Skip Navigation

This Article
Right arrow Abstract Freely available
Right arrow Print PDF (249K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (12)
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Portes-Sentis, S.
Right arrow Articles by Gruffat, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Portes-Sentis, S.
Right arrow Articles by Gruffat, H.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 1995 Oxford University Press 1347-1354

Footnote

A particular DNA structure is required for the function of a cis -acting component of the Epstein-Barr virus OriLyt origin of replication

A particular DNA structure is required for the function of a cis -acting component of the Epstein-Barr virus OriLyt origin of replication Stéphanie Portes-Sentis , Alain Sergeant and Henri Gruffat*

Laboratoire de Virologie Humaine, U412 INSERM, ENS-Lyon, 46 Allée d'Italie, F-69364 Lyon cedex 07, France

Received December 13, 1996; Revised and Accepted February 3, 1997

ABSTRACT

OriLyt, the cis -acting element of Epstein-Barr virus lytic origin of replication, consists of upstream and downstream components. The upstream component plays a dual role in transcription and replication. The downstream component contains a homopurine-homopyrimidine sequence which forms an H palindrome. We show that the downstream component can adopt a triple helix structure in vitro , that the 5 ' border of the homopyrimidine sequence is sensitive to P1 nuclease when carried by a supercoiled plasmid and that an oligonucleotide complementary to the homopyrimidine strand is taken up by a plasmid carrying the OriLyt H palindrome. We also show that all mutations which alter the H palindrome impair both oligonucleotide uptake and OriLyt-dependent replication. Interestingly, compensatory mutations which restore an H palindrome also restore oligonucleotide uptake by the mutated plasmids and their OriLyt-dependent replication. Thus, there is a strong correlation between the inability of the OriLyt H palindrome to form a non-B-DNA structure in vitro and impairment of OriLyt-dependent replication. This suggests that the presence of a non-B-DNA structure in the OriLyt downstream component is required for OriLyt-dependent replication.

INTRODUCTION

Epstein-Barr virus (EBV) is a human gammaherpesvirus which is associated with different malignancies, including Burkitt's lymphoma and nasopharyngeal carcinoma (for a review see 1 ). In vivo it is supposed that EBV infection is latent in B lymphocytes but productive in epithelial cells within the oropharynx ( 2 ). In vitro EBV infects and immortalizes human B lymphocytes. In such immortalized B cells the viral genome is maintained as an extrachromosomal, nuclear multiple copy episome ( 3 ), from which 11 genes are expressed, defined as type III latency. The products of these viral genes are implicated in B cell immortalization and in EBV genome persistance. The cis -acting elements which mediate DNA replication during type III latency have been identified and called Ori-P (for plasmid origin of replication). Viral DNA replication occurs once per cell cycle ( 4 ), proceeds bidirectionally from Ori-P ( 5 ) and is dependent both on cellular proteins and on the viral protein EBNA1 ( 6 , 7 ).

In vitro EBV production occurs spontaneously in a few infected B cells and can be induced in many cells by treatment with agents such as 12- O -tetradecanoylphorbol-13-acetate associated with butyrate or by cross-linking of surface immunoglobulin ( 8 , 9 ). The productive cycle is characterized by the sequential activation of viral gene expression, which begins with expression of the viral transcription factors EB1 (also called Z, ZEBRA or Zta) and R (also called Rta) ( 10 - 14 ). Once made, R and EB1 induce expression of the EBV early genes, whose products are implicated in amplification of the viral genome via an origin of replication different from Ori-P, called OriLyt (Fig. 1 ) ( 15 ). A co-transfection-replication assay has demonstrated that seven EBV proteins are required for OriLyt-dependent replication: BALF5 (DNA polymerase), BMRF1 (polymerase processivity factor), BALF2 (single-stranded DNA binding protein), BSLF1 (primase), BBLF4 (helicase), BBLF2/3 (helicase/primase-associated protein) and EB1 ( 16 , 17 ). Several non-essential proteins with enzymatic activities involved in the biochemical pathways of nucleotide synthesis and phosphorylation are also encoded by the virus. These replication proteins, with the exception of EB1, have homology with known herpes simplex virus (HSV) and cytomegalovirus (CMV) replication proteins ( 18 , 19 ). One factor that has not been identified in EBV is the origin binding protein, the equivalent of HSV UL9 protein. Although the viral trans -acting factors and the cis -acting elements have been identified, the mechanisms which underlie herpesvirus replication are not yet understood.

The EBV OriLyt overlaps with the divergent promoters of the BHLF1 and BHRF1 early genes (Fig. 1 ). OriLyt consists of two essential and several auxiliary elements which are required for full activity. Auxiliary elements which are only poorly defined are non-essential but influence the efficiency by which OriLyt- containing plasmids replicate in transient replication assays ( 15 ). The two essential or core elements, called respectively the upstream and the downstream components, constitute the minimal origin of DNA replication (Fig. 1 ) ( 20 ).

The upstream OriLyt component is localized within the promoter region of the BHLF1 gene, which contains four Z-responsive elements (ZREs). Mutation of these sites completely abolishes OriLyt-dependent replication, indicating that EB1 binding to these sites is required to mediate OriLyt-dependent replication ( 21 ). In effect, exchange of the OriLyt ZREs with other activator binding sites does not rescue replication, even if transcription of the BHLF1 gene is unaffected. The EB1 domain implicated in activation of replication seems to co-localize with the transcriptional activation domain ( 22 ).

