- Email: [email protected]
Construction of a swine artificial chromosome: a novel vector for transgenesis in the pig
Biochimie 84 (2002) 1143–1150 www.elsevier.com/locate/biochi
Construction of a swine artificial chromosome: a novel vector for transgenesis in the pig Paola Poggiali a, Gian Luca Scoarughi a, Marialuisa Lavitrano b, Pierluigi Donini a, Carmen Cimmino a,* a
Dipartimento di Biologia Cellulare e dello Sviluppo, University of Rome “La Sapienza”, Via dei Sardi, 70, 00185 Roma, Italy b Dipartimento di Medicina Sperimentale Ambientale e Biotecnologie Mediche, University of Milano-Bicocca, Via Cadore, 48, 20052 Monza, Milan, Italy Received 21 June 2002; accepted 18 October 2002
Abstract A de novo SAC was constructed by making use of YAC technology and a humanized yeast strain. The construct (about 310 kb) contained pig centromeric DNA and the Neo gene. The construct was introduced into a pig cell line by yeast-mammalian cell fusion and G418 resistant clones were obtained. One clone was characterized by FISH and shown to contain an episomally located microchromosome containing YAC, Neo and pig centromere sequences. FISH analysis over time showed that the SAC was mitotically stable for at least 34 generations in the absence of selection. The size of the SAC was determined by confocal microscopy of the SAC and shown to be approximately 7 Mb, which is about 25-fold greater than the size of the original YAC. From its behavior in pulsed field gel electrophoresis, FISH analysis of stretched DNA fibers, and its appearance under scanning confocal microscopy, it was concluded that the SAC is a circularized and multimerized derivative of the original YAC. Possible applications as vectors for animal transgenesis are discussed. © 2002 Éditions scientifiques et médicales Elsevier SAS and Société française de biochimie et biologie moléculaire. All rights reserved. Keywords: Pig; Microchromosome; Artificial chromosome; MAC; SAC
1. Introduction Current gene transfer systems make use of viral vectors and are only partially satisfactory for gene therapy and animal biotechnology: the vector remains in the transformed cell for a number of generations or else becomes randomly integrated in the host genome, possibly producing insertional mutagenesis, gene silencing, or loss of control of expression of the inserted gene. Conversely, mammalian artificial chromosomes (MACs) are a novel class of episomal vectors that rely exclusively on the replication and segregation mechanisms of natural chromosomes. They have a number of advantages over other vectors including the capacity for carrying much larger genes, proper control of gene expression,
Abbreviations: SAC, swine artificial chromosome; pCGS990MegaNeo plasmid. * Corresponding author. Tel.: +39-06-4991-7588; fax: +39-06-4991-7594. E-mail address: [email protected] (C. Cimmino)
pCG,
maintenance of gene copy number, and the absence of side effects such as immunological responses and cell transformation. Several groups have been successful in generating stable human artificial chromosomes (HACs) following two main strategies: “top down” and “bottom up”. The “top down” approach is based on fragmentation of natural linear mammalian chromosomes using telomere-directed fragmentation or exposure to irradiation to reduce their size, making them easier to manipulate with molecular biology techniques (minichromosomes) [1–3]. The de novo “bottom up” strategy consists in the in vitro assembly of all the components necessary for the stable maintenance of a linear mammalian chromosome, i.e. telomeres, a centromere, and one or more replication origins (microchromosomes). A functional centromere is of the utmost importance for the production of artificial chromosomes. Several studies in this field have focused on the use of human alpha-satellite (alphoid) DNA, a family of repetitive sequences specific to primate centromeres and constituting their major nucleic acid component [4–6]. It was thus possible to show that alphoid DNA can form a new centromere with origin function in a HAC.
© 2002 Éditions scientifiques et médicales Elsevier SAS and Société française de biochimie et biologie moléculaire. All rights reserved. PII: S 0 3 0 0 - 9 0 8 4 ( 0 2 ) 0 0 0 1 9 - 6
1144
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Several methods have been used to obtain microchromosomes: one is based on the fact that DNA molecules transfected into mammalian cells are efficiently joined by nonhomologous recombination. Harrington and co-workers [6] transfected human alphoid, telomeric, and genomic DNA into human cells, producing different combinations of DNA molecules, including some containing properly positioned elements necessary for autonomous replication and mitotic stability. Transfection into the HT1080 cell line produced microchromosomes with sizes ranging from 6 to 10 Mb. Some of the clones were analyzed in detail and it was shown that one contained only DNA that had been introduced by the transfection. Another technique used for the construction of minichromosomes is based on the use of yeast artificial chromosomes (YACs) appropriately modified in such a way as to contain human telomeric, centromeric, and genomic DNA. To construct microchromosomes following this approach, use is made of YACs containing human DNA alphoid sequences. The ends of the YACs are replaced with human telomeric sequences by yeast homologous recombination and a selective marker for mammalian cells is inserted. The resulting microchromosomes are mitotically stable and are 10 to 25 times larger than the original YAC [7–9]. It has recently been shown that microchromosomes can be obtained by assembly of chromosomal elements in circular form, in this case the presence of telomeres is superfluous. It is possible to obtain artificial chromosomes based on BAC or PAC bacterial plasmids. These circular prokaryotic vectors can accommodate DNA inserts up to a size of 300 kb. Circular BAC/PAC constructs have been used that lack telomeric sequences and that contain human centromeric inserts and a selective marker for mammalian cells. When these vectors were transferred into cultured human cells, they gave rise with high efficiency to MACs that were mitotically stable for many generations in the absence of selection [10–12]. The use of MACs as gene transfer vectors for the production of transgenic animals is in its initial stages. In this report, we describe the construction of the first SAC that can be used as a stable gene expression vector and will probably furnish an important contribution to the creation of transgenic pigs, organisms of great interest for their potential applications in several fields such as xenotransplantation and other fields of medicine, biotechnology and animal husbandry. The SAC was constructed according to the “bottom up” approach, making use of a YAC containing centromeric repeated porcine alphoid DNA obtained from a swine YAC library [13].
2. Materials and methods
domain of the telomerase RNA gene, TLC1, had been replaced with a human telomere template domain and is thus capable of adding human telomere sequences to the ends of yeast chromosomes [14]. The swine cell line PK15 was cultured in low glucose DMEM medium supplemented with 10% fetal calf serum, 2 mM L-glutamine, 50 U ml–1 penicillin and 50 mg ml–1 streptomycin (Gibco Brl). Plasmid pCG (16.6 kb), kindly provided by B. Grimes, contains 1.1 kb of mammalian TTAGGG telomere repeats flanked by 0.3 kb of yeast (TG1–3) telomere repeats, the CEN4 region derived from YAC4, the neomycin resistance gene (Neo) that confers resistance to the neomycin analog G418 for selection in mammalian cells and the yeast lys2 gene [8]. Plasmid pHutel, kindly furnished by E. Gilson, contains 10 repeats of the human telomeric sequence TTAGGG. 2.2. Southern analysis Southern hybridizations were carried out at 62 °C according to Sambrook et al. [15] and all the probes used were prepared by the Roche Nick Translation kit using [a32P] dCTP. 2.3. Transfer of the YAC into humanized yeast strain Saccharomyces cerevisiae strain AB1380 (Mat a ade2-1 can1-100 lys2-1 trp1 ura3 his5-u) carrying YAC 225D04B was crossed with the humanized haploid strain a4 (mat a) [14] and plated onto YMIN media to select for diploid colonies. Selected colonies were verified to be diploid by analysis of yeast chromosomes by pulsed field gel electrophoresis (PFGE). Diploid strains were sporulated, tetrads were dissected and haploid spores were regenerated on YMIN media lacking uracil and tryptophane to select for a4 clones carrying YAC 225D04B. Haploid status (a4/225D04B) was verified by genotyping at the Mat locus as described [16]. PFGE and Southern analysis with a human telomeric probe, obtained from double digested EcoRI–HindIII of the pHutel plasmid, confirmed the presence of TTAGGG repeats at the ends of YAC 225D04B and of the yeast chromosomes. 2.4. Introduction of a mammalian selectable marker by retrofitting The left arm replacement vector, pCG was linearized with SalI and transfected by the lithium acetate method into a4/225D04B. The transformants were plated onto YMIN medium lacking uracil and lysine to select for clones in which the left arm of YAC 225D04B had been replaced by retrofitting plasmid containing the Neo gene.
