Transfer of a Human Chromosomal Vector from a Hamster CellLine to a Mouse Embryonic Stem Cell LineMARIANNA PAULIS,a,b MIRELLA BENSI,a DONATA ORIOLI,c CHIARA MONDELLO,c GIULIANO MAZZINI,cMAURIZIO D’INCALCI,d CRISTIANO FALCIONI,d ENRICO RADAELLI,e EUGENIO ERBA,c ELENA RAIMONDI,aLUIGI DE CARLIa

aDipartimento di Genetica e Microbiologia “Adriano Buzzati Traverso” Universita` degli Studi di Pavia, Pavia, Italy;bIstituto di Tecnologie Biomediche del Consiglio Nazionale delle Ricerche di Segrate, Milan, Italy; cIstituto diGenetica Molecolare del Consiglio Nazionale delle Ricerche di Pavia, Pavia, Italy; d Department of OncologyIstituto di Ricerche Farmacologiche “Mario Negri” di Milano, Milan, Italy; e Dipartimento di Patologia Animale,Igiene e Sanita` Pubblica Veterinaria Universita` degli Studi di Milano, Milan, Italy

Key Words. Human chromosomal vector • Mouse embryonic stem cell • Cell fusionTranschromosomic mouse embryonic stem cell line • Ploidy variation • Pluripotent gene markers

ABSTRACTTwo transchromosomic mouse embryonic stem (ES) sub-lines (ESMClox1.5 and ESMClox2.1) containing a humanminichromosome (MC) were established from a sample ofhybrid colonies isolated in fusion experiments between anormal diploid mouse ES line and a Chinese hamster ovaryline carrying the MC. DNA cytometric and chromosomeanalyses of ESMClox1.5 and ESMClox2.1 indicated a mousechromosome complement with a heteroploid constitution ina subtetraploid range; the karyotypes showed various de-grees of polysomy for different chromosomes. A single copyof the MC was found in the majority of cells in all theisolated hybrid colonies and was stably maintained in theestablished sublines for more than 100 cell generations ei-ther with or without the selective agent. No significant dif-

ferences from the ES parental cells were observed in growthcharacteristics of the transchromosomic ES sublines.ESMClox1.5 cells were unable to grow in soft agar; whencultured in hanging drops, they formed embryoid bodies,and when inoculated in nude mice, they produced terato-mas. They were able to express the early development mark-ers Oct4 and Nanog, as demonstrated by reverse transcrip-tion-polymerase chain reaction assay. All these features arein common with the ES parental line. Further research usingthe transchromosomic ES sublines described here may allowgene expression studies on transferred human minichromo-somes and could shed light on the relationships amongploidy, pluripotency, cell transformation, and tumorigene-sis. STEM CELLS 2007;25:2543–2550

Disclosure of potential conflicts of interest is found at the end of this article.

INTRODUCTION

The transgenic approach for gene localization, gene expressionstudies, and gene therapy is based on transfer of genetic materialranging in size from small DNA stretches to whole chromo-somes. Fragmented or intact human chromosomes introducedinto embryonic stem cells and segregating as autonomous ele-ments in cell cultures or in living animals have been investigatedby different authors [1–7]. The method most commonly used,although limited by its low efficiency, is the microcell-mediatedchromosome transfer (MMCT), which allows the isolation ofembryonic stem cell lines carrying the extra chromosome; thefinal goal is the generation of transchromosomic mice by inject-ing the chromosomally modified cells into recipient blastocysts.Using this approach, O’Doherty et al. [8] produced a chimericmouse in which the majority of cells contained the humanchromosome 21. These studies revealed the utility of transchro-mosomic mice for analyzing gene dosage effects on a variety ofphenotypic traits and as models of disorders associated withchromosome number variation, such as Down syndrome.

Of particular interest for investigations on chromosometransfer in embryonic stem (ES) cells are human artificial chro-mosomes (HACs), which not only are useful for analyzing theeffects of gene imbalance on embryo development and expres-sion of tissue-specific functions but also can be exploited aspotential vectors for gene therapy. Two possible sources ofHACs are (a) the new formation of a functionally active chro-mosome from isolated minimum essential DNA sequences and(b) reduction of a pre-existing chromosome. Following this lastapproach, Shen et al. [2] succeeded in introducing a 4-Mb HACderived from the long arm of the human Y chromosome intomouse embryonic stem cells, to generate transchromosomicmice. Ren et al. [9] gave the first demonstration of lineage-specific expression induction of transgenes in human mesenchy-mal stem cells by a HAC vector from human chromosome 21.

Here, we describe the isolation of transchromosomic mouseES cell lines containing a HAC generated from a radiation-reduced chromosome 9-derived minichromosome (MC), whichhas the properties of a vector thanks to the presence of selectablemarkers and of the loxP sequence for site-specific recombina-tion [10 –12]. These cell lines were established from coloniesisolated in fusion experiments between a human/hamster mono-

Correspondence: Luigi De Carli, Ph.D., Dipartimento di Genetica e Microbiologia, Via Ferrata 1, 27100 Pavia, Italy. Telephone:39-0382-985554; Fax: 39-0382-528496; e-mail: decarli@ipvgen.unipv.it Received January 22, 2007; accepted for publication June 26,2007; first published online inSTEM CELLSEXPRESS July 5, 2007. ©AlphaMed Press 1066-5099/2007/$30.00/0 doi: 10.1634/stemcells.2007-0052

EMBRYONIC STEM CELLS

STEM CELLS2007;25:2543–2550 www.StemCells.com

chromosomic hybrid cell line (MClox) and a mouse ES cell line(E14). Our aim was to produce a mouse ES cell line capable ofexpressing human genes carried by a supernumerary, freelysegregating chromosome, whose mitotic stability and distribu-tion in the successive cell divisions could be followed over longculture periods.

