Human Molecular Genetics, 2000, Vol. 9, No. 4 617-629
© 2000 Oxford University Press
Expression of the human CFTR gene from episomal oriP-EBNA1-YACs in mouse cells
Division of Biomedical Sciences, Imperial College School of Medicine, Sir Alexander Fleming Building, South Kensington, London SW7 2AZ, UK
Received 2 November 1999; Revised and Accepted 14 December 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Plasmids carrying the origin of plasmid replication (oriP) and expressing the EBNA-1 protein from the Epstein–Barr virus replicate and segregate in human cells and are thus potentially useful vectors for gene therapy. As very large circular molecules, up to 660 kb in size, can be maintained episomally using this system, it is possible to include intact human genes with all their long-range controlling elements which might give high levels of tissue-specific and controlled gene expression. We have shown previously that a 320 kb yeast artificial chromosome (YAC) carrying the intact human CFTR gene can complement the Cambridge null cystic fibrosis mice as a transgene. We have now modified this YAC to a circular molecule carrying both oriP and the EBNA-1 gene. We show that this oriP-EBNA1-YAC can be stably maintained as unrearranged episomes in mouse LA-9 cells, which do not express endogenous cftr, and in mouse CMT-93 cells, which do express endogenous cftr. The human CFTR gene is expressed in some of the cell lines, but the level of expression is very variable between cell lines and is not related to the copy number of the elements.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Epstein–Barr virus (EBV) is a human herpes virus with a 172 kb circular, double-stranded genome which is maintained episomally during its latent cycle in human cells. The functional components required for the latent replication and maintenance of the EBV genome are the latent origin of replication oriP which spans ~1.5 kb (1) and the viral transactivator protein EBNA-1 which is encoded by a 1.9 kb gene (2,3). oriP consists of two components: a family of repeats which comprises 20 tandem copies of a 30 bp sequence and the 65 bp dyad symmetry which contains four copies of the repeat (4). EBNA-1 binds to both oriP elements (5). The dyad symmetry is the actual start site of latent replication (6), whereas the family of repeats confers nuclear retention (7) and is also an EBNA-1-dependent enhancer of transcription (8). The combination of replication once per cell cycle and nuclear retention by association with the chromatin leads to quite stable episomal maintenance of plasmids carrying oriP and EBNA-1.
Vectors carrying oriP and EBNA-1 could be useful for gene therapy as they allow long-term episomal maintenance of the DNA (9). Generally, DNA introduced into cells is lost during cell growth unless it is integrated. Integration is not only potentially mutagenic but is also an infrequent event unless a viral integration system such as those found in retroviruses or adeno-associated virus is included. The presence of oriP and EBNA-1 on a plasmid has been shown to produce much higher expression after transient transfection into cells in tissue culture (10) and also prolonged expression in rapidly dividing cells (11). Similar effects have also been observed in vivo. An oriP/EBNA-1 vector expressing luciferase and delivered to the liver with HVJ liposomes (a hybrid vector with liposome and Sendai virus) gave higher expression than a normal plasmid and expression also persisted to day 35, whereas expression from a normal plasmid was lost by day 14 (12). Similarly an oriP/EBNA-1 vector carrying recombinant dystrophin driven by the Rous sarcoma virus (RSV) promoter gave a significantly enhanced number of expressing muscle fibres and better long-term expression after intramuscular injection (13).
Mouse models are valuable for testing gene therapy applications. However, rodent cells do not support replication of small oriP/EBNA-1 vectors, although they are permissive for episomal maintenance (14). When large fragments of mammalian DNA are added to vectors containing the oriP family of repeats (which provide nuclear retention) and the EBNA-1 gene, they provide mammalian origins of replication (7) resulting in vectors which have the ability to replicate and segregate properly in mouse cells (14). Thus, the oriP/EBNA-1 system can also be used in mouse models.
Cystic fibrosis (CF) is a good candidate for gene therapy as it is recessive and the affected tissues (lung and gut) are fairly accessible for gene delivery. Several clinical trials using delivery of cDNA-based constructs have been carried out and have resulted in partial and transient correction of chloride transport (reviewed in ref. 15). Use of oriP/EBNA-1 vectors could provide long-term expression and thus reduce the frequency of treatment needed. In addition, use of the intact CFTR gene rather than a cDNA construct could give tissue-specific and controlled expression which may give better correction. We have shown previously that a 320 kb yeast artificial chromosome (YAC) carrying the intact 200 kb CFTR gene plus its associated control regions gives tissue-specific expression in transgenic mice and complements the null CF phenotype (16). This YAC is the only construct that has fully restored chloride transport in transgenic mice and therefore is potentially useful for gene therapy; cDNA constructs have only given partial correction in some tissues (17–19). Thus, the use of the intact CFTR gene with oriP and EBNA-1 could give a long-term vector which would give tissue-specific and controlled expression.
Here we describe the construction of circular YACs carrying EBNA-1, oriP and the intact human CFTR gene. These YACs have been transferred to two mouse, as opposed to human, cell lines where we show that they are stably maintained as episomes for long periods of time in the absence of selection. However, although the YAC contains all the long-range elements necessary for full levels of expression in transgenic mice, expression from the episomes is variable and not related to copy number.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Construction of oriP-EBNA1-YACs carrying the CFTR gene
The YAC yCFTR carries the intact human CFTR gene and previously has been shown to be able to complement the null CF mutation in transgenic mice (16). It has then been retrofitted with an internal ribosome entry site (IRES) in the 3'-untranslated region (3'-UTR), giving the YAC yCFIRES. This allows expression of ß-geo (a fusion between ß-galactosidase and the neomycin resistance gene) in addition to CFTR from the same mRNA (20). The IRES-ß-geo element in this YAC (yCFIRES) should provide a quick assay to check CFTR expression once the YAC is in mammalian cells.
The linear YAC yCFIRES was circularized using the plasmid pDH3 (Fig. 1A). pDH3 contains the oriP region and EBNA-1 gene from EBV plus the HIS5 yeast selectable marker, and the hygromycin resistance gene (hygr) for selection in mammalian cells. Restriction with ClaI gives free ends which homologously recombine with the ends of a linear YAC (Fig. 1B). The resulting circular YAC, yCFTR-oriP-EBNA1, is shown in Figure 1C.
|
To check that the YAC had been circularized but not rearranged, high molecular weight YAC DNA was treated with
-irradiation which randomly introduces double-stranded breaks into DNA. Subsequent pulsed-field gel electrophoresis (PFGE) and Southern analysis showed that without irradiation circular yCFTR-oriP-EBNA1 clone 11 remains in the wells (Fig. 2A, n), whereas after irradiation a 320 kb band is seen (Fig. 2A, ) which is the same size as the original linear yCFIRES. This result indicates that yCFTR-oriP-EBNA1 clone 11 is circular and of the expected size. Long-range restriction analysis also showed no rearrangements (data not shown), and the oriP region, hygr and CFTR genes were all present and non-rearranged as determined by Southern analysis.
|
The GlyAla repeat region of EBNA-1 is unstable
Rearrangements were found in the GlyAla region (also called IR3) present in EBNA-1. This region spans from amino acid 90 to 328 of EBNA-1, is rich in glycine and alanine and is rich in GGG, GCA and GGA repeats at the DNA level (21). After circularization with pDH3, nearly half (6/14) of the yCFTR-oriP-EBNA1 YACs were found to contain rearrangements of the GlyAla region as determined by rearrangement of the 1122 bp AvaI fragment spanning this region (Fig. 2B, rearranged clones indicated by *). Clone 10 has a deleted GlyAla region of EBNA-1 whereas clone 11 is slightly larger in this region.