2.1. Microorganisms, cell lines, plasmids 2.5. Transfer of YAC into swine cells Saccharomyces cerevisiae strains AB1380 (Mat a ade2-1 can1-100 lys2-1 trp1 ura3 his5-u) carrying YAC 225D04B containing swine centromeric DNA [13]. Humanized haploid strain a4 (mat a) in which the entire 16 bp template
Yeast strain d1Neo2 was grown on selective YMIN media lacking uracil and lysine. Cells were spheroplasted and fused with PK15 cells as described [17]. PK15 cells containing
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
YACs were selected in low glucose DMEM media containing 500 µg ml–1 G418 (Life Technologies). 2.6. PCR and primers PCR amplification of the Neo gene was performed using primers Neo1 (5'-gaacaagatggattgcacgcaggt-3') and Neo2 (5'-gaactcgtcaagaaggcgatagaag-3') (Gibco Brl) under the following conditions: 94 °C for 5 min (94 °C for 15 s, 64 °C for 45 s, 72 °C for 3 min 30 s) × 30 cycles, 72 °C for 10 min. Centromeric probe for FISH and Southern analysis was generated by PCR amplification of swine total genomic DNA with the primer pair SSCR2 (EMBL Accession number X62139). These primers amplify a 340 bp fragment under the following conditions: 96 °C for 5 min; (94 °C for 1 min, 50 °C for 1 min, 72 °C for 1 min 30 s) × 30 cycles; 72 °C for 10 min. 2.7. Fluorescence in situ hybridization (FISH) In situ hybridization was carried out by standard techniques. Cells were treated with 2 µg ml–1 colcemid for 4 h and fixed with methanol–acetic acid 3:1. Artificial chromosomes in PK15 clones were detected in the presence of 15 µg µl–1 of salmon sperm with a biotin and digoxigenin labeled pCG probe (Nick Translation kit, Roche diagnostic) and biotin labeled swine centromeric probe. The biotin labeled probes were detected using the sandwich method with avidin–fluorescein isothiocyanate (FITC) and biotinylated anti-avidin (Vector). Digoxigenin labeled probe was detected with a sheep monoclonal anti-digoxigenin rhodamine conjugate antibody followed by rabbit anti-sheep IgG antibody Texas red conjugate (Vector) and a final layer of goat antirabbit IgG antibody Texas red conjugate (Vector) for increasing sensitivity. Hybridizations were performed under low stringency conditions (42 °C). DNA was counterstained with DAPI or propidium iodide and the slides were mounted using Vectashield mounting medium (Vector). 2.8. DNA fiber FISH Stretched DNA fibers were prepared as described [18]. Ten micro liters of the cell suspension at 106 cells ml–1 in PBS (phosphate buffered saline) were placed on a microscope slide and air-dried at 45–50 °C. The slide was placed in a vertical position and the cells were lysed with 0.05 M NaOH in 28.6% ethanol. The resulting extended chromatin fibers were fixed with methanol and dehydrated with increasingly concentrated ethanol. DNA fibers were analyzed by FISH after denaturation in 70% formamide, 2X SSC for 2 min at 70 °C. 2.9. Fluorescence microscopy Biotin and digoxigenin detection and microscopic analysis of metaphase slides and DNA fiber FISH were performed with a Nikon fluorescence microscope. Propidium iodide
1145
stained chromosomes and metaphase FISH were also analyzed using a Biorad radians 2000 laser scanning confocal microscope with IP-Lab software and a Zeiss LSM510Meta laser-scanning confocal microscope with Zeiss LSM Image Browser software.
3. Results
3.1. Modification of a YAC containing pig alphoid sequences Telomere DNA sequences of all vertebrates studied share the same regular repeat (TTAGGG)n, whereas the telomere sequences in the budding yeast Saccharomyces cerevisiae is a GC rich irregular sequence (TG1–3). Humanized yeast strains are capable of adding human telomere sequences to the ends of yeast chromosomes and are therefore potentially useful for the assembly and propagation of mammalian artificial chromosomes. YAC 225D04B was transferred to the humanized yeast strain by crossing strains 225D04B and a4. Eight diploid strains, obtained from the 225D04B × a4 cross were sporulated and the tetrads were dissected. Three haploid strains growing on selective medium lacking uracil and tryptophane were obtained from the spores. All strains were analyzed by PFGE to verify presence, linearity and size of the YAC (about 300 kb) and subsequently hybridized with a human telomeric probe (Fig. 1A): one haploid strain, designated d1 and containing the YAC with the centromeric insert and human telomeric ends was chosen for further modification. The YAC contained in d1 was modified by retrofitting to introduce the Neo mammalian selectable marker Neo (Fig. 1B). Plasmid pCG, used for retrofitting, contains homology to the left arm of YAC d1 and carries the yeast lys2 gene and Neo gene. YACs containing d1 yeast Lys+ transformants were inspected for a Trp– phenotype, as the latter would be likely to have arisen as a result of specific integration of the retrofitting plasmid into the YAC arm. After introduction of linearized pCG into yeast d1, 10 transformants were obtained that grew on Lys+, Ura+ selective medium. Four strains were analyzed by PFGE and hybridized with a Neo probe. One of these, YAC d1Neo2, containing all of the appropriate components, was analyzed by digestion with several restriction endonucleases followed by hybridization with swine centromeric probe and total pig genomic DNA probe. This analysis revealed that the insert of the YAC consisted prevalently of repetitive centromeric DNA with a small fraction of genomic non-centromeric DNA ranging between 14 and 25 kb in size (data not shown). As a result of the retrofitting, the YACs increased in size by about 11 kb. YAC d1Neo2 appeared to remain stable during further propagation in yeast and was chosen for further analysis in swine cells.
1146
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Fig. 1. Modification of YAC 225D04B in S. cerevisiae. (A) PFGE and Southern analysis of S. cerevisiae haploid strains derived from a 225D04B × a4 cross with human telomeric probe. Size markers are expressed in kilobases. Lane M: PFGE markers 50–1000 kb; lane 1: strain AB1380 carrying YAC 225D04 as negative control; lane 2: haploid strain b3; lane 3: haploid strain d1; lane 4: haploid strain e1. TheYAC is indicated by the arrow on the right. (B) Schematic representation of the retrofitting procedure for insertion of the Neo gene into theYAC contained in the haploid strain d1. The left arm of theYAC was replaced with pCG plasmid linearized by digestion with ClaI. The construct obtained was designated d1Neo2.
