CLONING AND STEM CELLS Volume 10, Number 1, 2008 © Mary Ann Liebert, Inc. DOI: 10.1089/clo.2007.0072
Human Embryonic Stem Cells Passaged Using Enzymatic Methods Retain a Normal Karyotype and Express CD30 ALISON THOMSON, DAVINA WOJTACHA, ZOË HEWITT,* HELEN PRIDDLE,† VIRGINIE SOTTILE,‡ ALEX DI DOMENICO, JUDY FLETCHER,§ MARTIN WATERFALL, NÉSTOR LÓPEZ CORRALES,¶ RAY ANSELL, and JIM MCWHIR
ABSTRACT Human embryonic stem cells (hESCs) are thought to be susceptible to chromosomal rearrangements as a consequence of single cell dissociation. Compared in this study are two methods of dissociation that do not generate single cell suspensions (collagenase and EDTA) with an enzymatic procedure using trypsin combined with the calcium-specific chelator EGTA (TEG), that does generate a single cell suspension, over 10 passages. Cells passaged by single cell dissociation using TEG retained a normal karyotype. However, cells passaged using EDTA, without trypsin, acquired an isochromosome p7 in three replicates of one experiment. In all of the TEG, collagenase and EDTA-treated cultures, cells retained consistent telomere length and potentiality, demonstrating that single cell dissociation can be used to maintain karyotypically and phenotypically normal hESCs. However, competitive genomic hybridization revealed that subkaryotypic deletions and amplifications could accumulate over time, reinforcing that present culture regimes remain suboptimal. In all cultures the cell surface marker CD30, reportedly expressed on embryonal carcinoma but not karyoptically normal ESCs, was expressed on hESCs with both normal and abnormal karyotype, but was upregulated on the latter. hESCs, collagenase was used to passage the cells once established, and supported the long-term growth of karyotypically stable, pluripotent cells (Thomson et al., 1998). More recently, this group, in collaboration with others, did report karyotypic abnormalities in hESCs passaged using collagenase (Draper et al., 2004). Aneuploid cells
INTRODUCTION
A
CRITICAL ASPECT of practical hESC culture is method of passage. Most hESC lines require mechanical disaggregation in early passages, suggesting sensitivity at that stage to cell–cell disruption. In the original report of the isolation of
Division of Gene Function and Development, Roslin Institute, Roslin, Midlothian, Scotland. *Current address: Centre for Stem Cell Biology, The University of Sheffield, Alfred Denny Building, Western Bank, Sheffield, S10 2TN, UK. †Current address: D Floor East Block, Queen’s Medical Centre, University of Nottingham, Nottingham, NG7 2UH, UK. ‡Current address: Institute of Genetics, Queen’s Medical Centre, University of Nottingham, Nottingham, NG7 2UH, UK. §Current address: Centre for Regenerative Medicine, University of Edinburgh, 49 Little France Crescent, Edinburgh, EH16 4SB, UK. ¶Current address: Public University of Navarra, Department of Agriculture, Edif. Los Olivos, Campus de Arrosadía 31006, Pamplona, Spain.
1
2
THOMSON ET AL.
were reported in H1, H7, and H9 hESCs passaged with collagenase (Rosler et al., 2004). Others have reported aneuploidy when using enzymatic or chemical, but not mechanical, methods of passaging (Mitalipova et al., 2005). These authors and others (Brimble et al., 2004) have suggested that the poor clonal efficiency of hESCs leads to selective pressure for the accumulation of particular chromosomal rearrangements, and that this occurs as a consequence of generating a singlecell suspension. In this paper the karyoptypic stability and potentiality of hESCs were evaluated following passage using enzymatic and chemical-based methods of disaggregation (collagenase, trypsin in combination with EGTA, the Ca chelating agent, trypsin in combination with EDTA, a CaMg chelator, and EDTA alone). Both collagenase- and trypsin-based methods supported long-term culture without gross chromosomal changes. Furthermore, single-cell suspensions generated with trypsin could be transfected, and clones isolated and expanded, that retain normal karyotype. However, gross karyotypic normality does not exclude small deletions or duplications that can be detected over as few as 10 passages by competitive genomic hybridization. In one experiment, cells passaged with EDTA did become karyotypically abnormal (duplication of the short arm of chromosome 7). CD30 expression has been reported to mark transformed hESCs (Herszfeld et al., 2006). CD30 is a member of the tumor necrosis factor superfamily that is a surface marker for malignant cells in Hodgkin’s disease (Durkop et al., 1992), and is also expressed in embryonal carcinoma (EC) cells, that share many of the properties of ESCs but are karyotypically abnormal (Durkop et al., 2000; Pera et al. 1997). However, screening of both the normal and abnormal lines for CD30 expression revealed that although CD30 was upregulated in the karyotypically abnormal line, it was also detected in cultures of karyotypically normal hESCs.
MATERIALS AND METHODS Cells and media H1, H7, and H9 hESCs were a gift from Geron Corp. (Menlo Park, CA). Prior to the comparison of passage regime, hESCs were cultured and passaged with collagenase as described previously (McWhir et al., 2006; Xu et al., 2001).
K562 cells, a human erythroleukemia cell line, were cultured in RPMI 1640 (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (FBS) and 2 mM L-glutamine. The human embryonal carcinoma (hEC) cell line NTera 2 cells were cultured in DMEM (Invitrogen) supplemented with 10% FBS, 2 mM L-glutamine, 0.1 mM nonessential amino acids and 1 mM sodium pyruvate. Human trabecular bone (HTB) cells were isolated as described in (Sottile et al., 2002) and grown in the same medium as the NTera 2 cells. HM1 murine ESCs were cultured as described in McWhir et al. (1996).