This region of EBNA-1 has been found to be unstable in Escherichia coli and it is not surprising that it is unstable in yeast considering the high efficiency of homologous recombination in yeast and the repetitive nature of the region. To determine whether the region remained unstable during growth of the yeast, clone 11 was streaked out and 10 clones were analysed. Nine of the 10 clones were non-rearranged in the GlyAla region (data not shown), indicating that it is relatively stable during propagation in yeast, and the high level of rearrangement in the initial clones was therefore related to the transfection and homologous recom- bination event of circularization rather than simply growth in Saccharomyces cerevisiae. Deleted versions of EBNA-1 are usually deleted in-frame and function well for stable maintenance of vectors (3).
OriP-EBNA1-YACs are maintained episomally in mouse cells
The yCFTR-oriP-EBNA1 molecules were introduced into mouse LA-9 cells by fusion with yeast spheroplasts followed by selection for hygromycin resistance. Fusion with yCFTR-oriP-EBNA1 clone 11 (Fig. 2B) gave two cell lines named LA9-B2 and -B3, whereas fusion with clone 10 gave two cell lines called LA9-ori2 and -ori3. The frequency of colony formation in both cases was about two colonies from 106 cells fused. As this frequency was rather low, the yCFTR-oriP-EBNA1 clone 11 was retrofitted with pLUNA which introduces a strong neor gene. The resulting YAC yCFTR-oriP-EBNA1-neo was then fused with CMT-93 mouse cells using selection with G418. A total of 26 CMT-93 cell lines were generated (CMT-1 to -26), at a frequency of ~10 resistant colonies per 106 cells.
To determine whether the LA-9 cell lines carry episomal or integrated oriP-EBNA1-YACs, high molecular weight DNA was treated with -irradiation, resolved by PFGE and Southern blotted. An ~320 kb band is apparent in LA9-B3 and 640 kb bands are apparent in LA9-B2, -ori2 and -ori3 after irradiation (Fig. 3A), whereas no bands are observed in the absence of irradiation (data not shown). Mammalian DNA tends to run more slowly than yeast DNA and size markers on pulsed-field gels due to the greater amount loaded to compensate for the greater complexity. This indicates that all four cell lines contain circular episomal DNA; integrated DNA would give a smear of signal rather than a band, and linear episomes would resolve in the absence of -irradiation. The 320 kb molecule in LA9-B3 could be the non-rearranged yCFTR-oriP-EBNA1, whereas the 640 kb molecules in LA9-B2, -ori2 and -ori3 are consistent with dimers of the original YAC, although they could also be completely rearranged molecules. The signal in the non-resolved area of the pulsed-field gel is higher in LA9-ori2 (Fig. 3A) than in the other lanes, indicating that even larger molecules are probably present in this cell line.
|
Eleven of the 26 cell lines derived from CMT-93 cells were analysed in detail.
-irradiation resulted in bands of ~320 and 640 kb in seven of the cell lines (Fig. 3B), which probably correspond to monomers and dimers of yCFTR-oriP-EBNA1-neo. The copy number of the oriP-EBNA1-YACs in CMT-1, -9 and -16 is too low for detection by this methodology (linearization by -irradiation is inefficient as many molecules are cut several times or not at all). The intensity of the 320 kb band versus the 640 kb band is different among the cell lines and is not related to the overall copy number. CMT-6, -10 and -14 contain mainly 320 kb episomal molecules, whereas CMT-2, -3, -7 and -25 contain both sizes of molecule. The two bands in CMT-13 are somewhat larger than 320 and 640 kb, indicating that the oriP-EBNA1-YAC is probably rearranged in this cell line. The oriP-EBNA1-YACs are thus maintained as circular episomes in both the LA-9 and CMT-93 cell lines. However, they occur as both 320 and 640 kb molecules in varying proportions, independently of copy number.
The oriP-EBNA1-YAC episomes are largely non-rearranged
SalI cuts three times in the plasmid region of the yCFTR-oriP-EBNA1 YACs (Fig. 1A) and not at all in the inserts, resulting in a 320 kb fragment containing the whole CFTR gene. The sizes of the SalI fragments in the four cell lines LA9-B2, -B3, -ori2 and -ori3 vary slightly from the 320 kb SalI fragment observed in yCFTR-oriP-EBNA1 (Fig. 4A). Again, mammalian DNA tends to run more slowly than yeast DNA and size markers on pulsed-field gels due to the greater amount loaded to compensate for the greater complexity. LA9-B2 and -B3 probably contain non-rearranged YACs, whereas LA9-ori2 and -ori3 contain somewhat larger molecules, indicating that these two clones probably have rearranged inserts (Fig. 4A). To detect rearrangements within the CFTR gene itself, DNA digested with HindIII was hybridized with the CFTR cDNA probe. In all four cell lines, the fragments generated were identical to those from yCFTR-oriP-EBNA1 (Fig. 4B), indicating that there are few or no rearrangements around the CFTR exons.
|
Eight of the 11 CMT-93 cell lines (CMT-1, -2, -3, -6, -7, -10, -14 and -25) contain a 320 kb internal SalI fragment which is the same size as that from the original yCFTR-oriP-EBNA1 (Fig. 4C), indicating that the oriP-EBNA1-YACs remain grossly non-rearranged in these cell lines. CMT-13 gives a slightly larger fragment, CMT-9 gives a smaller fragment in addition to the 320 kb fragment, and CMT-16 gives a smaller fragment, indicating that these three clones contain rearranged forms of yCFTR-oriP-EBNA1-neo. There may be very small amounts of rearranged molecules in the other cell lines. Digestion with restriction enzymes is a more efficient process than
-irradiation, and SalI fragments from the oriP-EBNA1-YACs can be observed in CMT-1, -9 and -16 in which fragments could not be detected after -irradiation (Fig. 3B). Of these, CMT-1 has the non-rearranged 320 kb SalI fragment. When YACs not carrying oriP are introduced into mammalian cells by fusion, the resulting cell lines have often integrated a large amount of yeast genomic DNA in addition to the YAC DNA (22). This DNA can be detected by hybridization with a probe for the yeast repetitive element TY1. Using this test, the fusion cell lines were found not to contain yeast genomic DNA, except for line CMT-2 (data not shown). This low frequency of integration of yeast DNA, also observed with oriP-YACs not carrying EBNA-1 (23), is probably due to the efficient nuclear localization of the YAC, but not the yeast genomic DNA due to the interaction between oriP and EBNA-1.
The yCFTR-oriP-EBNA1 molecules are thus maintained as probably grossly non-rearranged monomers or dimers of the original molecules in two of four LA-9-derived cell lines and in eight of eleven CMT-93-derived cell lines.
The oriP-EBNA1-YACs are maintained at variable copy number
The copy number of the yCFTR-oriP-EBNA1 molecules in the mouse cell lines was determined by hybridization of a Southern blot with a human CFTR exon 4 probe. Different loading between the lanes with the different mouse-derived cell lines was normalized using a mouse cftr exon 10 probe. Genomic DNA from the human cell line Caco-2 was used as a standard for two copies of the human CFTR gene, and the relative loading of the human Caco-2 versus the mouse-derived lanes was determined by intensity of ethidium bromide staining.