3.2. Microchromosome d1Neo2 in swine PK15 cells YAC d1Neo2 was transferred into the swine cell line PK15 by means of yeast-mammalian cell fusion with the PEG1500 method. After fusion, yeast chromosomes were gradually eliminated from pig cells. Five days after cell fusion, transformant cells were selected with G418 and 5 resistant clones were obtained. One clone designated PK15d1Neo2 that grew for over six months under G418 selection was chosen for detailed analysis. PCR analysis for the Neo gene was performed using specific primers (Neo1 and Neo2) and using PK15d1Neo2 total genomic DNA as the template and PK15 total genomic DNA as a negative control. This analysis confirmed the presence in PK15d1Neo2 clonal cell line of the Neo gene located on the left arm of the original YAC d1Neo2 (data not shown). To determine whether the
modified alphoid YAC was maintained extrachromosomally in PK15 cells after G418 selection, the clonal cell line PK15d1Neo2 was subjected to FISH analysis (Fig. 2A). FISH was performed using the pCG probe, specific for the microchromosome, and corresponding to the left arm of the original YAC d1Neo2, and indicated that a proportion of metaphases analyzed contained a microchromosome detectable as overlapping signals with biotynilated probe and DAPI or propidium iodide signals. Microchromosome copy number per cell was consistently one, but the specific signals were detectable in less than 50% of the total signals from the metaphase spreads analyzed. No integration events were observed. FISH was also performed using the pCG probe and a centromeric probe obtained by PCR amplification of swine total genomic DNA with SSCSR2 primers, that in standard stringency conditions labels all swine chromosomes exceptY
Fig. 2. Merged image of FISH analysis of the clonal cell line PK15d1Neo2. (A) Metaphase chromosomes were hybridized with pCG probe (green signal) and counterstained with DAPI. The white arrow indicates the d1Neo2 microchromosome. Upper-right panel shows a 2 × magnification. (B) Hybridization with centromeric probe (green signal) obtained from PCR amplification at total genomic swine DNA with SSCR2 primers. Metaphase chromosomes were counterstained with DAPI and the microchromosome is indicated by the white arrow. Upper-right panel shows a 2 × magnification.
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
1147
Table 1 Microchromosome stability obtained by Fish-analysis of PK15d1Neo2 clonal cell line, in presence and in absence of G418 selection. PK15 clonal cell line was used as a negative control Cell line
Generations w/o selection
PK15 PK15d1Neo2 PK15d1Neo2 PK15 PK15d1Neo2 PK15d1Neo2
0 0 13 0 0 34
Fish analysis Metaphases analyzed
Microchr.
30 30 30 30 30 30
0 15 9 0 11 11
(Fig. 2B). It was possible to observe the overlapping of the signals relative to the two probes, thus confirming the episomal location of the microchromosome. 3.3. Mitotic stability PK15d1Neo2 cells were maintained in the presence or absence of G418 and analyzed at various times by PCR amplification of the Neo gene to verify the retention of the microchromosome and by FISH for its extrachromosomall maintenance. PCR analysis was performed after 13 and 26 generations in the absence of selection and produced no differences in the intensity of the specific bands (data not shown). The results of FISH analysis are shown in Table 1 and confirm the functional nature of the microchromosome centromere, since the microchromosome is efficiently retained in PK15d1Neo2 cells after 13 and 34 generations without selection. The screening was performed on 30 metaphase spreads for every time point considered and did not show significant differences in the fraction of cells that contained the microchromosome signal.
Microchr./cell
Positive cells
0
1
2
30 17 21 30 19 19
0 11 9 0 11 11
0 2 0 0 0 0
0% 43.3% 30% 0% 36.6% 36.6%
counted in the metaphase spreads and the DNA content of pig chromosomes 11, 12, 17, 18, and Y was obtained from published Karyotype data [19]. It was thus possible to estimate a size for the microchromosome ranging between 6.5 and 7.2 Mb, which is about 20–25 fold greater than the input YAC (about 310 kb). Since a linear 5.5 Mb molecule enters PFGE gels under the conditions used by us [3] it must be concluded that d1Neo2 possesses a non-linear, presumably circular structure. Additional information on the structure of d1Neo2 was obtained by FISH analysis of stretched DNA fibers using the pCG probe and the swine centromeric probe. The signals detected for the two probes appear as alternating consecutive blocks, strongly indicating that the large size of the minichromosome is due to multimerization of the original input YAC microchromosome DNA fiber as consecutive segments of irregular size (Fig. 4A). The structure of d1Neo2 was further analyzed by Zeiss LSM510Meta laser scanning confocal microscopy that enables visualization of the minichromosome at high magnifi-
3.4. Microchromosome size and structure To characterize the size and the structure of microchromosome d1Neo2, genomic DNA extracts of PK15d1Neo2 cells, undigested and digested with several restriction endonucleases, were embedded in agarose plugs and subjected to PFGE under conditions programmed for bands in the 500 kb–5 Mb range. No bands were visible in the dried gels after hybridization with Neo and swine centromeric probes. The explanation for this negative result could be either that the size of the d1Neo2 genome is greater than 7 Mb or that it does not have a linear structure. The size of the DNA contained in the microchromosome was therefore established by measuring the SAC with Biorad radians 2000 scanning confocal microscopy (Fig. 3). PK15d1Neo2 metaphase chromosomes were treated with Propidium iodide that confers a homogeneous red signal to DNA. The areas of the signals corresponding to the smallest swine endogenous chromosomes were evaluated using IPLab software that permits to calculate the number of pixels of a certain color in a selected area. The red pixels for 5 of each of these chromosomes were
Fig. 3. Metaphase spread of clonal cell line PK15d1Neo2 containing the d1Neo2 microchromosome visualized by laser scanning confocal microscopy. DNA was counterstained with propidium iodide.
1148
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Fig. 4. Analysis of the structure of d1Neo2 microchromosome. (A) Dual hybridization of stretched DNA fibers from the PK15d1Neo2 cell line with centromeric (green signal) and pCG (red signal) probes. The signals show that both sequences are consecutive not overlapping segments of irregular lengths. The gap between the segments is expected where the sequences are not represented in the probes used. (B–G) Laser scanning confocal microscopy analysis of d1Neo2 microchromosome with Zeiss LSM Image Browser software. Upper panels and lower panels are two different views of the d1Neo2 microchromosome. (B, E) Hybridization with pCG probe. (C, F) Hybridization with centromeric probe. (D, G) Merged image.
cation and determination of the exact position of the individual fluorescent signals. The results of this analysis, given in Fig. 4(B–G), show that d1Neo2 appears to have a spherical or ovoidal shape that is not consistent with a linear structure. Fig. 4(B–G) also shows that the hybridization signal for the two probes used can be visualized as large patches of approximately equal dimensions on the surface of the DNA structure. This result is very different from the one obtained with the linear MC1 minichromosome, where the centromeric region is distinctly located in the central region of the molecule, while the terminal telomeric regions are present only at the two ends, as expected. 4. Discussion The results reported in this study show that YAC technology can be used to construct SACs following the bottom up approach. The swine centromeric DNA contained in the YAC
has the same sequence organization as human alphoid DNA with monomers tandemly repeated head to tail, but have a different DNA composition. This type of sequence has not been found to contain CENP-B binding sites, which have been shown to be important for kinetochore seeding and centromere function in the formation of human artificial chromosomes [4]. In spite of this, the centromeric function provided by pig satellite DNA is normal as shown by FISH analysis of mitotic stability. This could mean that unidentified CENP-B-like sequences are present in the insert of the YAC used or that the repeat structures suffice by themselves for the assembly of active centromere/kinetochore components into SACs. However, the minichromosome is detected in approximately 40% of the metaphase spreads and is always considerably distant from the grouped chromosomes that make up the spread. The linear minichromosome MCI was shown to be detectable in metaphase spreads with 70% efficiency and was more centrally located in the spreads [20].