Plating efficiency of hESCs passaged under different regimes Three passage regimes were compared: (1) collagenase IV(200 units/mL) (2) trypsin/EGTA (TEG)—0.25% trypsin in Ca-Mg-free phosphate-buffered saline (PBS) containing 1 mM EGTA detailed in McWhir et al. (1996) and Priddle et al. (2004). (3) EDTA—0.2 mg/mL ethylenediaminotetra acetate (EDTA) in PBS. H9 hESCs were seeded at 1 105 cells/well in six-well plates at passage 40. The collagenase passage regime was as described in Xu et al. (2001). Prior to dissociation in all treatments, wells were washed with 2 mL KODMEM. For the TEG treatment, cells were incubated with 0.5 mL TEG at 37°C for approximately 2 min. KO-DMEM was then added, cells triturated, and centrifuged at 200 g for 2 min before resuspension in conditioned medium (CM). For the EDTA treatment cells were incubated with 0.5 mM EDTA in PBS at 37°C for approximately 3–5 min. After removal of the EDTA the cells were scraped into 1 mL CM. Clumps were gently triturated three times with a 5-mL pipette. After 10 passages the cells were expanded to generate stocks for cryopreservation, genomic DNA preparation, isolation, and karyotype analysis. Each treatment was replicated three times, and three wells were set up for each replicate. The initial seeding for all replicates of all three treatments was from a single collagenase-treated flask (passage 40). One well per replicate was counted 16–18 h after plating to allow calculation of the plating efficiency (number of cells plated down/ number of cells seeded 100%). Once the remaining two wells were confluent one well was treated with TEG to obtain a single-cell suspension that could be counted. On the basis of this cell count the sister well was then passaged.
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE
Three wells of a six-well plate were seeded with approximately 1 105 cells. It was unavoidable that those wells set up using collagenase would have greater variability in actual cell number because collagenase tretament results in cell clumps rather than dissociated cells. The experimental design is summarized schematically in Figure 1A. Replicates from Experiment 1 were designated e1 replicate A, B, or C. The experiment was repeated using earlier passage (P34) H9 cells as described above but with
A P40
46XX
Collagenase Collagenase
TEG
EDTA
Count1 Count2
Count1 Count2
P⫹2
P⫹2
Count1 Count2
Count1 Count2 P⫹2
Count1 Count2
X3 reps
P⫹3
Count1 Count2
P⫹3
Count1 Count2
X3 reps
P⫹3
Count1 Count2
X3 reps
3
the addition of a fourth passage regime similar to TEG, but replacing EGTA with EDTA (TED). Replicates from Experiment 2 were designated e2 replicate 1, 2, or 3.
Calculation of doubling time Cells were seeded at 1 105 cells/well in two, six-well plates for each of two replicates. At each time point two wells were trypsinized in each replicate and the cells were counted using a hemocytometer. Because collagenase-treated cells do not generate a single-cell suspension, additional sacrificial wells were seeded for counting following trypsinization. For each cell type, linear regressions of time were fitted to the logarithm of the cell numbers. Doubling time was estimated as log(2) divided by the slope of the regression. Confidence intervals for doubling times were obtained by replacing the slope by its estimated confidence limits from fit of the regression.
Count1 Count2
DNA preparation P⫹7 P⫹10
P⫹10
Count1 Count2
Count1 Count2
P⫹13
P⫹13
46XX
46,XX,i(7q)
H9-TEGe1repA/B/C
H9-EDTAe1repA/B/C
Plating Efficiency %
B 100 90 80 70 60 50 40 30 20 10 0
TEG EDTA Collagenase
Genomic DNA was prepared from cell cultures by overnight incubation at 55°C in lysis buffer (100 mM Tris–Cl, 5 mM EDTA, 200 mM NaCl, 0.2% SDS, 100 g/mL proteinase K, pH 8.5), followed by incubation with Rnase A (325 g/mL) at 37°C for 1 h, and further treatment with proteinase K (100 g/mL) at 55°C for 2 h. DNA was then precipitated by addition of isopropanol, washed with 70% ethanol, and resuspended in TE buffer (10 mM Tris–Cl, 1 mM EDTA pH 8.0). Genomic DNA for comparative genomic hybridization (CGH) analysis was further purified by phenol/chloroform extraction prior to precipitation.
Comparative genomic hybridization (CGH) analysis 1
2
3
4 5 6 7 Passage Number
8
9
10
FIG. 1. Plating efficiency of H9 hESCs passaged using different treatments. (A) Schematic representation of experimental design. All treatments originated from a single flask of H9 cells cultured for 40 passages (P40) using collagenase. A sister culture was karyotyped and confirmed 46XX. Passages within the experiment are numbered P1, P2, etc. Cells were counted 16–18 h after plating (count 1) and again at time of passage (count 2). (B) Cells were passaged using either collagenase (green), TEG (blue), or EDTA (pink). The plating efficiency values are the mean of the three replicates and the error bars indicate the standard deviation.
The CGH procedure was modified from the original methods of Kallioniemi et al. (1992, 1994). Target human metaphases were obtained from phytohemagglutinin-stimulated blood lymphocytes cultures (72 h). Chromosome spreads were prepared 1 day before hybridization and left overnight at 43°C. Before CGH analysis the slides were treated with pepsin (50 g/mL in 0.01 M HCl), denatured with formamide, and serially dehydrated in ethanol. hESC genomic DNA was labeled by nick translation with SpectralGreen (Vysis–Abbott Molecular, Des Plaines, IL). Spec-
4
tralRed (Vysis) labeled normal human genomic DNA was used as reference DNA. Labeled fragments were 500–1000 bp. Equal amounts (400 ng–1 g) of hES and reference DNA were mixed and precipitated with 20–30 g of unlabeled Cot-1 DNA (Roche, Indianapolis, IN). DNA samples were dissolved in hybridization solution and denatured at 75°C. Slides were hybridized for 72 h in a humidity chamber at 37°C. After hybridization slides were washed in SSC at high stringency and dehydrated through an ethanol series. Slides were counterstained with DAPI in Vectashield (Vector Laboratories, Burlingame, CA). The hybridized metaphase cells were examined using a workstation composed of an epifluorescence microscope coupled to a CCD camera (Olympus, Melville, NY) and a CytoVision 2.7 CGH software (Applied Imaging, San Jose, CA). CGH analysis was developed over 10 complete metaphases. For each metaphase three fluorochrome images (DAPI, SpectralGreen and SpectralRed) were acquired and processed using high-resolution CGH software. Statistical analysis was performed using three confidence levels (95, 99.5, and 99.9%).