The copy numbers of the episomes in the LA-9 cell lines were determined to be 6, 8, 7 and 9 (Fig. 5A), whereas in the CMT-93 cell lines the copy number varied between 2 and 56 (Fig. 5B). These values represent the numbers of episomes as well as any integrated copies of yCFTR-oriP-EBNA1. However, the copy numbers seem to correlate well with the relative intensity of signal after -irradiation (which was not normalized for loading), indicating that it represents mainly the episomal copy number. Also, very few integrations were observed by fluorescence in situ hybridization (FISH) (see below).
|
OriP-EBNA1-YACs are stably maintained in mouse cells
Replication and segregation ability of the oriP-EBNA1-YACs was assayed by growing the fusion cell lines with and without selection for 1–16 weeks, followed by Southern blotting to determine the amount of episomes present. The episomes in all four LA-9 fusion cell lines remained at the same copy number in the presence of selection (data not shown) and slowly dropped in the absence of selection (Fig. 6A). Using an exponential model of loss of elements, the rate of loss was determined to be 5, 2.6, 4.2 and 4.9% per cell division for the LA-9 cell lines (Table 1).
|
|
Of the CMT-93 cell lines, CMT-1, -3 and -7 were taken as examples of cells which have low (= 2), medium (= 10) and high (= 56) copy number for analysis up to 16 weeks of culture with or without selection (Fig. 6B). In CMT-1 and -3, the molecules remained at the same copy number in the presence of selection and were slowly lost at 0.4 and 1.8% per generation in the absence of selection. In CMT-7, the episomes had reduced significantly at 8 and 16 weeks both with and without selection—loss is possible in the presence of selection because of the initial very high copy number in this cell line. The rate of loss in the absence of selection was 1.6% per generation. Even after 16 weeks in culture (with or without selection), the oriP-EBNA1-YACs were still episomal as determined by
-irradiation and PFGE (data not shown).
The oriP-EBNA1-YACs are sometimes, but not always, attached to the host chromosomes
FISH using a cosmid from the human CFTR gene as probe was used to visualize the oriP-EBNA1-YACs in the mouse cell lines. This probe gave no signal on the parental LA-9 and CMT-93 cell lines (data not shown).
All four LA-9 cell lines clearly contain episomal molecules (Fig. 7). The LA-9 cell lines generally have a fairly homogeneous distribution of episomes between cells. However, whereas in LA9-B3 the oriP-EBNA1-YACs are clearly attached to the mouse host chromosomes (but generally not integrated, as the dots are not located symmetrically on the chromatids), the episomes in LA9-B2, -ori2 and -ori3 are not attached but are well spread out away from the host chromosomes (Fig. 7).
|
For the CMT-93 fusion cell lines, the distribution of elements between cells was quite variable as was the attachment or non-attachment of the elements to the host chromosomes. Fifty metaphases were analysed for each CMT-93 cell line, and the most common positioning of the elements is indicated in Table 1. For instance, CMT-7 has mostly attached episomes (yellow signal on top of chromosomes but usually not integrated as they are positioned at the side or diagonally on a chromosome) (Fig. 7), whereas CMT-14 has mostly episomes off the chromosomes (Fig. 7). In CMT-3, the episomes are usually bound as a large cluster to a host chromosome (
-irradiation showed that they are not integrated) (Fig. 7). Such clusters were also seen in a few metaphases of CMT-6 and -10. Occasional cells contained apparent integrations of the CFTR gene into a host chromosome as the pair of dots is arranged symmetrically on the chromatids (there is a clear example in CMT-7 of Fig. 7). However, this occurred in only 0–13% of metaphases in any cell line (‘Int.’ in Table 1), indicating that if some copies are integrated into the genome, it is a small minority. There was also a variable percentage of metaphases with no signal in some cell lines (‘None’ in Table 1), although this could be due to false-negative results which are fairly common in FISH experiments.
It was not clear from the -irradiation experiments whether CMT-1, -9 and -16 contained extrachromosomal or integrated DNA. FISH analysis showed a low percentage of metaphases with possible integrations in CMT-1 (4%), -9 (1.3%) and -16 (12.5%) versus extrachromosomal elements (16, 36 and 58%, respectively), indicating that these cell lines do have mostly extrachromosomal elements. These cell lines also have a high percentage of metaphases with no signal (CMT-1, 80%; -9, 63%; and -16, 47%). However, other cell lines such as CMT-2 (60%), -10 (54%) and -25 (49%) also have high percentages of cells with no detectable signal, indicating that the distribution of elements between cells within a line is very uneven.
The human CFTR transgene is expressed in some cell lines
RT–PCR was carried out to determine whether the human CFTR gene on the yCFTR-oriP-EBNA1 episomes was being expressed. LA-9 cells do not express endogenous cftr so expression from the yCFTR-oriP-EBNA1 molecules could not be compared with endogenous levels. However, using standard PCR conditions, a strong product was obtained from human CFTR mRNA in LA9-B2, -ori2 and -ori3 but not from -B3 (Fig. 8A, LA9-ori3 not shown).
|
As the YAC used carries an IRES-ß-geo cassette in the 3'-UTR of the CFTR gene, lacZ should be expressed in the cell lines where the CFTR mRNA is detected. Staining with X-gal gave fairly dark staining in many cells in LA9-B2, -ori2 and -ori3 but not in -B3 (Fig. 8B), and this correlates with the expression of CFTR mRNA as detected by RT–PCR in these cell lines.
All four cell lines apparently contain the intact CFTR gene (Fig. 4B) and have similar copy number. However, LA9-B3, which does not express CFTR, differs from the other three cell lines in that it has ~320 rather than 640 kb episomes and the episomes are attached to the host chromosomes rather than spread out.
CMT-93 cells express endogenous cftr mRNA, so the level of expression of the human CFTR gene could be compared directly with this. RT–PCR was carried out with primers which bind equally to highly homologous regions of the mouse and human CFTR genes in exons 2 and 6a/b to amplify a 610 bp region of the cDNA. Digestion with NruI cuts the human product to give bands of 370 and 240 bp, whereas digestion with HindIII cuts the mouse product to give fragments of 548 and 62 bp. The murine-specific 548 bp band is observed for CMT-93 and all CMT-93-derived fusion cell lines (Fig. 8A), whereas the 370 and 240 bp bands are observed with mRNA from human Caco-2 cells which express endogenous human CFTR mRNA (Fig. 8A). Bands corres- ponding to human CFTR expression were visible in CMT-1, -2, -3, -6, -9, -14 and -25 (Fig. 8A, and data not shown). The percentage of human versus murine expression was determined with a phosphorimager (Table 1). Only CMT-1 expresses high levels of human CFTR mRNA (220%), the rest of the cell lines express levels <25% of the murine cftr gene expression.
The expression in clones 1, 3 and 7 was also investigated after 16 weeks with or without selection (Fig. 8A). In clone 1, the high level of expression continues under both growth conditions, in clone 3 the low level of expression drops further under both conditions, whereas in clone 7 no expression is observed at any time point.
The level of expression is clearly not proportional to copy number, as clone 7 with the highest copy number shows no expression whereas clone 1 with the lowest copy number has the highest expression. However, there does appear to be a correlation between expression and the episomes being ‘off’ the endogenous chromosomes, rather than attached, as detected by FISH (Table 1).
Staining for lacZ was also carried out on the CMT-93 cell lines. A fairly high level of background staining was observed for this cell line (whereas none was observed in the LA-9 cells). In addition, no significant staining above background was seen in any of the CMT-93 fusion cell lines (data not shown). This could be due to the low levels of CFTR expression, or the IRES-ß-geo may not function in this cell line or the IRES-ß-geo cassette could have become mutated in the yeast prior to formation of the cell lines.
