R-expressing vector, 5/tg of reporter promoters, pSVO Production of HeLaN and HeLaR cells when required, 0,5/tg of pSV2/3 as internal control and pUC19 up to 15/tg. HeLa cells were transfected with plasmids p36/7polyA or p36/7polyAR (Figure 2A), by the calcium precipitate method RNA extraction and SI nuclease mapping (30). 20 hrs after transfection, the cells were exposed to G418 at a concentration of 800mg/L. Every three days, the cells were The cells transfected were lysed by NP40 as described elsewhere washed and placed in fresh medium plus G418. Several G418 (29). Nuclei were pelleted and RNA phenol extracted from the resistant clones were isolated, amplified and frozen. Two of them cytoplasmic fraction. 10 to 40 /tg of total cytoplasmic RNA was were selected and called HeLaR (for cells transfected with hybridized overnight at 30°C in 50% formamide, 0.3M NaCl, p36/7polyA R) and HeLaN (for cells transfected with 0.01M Tris-HCl pH 7.4 to 5'- 32P-labelled synthetic single- p36/7polyA). stranded DNA probes (Figure IB). The hybrids were digested 2 hours at 20°C with 5U of SI nuclease per 10 /xg of RNA. The HeLa cells nuclear extracts size of the SI protected DNA fragment was analysed on 8% (w/v) acrylamide/8.3 M urea gels. Quantification was made by cutting Nuclear extracts were prepared from HeLaN and HeLaR cells, the specific SI-protected bands out of the gel and counting the by the method of Dignam (31), with the modifications introduced radioactivity. The results were corrected as follows: (i) according by Wildeman et al. (32). to the efficiency of transfection as evaluated by counting the Production of R and mutant proteins in vitro radioactivity present in SI-protected probes corresponding to specific SV40 early RNA expressed from plasmid pSV2/3 and The BRLF1 ORF was cloned in plasmids pSPT18 or pSPT19 (ii) according to the activities of the different constructions in (Boehringer Mannheim). Cloned inserts within the polylinker the absence of R. region were transcribed from either the SP6 or the T7 promoters. Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6837 RESULTS |—^ft^n#*tft)pzi MH.FI The ORDyt RRE contains two regions responding to R EBI ** , „ „. ,„ U5 t. Tl Ui III U2IM US We have previously shown that the ORDyt enhancer sequences located between positions -741 to -899 (region B) upstream from the IR4 gene CAP site (Figure 1A), responded to R in a non cell-specific manner when linked to an heterologous promoter (22; 28). We show here, that 69bp within the R-responsive enhancer contains the R-responsive element. Enhancer deletion mutants placed 425 bp upstream of the rabbit /3-globin promoter in plasmid pG2 (Figure IB), were transfected in HeLa cells, and their response to increasing amounts of R-expressing vector was evaluated by quantitative SI analysis. We do not show that increasing amounts of R-expressing vector result in increasing amounts of R protein. However, since specific R-induced transcription increased in response to increasing amounts of R- expressing vector (see figure IB), we assume that this is rather due to increasing amounts of R-protein. The enhancer fragments inserted increased 2 fold the basal activity of the /3-globin promoter in plasmid pG2 (FigureIB, compare lane 1 with lanes 2 and 3), suggesting that cellular factors participate in the function of this enhancer. R highly stimulated the expression of specific /3-globin transcripts when region B (plasmid pG2-899/741), was inserted 5' to the /3-globin promoter (Fig. IB, lanes 7, 11 and I 0 tl 12 I ) T4 16 « T7 15). Deletion of sequences between bp -741 and -75 2 (plasmid pG2-899/752), decreased the R-activation effect by a factor 2, and this was seen at low R concentrations (Figure IB, lanes 8 and 12), but not at high R concentrations (Figure IB, lane 16). Therefore, the effect of the deletion can be compensated by the amount of R protein expressed. Deletion of sequences between bp -752 and -775 (plasmid pG2-899/775), impaired the induction of /3-globin transcripts by R, and this effect could not be compensated by the amount of R expressed (FigureIB, lanes Figure 1. An R-rcsponsive enhancer overlaps with ORDyt. (A) Two EBV ORFs 9, 13 and 17). However, the remaining sequences retained a low have been shown to code for transcription factors: BZLF1 codes for EBI and BRLF1 codes for R. ORTlyt overlaps with the DR promoter where four EBI but detectable R-inducibility. In conclusion, it seems that there binding sites have been mapped (ZRE1 to ZRE4). ORIlyt is