THOMSON ET AL.
Flow cytometry analysis for expression of stem cell surface markers Single cell suspensions of hESC were stained as described in Hewitt et al. (2006). Data for 40,000 events/sample were acquired and analyzed using CellQuestPro software (BD, UK). Three independent experiments were performed.
Flow cytometry analysis for the cell surface CD30 (Ber-H2) epitope Adherent hESCs washed with PBS were incubated with TEG to dissociate cells. Recovered cells were pelleted at 200 g and resuspended at between 2 106 and 1 107 cells/mL in PBS (Ca2 and Mg2 free) with 0.2% BSA. Aliquots (100 L) were stained with either CD30-PE (Santa Cruz, Santa Cruz, CA; sc19658) or isotype control (mIgG1-PE; BD Pharmingen, San Diego, CA; 555749). Cells were incubated in the dark at 4°C for 30 min, washed, and finally resuspended in PBS. Flow cytometry was performed using a Becton Dickinson FACSAria and Diva analysis software. Live cells were gated using forward- and side-scatter profiles. Data were acquired for 10,000 live events.
Karyotype analysis Exponentially growing cultures were treated with Karyomax colcemid solution (Invitrogen, UK) at 100 ng/mL for 2 h at 37°C. Karyotypes were prepared and analyzed as described in McWhir et al. (2006), Hewitt et al. (2007), and Gallimore and Richardson (1973).
Cell viability assay/ propidium iodide (PI) staining Cells were disaggregated using TEG and a single-cell suspension prepared in FACS buffer (PBS, 0.1% BSA, 0.1% Sodium Azide). Cells were aliquoted at 2–5 105 cells/sample, washed, and resuspended in 300 L FACS buffer. The flow cytometer was set up for autofluorescence signals in FL2 channel using unstained cells. Cells for PI staining were incubated with 15 L of staining solution (PI at 50 g/mL in PBS) with gentle mixing for 1 min. PI fluorescence data (FL2 channel; 575/42 nm) was then acquired for the stained cells. Data for 5–10,000 ungated events were acquired on the FACSAria and analyzed using FACSDiva software [Beckton Dickinson Immunocytometry Systems (BD), UK].
Fixation of cells Sample fixation was performed immediately following CD30 staining. Cells were incubated with 0.1% paraformaldehyde for 15 min at room temperature, washed, and finally resuspended in 100 L PBS.
Transfection A linearized plasmid containing a neomycin resistance gene flanked by a PGK promoter and poly-adenylation signal was used to test stable transfection efficiencies. Lipofection was carried out using Lipofectamine 2000 (Invitrogen) according to the manufacturers instructions. Subconfluent cells (1 106) in three wells of a sixwell plate were lipofected with 5 g of linear plasmid per well. Two different electroporation methods were compared, using 1 106 cells in the log phase of growth, treated with TEG to give a single-cell suspension. Using a Gene Pulser (BioRad, Hercules, CA), the cell/DNA mix was electroporated in PBS at room temperature in a 0.4-cm electrode cuvette at 940-F, 200 V. Using a Multiporator (Eppendorf, Westbury, NY) the
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE
cells were swelled in hypo-osmolar buffer (Eppendorf) for 20 min at room temperature, mixed with DNA, and pulsed in a 0.4-cm electrode cuvette at 300 V for 100 S. Following either electroporation method, cuvettes were left at room temperature for 10 min prior to plating the cells onto 15-cm matrigel-coated dishes. Selection in G418 at 150 g/mL was applied 48 h after transfection. G418 resistant colonies were fixed with methanol, stained with 10% Gurrs R66 Giemsa in phosphate-buffered saline pH6.8, and counted (Priddle, 2004).
RT-PCR RNA was isolated using an RNeasy mini kit (Qiagen, Valencia, CA). One-step reverse transcription–polymerase chain reaction (RT-PCR) was performed using Superscript One-Step RTPCR with Platinum Taq (Invitrogen). Forty-five cycles were used for each primer pair. Primers were designed to span exons and distinguish genomic DNA from cDNA products. CD30 mRNA was amplified using the forward primer 5-AGCTAGAGCTTGTGGATTCCA-3 and the reverse primer 5-GTCTTCTTTCCCTTCCTCTTCC-3 to give a product of 464 bp. -Actin mRNA was amplified using the forward primer 5-GCCACGGCTGCTTCCAGC-3 and the reverse primer 5CAAGATGAGATTGGCATGGCT-3 to give a product of 528 bp.
Cryopreservation Cells were resuspended in hESC medium and mixed with an equal volume of hESC medium supplemented with 10% serum replacement and 20% DMSO. Cells were transferred to cryovials, stored at 80°C overnight and transferred to 150°C the next day.
In vitro differentiation of hESCs Embryoid bodies were prepared by suspension culture as described in Hewitt et al. (2007). Embryoid bodies were plated onto gelatin-coated plates in differentiation medium [Knockout DMEM (Invitrogen), 10% FCS (Globepharm, Deerfield, IL), 2 mM L-glutamine (Invitrogen), 1 nonessential amino acids (Invitrogen), and 100 M -mercaptoethanol (Invitrogen)] and allowed to reattach to the culture surface. Differentiation was allowed to proceed for a further 2 weeks.