EBNA-1 protein is expressed at variable levels
We observed very variable copy numbers of the elements, variable attachment to the host chromosomes and also variable levels of CFTR expression in the different cell lines. As the EBNA-1 protein is involved in segregation of EBV-based vectors and is also a transcriptional transactivator, we performed immunodetection on western blots to detect both the size and abundance of EBNA-1 (Fig. 9).
|
EBNA-1 protein has a molecular weight of ~80 kDa, as is observed in the EBNA-1-expressing cell line Akata, used as a positive control, but not in the parental LA-9 and CMT-93 cell lines (Fig. 9A and C). EBNA-1 protein is detected in LA9-B2, -B3 and -ori2 (LA9-ori3 was not analysed) (Fig. 9A) but it is smaller than that in the control cell line Akata. The EBNA-1 gene present in these cell lines was investigated by Southern analysis of AvaI-digested DNA (Fig. 9B). In all four cell lines, the 1122 bp AvaI band corresponding to the GlyAla repetitive domain is missing, indicating that this region has rearranged. The 467, 916 and 1453 bp AvaI fragments are present, indicating that the rest of the EBNA-1 gene is non-rearranged and it has simply rearranged in the repetitive region, which would account for the reduction in size of the EBNA-1 protein detected on the western blot. The rearrangement seems to be more dramatic than in the original YACs (Fig. 2B), and we do not know when the rearrangement occurred.
The level of expression of the truncated version of EBNA-1 in the LA-9 fusion cell lines does not correlate with the level of CFTR expression (which is lacking in LA9-B3) or the sticking of the episomes to the host chromosomes (which does not occur in LA9-B3).
The CMT-93 cell lines mostly have an EBNA-1 protein of the same size as that observed in the Akata cell line. CMT-25 and -7 appear to have a larger protein, although this is the same size as a cross-reacting band in the parental CMT-93 cells. High levels of the normal EBNA-1 band are observed in CMT-1, -6, -9, -16 and -25. Less intense bands are visible in CMT-3 and -13, and no band, or only a very weak band, is observed in CMT-2 and -10 (Fig. 9C). The CMT-14 sample is underloaded and CMT-7 only has the larger band, which may be EBNA-1 or a cross-reacting protein.
To investigate the possibility that absence of EBNA-1 protein in some cell lines was due to rearrangements in the EBNA-1 gene, we again looked at the AvaI restriction fragments present (Fig. 9D). All the CMT-93 fusion cell lines except CMT-25 seem to contain intact EBNA-1 genes. CMT-25 appears to lack the 1122 bp band which contains the GlyAla repetitive region, even though it contains normal sized EBNA-1 protein on the western blot. CMT-25 and -7, which possibly have a larger version of EBNA-1 on the western blot, may have a larger version of the 1122 bp AvaI fragment co-migrating with the larger 1453 bp band.
In the CMT-93-derived cell lines, the amount of intact EBNA-1 expressed is independent of the copy number of the yCFTR-oriP-EBNA1-neo molecules, the stability of the molecules in the absence of selection, whether they are attached to the chromo- somes and the level of CFTR gene expression.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
We describe a vector, pDH3, which allows YACs to be circularized and the oriP and EBNA-1 regions from EBV to be introduced. This vector worked efficiently to circularize a 320 kb YAC carrying the intact CFTR gene. Generally, the YACs were otherwise non-rearranged, apart from the repetitive GlyAla region of EBNA-1 which was non-rearranged in ~50% of the clones. However, non-rearranged versions can be screened for and are then fairly stable.
The oriP-EBNA1-YACs were transferred into two different mouse cell lines, by fusion with yeast spheroplasts, followed by selection to establish stable cell lines. The YACs were generally non-rearranged in the mouse cell lines and were maintained as circular episomes of either 320 or 640 kb. As few other rearrangements were observed, these molecules probably correspond to monomers and dimers of the input molecules. The copy number varied from ~2 to 56 in the different cell lines, and the rate of loss in the absence of selection was found to be between 0.4 and 5% per cell generation.
The copy number and rate of loss of the oriP-EBNA1-YACs are almost exactly the same as for the oriP-YACs that we analysed previously in human 293 cells expressing EBNA-1 (23) and are also very similar to other reports with oriP/EBNA-1 plasmids in human cells (4). Thus, it makes little difference whether the EBNA-1 is expressed in cis or in trans, and the two mouse cell lines are just as permissive for these molecules as are human cell lines. It has been shown previously that the ability of oriP/EBNA-1 vectors to be stably maintained in rodent as opposed to human cells is dependent on providing a fairly large amount of mammalian (or other) DNA on the vector which probably allows the molecules to replicate (14). It has also been observed that large molecules with 95 and 105 kb of human DNA, and carrying oriP and EBNA-1, are able to form stable episomes in mouse LA-9 cells (24).
The repetitive GlyAla region (IR3 region) of EBNA-1 is largely deleted in the LA-9 cell lines, but the episomes are maintained extrachromosomally and are as stable as those in the CMT-93 cells which have intact EBNA-1, indicating that the deleted EBNA-1 works efficiently in this context. In-frame deletions of the EBNA-1 IR3 region can be found in different strains of EBV (25) and also occur in recA+ E.coli hosts (26). Such deleted EBNA-1 proteins generally are able to allow oriP-containing plasmids to replicate and segregate as episomes (3,27); however, the efficiency of maintenance over long periods of time is reduced (28). Thus, although this region of EBNA-1 rearranges at high frequency during retrofitting of the YACs, this does not necessarily affect maintenance of the extrachromosomal molecules, and intact versions can be screened for. However, in gene therapy applications, deletion of the IR3 region may make the EBNA-1 protein more immunogenic, as the GlyAla repeat region has been shown to be able to mask proteins from cytotoxic lymphocyte surveillance (29).
One of the key aspects of this work is the investigation of the level of expression of the intact human CFTR gene from the oriP/EBNA-1 episomes. We previously have shown that the original 320 kb YAC drives physiological levels of CFTR expression in transgenic mice which is tissue specific and able to complement the Cambridge null mice (16). This YAC also drives full levels of position-independent expression in human Caco-2 cells (30). We have also quantified the level of expression from this YAC when it is integrated into CMT-93 cell lines; expression is copy number dependent and each copy of the YAC drives ~20% of the level of each endogenous mouse gene—the low level is probably due to the human gene in a mouse environment (unpublished data). Finally, we have shown that the version used in these experiments containing the IRES-ß-geo cassette is well expressed in human Caco-2 cells (20). Thus, we would expect the CFTR gene to be well expressed from the yCFTR-oriP-EBNA1-neo constructs in CMT-93-derived cell lines, especially in those cell lines with high copy number.
We observed fairly high levels of expression of the CFTR gene, as detected by non-quantitative RT–PCR and lacZ staining, in three of the four LA-9-derived cell lines. LA-9 cells do not express endogenous cftr, so expression of the human CFTR gene would not necessarily be expected. However, CFTR expression from this YAC has been observed previously in CHO-derived cell lines which also do not express endogenous cftr (31), indicating that expression of the CFTR gene from integrated YAC DNA is not dependent on tissue-specific transcription factors in these cell lines.