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Apart from structure, an important difference between the two MACs is that MCI is a cut down version of a natural chromosome, presumably containing MAR sequences that anchor the chromosomes to the nuclear matrix. Thus d1Neo2 is likely to have a propensity to move away from the cluster of chromosomes in metaphase spreads, thus escaping detection. The results stemming from the experiments performed with stretched DNA fibers and laser scanning confocal microscopy strongly indicate that d1Neo2 has a circular molecular structure consisting of a highly metamerized array of alternating centromeric and non-centromeric sequences and that the DNA is compacted into an ovoidally shaped geometric solid. Hybridization signals corresponding at the two probes used in the latter experiments, the pCG probe and the swine centromeric probe, are grouped in large homogeneous patches, indicating that the DNA structure may consist of adjacent radially oriented loops having a periodicity that brings regions of similar sequence composition to the surface of the molecule. The periodicity may not coincide perfectly with the length of the molecule, causing transitions in the region of the initial input DNA that surfaces on the outside of the SAC molecule. The increase in size of the initial DNA construct that gives rise to an artificial chromosome that is considerably larger than input is a general rule for the generation of MACs by the bottom up approach [4–6]. The incomplete structural information presented here is consistent with the notion that the size increase is produced by a number of multimerization events of the original construct. One can speculate that MACs produced by the bottom up approach can either replicate efficiently or segregate efficiently only if their size is in the 5 Mb range or larger. Since it has been shown that centromeric DNA tracts with a size of 100 kb can provide efficient centromeric function in linear minichromosomes [5], it is likely that large MACs are selected for on the basis of increased replication efficiency. In fact, plasmids into which single mammalian replication origins are inserted unaccompanied by additional sequences such as MARs or specialized viral sequences do not replicate in mammalian cells [21]. In order to utilize the SAC constructed in this study for animal transgenesis, genes of interest will be added to the microchromosome. This can be done in the initial phases of the construction, before introduction into the mammalian cell, thus obtaining a gene copy number of about 15. A gene copy number of one can be obtained by inserting the gene into the SAC in its final form; this can be accomplished most easily if the SAC is transferred into a host endowed with an efficient homologous recombination system, such as the DT40 chicken strain, or conceivably after transfer and propagation in S. cerevisiae. Artificial chromosomes are currently being used in the mouse and other animals for the study of centromere function; human artificial chromosomes are being considered as possible vectors for gene therapy; pig artificial chromosomes, such as the SAC constructed and characterized in this
1149
study, will very likely be useful episomal vectors for producing transgenic pigs for xenotransplantation and other purposes important for medicine and biotechnology.
Acknowledgements We thank Patrick Chardon for furnishing YAC 225D04B. Thanks are also due to Fiorentina Ascenzioni for useful discussions, for helpful collaboration in the theory and practice of YAC technology and for preparing and providing the humanized yeast strain. We are grateful to Barbara Barboni, Mauro Mattioli, Manfred Brich and Carl Zeiss Jena GmbH for assistance with confocal microscopy. This work was supported by a grant from the MURST (Modalità alternative di transgenesi animale: ricerche scientifiche sulla costruzione di mini- e microcromosomi artificiali).
References [1]
M.A. Barnett, V.J. Buckle, E.P. Evans, A.C.G. Porter, D. Rout, A.G. Smith, W.R.A. Brown, Telomere directed fragmentation of mammalian chromosomes, Nucl. Acid. Res. 21 (1993) 27–36. [2] W. Mills, R. Critcher, C. Lee, C.J. Farr, Generation of an z2.4 Mb human X centromere-based minichromosome by targeted telomereassociated chromosome fragmentation in DT40, Hum. Mol. Gen. 8 (1999) 751–761. [3] C. Guiducci, F. Ascenzioni, C. Auriche, E. Piccolella, A.M. Guerrini, P. Donini, Use of a human minichromosome as a cloning and expression vector foe mammalian cells, Hum. Mol. Gen. 8 (1999) 1417–1424. [4] J.E. Mejia, A. Alazami, A. Willmott, P. Marschall, E. Levy, W.C. Earnshaw, Z. Larin, Efficiency of de novo centromere formation in human artificial chromosomes, Genomics 79 (2002) 297–304. [5] J.W. Yang, C. Pendon, J. Yang, N. Haywood, A. Chand, W.R.A. Brown, Human mini-chromosomes with minimal centromeres, Hum. Mol. Gen. 9 (2000) 1891–1902. [6] J.J. Harrington, G.V. Bokkelen, R.W. Mays, K. Gustashaw, H.F. Willard, Formation of de novo centromeres and construction of firstgeneration human artificial minichromosomes, Nature Gen. 15 (1997) 345–355. [7] M. Ikeno, B. Grimes, T. Okazaki, M. Nakano, K. Saitoh, H. Hoshino, N.I. McGill, H. Cooke, H. Masumoto, Construction of YAC-based mammalian artificial chromosomes, Nature Biotech. 16 (1998) 431–439. [8] K.A. Henning, E.A. Novotny, S.T. Compton, X.Y. Guan, P.P. Liu, M.A. Ashlock, Human artificial chromosomes generated by modification of a yeast artificial chromosome containing both human alpha satellite and single-copy DNA sequences, Proc. Natl. Acad. Sci. 96 (1999) 592–597. [9] B. Grimes, H. Cooke, Engineering mammalian chromosomes, Hum. Mol. Gen. 7 (1998) 1635–1640. [10] J.E. Mejia, A. Willmott, E. Levy, W.C. Earnshaw, Z. Larin, Functional complementation of a genetic deficiency with human artificial chromosomes, Am. J. Hum. Genet. 69 (2001) 315–326. [11] B. Grimes, D. Schindelhauer, N.I. McGill, A. Ross, T.A. Ebersole, H.J. Cooke, Stable gene expression from a mammalian artificial chromosome, EMBO Rep. 21 (2001) 910–914. [12] T.A. Ebersole, A. Ross, E. Clark, N. McGill, D. Schindelhauer, H. Cooke, B. Grimes, Mammalian artificial chromosome formation from circular alphoid input DNA does not require telomere repeats, Hum. Mol. Gen. 9 (2000) 1623–1631.
1150
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
[13] C. Rogel-Gaillard, N. Bourgeaux, J.C. Save, C. Renard, P. Coullin, P. Pinton, M. Yerle, M. Vaiman, P. Chardon, Construction of a swine YAC library allowing an efficient recovery of unique and centromeric repeated sequence, Mamm. Genome 8 (1997) 186–192. [14] K.A. Henning, N. Moskowitz, M.A. Ashlock, P. Liu Pu, Humanizing the yeast telomerase template, Proc. Natl. Acad. Sci. 95 (1998) 5667–5671. [15] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: a Laboratory Manual, 2nd ed, Cold Spring Harbor Laboratory Press, New York, 1989. [16] G.A. Silverman,YAC protocols, in: D. Marlie (Ed.), Humana, Totowa, NY, 1996, pp. 199–216. [17] V. Pachnis, L. Perny, R. Rothstein, F. Costantini, Transfer of a yeast artificial chromosome carrying human DNA from Saccharomyces cerevisiae into mammalian cells, Proc. Natl. Acad. Sci. 87 (1990) 5109–5113.
[18] T. Haaf, D.C. Ward, Structural analysis of a-satellite DNA and centromere proteins using extended chromatin and chromosomes, Hum. Mol. Gen. 3 (1994) 697–709. [19] A. Schmitz, B. Chaput, P. Fouchet, M.N. Guilly, G. Frelat, M. Vaiman, Swine chromosomal DNA quantification by bivariate flow karyotyping and karyotype interpretation, Cytometry 13 (7) (1992) 703–710. [20] C. Auriche, P. Donini, F. Ascenzioni, Molecular and cytological analysis of a 5.5 Mb minichromosome, EMBO Reports 2 (2) (2001) 102–107. [21] A. Baiker, C. Maercker, C. Piechaczek, S.B. Schmidt, J. Bode, C. Benham, H.J. Lipps, Mitotic stability of an episomal vector containing a human scaffold/matrix attached region is provided by association with nuclear matrix, Nature Cell. Biol. 2 (3) (2000) 182–184.
