Episomal Maintenance of Plasmids with Hybrid Origi
http://www.100md.com
病菌学杂志 2005年第24期
Department of Microbiology and Virology, Institute of Molecular and Cell Biology, Tartu University, 23 Riia St., Tartu 51010, Estonia
Estonian Biocentre, 23 Riia St., Tartu 51010, Estonia
Department of Biomedical Technology, Institute of Technology, Tartu University, 21 Vanemuise St., Tartu 51010, Estonia
ABSTRACT
Bovine papillomavirus type 1 (BPV1), Epstein-Barr virus (EBV), and human herpesvirus 8 genomes are stably maintained as episomes in dividing host cells during latent infection. The mitotic segregation/partitioning function of these episomes is dependent on single viral protein with specific DNA-binding activity and its multimeric binding sites in the viral genome. In this study we show that, in the presence of all essential viral trans factors, the segregation/partitioning elements from both BPV1 and EBV can provide the stable maintenance function to the mouse polyomavirus (PyV) core origin plasmids but fail to do so in the case of complete PyV origin. Our study is the first which follows BPV1 E2- and minichromosome maintenance element (MME)-dependent stable maintenance function with heterologous replication origins. In mouse fibroblast cell lines expressing PyV large T antigen (LT) and either BPV1 E2 or EBV EBNA1, the long-term episomal replication of plasmids carrying the PyV minimal origin together with the MME or family of repeats (FR) element can be monitored easily for 1 month under nonselective conditions. Our data demonstrate clearly that the PyV LT-dependent replication function and the segregation/partitioning function of the BPV1 or EBV are compatible in certain, but not all, configurations. The quantitative analysis indicates a loss rate of 6% per cell, doubling in the case of MME-dependent plasmids, and 13% in the case of FR-dependent plasmids in nonselective conditions. Our data clearly indicate that maintenance functions from different viruses are principally interexchangeable and can provide a segregation/partitioning function to different heterologous origins in a variety of cells.
INTRODUCTION
Several eukaryotic DNA viruses maintain their genomes as extrachromosomal multicopy nuclear episomes in proliferating host cells. Such episomal maintenance is characteristic of latent infection of bovine papillomavirus type 1 (BPV1), Epstein-Barr virus (EBV), and Kaposi's sarcoma-associated human herpesvirus 8 (HHV8). Two functions of the viral genome are absolutely critical for extrachromosomal maintenance in dividing cells: viral genome replication during the S phase and proper segregation and partitioning of the replicated genomes into daughter cells during host cell mitosis. For BPV1 and two members of the gammaherpesvirus family, EBV and HHV8, effective segregation of viral genomes into daughter cells and nuclear retention during mitosis are mediated through a single viral protein serving as a molecular linker, which attaches viral genomes to the host mitotic chromosomes (4, 11, 12, 19, 24, 35). This linker protein is viral regulatory protein E2 for BPV1 (18, 24, 35), viral transactivator EBNA1 for EBV (19), and viral transcription repressor LANA1 for HHV8 (4, 5).
For the initiation of the DNA replication of a BPV1-based replicon in vivo, the minimal origin region in cis and two viral proteins, E1 and E2, in trans, are absolutely essential (39, 40). However, the minimal origin is not sufficient for stable extrachromosomal replication in dividing cells (30). An additional element, the minichromosome maintenance element (MME), ensures the long-term episomal persistence of the genome in the presence of viral E1 and the E2 proteins in the dividing cells (30). In the BPV1 genome, in total 17 E2 protein binding sites (BS) with different affinities for E2 can be identified; 12 of these are located in the noncoding upstream regulatory region (URR) (26). We have shown that, for efficient partitioning/segregation of the episomal plasmid, MME activity is provided by a sufficient number of high-affinity E2 BS (30). The function of multimeric E2 BS in the stable maintenance of the BPV1 genomes is to provide the anchoring function for the E2 protein, which therefore tethers MME-containing plasmids to mitotic chromosomes (18, 24). This linkage between the BPV1 genome and host chromatin ensures also that the viral genome is maintained in the nucleus when the nuclear membrane is reassembled during mitosis. In the case of EBV, the stable maintenance of replicated genomes is achieved due to the EBNA1 protein and family of repeats (FR) element, which is composed of multimeric EBNA1 protein binding sites (19, 27).
We have shown that both the BPV1 E2 protein-dependent MME (1) and EBV EBNA1-dependent FR (A. M?nnik, K. Janikson, and M. Ustav, unpublished data) segregation/partitioning and chromatin attachment activities function independently from replication of the plasmids (18). The stable-maintenance function of EBNA1/FR has been used to ensure long-time episomal maintenance for non-OriP origins, usually the cellular replication origins (22, 41). In the case of OriP, the enzymatic activity required for initiation of replication is the same as in cellular origins (14, 37). The E2/MME-dependent stable-maintenance function has not been tested with heterologous replication origins. In the present study, we have further examined the compatibility of viral segregation/partitioning elements with heterologous replication origins. For this purpose, different reporter plasmids were constructed that combine the E2/MME- and EBNA1/FR-based stable-maintenance function and different variants of the mouse polyomavirus (PyV) replication origin. The mouse polyomavirus is a lytic virus, which replicates its DNA very fast during productive infection. Infected cells contain up to 200,000 molecules of viral DNA, and the maximal copy number is reached about 50 h postinfection (10). The replication origin of PyV contains the transcription/replication enhancer responsible for the high level of replication (13). We tested the stable maintenance of plasmids that contained E2/MME in conjunction with the wild-type (wt) PyV or core (enhancerless) origin of replication in cell lines expressing PyV large T antigen (LT) and E2 or its mutants. Also the stable maintenance of a plasmid containing the FR and the PyV core origin was tested in cell lines expressing LT and EBNA1. The results from these experiments show convincingly that the segregation/partitioning functions of BPV1 and EBV can effectively be used for stable episomal maintenance of the PyV core origin. In addition, efficient chromatin attachment rather than a high level of activation of replication is required for stable episomal maintenance.
MATERIALS AND METHODS
Plasmids. For constructing hybrid replicons (Fig. 1B and C) containing the PyV origin (wild type or core origin), we used vector pUC19 as the basic backbone, where we cloned 1, 5, or 10 head-to-tail copies of high-affinity E2 BS 9. PyV wt and the core origin were amplified by PCR from vectors pmu1046/CAT and pmu1047/CAT (29) using primers Py4963 (5'-AGGGAGCTACTCCTGATG-3') and Py174 (5'-CTACCACCACTCCGACTT-3'). Amplified PyV origin fragments were digested with enzymes EheI and BclI and inserted between BamHI and HincII sites of the pUC19 vector containing different numbers of BPV1 E2 BS.
Hybrid replicons containing a Geneticin resistance gene (Fig. 1D) were established by replacing the URR in plasmid pNeoBgl40 (30) with PyV wt origin, core origin, or core origin with 10 E2 BS, which were amplified by PCR and digested with enzymes BamHI and Ecl136II and cloned into BamHI and HindIII sites in pNeoBgl40.
Three types of the enhanced green fluorescent protein (EGFP) marker containing plasmids were designed. First, a fragment comprising the PyV minimal origin and 10 E2 BS was added to a plasmid containing a Geneticin resistance marker (expressed from the simian virus 40 promoter). Then either an EGFP or destabilized green fluorescent protein (d1EGFP) marker was added (named either pMMEG or pMMEG plasmid; see Fig. 6A). EGFP expression cassettes, which are under the control of the cytomegalovirus promoter, were taken either from pEGFP-C1 or pd1EGFP-N1 plasmids (Clontech). For the third plasmid, the first EBV FR element was added to the pUC19 plasmid containing the PyV core origin. Then the fragment containing the PyV minimal replication origin and 10 copies of E2 BS 9 from plasmid pMMEG was replaced by the fragment containing the PyV minimal origin and EBV FR element (plasmid pFRG; see Fig. 6A).
Construction of cell lines. For construction of cell lines which express BPV1 wt E2 protein and its mutant forms E39A and R68A, the vector pBabePuro (28) was linearized using enzyme SalI and was ligated with an equal amount of E2 expression vectors (pCGE2, pCGE2/R68, pCGE2/E39 [2, 39]), which were linearized with XhoI endonuclease. One microgram of ligated hybrid plasmids was electroporated into the COP5 cell line (38). Electroporation experiments were preformed with a Bio-Rad Gene Pulser with capacitance and voltage settings of 975 μF and 220 V. For selection puromycin (2 μg/ml) was added. The expression of the proteins was analyzed by Western blotting.
A cell line expressing wt E2 and carrying a Geneticin selection cassette was constructed by the same protocol described above using vector pBabeNeo (28) instead of pBabePuro.
A cell line expressing PyV T antigens and EBV EBNA1 protein was generated as a result of transfection of the NotI-linearized plasmid pBabePuro/EBNA1 (EBNA1 coding sequence inserted into EcoRI/SalI sites in the pBabePuro vector) into the COP5 cell line and selection for puromycin (2 μg/ml). The expression of the proteins was analyzed by Western blotting. The cell line was named COP5EBNA1/Puro.
Cells and transfection. COP5 cells (38) and their derivatives COP5E2/Puro, COP5E2/Neo, COP5R68/Puro, COP5E39/Puro, and COP5EBNA1/Puro expressing PyV T antigens and BPV1 wt E2 or its mutant forms or EBNA1 were grown in Iscove's modified Dulbecco's medium (IMDM) supplemented with 10% fetal calf serum. For selection Geneticin (500 μg/ml) or puromycin (2 μg/ml) was added, depending on the selection marker. Electroporation experiments were performed with a Bio-Rad Gene Pulser with capacitance and voltage settings of 975 μF and 220 V, respectively.
COP5E2/Puro cells transfected with neomycin constructs were selected with Geneticin at 500 μg/ml. COP5E2/Neo cells cotransfected with pBabePuro (28) were selected with puromycin at 2 μg/ml. After transfection with 500 ng of plasmids carrying the Geneticin resistance marker and EGFP coding sequence, the COP5EBNA1/Puro cell line was grown in IMDM containing 500 μg/ml Geneticin (medium contained no puromycin).
Southern blot analysis. Total DNA was extracted from cells by following a standard protocol (3). Extraction of low-molecular-weight DNA from cells and analysis of origin construct levels in both low-molecular-weight- and total-DNA preparations were performed as described previously (30, 39). All restriction reactions included DpnI to eliminate bacterially methylated input DNA. In the case of the nicking reaction 0.2 units of nicking enzyme Nb.Bpu10I (Fermentas, Vilnius, Lithuania) was added. During episomal- or total-DNA studies always equal numbers of cells or equal amounts of DNA were loaded onto each lane. Specific probes were labeled with [32P]dCTP by random-decamer-primed synthesis using the DecaLabel kit (Fermentas). PyV origin- and MME-specific probes were made by PCR with [32P]dCTP. Hybridizing species were visualized by autoradiography. Radioactive signals on the blots were quantified on PhosphorImagerSI using ImageQuant software (Molecular Dynamics, Amersham Biosciences, Little Chalfont, United Kingdom).
Immunoprecipitation. Cells (1.5 x 107) were lysed with ice-cold 1% sodium dodecyl sulfate (SDS)-phosphate-buffered saline on ice, collected in a 15-ml tissue culture tube, and sonicated. From this step an aliquot for the Bradford assay was taken. SDS was diluted to 0.1% by adding ice-cold radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% deoxycholate, 0.1 mM dithiothreitol [DTT], 0.5 mM phenylmethylsulfonyl fluoride, protease inhibitors). The insoluble fraction was sedimented by centrifugation at 5,000 x g for 15 min. The soluble fraction was transferred to a new tube and incubated with 5H4, 3E8, 1E4, and 3F12 antibodies (22) overnight at 4°C. Then protein G-Sepharose (Amersham Biosciences) was added and incubated for 1 h. Sepharose beads were washed three times with RIPA buffer and resuspended in SDS loading buffer and subjected to immunoblotting analysis with horseradish peroxidase-conjugated 5E11 (subclone of MAb 3F12) antibody (Quattromed AS, Tartu, Estonia).
Immunoblotting. Total protein from the same number of cells lysed in standard loading buffer supplemented with 100 mM DTT was separated by electrophoresis on an 8% polyacrylamide-SDS gel and transferred to an Immobilon-P membrane (Millipore). Antibody 1E4 (23) was used to detect E2 proteins. Antibodies BM3167 and BM1083 (DPC Biermann) were used to detect EBNA1 protein. Peroxidase-conjugated goat anti-mouse antibody and the enhanced chemiluminescence detection kit (ECL Western blotting reagents; Amersham Biosciences) were used for subsequent development of the blot, using a standard protocol provided by the supplier.
Plasmid rescue assay. Two micrograms of uncut genomic DNA was electrotransformed into Escherichia coli strain DH10B. The electrocompetent cells were prepared as described previously (34), and the transformations were performed using a Pulser apparatus and 2-mm electroporation cuvette (Bio-Rad Laboratories, Hercules, CA) according to the manufacturer's instructions. The cells were recovered by centrifugation and were grown on medium containing ampicillin at 100 μg/ml. Plasmid DNA from single colonies was purified and analyzed using restriction endonucleases.
Flow cytometry analysis. EGFP expression was analyzed by flow cytometry using a Becton Dickinson FACSCalibur flow cytometer with associated CellQuest software. One hundred thousand to 200,000 signals were analyzed from each sample. The threshold for autofluorescence was set to 99% of the signals from the mock-transfected control cells. All the signals above the threshold were considered to correspond to EGFP-positive cells. For calculating the episomal rates of loss in Table 1, EGFP expression data were analyzed on days 0 and 12 (pEGFP-C1 and pd1EGFP-N1), on days 0 and 55 for pMMEG, on days 0 and 37 for pMMEG, and on days 0 and 30 for pFRG (day 0 is the time point when selection was removed). For this calculation a first-order rate-of-loss model was used: rate of loss () = (–1/t)(ln Nt/N0) (41), where N0 is the percentage of green cells at the beginning of the experiment at nonselective conditions and Nt is the percentage of green cells after t generations.