We also observe some CFTR gene expression in seven of the eleven CMT-93 cell lines but, apart from line CMT-1 where expression is twice as high as endogenous expression, the level of expression is only ~5–20% that of endogenous, and in two cell lines with non-rearranged copies of the YAC (CMT-7 and -10) there was no detectable expression. If the CFTR gene were being expressed in a copy number-dependent manner from the episomes, we would expect any line with a copy number > 5 to give as much human as mouse mRNA (the unmodified human YAC gave 20% as much expression per endogenous mouse gene copy), and in line CMT-7, which has a copy number of 56, we would expect ~10 times as much human as endogenous expression. Clearly the expression in the CMT-93 cell lines is not copy number dependent and is generally much lower than for the original YAC integrated into the genome.
The vast majority of experiments using oriP/EBNA-1 vectors for gene expression have used cDNA constructs driven by heterologous promoters. For example, good expression over several weeks has been observed for the human CFTR cDNA driven by an RSV promoter (32), the FACC cDNA driven from the RSV promoter (33), for the hypoxanthine phosphoribosyl- transferase (HPRT) cDNA driven from a retroviral promoter (34) and for granulocyte–macrophage colony-stimulating factor, interleukin-6 and tumour necrosis factor (TNF)-
driven by the cytomegalovirus promoter (35). Clearly strong, particularly viral, promoters can work well in the context of oriP/EBNA-1 vectors, and indeed the neor and hygr genes on our oriP-EBNA1-YACs work well.
In contrast, there are very few reports where intact mammalian genes have been expressed off such vectors. The intact TNF gene on a 35 kb cosmid, with oriP and EBNA-1, was introduced into human fibrosarcoma cells (which do not express endogenous TNF) and the correct sized mRNA was expressed, although at variable levels which were not related to copy number of the episomal DNA (36). The intact ß-globin cluster on a 200 kb bacterial artificial chromosome carrying oriP and EBNA-1 was expressed in D98/Raji cells as determined by RT–PCR, but these cells do not express endogenous ß-globin, so the relative level and level versus copy number of the elements were not determined (37). Thus, no one has related the levels of expression of an intact gene on oriP/EBNA-1 episomes with the copy number and the level of endogenous gene expression.
The variable levels of expression of the CFTR gene in the CMT cell lines does not seem to be related to the overall level of EBNA-1 expression, as both expressing and non-expressing cell lines have low and high levels of EBNA-1. It is possible that there is a trivial reason for the low expression of CFTR in the CMT-93 cell lines such as a subtle mutation in the CFTR gene or the IRES-ß-geo cassette severely affecting expression in these cell lines. However, it is also possible that there is a general effect of oriP and/or EBNA-1 on the CFTR promoter which causes down-regulation of expression. Such down-regulation could be related to chromatin structure and whether or not the extrachromosomal elements are physically attached to the chromosomes or are somewhat separate from the chromosomes, as observed by FISH. Although the correlation is not perfect (CMT-6 is the exception), all the cell lines where the majority of the elements are off the chromosomes have some expression, whereas those with most of the elements attached to the chromosomes have no expression (the clusters in CMT-3 seem to be associated with gene expression).
Vectors carrying oriP and EBNA-1 have the potential advantage for gene therapy of being fairly stably maintained in mammalian cells over many cell divisions without integration. They also have the capacity to carry very large regions of DNA; in this case, the intact 200 kb CFTR gene with enough flanking DNA to give tissue-specific and physiological levels of expression in transgenic mice. However, although we observe good expression in LA-9 cells which do not express endogenous cftr, expression was disappointingly low in CMT-93 cells which express endogenous cftr, and the expression was clearly not copy number dependent.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Yeast culture and oriP-EBNA1-YAC construction
The YAC constructs used here are based on a 310 kb YAC, 37AB12, carrying the intact human CFTR gene (38). A picornaviral IRES-ßgeo cassette was inserted into the 3' end of the CFTR gene, allowing lacZ expression to be driven by the CFTR promoter (20). This YAC (yCFIRES) was circularized with the plasmid pDH3.
pDH3 (Fig. 1A) contains the EBV oriP domain and EBNA-1 gene, as well as regions of homology to both ends of a linear YAC, the yeast selectable marker HIS5 and a mammalian hygromycin resistance cassette (hygr) for selection in mammalian cells. To construct pDH3, a 2.066 kb SalI fragment containing the yeast HIS5 gene from pCH45 (39) was inserted into SalI-digested pCEP4 (contains oriP, EBNA-1 and hygr; Invitrogen, Carlsbad, CA), replacing its polylinker to give pCEP4-HIS5. pPM680 (40) was digested with HindIII and SalI, blunt ended and ligated to eliminate undesired restriction sites. A 3.6 kb StuI–BsaAI fragment from pPM680, containing regions of homology to both arms of YACs based on pYAC4 (41), was ligated into pCEP4-HIS5 digested with ClaI and NruI and blunt ended.
pDH3 was linearized with ClaI and transfected into the yeast strain containing the linear YAC (yCFIRES) by spheroplast transformation (42). Transformants originally selected on His– plates were screened to find Trp+His+Lys– clones which arose at a frequency of 26 out of 100 colonies. Twelve clones were then analysed, of which 11 contained a circular YAC of the expected size.
The circular oriP-EBNA1-YAC clone number 11 was retrofitted with the plasmid pLUNA (43) to introduce a neomycin resistance (neor) gene as a selectable marker in mammalian cells. pLUNA was partially digested with ScaI to obtain a fragment with ends homologous to the ampicillin gene present in the oriP-EBNA1-YAC and transfected into the yeast by spheroplast transformation. Thirteen of 21 clones had the correct Ura+Lys+ phenotype.
Cell culture and yeast spheroplast fusion
CMT-93, a mouse epithelial rectal carcinoma cell line which expresses the murine cftr gene, was obtained from the American Type Culture Collection (Rockville, MD) (ATCC CCL-223). LA-9, a mouse fibroblast cell line which is negative for the hprt gene and does not express the murine cftr gene, was obtained from the NIGMS Human Genetic Mutant Cell Repository (Camden, NJ) (accession no. GM00346B). Caco-2, a human colon adeno- carcinoma cell line that expresses human CFTR, was obtained from the ATCC (HTB-37). CMT-93 and LA-9, and derived cell lines, were cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine serum at 37°C with 5% CO2.
Fusion of mammalian cells with yeast spheroplasts was carried out as previously described (22) to transfect the circular oriP-EBNA1-YACs into both cell lines. yCFTR-oriP-EBNA1 clones 10 and 11 (retrofitted with pDH3 but not pLUNA) were transfected into LA-9 cells and transformants were selected with hygromycin. The YAC yCFTR-oriP-EBNA1-neo (clone 11 retrofitted with pLUNA) was transfected into CMT-93 cells and transformants were selected with G418.
At 48 h after spheroplast fusion, the medium was replaced with fresh medium containing 600 µg/ml of hygromycin (Calbiochem-Novabiochem, Nottingham, UK) for LA-9 cells and 400 µg/ml of G418 (Gibco BRL, Paisley, UK) for CMT-93 cells. The medium was changed every week. After 3 weeks, single colonies were isolated using cloning cylinders and placed into 96-well plates containing selective medium. They were passaged sequentially to larger dishes until reaching >107 cells, at which point DNA was prepared.