5
In all cases differentiated cells were washed with PBS, fixed with 4% paraformaldehyde (PFA) at room temperature for 20 min, washed two times with PBS followed by a 15-min incubation with PBS and 0.1% Triton X-100, and two washes with PBS. Nonspecific protein binding was blocked with 2% bovine serum albumin (BSA) for 30 min at room temperature. Primary antibodies were bound to their antigens in PBS with 2% BSA for 1 h at room temperature. Antibodies and dilutions were: monoclonal -sarcomeric actinin (Sigma, St. Louis, MO) at a 1 in 500, monoclonal anticardiac troponin 1 (Chemicon International, Temecula, CA) at a 1 in 1000, monoclonal anticardiac troponinT (Neomarkers, Freemont, CA) at a 1 in 100, monoclonal anti--tubulin III (Sigma) 1 in 200, rabbit polyclonal anti-NF200 (Sigma) 1 in 100 dilution, monoclonal anti--fetoprotein (Sigma) 1 in 500. Unbound antibody was removed by two 10-min room temperature washes with PBS followed by a 10-min incubation with PBS and 0.05% Tween-20. FITC-conjugated goat antimouse IgG (Sigma) and FITC conjugated goat antirabbit (Vector Laboratories) were used as secondary antibodies (1 in 1000) in PBS with 2% BSA by incubation at room temperature for 1 h. Coverslips were mounted in Vectashield (Vector Laboratories), and viewed with a fluorescence microscope (Zeiss, Thornwood, NY). Control reactions with no primary antibody were performed to confirm that the secondary antibody did not stain cells nonspecifically. Staining on undifferentiated hESCs was also performed to assess level of background staining and found to be negligible.
Formation of tumors in severe combined immunodeficiency disease (SCID) mice and histological analysis SCID mice were obtained from Harlan UK Ltd. (Bicester, UK) and maintained in a sterile environment. Human ESCs for injection into SCID mice were disaggregated by treatment with collagenase or TEG, washed once in PBS, and resuspended in PBS at 1 108 cells/mL. For each population tested, a 100-L aliquot was injected into the leg muscle of each of three SCID mice. Three to 5 months later, the mice were sacrificed and the tumors removed, fixed with 4% PFA for 20 min at room temperature, embedded in paraffin wax using a Shandon Hypercentre XP
6
THOMSON ET AL.
processor (Shandon Scientific UK Ltd), cut into 10-m sections using a Microm HM325 rotary microtome (MICROM international GmbH), and stained using standard hematoxylin and eosin (H&E) protocols.
RESULTS
similar doubling time (13.3 h) to mESCs (13.1 h) when measured at subconfluent densities. Doubling time increased as TEG-treated H9 cells attained confluence. Collagenase-treated H9 cells did not attain confluence over the course of the experiment. When the growth rates of collagenase-passaged cells and TEG-passaged cells were compared at lower initial density (Fig. 2b) TEGpassaged cells again reached confluence earlier.
Plating efficiency
FIG. 2. Doubling time of hESCs. (A) Growth curves calculated for murine HM1 ESCs and human H9 ESCs by seeding 104 to 105 cells per well of a series of six-well plates. For each cell type, linear regressions of time were fitted to the logarithm of the cell numbers. Doubling time was estimated as log(2) divided by the slope of the regression. Confidence intervals for doubling times were obtained by replacing the slope by its estimated confidence limits from fit of the regression. The y-axis shows logarithmically transformed cell numbers for: collagenase-treated H9 cells (black), murine HM1 ESC (blue), and H9 hESCs following long-term passage using TEG (red). TEG treated cells were greater in number at the start of the experiment and reached confluence faster as indicated by the nonlinearity of those data. Collagenase treated cells grew with a constant doubling time of 17.3 h while murine ES cells had a doubling time of 13.3 h. When the experiment was repeated for TEG-treated and collagenase-treated H9 cells at lower seeding density (B) TEG-treated cells again reached confluence more rapidly. (C) The phenotype of H9 cells 90 h after plating at 104 per well of six-well plate. TEG-treated cells (Ci) showed less differentiation and formed larger colonies than did collagenase-treated cells (Cii).
A 15.5
HM1 H9-TEG
Log (cell count)
14.5 13.5
H9-Collagenase 12.5 11.5 10.5 0
50
100
Time in hours
B H9-TEG
13
Log (cell no.)
The response of H9 hESCs to three methods of disaggregation (collagenase, TEG, and EDTA) was followed over 10 passages. Plating efficiencies were obtained for each treatment (Fig. 1b). All three replicates of cells cultured by passage with collagenase failed to survive the seventh passage. Although in our experience it is unusual to lose a culture, this exemplifies the erratic nature of the plating efficiencies we have observed with this method of passage particulary when plating at the relatively low cell densities used in this experiment. The TEG passage regime was less variable than either EDTA or collagenase, yielding an average plating efficiency of 32%. All three replicates of the EDTA-treated cells passaged at particularly poor efficiency at passage 8 (Fig. 1b) H9 hESCs passaged with collagenase and TEG proliferated with significantly different (p 001) doubling times of 17.3 and 13.3 h, respectively (Fig. 2). Human ESCs passaged with TEG had
12
11
H9-Collagenase
10
0
50 Time in hours
C i
ii
20 M
20 M
100
7
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE
As initial plating densities were identical, this confirms an accelerated growth rate in TEGtreated cells leading to earlier confluence. Figure 2C shows the morphology of collagenase- and TEG-passaged cells 90 h postplating. TEG-passaged cells typically formed larger colonies, while collagenase-passaged cultures, in addition to undifferentiated hESCs, contained a second, differentiated cell type of fibroblastic morphology.
cultured for 23 passages with collagenase, and had a diploid karyotype (46XX) (Fig. 3) at passage 40 in this experiment (Fig. 1A). Cells disaggregated for 13 passages using TEG retained a normal karyotype. By contrast, cells disaggregated using EDTA had acquired an abnormal karyotype with all three replicates having a duplication of the short-arm of chromosome 7 and deletion of the long arm [46,XX,i(7q)]. This result was confirmed by comparative genomic hybridization (CGH) (Fig. 4). Although TEG-treatment led to no gross karyotypic abnormalities, CGH suggested the presence of random deletions clustered near the telomeres of many chromo-
Karyotypes The H9 hESCs used to set up the comparison of different disaggregation treatments had been
a.
b.