RESULTS
BPV1 E2 protein and its multimeric binding sites activate the replication of PyV core origin and provide the segregation/partitioning function to the origin plasmid. The BPV1 E2 protein is a multifunctional protein which is involved in transcriptional regulation and viral DNA replication and segregation. It has been shown that, for stable episomal replication of BPV1, E1 and E2 proteins and the MME, which consists of multimeric E2 BS, are required (30). E2 protein can also activate the transient replication of the PyV core origin in vivo in E2 multimeric BS-dependent fashion (29). We decided to study if the BPV1 E2 BS in the hybrid PyV origin can, in addition to the activation of the initiation of replication, also provide the long-term maintenance function to the PyV-derived replicator in cells expressing LT and E2 protein.
Replication of the PyV origin requires LT as the only viral replication factor; all other components are derived from the host cell (9, 15). LT is an origin recognition factor and DNA helicase, thus directly participating in initiation and elongation of the replication of the viral origin (43). We constructed mouse cell lines expressing PyV LT and the BPV1 E2 protein using the cell line COP5, which constitutively produces LT from the integrated replication-defective PyV genome (38). Individual colonies were allowed to expand in the presence of selection (puromycin or Geneticin), and PyV LT- and BPV1 E2-positive double-expression cell lines were identified and characterized. The cell lines expressing E2 protein at the highest level were used in further assays (referred to as COP5/E2/Puro or COP5/E2/Neo, selected for puromycin or Geneticin selection markers, respectively). The same approach was used for construction of cell lines which express mutant forms of the E2 proteins, E39A and R68A (referred to as COP5/E39/Puro and COP5/R68/Puro). As described earlier, both these mutants are at least partially functional in E2 BS-dependent transcriptional activation and initiation of the BPV1 origin replication as well as in activation of initiation of PyV core origin replication; however, they fail to attach to the host cell chromosomes and do not support segregation/partitioning of the MME plasmids (1, 2; A. Abroi et al., submitted for publication). Expression of the wt E2 protein in the cell lines COP5E2/Neo and COP5E2/Puro was verified using Western blot analysis (Fig. 2B). In both cell lines the expression of the wt E2 protein was maintained at a detectable level for a prolonged period without selection, which is essential for the study of the maintenance of the episomal plasmids. We determined the expression level of E2 proteins in the constructed COP5E2/Puro, COP5E39/Puro, and COP5R68/Puro cell lines, as well as the E2 expression level from C127 cells stably maintaining the BPV1 genome as an episome. Immunoprecipitation with E2-specific antibodies and following normalized immunoblotting showed that the expression level of E2 proteins in constructed cell lines is higher than the full-length-E2 expression level in the BPV1-transformed cell line (Fig. 2C). Thus, the E2 protein expression level in our constructed cell lines is not limiting in stable-maintenance experiments.
The constructed cell lines were used in further experiments to study the effect of the BPV1 E2 protein and E2 BS on initiation of replication and maintenance of the constructed plasmids. At first, we examined the chimeric origins, comprising the PyV wt origin (Fig. 1B) or the core origin (Fig. 1C) linked to different numbers of E2 BS in cell lines COP5/E2/Neo (expressing constitutively PyV LT and BPV1 E2) and COP5 (expressing constitutively PyV LT), respectively. Ninety-six hours after transfection strong replication signals of wt origin plasmids were detected in both cell lines (Fig. 3A and B, 4-day time points, lanes 1 to 4). Additional E2 BS had rather an inhibitory effect on the replication of the wt PyV origin. However, the E2 protein-dependent activation of replication was clearly detected in the cases when the PyV enhancerless core origin was linked to different numbers of E2 BS. Addition of one E2 BS had no effect on the initiation of replication of the core origin; however, the addition 5 or 10 E2 BS activated core origin replication to almost the wt origin replication level in an E2 protein-dependent fashion (compare Fig. 3A and B, lanes 5 to 9). In the COP5 cell line lacking E2 protein, the replication enhancer function of the E2 BS to the core origin cannot be detected (Fig. 3B). These results are in principal agreement with the data previously published by Nilsson et al. (29), showing that the replication of the PyV enhancerless origin can be activated by BPV1 E2 and its BS.
We further studied the stable episomal maintenance of different PyV origin-containing constructs in two experimental settings—first, without any selective pressure, and second, under puromycin selection—after cotransfection of the origin plasmids together with plasmid pBabePuro, which encodes the puromycin resistance marker, to select out transfected cells (see Materials and Methods for details). The episomal persistence of the PyV origin-containing plasmids was analyzed by Southern blotting. wt origin plasmids were lost from the cells under selective and nonselective conditions very fast in COP5E2/Neo and COP5 cells (Fig. 3A and B). However, the hybrid origins comprising the core origin and 5 or 10 E2 BS were capable of long-term persistence (11 and 34 days, at least 27 doublings) in the COP5E2/Neo cells, as analyzed by episomal DNA extraction or analysis of total DNA from the transfected cells. After 21 and 34 days without selective pressure, the only origin construct that was efficiently maintained as an episome was the hybrid of the core origin with 10 E2 BS (Fig. 3A, 21- and 34-day time points without selection, lane 5). For the PyV origin with five E2 BS, a weak replication signal was detected only at longer exposure (data not shown). We estimated the average copy number of the episomal plasmids in the culture using total-DNA Southern blotting (Fig. 3A, total DNA time points). The plasmid with 5 and 10 E2 BS had, on average, 5 and 17 copies per cell, respectively, after 34 days (with the assumption that all cells contain a replicon). These data indicate that E2 and its BS can provide the episomal maintenance function for chimeric PyV origin constructs that are otherwise lost from the cell population during cell growth.
Somewhat unexpectedly, we found that when the BPV1 segregation/partitioning element is linked to the PyV wt origin (short-term replication signals on Fig. 3A, 4-day time points, lanes 1 to 4) these replicons are not able to survive despite of their high-level replication (Fig. 3A, lanes 1 to 4). This could be due to the overreplication of the intact enhancer-containing origin plasmid, which could lead to cell death. Inspection of the transfected culture indicated that indeed the wt PyV origin plasmids induced extensive cell death at the later time points (data not shown).
Replication of the origin plasmids carrying an episomal selection marker. COP5E2/Puro cells were transfected with three different plasmids carrying, in addition to the origin, a Geneticin selection marker (Fig. 1D). The eukaryotic selection cassette in the plasmid makes it possible to select for cells carrying reporter plasmids in transfected cells in the presence of Geneticin. Transient transfection of COP5E2/Puro with neomycin reporter plasmids resulted in a strong replication signal for the wt origin construct (Fig. 4A, time points at 48 and 72 h, lane 1) compared to a much lower replication signal for the core origin construct (Fig. 4A, 48- and 72-h time points, lane 2). As expected, addition of 10 E2 BS to the core origin increased the transient-replication signal (Fig. 4A, 48- and 72-h time points, lane 3). The transfected cells were then grown in selective medium containing Geneticin, followed by a series of cell divisions comparatively with and without selection. After 2 months of cultivation, these pooled cell lines were analyzed for stable maintenance of the reporter plasmids using Southern blotting with a radioactively labeled probe against the PyV origin. Ten E2 BS containing the reporter plasmid could establish the extrachromosomal maintenance of autonomous episomes in E2-positive cells (Fig. 4B, lanes 3, analysis after 2 months). Removal of the selection reduced the replication signal, but it was still detectable in the episomal fraction after 2 months (Fig. 4B, lanes 3, 2-month time points), and even after 5 months (data not shown).
Efficient partitioning/segregation rather than a high level of activation of replication is required for stable episomal maintenance. We compared the stable episomal maintenance of the hybrid origins in the cell lines expressing wt E2 with that in cell lines expressing mutant forms of E2 carrying alanine substitutions of the conserved charged residues in the N-terminal domain. These mutants have been previously characterized in BPV1 replication, transactivation, sequence-specific DNA binding, and partitioning assays (1, 2). E2 mutants E39A and R68A (Fig. 2A) are inactive in the chromatin attachment functions and failed to mediate the segregation/partitioning of the BPV1 URR reporter plasmids but were still active in initiation of transient replication and in transcription, where their relative activity was comparable to wt E2 (1). We transfected the COP5 cells with expression constructs for the BPV1 E2 mutant forms R68A or E39A as well as a puromycin resistance marker; the puromycin-resistant clones were picked, expanded, and characterized for expression of the desired proteins (Fig. 2C). The cells with the best expression were selected for the subsequent assays. In the following short- and long-term replication assays, we used the reporter constructs carrying the selection marker that confers resistance to Geneticin selection (Fig. 1D). Both E2 mutant forms R68A and E39A activated PyV core origin replication in an E2 BS-dependent fashion in established cell lines (Fig. 4A, lane 3, 48- and 72-h time points). This suggested that E2 mutant forms R68A and E39A behave as efficiently in replication activation as wt E2 protein. The transfected cells were grown in the media with and without Geneticin selection for time periods up to 2 months. By this time, only the replication of reporter plasmid with 10 E2 BS added to PyV the core origin was detectable in cells (Fig. 4B, lanes 1 to 3). In wt E2-expressing cells, this signal was present in Geneticin-selected cells as well as in control cells without selection. On the other hand, in the case of E2 mutant forms E39A and R68A, only a very weak replication signal was observed in cells grown under Geneticin selection (Fig. 4B, lane 3). It is important to note that further cultivation up to 5 months resulted in the complete loss of the episomal signal in mutant E2 cell lines (data not shown). The same results as in cell lines expressing mutant E2 proteins R68A and E39A were obtained from the experiments with cell lines which express hybrid proteins VP16/E2 and p53/E2, where the whole transactivation domain of the E2 protein is replaced with the respective activation domain from VP16 or p53 protein, respectively (data not shown). These transactivation domains have been shown to activate PyV replication very efficiently (6, 16; A. Abroi unpublished data) and at least the VP16 activation domain does not support the plasmid partitioning function (1). These results showed that the chromatin attachment function of E2 protein is required to ensure stable maintenance of the chimeric PyV origin and that the replication activation function alone is not sufficient for stable episomal maintenance.
Episomal state of chimeric origins. A high mutation frequency, especially for recombination, is often associated with replication from the papillomavirus and polyomavirus origin-based vector systems (8, 42). Therefore, we decided to check for this possibility in our experimental model. Episomal DNA was extracted from COP5/E2/Puro-derived cell lines to analyze the presence and the status of episomally maintained hybrid origin plasmids (Fig. 5A). Hybridization analysis with a neomycin gene-specific probe of the linearized DNA from the PyV MME reporter-carrying cell line revealed mostly one very discrete band that migrated similarly to the unit size marker on the agarose gel (Fig. 5A, compare lane 1 to lane 9). The sample digested with a noncutter (enzyme with no restriction sites in plasmid DNA) gave a pattern where open circular (OC) and covalently closed circular (CCC) forms can be detected (Fig. 5A, compare lane 2 with marker lanes 10 and 11). However, the additional slower-moving reporter-specific bands were observed (Fig. 5A, lane 2 and 3). We suggest that the signals correspond to the oligomerized episomes; both forms have been shown to appear, for example, during the episomal maintenance of papillomavirus full-length genomes in several cell lines (32) and URR-containing plasmids in an E1/E2-positive cell line (30). To confirm that hybridization signals are not from integrated material, the samples were digested with a noncutter enzyme together with a nicking enzyme Nb.Bpu10I (Fig. 5A, lane 3). Nb.Bpu10I is a site- and strand-specific endonuclease that cleaves only one strand of DNA within its recognition sequence on a double-stranded DNA substrate. Thus, by nicking enzyme CCC DNA transfers to the OC form, a DNA mobility shift on agarose gel is observed (Fig. 5A, lane 12). Nb.Bpu10I does not change linear DNA mobility as can be observed in Fig. 5A, lane 8, which represents circular DNA digestion with a linearizing enzyme together with a nicking enzyme. However, no hybridization signal was observed on lanes containing the wt PyV origin (Fig. 5A, lanes 4 to 6). The results of this experiment show that the analyzed episomal DNA fraction contained a reporter plasmid which was sensitive to the nicking enzyme so hybridization signals were not from integrated material.
The presence of episomal DNA was also confirmed by plasmid rescue into E. coli, using the uncut total DNA from the E2- and LT-expressing cells that were carrying a reporter plasmid with the PyV core origin and 10 E2 BS. Analysis of uncut rescued plasmids showed an oligomerized pattern compared to input DNA (Fig. 5B). Restriction analysis of rescued plasmids by endonuclease BglI showed that, compared to input DNA some rearrangements in the plasmid backbone can be observed (Fig. 5C, lanes 1 to 3, 5, 6, 10, and 12). Thus, some cells carried plasmids with rearrangements. In addition, intact unarranged DNA forms were also detected (Fig. 5C, lanes 4, 7 to 9, and 11). To confirm that rescued plasmids still contain the PyV origin and MME, we analyzed the BglI digestion pattern by Southern blot analysis with an MME- or PyV-specific probe (Fig. 5D and E, respectively). Southern blot analysis showed that all rescued reporter plasmids contained the MME and PyV origin fragment (Fig. 5D and E). A plasmid rescue assay with total DNA (total DNA was extracted from cells whose episomal DNA is analyzed in Fig. 5A, lanes 4 to 6) from cells carrying the reporter plasmid with the wt PyV origin revealed only one colony, which we analyzed for the existence of the MME or PyV origin fragment (Fig. 5D and E, lanes 13). Southern blot analysis indicated that plasmid DNA from this colony did not contain MME or the PyV origin (Fig. 5D and E, compare lane 13 with lanes wt input and input). The results of these experiments strongly suggest that the cell lines we analyzed carry the input vectors as an extrachromosomal element.