Construction of a swine artificial chromosome: a novel vector for transgenesis in the pig Paola Poggiali a, Gian Luca Scoarughi a, Marialuisa Lavitrano b, Pierluigi Donini a, Carmen Cimmino a,* a
Dipartimento di Biologia Cellulare e dello Sviluppo, University of Rome “La Sapienza”, Via dei Sardi, 70, 00185 Roma, Italy b Dipartimento di Medicina Sperimentale Ambientale e Biotecnologie Mediche, University of Milano-Bicocca, Via Cadore, 48, 20052 Monza, Milan, Italy Received 21 June 2002; accepted 18 October 2002
Abstract A de novo SAC was constructed by making use of YAC technology and a humanized yeast strain. The construct (about 310 kb) contained pig centromeric DNA and the Neo gene. The construct was introduced into a pig cell line by yeast-mammalian cell fusion and G418 resistant clones were obtained. One clone was characterized by FISH and shown to contain an episomally located microchromosome containing YAC, Neo and pig centromere sequences. FISH analysis over time showed that the SAC was mitotically stable for at least 34 generations in the absence of selection. The size of the SAC was determined by confocal microscopy of the SAC and shown to be approximately 7 Mb, which is about 25-fold greater than the size of the original YAC. From its behavior in pulsed field gel electrophoresis, FISH analysis of stretched DNA fibers, and its appearance under scanning confocal microscopy, it was concluded that the SAC is a circularized and multimerized derivative of the original YAC. Possible applications as vectors for animal transgenesis are discussed. © 2002 Éditions scientifiques et médicales Elsevier SAS and Société française de biochimie et biologie moléculaire. All rights reserved. Keywords: Pig; Microchromosome; Artificial chromosome; MAC; SAC
1. Introduction Current gene transfer systems make use of viral vectors and are only partially satisfactory for gene therapy and animal biotechnology: the vector remains in the transformed cell for a number of generations or else becomes randomly integrated in the host genome, possibly producing insertional mutagenesis, gene silencing, or loss of control of expression of the inserted gene. Conversely, mammalian artificial chromosomes (MACs) are a novel class of episomal vectors that rely exclusively on the replication and segregation mechanisms of natural chromosomes. They have a number of advantages over other vectors including the capacity for carrying much larger genes, proper control of gene expression,
Abbreviations: SAC, swine artificial chromosome; pCGS990MegaNeo plasmid. * Corresponding author. Tel.: +39-06-4991-7588; fax: +39-06-4991-7594. E-mail address: [email protected] (C. Cimmino)
pCG,
maintenance of gene copy number, and the absence of side effects such as immunological responses and cell transformation. Several groups have been successful in generating stable human artificial chromosomes (HACs) following two main strategies: “top down” and “bottom up”. The “top down” approach is based on fragmentation of natural linear mammalian chromosomes using telomere-directed fragmentation or exposure to irradiation to reduce their size, making them easier to manipulate with molecular biology techniques (minichromosomes) [1–3]. The de novo “bottom up” strategy consists in the in vitro assembly of all the components necessary for the stable maintenance of a linear mammalian chromosome, i.e. telomeres, a centromere, and one or more replication origins (microchromosomes). A functional centromere is of the utmost importance for the production of artificial chromosomes. Several studies in this field have focused on the use of human alpha-satellite (alphoid) DNA, a family of repetitive sequences specific to primate centromeres and constituting their major nucleic acid component [4–6]. It was thus possible to show that alphoid DNA can form a new centromere with origin function in a HAC.
© 2002 Éditions scientifiques et médicales Elsevier SAS and Société française de biochimie et biologie moléculaire. All rights reserved. PII: S 0 3 0 0 - 9 0 8 4 ( 0 2 ) 0 0 0 1 9 - 6
1144
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Several methods have been used to obtain microchromosomes: one is based on the fact that DNA molecules transfected into mammalian cells are efficiently joined by nonhomologous recombination. Harrington and co-workers [6] transfected human alphoid, telomeric, and genomic DNA into human cells, producing different combinations of DNA molecules, including some containing properly positioned elements necessary for autonomous replication and mitotic stability. Transfection into the HT1080 cell line produced microchromosomes with sizes ranging from 6 to 10 Mb. Some of the clones were analyzed in detail and it was shown that one contained only DNA that had been introduced by the transfection. Another technique used for the construction of minichromosomes is based on the use of yeast artificial chromosomes (YACs) appropriately modified in such a way as to contain human telomeric, centromeric, and genomic DNA. To construct microchromosomes following this approach, use is made of YACs containing human DNA alphoid sequences. The ends of the YACs are replaced with human telomeric sequences by yeast homologous recombination and a selective marker for mammalian cells is inserted. The resulting microchromosomes are mitotically stable and are 10 to 25 times larger than the original YAC [7–9]. It has recently been shown that microchromosomes can be obtained by assembly of chromosomal elements in circular form, in this case the presence of telomeres is superfluous. It is possible to obtain artificial chromosomes based on BAC or PAC bacterial plasmids. These circular prokaryotic vectors can accommodate DNA inserts up to a size of 300 kb. Circular BAC/PAC constructs have been used that lack telomeric sequences and that contain human centromeric inserts and a selective marker for mammalian cells. When these vectors were transferred into cultured human cells, they gave rise with high efficiency to MACs that were mitotically stable for many generations in the absence of selection [10–12]. The use of MACs as gene transfer vectors for the production of transgenic animals is in its initial stages. In this report, we describe the construction of the first SAC that can be used as a stable gene expression vector and will probably furnish an important contribution to the creation of transgenic pigs, organisms of great interest for their potential applications in several fields such as xenotransplantation and other fields of medicine, biotechnology and animal husbandry. The SAC was constructed according to the “bottom up” approach, making use of a YAC containing centromeric repeated porcine alphoid DNA obtained from a swine YAC library [13].
2. Materials and methods
domain of the telomerase RNA gene, TLC1, had been replaced with a human telomere template domain and is thus capable of adding human telomere sequences to the ends of yeast chromosomes [14]. The swine cell line PK15 was cultured in low glucose DMEM medium supplemented with 10% fetal calf serum, 2 mM L-glutamine, 50 U ml–1 penicillin and 50 mg ml–1 streptomycin (Gibco Brl). Plasmid pCG (16.6 kb), kindly provided by B. Grimes, contains 1.1 kb of mammalian TTAGGG telomere repeats flanked by 0.3 kb of yeast (TG1–3) telomere repeats, the CEN4 region derived from YAC4, the neomycin resistance gene (Neo) that confers resistance to the neomycin analog G418 for selection in mammalian cells and the yeast lys2 gene [8]. Plasmid pHutel, kindly furnished by E. Gilson, contains 10 repeats of the human telomeric sequence TTAGGG. 2.2. Southern analysis Southern hybridizations were carried out at 62 °C according to Sambrook et al. [15] and all the probes used were prepared by the Roche Nick Translation kit using [a32P] dCTP. 2.3. Transfer of the YAC into humanized yeast strain Saccharomyces cerevisiae strain AB1380 (Mat a ade2-1 can1-100 lys2-1 trp1 ura3 his5-u) carrying YAC 225D04B was crossed with the humanized haploid strain a4 (mat a) [14] and plated onto YMIN media to select for diploid colonies. Selected colonies were verified to be diploid by analysis of yeast chromosomes by pulsed field gel electrophoresis (PFGE). Diploid strains were sporulated, tetrads were dissected and haploid spores were regenerated on YMIN media lacking uracil and tryptophane to select for a4 clones carrying YAC 225D04B. Haploid status (a4/225D04B) was verified by genotyping at the Mat locus as described [16]. PFGE and Southern analysis with a human telomeric probe, obtained from double digested EcoRI–HindIII of the pHutel plasmid, confirmed the presence of TTAGGG repeats at the ends of YAC 225D04B and of the yeast chromosomes. 2.4. Introduction of a mammalian selectable marker by retrofitting The left arm replacement vector, pCG was linearized with SalI and transfected by the lithium acetate method into a4/225D04B. The transformants were plated onto YMIN medium lacking uracil and lysine to select for clones in which the left arm of YAC 225D04B had been replaced by retrofitting plasmid containing the Neo gene.