The downstream OriLyt component is located 440 bp from the upstream component and is 90 bp long. In contrast to the upstream component, the downstream component is only dedicated to activation of OriLyt-mediated DNA replication and does not appear to be involved in transcriptional activation of the BHLF1 and BHRF1 genes. Several cellular proteins, including SP1, bind specifically to this element in vitro , but they are probably not implicated in OriLyt-dependent replication ( 23 ).

The contribution of the downstream component in OriLyt- dependent replication is as yet unknown. It contains a pyrimidine-rich sequence which can be divided into two parts (Fig. 2 A). The first part represents a homopurine-homopyrimidine sequence, made up of a mirror repeat, also called an H palindrome, in which any introduced mutation impairs OriLyt-dependent replication in a transient replication assay ( 20 , 23 ). The second part, located 3' of the homopurine-homopyrimidine sequence is C rich. We have focused our attention on the homopurine-homopyrimidine sequence element and we show that in vitro it can form a triple helix. We also present experimental evidence demonstrating that every mutation in the H palindrome that impairs formation of the triple helix in vitro also impairs OriLyt-dependent replication in a transient replication assay and that mutations which do not dramatically alter folding of the H palindrome in vitro allow some replication of plasmids carrying these mutations. Thus, OriLyt-dependent replication of the EBV genome appears to require that the homopurine-homopyrimidine sequence of the OriLyt downstream component adopts a non-B-DNA structure. It also appears probable that a specific, as yet unknown, trans -acting factor contributes to activation of this element.

MATERIALS AND METHODS

Cell lines

D98HR1 cells were derived from a somatic cell hybrid made between the EBV genome-positive Burkitt lymphoma cell line P3HR1 and the human epithelial cell line D98 ( 24 ). This adherent cell line was maintained in Dulbecco's modified Eagle's medium containing 10% fetal calf serum.

Plasmids and oligonucleotides

Plasmid DNA was prepared by alkali lysis and supercoiled DNA molecules were purified by ethidium bromide/cesium chloride centrifugation. Plasmid p968.22, which carries the complete OriLyt element (EBV B95-8 nucleotide coordinates 48848- 56084; 25 ) has been described in detail elsewhere ( 20 ). Oligonucleotide-directed mutations were introduced into this plasmid as described elsewhere ( 20 ). The mutations relevant to this study are shown in the figures, which include sequence information on specific mutations. Plasmid p1562 contains the OriLyt downstream element (EBV B95-8 nucleotide coordinates 53337-53428; 25 ) cloned in the pBluescript SK- (Stratagene) vector.

The different single-stranded oligonucleotides were radiolabeled with 32 P using T4 polynucleotide kinase and then purified on a 10% polyacylamide gel according to standard procedures. Cstr refers to the pyrimidine-rich DNA strand of the OriLyt downstream element. Gstr refers to the other purine-rich DNA strand. The different oligonucleotides used in the oligonucleotide uptake assays have been previously described, as have the mutant sequences and wild-type plasmids containing the OriLyt downstream element ( 23 ). The M13 sequencing primers -40 and reverse sequencing primer were purchased from Pharmacia LKB.

P1 nuclease digestion

Supercoiled plasmid pBSK or p1562 was incubated with different amounts (between 1 * 10 -4 mg/ml and 1 * 10 -2 mg/ml) of P1 nuclease (Boehringer) in 50 [mu]l P buffer (10 mM Tris-HCl, pH 7.6, 10 mM MgCl 2 , 50 mM NaCl) for 10 min at 37oC. The reaction was stopped by addition of EDTA (37 mM final concentration). The DNA was purified by phenol/chloroform extraction and digested with Pvu II. The resulting DNA fragments were separated on 1% agarose gels. Gels were stained with ethidium bromide and the DNA fragments visualized with UV light.

High resolution mapping of P1 nuclease sensitivity

The P1-treated DNAs were incubated in 17 [mu]l buffer H (25 mM Tris-HCl, pH 7.2, 5 mM MgCl 2 ) with 2 * 10 5 c.p.m. of a 5'-end-labeled primer for 4 min at 90oC and then for 20 min at 45oC. The primer extension reaction was performed on the C-rich strand using the reverse sequencing primer and on the G-rich strand using the -40 primer. After addition of 5 mM dithiothreitol and 4 mM dNTP, 5 U Klenow enzyme were added and the DNA was allowed to polymerize for 10 min at 37oC. The reaction was stopped by addition of 10 [mu]l of a solution containing 20 mM EDTA and 0.5 M NH 4 Ac. The primer-extended products were resolved on a 6% polyacrylamide-urea denaturing gel.

Oligonucleotide uptake by supercoiled plasmid

Aliquots of 200 ng supercoiled (S) or linearized (L) plasmid DNA were incubated for 2 h at 25oC in 10 [mu]l Hyb. buffer (10 mM Tris-HCl, pH 7.5, 50 mM sodium acetate, 2.5 mM MgCl 2 ) containing 1 * 10 4 c.p.m. radiolabeled oligonucleotide. The reaction mixtures were then analyzed on a 0.8% agarose gel in TAE buffer containing 20 mM sodium acetate, pH 7.5. After ethidium bromide staining, the positions of the supercoiled and linearized plasmids were visualized with UV light. The gel was then dried under vacuum onto DE81 paper (Whattman) and autoradiographed to detect labeled oligonucleotides.