To move the chromosome from the donor to the host cells,we used a simplified procedure of cell fusion-mediated chro-mosome transfer without separation of microcells. Using a com-bined pretreatment with colcemid and cytochalasin B of theMClox parental cell line, we succeeded in isolating in selectivemedium ES hybrid colonies retaining a single copy of the MC inthe absence of hamster chromosomes. The heteroploid consti-tution found in all the isolated hybrid colonies and in theestablished cell lines would allow investigation of the effects ofchromosome number variation on control of pluripotency, celltransformation, and tumorigenesis.

MATERIALS AND METHODS

Cell CultureMClox is a monochromosomic human/hamster somatic hybrid orig-inally obtained by cell fusion between hypoxanthine-quanine-phos-phoribosyl-transferase–negative (HPRT⫺) Chinese hamster ovary(CHO) and a lymphoblastoid line from a patient carrying an MCidentified as a chromosome 9 derivative in a complex mosaickaryotype [10]. The centromeric region of MC contains the neomy-cin gene (neo) and five copies of the pG12 plasmid carrying a loxPsite plus the puromycin resistance gene (puro) [11]. The cell linewas maintained by standard culture procedures in RPMI 1640medium (Euroclone, Pero, Italy, http://www.euroclone.net), supple-mented with 10% fetal calf serum (Euroclone), and incubated at37°C with 5% CO 2.

E14 is a mouse embryonic stem cell line from 129/Ola mice[13], kindly supplied by R. Klein (Max Plank Institute, Martin-sried, Germany). The cells were grown on 100-mm Petri dishescoated with 0.1% gelatin (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) in high-glucose Dulbecco’s modified Eagle’smedium (DMEM) (Euroclone) containing 15% fetal bovine serum(Invitrogen, Carlsbad, CA, http://www.invitrogen.com) and 1,000U/ml leukemia inhibitory factor (ESGRO; Chemicon International,Temecula, CA, http://www.chemicon.com); the medium was con-ditioned with mouse embryonic fibroblasts (MEFs) grown for 2days in log phase. MEFs were obtained by mincing and dissociating13.5 days post coitum CD-1 embryos (Charles River Laboratories,Wilmington, MA, http://www.criver.com) by collagenase (0.25% inphosphate-buffered saline [PBS] plus 20% fetal calf serum); pri-mary cultures were maintained in DMEM (Euroclone) supple-mented with 10% fetal calf serum (Euroclone) and incubated at37°C with 7% CO 2.

Cell Fusion and Colony IsolationMClox cells arrested in mitosis and E14 unsynchronized cells werefused. MClox cells (4 ⫻ 106) were inoculated in four T25 flasks(tissue culture flasks, 25 cm); colcemid was added to the cultures ata final concentration of 0.2 g/ml when the cells reached approx-imately 80% confluence. After 24 hours of incubation, mitoses weremechanically removed from monolayers, incubated in growth me-dium containing 20 g/ml cytochalasin B at 37°C for 4 hours withconstant stirring, and then mixed with an equal amount of E14monodispersed cells. The mixture with a total of 107 cells wasresuspended in 10 ml of 2.5% polyethylene glycol (PEG) (mol. wt.1500) in DMEM and kept in this solution for 20 minutes at 4°C.After centrifugation at 160g, 1 ml of a prewarmed solution of 50%PEG was poured onto the cell pellet over 2 minutes; 10 ml of freshDMEM was then gradually added to the cell suspension over 10minutes. After washing with PBS, the cells were resuspended inMEF-conditioned ES medium, distributed in two 60-mm Petridishes, and incubated for a recovery period of 48 hours at 37°C. The

cells were finally plated, 10 5 cells per 100-mm Petri dish in selec -tive medium containing HAT (100 M hypoxanthine, 0.4 g/mlaminopterin, and 16 M thymidine) and puromycin at differentconcentrations (0.3–10 g/ml).

Chromosome AnalysisChromosome analysis was done either on slide preparations of cellsuspensions or on cells grown on coverslips; for the first method,monolayer cell cultures were treated with colcemid at a final con-centration of 0.1 g/ml for 2 hours at 37°C, and mitoses weremechanically removed. After hypotonic treatment with 0.075 MKCl and fixation in methanol:acetic acid (3:1 vol/vol), the cellsuspension was dropped onto a slide and air-dried. Cells grownon coverslips were treated the same way except that the colcemidconcentration was 0.3 g/ml. Chromosome counts and karyotypeanalyses were done on metaphases stained with a standard Qbanding.

Fluorescence In Situ HybridizationCHO genomic DNA and pMR9A plasmid, identifying a chromo-some 9-specific ␣-satellite DNA subfamily [14], were used as DNAprobes. The probes were labeled using a nick-translation reagentsystem (Invitrogen) and Bio-16-dUTP (Roche Diagnostics S.p.A,Milano, Italy, http://www.roche-applied-science.com) according tothe manufacturers’ protocols. The labeled probes were resuspendedin hybridization buffer (50% formamide, 10% dextran sulfate, 1⫻Denhart’s solution, 0.1% SDS, 40 mM Na 2HPO4, pH 6.8, 2⫻standard saline citrate [SSC]) containing a 10 ⫻ excess of salmonsperm DNA and denatured at 80°C for 10 minutes. In situ hybrid-ization was performed essentially as previously described [10]. Inbrief, slides were treated with RNase (type III) at 37°C for 1 hourand dehydrated through the ethanol series before denaturation in70% formamide/2 ⫻ SSC. Hybridization was done overnight at42°C. Stringent washings were done in 50% formamide/2⫻ SSC at42°C.