2.1. Microorganisms, cell lines, plasmids 2.5. Transfer of YAC into swine cells Saccharomyces cerevisiae strains AB1380 (Mat a ade2-1 can1-100 lys2-1 trp1 ura3 his5-u) carrying YAC 225D04B containing swine centromeric DNA [13]. Humanized haploid strain a4 (mat a) in which the entire 16 bp template
Yeast strain d1Neo2 was grown on selective YMIN media lacking uracil and lysine. Cells were spheroplasted and fused with PK15 cells as described [17]. PK15 cells containing
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
YACs were selected in low glucose DMEM media containing 500 µg ml–1 G418 (Life Technologies). 2.6. PCR and primers PCR amplification of the Neo gene was performed using primers Neo1 (5'-gaacaagatggattgcacgcaggt-3') and Neo2 (5'-gaactcgtcaagaaggcgatagaag-3') (Gibco Brl) under the following conditions: 94 °C for 5 min (94 °C for 15 s, 64 °C for 45 s, 72 °C for 3 min 30 s) × 30 cycles, 72 °C for 10 min. Centromeric probe for FISH and Southern analysis was generated by PCR amplification of swine total genomic DNA with the primer pair SSCR2 (EMBL Accession number X62139). These primers amplify a 340 bp fragment under the following conditions: 96 °C for 5 min; (94 °C for 1 min, 50 °C for 1 min, 72 °C for 1 min 30 s) × 30 cycles; 72 °C for 10 min. 2.7. Fluorescence in situ hybridization (FISH) In situ hybridization was carried out by standard techniques. Cells were treated with 2 µg ml–1 colcemid for 4 h and fixed with methanol–acetic acid 3:1. Artificial chromosomes in PK15 clones were detected in the presence of 15 µg µl–1 of salmon sperm with a biotin and digoxigenin labeled pCG probe (Nick Translation kit, Roche diagnostic) and biotin labeled swine centromeric probe. The biotin labeled probes were detected using the sandwich method with avidin–fluorescein isothiocyanate (FITC) and biotinylated anti-avidin (Vector). Digoxigenin labeled probe was detected with a sheep monoclonal anti-digoxigenin rhodamine conjugate antibody followed by rabbit anti-sheep IgG antibody Texas red conjugate (Vector) and a final layer of goat antirabbit IgG antibody Texas red conjugate (Vector) for increasing sensitivity. Hybridizations were performed under low stringency conditions (42 °C). DNA was counterstained with DAPI or propidium iodide and the slides were mounted using Vectashield mounting medium (Vector). 2.8. DNA fiber FISH Stretched DNA fibers were prepared as described [18]. Ten micro liters of the cell suspension at 106 cells ml–1 in PBS (phosphate buffered saline) were placed on a microscope slide and air-dried at 45–50 °C. The slide was placed in a vertical position and the cells were lysed with 0.05 M NaOH in 28.6% ethanol. The resulting extended chromatin fibers were fixed with methanol and dehydrated with increasingly concentrated ethanol. DNA fibers were analyzed by FISH after denaturation in 70% formamide, 2X SSC for 2 min at 70 °C. 2.9. Fluorescence microscopy Biotin and digoxigenin detection and microscopic analysis of metaphase slides and DNA fiber FISH were performed with a Nikon fluorescence microscope. Propidium iodide
1145
stained chromosomes and metaphase FISH were also analyzed using a Biorad radians 2000 laser scanning confocal microscope with IP-Lab software and a Zeiss LSM510Meta laser-scanning confocal microscope with Zeiss LSM Image Browser software.
3. Results
3.1. Modification of a YAC containing pig alphoid sequences Telomere DNA sequences of all vertebrates studied share the same regular repeat (TTAGGG)n, whereas the telomere sequences in the budding yeast Saccharomyces cerevisiae is a GC rich irregular sequence (TG1–3). Humanized yeast strains are capable of adding human telomere sequences to the ends of yeast chromosomes and are therefore potentially useful for the assembly and propagation of mammalian artificial chromosomes. YAC 225D04B was transferred to the humanized yeast strain by crossing strains 225D04B and a4. Eight diploid strains, obtained from the 225D04B × a4 cross were sporulated and the tetrads were dissected. Three haploid strains growing on selective medium lacking uracil and tryptophane were obtained from the spores. All strains were analyzed by PFGE to verify presence, linearity and size of the YAC (about 300 kb) and subsequently hybridized with a human telomeric probe (Fig. 1A): one haploid strain, designated d1 and containing the YAC with the centromeric insert and human telomeric ends was chosen for further modification. The YAC contained in d1 was modified by retrofitting to introduce the Neo mammalian selectable marker Neo (Fig. 1B). Plasmid pCG, used for retrofitting, contains homology to the left arm of YAC d1 and carries the yeast lys2 gene and Neo gene. YACs containing d1 yeast Lys+ transformants were inspected for a Trp– phenotype, as the latter would be likely to have arisen as a result of specific integration of the retrofitting plasmid into the YAC arm. After introduction of linearized pCG into yeast d1, 10 transformants were obtained that grew on Lys+, Ura+ selective medium. Four strains were analyzed by PFGE and hybridized with a Neo probe. One of these, YAC d1Neo2, containing all of the appropriate components, was analyzed by digestion with several restriction endonucleases followed by hybridization with swine centromeric probe and total pig genomic DNA probe. This analysis revealed that the insert of the YAC consisted prevalently of repetitive centromeric DNA with a small fraction of genomic non-centromeric DNA ranging between 14 and 25 kb in size (data not shown). As a result of the retrofitting, the YACs increased in size by about 11 kb. YAC d1Neo2 appeared to remain stable during further propagation in yeast and was chosen for further analysis in swine cells.
1146
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Fig. 1. Modification of YAC 225D04B in S. cerevisiae. (A) PFGE and Southern analysis of S. cerevisiae haploid strains derived from a 225D04B × a4 cross with human telomeric probe. Size markers are expressed in kilobases. Lane M: PFGE markers 50–1000 kb; lane 1: strain AB1380 carrying YAC 225D04 as negative control; lane 2: haploid strain b3; lane 3: haploid strain d1; lane 4: haploid strain e1. TheYAC is indicated by the arrow on the right. (B) Schematic representation of the retrofitting procedure for insertion of the Neo gene into theYAC contained in the haploid strain d1. The left arm of theYAC was replaced with pCG plasmid linearized by digestion with ClaI. The construct obtained was designated d1Neo2.