Measurement of the episomal plasmid loss using flow cytometry analysis. Maintenance of plasmids containing the PyV core origin, MME, selection marker (Geneticin resistance), and green fluorescent protein marker (either long-half-life EGFP or short-half-life d1EGFP) was analyzed by flow cytometry. Transfection of these plasmids (schematically presented in Fig. 6A), into the COP5E2/Puro cell line resulted in efficient transient replication of these plasmids, which could be detected by Southern blot analysis (data not shown) as well as followed indirectly by measuring the fluorescence of plasmid-encoded EGFP. Two different variants of the EGFP protein marker were used comparatively to avoid potential problems coming from by-fluorescence of long-half-life EGFP in the case of short-term experiments. Transfected cells were grown in continuous culture in the presence or absence of Geneticin for up to 96 days. The cells were passaged every second day (every day when grown without selection), assuring active cell division. During each passage 100,000 to 200,000 cells were taken for analysis and the proportion of EGFP-positive cells was measured by flow cytometry. The percentage of cells (COP5E2/Puro) expressing EGFP above background (the fluorescence signal in the EGFP detection channel is higher than the cellular autofluorescence) was calculated for each transfected cell culture at each time point. Without Geneticin selection the percentage of the EGFP-fluorescent cells decreased quite rapidly. Eleven days after transfection without selection few EGFP-positive cells could be detected using fluorescence-activated cell sorter analysis compared to the initial approximately 50% EGFP-positive cells (Fig. 6B and C). Selection of the COP5E2/Puro cells transfected with the plasmid carrying the Geneticin resistance marker resulted in a cell culture which had nearly 100% EGFP-positive cells in the case of the plasmid expressing long-half-life EGFP and approximately 50% when the plasmid expressed short-half-life d1EGFP (Fig. 6B and C). The percentage of EGFP-positive cells stayed constant for more than 20 cell generations, indicating that these cells are capable of long-term maintenance of episomal genetic elements that contain the PyV core origin and MME. When the Geneticin selection was removed, the percentage of EGFP-positive cells decreased from 90% to approximately 1% in 55 days (from 64% to 2.4% in the case of d1EGFP in 37 days). In the case of integration of the episome the percentage of the EGFP-fluorescent cells remains constant even when the selection is removed (41). In order to characterize the kinetics of loss of the episomes, the rate of loss for each episomal construct during nonselective conditions was calculated for two independent experiments (Table 1, series 1 and 2). Two control plasmids, pEGFP-C1 and pd1EGFP-N1 (control plasmids from Clontech lacking the replication origin and MME), were used in the flow cytometry study to provide a comparison to the normal rate of loss of the episomes in the COP5E2/Puro cell line. After COP5E2/PuroMMEG cells were grown for 55 days and COP5E2/PuroMMEG cells for 37 days without selection, 1% of the cells still contained the episome, as indicated by the flow cytometry analysis. The reapplication of Geneticin selection at this point soon restored the proportion of EGFP-expressing cells to the initial level (Fig. 6B and C).
Comparison of the segregation/partitioning effects provided by the BPV1 MME and EBV FR elements to the PyV core origin plasmid. To compare the effects of BPV1 MME and EBV FR-based elements on the segregation/partitioning of the PyV replication origin construct, we constructed the EBNA1-expressing COP5 cell line and PyV core origin and FR-containing reporter plasmid that was similar to those used in the E2/MME analysis described above (Fig. 6A). EBNA1 activated PyV core origin replication in an FR-dependent manner (data not shown). The long-term maintenance of a transfected reporter plasmid (pFRG) containing the PyV core origin, FR element, Geneticin selection marker, and expression cassette for d1EGFP was monitored by flow cytometry. In this case, the replication function of the plasmid is provided by the PyV core origin and LT protein and the segregation/partitioning function is provided by the FR element and EBNA1 protein. Transfected cells were grown in continuous culture in the presence or absence of Geneticin for up to 75 days. Selection of the transfected COP5EBNA1/PuroFRG for Geneticin resulted in a cell culture which had approximately 40% d1EGFP-positive cells (Fig. 6D). When the Geneticin selection was removed, the percentage of d1EGFP-positive cells decreased from 40% to 1% in 30 days. When Geneticin selection on the COP5E2/PuropFRG cell line was restored at this point, the proportion of EGFP-expressing cells increased back to the initial level (Fig. 6D). These results are, in principle, identical to those obtained from the similar experiments with the E2/MME-dependent segregation/partitioning system described in the previous section (Fig. 6B and C). Therefore, EBNA1/FR elements and E2/MMEs confer comparable segregation/partitioning functions on the PyV core origin reporter plasmids in the analyzed cell model.
To exclude the possibility that the loss of EGFP fluorescence is due to inactivation of the promoter of EGFP, we also analyzed the DNA content in the cells. After removal of Geneticin selection total DNA was extracted from cells and digested with MluI (linearizes pMMEG and pFRG plasmids) and DpnI. Equal amounts of total DNA were then analyzed using Southern blotting with a radioactively labeled probe against the pMMEG or pFRG plasmid. As presented in Fig. 7 the loss of the episomal plasmid DNA from the cells grown without Geneticin selection correlates with the flow cytometry analysis. On the other hand, these results indicate that EGFP fluorescence was indeed measured from plasmids which exist in the episomal state. In the case of plasmid integration the hybridization signals remains constant.
DISCUSSION
The E2/MME works efficiently as a partitioning/segregation determinant also with a heterologous replicon. Prior to this study the only episomal maintenance element tested together with its nonnative replication origin was EBNA1/FR linked to different cellular replication origins. In these cases as well as in the case of replication of EBV OriP all the enzymatic activities required for replication are provided by the host cell. We have developed a system to study the mechanism of stable replication of a plasmid containing different maintenance elements in combination with their nonnative origins. More specifically, we have examined the functioning of E2/MME and EBNA1/FR maintenance elements in conjunction with the PyV replicon in the present study. As shown on Fig. 3A, the establishment of stable maintenance in this model depends on the simultaneous presence of E2 and MME. The plasmid containing 10 E2 BS and the PyV core origin is still present in the cell population 1 month after the initial enrichment of the transfected-reporter-carrying population based on antibiotic resistance selection. The same plasmid is efficiently maintained in the cells even if no selection is applied. Analysis of episomal DNA with nicking enzyme Nb.Bpu10I and rescue onto bacteria (Fig. 5) indicate that at least a majority of the stably maintained plasmids are not integrated and exist in an episomal state in the cells. Southern blot analysis and plasmid rescue assay indicated that some cells carry an oligomerized reporter plasmid (Fig. 5). Analysis of rescued plasmids showed that, in roughly one-half of the analyzed colonies, some plasmid rearrangement occurred (Fig. 5C). It has been shown that spontaneous sequence deletion by homologous recombination in the bacterial cell may occur (46). We have also noticed that highly palindromic sequences, like those of E2 BS-containing plasmids, are unstable during bacterial manipulations (A. Abroi, I. Ilves, and K. Janikson, personal communications). Whether these rearrangements in our present experiment originated in eukaryotic or in bacterial cells remains unclear; it is, however, important to note that all rearranged plasmids still carried the PyV minimal origin and BPV1 maintenance element (Fig. 5D and E). In addition, a control plasmid rescue assay with total DNA from cells carrying reporter plasmids with the wt PyV origin revealed only one colony whose low-molecular-weight DNA did not contain an MME or PyV origin (Fig. 5D and E). Thus, the E2/MME provides the stable episomal maintenance function not only to a BPV1-based replicon but also to a PyV-based replicon. Our study shows that mitotic chromatin attachment determinants of both BPV1 and EBV can provide the stable episomal maintenance function to heterologous viral replicons in addition to their native ones.
The episomally maintained reporter plasmids analyzed in our study do not cause major growth disadvantages for transfected cells and have a relatively low loss rate (Fig. 3 and Table 1). Previous studies have shown that, in the established cell lines carrying OriP, the plasmid loss is 2 to 8% per cell generation. However, the average plasmid copy number of OriP decreases more than 100-fold during the first 2 weeks after transfection into the cells expressing EBNA1 (25). In our system, certainly some decrease in the average copy number can be observed, but definitely not as fast. The quantitative aspects of the establishment of BPV1 stable episomal maintenance were not addressed in the present study. However, our data suggest that this process may be even more effective than in the case of EBNA1/OriP (25).
Overreplication of the wt PyV origin disrupts the establishment of stable maintenance. On the other hand, E2/MME cannot provide stable episomal maintenance to plasmids with the PyV wt origin, even under selective pressure (Fig. 3 and 4B). PyV exhibits a replication pattern that is uncoupled from the regulatory mechanisms of the host cell, so that each viral genome replicates many times within each cell cycle. The complete PyV origin includes transcriptional and replicational enhancer sequences, which dictate the origin activity and the efficiency of replication in specific cells by determining the availability of the replication factors and nucleotides. Papillomavirus origin replication control is similar to PyV replication in the first, amplificational phase of replication. However, in the latent-replication phase the copy number control mechanism is applied, which assures the controlled initiation of replication of the episomal viral genome in the latent-replication phase. We show that replacement of the wt PyV enhancer with 5 or 10 synthetic BS for the BPV1 E2 protein can replace replication enhancer function and makes it dependent on E2 protein. These results are in accordance with earlier reports by Nilsson et al. (29) and Abroi et al. (submitted) that at least two E2 BS are required to activate the PyV core origin. It is interesting to note that adding 5 or 10 E2 BS to the PyV wt origin did not cause additional replication activation. This fact supports the idea that replication from the episomal viral replication origins has a certain maximal threshold level in the host cell, which can be achieved by the presence (or addition) of strong enhancer elements. Further enhancement of the replication is not possible, even if more enhancer elements are added. It could be explained by the limiting levels of cellular replication factors or by the saturation of the nucleus with active genetic elements in the form of replication intermediates. This observation is also supported by the replication kinetics data, showing that the maximal level of replication by the wt PyV origin is achieved already 24 h posttransfection and that there is no more increase later; rather some decrease is observed (A. Abroi, unpublished data). At the same time, the replication signal of the PyV core origin or E2-dependent core origin increases in time. The toxic effect of the overreplication of the episomes on the cell can be suggested, as we observed many floating dead cells after transfection with PyV wt origin constructs. Thus, even though additional experimental data are needed to clarify this point, the most likely explanation for the inability of E2/MME to provide stable episomal maintenance to PyV wt origins is the cellular response against the high level of replication intermediates and/or the high level of replication itself.
The stable maintenance element from an EBV replicon that replicates strictly once per cell cycle can confer stable episomal maintenance properties to replication origin constructs derived from lytically replicating virus. The replication origins of BPV1 and PyV are fired several times during their amplificational replication in the single S phase of the host cell cycle, and the respective initiator proteins, E1 and LT, have many biochemical and structural similarities. However, these viruses have different time courses of productive infection. PyV is a lytic virus, and thus the viral DNA does not need to be stably maintained. During the stable replication of the BPV1 genome or URR, the origin is not restricted to precisely once replication round in each cell cycle (30, 31, 36). At the same time the EBV latent origin OriP replicates strictly once per cell cycle, exactly the same way as chromosomal DNA. Thus, in these terms, the replication modes of PyV and OriP are completely different. As shown on Fig. 6, the BPV1 E2/MME and EBV EBNA1/FR element can provide a stable maintenance function to the PyV core origin plasmids in the presence of viral trans factors. Our data presented here suggest that stable maintenance of the episomes provided by the function of MME orthe FR element is not connected to the mode of replication of the episome. The FR element can provide a stable maintenance function to several types of origins, in its natural context within the EBV latent origin OriP, in the plasmids where the chromosomal origin of replication from cellular DNA is linked to the FR element, and in our hybrid replicon together with the PyV core origin (22, 41). These data also show that the replication function is not connected to the stable-maintenance function of the virus; replication origins of different viruses can be combined with heterologous stable-maintenance elements without the loss of either function. The cellular receptors of BPV1 E2 protein and EBV EBNA1 protein, which link the episomes to mitotic host chromatin and therefore provide the stable-maintenance function, are different (20, 21, 33, 44, 45, 47). And, most likely, E2/MME- and EBNA1/FR-dependent plasmids are localized on chromosomes in different places. Our data presented here indicate that the different localizations of the episome on mitotic chromosomes do not interfere with the replication of the PyV minimal replication origin.
The chromatin attachment/partitioning function, not activation of replication, is responsible for stable episomal maintenance of a heterologous origin. Structural intactness of the E2 protein is very important in order to provide the MME-dependent partitioning function. A recent study from our laboratory showed that single point mutations might affect the chromatin attachment of E2 protein or its ability to mediate chromatin tethering of URR reporters even if the effect on replication initiation or transcription activation is relatively modest (1). E2 mutant proteins E39A and R68A were used to analyze the role of the transcription activation properties of the E2 protein in its functioning as a trans factor for stable episomal maintenance of a PyV-derived replicon. These E2 mutants were shown to be inactive in chromatin attachment, URR plasmid tethering, and segregation/partitioning assays in our studies using Chinese hamster ovary cells (1); they were, however, functional in Brd4 binding and chromatin binding assays using CV1 cells derived from African green monkey kidney cells (7). These mutant E2 proteins though seemed to be nonfunctional in supporting long-term episomal maintenance of the hybrid MME/PyV core ori plasmids in C127 mouse cells, as shown in this paper. The apparent contradiction in the features of these E2 mutants may be attributed to the difference in the species of cells used to measure the functions of the E2 mutants. The interaction of these E2 mutants could be different with mouse, hamster, and monkey Brd4 proteins, which could be due to the difference in the sequences between Brd4 proteins of the different species. The cell lines used in our assays have been shown to be functional in supporting BPV1 origin replication, segregation/partitioning, and long-term maintenance of episomal replication, which is not necessarily the case with CV1 cells.