For single signal detection, the slides were incubated withfluorescein isothiocyanate (FITC)-conjugated cell sorting grade avi-din D (avidin DCS) (Molecular Probes Inc., Eugene, OR, http://probes.invitrogen.com; Invitrogen), then with biotin-conjugated an-ti-avidin D antibody (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com), and finally with FITC-conjugated avidinDCS. Avidin and all the antibodies were used at a final concentra-tion of 5 g/ml.

For double signal detection, the slides were incubated firstwith FITC-conjugated avidin DCS and tetramethylrhodamine Bisothiocyanate (TRITC)-conjugated sheep anti-digoxigenin anti-body (Roche Diagnostics), then with biotin-conjugated anti-avidinD antibody and TRITC-conjugated rabbit anti-sheep antibody(Chemicon International), and finally with FITC-conjugated avidin andTRITC-conjugated anti-rabbit antisera (Calbiochem,San Diego, CA,http://www.emdbiosciences.com).Avidin and all the antibodies wereused at a final concentration of 5g/ml.

All the slides were counterstained with 4 ⬘,6-diamidino-2-phenylindole (DAPI) (0.01 g/ml) (Sigma-Aldrich) and mountedin Tris-HCl (pH 7.5) 90% glycerol containing 2% 1,4-diazabicyclo(2.2.2)octane antifade (Sigma-Aldrich). Slides werescored under a Zeiss Axioplan fluorescent photomicroscope(Carl Zeiss, Jena, Germany, http://www.zeiss.com) equippedwith a cooled CCD camera (Photometrics, Tucson, AZ, http://www.photomet.com). Images were captured with IPlab spectrumP software (BD Biosciences Bioimaging, Rockville, MD, http://www.scanalytics.com).

Indirect ImmunofluorescenceCells were fixed in 4% paraformaldehyde in PBS for 15 minutes at4°C; permeabilized in 0.05% Tween 20, 0.5% bovine serum albu-min in PBS for 10 minutes at room temperature; and then incubatedfor 1 hour at 37°C with the polyclonal anti-Oct4 antibody (SantaCruz) diluted 1:500 in PBS. The binding of the antibody wasrevealed with a FITC-conjugated antibody (Sigma-Aldrich), diluted1:500. Nuclei were counterstained with 200 g/ml DAPI. Slideswere observed using an Olympus IX71 optical microscope equipped

2544 A Human Minichromosome in Mouse ES Cells

with a ⫻60 objective (Olympus, Tokyo, http://www.olympus-global.com). Images were taken with a Cool SNAPES digital camera(Photometrics) using the MetaMorph software.

Flow Cytometric AnalysisPellets from 5 ⫻ 105 to 1 ⫻ 106 cells, carefully dispersed in 100 lof cold PBS, were fixed in 5 ml of 70% cold ( ⫺20°C) ethanol andstored at 4°C until analysis. After washing again in PBS cells werestained with 1.5 ml of propidium iodide (PI) (50 g/ml) containingRNase (100 U/ml) and Nonidet P40 (0.05%). Samples were storedovernight at 4°C in the dark and analyzed by flow cytometry.Measurements were done with a Partec PAS II flow cytometer(Munster, Germany, http://www.partec.com) equipped with a dualexcitation system (argon ion laser and HBO 100-W arc lamp). The488 nm blue line of the laser was used to excite PI intercalated intoDNA complex. A preliminary instrument alignment and control wasalways set up (with rat thymocytes stained with PI) to ensure thebest instrumental analytical performance. Immediately before mea-surement, each sample was filtered through Filcons 100 (ConsulTS,Turin, Italy, http://www.consul-ts.com) to remove cell clusters. Fora sample measurement, a minimum of 20,000 events were acquired.The red fluorescence emission band over 610 nm (FL3) was col-lected, converted, and stored as DNA distribution values by adedicated computer integrated into the instrument.

Embryoid BodiesA total of 400 ES E14 or ESMClox1.5 cells were inoculated inhanging drops (30 l) on covers of 120 mm hydrophobic Petridishes in differentiation medium (MEF-conditioned ES mediumwithout leukemia inhibitory factor [LIF]). After 3 days of incuba-tion, the drops were transferred to the bottom of the plates with theaddition of 1 ml of medium and further incubated for 5 days. Thenascent EBs were plated separately onto gelatin-coated 24-micro-well plates.

Colony Formation in Soft AgarA 3-ml layer of 2% agar (wt/vol) in MEF-conditioned medium waspoured in 60-mm Petri dishes. ES E14, ESMClox1.5, or HeLa cells,used as positive control in the assay, were resuspended in 0.33%agar (wt/vol) in MEF-conditioned medium at a density of 10 4 cellsper 3 ml. Cell suspension (3 ml) was poured on the top of the baselayer, allowed to solidify, and incubated at 37°C with 5% CO 2.Cells were fed twice a week with 0.3 ml of fetal calf serum(Euroclone) and observed weekly under the microscope for colonyformation.