Transient replication assays

Transient replication assays were performed in D98HR1 cells by co-transfections of plasmids carrying either the wild-type (p968.22) or the mutated OriLyts together with plasmid pCMV-BZLF1, which efficiently expresses EB1 protein ( 15 ). Two days after co-transfection, DNA was prepared by the Hirt technique, digested with Dpn I and Bam HI, subjected to electrophoresis through a 0.7% agarose gel, transferred to a nylon N+ membrane (Amersham) and probed with random primer 32 P-labeled pUC19. The replication efficiency of plasmids carrying OriLyt was quantified by scanning the Southern blot autoradiogram.


Figure 1 . EBV OriLyt structure. The position of the different origins of replication of EBV (Ori-P and two OriLyt) are localized in a schematic representation of the EBV genome (TR, terminal repeats; IR, internal repeats). Ori-P is used to maintain the EBV genome as an episome in the cell nucleus and is transactivated by the EBNA1 protein. In most EBV strains a part of OriLyt is present twice in two functional copies which are >100 kb apart. OriLyt is used to amplify the copy number of the EBV genome during the productive cycle and is transactivated by the EBV immediate early gene EB1. An EBV subgenomic fragment encompassing OriLyt within the left duplicated segment (IR2) is shown schematically in the middle part of the figure. OriLyt is flanked by two divergent early genes, BHLF1 and BHRF1, which are shown together with their promoter localizations (gray boxes) and mRNA (hatched rectangles). The localization of the two essential elements of OriLyt, the upstream and downstream components, are shown in the lower part of the figure, together with those of the auxiliary regions (shaded areas), which increase the efficiency with which OriLyt replicates. An enlarged view of OriLyt is given at the bottom of the figure. Known functional elements of the upstream component are indicated (TATA box and four ZREs). The upstream component co-localizes with the promoter of the BHLF1 gene. The downstream component is located 440 bp from the upstream component. The coordinates for the boundaries of the two essential components are shown according to Baer et al. (25)


Figure 2 . Low resolution mapping of P1 nuclease sensitivity. ( A ) Nucleotide sequence of the OriLyt downstream component from nucleotide coordinates 53337 to 53428 (25). The C-rich and the mirror repeat elements are indicated. The mirror symmetry axis is underlined. ( B ) Graphic representation of the p1562 plasmid used in the P1 nuclease assays. This contains the OriLyt downstream component (gray box) cloned into the Sma I site of the pBluescript KSII (pBSK) vector. The position of the two Pvu II sites are indicated together with the expected fragment size. ( C ) pBSK or p1562 plasmid DNA was treated with increasing amounts of P1 nuclease before (S) (lanes 1-3 and 5-7) or after (L) (lanes 4 and 8) digestion with Pvu II. The digests were resolved on a 1% agarose gel and visualized by ethidium bromide staining. The molecular size marker (M) is a 1 kb ladder (Gibco BRL). The sizes of the fragments obtained are indicated by arrows.

RESULTS

The OriLyt downstream component is sensitive to single- strand DNA-specific nuclease

The OriLyt downstream component contains a homopurine- homopyrimidine sequence which forms a mirror repeat element, also called the H palindrome (Fig. 2 A). Analysis of the nucleotide sequence of the H palindrome suggested that it might form a DNA triplex structure. Formation of an intramolecular triplex structure in vitro occurs on supercoiled DNA and induces a single-stranded DNA loop sensitive to specific nucleases ( 26 ). We thus tested whether the homopurine-homopyrimidine sequence of the OriLyt downstream component present in supercoiled DNA (p1562) is sensitive to the single-strand-specific nuclease P1. The supercoiled plasmid p1562 was treated with increasing amounts of P1 nuclease and then digested with Pvu II (Fig. 2 B). As shown in Figure 2 C, two fragments of 2.5 and 0.5 kb were generated by digestion of plasmid p1562 with Pvu II (lane 5). In the presence of P1 nuclease two additional fragments of ~0.2 kb were also generated, as expected if the OriLyt sequence carried by the supercoiled plasmid contains a single-stranded DNA region (lanes 6 and 7). With the pBSK supercoiled vector, devoid of the OriLyt downstream component, only the two fragments of 2.5 and 0.45 kb were detected (lanes 1-3), indicating that the OriLyt downstream component cloned in the pBSK vector was necessary for P1 nuclease sensitivity. When the p1562 or pBSK plasmids were digested with Pvu II before treatment with P1 nuclease, the two small 0.2 kb fragments were not observed (lanes 4 and 8), demonstrating that sensitivity of the plasmid to the action of P1 nuclease was dependent on supercoiling. These results suggest that single-stranded DNA is present in the OriLyt downstream component carried by a supercoiled plasmid.


Figure 3 . High resolution mapping of P1 nuclease sensitivity. Supercoiled or linearized p1562 plasmid DNA was submitted to increasing amounts of P1 nuclease. The partially digested DNA was used as a DNA matrix for the primer extension assay. ( A ) Results from mapping on the G strand. ( B ) Results from mapping on the C strand. A, C, G and T are a sequencing reaction using the same primer used for the primer reaction. A schematic representation of the sequence with the position of the OriLyt downstream component is drawn to the left of the sequencing reaction. ( C ) Nucleotide sequence of the OriLyt downstream component indicating major (|) and minor (|) cleavages by P1 nuclease on the C and G strands.