In addition, we tested the stable replication of the E2/MME-dependent PyV replicon also in cell lines expressing LT and VP16-E2 or p53-E2 (containing the activation domains from VP16 and p53, respectively). They both are very potent transactivators and capable of activating the replication of the PyV core origin in transient assays. However, none of these mutant or hybrid E2 variants was capable of ensuring effective stable episomal maintenance, in the case of VP16-E2 and p53-E2 not even under selective conditions (Fig. 4B and data not shown). Our results show clearly that chromatin attachment, but not transactivation and the consequent replication initiation activity of E2 protein, is essential to provide stable maintenance for chimeric constructs used in this study and that random partitioning of the episomal plasmids cannot provide a reliable mechanism for stable episomal maintenance of the plasmids even in the presence of selection for the episomal selection marker. These results suggest that MME-mediated partitioning in conjunction with the PyV origin or its natural BPV1 origin is achieved by using the same strategy, i.e., through chromatin attachment.
Rate of loss of episomal plasmids. In the present study we have analyzed the episomal maintenance of plasmids containing the PyV minimal replication origin and either BPV1 MME or the EBV FR element (Fig. 6A) in cells where the appropriate viral trans factors (either PyV LT and BPV1 E2 or PyV LT and EBV EBNA1 protein) were stably expressed. In the case of plasmids containing the PyV minimal replication origin and BPV1 MME, the rate of episomal loss was 6% per cell division in the absence of selection. For plasmids containing the PyV minimal origin and EBV FR element, the rate of episomal plasmid loss was higher (13%), but it is still significantly lower than 22 to 30%, which we observed in the case of control plasmids (pEGFP-C1 and pd1EGFP-N1) without a eukaryotic replication origin and segregation elements. The rate of loss of plasmids containing the PyV minimal replication origin and FR element (pFRG) is also different from the previously published results on the rate of loss of several replicating plasmids that contained the FR element as a stable maintenance factor, where it was 2.1 to 7.8% (41); however, it is very similar to the 15% rate of loss previously estimated for OriP-containing plasmids (17). The difference in the rate of loss may be due to the differences in the expression level of EBNA1, the configuration of the test plasmids used, or the nature of the chromatin receptor for EBNA1, because in our experiments the mouse cell line COP5, not human cells, was used. The difference between the d1EGFP-positive and EGFP-positive cells in cell culture grown under Geneticin selection is probably due to the sensitivity of detection. As the half-life of the d1EGFP protein is 1 hour and as d1EGFP does not accumulate in the cells, the level of d1EGFP in cells with lower reporter plasmid copy number may probably be insufficient for its detection from the autofluorescence background and thus a certain fraction of cells which in fact carry the reporter will probably be regarded as "EGFP negative". However, the fact that both long- and short-half-life EGFP reporters gave similar rates of loss indicates that our method for measuring the rate of plasmid loss is adequate.
In conclusion, all these data together indicate that the maintenance elements from different DNA viruses are interchangeable with each other and can work in conjunction with different replicons, even with those from lytically replicating viruses. However, in heterologous systems as well as in native configurations, a certain loss of plasmid exists. In order to compensate the loss of episomes, viruses have evolved systems to accelerate the host cell proliferation compared to uninfected cells.
ACKNOWLEDGMENTS
We thank Anne Kalling for technical assistance, Ivar Ilves for careful reading of the manuscript, Kertu Rünkorg for constructing the COP5E2/Neo cell line, Michael Botchan for VP16-E2 expression construct, Mari Sepp for the BPV1-transfected C127 cell line, and G?ran Magnusson for the COP5 cell line and PyV origin constructs.
This study was supported in part by grants 4475, 4476, 5999, and 5998 from the Estonian Science Foundation, grant INTNL 55000339 from the Howard Hughes Medical Institute, grant CT96-0918 from the European Union, and target financial project 0182566s03.
REFERENCES
Abroi, A., I. Ilves, S. Kivi, and M. Ustav. 2004. Analysis of chromatin attachment and partitioning functions of bovine papillomavirus type 1 E2 protein. J. Virol. 78:2100-2113.
Abroi, A., R. Kurg, and M. Ustav. 1996. Transcriptional and replicational activation functions in the bovine papillomavirus type 1 E2 protein are encoded by different structural determinants. J. Virol. 70:6169-6179.
Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, and K. Struhl (ed.). 1998. Current protocols in molecular biology. John Wiley & Sons, Inc., New York, N.Y.
Ballestas, M. E., P. A. Chatis, and K. M. Kaye. 1999. Efficient persistence of extrachromosomal KSHV DNA mediated by latency-associated nuclear antigen. Science 284:641-644.
Ballestas, M. E., and K. M. Kaye. 2001. Kaposi's sarcoma-associated herpesvirus latency-associated nuclear antigen 1 mediates episome persistence through a cis-acting terminal repeat (TR) sequence and specifically binds TR DNA. J. Virol. 75:3250-3258.
Baru, M., M. Shlissel, and H. Manor. 1991. The yeast GAL4 protein transactivates the polyomavirus origin of DNA replication in mouse cells. J. Virol. 65:3496-3503.
Baxter, M. K., M. G. McPhillips, K. Ozato, and A. A. McBride. 2005. The mitotic chromosome binding activity of the papillomavirus E2 protein correlates with interaction with the cellular chromosomal protein, Brd4. J. Virol. 79:4806-4818.
Bostock, C. J., and R. C. Allshire. 1986. Comparison of methods for introducing vectors based on bovine papillomavirus-1 DNA into mammalian cells. Somatic Cell Mol. Genet. 12:357-366.
Clertant, P., and F. Cuzin. 1980. Initiation of polyomavirus DNA replication in vitro and its dependence on the viral gene A protein. Nucleic Acids Res. 8:4377-4392.
Cole, C. N. C., and S. D. Conzen. 2001. Polyomaviridae: the viruses and their replication, p. 2141-2174. In D. M. Knipe and P. M. Howley (ed.), Fields' virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Cotter, M. A., II, and E. S. Robertson. 1999. The latency-associated nuclear antigen tethers the Kaposi's sarcoma-associated herpesvirus genome to host chromosomes in body cavity-based lymphoma cells. Virology 264:254-264.
Delecluse, H. J., S. Bartnizke, W. Hammerschmidt, J. Bullerdiek, and G. W. Bornkamm. 1993. Episomal and integrated copies of Epstein-Barr virus coexist in Burkitt lymphoma cell lines. J. Virol. 67:1292-1299.
de Villiers, J., W. Schaffner, C. Tyndall, S. Lupton, and R. Kamen. 1984. Polyomavirus DNA replication requires an enhancer. Nature 312:242-246.
Dhar, S. K., K. Yoshida, Y. Machida, P. Khaira, B. Chaudhuri, J. A. Wohlschlegel, M. Leffak, J. Yates, and A. Dutta. 2001. Replication from oriP of Epstein-Barr virus requires human ORC and is inhibited by geminin. Cell 106:287-296.
Francke, B., and W. Eckhart. 1973. Polyoma gene function required for viral DNA synthesis. Virology 55:127-135.
Guo, Z. S., and M. L. DePamphilis. 1992. Specific transcription factors stimulate simian virus 40 and polyomavirus origins of DNA replication. Mol. Cell. Biol. 12:2514-2524.
Hung, S. C., M. S. Kang, and E. Kieff. 2001. Maintenance of Epstein-Barr virus (EBV) oriP-based episomes requires EBV-encoded nuclear antigen-1 chromosome-binding domains, which can be replaced by high-mobility group-I or histone H1. Proc. Natl. Acad. Sci. USA 98:1865-1870.
Ilves, I., S. Kivi, and M. Ustav. 1999. Long-term episomal maintenance of bovine papillomavirus type 1 plasmids is determined by attachment to host chromosomes, which is mediated by the viral E2 protein and its binding sites. J. Virol 73:4404-4412.
Kanda, T., M. Otter, and G. M. Wahl. 2001. Coupling of mitotic chromosome tethering and replication competence in Epstein-Barr virus-based plasmids. Mol. Cell. Biol. 21:3576-3588.
Kapoor, P., and L. Frappier. 2003. EBNA1 partitions Epstein-Barr virus plasmids in yeast cells by attaching to human EBNA1-binding protein 2 on mitotic chromosomes. J. Virol. 77:6946-6956.
Kapoor, P., K. Shire, and L. Frappier. 2001. Reconstitution of Epstein-Barr virus-based plasmid partitioning in budding yeast. EMBO J. 20:222-230.
Krysan, P. J., and M. P. Calos. 1993. Epstein-Barr virus-based vectors that replicate in rodent cells. Gene 136:137-143.
Kurg, R., J. Parik, E. Juronen, T. Sedman, A. Abroi, I. Liiv, U. Langel, and M. Ustav. 1999. Effect of bovine papillomavirus E2 protein-specific monoclonal antibodies on papillomavirus DNA replication. J. Virol. 73:4670-4677.
Lehman, C. W., and M. R. Botchan. 1998. Segregation of viral plasmids depends on tethering to chromosomes and is regulated by phosphorylation. Proc. Natl. Acad. Sci. USA 95:4338-4343.
Leight, E. R., and B. Sugden. 2001. Establishment of an oriP replicon is dependent upon an infrequent, epigenetic event. Mol. Cell. Biol. 21:4149-4161.
Li, R., J. Knight, G. Bream, A. Stenlund, and M. Botchan. 1989. Specific recognition nucleotides and their DNA context determine the affinity of E2 protein for 17 binding sites in the BPV-1 genome. Genes Dev. 3:510-526.
Lupton, S., and A. J. Levine. 1985. Mapping genetic elements of Epstein-Barr virus that facilitate extrachromosomal persistence of Epstein-Barr virus-derived plasmids in human cells. Mol. Cell. Biol. 5:2533-2542.
Morgenstern, J. P., and H. Land. 1990. Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res. 18:3587-3596.
Nilsson, M., M. Forsberg, Z. Y. You, G. Westin, and G. Magnusson. 1991. Enhancer effect of bovine papillomavirus E2 protein in replication of polyomavirus DNA. Nucleic Acids Res. 19:7061-7065.
Piirsoo, M., E. Ustav, T. Mandel, A. Stenlund, and M. Ustav. 1996. Cis and trans requirements for stable episomal maintenance of the BPV-1 replicator. EMBO J. 15:1-11.
Ravnan, J. B., and S. N. Cohen. 1995. Transformed mouse cell lines that consist predominantly of cells maintaining bovine papilloma virus at high copy number. Virology 213:526-534.
Schvartzman, J. B., S. Adolph, L. Martin-Parras, and C. L. Schildkraut. 1990. Evidence that replication initiates at only some of the potential origins in each oligomeric form of bovine papillomavirus type 1 DNA. Mol. Cell. Biol. 10:3078-3086.
Sears, J., M. Ujihara, S. Wong, C. Ott, J. Middeldorp, and A. Aiyar. 2004. The amino terminus of Epstein-Barr virus (EBV) nuclear antigen 1 contains AT hooks that facilitate the replication and partitioning of latent EBV genomes by tethering them to cellular chromosomes. J. Virol. 78:11487-11505.
Sharma, R. C., and R. T. Schimke. 1996. Preparation of electrocompetent E. coli using salt-free growth medium. BioTechniques 20:42-44.
Skiadopoulos, M. H., and A. A. McBride. 1998. Bovine papillomavirus type 1 genomes and the E2 transactivator protein are closely associated with mitotic chromatin. J. Virol. 72:2079-2088.
Ten Hagen, K. G., J. B. Ravnan, and S. N. Cohen. 1995. Disparate replication properties of integrated and extrachromosomal forms of bovine papilloma virus in ID13 cells. J. Mol. Biol. 254:119-129.
Tsurumi, T., M. Fujita, and A. Kudoh. 2005. Latent and lytic Epstein-Barr virus replication strategies. Rev. Med. Virol. 15:3-15.
Tyndall, C., G. La Mantia, C. M. Thacker, J. Favaloro, and R. Kamen. 1981. A region of the polyoma virus genome between the replication origin and late protein coding sequences is required in cis for both early gene expression and viral DNA replication. Nucleic Acids Res. 9:6231-6250.
Ustav, M., and A. Stenlund. 1991. Transient replication of BPV-1 requires two viral polypeptides encoded by the E1 and E2 open reading frames. EMBO J. 10:449-457.
Ustav, M., E. Ustav, P. Szymanski, and A. Stenlund. 1991. Identification of the origin of replication of bovine papillomavirus and characterization of the viral origin recognition factor E1. EMBO J. 10:4321-4329.
Wade-Martins, R., J. Frampton, and M. R. James. 1999. Long-term stability of large insert genomic DNA episomal shuttle vectors in human cells. Nucleic Acids Res. 27:1674-1682.
Waldenstrom, M., K. Schenstrom, K. Sollerbrant, and L. Hansson. 1992. Replication of bovine papillomavirus vectors in murine cells. Gene 120:175-181.
Wang, E. H., and C. Prives. 1991. DNA helicase and duplex DNA fragment unwinding activities of polyoma and simian virus 40 large T antigen display similarities and differences. J. Biol. Chem. 266:12668-12675.
Wu, H., D. F. Ceccarelli, and L. Frappier. 2000. The DNA segregation mechanism of Epstein-Barr virus nuclear antigen 1. EMBO Rep. 1:140-144.
Yates, J. L., and N. Guan. 1991. Epstein-Barr virus-derived plasmids replicate only once per cell cycle and are not amplified after entry into cells. J. Virol. 65:483-488.
Yates, J. L., N. Warren, and B. Sugden. 1985. Stable replication of plasmids derived from Epstein-Barr virus in various mammalian cells. Nature 313:812-815.