1
2
3
6
7
8
13
14
15
19
20
9
21
4
5
1
2
3
10
11
12
6
7
8
16
17
18
13
14
15
x
19
20
22
9
21
4
5
10
11
12
16
17
18
22
x
c.
d. 1
2
3
6
7
8
13
14
15
19
20
9
21
4
5
10
11
12
16
17
18
22
x
x
7
7
7
y
FIG. 3. Karyotypes of H9 cells. (a) Collagenase-treated cells at the start of the experiment (passage 40, 23 passages with collagenase in our lab); (b) passaged 13 times with TEG (H9-TEGe1repA); (c) passaged 13 times with EDTA(H9EDTAe1repA). Karyotypes in b and c are representative of 10 spreads fully analyzed for each of three independent replicates. No abnormalities were apparent in the parental collagenase-treated cells (a) nor in any of three replicates of the TEG-treated cells(H9-TEGe1reps A, B, and C) (b). However, all three replicates of the EDTA protocol (H9-EDTAe1reps A, B, and C) carried the same isochromosome isop7 (c and d). (d) Chromosome 7 pairs from three different EDTA replicates.
A.i. ? ? ?
1 n18
6 n19
2 n18
7 n22
3 n21
8 n19
4 n20
9 n21 10 n21 11 n21 12 n20
13 n20 14 n19 15 n21 16 n17 17 n22
19 n23
20 n23
5 n19
21 n20
18 n22
22 n21
y
x n19
ii. ? ? ?
1 n23
2 n18
3 n19
4 n18
6 n19
7 n18
8 n21
9 n22
13 n19
14 n21
15 n21
16 n21 17 n21
19 n20
20 n22
21 n21
10 n20 11 n22
22 n19
B.
0.5
7
5 n20
12 n20
18 n21
y
x n17
1.0
1.5
n23
FIG. 4. Comparative genome analysis (CGH) of H9 hESCs. (A) Summarizes two independent CGH experiments for one replicate of TEG-passaged H9 cells showing a tendency to accumulate small subtelomeric deletions. When CGH is repeated with the same DNA samples (Ai and Aii) similar, but not identical subtelomeric deletions also appear. Circles are areas showing a tendency for small deletions to accumulate in similar subtelomeric regions. Similar, but not identical subtelomeric deletions were also associated with the other two TEG replicates but not with EDTA or collagenase (data not shown). (B) Summarizes CGH experiments for EDTA-passaged H9 hESCs showing an abnormality on chromosome 7 with duplication of the p arm and loss of the q arm.
9
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE
A) H7 Collagenase SSEA-1
100
101
102 FL1-H
103
Events 0
0
0
Events
Events
80
80
Tra-1-60
80
SSEA-4
100
104
101
102 FL2-H
103
104
101
102 FL2-H
103
104
100
101
102 FL2-H
103
104
102 FL2-H
103
104
CD9
0
0
Events
Events
80
80
Tra-1-81
100
101
102 FL2-H
103
100
104
B) H7 TEG SSEA-1
100
101
102 FL1-H
103
Events 0
0
0
Events
Events
80
80
Tra-1-60
80
SSEA-4
100
104
101
102 FL2-H
103
104
101
102 FL2-H
103
104
100
101
CD9
0
0
Events
Events
80
80
Tra-1-81
100
101
102 FL2-H
103
104
100
FIG. 5. Flow cytometric analysis of stem cell markers on H7 hESCs passaged with. (A) collagenase or (B) TEG. Shaded areas are indicative of staining with the stem cell marker antibody (green FITC conjugated and red R PE conjugated secondaries). Data represents 40,000 events; representative plots are shown from three independent experiments.
10
THOMSON ET AL.
6
7
8
9
re p
B ED TA
H TB
5
H 9
4
c1
re p
2 c1
re p c2
H 9
ED TA
H 9
ED TA
H 9
TE G
H 9
TE G
re pA
re pC
e as en co
TE G
lla g
3
H 1
2
H 9
1 CD30
K5 62
N Te ra 2
2O
H
M
C
A
10
11
1000 bp 800 bp 600 bp 400 bp 200 bp
-Actin
1000 bp 800 bp 600 bp 400 bp 200 bp
B 120 102 PE-A
103
104
57.1% M1
0 102 PE-A
103
104
41.1% M1
101
102 PE-A
103
104
viii) H9 EDTA e1repB
69.9% M1
0
0
Events
100
120
120
vii) H9 EDTA e2rep2
100
101
102 PE-A
103
104
101
102 PE-A
103
104
vi) H1 TEG
35.4% M1
100
101
102 PE-A
103
104
ix) H9 EDTA e1repC
79.4% M1
0
101
Events
100
100
120
0
Events
44.8% M1
101
v) H9 TEGe1repA
120
120
iv) H9 collagenase
100
120
104
M1
0
103
Events
102 PE-A
Events
101
Events
100
0.6%
0
90.6% M1
0
M1
0
95%
iii) HTB
Events
120
ii) NTera2
Events
Events
120
i) K562
100
101
102 PE-A
103
FIG. 6A & B.