To confirm these observations and to map precisely which strand and which bases are cleaved by P1 nuclease, we performed primer extension assays in the OriLyt region using specific primers for each DNA strand of plasmid p1562. When supercoiled plasmid p1562 was treated with limiting amounts of P1 nuclease at pH 7.6, several cleavage sites were observed on the G strand located in the OriLyt downstream component and also at different positions in the vector [Fig. 3 A (lanes 1-4) and C]. The P1-hypersensitive sites were specific for supercoiled plasmid, since they were not detected when the plasmid was digested with Pvu II before P1 treatment (Fig. 3 A, lanes 5-8). On the C strand, P1 nuclease cleavage sites accumulated in the 5'-half of the mirror repeat, which contains a homopyrimidine track (Fig. 3 B, lanes 1-6), but were not detected in the Pvu II digested plasmids (Fig. 3 B, lanes 7-12). With low amounts of P1 nuclease, different strong hypersensitive sites located in the OriLyt H palindrome were detected (Fig. 3 B, lane 2). Increasing the amount of P1 nuclease did not modify the relative intensity of these sites (Fig. 3 B, compare lanes 4-6 with 2), but increased the relative intensity of the sites surrounding the H palindrome (Fig. 3 B, lanes 4-6), suggesting that the first sites which were accessible to the nuclease were located in the 5'-border of the mirror repeat. These differences in the relative intensity of the bands were not observed on the G strand, suggesting that only the C strand is single stranded. Similar results were observed with plasmid p1562 digested with either nuclease S1 or P1 at pH 4.6 (not shown). The asymmetrical distribution of P1 nuclease-sensitive sites on the C strand of the H palindrome (Fig. 3 C) suggests that the DNA in the OriLyt downstream component formed a triplex structure in which part of the C strand was in a single-stranded state, whereas the G strand was folded into the major groove of the double helix where it was protected from P1 nuclease attack.

Oligonucleotide uptake by supercoiled plasmid p1562

If the 5'-part of the C strand in the H palindrome is single stranded, then it should hybridize with a radiolabeled complementary single-stranded oligonucleotide. To test this hypothesis, several radiolabeled single-stranded oligonucleotides complementary to the OriLyt downstream component (Fig. 4 A) were incubated with supercoiled or linearized plasmid p1562. In an agarose gel the hybridized radiolabeled oligonucleotide should co-migrate with the supercoiled p1562 plasmid but not with the linearized plasmid. As shown in Figure 4 B, only two single-stranded oligonucleotides, T1.2.G (lane 2) and T2.G (lane 3), detectably hybridized with the supercoiled plasmid p1562. However, T2.G and T2.C did not hybridize with linearized plasmid p1562 (Fig. 4 B, lanes 11 and 12) nor with supercoiled plasmid pBSK (Fig. 4 B, lanes 13 and 14). Moreover, uptake of oligonucleotide T1.2.G by the supercoiled plasmid p1562 was less efficient than that of oligonucleotide T2.G (Fig. 4 B, compare lanes 2 and 3). It should be noted that the sequences of the T1.2.G and T2.G oligonucleotides are complementary to a region in the OriLyt downstream component that was strongly sensitive to P1 nuclease (Fig. 3 C). These results strongly indicate that the 5'-border of the C-rich sequence of the OriLyt downstream component is single stranded when carried by a supercoiled plasmid. Taken together, these results suggest that DNA in the OriLyt downstream component has the potential to form a triplex structure in vitro when carried by a supercoiled plasmid and that the 5'-part of the C strand is likely to be single stranded (Fig. 4 C).


Figure 4 . Oligonucleotide uptake by plasmid p1562. ( A ) Supercoiled or linearized p1562 plasmids were incubated with different radioactive oligonucleotides corresponding to the G or the C strand of the OriLyt downstream component. ( B ) The plasmid oligonucleotide mixtures were resolved on a 1% agarose gel. The position of the different forms of the plasmid (-S-, supercoiled; -R-, relaxed) were visualized by ethidium bromide staining of the gel prior to drying and autoradiography. ( C ) A schematic representation of the possible non-B-DNA structure adopted by the OriLyt downstream component consistent with the results shown in Figure 3, together with the hybridized T2.G oligonucleotide (gray sequence).

Oligonucleotide uptake by p1562 mutants

We next evaluated whether mutations in the H palindrome of the OriLyt downstream component that impaired replication in a transient replication assay (Fig. 5 A; 23 ) also impaired uptake of oligonucleotide T1.2.G in vitro . As shown in Figure 5 B, oligonucleotide T1.2.G hybridized very efficiently with supercoiled plasmid p1562 (lane 1) but not with H palindrome mutants that were inactive in the replication assay (lanes 2-4 and 7). Mutant p1568, whose replication efficiency was only 5% of that of plasmid p1562, also hybridized weakly with oligonucleotide T1.2.G (lane 6). In the replication assay, mutant p1567 was 80% less efficient than the wild-type OriLyt plasmid p1562 and interestingly uptake of radiolabeled oligonucleotide T1.2.G by this mutated plasmid was ~20% of that observed with the wild-type OriLyt plasmid p1562 (lane 5). Thus there seems to be a good correlation between efficiency of replication of the H palindrome mutants and efficiency of uptake of oligonucleotide T1.2.G by these mutants.