You, J., J. L. Croyle, A. Nishimura, K. Ozato, and P. M. Howley. 2004. Interaction of the bovine papillomavirus E2 protein with Brd4 tethers the viral DNA to host mitotic chromosomes. Cell 117:349-360.(Toomas Silla, Ingrid H??l)
Estonian Biocentre, 23 Riia St., Tartu 51010, Estonia
Department of Biomedical Technology, Institute of Technology, Tartu University, 21 Vanemuise St., Tartu 51010, Estonia
ABSTRACT
Bovine papillomavirus type 1 (BPV1), Epstein-Barr virus (EBV), and human herpesvirus 8 genomes are stably maintained as episomes in dividing host cells during latent infection. The mitotic segregation/partitioning function of these episomes is dependent on single viral protein with specific DNA-binding activity and its multimeric binding sites in the viral genome. In this study we show that, in the presence of all essential viral trans factors, the segregation/partitioning elements from both BPV1 and EBV can provide the stable maintenance function to the mouse polyomavirus (PyV) core origin plasmids but fail to do so in the case of complete PyV origin. Our study is the first which follows BPV1 E2- and minichromosome maintenance element (MME)-dependent stable maintenance function with heterologous replication origins. In mouse fibroblast cell lines expressing PyV large T antigen (LT) and either BPV1 E2 or EBV EBNA1, the long-term episomal replication of plasmids carrying the PyV minimal origin together with the MME or family of repeats (FR) element can be monitored easily for 1 month under nonselective conditions. Our data demonstrate clearly that the PyV LT-dependent replication function and the segregation/partitioning function of the BPV1 or EBV are compatible in certain, but not all, configurations. The quantitative analysis indicates a loss rate of 6% per cell, doubling in the case of MME-dependent plasmids, and 13% in the case of FR-dependent plasmids in nonselective conditions. Our data clearly indicate that maintenance functions from different viruses are principally interexchangeable and can provide a segregation/partitioning function to different heterologous origins in a variety of cells.
INTRODUCTION
Several eukaryotic DNA viruses maintain their genomes as extrachromosomal multicopy nuclear episomes in proliferating host cells. Such episomal maintenance is characteristic of latent infection of bovine papillomavirus type 1 (BPV1), Epstein-Barr virus (EBV), and Kaposi's sarcoma-associated human herpesvirus 8 (HHV8). Two functions of the viral genome are absolutely critical for extrachromosomal maintenance in dividing cells: viral genome replication during the S phase and proper segregation and partitioning of the replicated genomes into daughter cells during host cell mitosis. For BPV1 and two members of the gammaherpesvirus family, EBV and HHV8, effective segregation of viral genomes into daughter cells and nuclear retention during mitosis are mediated through a single viral protein serving as a molecular linker, which attaches viral genomes to the host mitotic chromosomes (4, 11, 12, 19, 24, 35). This linker protein is viral regulatory protein E2 for BPV1 (18, 24, 35), viral transactivator EBNA1 for EBV (19), and viral transcription repressor LANA1 for HHV8 (4, 5).
For the initiation of the DNA replication of a BPV1-based replicon in vivo, the minimal origin region in cis and two viral proteins, E1 and E2, in trans, are absolutely essential (39, 40). However, the minimal origin is not sufficient for stable extrachromosomal replication in dividing cells (30). An additional element, the minichromosome maintenance element (MME), ensures the long-term episomal persistence of the genome in the presence of viral E1 and the E2 proteins in the dividing cells (30). In the BPV1 genome, in total 17 E2 protein binding sites (BS) with different affinities for E2 can be identified; 12 of these are located in the noncoding upstream regulatory region (URR) (26). We have shown that, for efficient partitioning/segregation of the episomal plasmid, MME activity is provided by a sufficient number of high-affinity E2 BS (30). The function of multimeric E2 BS in the stable maintenance of the BPV1 genomes is to provide the anchoring function for the E2 protein, which therefore tethers MME-containing plasmids to mitotic chromosomes (18, 24). This linkage between the BPV1 genome and host chromatin ensures also that the viral genome is maintained in the nucleus when the nuclear membrane is reassembled during mitosis. In the case of EBV, the stable maintenance of replicated genomes is achieved due to the EBNA1 protein and family of repeats (FR) element, which is composed of multimeric EBNA1 protein binding sites (19, 27).
We have shown that both the BPV1 E2 protein-dependent MME (1) and EBV EBNA1-dependent FR (A. M?nnik, K. Janikson, and M. Ustav, unpublished data) segregation/partitioning and chromatin attachment activities function independently from replication of the plasmids (18). The stable-maintenance function of EBNA1/FR has been used to ensure long-time episomal maintenance for non-OriP origins, usually the cellular replication origins (22, 41). In the case of OriP, the enzymatic activity required for initiation of replication is the same as in cellular origins (14, 37). The E2/MME-dependent stable-maintenance function has not been tested with heterologous replication origins. In the present study, we have further examined the compatibility of viral segregation/partitioning elements with heterologous replication origins. For this purpose, different reporter plasmids were constructed that combine the E2/MME- and EBNA1/FR-based stable-maintenance function and different variants of the mouse polyomavirus (PyV) replication origin. The mouse polyomavirus is a lytic virus, which replicates its DNA very fast during productive infection. Infected cells contain up to 200,000 molecules of viral DNA, and the maximal copy number is reached about 50 h postinfection (10). The replication origin of PyV contains the transcription/replication enhancer responsible for the high level of replication (13). We tested the stable maintenance of plasmids that contained E2/MME in conjunction with the wild-type (wt) PyV or core (enhancerless) origin of replication in cell lines expressing PyV large T antigen (LT) and E2 or its mutants. Also the stable maintenance of a plasmid containing the FR and the PyV core origin was tested in cell lines expressing LT and EBNA1. The results from these experiments show convincingly that the segregation/partitioning functions of BPV1 and EBV can effectively be used for stable episomal maintenance of the PyV core origin. In addition, efficient chromatin attachment rather than a high level of activation of replication is required for stable episomal maintenance.
MATERIALS AND METHODS
Plasmids. For constructing hybrid replicons (Fig. 1B and C) containing the PyV origin (wild type or core origin), we used vector pUC19 as the basic backbone, where we cloned 1, 5, or 10 head-to-tail copies of high-affinity E2 BS 9. PyV wt and the core origin were amplified by PCR from vectors pmu1046/CAT and pmu1047/CAT (29) using primers Py4963 (5'-AGGGAGCTACTCCTGATG-3') and Py174 (5'-CTACCACCACTCCGACTT-3'). Amplified PyV origin fragments were digested with enzymes EheI and BclI and inserted between BamHI and HincII sites of the pUC19 vector containing different numbers of BPV1 E2 BS.
Hybrid replicons containing a Geneticin resistance gene (Fig. 1D) were established by replacing the URR in plasmid pNeoBgl40 (30) with PyV wt origin, core origin, or core origin with 10 E2 BS, which were amplified by PCR and digested with enzymes BamHI and Ecl136II and cloned into BamHI and HindIII sites in pNeoBgl40.
Three types of the enhanced green fluorescent protein (EGFP) marker containing plasmids were designed. First, a fragment comprising the PyV minimal origin and 10 E2 BS was added to a plasmid containing a Geneticin resistance marker (expressed from the simian virus 40 promoter). Then either an EGFP or destabilized green fluorescent protein (d1EGFP) marker was added (named either pMMEG or pMMEG plasmid; see Fig. 6A). EGFP expression cassettes, which are under the control of the cytomegalovirus promoter, were taken either from pEGFP-C1 or pd1EGFP-N1 plasmids (Clontech). For the third plasmid, the first EBV FR element was added to the pUC19 plasmid containing the PyV core origin. Then the fragment containing the PyV minimal replication origin and 10 copies of E2 BS 9 from plasmid pMMEG was replaced by the fragment containing the PyV minimal origin and EBV FR element (plasmid pFRG; see Fig. 6A).
Construction of cell lines. For construction of cell lines which express BPV1 wt E2 protein and its mutant forms E39A and R68A, the vector pBabePuro (28) was linearized using enzyme SalI and was ligated with an equal amount of E2 expression vectors (pCGE2, pCGE2/R68, pCGE2/E39 [2, 39]), which were linearized with XhoI endonuclease. One microgram of ligated hybrid plasmids was electroporated into the COP5 cell line (38). Electroporation experiments were preformed with a Bio-Rad Gene Pulser with capacitance and voltage settings of 975 μF and 220 V. For selection puromycin (2 μg/ml) was added. The expression of the proteins was analyzed by Western blotting.
A cell line expressing wt E2 and carrying a Geneticin selection cassette was constructed by the same protocol described above using vector pBabeNeo (28) instead of pBabePuro.
A cell line expressing PyV T antigens and EBV EBNA1 protein was generated as a result of transfection of the NotI-linearized plasmid pBabePuro/EBNA1 (EBNA1 coding sequence inserted into EcoRI/SalI sites in the pBabePuro vector) into the COP5 cell line and selection for puromycin (2 μg/ml). The expression of the proteins was analyzed by Western blotting. The cell line was named COP5EBNA1/Puro.
Cells and transfection. COP5 cells (38) and their derivatives COP5E2/Puro, COP5E2/Neo, COP5R68/Puro, COP5E39/Puro, and COP5EBNA1/Puro expressing PyV T antigens and BPV1 wt E2 or its mutant forms or EBNA1 were grown in Iscove's modified Dulbecco's medium (IMDM) supplemented with 10% fetal calf serum. For selection Geneticin (500 μg/ml) or puromycin (2 μg/ml) was added, depending on the selection marker. Electroporation experiments were performed with a Bio-Rad Gene Pulser with capacitance and voltage settings of 975 μF and 220 V, respectively.
COP5E2/Puro cells transfected with neomycin constructs were selected with Geneticin at 500 μg/ml. COP5E2/Neo cells cotransfected with pBabePuro (28) were selected with puromycin at 2 μg/ml. After transfection with 500 ng of plasmids carrying the Geneticin resistance marker and EGFP coding sequence, the COP5EBNA1/Puro cell line was grown in IMDM containing 500 μg/ml Geneticin (medium contained no puromycin).
Southern blot analysis. Total DNA was extracted from cells by following a standard protocol (3). Extraction of low-molecular-weight DNA from cells and analysis of origin construct levels in both low-molecular-weight- and total-DNA preparations were performed as described previously (30, 39). All restriction reactions included DpnI to eliminate bacterially methylated input DNA. In the case of the nicking reaction 0.2 units of nicking enzyme Nb.Bpu10I (Fermentas, Vilnius, Lithuania) was added. During episomal- or total-DNA studies always equal numbers of cells or equal amounts of DNA were loaded onto each lane. Specific probes were labeled with [32P]dCTP by random-decamer-primed synthesis using the DecaLabel kit (Fermentas). PyV origin- and MME-specific probes were made by PCR with [32P]dCTP. Hybridizing species were visualized by autoradiography. Radioactive signals on the blots were quantified on PhosphorImagerSI using ImageQuant software (Molecular Dynamics, Amersham Biosciences, Little Chalfont, United Kingdom).
Immunoprecipitation. Cells (1.5 x 107) were lysed with ice-cold 1% sodium dodecyl sulfate (SDS)-phosphate-buffered saline on ice, collected in a 15-ml tissue culture tube, and sonicated. From this step an aliquot for the Bradford assay was taken. SDS was diluted to 0.1% by adding ice-cold radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% deoxycholate, 0.1 mM dithiothreitol [DTT], 0.5 mM phenylmethylsulfonyl fluoride, protease inhibitors). The insoluble fraction was sedimented by centrifugation at 5,000 x g for 15 min. The soluble fraction was transferred to a new tube and incubated with 5H4, 3E8, 1E4, and 3F12 antibodies (22) overnight at 4°C. Then protein G-Sepharose (Amersham Biosciences) was added and incubated for 1 h. Sepharose beads were washed three times with RIPA buffer and resuspended in SDS loading buffer and subjected to immunoblotting analysis with horseradish peroxidase-conjugated 5E11 (subclone of MAb 3F12) antibody (Quattromed AS, Tartu, Estonia).
Immunoblotting. Total protein from the same number of cells lysed in standard loading buffer supplemented with 100 mM DTT was separated by electrophoresis on an 8% polyacrylamide-SDS gel and transferred to an Immobilon-P membrane (Millipore). Antibody 1E4 (23) was used to detect E2 proteins. Antibodies BM3167 and BM1083 (DPC Biermann) were used to detect EBNA1 protein. Peroxidase-conjugated goat anti-mouse antibody and the enhanced chemiluminescence detection kit (ECL Western blotting reagents; Amersham Biosciences) were used for subsequent development of the blot, using a standard protocol provided by the supplier.
Plasmid rescue assay. Two micrograms of uncut genomic DNA was electrotransformed into Escherichia coli strain DH10B. The electrocompetent cells were prepared as described previously (34), and the transformations were performed using a Pulser apparatus and 2-mm electroporation cuvette (Bio-Rad Laboratories, Hercules, CA) according to the manufacturer's instructions. The cells were recovered by centrifugation and were grown on medium containing ampicillin at 100 μg/ml. Plasmid DNA from single colonies was purified and analyzed using restriction endonucleases.
Flow cytometry analysis. EGFP expression was analyzed by flow cytometry using a Becton Dickinson FACSCalibur flow cytometer with associated CellQuest software. One hundred thousand to 200,000 signals were analyzed from each sample. The threshold for autofluorescence was set to 99% of the signals from the mock-transfected control cells. All the signals above the threshold were considered to correspond to EGFP-positive cells. For calculating the episomal rates of loss in Table 1, EGFP expression data were analyzed on days 0 and 12 (pEGFP-C1 and pd1EGFP-N1), on days 0 and 55 for pMMEG, on days 0 and 37 for pMMEG, and on days 0 and 30 for pFRG (day 0 is the time point when selection was removed). For this calculation a first-order rate-of-loss model was used: rate of loss () = (–1/t)(ln Nt/N0) (41), where N0 is the percentage of green cells at the beginning of the experiment at nonselective conditions and Nt is the percentage of green cells after t generations.