composed of three are at least two R-responsive elements within or overlapping with regions, indicated by open rectangles (27). Numbers over the rectangles indicate the sequences located between positions -742 and —775. This the end-points of the regions with respect to position 52787 as + 1 on the EBV is likely to be the case, since a DNA fragment overlapping this B95-8 sequence (34). Horizontal arrows indicate palindromic se<juences. ori region and placed upstream of the /3-globin gene (plasmid indicates the origin of replication. Oriryt is overlapping with a promoter controlling the expression of a short repeted sequence called IR4. Proximal to the CAP site, pG2-737/805), rendered this promoter responsive to R (Figure there are four EB1 binding sites called ZRE1 to 4 . Distal to the CAP site, region IB, lanes 6, 10 and 14), and the level of R-activation was -63 9 to -79 4 is the ORIlyt enhancer. The sequence of the R-rcsponsive region comparable with that obtained with region B at all R in the ORIlyt enhancer and denoted domain B is shown, and has been mutated concentrations tested (compare lanes 7,11 and 15 with lanes 6, by progressive deletion. The structures of the mutants are shown (thick lines). 10 and 14). However, there could be only one R-responsive Numbers at both ends of the thick lines indicate the end points of the deletions in domain B, with respect to position 52787 as +1 on the EBV B95-8 sequence. element, and the stimulatory effect might be due to a stabilizing These mutants have been linked to the 0-globin promoter (plasmid pG2). The effect of flanking sequences (-74 1 to -752). Plasmid pSV2/3, schematic structure of SI nuclease DNA probes and the size of the SI protected expressing a chimeric SV40-/3-globin RNA under the control of DNA fragments are also presented. (B) HeLa cells were transfected with the the SV40 early promoter-enhancer sequences, was cotransfected different constructs in the conditions indicated in the top panel. Transcnptional activation was determined by quantitative SI analysis of total cellular RNA isolated as a control for transfection efficiency (see figure IB, pSV2/S from transfected cells . Plasmid pSV2/3 expresses an SV4O-/3-globin hybrid RNA internal control). An equal amount of specifically initiated RNA under the control of the SV40 early promoter-enhancer was cotransfected as an was found in every transfection, indicating that the results internal control (28). The specific start sites of 0-globin and the early-early start described above could be compared. sites of SV40 are indicated by 0 and SV respectively. HeLaR cells constitutively produce a functional R protein polypeptide recognized by a rabbit anti-R polyclonal antibody Having shown that at least two R-responsive elements are located (Figure 2B, lane 2), not detected in HeLaN cells containing only between positions -737 and -805 , we wanted to determine if the plasmid p36/7polyA (Figure 2B, lane 1). This polypeptide R binds directly to this region and where. First, we made a HeLa migrated with the same apparent molecular weight (90 Kda), as cell line (HeLaR) constitutively producing the R protein. HeLaR the R protein expressed in Raji cells treated by TPA (Figure 2B, cells contain an integrated plasmid (p36/7polyAR, Figure 2A) compare lanes 2 and 4). We transfected the rabbit /3-globin gene carrying the neomycin gene as a selection marker, and the R gene into HeLaR cells, either enhancerless (plasmid pG2), or linked (Figure 2A, ORF BRLF1), fused to the metal-inducible human to the R-responsive enhancer (plasmid pG2-737/805) (Figure 2C). metallothionein HA promoter. These HeLaR cells produce a Treatment of the HeLaR cells by ZnCl , did not lead to 2 Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6838 Nucleic Acids Research, Vol. 18, No. 23 2 and 4). An equal amount of RNA was expressed from plasmid pSV2/J cotransfected as an internal control, suggesting that the p36/7 results described above could be compared (Figure 2C, lanes 5 to 8). In conclusion, HeLaR cells produce a functional R polypeptide, and the amount of R produced can be increased by ZnCl . Wild type R protein and a C-terminal deleted mutant of R (Manet et al, unpublished data) were also produced by in vitro translation. RNAs transcribed from the SP6 promoter located upstream from the R gene cloned in plasmid pSPT19 (Figure 2A, plasmid pSPT19R) or of the deleted mutant gene cloned in the same plasmid, were used to programm protein synthesis in messenger-dependent rabbit