Reverse Transcription-Polymerase Chain ReactionFirst-strand cDNAs were synthesized directly from ES E14, MEF,and ESMClox1.5 cells and from ES E14 EBs and ESMClox1.5 EBsusing the Superscript CellsDirect cDNA Synthesis Kit (Invitrogen)according to the manufacturer’s protocol. Polymerase chain reac-tions (PCRs) were performed using specific primer pairs (supple-mental online Table 1). PCRs were performed under the followingconditions: denaturing for 30s at 94°C, annealing temperature for30s, and extension for 30s at 72°C, repeated 30 times. The PCRproducts were visualized by 2% agarose gel electrophoresis stainedwith ethidium bromide.

In Vivo TumorigenicityFour- to 6-week-old female Swiss Ncr nu/nu mice (Charles RiverLaboratories) weighing 20 –25 g were used. Mice were maintainedunder specific pathogen-free conditions, with food and water adlibitum. Samples of 107 cells harvested from cultures of tested celllines were inoculated subcutaneously into the flank of the recipientmouse. When the mass grown from transplanted cells was palpable(50 –100 mg), it was measured weekly with calipers. Tumor volumewas calculated as L2 ⫺ D/2, where L is the length and D the widthof the mass. The volume on day n was expressed as relative tumorvolume (RTV) according to the formula RTV ⫽ TVn/TVo, whereTVn is the tumor volume on day n and TVo is the volume on day 0.

Procedures involving animals and their care were conducted inconformity with the institutional guidelines that are in compliancewith national (D.L.n.116, G.U., suppl. 40, 18 febbraio 1992, Cir-colare No. Eight, G.U., 14 luglio 1994) and international (EECCouncil Directive 86/609, OJ L 358, 1, Dec. 12, 1987; Guide for theCare and Use of Laboratory Animals, U.S. National ResearchCouncil, 1996) laws and policies.

RESULTS

Characterization of the Parental Cell LinesThe mouse ES cell line E14 used as the MC recipient has beenadapted in our laboratory to grow in the absence of a feederlayer of MEFs: it has acquired independence from the feederafter repeated passages in MEF-conditioned medium. Underthese conditions, it maintains a normal diploid karyotype; thegeneration time is 24 hours, and the plating efficiency is close to10%.

The hamster cell line containing the MC (MClox) has anear-diploid/pseudodiploid chromosome constitution, with amodal chromosome number of 21. The MC is present in ap-proximately 60% of the cells. The generation time is 12 hours,and the plating efficiency may reach 80%.

Fusion Experiments and Isolation of ESMCloxHybrid ColoniesFusion experiments followed the procedure described in Mate-rials and Methods. To facilitate the loss of hamster chromo-somes in hybrid cells, the parental MClox cell line was pre-treated with colcemid for 24 hours, and the collected mitoseswere exposed to cytochalasin B for 4 hours.

Three weeks after fusion, a total of 31 colonies, of likelyclonal origin, grown in HAT medium with puromycin, wereisolated and expanded in culture for determination of the DNAcontent and for cytological preparations. DNA flow-cytometricanalysis was done on samples of 5⫻ 105 cells from 23 colonies.Table 1 reports the data on the DNA content expressed as themean value of the G0/G1 peak, with the corresponding coeffi-cient of variation (CV), for 23 hybrid colonies, for ES parentalcells, and for an MEF control. Whereas in the parental and MEFcells, which have a normal diploid constitution, the position ofthe G0/G1 peak was around 50, in the majority of the hybridcolonies (13 of 23), the peak position was between 86 and 100,corresponding to a “near-tetraploid” DNA content. In the re-maining colonies, with one exception (ESMClox1.5), in addi-tion to the main peak between 90 and 97, there was a secondpeak between 59 and 71, indicating a “near-triploid” DNAvalue. A minor peak, corresponding to a diploid DNA, valuewas observed in two colonies (ESMClox1.20 and ESMClox1.5),which may suggest that the cell populations contain rare normaldiploid cells carrying the MC. The possibility of sorting outthese cells to establish a euploid ESMClox line has been takeninto consideration.

To detect the MC and the residual CHO chromosomalmaterial, metaphases from 30 colonies were hybridized in situwith chromosome 9-specific ␣-satellite DNA (pMR9A) andCHO genomic DNA probes, respectively. Twenty metaphasespreads for each colony were analyzed. The results are summa-rized in Table 1. All except two of the colonies contained theMC free or translocated on hamster residual chromosomes, theproportion of MC-positive mitoses varying from 10% to 95%.Interestingly, in 14 of 30 colonies (47%), the MC was the onlychromosome transferred from the donor cell line.

One of the colonies with the highest frequency of MC-positive cells and with the typical ES morphology was expanded

2545Paulis, Bensi, Orioli et al.

www.StemCells.com

to establish a subline designated as ESMClox1.5. Another sub-line (ESMClox2.1) with the same characteristics was estab-lished from one of eight colonies isolated in an independentfusion experiment, following the same protocol as before. Allthe colonies showed a heteroploid mouse complement withmodal chromosome numbers comparable to those in the firstfusion experiment, with the majority of cells containing the MC(data not shown).

DNA and Chromosome Analyses of ESMClox1.5and ESMClox2.1 SublinesThe DNA flow-cytometric profiles, together with the distribu-tions of chromosome numbers of the two sublines and of the EScontrol line are, shown in Figure 1. The control showed a DNAdistribution typical of diploid cells along the cell cycle phases.The main peak, representing G0/G1 cells, was around channel50, whereas the corresponding G2 and mitoses were evidentin channels 95–105. The DNA distribution of the sublinesESMClox1.5 and ESMClox2.1 clearly showed a shift towarddouble or nearly double the control values. The first G0/G1peak was in fact located around channels 90 –100, and the

corresponding G2 and mitoses were between channels 180 and200. A small subfraction of “residual” diploid population wasnoted in the first part of the DNA histogram of ESMClox1.5(Fig. 1), confirming the result of the DNA analysis on theoriginal hybrid colony (Table 1).