RESULTS
BPV1 E2 protein and its multimeric binding sites activate the replication of PyV core origin and provide the segregation/partitioning function to the origin plasmid. The BPV1 E2 protein is a multifunctional protein which is involved in transcriptional regulation and viral DNA replication and segregation. It has been shown that, for stable episomal replication of BPV1, E1 and E2 proteins and the MME, which consists of multimeric E2 BS, are required (30). E2 protein can also activate the transient replication of the PyV core origin in vivo in E2 multimeric BS-dependent fashion (29). We decided to study if the BPV1 E2 BS in the hybrid PyV origin can, in addition to the activation of the initiation of replication, also provide the long-term maintenance function to the PyV-derived replicator in cells expressing LT and E2 protein.
Replication of the PyV origin requires LT as the only viral replication factor; all other components are derived from the host cell (9, 15). LT is an origin recognition factor and DNA helicase, thus directly participating in initiation and elongation of the replication of the viral origin (43). We constructed mouse cell lines expressing PyV LT and the BPV1 E2 protein using the cell line COP5, which constitutively produces LT from the integrated replication-defective PyV genome (38). Individual colonies were allowed to expand in the presence of selection (puromycin or Geneticin), and PyV LT- and BPV1 E2-positive double-expression cell lines were identified and characterized. The cell lines expressing E2 protein at the highest level were used in further assays (referred to as COP5/E2/Puro or COP5/E2/Neo, selected for puromycin or Geneticin selection markers, respectively). The same approach was used for construction of cell lines which express mutant forms of the E2 proteins, E39A and R68A (referred to as COP5/E39/Puro and COP5/R68/Puro). As described earlier, both these mutants are at least partially functional in E2 BS-dependent transcriptional activation and initiation of the BPV1 origin replication as well as in activation of initiation of PyV core origin replication; however, they fail to attach to the host cell chromosomes and do not support segregation/partitioning of the MME plasmids (1, 2; A. Abroi et al., submitted for publication). Expression of the wt E2 protein in the cell lines COP5E2/Neo and COP5E2/Puro was verified using Western blot analysis (Fig. 2B). In both cell lines the expression of the wt E2 protein was maintained at a detectable level for a prolonged period without selection, which is essential for the study of the maintenance of the episomal plasmids. We determined the expression level of E2 proteins in the constructed COP5E2/Puro, COP5E39/Puro, and COP5R68/Puro cell lines, as well as the E2 expression level from C127 cells stably maintaining the BPV1 genome as an episome. Immunoprecipitation with E2-specific antibodies and following normalized immunoblotting showed that the expression level of E2 proteins in constructed cell lines is higher than the full-length-E2 expression level in the BPV1-transformed cell line (Fig. 2C). Thus, the E2 protein expression level in our constructed cell lines is not limiting in stable-maintenance experiments.
The constructed cell lines were used in further experiments to study the effect of the BPV1 E2 protein and E2 BS on initiation of replication and maintenance of the constructed plasmids. At first, we examined the chimeric origins, comprising the PyV wt origin (Fig. 1B) or the core origin (Fig. 1C) linked to different numbers of E2 BS in cell lines COP5/E2/Neo (expressing constitutively PyV LT and BPV1 E2) and COP5 (expressing constitutively PyV LT), respectively. Ninety-six hours after transfection strong replication signals of wt origin plasmids were detected in both cell lines (Fig. 3A and B, 4-day time points, lanes 1 to 4). Additional E2 BS had rather an inhibitory effect on the replication of the wt PyV origin. However, the E2 protein-dependent activation of replication was clearly detected in the cases when the PyV enhancerless core origin was linked to different numbers of E2 BS. Addition of one E2 BS had no effect on the initiation of replication of the core origin; however, the addition 5 or 10 E2 BS activated core origin replication to almost the wt origin replication level in an E2 protein-dependent fashion (compare Fig. 3A and B, lanes 5 to 9). In the COP5 cell line lacking E2 protein, the replication enhancer function of the E2 BS to the core origin cannot be detected (Fig. 3B). These results are in principal agreement with the data previously published by Nilsson et al. (29), showing that the replication of the PyV enhancerless origin can be activated by BPV1 E2 and its BS.
We further studied the stable episomal maintenance of different PyV origin-containing constructs in two experimental settings—first, without any selective pressure, and second, under puromycin selection—after cotransfection of the origin plasmids together with plasmid pBabePuro, which encodes the puromycin resistance marker, to select out transfected cells (see Materials and Methods for details). The episomal persistence of the PyV origin-containing plasmids was analyzed by Southern blotting. wt origin plasmids were lost from the cells under selective and nonselective conditions very fast in COP5E2/Neo and COP5 cells (Fig. 3A and B). However, the hybrid origins comprising the core origin and 5 or 10 E2 BS were capable of long-term persistence (11 and 34 days, at least 27 doublings) in the COP5E2/Neo cells, as analyzed by episomal DNA extraction or analysis of total DNA from the transfected cells. After 21 and 34 days without selective pressure, the only origin construct that was efficiently maintained as an episome was the hybrid of the core origin with 10 E2 BS (Fig. 3A, 21- and 34-day time points without selection, lane 5). For the PyV origin with five E2 BS, a weak replication signal was detected only at longer exposure (data not shown). We estimated the average copy number of the episomal plasmids in the culture using total-DNA Southern blotting (Fig. 3A, total DNA time points). The plasmid with 5 and 10 E2 BS had, on average, 5 and 17 copies per cell, respectively, after 34 days (with the assumption that all cells contain a replicon). These data indicate that E2 and its BS can provide the episomal maintenance function for chimeric PyV origin constructs that are otherwise lost from the cell population during cell growth.
Somewhat unexpectedly, we found that when the BPV1 segregation/partitioning element is linked to the PyV wt origin (short-term replication signals on Fig. 3A, 4-day time points, lanes 1 to 4) these replicons are not able to survive despite of their high-level replication (Fig. 3A, lanes 1 to 4). This could be due to the overreplication of the intact enhancer-containing origin plasmid, which could lead to cell death. Inspection of the transfected culture indicated that indeed the wt PyV origin plasmids induced extensive cell death at the later time points (data not shown).
Replication of the origin plasmids carrying an episomal selection marker. COP5E2/Puro cells were transfected with three different plasmids carrying, in addition to the origin, a Geneticin selection marker (Fig. 1D). The eukaryotic selection cassette in the plasmid makes it possible to select for cells carrying reporter plasmids in transfected cells in the presence of Geneticin. Transient transfection of COP5E2/Puro with neomycin reporter plasmids resulted in a strong replication signal for the wt origin construct (Fig. 4A, time points at 48 and 72 h, lane 1) compared to a much lower replication signal for the core origin construct (Fig. 4A, 48- and 72-h time points, lane 2). As expected, addition of 10 E2 BS to the core origin increased the transient-replication signal (Fig. 4A, 48- and 72-h time points, lane 3). The transfected cells were then grown in selective medium containing Geneticin, followed by a series of cell divisions comparatively with and without selection. After 2 months of cultivation, these pooled cell lines were analyzed for stable maintenance of the reporter plasmids using Southern blotting with a radioactively labeled probe against the PyV origin. Ten E2 BS containing the reporter plasmid could establish the extrachromosomal maintenance of autonomous episomes in E2-positive cells (Fig. 4B, lanes 3, analysis after 2 months). Removal of the selection reduced the replication signal, but it was still detectable in the episomal fraction after 2 months (Fig. 4B, lanes 3, 2-month time points), and even after 5 months (data not shown).
Efficient partitioning/segregation rather than a high level of activation of replication is required for stable episomal maintenance. We compared the stable episomal maintenance of the hybrid origins in the cell lines expressing wt E2 with that in cell lines expressing mutant forms of E2 carrying alanine substitutions of the conserved charged residues in the N-terminal domain. These mutants have been previously characterized in BPV1 replication, transactivation, sequence-specific DNA binding, and partitioning assays (1, 2). E2 mutants E39A and R68A (Fig. 2A) are inactive in the chromatin attachment functions and failed to mediate the segregation/partitioning of the BPV1 URR reporter plasmids but were still active in initiation of transient replication and in transcription, where their relative activity was comparable to wt E2 (1). We transfected the COP5 cells with expression constructs for the BPV1 E2 mutant forms R68A or E39A as well as a puromycin resistance marker; the puromycin-resistant clones were picked, expanded, and characterized for expression of the desired proteins (Fig. 2C). The cells with the best expression were selected for the subsequent assays. In the following short- and long-term replication assays, we used the reporter constructs carrying the selection marker that confers resistance to Geneticin selection (Fig. 1D). Both E2 mutant forms R68A and E39A activated PyV core origin replication in an E2 BS-dependent fashion in established cell lines (Fig. 4A, lane 3, 48- and 72-h time points). This suggested that E2 mutant forms R68A and E39A behave as efficiently in replication activation as wt E2 protein. The transfected cells were grown in the media with and without Geneticin selection for time periods up to 2 months. By this time, only the replication of reporter plasmid with 10 E2 BS added to PyV the core origin was detectable in cells (Fig. 4B, lanes 1 to 3). In wt E2-expressing cells, this signal was present in Geneticin-selected cells as well as in control cells without selection. On the other hand, in the case of E2 mutant forms E39A and R68A, only a very weak replication signal was observed in cells grown under Geneticin selection (Fig. 4B, lane 3). It is important to note that further cultivation up to 5 months resulted in the complete loss of the episomal signal in mutant E2 cell lines (data not shown). The same results as in cell lines expressing mutant E2 proteins R68A and E39A were obtained from the experiments with cell lines which express hybrid proteins VP16/E2 and p53/E2, where the whole transactivation domain of the E2 protein is replaced with the respective activation domain from VP16 or p53 protein, respectively (data not shown). These transactivation domains have been shown to activate PyV replication very efficiently (6, 16; A. Abroi unpublished data) and at least the VP16 activation domain does not support the plasmid partitioning function (1). These results showed that the chromatin attachment function of E2 protein is required to ensure stable maintenance of the chimeric PyV origin and that the replication activation function alone is not sufficient for stable episomal maintenance.
Episomal state of chimeric origins. A high mutation frequency, especially for recombination, is often associated with replication from the papillomavirus and polyomavirus origin-based vector systems (8, 42). Therefore, we decided to check for this possibility in our experimental model. Episomal DNA was extracted from COP5/E2/Puro-derived cell lines to analyze the presence and the status of episomally maintained hybrid origin plasmids (Fig. 5A). Hybridization analysis with a neomycin gene-specific probe of the linearized DNA from the PyV MME reporter-carrying cell line revealed mostly one very discrete band that migrated similarly to the unit size marker on the agarose gel (Fig. 5A, compare lane 1 to lane 9). The sample digested with a noncutter (enzyme with no restriction sites in plasmid DNA) gave a pattern where open circular (OC) and covalently closed circular (CCC) forms can be detected (Fig. 5A, compare lane 2 with marker lanes 10 and 11). However, the additional slower-moving reporter-specific bands were observed (Fig. 5A, lane 2 and 3). We suggest that the signals correspond to the oligomerized episomes; both forms have been shown to appear, for example, during the episomal maintenance of papillomavirus full-length genomes in several cell lines (32) and URR-containing plasmids in an E1/E2-positive cell line (30). To confirm that hybridization signals are not from integrated material, the samples were digested with a noncutter enzyme together with a nicking enzyme Nb.Bpu10I (Fig. 5A, lane 3). Nb.Bpu10I is a site- and strand-specific endonuclease that cleaves only one strand of DNA within its recognition sequence on a double-stranded DNA substrate. Thus, by nicking enzyme CCC DNA transfers to the OC form, a DNA mobility shift on agarose gel is observed (Fig. 5A, lane 12). Nb.Bpu10I does not change linear DNA mobility as can be observed in Fig. 5A, lane 8, which represents circular DNA digestion with a linearizing enzyme together with a nicking enzyme. However, no hybridization signal was observed on lanes containing the wt PyV origin (Fig. 5A, lanes 4 to 6). The results of this experiment show that the analyzed episomal DNA fraction contained a reporter plasmid which was sensitive to the nicking enzyme so hybridization signals were not from integrated material.
The presence of episomal DNA was also confirmed by plasmid rescue into E. coli, using the uncut total DNA from the E2- and LT-expressing cells that were carrying a reporter plasmid with the PyV core origin and 10 E2 BS. Analysis of uncut rescued plasmids showed an oligomerized pattern compared to input DNA (Fig. 5B). Restriction analysis of rescued plasmids by endonuclease BglI showed that, compared to input DNA some rearrangements in the plasmid backbone can be observed (Fig. 5C, lanes 1 to 3, 5, 6, 10, and 12). Thus, some cells carried plasmids with rearrangements. In addition, intact unarranged DNA forms were also detected (Fig. 5C, lanes 4, 7 to 9, and 11). To confirm that rescued plasmids still contain the PyV origin and MME, we analyzed the BglI digestion pattern by Southern blot analysis with an MME- or PyV-specific probe (Fig. 5D and E, respectively). Southern blot analysis showed that all rescued reporter plasmids contained the MME and PyV origin fragment (Fig. 5D and E). A plasmid rescue assay with total DNA (total DNA was extracted from cells whose episomal DNA is analyzed in Fig. 5A, lanes 4 to 6) from cells carrying the reporter plasmid with the wt PyV origin revealed only one colony, which we analyzed for the existence of the MME or PyV origin fragment (Fig. 5D and E, lanes 13). Southern blot analysis indicated that plasmid DNA from this colony did not contain MME or the PyV origin (Fig. 5D and E, compare lane 13 with lanes wt input and input). The results of these experiments strongly suggest that the cell lines we analyzed carry the input vectors as an extrachromosomal element.