Figure 5 . Oligonucleotide uptake by mutated p1562 plasmids. ( A ) Schematic representation of the localization of the mutants in p1562. Transversion mutants which carried two base exchanges (GG) compared with the wild-type p1562 sequence have been described elsewhere (23). The replication efficiencies of these different mutants have been described elsewhere (23). ( B ) Supercoiled plasmids were incubated with radioactive oligonucleotide T1.2.G. The plasmid oligonucleotide mixtures were resolved on a 1% agarose gel. The position of the different forms of the plasmid (-S-, supercoiled; -R-, relaxed) were visualized by ethidium bromide staining of the gel prior to drying and autoradiography.

OriLyt H-DNA mutants: oligonucleotide uptake and replication assay

If our conclusion is valid, any mutation introduced into the 5'-part or into the 3'-part of the mirror repeat of the OriLyt downstream component should impair formation of a triple helix in vitro and the function of OriLyt in vivo . However, a compensatory mutation which recreates a mirror repeat but does not modify the homopurine-homopyrimidine nature of the sequence should restore the structure and possibly the function of OriLyt. We generated the mutants depicted in Figure 6 A and first analyzed the efficiency of oligonucleotide T1.2.G uptake by supercoiled plasmids carrying the mutations. As already shown, an OriLyt downstream component containing the wild-type mirror repeat cloned in replication test plasmid p968 took up oligonucleotide T1.2.G (Fig. 6 B, lane 1). Mutants with a C -> T transition in either the left (p1648 and p1658) or right (p1637 and p1662) half of the mirror repeat (Fig. 6 A) did not hybridize with oligonucleotide T1.2.G (Fig. 6 B, lanes 2, 3, 5 and 6). In agreement with our prediction, mutants in which the mirror symmetry was restored (p1649 and p1659; Fig. 6 A) took up oligonucleotide T1.2.G (Fig. 6 B, lanes 4 and 7). However, uptake of T1.2.G by mutants p1649 and p1659 was less efficient than that observed with the wild-type mirror repeat. Nevertheless, our results demonstrate that homopurine-homopyrimidine mirror sequences similar to that found in the OriLyt downstream component could form a triplex in vitro .


Figure 6 . H-DNA mutants: oligonucleotide uptake and functional analysis. ( A ) The sequence of the different mutants in the OriLyt downstream component is shown, with substituted bases underlined. The sequence of oligonucleotide T1.2.G is also indicated. ( B ) Supercoiled plasmids were incubated with radiolabeled oligonucleotide T1.2.G. The plasmid oligonucleotide mixtures were resolved on a 1% agarose gel. The position of the supercoiled (S) form of the plasmid was visualized by ethidium bromide staining of the gel prior to drying and autoradiography. ( C ) The mutants were analyzed for their replication efficiency in a transient replication assay. The specific signal for the replicated plasmid is indicated by an arrow. p968.22 represents the wild-type OriLyt plasmid. The replication efficiencies of the p1649 and p1659 mutants were 5% as measured with the aid of a phosphorimager and normalized against the replication efficiency of p968.22, which was set as 100%.

The different OriLyt H-DNA mutants were also analyzed for their replication efficiencies in a transient replication assay. As shown in Figure 6 C, mutants with an altered mirror repeat sequence (p1648, p1637, p1658 and p1662) were completely inactive in the replication assay (Fig. 6 C, lanes 2, 3, 5 and 6). The two mutants in which a mirror repeat sequence was restored by a compensatory mutation (p1649 and p1659) supported replication, although only at a level of 5-10% compared with the wild-type OriLyt in p968.22 (Fig. 6 C, compare lanes 1, 4 and 7). Our results strongly suggest that mutations in the H palindrome which impair formation of a triple helix in vitro also impair replication of plasmids carrying these mutated OriLyt in vivo .

DISCUSSION

Chromosomal or viral origins of replication are composed of different cis -acting elements consisting of a putative DNA unwinding element aligned with clusters of scaffold-associated region, autonomously replicating sequence and pyrimidine tracts ( 27 , 28 ). These cis -acting elements are sites of interaction with origin-specific DNA binding proteins, which initiate DNA replication in this region ( 29 ). DNA replication origins are often associated with transcriptional units ( 30 ). The origin of replication OriLyt of EBV seems to be structurally slightly different from other eukaryotic origins of replication. It is composed of two essential core elements which lie >400 bp apart and are unique in their DNA sequence. However, their function in OriLyt-dependent replication is still unknown. The upstream component co-localizes with a basal promoter element which is concomitantly activated with OriLyt. Induction of this component for both transcription and replication is transactivated by EB1 ( 20 , 21 ). The downstream component has been shown previously to contain binding sites for several cellular proteins, including Sp1 ( 23 ). However, these factors do not appear to be directly implicated in OriLyt-dependent replication. In an attempt to understand the function of the downstream component, we show here that the downstream component contains an H palindrome capable of forming a non-B-DNA structure in vitro which is most likely a triple helix. In addition, we found that all mutations which disturb the structure of the OriLyt downstream component in a supercoiled plasmid in vitro also impair the replication efficiency in vivo of a plasmid carrying an OriLyt that contains these mutations. Thus it seems that the OriLyt downstream component has to adopt a particular structure in order to be functional.