The chromosome complements, as evidenced by the chro-mosome counts, varied in a subtetraploid range from 40 to 70,with modal values respectively of 65 and 60 for ESMClox1.5and ESMClox2.1. The cytological and cytometric data do notappear to be fully concordant, because the DNA content wasnear that expected for tetraploidy, whereas chromosome countswere definitely lower. This might be explained by a selection ofmitoses in cultures for chromosome preparations: cells with atetraploid or near-tetraploid complement would have a slowergrowth rate and a lower probability of entering division andprogressing to mitosis. Detailed morphological analysis of chro-mosomes on samples of 10 cells taken from the modal classrevealed various degrees of polysomy for different chromo-somes. The presence of the MC was confirmed in ESMClox1.5and ESMClox2.1 by fluorescence in situ hybridization (FISH)analysis using the pMR9A probe. A single copy of the MC wasfound in more than 95% of the cells of the two lines. Metaphasespreads from ESMClox1.5 and ESMClox2.1 with the labeledMC, together with a representative banded karyotype of theESMClox1.5, are shown in Figure 2.

Growth Characteristics and Pluripotent Phenotypeof ESMClox1.5 and ESMClox2.1 SublinesThere was a slight increase in the growth rate and platingefficiency of the ESMClox1.5 and the ESMClox2.1 sublinescompared with the E14 line. The cells and the colonies alsoshowed the typical ES morphology (Fig. 3A, 3C). When incu-bated in hanging drops in the absence of LIF and then trans-ferred to hydrophobic Petri dishes, they formed tight clusterscomparable in appearance to the EBs inducible in ES cells underthe same conditions, indicating their ability to differentiate invitro (Fig. 3D). Evidence of cell differentiation in the EBs hasbeen provided by reverse transcription (RT)-PCR assay on thefollowing germ layer-specific genes: Sox1 (SRY box-containinggene 1) for ectoderm, BMP2 (bone morphogenetic protein 2) formesoderm, and AFP (␣-fetoprotein) for endoderm. The capacityof the ESMClox1.5 cells to grow in soft agar was also tested todetermine their anchorage dependence (data not shown). Thisassay was negative, as in ES control cells.

To verify the maintenance of the pluripotent phenotype ofhybrid cells, the expression of the early embryonic developmen-tal markers Oct4 and Nanog was examined by RT-PCR. TotalcDNA was amplified from ESMClox1.5 and ESMClox2.1 cells,from parental ES cells as a positive control, and from MEF cellsas a negative control. Fragments of the expected sizes wereobtained in samples from the two hybrid lines and from theparental ES. Primers for ␤-actin were included in the experi-ments as a internal control. The results are shown in Figure 4.To study the expression of the pluripotency marker Oct4 at thesingle cell level in the ESMClox1.5 subline, we performedindirect immunofluorescence experiments using an antibodyagainst Oct4. We found that the vast majority of the cells in thepopulation were positive to the staining with the antibody.However, the intensity of the signal was variable, suggestingthat the level of expression could vary among cells (Fig. 4B). Itis worth noting that such a variability was even observed amongcells of the control ES population. The ESMClox1.5 sublineability to form teratomas, which is common to all mouse ES celllines, was also tested. The ESMClox1.5 cells were transplantedinto four female nude mice (10 7 cells per mouse); a parallelcontrol was set up with E14 cells. Three weeks after cell

Table 1. Results of flow cytometric and FISH analyses onESMClox hybrid colonies

Samples

Flow cytometry FISH

G0/G1 CV MCa CHO

MClox NA NA 60% ⫹Ctr(MEF) 51 9 NA NACtr(ES-E14) 53 5 NA NAESMClox1.1 99 7 NA NAESMClox1.2 100 6 30% ⫹ESMClox1.3 98 8 10% ⫺ESMClox1.4 98 7 40% ⫺ESMClox1.5 50/97 6 95% ⫺ESMClox1.6 96 6 85% ⫺ESMClox1.7 90 6 10% ⫺ESMClox1.8 91 6 40% ⫹ESMClox1.9 86 5 20% ⫺ESMClox1.10 86 5 40% ⫺ESMClox1.11 89 4 95% tCHO ⫹ESMClox1.12 91 5 30% ⫹ESMClox1.13 93 8 90% tCHO ⫹ESMClox1.14 71/93 ND 90% tCHO ⫹ESMClox1.15 93 6 50% ⫺ESMClox1.16 NA NA 50% ⫹ESMClox1.17 64/93 ND 70% ⫹ESMClox1.18 NA NA 95% ⫺ESMClox1.20 50/68/95 10 60% ⫹ESMClox1.21 64/94 8 10% ⫺ESMClox1.22 NA NA 90% ⫹ESMClox1.23 65/93 8 90% tCHO ⫹ESMClox1.24 NA NA 40% ⫺ESMClox1.25 63/95 9 40% ⫹ 40% tCHO ⫹ESMClox1.26 59/90 10 70% ⫹ 90% tCHO ⫹ESMClox1.27 NA NA 70% ⫺ESMClox1.28 NA NA 85% ⫹ESMClox1.29 68/95 8 90% ⫹ESMClox1.30 NA NA 80% ⫹ESMClox1.31 NA NA 50% ⫺ESMClox1.32 60/94 8 10% ⫺aPercentage of mitoses containing the MC free or translocated to ahamster chromosome.Abbreviations: CHO, Chinese hamster ovary; Ctr, control; CV,coefficient of variation; ES, embryonic stem; FISH, fluorescencein situ hybridization; MC, minichromosome; MEF, mouseembryonic fibroblast; NA, not analyzed; ND, not determined;tCHO, to a hamster chromosome.