3.2. Microchromosome d1Neo2 in swine PK15 cells YAC d1Neo2 was transferred into the swine cell line PK15 by means of yeast-mammalian cell fusion with the PEG1500 method. After fusion, yeast chromosomes were gradually eliminated from pig cells. Five days after cell fusion, transformant cells were selected with G418 and 5 resistant clones were obtained. One clone designated PK15d1Neo2 that grew for over six months under G418 selection was chosen for detailed analysis. PCR analysis for the Neo gene was performed using specific primers (Neo1 and Neo2) and using PK15d1Neo2 total genomic DNA as the template and PK15 total genomic DNA as a negative control. This analysis confirmed the presence in PK15d1Neo2 clonal cell line of the Neo gene located on the left arm of the original YAC d1Neo2 (data not shown). To determine whether the
modified alphoid YAC was maintained extrachromosomally in PK15 cells after G418 selection, the clonal cell line PK15d1Neo2 was subjected to FISH analysis (Fig. 2A). FISH was performed using the pCG probe, specific for the microchromosome, and corresponding to the left arm of the original YAC d1Neo2, and indicated that a proportion of metaphases analyzed contained a microchromosome detectable as overlapping signals with biotynilated probe and DAPI or propidium iodide signals. Microchromosome copy number per cell was consistently one, but the specific signals were detectable in less than 50% of the total signals from the metaphase spreads analyzed. No integration events were observed. FISH was also performed using the pCG probe and a centromeric probe obtained by PCR amplification of swine total genomic DNA with SSCSR2 primers, that in standard stringency conditions labels all swine chromosomes exceptY
Fig. 2. Merged image of FISH analysis of the clonal cell line PK15d1Neo2. (A) Metaphase chromosomes were hybridized with pCG probe (green signal) and counterstained with DAPI. The white arrow indicates the d1Neo2 microchromosome. Upper-right panel shows a 2 × magnification. (B) Hybridization with centromeric probe (green signal) obtained from PCR amplification at total genomic swine DNA with SSCR2 primers. Metaphase chromosomes were counterstained with DAPI and the microchromosome is indicated by the white arrow. Upper-right panel shows a 2 × magnification.
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
1147
Table 1 Microchromosome stability obtained by Fish-analysis of PK15d1Neo2 clonal cell line, in presence and in absence of G418 selection. PK15 clonal cell line was used as a negative control Cell line
Generations w/o selection
PK15 PK15d1Neo2 PK15d1Neo2 PK15 PK15d1Neo2 PK15d1Neo2
0 0 13 0 0 34
Fish analysis Metaphases analyzed
Microchr.
30 30 30 30 30 30
0 15 9 0 11 11
(Fig. 2B). It was possible to observe the overlapping of the signals relative to the two probes, thus confirming the episomal location of the microchromosome. 3.3. Mitotic stability PK15d1Neo2 cells were maintained in the presence or absence of G418 and analyzed at various times by PCR amplification of the Neo gene to verify the retention of the microchromosome and by FISH for its extrachromosomall maintenance. PCR analysis was performed after 13 and 26 generations in the absence of selection and produced no differences in the intensity of the specific bands (data not shown). The results of FISH analysis are shown in Table 1 and confirm the functional nature of the microchromosome centromere, since the microchromosome is efficiently retained in PK15d1Neo2 cells after 13 and 34 generations without selection. The screening was performed on 30 metaphase spreads for every time point considered and did not show significant differences in the fraction of cells that contained the microchromosome signal.
Microchr./cell
Positive cells
0
1
2
30 17 21 30 19 19
0 11 9 0 11 11
0 2 0 0 0 0
0% 43.3% 30% 0% 36.6% 36.6%
counted in the metaphase spreads and the DNA content of pig chromosomes 11, 12, 17, 18, and Y was obtained from published Karyotype data [19]. It was thus possible to estimate a size for the microchromosome ranging between 6.5 and 7.2 Mb, which is about 20–25 fold greater than the input YAC (about 310 kb). Since a linear 5.5 Mb molecule enters PFGE gels under the conditions used by us [3] it must be concluded that d1Neo2 possesses a non-linear, presumably circular structure. Additional information on the structure of d1Neo2 was obtained by FISH analysis of stretched DNA fibers using the pCG probe and the swine centromeric probe. The signals detected for the two probes appear as alternating consecutive blocks, strongly indicating that the large size of the minichromosome is due to multimerization of the original input YAC microchromosome DNA fiber as consecutive segments of irregular size (Fig. 4A). The structure of d1Neo2 was further analyzed by Zeiss LSM510Meta laser scanning confocal microscopy that enables visualization of the minichromosome at high magnifi-
3.4. Microchromosome size and structure To characterize the size and the structure of microchromosome d1Neo2, genomic DNA extracts of PK15d1Neo2 cells, undigested and digested with several restriction endonucleases, were embedded in agarose plugs and subjected to PFGE under conditions programmed for bands in the 500 kb–5 Mb range. No bands were visible in the dried gels after hybridization with Neo and swine centromeric probes. The explanation for this negative result could be either that the size of the d1Neo2 genome is greater than 7 Mb or that it does not have a linear structure. The size of the DNA contained in the microchromosome was therefore established by measuring the SAC with Biorad radians 2000 scanning confocal microscopy (Fig. 3). PK15d1Neo2 metaphase chromosomes were treated with Propidium iodide that confers a homogeneous red signal to DNA. The areas of the signals corresponding to the smallest swine endogenous chromosomes were evaluated using IPLab software that permits to calculate the number of pixels of a certain color in a selected area. The red pixels for 5 of each of these chromosomes were
Fig. 3. Metaphase spread of clonal cell line PK15d1Neo2 containing the d1Neo2 microchromosome visualized by laser scanning confocal microscopy. DNA was counterstained with propidium iodide.
1148
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Fig. 4. Analysis of the structure of d1Neo2 microchromosome. (A) Dual hybridization of stretched DNA fibers from the PK15d1Neo2 cell line with centromeric (green signal) and pCG (red signal) probes. The signals show that both sequences are consecutive not overlapping segments of irregular lengths. The gap between the segments is expected where the sequences are not represented in the probes used. (B–G) Laser scanning confocal microscopy analysis of d1Neo2 microchromosome with Zeiss LSM Image Browser software. Upper panels and lower panels are two different views of the d1Neo2 microchromosome. (B, E) Hybridization with pCG probe. (C, F) Hybridization with centromeric probe. (D, G) Merged image.
cation and determination of the exact position of the individual fluorescent signals. The results of this analysis, given in Fig. 4(B–G), show that d1Neo2 appears to have a spherical or ovoidal shape that is not consistent with a linear structure. Fig. 4(B–G) also shows that the hybridization signal for the two probes used can be visualized as large patches of approximately equal dimensions on the surface of the DNA structure. This result is very different from the one obtained with the linear MC1 minichromosome, where the centromeric region is distinctly located in the central region of the molecule, while the terminal telomeric regions are present only at the two ends, as expected. 4. Discussion The results reported in this study show that YAC technology can be used to construct SACs following the bottom up approach. The swine centromeric DNA contained in the YAC
has the same sequence organization as human alphoid DNA with monomers tandemly repeated head to tail, but have a different DNA composition. This type of sequence has not been found to contain CENP-B binding sites, which have been shown to be important for kinetochore seeding and centromere function in the formation of human artificial chromosomes [4]. In spite of this, the centromeric function provided by pig satellite DNA is normal as shown by FISH analysis of mitotic stability. This could mean that unidentified CENP-B-like sequences are present in the insert of the YAC used or that the repeat structures suffice by themselves for the assembly of active centromere/kinetochore components into SACs. However, the minichromosome is detected in approximately 40% of the metaphase spreads and is always considerably distant from the grouped chromosomes that make up the spread. The linear minichromosome MCI was shown to be detectable in metaphase spreads with 70% efficiency and was more centrally located in the spreads [20].