reticulocyte lysates. In these lysates, a 90Kda 14C-Leucine-labelled polypeptide corresponding to the wild type R was detected after SDS-PAGE and autoradiogaphy (Figure 2B, lane 7) whereas a 50 Kda 14C-Leucine-labelled polypeptide was detected for the C-terminal deleted mutant. R binds to the RRE-DR in vitro Four sources of proteins, HeLaR and HeLaN cells nuclear extracts, and reticulocyte lysates containing R or BMV (Bromomosaic virus) proteins translated in vitro were used in conjunction with 5'-end-labeled double stranded DNA probes covering the RRE-DR to determine the sequence-specific DNA P62-B05/737 properties of R by DNAase I footprinting. One probe was 5'-end- labeled close to position -741 and called pG2ENC (Figure 3A), and the other probe was 5'-end-labeled on the opposite strand close to position -899 and called pG2EC (Figure 3C). Several pG2 regions of probe pG2ENC, called I, II, HI, IV and V, were tokttt e-Sl.H« protected against DNAase I digestion by a HeLaN nuclear extract Inductd tin (Figure 3B, lanes 5 and 6), as compared to the DNAase I Non Induced ISOun ZnCl digestion pattern of the probe in absence of protein extract (Figure 3B, lanes 2 to 4). Moreover, two DNAase I hypersensitive sites pSVjB called b and c, were also induced. However, the HeLaR nuclear (Intaraal control) extract strongly increased the protection of regions I ( — 741 to 9 6 7 8 -760 ) and II (-76 0 to -800) (Figure 3D, lanes 7 and 8), and induced the DNAase I hypersensitive site a. These results suggest that R is binding directly to regions I and II. They are also compatible with the idea that the HeLaR extracts contain cellular factors induced by R that bind to region I and II. The same factor sv*- might be present in HeLaN cells at lower concentrations. To ff determine further if R directly binds to these regions, DNAase I footprints were performed with extracts containing in vitro Figure 2. HeLaR cells produce a functional R polypeptide (A) Schematic translated R. In this case, only region I and a shorter segment representation of the plasmids used to generate HeLaR and HeLaN cells, and of region II were protected against DNAase I digestion (Figure to produce R by in vitro translation. (B) HeLaR cells but not HeLaN cells induced 3B, lanes 16 and 17). As such a footprint is not observed with by ZnCl , produce a 90 Kda polypeptide immuno-related to R produced in Raji cells where the EBV early gene products EBl and R have been induced by TPA extracts containing proteins translated in vitro from BMV RNA (lanes 1 to 4). The proteins were revealed b y western Wot using two rabbits antisera, in the same conditions (Figure 3B, lanes 13 to 15), the DNAase one ami R (35) and one ami EBl, mixed together. R and a truncated version I footprint obtained with R in vitro translated can be considered of R (RD97st) were also produced by in vitro translation and separated on PA- as specific. As shown in Figure 3A, regions I and II overlap the GE (lanes 5 to 8). (C) To determine if the 90 Kda polypeptide produced by HeLaR sequences located between positions -737 and -775, which cells was transactivating the RRE-DR linked to the 0-globin promoter, HeLaR cells were transfected with plasmid pG2-805/737 as indicated on the top panel. are those that mediate R-activation of the ORIlyt enhancer (see The level of correctly initiated /3-globin RNA (fi) was determined by SI nuclease Figure 1). analysis (lanes 1 to 4). The amount of RNA expressed from pSV2)3 (SV) cotransfected as an internal control, is presented in lanes 5 to 8, and corresponds DNAase I footprints were also performed on probe pG2EC to the RNA samples presented in lanes 1 to 4. (Figure 3C). No clear footprint was detected on this strand with either the HeLaN or the HeLaR nuclear extract (Figure 3D, lanes 1 to 9). However, a strong DNAase I hypersensitive site called an increased expression of specifically initiated |3-globin mRNA a, was detected with the HeLaR nuclear extract (Figure 3D, lanes from plasmid pG2 (Figure 2C, lanes 1 and 3). However, in 7 to 9). This strong hypersensitive site was also induced by the in vitro translated R protein, but this time there was a clear HeLaR cells, /3-globin RNA levels were increased when the R- footprint overlapping the sequences located between positions responsive enhancer was inserted in front of the /3-globin promoter (Figure 2C, compare lanes 1 and 2). Moreover, this -73 7 and -77 5 (Figure 3D, lanes 14 and 15). As previously, R-enhanced /3-globin RNA transcription was further increased such a footprint is not seen with extracts containing proteins in by treatment of the cells by ZnC12 (Figure 2C, compare lanes vitro translated from BMV RNA (Figure 3D, lanes 12 and 13). Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6839 la vttr* tranalitM R HaiaB Htlii r aitrKt HH M MClaar <itr*cl ia vitro truslatt d R 31ca I I -COAUTCGACBSAACASeSCACCTIlTTACWEMGCTCECACCACCBAC -i- -741 -737 -775 S' ICO •l*-GCTC6*GCTGCCXrGTCCC6T66ACAA|6TCCCTCCASCGTG^TCeCTQ -74 1 -751 -775 CCTTTGGGATGCATC ACTTT 6 A DCC AC T AAGC C C CC 6T T GCT C G CCT T GC CT 6C C T G6AAACCCTACSTA8TBAAACTCG6TBATTCOSSe6CAACGAGCGGAACG6ACBBA -ooa -a'u iy -OO9 c -012 V CACCAT6ACACACTAAGCCCCT6CTAATCCAT6AGCCCCGCCTTTAG6AAGCACC STHTMTSTeTMTTCtG6«ACSArrAGCTACTCE£ECCECAAATCCTTCSTGCTB CASSfCCCTWCAWTCT - HIMIII* ACGTCCCGGGATCCTCTAGA-Hlndlll -m In vttr* trwulattos ff»cl««r Cntr*ct !• vitro trHnUtion Htt«JI HtiiR 1 2 3 4 S 6 7 6 •• * «!;:; • Figure 3. R binds to the RRE-DR. (A) and (C) Schematic representation of the sequences protected against DNAase I digestion by the different protein extracts used in the footprint experiment. Horizontal braids indicate footprints induced by in vitro translated R. Horizontal closed bars indicate footprints induced by HeLaR extracts. Open bars indicate footprints induced by HeLaN extracts. Vertical arrows and letters indicate DNAase I hypersensitive sites. (B) and (D) DNAase I footprints with decreasing amounts of DNAase I, on both strands of a DNA probe located between positions -741 to -899 . The protein extracts used are indicated in the top panel. Vertical bars and numbers indicate footprints. Horizontal arrows and letters indicate DNAase I hypersensitive sites. Closed triangles indicate decreasing amounts of DNAase I. Thus, in conclusion, R seems to bind directly to the sequences protein in conjunction with three synthetic double-stranded (RRE-DR) mediating the R-activation in a transient expression oligonucleotides overlapping regions I and II, and called BO, Bl and BR (Figure 4A). We also made use of a fourth double assay, as defined by deletion mutagenesis (Figure IB). stranded oligonucleotide called B2 (Figure 4A), which The R DNA-binding domain is located in the N-terminal 355 corresponds to a region protected by HeLaR extracts but not by amino-acids in vitro translated R (-78 7 to -807). Probe B2 should not therefore bind in vitro translated R. As judged by mobility shift As discussed above, R seems to bind in vitro to a region assays and competitions, specific binding of R was observed to overlapping positions -741 and —780 in the RRE-DR, and probes BO, Bl and BR, but not to probe B2. The in vitro possibly to two sites within region I and n . Interestingly, deletion mutagenesis in this region also showed a two-step decrease in translated R bound with a higher affinity to BO (Figure 4B, lane the R-responsiveness, corresponding approximately to regions 2) and to BR (Figure 4B, lane 14), than to Bl (Figure 4B, lane I and II. To localize further where R bound in the RRE-DR, 8). This was also seen when a 400 fold molar excess of unlabeled we made mobility shift assays using the in vitro translated R B0, Bl and BR probes were used as competitors. Probe B0 Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6840 Nucleic Acids Research, Vol. 18, No. 23 80 S I I '• -Induced footprint I RRE CCTGTGCCT1GTCCCGTGGACAATGTCCCTCCAGCGTGGTGGCTGCCTTT66GATGCATCACTTT6A6CC BO CCTGTGCCTTGTCCC6T66ACAATGTCCC B l GTCCCTCCAGCGTG6T66CT6CC B2 GGGATGCATCACTTT6A6CC BR TTGTCCCGTGGACAATGTCCCTCCA6CGT6GT6GCT6 probes -80- \- -B0- H h H proteins / L | R- ' L r- _|L | KD97 |L -R- competitors / / / BO Bl B2 / / / BR BO B2 / BO B2 / / BO B2 / 12 13 M E » 17 6 7 8 9 10 II 1 2 3 4 5 6 7 8 Figure 4. Mapping of R binding sites by gel retardation assay and DNA competition (A) Schematic representation of DNA probes used in the binding assays and competitions. (B) Probes BO, Bl and BR were incubated with protein extracts and with DNA competitors as indicated in the top panel. (C) In vitro translated R and RD97st proteins were incubated with probe BO and DNA competitors as indicated in the top panel. In (B) and (C), /, indicates no protein added; L, indicates lysate translated without exogenous RNA; DNA competitors were added at a mass ratio of 100; Free (F) and bound (B) DNA probe are indicated on the left of the autoradiographies. competed efficiently for the binding of R on probes BO (Figure 5'-GTGGAC-3' (Figure 5B). Such strong interferences suggest 4B, lane 3), Bl (Figue 4B, lane 9) and BR (Figure 4B, lane 12). that R contacts simultaneously both sequences rather than binding to them separately. Similar results were observed with probe BR, However, probe Bl competed less efficiently than probe BO for where strong interferences were detected on guanine residues in the binding of R on probe BO (Figure 4B, lane 4). As expected no R binding to the non-specific probe B2 could be detected (not the sequences 5'-GTCCC-3' and 5'-CGTGGGTG-3' (Figure shown), and unlabeled B2 probe did not detectably compete for 5C). Again, such strong interferences suggest that R contacts the the binding of R on labeled probes BO (Figure 4B, lane 5), Bl two sequences simultaneously. This was strengthened by the fact that there was no interference on the sequence 5'-GTGGAC-3' (Figure 4B, lane 11), and BR (Figure 4B, lane 17). The truncated present alone in probe BR, whereas interferences on this sequence R protein (RD97st) generated by in vitro-transcription/translation of the C-terminal deleted BRLF1 mutant construct (Figure 2B, were observed in probe BO (see Figure 5B). Similarly, there was lane 8), still bound to probe BO. RD97st contains the N-terminal no interference on the sequence 5'-GTCCC-3' located between 355 amino-acids of R, and has a higher affinity for probe BO positions —760 and -76 5 and present alone in probe BO (Figure 5B), whereas interferences were found on this sequence in probe than the wild type R protein (Figure 4C, compare lanes 1 and BR (Figure 5C). Moreover, on both the coding and the noncoding 5). Again, the interaction was specific since binding of RD97st to probe BO was competed out by an excess of unlabelled probe strands of a double stranded oligonucleotide covering the entire BO (Figure 4C, lane 6), but not by an excess of probe B2 (Figure RRE-DR, interferences were observed on the four sequences 4C, lane 7). Similar results were obtained with probes Bl and described above but with a much weaker intensity (not shown). The results of the methylation interferences are summarized in BR (not shown). These results suggest that there are at least two Figure 5A, and suggest that R binds to two contiguous sites in R binding sites in the RRE-DR. Furthermore, specific binding to these sites is still observed when the R protein is reduced to the RRE-DR. Each site probably consists of two regions directly the N-terminal 355 amino-acids. contacted by R and separated by 6 or 7 bp, 5'-GTGCCTTGT- CCCGTGGAC-3' and 5-GTCCCTCCAGCGTGGTG-3', called RRE-DR 1 and RRE-DR2 respectively. Similar results have R binds to two contiguous sites in the RRE-DR been obtained with the R deletion mutant RD97st (not shown). Since in vitro translated R bound to probes BO and BR, the contact points of R with these double-stranded oligonucleotides were The RRE-DR1 and the RRE-DR2 do not respond probed chemically by modification of the DNA with dimethyl cooperatively to R sulfate (methylation interference analysis). After partial methylation at purine residues, the modified probes were used In order to further evaluate the contribution of each R-binding in the gel retardation assay. Free (F) and bound (B) probe DNA site in the R-induced activity of the RRE-DR, we mutated the was recovered, subjected to chemical cleavage at methylation RRE-DR1 (Figure 6A, plasmid pG2M2), the RRE-DR2 (Figure sites, and the products were resolved by gel electrophoresis under 6A, plasmid pG2M3) or both (Figure 6A, plasmid pG2M4). denaturing conditions. On both the coding (C) and the non-coding These mutations impaired totally the binding of R on the isolated (N Q strands of probe BO, strong interferences were detected on RRE-DR 1 or RRE-DR2 in vitro (not shown). Each mutant and guanine residues in the sequences 5'-GTGCC-3' and the wild type RRE were placed upstream of the /3-globin promoter Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6841 -737 -74S -7«« -Tit TCTCCCTCCMCtTWTCSCTaXTTTCSUTCCATUCTTTSMCC »62RRE CCrCTGCCTTGTCCCGTCCACAItrcTCCtrCCMCtTeSTBSCTBCCTTTCCCATGCATCACTTTCMCC i R-lnducad footprint prob t puC I B 0 0 *•« -7?7.. ._ - r C 5-<Mcrc<iCTKCCCCTCTCCCTTCTCCCSTSS*CMTCTCCCCaG«TCCTCTMMCTCGU »C6AATTCCC666TTCTA6A6CTC 6 prob« »62 BR , EcoRi Sm«l xtai SK I Xhoi 45 7i0 Xh» I I , E» ftl rabbit B-globin C S'-CTCSMTTSTCCCCTCC*CMT{TCCtTCCACcSTSTGSCTSWn 1*fTTr HIWntTtTI>fTTr r Tnrny|*j|j'"'g''"' r "—**""' • f «C e 50»« B C lp62RREn2 t13 rul l pG2 RRE H2 113 rwl 1*62 RRE M2 H3 tw l prooa pUCIBBO pro»« p62BR fold induction I 17 8 7 13 2 32 19 21 6 C MC C He I 2 