Measurement of the episomal plasmid loss using flow cytometry analysis. Maintenance of plasmids containing the PyV core origin, MME, selection marker (Geneticin resistance), and green fluorescent protein marker (either long-half-life EGFP or short-half-life d1EGFP) was analyzed by flow cytometry. Transfection of these plasmids (schematically presented in Fig. 6A), into the COP5E2/Puro cell line resulted in efficient transient replication of these plasmids, which could be detected by Southern blot analysis (data not shown) as well as followed indirectly by measuring the fluorescence of plasmid-encoded EGFP. Two different variants of the EGFP protein marker were used comparatively to avoid potential problems coming from by-fluorescence of long-half-life EGFP in the case of short-term experiments. Transfected cells were grown in continuous culture in the presence or absence of Geneticin for up to 96 days. The cells were passaged every second day (every day when grown without selection), assuring active cell division. During each passage 100,000 to 200,000 cells were taken for analysis and the proportion of EGFP-positive cells was measured by flow cytometry. The percentage of cells (COP5E2/Puro) expressing EGFP above background (the fluorescence signal in the EGFP detection channel is higher than the cellular autofluorescence) was calculated for each transfected cell culture at each time point. Without Geneticin selection the percentage of the EGFP-fluorescent cells decreased quite rapidly. Eleven days after transfection without selection few EGFP-positive cells could be detected using fluorescence-activated cell sorter analysis compared to the initial approximately 50% EGFP-positive cells (Fig. 6B and C). Selection of the COP5E2/Puro cells transfected with the plasmid carrying the Geneticin resistance marker resulted in a cell culture which had nearly 100% EGFP-positive cells in the case of the plasmid expressing long-half-life EGFP and approximately 50% when the plasmid expressed short-half-life d1EGFP (Fig. 6B and C). The percentage of EGFP-positive cells stayed constant for more than 20 cell generations, indicating that these cells are capable of long-term maintenance of episomal genetic elements that contain the PyV core origin and MME. When the Geneticin selection was removed, the percentage of EGFP-positive cells decreased from 90% to approximately 1% in 55 days (from 64% to 2.4% in the case of d1EGFP in 37 days). In the case of integration of the episome the percentage of the EGFP-fluorescent cells remains constant even when the selection is removed (41). In order to characterize the kinetics of loss of the episomes, the rate of loss for each episomal construct during nonselective conditions was calculated for two independent experiments (Table 1, series 1 and 2). Two control plasmids, pEGFP-C1 and pd1EGFP-N1 (control plasmids from Clontech lacking the replication origin and MME), were used in the flow cytometry study to provide a comparison to the normal rate of loss of the episomes in the COP5E2/Puro cell line. After COP5E2/PuroMMEG cells were grown for 55 days and COP5E2/PuroMMEG cells for 37 days without selection, 1% of the cells still contained the episome, as indicated by the flow cytometry analysis. The reapplication of Geneticin selection at this point soon restored the proportion of EGFP-expressing cells to the initial level (Fig. 6B and C).
Comparison of the segregation/partitioning effects provided by the BPV1 MME and EBV FR elements to the PyV core origin plasmid. To compare the effects of BPV1 MME and EBV FR-based elements on the segregation/partitioning of the PyV replication origin construct, we constructed the EBNA1-expressing COP5 cell line and PyV core origin and FR-containing reporter plasmid that was similar to those used in the E2/MME analysis described above (Fig. 6A). EBNA1 activated PyV core origin replication in an FR-dependent manner (data not shown). The long-term maintenance of a transfected reporter plasmid (pFRG) containing the PyV core origin, FR element, Geneticin selection marker, and expression cassette for d1EGFP was monitored by flow cytometry. In this case, the replication function of the plasmid is provided by the PyV core origin and LT protein and the segregation/partitioning function is provided by the FR element and EBNA1 protein. Transfected cells were grown in continuous culture in the presence or absence of Geneticin for up to 75 days. Selection of the transfected COP5EBNA1/PuroFRG for Geneticin resulted in a cell culture which had approximately 40% d1EGFP-positive cells (Fig. 6D). When the Geneticin selection was removed, the percentage of d1EGFP-positive cells decreased from 40% to 1% in 30 days. When Geneticin selection on the COP5E2/PuropFRG cell line was restored at this point, the proportion of EGFP-expressing cells increased back to the initial level (Fig. 6D). These results are, in principle, identical to those obtained from the similar experiments with the E2/MME-dependent segregation/partitioning system described in the previous section (Fig. 6B and C). Therefore, EBNA1/FR elements and E2/MMEs confer comparable segregation/partitioning functions on the PyV core origin reporter plasmids in the analyzed cell model.
To exclude the possibility that the loss of EGFP fluorescence is due to inactivation of the promoter of EGFP, we also analyzed the DNA content in the cells. After removal of Geneticin selection total DNA was extracted from cells and digested with MluI (linearizes pMMEG and pFRG plasmids) and DpnI. Equal amounts of total DNA were then analyzed using Southern blotting with a radioactively labeled probe against the pMMEG or pFRG plasmid. As presented in Fig. 7 the loss of the episomal plasmid DNA from the cells grown without Geneticin selection correlates with the flow cytometry analysis. On the other hand, these results indicate that EGFP fluorescence was indeed measured from plasmids which exist in the episomal state. In the case of plasmid integration the hybridization signals remains constant.
DISCUSSION
The E2/MME works efficiently as a partitioning/segregation determinant also with a heterologous replicon. Prior to this study the only episomal maintenance element tested together with its nonnative replication origin was EBNA1/FR linked to different cellular replication origins. In these cases as well as in the case of replication of EBV OriP all the enzymatic activities required for replication are provided by the host cell. We have developed a system to study the mechanism of stable replication of a plasmid containing different maintenance elements in combination with their nonnative origins. More specifically, we have examined the functioning of E2/MME and EBNA1/FR maintenance elements in conjunction with the PyV replicon in the present study. As shown on Fig. 3A, the establishment of stable maintenance in this model depends on the simultaneous presence of E2 and MME. The plasmid containing 10 E2 BS and the PyV core origin is still present in the cell population 1 month after the initial enrichment of the transfected-reporter-carrying population based on antibiotic resistance selection. The same plasmid is efficiently maintained in the cells even if no selection is applied. Analysis of episomal DNA with nicking enzyme Nb.Bpu10I and rescue onto bacteria (Fig. 5) indicate that at least a majority of the stably maintained plasmids are not integrated and exist in an episomal state in the cells. Southern blot analysis and plasmid rescue assay indicated that some cells carry an oligomerized reporter plasmid (Fig. 5). Analysis of rescued plasmids showed that, in roughly one-half of the analyzed colonies, some plasmid rearrangement occurred (Fig. 5C). It has been shown that spontaneous sequence deletion by homologous recombination in the bacterial cell may occur (46). We have also noticed that highly palindromic sequences, like those of E2 BS-containing plasmids, are unstable during bacterial manipulations (A. Abroi, I. Ilves, and K. Janikson, personal communications). Whether these rearrangements in our present experiment originated in eukaryotic or in bacterial cells remains unclear; it is, however, important to note that all rearranged plasmids still carried the PyV minimal origin and BPV1 maintenance element (Fig. 5D and E). In addition, a control plasmid rescue assay with total DNA from cells carrying reporter plasmids with the wt PyV origin revealed only one colony whose low-molecular-weight DNA did not contain an MME or PyV origin (Fig. 5D and E). Thus, the E2/MME provides the stable episomal maintenance function not only to a BPV1-based replicon but also to a PyV-based replicon. Our study shows that mitotic chromatin attachment determinants of both BPV1 and EBV can provide the stable episomal maintenance function to heterologous viral replicons in addition to their native ones.
The episomally maintained reporter plasmids analyzed in our study do not cause major growth disadvantages for transfected cells and have a relatively low loss rate (Fig. 3 and Table 1). Previous studies have shown that, in the established cell lines carrying OriP, the plasmid loss is 2 to 8% per cell generation. However, the average plasmid copy number of OriP decreases more than 100-fold during the first 2 weeks after transfection into the cells expressing EBNA1 (25). In our system, certainly some decrease in the average copy number can be observed, but definitely not as fast. The quantitative aspects of the establishment of BPV1 stable episomal maintenance were not addressed in the present study. However, our data suggest that this process may be even more effective than in the case of EBNA1/OriP (25).
Overreplication of the wt PyV origin disrupts the establishment of stable maintenance. On the other hand, E2/MME cannot provide stable episomal maintenance to plasmids with the PyV wt origin, even under selective pressure (Fig. 3 and 4B). PyV exhibits a replication pattern that is uncoupled from the regulatory mechanisms of the host cell, so that each viral genome replicates many times within each cell cycle. The complete PyV origin includes transcriptional and replicational enhancer sequences, which dictate the origin activity and the efficiency of replication in specific cells by determining the availability of the replication factors and nucleotides. Papillomavirus origin replication control is similar to PyV replication in the first, amplificational phase of replication. However, in the latent-replication phase the copy number control mechanism is applied, which assures the controlled initiation of replication of the episomal viral genome in the latent-replication phase. We show that replacement of the wt PyV enhancer with 5 or 10 synthetic BS for the BPV1 E2 protein can replace replication enhancer function and makes it dependent on E2 protein. These results are in accordance with earlier reports by Nilsson et al. (29) and Abroi et al. (submitted) that at least two E2 BS are required to activate the PyV core origin. It is interesting to note that adding 5 or 10 E2 BS to the PyV wt origin did not cause additional replication activation. This fact supports the idea that replication from the episomal viral replication origins has a certain maximal threshold level in the host cell, which can be achieved by the presence (or addition) of strong enhancer elements. Further enhancement of the replication is not possible, even if more enhancer elements are added. It could be explained by the limiting levels of cellular replication factors or by the saturation of the nucleus with active genetic elements in the form of replication intermediates. This observation is also supported by the replication kinetics data, showing that the maximal level of replication by the wt PyV origin is achieved already 24 h posttransfection and that there is no more increase later; rather some decrease is observed (A. Abroi, unpublished data). At the same time, the replication signal of the PyV core origin or E2-dependent core origin increases in time. The toxic effect of the overreplication of the episomes on the cell can be suggested, as we observed many floating dead cells after transfection with PyV wt origin constructs. Thus, even though additional experimental data are needed to clarify this point, the most likely explanation for the inability of E2/MME to provide stable episomal maintenance to PyV wt origins is the cellular response against the high level of replication intermediates and/or the high level of replication itself.
The stable maintenance element from an EBV replicon that replicates strictly once per cell cycle can confer stable episomal maintenance properties to replication origin constructs derived from lytically replicating virus. The replication origins of BPV1 and PyV are fired several times during their amplificational replication in the single S phase of the host cell cycle, and the respective initiator proteins, E1 and LT, have many biochemical and structural similarities. However, these viruses have different time courses of productive infection. PyV is a lytic virus, and thus the viral DNA does not need to be stably maintained. During the stable replication of the BPV1 genome or URR, the origin is not restricted to precisely once replication round in each cell cycle (30, 31, 36). At the same time the EBV latent origin OriP replicates strictly once per cell cycle, exactly the same way as chromosomal DNA. Thus, in these terms, the replication modes of PyV and OriP are completely different. As shown on Fig. 6, the BPV1 E2/MME and EBV EBNA1/FR element can provide a stable maintenance function to the PyV core origin plasmids in the presence of viral trans factors. Our data presented here suggest that stable maintenance of the episomes provided by the function of MME orthe FR element is not connected to the mode of replication of the episome. The FR element can provide a stable maintenance function to several types of origins, in its natural context within the EBV latent origin OriP, in the plasmids where the chromosomal origin of replication from cellular DNA is linked to the FR element, and in our hybrid replicon together with the PyV core origin (22, 41). These data also show that the replication function is not connected to the stable-maintenance function of the virus; replication origins of different viruses can be combined with heterologous stable-maintenance elements without the loss of either function. The cellular receptors of BPV1 E2 protein and EBV EBNA1 protein, which link the episomes to mitotic host chromatin and therefore provide the stable-maintenance function, are different (20, 21, 33, 44, 45, 47). And, most likely, E2/MME- and EBNA1/FR-dependent plasmids are localized on chromosomes in different places. Our data presented here indicate that the different localizations of the episome on mitotic chromosomes do not interfere with the replication of the PyV minimal replication origin.
The chromatin attachment/partitioning function, not activation of replication, is responsible for stable episomal maintenance of a heterologous origin. Structural intactness of the E2 protein is very important in order to provide the MME-dependent partitioning function. A recent study from our laboratory showed that single point mutations might affect the chromatin attachment of E2 protein or its ability to mediate chromatin tethering of URR reporters even if the effect on replication initiation or transcription activation is relatively modest (1). E2 mutant proteins E39A and R68A were used to analyze the role of the transcription activation properties of the E2 protein in its functioning as a trans factor for stable episomal maintenance of a PyV-derived replicon. These E2 mutants were shown to be inactive in chromatin attachment, URR plasmid tethering, and segregation/partitioning assays in our studies using Chinese hamster ovary cells (1); they were, however, functional in Brd4 binding and chromatin binding assays using CV1 cells derived from African green monkey kidney cells (7). These mutant E2 proteins though seemed to be nonfunctional in supporting long-term episomal maintenance of the hybrid MME/PyV core ori plasmids in C127 mouse cells, as shown in this paper. The apparent contradiction in the features of these E2 mutants may be attributed to the difference in the species of cells used to measure the functions of the E2 mutants. The interaction of these E2 mutants could be different with mouse, hamster, and monkey Brd4 proteins, which could be due to the difference in the sequences between Brd4 proteins of the different species. The cell lines used in our assays have been shown to be functional in supporting BPV1 origin replication, segregation/partitioning, and long-term maintenance of episomal replication, which is not necessarily the case with CV1 cells.