DNA triplexes are formed by polypyrimidine/polypurine sequences. In this structure a DNA strand (donor strand) from one half of the sequence folds into the major groove of the other half duplex, forming Hoogsteen base pairs and leaving the other strand in a single-stranded state. Some triplex structures are formed only at acidic pH, while others are stabilized by the presence of magnesium in the buffer and can form at neutral pH. Triplex DNA formation in vitro always requires DNA supercoiling to compensate for the high nucleation energy required (reviewed in 26 , 31 ). Our results indicate that the OriLyt downstream component can adopt a non-B-DNA structure in vitro in a supercoiled plasmid. The structure formed at pH 7.6, which is an indication that triplex formation does not require protonation of the C-rich strand to become the donor strand. The DNA triplex is of a Pu-Py/Pu type. Confirmation of this hypothesis was given by fine mapping of the P1 nuclease-hypersensitive sites and by oligonucleotide uptake by supercoiled plasmids. Indeed, the pyrimidine-rich strand was more sensitive to P1 nuclease digestion than the purine-rich strand and might therefore be single stranded. This was confirmed by the observation that only oligonucleotides complementary to the pyrimidine strand were taken up by a supercoiled plasmid carrying the H palindrome. One interpretation of these results is that the DNA triplex structure formed utilizes the homopurine strand as the donor strand and leaves the homopyrimidine strand single stranded, as shown in Figure 4 C. Surprisingly, oligonucleotide T3.G, which shares some sequences with oligonucleotide T2.G, was not, or at least only very weakly, taken up by supercoiled plasmid p1562 (Fig. 4 B, lane 4). This oligonucleotide is composed of a guanine-rich DNA sequence which in vitro has been shown to be capable of associating to form a stable, parallel four-stranded structure termed G4 DNA under our assay conditions ( 32 ). Folding of oligonucleotide T3.G could explain why it could not interact with supercoiled plasmid p1562. Formation of the triplex DNA structure required only the presence of the mirror repeat sequence and did not appear to be influenced by the C-rich region which is located near the H palindrome. In effect, a plasmid carrying only the H palindrome sequence took up oligonucleotide T1.2.G as efficiently as a plasmid carrying the H palindrome sequence together with the C-rich sequence (not shown). However, The C-rich sequence is also important for function of the downstream component and it is possible that these two elements contribute differently to OriLyt-dependent replication. We estimate that <10% of the supercoiled plasmids used in the oligonucleotide uptake assay were in the triple-stranded form. This low proportion can be explained by the short length of the homopurine-homopyrimidine sequence implicated in the triplex structure and by the fact that the triplex structures are in rapid equilibrium with normal B-DNA structures. This rapid equilibrium could be detrimental to interaction between the plasmid and the oligonucleotide.

We do not, however, have strong evidence that the DNA triplex exists in vivo . In cells latently infected with EBV, no S1-hypersensitive sites were detected in the OriLyt downstream component (not shown). This does not mean that this region is not folded like H-DNA. It could be that the triplex is present but protected from nuclease attack by a cellular(s) factor(s). It is also conceivable that the DNA triplex structure is only transient. The fact that the OriLyt cis -acting elements co-localize with two divergent promoters could suggest that the transcription process can influence and/or stabilize formation of a non-B-DNA structure. It is well established that negative supercoiling is an energetic prerequisite for triplex formation under physiological conditions. Moreover, it has been shown that actively transcribing RNA polymerase complexes generate positive supercoiling in front of and negative supercoiling behind the elongation complex: this has been observed in vitro and in vivo in both prokaryotes and eukaryotes ( 33 - 38 ). Thus, transcription induced at the BHLF1 or BHRF1 promoter could facilitate DNA replication from OriLyt by introducing negative supercoiling favoring DNA triplex formation in the OriLyt downstream component.

DNA secondary structure has been shown to be important for replication origin activity of some bacterial plasmids, phages and eukaryotic viruses ( 39 - 42 ) and may be involved in initiation of DNA synthesis in mammalian cells ( 43 ). In this study we have observed that all mutations which disturb the DNA sequence of the H palindrome and impair formation of the DNA triplex in the OriLyt downstream component in vitro also alter the replication efficiency in vivo of plasmids carrying the mutated OriLyt. These results suggest that the structure of this component is probably important for its function in replication. A direct implication of a non-B-DNA structure in EBV OriLyt replication is strongly suggested by comparing the effects of mutations that change both the sequence and the structure of the H palindrome (p1637, p1648, p1662 and p1658) with those that alter only the sequence of the H palindrome (p1649 and p1659). Replication of plasmids which contain a 1 bp change in the left or right side of the mirror repeat and modify both the sequence and structure is impaired. Compensatory double mutants in the left and the right side of the mirror repeat retain part of the DNA triplex structure and replicate. However, the replication efficiency of these mutants is greatly reduced, suggesting that the sequence of the H palindrome, in addition to its structure, is important for efficient replication. One possibility is that the mutations reduce the stability of base pairing in the triplex compared with the wild-type sequence. In effect, in the mutants we have substituted two cytosines by two thymines, which is known to reduce folding stability. In vitro these compensatory mutations are also detrimental to the structure. Alternatively, the structured sequence may contain a recognition motif involved in origin function. This is consistent with the observation that exchange of the OriLyt H palindrome for a different homopurine/homopyrimidine sequence is detrimental to replication (not shown).