Genomic DNA preparation and analysis
High molecular weight genomic DNA was prepared in agarose plugs as described previously for mammalian cell lines (23) and for yeast (22). -irradiation of DNA was performed in a Gamma- cell 1000 Elite apparatus on agarose plugs which had been equilibrated in 20% NDS (0.2% laurylsarcosine, 100 mM EDTA, 2 mM Tris pH 9). PFGE was carried out with 1% agarose gels cast and run in 0.5x TBE using a Bio-Rad (Hercules, CA) CHEF DRII or III apparatus at 14°C. DNA was transferred from agarose gels to Hybond N+ membrane (Amersham, Little Chalfont, UK) in alkaline solution as described by the manufacturer. DNA probes were labelled using the Megaprime labelling kit (Amersham), and hybridization and washing were carried out as previously described (23). Radioactivity was quantified with a phosphor- imager (Molecular Dynamics, Cresham, UK).
The oriP probe is a 0.9 kb SmaI fragment from the plasmid pCEP4 (Invitrogen). The oriP-EBNA-1 probe is the XbaI–ClaI fragment from the plasmid pDH3 which contains the EBNA-1 gene and the oriP region from EBV. The CFTR cDNA probe is the 4.7 kb PstI fragment from the plasmid pSVK4.7 (44). The human CFTR exon 4 probe is a 438 bp EcoRI fragment of subcloned PCR product generated using primers 4i-5' and 4i-3' (45). The mouse cftr exon 10 probe is a 450 bp EcoRI fragment of subcloned PCR product generated using the primers mEx105' (5'-GAACCTTAGTCCTATGTTGC-3') and mEx103' (5'-CTGACACAAGTAGCTAACAC-3'). The neor probe is the 1.1 kb XhoI–HindIII fragment from pMC1NeoPolyA (Stratagene, La Jolla, CA).
RNA analysis
Total RNA was extracted from different clones of CMT-93 and LA-9 mouse cells using RNA Isolator (Genosys, Pampisford, UK) according to the manufacturer’s instructions. To carry out the first strand cDNA synthesis, 2 µg of total RNA and 1 µg of random primers in a volume of 11 µl were incubated at 70°C for 10 min to remove secondary structure. Then 1 µl (40 U) of RNaseOUT ribonucleotide inhibitor (Gibco BRL), 4 µl of 5x first strand buffer (250 mM Tris–HCl pH 8.3, 375 mM KCl, 15 mM MgCl2), 2 µl of dithiothreitol (0.1 M) and 1 µl of dNTPs (10 mM each of dATP, dGTP, dCTP and dTTP) were added and the mixture incubated at 25°C for 10 min to allow the random primers to anneal. The mixture was equilibrated to 42°C for 2 min, then 1 µl (200 U) of SUPERSCRIPT II RNase H-reverse transcriptase (Gibco BRL) was added and the mixture incubated at 42°C for 50 min. The reaction was inactivated by 15 min incubation at 70°C.
The cDNA was then used to perform semi-quantitative RT–PCR with the primers h/mEx2-5' (5'-CTTCTGYTGATTCWGCTGAC-3') and h/mEx6a/b (5'-GATCTCTGTACTTCAYCATC-3') which anneal equally to the mouse and human sequences and amplify from exon 2 to exon 6a/b of the CFTR gene. A 2 µl aliquot of cDNA was used in a PCR of 50 µl with 1x PCR buffer (Qiagen, Crawley, UK) supplemented to 2.0 mM MgCl, 0.5 µM primers and 200 µM each dNTP. PCR was carried out with 30 cycles of 96°C for 20 s, 58°C for 1 min, 72°C for 1 min. For semi-quantitative RT–PCR, 0.01 µCi of [33P]dATP was included in the reaction mix.
These primers amplify a 610 bp fragment of both mouse and human CFTR mRNA. HindIII cuts only the mouse PCR product, generating fragments of 548 and 62 bp. NruI cuts only the human PCR product, generating fragments of 240 and 370 bp. The resulting fragments were resolved on a 1.5% agarose gel and visualized by ethidium bromide staining. For semi-quantitative RT–PCR, the gel was dried down and the radioactivity (sum above background) in the mouse 548 bp and the human 370 bp bands was determined with a phosphorimager. In order to account for the different size and AT content of the bands, the value for the 370 bp human band was divided by 221 and the value for the mouse 548 bp band was divided by 322 (the number of As and Ts in each fragment, respectively). The ratio of human to mouse expression is then given by the ratio of these final values. As the mouse and human sequences are so similar and the amplification uses a single pair of primers, we expect amplification of the two sequences to be at very similar rates and the proportions not to change even if the PCR is in the stationary phase. We used 30 cycles, which is just at the beginning of the stationary phase for all quantifications.
X-gal staining of CMT-93 cells to determine levels of ß-galactosidase expression
Cells were grown to confluency in 6-well plates before being washed in phosphate-buffered saline (PBS). Fix solution [5% formaldehyde (40%), 0.8% glutaraldehyde] was applied to the cells for 30 min at 4°C. This was followed by three 15 min washes at room temperature in wash buffer [100 µl of MgCl2 (1 M), 50 mg of deoxycholate, 400 µl of NP-40 (25%), in 50 ml of PBS]. X-gal stain [66 mg of potassium ferricyanide, 84 mg of potassium ferrocyanide and 230 µl of X-gal (40 mg/ml in dimethylsulfoxide) in 10 ml of wash buffer] was then applied for 14 h at 30°C with the 6-well plate wrapped in foil to prevent light exposure. Stained cells were visualized and photographed on an inverted light microscope.
FISH
Metaphase spreads were prepared from cell lines by standard cytogenetic techniques. LA-9 cells were incubated with 0.1 µg/ml of colcemid for 1 h prior to harvesting; however, no colcemid was added to CMT-93 cells because it prevents good chromosome spreading in this particular cell line. FISH was carried out as described elsewhere (46) with various modifications. The probe was a CFTR cosmid, L0421, kindly provided by Dr Ann Harris (Institute of Molecular Medicine, Oxford, UK), which contains the region from +14 to +50 kb of the human CFTR gene. The CFTR cosmid was labelled with biotin-14-dATP using the BioNick Labelling system (Gibco BRL).
Slides were denatured by incubation in 0.8 M NaOH, 70% ethanol for 4 min at room temperature. A 400 ng aliquot of CFTR cosmid probe and 20 µg of E.coli tRNA in 50% formamide, 2x SSC, 10% dextran sulfate were hybridized per slide at 37°C overnight. Washing and detection were carried out as follows. Slides were washed twice in 50% formamide, 2x SSC at 45°C for 5 min, then three times in 2x SSC at 45°C for 5 min. A blocking step was followed by incubation in 200 µl of 3% bovine serum albumin (BSA, fraction V; Sigma, Poole, UK), 4x SSC for 30 min at 37°C. Probe detection was with avidin–FITC [200 µl of 5 µg/ml avidin–fluorescein isothiocyanate (Sigma), 1% BSA, 4x SSC, 0.1% Tween-20] for 30 min in a moist chamber at 37°C, followed by three washes in 4x SSC, 0.1% Tween-20 for 5 min at 42°C. Amplification of the signal was with 1 µg of anti-avidin D (Vector Laboratories, Burlingame, CA) in 1% BSA, 4x SSC, 0.1% Tween-20 for 30 min at 37°C, followed by three washes in 4x SSC, 0.1% Tween-20 for 5 min at 42°C. An avidin–FITC incubation was repeated, followed by a wash step as described above. Chromosomes were counterstained with a commercial propidium iodide/antifade solution (Sigma). The images were detected and captured with a Leica (Milton Keynes, UK) TCS NT Laser confocal microscope, with Kr/Ar laser.