3 4 5 6 7 B 9 10 II 12 13 14 IS G*A F B S.A F B 6»A F B S'A F B PSV2B (liit«m«l cmtr*l> I 2 3 4 5 6 7 B 9 10 II 12 13 14 IS Figure 6. The paired non-homologous RRE sites in the RRE-DR do not respond G • • <•. synergkally to R. (A) Schematic representation of the /3-globin gene constructions in which the RRE-DR (pG2RRE), or mutants in the RRE-DR1 (pG2M2) in the RRE-DR2 (pG2M3), and in both RREs (pG2M4), have been inserted. (B) The various constructs were transfected in HeLa cells as indicated in the top panel, and the level of correctly initiated /3-globin transcripts(/3) was evaluated by quantitative SI nuclease assay. Fold induction represents the ratio of the radioactivity counted in each SI-protected band obtained in presence of R on the radioactivity counted in the corresponding SI-protected band obtained in absence of R. This has been done for the two R concentrations. ( Q SI nuclease analysis of RNA initiated from plasmid pSV2j3 (SV) cotransfected as an internal control. RNA samples 1 to 15 are the same as in (B). saturating (lanes 8 and 9) than saturating amounts of R Qanes 13 and 14). Finally, when both the RRE-DR1 and the RRE-DR2 were mutated (mutant M4), the R-induced activation was strongly Figure 5. Fine mapping of R binding sites by methylation interference (A) impaired, and this was seen at 500 ng as well as 50 ng of R- Sequences of the R-Responsive Enhancer (RRE) and of double stranded DNA expression vector (lanes 10 and 15). These results suggest that probes used for formation of protein-DNA complexes (pUC18BO and pG2BR) the RRE-DR1 and the RRE-DR2 do not respond cooperatively The R binding site is indicated by an open bar under RRE. The G residues involved to R. An equal amount of RNA was expressed from plasmid in the formation of specific nucleoprotein complexes on the probes are shown by black squares. DNA binding assays were performed under standard conditions pSV2/S (Figure 6C, lanes 1 to 15), indicating that the results except that probes pUC 18 BO and pG2 BR (5'-end-labeled either on the coding presented above could be compared. However, it should be noted (C) or the noncoding (NC) strand) were treated with dimethyl sulfate before that the /3-globin promoter, which lacks R-binding sites, is weakly incubation with the in vitro translated protein R. (B) and (C) G+ A is a Maxam stimulated by R at high concentrations of the factor, suggesting and Gilbert sequence ladder. F, is the non-retarded probe, and B is the retarded that R present at high concentrations can activate by binding non complex. Black squares indicate the G residues whose methylation strongly prevents the formation of complexes, open squares those which are weakly involved specifically to the /3-globin promoter. Moreover, /3-globin specific transcription from mutant M4 was still weakly activated by high concentrations of R (Figure 6B, compare lanes 2 and 6), suggesting that the mutations did not impair totally R binding in plasmid pG2 (Figure 6A). These reporter plasmids were in vivo. transfected into HeLa cells either alone or along with various amounts of R-expression vector, and their transcriptional activity was evaluated by quantitative SI analysis. Representative results DISCUSSION are shown in Figure 6B. As expected, the RRE-DR located upstream of the /3-globin promoter resulted in a strong In this report, we provide the first evidence that the N-terminal transcriptional induction dependent on the presence of R (lane 355 amino-acids of the EBV-encoded transcription factor R, 7), and this induction reached a plateau at 500ng of R-expression confer sequence-specific DNA binding properties to this protein vector (lane 12). The three base-pairs change in the RRE-DR1 in vitro. (mutant M2, lane 8), or in the RRE-DR2 (mutant M3, lane 9), Furthermore, using DNAase I footprints we have also shown resulted in a reduction of about 50% of the R-induced activation that HeLa cell proteins bound to regions I and II in the RRE- of the /3-globin promoter, and this was more obvious at non- DR, where two R-binding sites have also been identified. The Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 6842 Nucleic Acids Research, Vol. 18, No. 23 protection against DNAase I digestion of regions I and II was early promoter PM, controlling the expression of a factor acting at the post-transcriptional level (23b; 24). Bidirectional deletion much more pronounced with HeLa cell extracts containing R than mutagenesis and DNAase I footprints indicate that there could with HeLa