Taken together, our results suggest that both the sequence and structure of the OriLyt downstream component contribute to EBV replication during the lytic cycle. A similar component to the OriLyt downstream component, called the Y block, is present in the OriLyt of human cytomegalovirus, which, like EBV, does not have an origin binding protein. Interestingly, Huang et al . have shown that some transcripts, called SRT, overlap the Y block and are complementary to the C-rich strand ( 44 ). These RNAs might cooperate with the Y block to initiate or stabilize strand separation, as observed in mitochondrial heavy strand replicator, in which a similar C-rich component, called the CSBII box, promotes formation of a stable RNA-DNA hybrid forming a locally opened region, a D loop, coincident with the origin of DNA synthesis ( 45 , 46 ). This type of system could explain, at least in part, the relative complexity of OriLyt and the lack of a requirement for a specific origin binding protein.

ACKNOWLEDGEMENTS

We thank Dr W.Hammerschmidt for continuous interest and helpful suggestions. We thank Conrad B.Bluink for reading the manuscript and our colleagues for discussions. S.P.-S. was supported by an MRT fellowship. Research in the laboratory is financially supported by INSERM and the Association pour la Recherche contre le Cancer (contract ARC 2049 to H.G).

REFERENCES

1 Rickinson,A.B. and Kieff,E. (1996) In Knipe,B.N.F.a.D.M. (ed.), Fields Virology, 2nd Edn. Raven Press, New York, NY, Vol. 3, pp. 1889-1920.

2 Sixbey,J.W. (1989) Adv. Viral Oncol., 8, 187-202.

3 Liebowitz,D. and Kieff,E. (1993) In Roizman,B., R.J.W. and Lopez,C. (eds), The Human Herpesvirus, 2nd Edn. Raven Press, New York, NY, pp. 107-172.

4 Yates,J.L. and Guan,N. (1991) J. Virol., 65, 483-488. MEDLINE Abstract

5 Gahn,T.A. and Schildkraut,C.L. (1989) Cell, 58, 527-535. MEDLINE Abstract

6 Yates,J., Warren,N., Reisman,D. and Sugden,B. (1984) Proc. Natl. Acad. Sci. USA, 81, 3806-3810. MEDLINE Abstract

7 Yates,J.L., Warren,N. and Sugden,B. (1985) Nature, 313, 812-815. MEDLINE Abstract

8 Zur Hausen,H., O'Neil,F.J., Freese,U.K.and Hecker,E. (1978) Nature, 27, 373-375.

9 Takada,K. and Ono,Y. (1989) J. Virol., 63, 445-449. MEDLINE Abstract

10 Chevallier-Greco,A., Manet,E., Chavrier,P., Mosnier,C., Daillie,J. and Sergeant,A. (1986) EMBO J., 5, 3243-3249.

11 Countryman,J. and Miller,G. (1985) Proc. Natl. Acad. Sci. USA, 82, 4085-4089. MEDLINE Abstract

12 Hardwick,J.M., Lieberman,P.M. and Hayward,S.D. (1988) J. Virol., 62, 2274--2284. MEDLINE Abstract

13 Chevallier-Greco,A., Gruffat,H., Manet,E., Calender,A. and Sergeant,A. (1989) J. Virol., 63, 615-623.

14 Gruffat,H. and Sergeant,A. (1994) Nucleic Acids Res., 22, 1172-1178. MEDLINE Abstract

15 Hammerschmidt,W. and Sugden,B. (1988) Cell, 55, 427-433. MEDLINE Abstract

16 Fixman,E.D., Hayward,G.S. and Hayward,S.D. (1992) J. Virol., 66, 5030-5039. MEDLINE Abstract

17 Fixman,E.D., Hayward,G.S. and Hayward,S.D. (1995) J. Virol., 69, 2998-3006. MEDLINE Abstract

18 McGeoch,D.J., Dalrymple,M.A., Dolan,A., McNab,D., Perry,L.J., Taylor,P. and Challberg,M.D. (1988) J. Virol., 62, 444-453. MEDLINE Abstract