Detection of EBNA-1 protein by western blot analysis
A total of 107 cells were resuspended in 300 µl of protein sample buffer [PSB (60 mM Tris pH 6.8, 2% PBS, 10% glycerol, 1% ß-mercaptoethanol, bromophenol blue)], then sonicated for 5 min and heated to 95°C for 5 min prior to loading. A 50 µl aliquot of each sample was loaded on a 10% polyacrylamide gel and run overnight at 50 V in protein running buffer (3 g/l Tris–HCl, 14.4 g/l glycine, 1% SDS).
Blotting of the gel was carried out in a Bio-Rad blotting system for 4 h at 250 mA at room temperature. Then, a 1 h blocking step in 5% dried skimmed milk powder, 1x PBS was carried out at room temperature. EBNA-1 was detected using a human serum (PF) from an EBV carrier (courtesy of Professor Paul Farrell, Ludwig Institute for Cancer Research, Imperial College School of Medicine, London, UK) at a dilution of 1:100 in 5% dried skimmed milk powder, incubated at 4°C overnight in a shaker, followed by three 15 min washes in 0.1% NP-40, 1x PBS at room temperature. Incubation with a secondary horseradish peroxidase-coupled rabbit anti-human antibody was carried out at a dilution of 1:1000 in 5% dried skimmed milk powder for 1 h at room temperature. Washes in 0.1% NP-40 were repeated as described above. Detection of secondary antibody binding was with protein A-conjugated avidin (ECL; Amersham) following the manufacturer’s instructions. The Akata cell line (47) which expresses the viral EBNA-1 protein was used as a positive control.
|
ACKNOWLEDGEMENTS |
|---|
We would like to thank Prof. Paul Farrell and Barbara Wensing for assistance with the analysis of EBNA-1 expression by western blotting, and Prof. F. Azorin for his support during this work. D.H. was an EU Fellow, and this work was supported by the MRC.
|
FOOTNOTES |
|---|
+ Present address: Department of Molecular and Cell Biology, CID-CSIC, Barcelona, Spain
§ To whom correspondence should be addressed. Tel: +44 171 549 3028; Fax: +44 171 594 3015; Email: c.huxley@ic.ac.uk
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Yates, J., Warren, N., Reisman, D. and Sugden, B. (1984) A cis-acting element from the Epstein–Barr viral genome that permits stable replication of recombinant plasmids in latently infected cells. Proc. Natl Acad. Sci. USA, 81, 3806–3810.
2 Lupton, S. and Levine, A.J. (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.
3 Yates, J.L., Warren, N. and Sugden, B. (1985) Stable replication of plasmids derived from Epstein–Barr virus in various mammalian cells. Nature, 313, 812–815.[Medline]
4 Reisman, D., Yates, J. and Sugden, B. (1985) A putative origin of replication of plasmids derived from Epstein–Barr virus is composed of two cis-acting components. Mol. Cell. Biol., 5, 1822–1832.
5 Rawlins, D.R., Milman, G., Hayward, S.D. and Hayward, G.S. (1985) Sequence-specific DNA binding of the Epstein–Barr virus nuclear antigen (EBNA-1) to clustered sites in the plasmid maintenance region. Cell, 42, 859–868.[Web of Science][Medline]
6 Gahn, T.A. and Schildkraut, C.L. (1989) The Epstein–Barr virus origin of plasmid replication, oriP, contains both the initiation and termination sites of DNA replication. Cell, 58, 527–535.[Web of Science][Medline]
7 Krysan, P.J., Haase, S.B. and Calos, M.P. (1989) Isolation of human sequences that replicate autonomously in human cells. Mol. Cell. Biol., 9, 1026–1033.
8 Reisman, D. and Sugden, B. (1986) Trans activation of an Epstein–Barr viral transcriptional enhancer by the Epstein–Barr viral nuclear antigen 1. Mol. Cell. Biol., 6, 3838–3846.
9 Calos, M.P. (1996) The potential of extrachromosomal replicating vectors for gene therapy. Trends Genet., 12, 463–466.[Web of Science][Medline]
10 Längle-Rouault, F., Patzel, V., Benavente, A., Taillez, M., Silvestre, N., Bompard, A., Sczakiel, G., Jacobs, E. and Rittner, K. (1998) Up to 100-fold increase of apparent gene expression in the presence of Epstein–Barr virus oriP sequences and EBNA1: implications of the nuclear import of plasmids. J. Virol., 72, 6181–6185.
11 Wohlgemuth, J.G., Kang, S.H., Bulboaca, G.H., Nawotka, K.A. and Calos, M.P. (1996) Long-term gene expression from autonomously replicating vectors in mammalian cells. Gene Ther., 3, 503–512.[Web of Science][Medline]
12 Saeki, Y., Wataya-Kaneda, M., Tanaka, K. and Kaneda, Y. (1998) Sustained transgene expression in vitro and in vivo using an Epstein–Barr virus replicon vector system combined with HVJ liposomes. Gene Ther., 5, 1031–1037.[Web of Science][Medline]
13 Tsukamoto, H., Wells, D.J., Brown, S.C., Serpente, P., Strong, P.N., Drew, J., Inui, K., Okada, S. and Dickson, G. (1999) Enhanced expression of recombinant dystrophin following intramuscular injection of Epstein–Barr virus (EBV)-based mini-chromosome vectors in mdx mice. Gene Ther., 6, 1331–1335.[Web of Science][Medline]
14 Krysan, P.J. and Calos, M.P. (1993) Epstein–Barr virus-based vectors that replicate in rodent cells. Gene, 136, 137–143.[Web of Science][Medline]
15 Alton, E.W.F.W., Geddes, D.M., Gill, D.R., Higgins, C.F., Hyde, S.C., Innes, J.A. and Porteous, D.J. (1998) Towards gene therapy for cystic fibrosis: a clinical progress report. Gene Ther., 5, 291–292.[Web of Science][Medline]
16 Manson, A.L., Trezise, A.E.O., MacVinish, L.J., Kasschau, K.D., Birchall, N., Episkopou, V., Vassaux, G., Evans, M.J., Colledge, W.H., Cuthbert, A.W. and Huxley, C. (1997) Complementation of null CF mice with a human CFTR YAC transgene. EMBO J., 16, 4238–4249.[Web of Science][Medline]
17 Whitsett, J.A., Dey, C.R., Stripp, B.R., Wikenheiser, K.A., Clark, J.C., Wert, S.E., Gregory, R.J., Smith, A.E., Cohn, J.A., Wilson, J.M. and Engelhardt, J. (1992) Human cystic fibrosis transmembrane conductance regulator directed to respiratory epithelial cells of transgenic mice. Nature Genet., 2, 13–20.[Web of Science][Medline]
18 Zhou, L., Dey, C.R., Wert, S.E., DuVall, M.D., Frizzell, R.A. and Whitsett, J.A. (1994) Correction of lethal intestinal defect in a mouse model of cystic fibrosis by human CFTR. Science, 266, 1705–1708.
19 Rozmahel, R., Gyömörey, K., Plyte, S., Nguyen, V., Wilschanski, M., Durie, P., Bear, C.E. and Tsui, L.-C. (1997) Incomplete rescue of cystic fibrosis transmembrane conductance regulator deficient mice by the human CFTR cDNA. Hum. Mol. Genet., 6, 1153–1162.
20 Vassaux, G. and Huxley, C. (1997) A dicistronic construct allows easy detection of human CFTR expression from YAC DNA in human cells. Nucleic Acids Res., 25, 4167–4168.