2546 A Human Minichromosome in Mouse ES Cells

transplantation, solid masses were excised and histological stud-ies were conducted by conventional methods (Fig. 5). The tumormasses from E14 and ESMClox1.5 clones showed the typicalfeatures of teratomas. Masses were composed of abundant mes-enchyma surrounding areas of immature and disorganized os-teocartilaginous tissue, with both smooth and striated muscula-ture. Multiple cystic dilations lined by epidermis-like stratifiedsquamous keratinized epithelium and ectatic disarranged tubulo-papillary units lined by pseudostratified ciliated respiratory ep-ithelium were scattered in the masses.

To establish whether genes located on the transferred MCare active during undifferentiated growth and/or during differ-entiation in vitro of the ESMClox1.5 subline, we analyzed theexpression of the following genes: the housekeeping genesCLTA and TLN1 and the resistance marker for puromycin

(puro). The RT-PCR assays were negative for the two house-keeping genes but positive for the puro resistance gene, in bothcases (data not shown).

Propagation of the ESMClox1.5 SublineAfter the first passages in culture, the ESMClox1.5 subline wascryoconserved in liquid nitrogen. Cell samples were thawed atintervals of 3– 4 months. The overall period of culture was 6months, with 50 – 60 passages resulting in 100 –120 cell popu-lation doublings. The growth characteristics remained constant,and no changes were seen in the morphology of cells andcolonies. The distribution of chromosome numbers was thesame as in the subline at the first passages, with a modal classof 65. The MC was found in up to 95% of cells at passage 40,either with or without the selective agent.

A derivative of ESMClox1.5 was established in vitro fromdissociated fragments of teratomas produced in mice. The chro-mosome number distribution and the growth characteristics ofthis newly isolated subline from teratoma tissue were the same.The MC was retained in 80% of the cells, as indicated by FISHanalysis with the pMR9A probe (Fig. 6).

DISCUSSION

We have demonstrated that a human minichromosome can betransferred from a hamster cell line to a mouse cell line byfusing recipient cells with donor parental cells pretreated withcolcemid and cytocalasin B; the resulting product is a secondarymonochromosomic human/mouse hybrid. The donor cell linewas a CHO with a near-diploid/pseudodiploid constitution, andthe recipient was a mouse ES cell line with a diploid chromo-some complement. In our experiments, the efficiency of chro-mosome transfer, measured as the ratio between the number ofMC-positive hybrid colonies and the number of donor cells inthe fusion mixture, was on the order of 10⫺4. To our knowledge,this value is 5–10 times higher than that observed with theMMCT method [15].

In view of its pluripotency, the recipient cell line mayconstitute a suitable host for the expression of exogenous genestransferred into the human minichromosome, used as a tool forgenetic analysis or for physiological studies.

Since the fusion method that we used did not involve theseparation of microcells from the donor cells arrested in mitosis,these donor cells should have contained an intact chromosomecomplement. We were then faced with the problem of direc-tional chromosome segregation in mouse/hamster hybrids. In

Figure 1. DNA cytometric profiles and chromosome number distributions (insets) of E14 line (A), ESMClox1.5 (B), and ESMClox2.1 (C) sublines.Abbreviation: ES, embryonic stem.

Figure 2. Cytogenetic analysis. Fluorescence in situ hybridization onmetaphase spreads from ESMClox1.5 (A) and ESMClox2.1 (B) cells wasperformed with pMR9A probe labeled with digoxigenin and detected withtetramethylrhodamine B isothiocyanate (TRITC)-conjugated sheep anti-digoxigenin antibody followed by TRITC-conjugated rabbit anti-sheepantibody (red signal); the chromosomes were counterstained with 4⬘,6-diamidino-2-phenylindole (blue). (C): Representative Q-banded karyo-type of ESMClox1.5 cells.

2547Paulis, Bensi, Orioli et al.

www.StemCells.com

this regard, the literature is conflicting, although it indicates atendency to loss of mouse chromosomes [16 –18]. In our exper-iments, the construction of a hybrid containing only theminichromosome would have required the opposite dynamics,leading to the loss of hamster chromosomes. This reverse effectis indeed what we obtained, presumably because of the com-bined treatment of the MClox cell line with a spindle formationinhibitor and a microfilament disrupting agent coupled to apotent selective system, which proved highly effective in induc-ing the loss of hamster chromosomes. The process took place inone or a few steps, presumably before the formation of thehybrid colonies, between three and four cell divisions afterfusion. This suggests a mechanism of massive loss of hamsterchromosomes, possibly through unbalanced distribution ofchromosomes in multipolar mitoses, with the persistence of theMC carrying the selective marker.

The human MC was identified in the majority of cells of allthe isolated hybrid colonies. In the sublines maintained in cul-ture, it was detected in more than 90% of the cells in the periodiccontrols performed every 10 passages, in either the presence or

absence of the selective agent. This points to a high mitoticstability of the MC, probably thanks to a normal centromericfunction and regular interactions with the spindle apparatus. Nostructural alteration could be detected by cytogenetic analysis.These data agree with the report by Kazuki et al. [19] on ahuman chromosome 21 fragment transferred in mice. On theother hand, Shen et al. [2] found that a minichromosome derivedfrom the human chromosome Y was unstable in mouse ES cellsbut could be stabilized by exchanges of mouse centromericsequences. These contrasting findings might be explained bystructural differences between individual chromosomes and/orbetween methods used to incorporate chromosomes into the hostcell.