19 Pari,G. and Anders,D. (1993) J. Virol., 67, 6979-6988. MEDLINE Abstract

20 Schepers,A., Pich,D., Mankertz,J. and Hammerschmidt,W. (1993) J. Virol., 67, 4237-4245. MEDLINE Abstract

21 Schepers,A., Pich,D. and Hammerschmidt,W. (1993) EMBO J., 12, 3921-3929. MEDLINE Abstract

22 Schepers,A., Pich,D. and Hammerschmidt,W. (1996) Virology, 220, 367-376. MEDLINE Abstract

23 Gruffat,H., Renner,O., Pich,D. and Hammerschmidt,W. (1995) J. Virol., 69, 1878-1886. MEDLINE Abstract

24 Glaser,R., Nonoyama,M., Whitman,M. and Forcino,T. (1979) Virology, 98, 272-274. MEDLINE Abstract

25 Baer,R. et al. (1984) Nature, 310, 207-211. MEDLINE Abstract

26 Frank-Kamenetskii,M.D. and Mirkin,S.M. (1995) Annu. Rev. Biochem., 64, 65-95.

27 Benbow,R.M., Zhao,J. and Larson,D.D. (1992) BioEssays, 14, 661-670. MEDLINE Abstract

28 Dobbs,D., Shaiu,W. and Benbow,R. (1994) Nucleic Acids Res., 22, 2479-2489. MEDLINE Abstract

29 Challberg,M. and Kelly,T. (1989) Annu. Rev. Biochem., 58, 671-717. MEDLINE Abstract

30 DePamphilis,M.L. (1988) Cell, 52, 635-638. MEDLINE Abstract

31 Mirkin,S.M. and Frank-Kamenetskii,M.D. (1994) Annu. Rev. Biophys. Biomol. Struct., 23, 541-576. MEDLINE Abstract

32 Sen,D. and Gilbert,W. (1992) Methods Enzymol., 211, 191-199. MEDLINE Abstract

33 Liu,L.F. and Wang,J.C. (1987) Proc. Natl. Acad. Sci. USA, 84, 7024-7027. MEDLINE Abstract

34 Wu,H.Y., Shyy,S.H., Wang,J.C. and Liu,L.F. (1988) Cell, 53, 433-440. MEDLINE Abstract

35 Brill,S.J. and Sternglanz,R. (1988) Cell, 54, 403-411. MEDLINE Abstract

36 Figueroa,N. and Bossi,L. (1988) Proc. Natl. Acad. Sci. USA, 85, 9416-9420. MEDLINE Abstract

37 Giaever,G.N. and Wang,J.C. (1988) Cell, 55, 849-856. MEDLINE Abstract

38 Tsao,Y.P., Wu,H.Y. and Liu,L.F. (1989) Cell, 56, 111-118. MEDLINE Abstract

39 Wing,D., Koncz,C. and Schell,J. (1989) Mol. Gen. Genet., 219, 9-16. MEDLINE Abstract

40 Miao,D.M., Honda,Y., Tanaka,K., Higashi,A., Nakamura,T., Taguchi,Y., Sakai,H., Komano,T. and Bagdasarian,M. (1993) Nucleic Acids Res., 21, 4900-4903. MEDLINE Abstract

41 Lefebvre,R.B., Riva,S. and Berns,K.I. (1984) Mol. Cell. Biol., 4, 1416-1419. MEDLINE Abstract

42 Hiasa,H., Tanaka,K., Sakai,H., Yoshida,K., Honda,Y., Komano,T. and Godson,G.N. (1989) Gene, 84, 17-22. MEDLINE Abstract

43 Zannis-Hadjopoulos,M., Frappier,L., Khoury,M. and Price,G.B. (1988) EMBO J., 7, 1837-1844.

44 Huang,L., Zhu,Y. and Anders,D.G. (1996) J. Virol., 70, 5272-5281. MEDLINE Abstract

45 Xu,B. and Clayton,D.A. (1995) Mol. Cell. Biol., 15, 580-589. MEDLINE Abstract

46 Xu,B. and Clayton,D.A. (1996) EMBO J., 15, 3135-3143. MEDLINE Abstract


Return

*To whom correspondence should be addressed. Tel: +33 472 72 81 76; Fax: +33 472 72 86 86; Email: hgruffat@popserver.ens-lyon.fr
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
J. Virol.Home page
A. Wiedmer, P. Wang, J. Zhou, A. J. Rennekamp, V. Tiranti, M. Zeviani, and P. M. Lieberman
Epstein-Barr Virus Immediate-Early Protein Zta Co-Opts Mitochondrial Single-Stranded DNA Binding Protein To Promote Viral and Inhibit Mitochondrial DNA Replication
J. Virol., May 1, 2008; 82(9): 4647 - 4655.
[Abstract] [Full Text] [PDF]


Home page
Nucleic Acids ResHome page
S.-A. Xue and B. E. Griffin
Complexities associated with expression of Epstein-Barr virus (EBV) lytic origins of DNA replication
Nucleic Acids Res., May 11, 2007; 35(10): 3391 - 3406.
[Abstract] [Full Text] [PDF]


Home page
J. Virol.Home page
G. Liao, J. Huang, E. D. Fixman, and S. D. Hayward
The Epstein-Barr Virus Replication Protein BBLF2/3 Provides an Origin-Tethering Function through Interaction with the Zinc Finger DNA Binding Protein ZBRK1 and the KAP-1 Corepressor
J. Virol., January 1, 2005; 79(1): 245 - 256.
[Abstract] [Full Text] [PDF]


Home page
J. Virol.Home page
D. P. AuCoin, K. S. Colletti, Y. Xu, S. A. Cei, and G. S. Pari
Kaposi's Sarcoma-Associated Herpesvirus (Human Herpesvirus 8) Contains Two Functional Lytic Origins of DNA Replication
J. Virol., June 27, 2002; 76(15): 7890 - 7896.
[Abstract] [Full Text] [PDF]


Home page
J. Virol.Home page
G. Liao, F. Y. Wu, and S. D. Hayward
Interaction with the Epstein-Barr Virus Helicase Targets Zta to DNA Replication Compartments
J. Virol., September 15, 2001; 75(18): 8792 - 8802.
[Abstract] [Full Text] [PDF]


Home page
J. Virol.Home page
Y. Zhu, L. Huang, and D. G. Anders
Human Cytomegalovirus oriLyt Sequence Requirements
J. Virol., June 1, 1998; 72(6): 4989 - 4996.
[Abstract] [Full Text] [PDF]


Home page
J. Virol.Home page
H. Deng and S. Dewhurst
Functional Identification and Analysis of cis-Acting Sequences Which Mediate Genome Cleavage and Packaging in Human Herpesvirus 6
J. Virol., January 1, 1998; 72(1): 320 - 329.
[Abstract] [Full Text] [PDF]


This Article
Right arrow Abstract Freely available
Right arrow Print PDF (249K) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (12)
Right arrow Commercial Re-use Guidelines
for Open Access NAR Content
Google Scholar
Right arrow Articles by Portes-Sentis, S.
Right arrow Articles by Gruffat, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Portes-Sentis, S.
Right arrow Articles by Gruffat, H.
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?