cell protein extracts. At present, we do not know if be four R-binding sites, with sequences only partially related to this reflects the induction by R of such cellular factor, or if there the one described in this publication (H. Gruffat and M. Buisson, is cooperativity between R and these cellular factors, or if R competes with these cellular factors for binding on the RRE-DR. unpublished results). Therefore, as in the case of the EBV- encoded upstream element factor EB1 (18), the EBV enhancer Whether these cellular factors are contacting DNA at the same factor R could bind to many sites which are imperfectly sites as R, and if such proteins contribute to the R-activation of homologous and rather degenerated. This possibility is interesting the enhancer is also at present unknown. However, mutations if one considers the regulation of EBV early genes expression. which affect the purine residues contacted by R, impaired R- binding in vitro and also impaired the R-induced transcriptional Two transcription factors, EB1 and R, seem to be involved in the activation of EBV early genes expression. EB1 expressed in activation in a transient expression assay (Figure 6B). This EBV latently infected B cells, activates the expression of R and suggests that R-mediated transcriptional activation requires direct of many EBV early promoters (12; 22). Many EB1 binding sites interaction between R and its cognate binding sites. Nevertheless, but fewer EB1 responsive elements (ZRE) have been identified. one cannot rule out the possibility that although R can directly bind in vitro to the RRE-DR, in vivo, the function of R would These ZRE are different but partially conserved (18). R expressed in EBV latently infected cells activates as many promoters as be to stabilize the binding of cellular factors to the RRE-DR, does EB1, but R does not induce the expression of EB1 (A. or to induce the expression of such cellular factors. In that case, Chevallier-Greco, personal communication). This suggests, that transcriptional activation mediated by R in vivo would be due there must be many R-responsive elements in the EBV genome, solely to the cellular factors stably bound to the RRE-DR. Therefore, it could be that R is not a transcription factor per se, and the multiplicity of these elements calls for imperfect conservation of the binding sites. and in addition to the domain required for specific DNA binding and/or stabilization of cellular protein interaction with DNA, it could be that R does not have a domain responsible for ACKNOWLEDGEMENTS transcriptional activation. We have shown here that the R N- terminal 355 amino-acids are sufficient for specific DNA binding. We wish to thank Conrad B. Bluink for editorial assistance, However, this part of the R protein did not permit transcriptional P. Jalinot for plasmid pG2, B. Verrier for plasmid p36/7polyA, activation from the RRE-DR (E. Manet, unpublished data). It H. Wolf and H. Miller for the gift of the R and Z antisera. This remains to be established if the C-terminal part of R linked to work was financially supported by the Federation Nationale des an heterologous DNA binding domain will allow transcriptional Centres de Lutte centre le Cancer (F.N.C.L.C.C), by activation from the appropriate binding sites, demonstrating that I.N.S.E.R.M. (contrat n° 871015), and by the Association pour R is a transcription factor per se. la Recherche sur le Cancer (A.R.C., n° 6810). H.G. is a recipient of an MRT fellowship. In this report we also show that the contiguous site RRE-DR 1 and RRE-DR2 act additively but not synergically at low R concentrations (Figure 6B). However we have previously REFERENCES reported that the probe BO containing the RRE-DR 1 mediated 1. Lerner, M. , Andrews, N., Miller, G. and Steitz, J. (1981) Proc. Natl. R-activation of an heterologous promoter and that probe Bl Acad. Sci. USA, 78, 805-809. containing the RRE-DR2 did not, whereas when linked to each 2. Miller, G. (1985) In B. Fields (ed.), Virology. Raven Press, Publishers, other the paired RRE-DR sites responded synergically to R (28). New York, 563-590. These previous results can now be explained by the fact that R 3. Bodescot, M., and Perricaudet, M. (1986) Nucleic Acids Res., 14, 7103-7114. binds poorly to the RRE-DR2 in probe Bl (Figure 4B) due to 4. Pfitzner, A. J., Strominger, J. L. and Speck, S. H. (1978) J. Virol., 61 , the absence of flanking sequences which probably stabilize the 2943-2946. binding. In effect, in probe BR, which contains these flanking 5. Hermessy, K. , Fennewald, S., Hummel, M., Cole, T. and Kieff., E. (1984) sequences, R binds as efficiently to the RRE-DR2 as it does to Proc Nad. Acad. Sci. USA, 81, 7207-7211 . the RRE-DR1 in probe BO (Figure 4B). Therefore, in our original 6. Laux, G. , Perricaudet, M. and Farrell, P. J. (1988) EMBO J., 7, 769-744. 