As far as the origin of the subtetraploid chromosome set inthe hybrids containing the MC is concerned, it may be supposedthat the initial hybrid cell was derived from a polykaryocyte

Figure 3. Phase contrast photographs ofcultured cells and expression of differentia-tion markers. (A): Colonies of E14 cells.(B): Monolayer of MClox cells. (C): Colo-nies of ESMClox1.5 cells. (D): ESMClox1.5embryoid body (EB). Scale bar ⫽ 10 m(A–C) and 300 m (D). (E): Reverse tran-scription-polymerase chain reaction analysison germ layer markers. Lane 1, E14 cells;lane 2, E14 EBs; lane 3, ESMClox1.5 EBs.

Figure 4. Analysis of pluripotent markers. (A): Oct4 and Nanogexpression analysis by RT-PCR; ␤-actin, positive control. (B): Immu-nofluorescence analysis. Cells were stained with an anti-Oct4 antibody(yellow signal; 1, 3, and 5) and counterstained with 4 ⬘,6-diamidino-2-phenylindole (2, 4, and 6). MEF cells were negative to the antibody (1);E14 (2) and ESMClox1.5 (3) cells were positive. Abbreviations: ES,embryonic stem; MEF, mouse embryonic fibroblast.

Figure 5. Histology of teratomas developed in nude mice after injec-tion of ESMClox1.5 cells. The mesenchymal component consists ofabundant fibro-adipose tissue embedding portions of well-differentiatedcartilaginous tissue (ⴱ) peripherally associated with immature haphaz-ardly arranged striated muscular fibers (arrowhead). H&E staining;magnification, ⫻100.

Figure 6. Fluorescence in situ hybridization (FISH) analysis ofmetaphase spreads prepared from cells of the ESMClox1.5 sublinederived from teratoma fragments. FISH was performed as indicatedin Figure 2.

2548 A Human Minichromosome in Mouse ES Cells

containing two nuclei of the parental ES and one nucleus ofthe CHO line. Alternatively, the initial fusion product mighthave been a heterodikaryon with one hamster nucleus and onemouse nucleus; then, the mouse chromosome set would havebeen doubled by endoreduplication, which is occasionally ob-served in the cell lines used in our experiments. Successivelosses and gains of individual chromosomes, following themassive segregation of hamster chromosomes, would have gen-erated a hybrid with the chromosome distribution observed.Whatever the mechanism of origin of the heteroploidy in the EScells carrying the MC, cell fusion is likely to be the causativeevent. Notably, the same chromosome pattern was observed inthe vast majority of colonies isolated in two independent exper-iments. Reports in the literature on the dynamics of chromosomecomplements in mouse ES cell lines point to a marked stabilityof the karyotype, although karyotypic abnormalities are notinfrequent and prolonged cell culturing may affect the diploidchromosome constitution [20, 21]. In contrast, mouse cell cul-tures from a variety of somatic tissues, after a period of crisis,gradually turn into stable cell lines with a heteroploid chromo-some constitution associated with karyotype instability [22, 23].A similar chromosomal pattern is exhibited by our ESMCloxsublines. As ascertained by the flow cytometric analysis, cellswith a diploid DNA content are present in the hybrid cellpopulation but tend to be eliminated because the hyperdiploidcells are at a selective advantage. The probability of recoveringdiploid cells by sorting and establishing a euploid cell linewould greatly increase if the isolation were made a few gener-ations after fusion, when the cell population displays a widespectrum of chromosome sets. The ploidy variation and theincorporation of the MC in the ESMClox1.5 subline apparentlyhad no consequences on the pluripotent phenotype or on theability to differentiate in vitro as shown by RT-PCR analysesand by immunofluorescence studies on Oct4 marker. This find-ing appears unusual for a cell line that has undergone hetero-ploid transformation, an event generally associated with the lossof many of the differentiated properties that distinguish thetissue from which the cells were isolated [24]. Apparently, EScells do not follow these dynamics. Similar conclusions can bedrawn for the ability to produce teratomas, which was retainedin the ESMClox1.5 subline. An evidence against occurrence ofcell transformation is also provided by the lack of an anchorage-

independent growth. In this connection, it is worth mentioningthat transfection with a dominant-negative allele of the MKK4gene, a putative suppressor gene implicated in several trans-forming functions, makes ES cells, normally unable to grow insoft agar, anchorage-independent and more tumorigenic [25].

There is no doubt that the observed heteroploid constitutionof our transchromosomic mouse embryonic stem cell lineschanges the prospects of in vivo experimentation originallyaimed at generating mice exploitable as models for humanchromosomal pathology; however, the ESMClox hybrid cellshold promise for use in in vivo studies on the effects of chro-mosome number variation on early embryogenesis, cell trans-formation, and tumor development. In this regard, it is worthmentioning that Pralong et al. [26] demonstrated that tetraploidmouse ES cells produced by fusion are capable of integratinginto inner cell mass of diploid preimplantation blastocysts.

Moreover, the established mouse ES cell lines containing ahuman chromosomal vector may provide an in vitro modelsystem suitable for analyzing the expression and regulation ofcell lineage-specific genes. These genes may either be alreadypresent in the MC or be inserted by gene targeting throughhomologous recombination, with a view to possible applicationsin stem cell-mediated gene therapy.