104
100
101
102 PE-A
103
104
11
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE
C without fixation
103
104
0 102 PE-A
103
104
Events
120
101
101
102 PE-A
103
104
101
102 PE-A
103
104
Events
120
100
100
101
102 PE-A
103
104
102 PE-A
103
104
61.1% M1
100
101
102 PE-A
103
104
57.9% M1
0
62.8% M1
0
0
Events
120
iii) H9 EDTAe1repC
101
120
100
100
0
63.3% M1
0
0
Events
120
ii) H9 TEGe1repA
100
120
102 PE-A
Events
101
Events
100
45.8% M1
Events
120
Events
75.2% M1
0
0
Events
120
i) NTera2
with fixation 120
overlay
100
101
102 PE-A
103
104
100
101
102 PE-A
103
104
FIG. 6. Expression of CD30 in hESCs. (A) RT-PCR analysis of expression of CD30 transcript. Cells with normal karyotype as well as those with an abnormal karyotype express CD30 mRNA. Detection of beta-actin product as a loading control. Lanes 1, no RNA control; lane 2, positive control NTera2; lane 3, positive control K562; lane 4, H9 hESCs passaged with collagenase (H9-Collagenase e2rep1); lane 5, H1 hESCs passaged with TEG; lane 6, H9 hESCs passaged with TEG, replicate A from the first passaging experiment (H9-TEGe1repA); lane 7, H9 hESCs passaged with TEG, replicate C from first passaging experiment (H9-TEGe1repC); lane 8, H9 hESCs passaged with EDTA, replicate 2 from the second passaging experiment (H9-EDTAe2rep2); lane 9, H9 hESCs passaged with EDTA, replicate B from the first passaging experiment (H9-EDTAe1repB, abnormal karyotype); lane 10, H9 hESCs passaged with EDTA, replicate C from the first passaging experiment (H9-EDTAe1repC, abnormal karyotype); lane 11, negative control HTB cells. Position of molecular weight markers are shown on the left. (B) Flow cytometric analysis of CD30 expression. Cells with normal karyotype as well as those with an abnormal karyotype express the CD30 epitope. (i) K562, (ii) NTera 2, (iii) HTB, (iv) H9 hESCs passaged with collagenase, (v) H9 hESCs passaged with TEG (H9 TEGe1repA), (vi) H1 hESCs passaged with TEG, (vii) H9 hESCs passaged with EDTA; these cells were from replicate 2 of the second passaging experiment (H9-EDTAe2rep2), (viii) H9 hESCs passaged with EDTA; these cells were from replicate B of the first passaging experiment (H9-EDTAe1repB, abnormal karyotype), (ix) H9 hESCs passaged with EDTA; these cells were from replicate C of the first passaging experiment (H9-EDTAe1repC, abnormal karyotype). The analysis was performed on at least three separate occassions; representative plots are shown. (C) CD30 staining is affected by fixation treatment of the cells. CD30 staining was analyzed by flow cytometry. (i) N-Tera2; (ii) H9 TEGe1repA, (iii) H9 EDTAe1repC. The analysis was performed on at least two separate occasions; representative plots are shown.
12
somes. No similar pattern was observed with collagenase-treated cells or with EDTA-treated cells (data not shown). No specific deletions achieved statistical significance; however, the clustered pattern of subtelomeric deletion was consistent across three independent replicates, and suggests a mixed population containing cells with many different small deletions. The different passage regimes were repeated using earlier passage cells (P34), and extended to include TED. Cells passaged with TEG, TED, trypsin in the absence of a chelating agent, or collagenase were found to be karyotypically normal (data not shown). By contrast to the first experiment, cells passaged with EDTA were also found to be karyotypically normal (data not shown). H1 and H7 cells previously passaged with collagenase have also been switched to the TEG-based protocol and shown to retain a normal karyotype (data not shown). Analysis of telomere length showed that cells subjected to all four treatments showed similar telomere lengths, suggesting that the various methods of passage had no differential effect on telomere length, and that the telomeres were being maintained (data not shown).
Stem cell markers The expression of surface markers characteristic of undifferentiated (SSEA3, SSEA- 4,TRA-1-60 and TRA-1-81) and differentiated (SSEA-1) hESCs by TEG-passaged H7 hESCs (Fig. 5) was compared by flow cytometry with that of collagenase passaged cells. Cells passaged using TEG retained a high level of expression of markers characteristic of undifferentiated hESCs and a reduction in expression of SSEA-1, a marker of differentiation. This was consistent with the morphological observation that TEG-passaged cells are predominantly of undifferentiated phenotype. Similar results were obtained for H1 and H9 cells, and H9 TEG-passaged cells were also positive for expression of Oct-4 and alkaline phosphatase (data not shown). As concerns have been raised about the viability of hESCs once in a single cell suspension, cell
THOMSON ET AL.
viability was assessed by propidium iodide staining after passaging with TEG. The average cell viability, was 93.1 2.0% (SD) (n 8).