21 Heller, M., van Santen, V. and Kieff, E. (1982) Simple repeat sequence in Epstein–Barr virus DNA is transcribed in latent and productive infections. J. Virol., 44, 311–320.
22 Huxley, C., Hagino, Y., Schlessinger, D. and Olson, M.V. (1991) The human HPRT gene on a yeast artificial chromosome is functional when transferred to mouse cells by cell fusion. Genomics, 9, 742–750.[Web of Science][Medline]
23 Simpson, K., McGuigan, A. and Huxley, C. (1996) Stable episomal maintenance of yeast artificial chromosomes in human cells. Mol. Cell. Biol., 16, 5117–5126.[Abstract]
24 Kelleher, Z.T., Fu, H., Livanos, E., Wendelburg, B., Gulino, S. and Vos, J.-M. (1998) Epstein–Barr-based episomal chromosomes shuttle 100 kb of self-replicating circular human DNA in mouse cells. Nature Biotechnol., 16, 762–768.[Web of Science][Medline]
25 Hennessy, K., Heller, M., van Santen, V. and Kieff, E. (1983) Simple repeat array in Epstein–Barr virus DNA encodes part of the Epstein–Barr nuclear antigen. Science, 220, 1396–1398.
26 Buell, G.N., Reisman, D., Kintner, C., Crouse, G. and Sugden, B. (1981) Cloning overlapping DNA fragments from the B95-8 strain of Epstein–Barr virus reveals a site of homology to the internal repetition. J. Virol., 40, 977–982.
27 Yates, J.L. and Camiolo, S.M. (1988) Dissection of DNA replication and enhancer activation functions of Epstein–Barr virus nuclear antigen 1. Cancer Cells, 6, 197–205.
28 Wendelburg, B.J. and Vos, J.-M.H. (1998) An enhanced EBNA1 variant with reduced IR3 domain for long-term episomal maintenance and transgene expression of oriP-based plasmids in human cells. Gene Ther., 5, 1389–1399.[Web of Science][Medline]
29 Levitskaya, J., Coram, M., Levitsky, V., Imreh, S., Steigerwald-Mullen, P.M., Klein, G., Kurilla, M.G. and Masucci, M.G. (1995) Inhibition of antigen processing by the internal repeat region of the Epstein–Barr virus nuclear antigen-1. Nature, 375, 685–688.[Medline]
30 Vassaux, G., Manson, A.L. and Huxley, C. (1997) Copy number-dependent expression of a YAC-cloned human CFTR gene in a human epithelial cell line. Gene Ther., 4, 618–623.[Web of Science][Medline]
31 Mogayzel, P.J., Henning, K.A., Bittner, M.L., Novotny, E.A., Schwiebert, E.M., Guggino, W.B., Jiang, Y. and Rosenfeld, M.A. (1997) Functional human CFTR produced by stable Chinese hamster ovary cell lines derived using yeast artificial chromosomes. Hum. Mol. Genet., 6, 59–68.
32 Lei, D.C., Kunzelmann, K., Koslowsky, T., Yezzi, M.J., Escobar, L.C., Xu, Z., Ellison, A.R., Rommens, J.M., Tsui, L.-C., Tykocinski, M. and Gruenert, D.C. (1996) Episomal expression of wild-type CFTR corrects cAMP-dependent chloride transport in respiratory epithelial cells. Gene Ther., 3, 427–436.[Web of Science][Medline]
33 Banerjee, S., Livanos, E. and Vos, J.-M.H. (1995) Therapeutic gene delivery in human B-lymphoblastoid cells by engineered non-transforming infectious Epstein–Barr virus. Nature Med., 1, 1303–1308.[Web of Science][Medline]
34 Sun, T.-Q., Livanos, E. and Vos, J.-M.H. (1996) Engineering a mini-herpesvirus as a general strategy to transduce up to 180 kb of functional self-replicating human mini-chromosomes. Gene Ther., 3, 1081–1088.[Web of Science][Medline]
35 Mücke, S., Polack, A., Pawlita, M., Zehnpfennig, D., Massoudi, N., Bohlen, H., Doerfler, W., Bornkamm, G., Diehl, V. and Wolf, J. (1997) Suitability of Epstein–Barr virus-based episomal vectors for expression of cytokine genes in human lymphoma cells. Gene Ther., 4, 82–92.[Web of Science][Medline]
36 Kioussis, D., Wilson, F., Daniels, C., Leveton, C., Taverne, J. and Playfair, J.H.L. (1987) Expression and rescuing of a cloned human tumour necrosis factor gene using an EBV-based shuttle cosmid vector. EMBO J., 6, 355–361.[Web of Science][Medline]
37 Westphal, E.M., Sierakowska, H., Livanos, E., Kole, R. and Vos, J.-M. (1998) A system for shuttling 200-kb BAC/PAC clones into human cells: stable extrachromosomal persistence and long-term ectopic gene activation. Hum. Gene Ther., 9, 1863–1873.[Web of Science][Medline]
38 Anand, R., Ogilvie, D.J., Butler, R., Riley, J.H., Finniear, R.S., Powell, S.J., Smith, J.C. and Markham, A.F. (1991) A yeast artificial chromosome contig encompassing the cystic fibrosis locus. Genomics, 9, 124–130.[Web of Science][Medline]
39 Tolmachova, T., Simpson, K. and Huxley, C. (1999) Analysis of a YAC with human telomeres and oriP from Epstein–Barr virus in yeast and 293 cells. Nucleic Acids Res., 27, 3736–3744.
40 Garza, D., Ajioka, J.W., Burke, D.T. and Hartl, D.L. (1989) Mapping the Drosophila genome with yeast artificial chromosomes. Science, 246, 641–646.
41 Burke, D.T., Carle, G.F. and Olson, M.V. (1987) Cloning of large segments of exogenous DNA into yeast by means of artificial chromosome vectors. Science, 236, 806–812.
42 Burgers, P.M.J. and Percival, K.J. (1987) Transformation of yeast spheroplasts without cell fusion. Anal. Biochem., 163, 391–397.[Web of Science][Medline]
43 Davies, N.P., Rosewell, I.R. and Brüggemann, M. (1992) Targeted alterations in yeast artificial chromosomes for inter-species gene transfer. Nucleic Acids Res., 20, 2693–2698.
44 Alton, E.W.F.W., Middleton, P.G., Caplen, N.J., Smith, S.N., Steel, D.M., Munkonge, F.M., Jeffery, P.K., Geddes, D.M., Hart, S.L., Williamson, R. et al. (1993) Non-invasive liposome-mediated gene delivery can correct the ion transport defect in cystic fibrosis mutant mice. Nature Genet., 5, 135–142.[Web of Science][Medline]
45 Zielenski, J., Rozmahel, R., Bozon, D., Kerem, B.-S., Grzelczak, Z., Riordan, J.R., Rommens, J. and Tsui, L.-C. (1991) Genomic DNA sequence of the cystic fibrosis transmembrane conductance regulator (CFTR) gene. Genomics, 10, 214–228.[Web of Science][Medline]
46 Johnson, C.V., Singer, R.H. and Lawrence, J.B. (1991) Fluorescent detection of nuclear RNA and DNA: implications for genome organization. Methods Cell Biol., 35, 73–99.[Medline]
47 Takada, K. and Ono, Y. (1989) Synchronous and sequential activation of latently infected Epstein–Barr virus genomes. J. Virol., 63, 445–449.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||