ACKNOWLEDGMENTS

We thank Prof. Silvia Garagna and Dr. Paola Rebuzzini (Uni-versity of Pavia) for providing the anti-Oct4 antibody and forhelp in the immunofluorescence experiments. The work wassupported by grants from Ministero dell’Universita` e dellaRicerca Scientifica e Tecnologica (Cofinanziamento 2004),from the University of Pavia (FAR), and from Eurostells(STELLAR). M.P. is a fellow of the Fondazione A. BuzzatiTraverso.

DISCLOSURE OF POTENTIAL CONFLICTSOF INTEREST

The authors indicate no potential conflicts of interest.

REFERENCES

1 Kuroiwa Y, Tomizuka K, Shinohara T et al. Manipulation of humanminichromosomes to carry greater than megabase-sized chromosomeinserts. Nat Biotechnol 2000;18:1086 –1090.

2 Shen MH, Mee PJ, Nichols J et al. A structurally defined mini-chromo-some vector for the mouse germ line. Curr Biol 2000;10:31–34.

3 Shinohara T, Tomizuka K, Takehara S et al. Stability of transferredhuman chromosome fragments in cultured cells and in mice. Chromo-some Res 2000;8:713–725.

4 Tomizuka K, Yoshida H, Uejima H et al. Functional expression andgermline transmission of a human chromosome fragment in chimaericmice. Nat Genet 1997;16:133–143.

5 Tomizuka K, Shinohara T, Yoshida H et al. Double trans-chromosomicmice: Maintenance of two individual human chromosome fragmentscontaining Ig heavy and kappa loci and expression of fully humanantibodies. Proc Natl Acad Sci U S A 2000;97:722–727.

6 Voet T, Vermeesch J, Carens A et al. Efficient male and female germlinetransmission of a human chromosomal vector in mice. Genome Res2001;11:124 –136.

7 Robl JM, Kasinathan P, Sullivan E et al. Artificial chromosome vectorsand expression of complex proteins in transgenic animals. Theriogenol-ogy 2003;59:107–113.

8 O’Doherty A, Ruf S, Mulligan C et al. An aneuploid mouse straincarrying human chromosome 21 with Down syndrome phenotypes. Sci-ence 2005;309:2033–2037.

9 Ren X, Katoh M, Hoshiya H et al. A novel human artificial chromo-some vector provides effective cell lineage-specific transgene expres-sion in human mesenchymal stem cells. STEM CELLS 2005;23:1608 –1616.

10 Raimondi E, Balzaretti M, Moralli D et al. Gene targeting to the cen-tromeric DNA of a human minichromosome. Hum Gene Ther 1996;7:1103–1109.

11 Ramirez-Solis R, Liu P, Bradley A. Chromosome engineering in mice.Nature 1995;378:720 –724.

12 Moralli D, Vagnarelli P, Bensi M et al. Insertion of a loxP site in asize-reduced human accessory chromosome. Cytogenet Cell Genet 2001;94:113–120.

13 Hooper M, Hardy K, Handyside A et al. HPRT-deficient (Lesch-Nyhan)mouse embryos derived from germline colonization by cultured cells.Nature 1987;326:292–295.

14 Rocchi M, Archidiacono N, Ward DC et al. A human chromosome9-specific alphoid DNA repeat spatially resolvable from satellite 3 DNAby fluorescent in situ hybridization. Genomics 1991;9:517–523.

15 Irvine DV, Shaw ML, Choo KHA et al. Engineering chromosomes fordelivery of therapeutic genes. Trends Biotechnol 2005;23:575–583.

16 Lasserre C, Jami J, Aviles D. Patterns of chromosome segregation inChinese hamster⫻ mouse cell hybrids between permanent cell lines andthymus cells. J Cell Physiol 1980;104:403– 413.

17 Graves JA. Chromosome segregation from cell hybrids. I. The effect ofparent cell ploidy on segregation from mouse-Chinese hamster hybrids.Can J Genet Cytol 1984;26:557–563.

18 Graves JA, Wrigley JM. Chromosome segregation from cell hybrids. II.

2549Paulis, Bensi, Orioli et al.

www.StemCells.com

Do differences in parental cell growth rates and phase times determinedirection of loss? Can J Genet Cytol 1986;28:735–743.

19 Kazuki Y, Shinohara T, Tomizuka K et al. Germline transmission of atransferred human chromosome 21 fragment in transchromosomal mice.J Hum Genet 2001;46:600 – 603.

20 Hadjantonakis AK, Papaioannou VE. The stem cells of early embryos.Differentiation 2001;68:159 –166.

21 Guo J, Jauch A, Heidi HG et al. Multicolor karyotype analyses of mouseembryonic stem cells. In Vitro Cell Dev Biol Anim 2005;41: 278 –283.

22 Todaro GJ, Green H. Quantitative studies of the growth of mouse embryocells in culture and their development into established lines. J Cell Biol1963;17:299 –313.

23 Macieira-Coelho A, Azzarone B. The transition from primary culture tospontaneous immortalization in mouse fibroblast populations. AnticancerRes 1988;8:669 – 676.

24 Paul J. Achievement and challenge. In: Barigozzi C, ed. Origin andNatural History of Cell Lines. New York: Alan R. Liss, Inc., 1978:3–10.

25 Cazillis M, Bringuier AF, Delautier D et al. Disruption of MKK4signalling reveals its tumor-suppressor role in embryonic stem cells.Oncogene 2004;23:4735– 4744.

26 Pralong D, Lim ML, Vassiliev I et al. Tetraploid embryonic stem cellscontribute to the inner cell mass of mouse blastocysts. Cloning StemCells 2005;7:272–278.

See www.StemCells.com for supplemental material available online.

2550 A Human Minichromosome in Mouse ES Cells