CD30 expression CD30 expression has been reported in transformed hESCs, but not in karyotypically normal sister cultures (Herszfeld et al., 2006), suggesting that its expression may be diagnostic for karyotypic abnormality. hESC lines were screened for expression of CD30 transcripts by RT-PCR (Fig. 6A). Because CD30 expression is characteristic of EC cells and of human erythroid progenitors, the human EC line NTera 2 and the human erythroblastoid leukemic cell line K562 were included as positive controls and human trabecular bone cells (HTBs), as a negative control. Expression was detected in the positive control cell lines NTera 2 and K562 but not in HTBs. The lines H9EDTA e1repB and C (EDTA passaged cells from the first experiment, lanes 9 and 10), identified as being karyotypically abnormal, also showed expression of CD30 mRNA. Surprisingly, other hESC lines that had been shown to have a normal diploid karyotype at a similar passage also expressed the transcript. This included cells passaged with collagenase, EDTA, and TEG (lanes 4–8). Screening these cell lines for expression of CD30 protein by flow cytometry using the same monoclonal antibody used by Herszfeld et al. (2006), showed, as expected, that a high proportion of cells in both K562 and N-Tera 2 cultures were CD30 positive (Fig. 6B), and that HTBs lacked the CD30 epitope. Consistent with the RTPCR data, CD30 expression was detected in all the hESC lines tested. The abnormal cell lines H9EDTA e1repB and C (viii and ix) did express higher levels than the “normal” hESCs (iv–vii). Herszfeld et al. (2006) had reported CD30 flow cytometry data using fixed cells. To test whether fixation steps affected the level of CD30 staining several cell lines were examined using different staining protocols either with or without fixation. Postfixation of samples reduced the number of cells that stained positive for CD30 expression
FIG. 7. Differentiation of H9 cells in vivo and in vitro. (A) Sections from tumors generated from, collagenase- (a–f) or TEG-treated (g–n) H9 cells. Cartilage (a, g), mucus-secreting epithelium (b, m), pigmented epithelium (c), neuroglia (d), sebaceous gland (e, n), adipose tissue (h), smooth muscle (i), primitive neural tissue (j), ganglia (k), sweat glands (l). Bar 50 m. (B) Immunostaining of in vitro differentiated cultures from TEG-passaged H9 cells using primary antibodies against -sarcomeric Actinin (a), Troponin 1 (b), cardiac Troponin T (c), 3-Tubulin (d), Neurofilament 200 (e), -Fetoprotein (f).
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE
FIG. 7.
13
14
THOMSON ET AL.
(Fig. 6C), as did prefixation (data not shown). Monitoring expression of CD30 on fixed cells, however, still showed that cultures of hESCs passaged with collagenase, TEG, and EDTA all contained CD30-positive cells. Although the proportion of positive cells was similar, the abnormal hESC lines H9-EDTAe1rep B and C displayed higher levels of CD30 expression than the normal hESC lines. We have also detected expression of the CD30 epitope on another hESC line, RH1, derived at Roslin Institute (Fletcher et al., 2006) and passaged with collagenase.
Potentiality in vivo and in vitro To test the potentiality of H9 hESCs disaggregated by TEG or collagenase, cells were injected into SCID mice. Four animals were injected with each cell type, giving rise to tumors in 1/4 of the recipients of collagenase-treated cells and all four recipients of the TEG-treated cells. Tumors generated from TEG- and collagenase-dissociated cells gave rise in each case to a variety of tissues derived from the three germ layers (Fig. 7A). There was no apparent accumulation of undifferentiated cells in recipients of cells from either treatment. When submitted to in vitro differentiation protocols, the TEG-treated cells gave rise to cell types arising from all three germ layers (Fig. 7B). TEG-treated H9 cells were tested for stable transfection efficiency using electroporation, multiporation, and lipofection procedures. A set of conditions was defined using a multiporator (Eppendorf), which gave transfection efficiencies some fivefold higher (4 104) than were achieved by lipofection or with a Gene Pulser electroporator (Table 1). The TEG-passaged cells have subsequently been used successfully in several transfection exTABLE 1. COMPARISON OF STABLE TRANSFECTION EFFICIENCIES Method Lipofection Gene Pulser Electroporator Multiporator
Number of clones
Transfection efficiency
41 36
4.1 105 3.6 105
231
4.1 104
Transfection efficiencies were obtained for H9 hESCs using lipofection, a gene pulser electroporator and a multiporator.
periments. The TEG, EDTA, and collagenase methods of disaggregation were compared for their effects on the proportion of frozen cells surviving thaw and replating. TEG treatment led to a significant improvement in freeze/thaw survival over collagenase (76.2 vs. 15.8%, n 3).
DISCUSSION In this study hESCs passaged by single-cell dissociation using trypsin/EGTA (TEG) were maintained and retained stable morphology, karyotype, and growth rate over 10 passages. Single-cell dissociation of hESCs did not lead to karyotypic instability. However, subkaryotypic analysis by CGH suggests that TEG passage may be associated with small subtelomeric deletions and amplifications.
Plating efficiency and doubling time A doubling time of 17.3 h was calculated for collagenase-passaged H9 hESCs, compared with 13.3 h for TEG-passaged cells. A previous report calculated the doubling time of H9 hESCs to be 35.3 h (Amit et al., 2000). This discrepancy may reflect the differences in culture conditions between labs; notably the growth of cells on matrigel versus a fibroblast feeder layer. The shorter doubling time of TEG passaged cells may be a consequence of the reduction in differentiated companion cells when compared with sister cultures passaged with collagenase. The high plating efficieny and short doubling time of the TEG-treated cells contrasts with reports that hESCs are sensitive to single-cell dissociation, for example (Amit et al., 2000; Draper et al., 2004; Hasegawa et al., 2006). However, Hasegawa and colleagues were able to subclone hESCs that they then showed were adapted to single-cell dissociation (Hasegawa et al., 2006). Furthermore, the ability to maintain and recover cells as single-cell suspensions affords the opportunity for cell sorting by flow cytometry (Hewitt et al., 2006). The efficiency of clonal isolation can also be increased by culturing the cells in 2% oxygen (Forsyth et al., 2006; Hewitt et al., 2006). Hence, it appears that hESCs can adapt in culture to overcome sensitivity to single-cell dissociation without gross karyotypic abnormality.