In addition, we tested the stable replication of the E2/MME-dependent PyV replicon also in cell lines expressing LT and VP16-E2 or p53-E2 (containing the activation domains from VP16 and p53, respectively). They both are very potent transactivators and capable of activating the replication of the PyV core origin in transient assays. However, none of these mutant or hybrid E2 variants was capable of ensuring effective stable episomal maintenance, in the case of VP16-E2 and p53-E2 not even under selective conditions (Fig. 4B and data not shown). Our results show clearly that chromatin attachment, but not transactivation and the consequent replication initiation activity of E2 protein, is essential to provide stable maintenance for chimeric constructs used in this study and that random partitioning of the episomal plasmids cannot provide a reliable mechanism for stable episomal maintenance of the plasmids even in the presence of selection for the episomal selection marker. These results suggest that MME-mediated partitioning in conjunction with the PyV origin or its natural BPV1 origin is achieved by using the same strategy, i.e., through chromatin attachment.
Rate of loss of episomal plasmids. In the present study we have analyzed the episomal maintenance of plasmids containing the PyV minimal replication origin and either BPV1 MME or the EBV FR element (Fig. 6A) in cells where the appropriate viral trans factors (either PyV LT and BPV1 E2 or PyV LT and EBV EBNA1 protein) were stably expressed. In the case of plasmids containing the PyV minimal replication origin and BPV1 MME, the rate of episomal loss was 6% per cell division in the absence of selection. For plasmids containing the PyV minimal origin and EBV FR element, the rate of episomal plasmid loss was higher (13%), but it is still significantly lower than 22 to 30%, which we observed in the case of control plasmids (pEGFP-C1 and pd1EGFP-N1) without a eukaryotic replication origin and segregation elements. The rate of loss of plasmids containing the PyV minimal replication origin and FR element (pFRG) is also different from the previously published results on the rate of loss of several replicating plasmids that contained the FR element as a stable maintenance factor, where it was 2.1 to 7.8% (41); however, it is very similar to the 15% rate of loss previously estimated for OriP-containing plasmids (17). The difference in the rate of loss may be due to the differences in the expression level of EBNA1, the configuration of the test plasmids used, or the nature of the chromatin receptor for EBNA1, because in our experiments the mouse cell line COP5, not human cells, was used. The difference between the d1EGFP-positive and EGFP-positive cells in cell culture grown under Geneticin selection is probably due to the sensitivity of detection. As the half-life of the d1EGFP protein is 1 hour and as d1EGFP does not accumulate in the cells, the level of d1EGFP in cells with lower reporter plasmid copy number may probably be insufficient for its detection from the autofluorescence background and thus a certain fraction of cells which in fact carry the reporter will probably be regarded as "EGFP negative". However, the fact that both long- and short-half-life EGFP reporters gave similar rates of loss indicates that our method for measuring the rate of plasmid loss is adequate.
In conclusion, all these data together indicate that the maintenance elements from different DNA viruses are interchangeable with each other and can work in conjunction with different replicons, even with those from lytically replicating viruses. However, in heterologous systems as well as in native configurations, a certain loss of plasmid exists. In order to compensate the loss of episomes, viruses have evolved systems to accelerate the host cell proliferation compared to uninfected cells.
ACKNOWLEDGMENTS
We thank Anne Kalling for technical assistance, Ivar Ilves for careful reading of the manuscript, Kertu Rünkorg for constructing the COP5E2/Neo cell line, Michael Botchan for VP16-E2 expression construct, Mari Sepp for the BPV1-transfected C127 cell line, and G?ran Magnusson for the COP5 cell line and PyV origin constructs.
This study was supported in part by grants 4475, 4476, 5999, and 5998 from the Estonian Science Foundation, grant INTNL 55000339 from the Howard Hughes Medical Institute, grant CT96-0918 from the European Union, and target financial project 0182566s03.
REFERENCES
Abroi, A., I. Ilves, S. Kivi, and M. Ustav. 2004. Analysis of chromatin attachment and partitioning functions of bovine papillomavirus type 1 E2 protein. J. Virol. 78:2100-2113.
Abroi, A., R. Kurg, and M. Ustav. 1996. Transcriptional and replicational activation functions in the bovine papillomavirus type 1 E2 protein are encoded by different structural determinants. J. Virol. 70:6169-6179.
Ausubel, F. M., R. Brent, R. E. Kingston, D. D. Moore, J. G. Seidman, and K. Struhl (ed.). 1998. Current protocols in molecular biology. John Wiley & Sons, Inc., New York, N.Y.
Ballestas, M. E., P. A. Chatis, and K. M. Kaye. 1999. Efficient persistence of extrachromosomal KSHV DNA mediated by latency-associated nuclear antigen. Science 284:641-644.
Ballestas, M. E., and K. M. Kaye. 2001. Kaposi's sarcoma-associated herpesvirus latency-associated nuclear antigen 1 mediates episome persistence through a cis-acting terminal repeat (TR) sequence and specifically binds TR DNA. J. Virol. 75:3250-3258.
Baru, M., M. Shlissel, and H. Manor. 1991. The yeast GAL4 protein transactivates the polyomavirus origin of DNA replication in mouse cells. J. Virol. 65:3496-3503.
Baxter, M. K., M. G. McPhillips, K. Ozato, and A. A. McBride. 2005. The mitotic chromosome binding activity of the papillomavirus E2 protein correlates with interaction with the cellular chromosomal protein, Brd4. J. Virol. 79:4806-4818.
Bostock, C. J., and R. C. Allshire. 1986. Comparison of methods for introducing vectors based on bovine papillomavirus-1 DNA into mammalian cells. Somatic Cell Mol. Genet. 12:357-366.
Clertant, P., and F. Cuzin. 1980. Initiation of polyomavirus DNA replication in vitro and its dependence on the viral gene A protein. Nucleic Acids Res. 8:4377-4392.
Cole, C. N. C., and S. D. Conzen. 2001. Polyomaviridae: the viruses and their replication, p. 2141-2174. In D. M. Knipe and P. M. Howley (ed.), Fields' virology, 4th ed. Lippincott Williams & Wilkins, Philadelphia, Pa.
Cotter, M. A., II, and E. S. Robertson. 1999. The latency-associated nuclear antigen tethers the Kaposi's sarcoma-associated herpesvirus genome to host chromosomes in body cavity-based lymphoma cells. Virology 264:254-264.
Delecluse, H. J., S. Bartnizke, W. Hammerschmidt, J. Bullerdiek, and G. W. Bornkamm. 1993. Episomal and integrated copies of Epstein-Barr virus coexist in Burkitt lymphoma cell lines. J. Virol. 67:1292-1299.
de Villiers, J., W. Schaffner, C. Tyndall, S. Lupton, and R. Kamen. 1984. Polyomavirus DNA replication requires an enhancer. Nature 312:242-246.
Dhar, S. K., K. Yoshida, Y. Machida, P. Khaira, B. Chaudhuri, J. A. Wohlschlegel, M. Leffak, J. Yates, and A. Dutta. 2001. Replication from oriP of Epstein-Barr virus requires human ORC and is inhibited by geminin. Cell 106:287-296.
Francke, B., and W. Eckhart. 1973. Polyoma gene function required for viral DNA synthesis. Virology 55:127-135.
Guo, Z. S., and M. L. DePamphilis. 1992. Specific transcription factors stimulate simian virus 40 and polyomavirus origins of DNA replication. Mol. Cell. Biol. 12:2514-2524.
Hung, S. C., M. S. Kang, and E. Kieff. 2001. Maintenance of Epstein-Barr virus (EBV) oriP-based episomes requires EBV-encoded nuclear antigen-1 chromosome-binding domains, which can be replaced by high-mobility group-I or histone H1. Proc. Natl. Acad. Sci. USA 98:1865-1870.
Ilves, I., S. Kivi, and M. Ustav. 1999. Long-term episomal maintenance of bovine papillomavirus type 1 plasmids is determined by attachment to host chromosomes, which is mediated by the viral E2 protein and its binding sites. J. Virol 73:4404-4412.
Kanda, T., M. Otter, and G. M. Wahl. 2001. Coupling of mitotic chromosome tethering and replication competence in Epstein-Barr virus-based plasmids. Mol. Cell. Biol. 21:3576-3588.
Kapoor, P., and L. Frappier. 2003. EBNA1 partitions Epstein-Barr virus plasmids in yeast cells by attaching to human EBNA1-binding protein 2 on mitotic chromosomes. J. Virol. 77:6946-6956.
Kapoor, P., K. Shire, and L. Frappier. 2001. Reconstitution of Epstein-Barr virus-based plasmid partitioning in budding yeast. EMBO J. 20:222-230.
Krysan, P. J., and M. P. Calos. 1993. Epstein-Barr virus-based vectors that replicate in rodent cells. Gene 136:137-143.
Kurg, R., J. Parik, E. Juronen, T. Sedman, A. Abroi, I. Liiv, U. Langel, and M. Ustav. 1999. Effect of bovine papillomavirus E2 protein-specific monoclonal antibodies on papillomavirus DNA replication. J. Virol. 73:4670-4677.
Lehman, C. W., and M. R. Botchan. 1998. Segregation of viral plasmids depends on tethering to chromosomes and is regulated by phosphorylation. Proc. Natl. Acad. Sci. USA 95:4338-4343.
Leight, E. R., and B. Sugden. 2001. Establishment of an oriP replicon is dependent upon an infrequent, epigenetic event. Mol. Cell. Biol. 21:4149-4161.
Li, R., J. Knight, G. Bream, A. Stenlund, and M. Botchan. 1989. Specific recognition nucleotides and their DNA context determine the affinity of E2 protein for 17 binding sites in the BPV-1 genome. Genes Dev. 3:510-526.
Lupton, S., and A. J. Levine. 1985. Mapping genetic elements of Epstein-Barr virus that facilitate extrachromosomal persistence of Epstein-Barr virus-derived plasmids in human cells. Mol. Cell. Biol. 5:2533-2542.
Morgenstern, J. P., and H. Land. 1990. Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res. 18:3587-3596.
Nilsson, M., M. Forsberg, Z. Y. You, G. Westin, and G. Magnusson. 1991. Enhancer effect of bovine papillomavirus E2 protein in replication of polyomavirus DNA. Nucleic Acids Res. 19:7061-7065.
Piirsoo, M., E. Ustav, T. Mandel, A. Stenlund, and M. Ustav. 1996. Cis and trans requirements for stable episomal maintenance of the BPV-1 replicator. EMBO J. 15:1-11.
Ravnan, J. B., and S. N. Cohen. 1995. Transformed mouse cell lines that consist predominantly of cells maintaining bovine papilloma virus at high copy number. Virology 213:526-534.
Schvartzman, J. B., S. Adolph, L. Martin-Parras, and C. L. Schildkraut. 1990. Evidence that replication initiates at only some of the potential origins in each oligomeric form of bovine papillomavirus type 1 DNA. Mol. Cell. Biol. 10:3078-3086.
Sears, J., M. Ujihara, S. Wong, C. Ott, J. Middeldorp, and A. Aiyar. 2004. The amino terminus of Epstein-Barr virus (EBV) nuclear antigen 1 contains AT hooks that facilitate the replication and partitioning of latent EBV genomes by tethering them to cellular chromosomes. J. Virol. 78:11487-11505.
Sharma, R. C., and R. T. Schimke. 1996. Preparation of electrocompetent E. coli using salt-free growth medium. BioTechniques 20:42-44.
Skiadopoulos, M. H., and A. A. McBride. 1998. Bovine papillomavirus type 1 genomes and the E2 transactivator protein are closely associated with mitotic chromatin. J. Virol. 72:2079-2088.
Ten Hagen, K. G., J. B. Ravnan, and S. N. Cohen. 1995. Disparate replication properties of integrated and extrachromosomal forms of bovine papilloma virus in ID13 cells. J. Mol. Biol. 254:119-129.
Tsurumi, T., M. Fujita, and A. Kudoh. 2005. Latent and lytic Epstein-Barr virus replication strategies. Rev. Med. Virol. 15:3-15.
Tyndall, C., G. La Mantia, C. M. Thacker, J. Favaloro, and R. Kamen. 1981. A region of the polyoma virus genome between the replication origin and late protein coding sequences is required in cis for both early gene expression and viral DNA replication. Nucleic Acids Res. 9:6231-6250.
Ustav, M., and A. Stenlund. 1991. Transient replication of BPV-1 requires two viral polypeptides encoded by the E1 and E2 open reading frames. EMBO J. 10:449-457.
Ustav, M., E. Ustav, P. Szymanski, and A. Stenlund. 1991. Identification of the origin of replication of bovine papillomavirus and characterization of the viral origin recognition factor E1. EMBO J. 10:4321-4329.
Wade-Martins, R., J. Frampton, and M. R. James. 1999. Long-term stability of large insert genomic DNA episomal shuttle vectors in human cells. Nucleic Acids Res. 27:1674-1682.
Waldenstrom, M., K. Schenstrom, K. Sollerbrant, and L. Hansson. 1992. Replication of bovine papillomavirus vectors in murine cells. Gene 120:175-181.
Wang, E. H., and C. Prives. 1991. DNA helicase and duplex DNA fragment unwinding activities of polyoma and simian virus 40 large T antigen display similarities and differences. J. Biol. Chem. 266:12668-12675.
Wu, H., D. F. Ceccarelli, and L. Frappier. 2000. The DNA segregation mechanism of Epstein-Barr virus nuclear antigen 1. EMBO Rep. 1:140-144.
Yates, J. L., and N. Guan. 1991. Epstein-Barr virus-derived plasmids replicate only once per cell cycle and are not amplified after entry into cells. J. Virol. 65:483-488.
Yates, J. L., N. Warren, and B. Sugden. 1985. Stable replication of plasmids derived from Epstein-Barr virus in various mammalian cells. Nature 313:812-815.
You, J., J. L. Croyle, A. Nishimura, K. Ozato, and P. M. Howley. 2004. Interaction of the bovine papillomavirus E2 protein with Brd4 tethers the viral DNA to host mitotic chromosomes. Cell 117:349-360.(Toomas Silla, Ingrid H??l)