7. Sample, J., Liebowitz, D. and Kiefl", E. (1989) J. Virol., 63, 933-937. report (28), the apparent synergy between the paired RRE sites 8. Zur Hauscn, H. , J. O'Neil, E. , Freese, U. K. and Hecher, E. (1978) Nature probably reflected stabilization of R binding to the RRE-DR2 (London), 272, 373-375. by flanking sequences. 9. Miller, G., Rabson, M. and Heston, L. (1984) J. Virol., 50, 174-182. 10. Countryman, J. and Miller, G. (1985) Proc. Natl. Acad. Sci. U. S. A., 82, The additive effect described above between the RRE-DR 1 and 4085-4089. the RRE-DR2 is not seen anymore at high R concentrations. This 11. Countryman, J. K., Jenson, H., Grogan, E. and Miller, G. (1986) Cancer could be due to the fact that binding of R to a single RRE, or Cells, 4, 517-523. R-induced binding of a cellular factor to a single RRE, results 12. Chevallier-Greco, A. , Manet, E., Chavrier, P., Mosnier, C , Daillie, J. in maximal stimulation of transcription. and Sergeant, A. (1986) EMBO J., 5, 3243-3249. 13. Takada, K., Shimizu, N., Sakuma, S. and Ono, Y. (1986) J. Virol., 57, The two R-binding sites characterized in this report are partially 1016-1022. homologous. Each seems to be composed of two regions 14. Chevallier-Greco, A., Manet, E., Chavrier, P., Urier, G., Buisson, M. contacted directly by the protein, and separated by 7 bp in the Daillie, J. and Sergeant, A. (1987) In D. L. Ablashi (ed), Epstein-Barr virus case of the RRE-DR 1, CCTGTGCCTTGTCCCGTGGACA- and Human diseases. The Humana Press, Clifton, N. J., 157-161. 15. Hardwick, J. M., Ueberman, P. M. and Hayward, S. D. (1988) J. Virol., ATGTCCC, and by 6bp in the case of the RRE-DR2, CAATGT- 62, 2274-2284. CCCTCCAGCGTGGTGGCTG. The binding sites are not 16. Manet, E. , Gruffat, H., Trescol-Biemont, M-C., Moreno, N., Chambard, palindromic and whether R interacts with the RRE sites as a P., Giot, J. F. and Sergeant, A. (1989) EMBO J., 8, 1819-1826. monomer or as a dimer is not as yet known. We, and others, 17. Farrell, P. J., Rowe, D. T., Rooney, M. C. and Kouzarides, T. (1989) EMBO have characterized another R-responsive enhancer in the EBV J., 8, 127-132. Downloaded from https://academic.oup.com/nar/article/18/23/6835/2388765 by DeepDyve user on 14 August 2020 Nucleic Acids Research, Vol. 18, No. 23 6843 18. Lieberman, P. M. and Berk, A. J. (1990) J. Virol., 64, 2560-2568. 19. Chang, Y. N., Dong, D. L. Y., Hayward, G. S. and Hayward, S. D. (1990) J. Virol., 64, 3358-3369. 20. Urier, G. , Buisson, M., Chambard, P. and Sergeant, A. (1989) EMBO J., 8, 1447-1453. 21. Remington, E. and Speck, S. H. (1990) J. Virol., 64, 1227-1232. 22. Chevallier-Greco, A. , GrufTat, H., Manet, E., Calender, A. and Sergeant, A. (1989) J. Virol., 63, 615-623. 23a. Kenney, S., Holley-Guthrie, E., Mar, E. C. and Smith M. (1989) J. Virol., 63, 3878-3883. 23b. Kenney, S., Kamine, J., Holley-Guthrie, E., Mar, E. C , Lin, J. C , Markovitz, D. and Pagano, J. (1989). J. Virol., 63: 3870-3877. 24. Buisson, M., Manet, E., Biemont, M. C , GrufTat, H., Durand, B. and Sergeant, A. (1989) J. Virol., 63, 5276-5284. 25. Cox, M. Leahy, J. and Hardwick, M. (1990) J. Virol., 64, 313-321. 26. Laux, G., Freese, U. K. and Bornkamm, G. (1985) J. Virol., 56, 987-995 . 27. Hammerschmidt, W., and Sugden, B. (1988) Cell, 55, 427-433 . 28. GrurTat, H., Moreno, N. and Sergeant, A. (1990) J. Virol., 64, 2810-2818. 29. Jalinot, P. and Kedinger, C. (1986) Nuc. Acids Res., 14, 2651-2669. 30. Graham, F. L. and Van Der Eb, A. J. (1973) Virology, 52, 456-467. 31 Dignam, J. D., Lebowitz, R. M. and Roeder, R. G. (1983) Nucl. Acids Res., 11, 1475-1489. 32 Wildeman, A. G., Sassone-Corsi, P., Grudstrom, T., Zenke, M. and Chambon, P. (1984) EMBO J., 3, 3129-3133. 33. Galas, D. and Schmitz, A. (1978) Nucl. Acids Res., 5, 3157-3170. 34. Baer, R , Bankier, A. T., Biggin, M. D., Deininger, P. L., Farrell, P. J., Gibson, T. J., Hatfull, G., Hudson, G. S., Satchwell, S.C., Seguin, C , Tuffnell, P. S. and Barrell, B. G. (1984) Nature , 310, 207-211. 35. Marshall, M., Leser, U , Seibl, R. and Wolf, H. (1989) J. Virol. 63, 938-942 .

Journal

Nucleic Acids Research – Oxford University Press

Published: Dec 1, 1990

There are no references for this article.