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
Apart from structure, an important difference between the two MACs is that MCI is a cut down version of a natural chromosome, presumably containing MAR sequences that anchor the chromosomes to the nuclear matrix. Thus d1Neo2 is likely to have a propensity to move away from the cluster of chromosomes in metaphase spreads, thus escaping detection. The results stemming from the experiments performed with stretched DNA fibers and laser scanning confocal microscopy strongly indicate that d1Neo2 has a circular molecular structure consisting of a highly metamerized array of alternating centromeric and non-centromeric sequences and that the DNA is compacted into an ovoidally shaped geometric solid. Hybridization signals corresponding at the two probes used in the latter experiments, the pCG probe and the swine centromeric probe, are grouped in large homogeneous patches, indicating that the DNA structure may consist of adjacent radially oriented loops having a periodicity that brings regions of similar sequence composition to the surface of the molecule. The periodicity may not coincide perfectly with the length of the molecule, causing transitions in the region of the initial input DNA that surfaces on the outside of the SAC molecule. The increase in size of the initial DNA construct that gives rise to an artificial chromosome that is considerably larger than input is a general rule for the generation of MACs by the bottom up approach [4–6]. The incomplete structural information presented here is consistent with the notion that the size increase is produced by a number of multimerization events of the original construct. One can speculate that MACs produced by the bottom up approach can either replicate efficiently or segregate efficiently only if their size is in the 5 Mb range or larger. Since it has been shown that centromeric DNA tracts with a size of 100 kb can provide efficient centromeric function in linear minichromosomes [5], it is likely that large MACs are selected for on the basis of increased replication efficiency. In fact, plasmids into which single mammalian replication origins are inserted unaccompanied by additional sequences such as MARs or specialized viral sequences do not replicate in mammalian cells [21]. In order to utilize the SAC constructed in this study for animal transgenesis, genes of interest will be added to the microchromosome. This can be done in the initial phases of the construction, before introduction into the mammalian cell, thus obtaining a gene copy number of about 15. A gene copy number of one can be obtained by inserting the gene into the SAC in its final form; this can be accomplished most easily if the SAC is transferred into a host endowed with an efficient homologous recombination system, such as the DT40 chicken strain, or conceivably after transfer and propagation in S. cerevisiae. Artificial chromosomes are currently being used in the mouse and other animals for the study of centromere function; human artificial chromosomes are being considered as possible vectors for gene therapy; pig artificial chromosomes, such as the SAC constructed and characterized in this
1149
study, will very likely be useful episomal vectors for producing transgenic pigs for xenotransplantation and other purposes important for medicine and biotechnology.
Acknowledgements We thank Patrick Chardon for furnishing YAC 225D04B. Thanks are also due to Fiorentina Ascenzioni for useful discussions, for helpful collaboration in the theory and practice of YAC technology and for preparing and providing the humanized yeast strain. We are grateful to Barbara Barboni, Mauro Mattioli, Manfred Brich and Carl Zeiss Jena GmbH for assistance with confocal microscopy. This work was supported by a grant from the MURST (Modalità alternative di transgenesi animale: ricerche scientifiche sulla costruzione di mini- e microcromosomi artificiali).
References [1]
M.A. Barnett, V.J. Buckle, E.P. Evans, A.C.G. Porter, D. Rout, A.G. Smith, W.R.A. Brown, Telomere directed fragmentation of mammalian chromosomes, Nucl. Acid. Res. 21 (1993) 27–36. [2] W. Mills, R. Critcher, C. Lee, C.J. Farr, Generation of an z2.4 Mb human X centromere-based minichromosome by targeted telomereassociated chromosome fragmentation in DT40, Hum. Mol. Gen. 8 (1999) 751–761. [3] C. Guiducci, F. Ascenzioni, C. Auriche, E. Piccolella, A.M. Guerrini, P. Donini, Use of a human minichromosome as a cloning and expression vector foe mammalian cells, Hum. Mol. Gen. 8 (1999) 1417–1424. [4] J.E. Mejia, A. Alazami, A. Willmott, P. Marschall, E. Levy, W.C. Earnshaw, Z. Larin, Efficiency of de novo centromere formation in human artificial chromosomes, Genomics 79 (2002) 297–304. [5] J.W. Yang, C. Pendon, J. Yang, N. Haywood, A. Chand, W.R.A. Brown, Human mini-chromosomes with minimal centromeres, Hum. Mol. Gen. 9 (2000) 1891–1902. [6] J.J. Harrington, G.V. Bokkelen, R.W. Mays, K. Gustashaw, H.F. Willard, Formation of de novo centromeres and construction of firstgeneration human artificial minichromosomes, Nature Gen. 15 (1997) 345–355. [7] M. Ikeno, B. Grimes, T. Okazaki, M. Nakano, K. Saitoh, H. Hoshino, N.I. McGill, H. Cooke, H. Masumoto, Construction of YAC-based mammalian artificial chromosomes, Nature Biotech. 16 (1998) 431–439. [8] K.A. Henning, E.A. Novotny, S.T. Compton, X.Y. Guan, P.P. Liu, M.A. Ashlock, Human artificial chromosomes generated by modification of a yeast artificial chromosome containing both human alpha satellite and single-copy DNA sequences, Proc. Natl. Acad. Sci. 96 (1999) 592–597. [9] B. Grimes, H. Cooke, Engineering mammalian chromosomes, Hum. Mol. Gen. 7 (1998) 1635–1640. [10] J.E. Mejia, A. Willmott, E. Levy, W.C. Earnshaw, Z. Larin, Functional complementation of a genetic deficiency with human artificial chromosomes, Am. J. Hum. Genet. 69 (2001) 315–326. [11] B. Grimes, D. Schindelhauer, N.I. McGill, A. Ross, T.A. Ebersole, H.J. Cooke, Stable gene expression from a mammalian artificial chromosome, EMBO Rep. 21 (2001) 910–914. [12] T.A. Ebersole, A. Ross, E. Clark, N. McGill, D. Schindelhauer, H. Cooke, B. Grimes, Mammalian artificial chromosome formation from circular alphoid input DNA does not require telomere repeats, Hum. Mol. Gen. 9 (2000) 1623–1631.
1150
P. Poggiali et al. / Biochimie 84 (2002) 1143–1150
[13] C. Rogel-Gaillard, N. Bourgeaux, J.C. Save, C. Renard, P. Coullin, P. Pinton, M. Yerle, M. Vaiman, P. Chardon, Construction of a swine YAC library allowing an efficient recovery of unique and centromeric repeated sequence, Mamm. Genome 8 (1997) 186–192. [14] K.A. Henning, N. Moskowitz, M.A. Ashlock, P. Liu Pu, Humanizing the yeast telomerase template, Proc. Natl. Acad. Sci. 95 (1998) 5667–5671. [15] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: a Laboratory Manual, 2nd ed, Cold Spring Harbor Laboratory Press, New York, 1989. [16] G.A. Silverman,YAC protocols, in: D. Marlie (Ed.), Humana, Totowa, NY, 1996, pp. 199–216. [17] V. Pachnis, L. Perny, R. Rothstein, F. Costantini, Transfer of a yeast artificial chromosome carrying human DNA from Saccharomyces cerevisiae into mammalian cells, Proc. Natl. Acad. Sci. 87 (1990) 5109–5113.
[18] T. Haaf, D.C. Ward, Structural analysis of a-satellite DNA and centromere proteins using extended chromatin and chromosomes, Hum. Mol. Gen. 3 (1994) 697–709. [19] A. Schmitz, B. Chaput, P. Fouchet, M.N. Guilly, G. Frelat, M. Vaiman, Swine chromosomal DNA quantification by bivariate flow karyotyping and karyotype interpretation, Cytometry 13 (7) (1992) 703–710. [20] C. Auriche, P. Donini, F. Ascenzioni, Molecular and cytological analysis of a 5.5 Mb minichromosome, EMBO Reports 2 (2) (2001) 102–107. [21] A. Baiker, C. Maercker, C. Piechaczek, S.B. Schmidt, J. Bode, C. Benham, H.J. Lipps, Mitotic stability of an episomal vector containing a human scaffold/matrix attached region is provided by association with nuclear matrix, Nature Cell. Biol. 2 (3) (2000) 182–184.