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE
Single-cell dissociation using TEG supports stable karyotype In the original report of the isolation of hESCs, long-term karyotypic stability was demonstrated when collagenase was used for routine passage of established lines (Thomson et al., 1998). Since then, several reports have described the detection of aneuploid hESCs using similar, collagenasebased, passage regimes (Amit et al., 2000; Draper et al., 2004; Inzunza et al., 2004; Mitalipova et al., 2005). Of particular note, the abnormalities in chromsomes 17q and 12 have been shown to occur on several independent occasions (Draper et al., 2004; Ludwig et al., 2006). Mitalipova et al. (2005) reported abnormal karyotypes after extended passage in culture using enzymatic- or chemical dissociation-based methods, but not when physical methods of disaggregation were used. Buzzard et al. (2004) hypothesized that the use of mechanical disaggregation may prevent the types of chromosomal abnormalities previously reported. H9 hESCs were passaged using two enzymatic methods (collagenase and TEG) for similar numbers of passages to the Mitalipova et al. (2005) report, and maintained a normal gross karyotype with both protocols. Of the three methods compared in this study only trypsin-based protocols generated a single-cell suspension, and the karyotypic stability of these trypsin-passaged cells does not support the hypothesis that single-cell cloning gives rise to karyotypic instability. Notwithstanding concerns about subtelomeric stability, this study indicates that the use of the TEG regime for rapid cell amplification, electroporation, and freezing offers important advantages. Regardless of the method of passage, caution is advisable when cells are cultured at low density, as this may lead to the selection of cells harboring abnormal karyotypes.
Subkaryotypic abnormalities can accumulate in TEG-passaged hESCs Comparative genomic analysis confirmed the diploid karyotypes of cells passaged with either TEG or collagenase. Pluripotency of TEGtreated cells was not obviously different from that of collagenase-treated cells when assayed both in vivo (tumors in SCID mice) and in vitro (response to embryoid body formation and removal of bFGF and conditioned medium). This
15
study has not addressed the possible accumulation of point mutations as reported for latepassage hESCs (Maitra et al., 2005). Analysis of subkaryotypic abnormalities at the level of CGH have demonstrated small subtelomeric deletions and duplications when using the TEG passage regime, although these were statistically significant only up to 95% confidence levels. This technique detects only those aberrations that are shared by a large proportion of the population. Hence, random changes that are not fixed in the population would not be detected. The tendency for accumulation of subtelomeric deletions must give rise to concern about using the TEG passage regime for the preparation of cells for therapeutic use.
EDTA-treated cells can acquire abnormal karyotype In the first experiment (Fig. 1a), EDTA treatment was associated with the appearance, in three separate replicates, of the same duplication of the short arm of chromosome 7. The complete absence of cells with normal karyotype in all three replicates argues strongly for a competitive advantage of this rearrangement when cells are passaged with EDTA. As the same abnormality was detected in all three replicates it is likely that the duplication was present in the starting population. It was noted that during the culture of the cells that acquired the karyotypic abnormality there was one exceptionally poor passage affecting all three replicates (Fig. 1, plating efficiency of 3% compared with an average of 24%). Those cells harboring the duplication of the short arm of chromosome 7 may have had a selective advantage at this point. Repeating the passage regime using EDTA with an earlier passage (p34) of collagenase-passaged cells, no karyotypic abnormalities were detected. Draper et al. (2004) reported that the gain of chromosome 12 and amplification of chromosome 17q in H1 and H14 hESCs was observed after the cells had been through clonal selection or switched from a feeder-layer to a feeder-free culture system. Both of these treatments could result in low cell densities that may allow a competitive advantage for cells of abnormal karyotypes. Observation of a particularly poor recovery among the three replicates of EDTA-treated cells at p8, followed by fixation of the isoP7 rearrangement, would be consistent with the presence of a minority population
16
THOMSON ET AL.
of isoP7 cells that preferentially survived under the conditions of that particular passage regime.
SUMMARY The TEG protocol offers advantages of convenience and repeatability, leading to reduction in the time required to amplify cells for experimental purposes. TEG-treated cells are amenable to electroporation, leading to stable transfection efficiencies approximately one order of magnitude higher than those previously reported (Siemen et al., 2005; Zwaka and Thomson, 2003). TEG treatment provides fivefold freeze/thaw efficiency gains over the collagenase protocol. Cells switched from collagenase- to TEG-passage have now been used for several projects. A clonal line, M2, derived from H9 hESCs that had been passaged using collagenase for 23 passages, and then TEG for 57 passages, also retained a diploid karyotype (Hewitt et al., 2006). The TEG protocol has been used successfully to passage cells for single-cell sorting (Hewitt et al., 2007), and clones from these experiments have been expanded using TEG-passage and retained a normal karyotype. Spontaneous chromosomal aberrations (SCA) were previously shown to be reduced by culture of TEG-passaged hESCs at physiologicaly relevant oxygen partial pressure (Forsyth et al., 2006). SCAs are precursors to karyotype abnormalities. Of concern was the observation that the rate of SCAs was higher in later passage cells. Others have reported an increase in mutation rate with passage of hESCs (Maitra et al., 2005). These data reinforce the need to thoroughly characterize cells prior to therapeutic use, and to provide failsafe protection for cancerous phenotypes by such means as conditional suicide genes (Hewitt et al., 2007; Schuldiner et al., 2003). Comparative gene expression profiling of early- and late-passage hESCs reveals differences in gene expression due to both karyotypic and epigenetic adaptation to culture (Enver et al., 2005). Together with the present observation that culture regime can lead to detectable levels of gene amplification and deletion, these observations also underline the need to establish a definitive assay of “normality” that does not rely solely on karyotype. CD30 was found to be expressed by all hESCs examined in this study, but its level of expression was greater in the karyotypically abnormal hESCs. Hence, the use of CD30 as a diagnostic
tool for normality of hESCs as suggested by Herszfeld et al. (2006) may require quantitative analysis and will need to be supported by examination of additional markers. All models of the application of hESCs in regenerative medicine are based upon massive ex vivo expansion of a few or possibly even a single cell(s) from the inner cell mass into very large cell numbers in vitro. To ensure the safety of the resulting graft a great deal remains to be done to optimize culture conditions, define and identify “normal cells,” and to provide failsafe features that protect against cellular overgrowth postengraftment (Hewitt et al., 2007) .
ACKNOWLEDGMENTS We thank Nick Forsyth for the primers used in the RT-PCR analysis and discussions on the CD30 work, Laura Dick for tumor embedding, Dr. Edward Duvall for the tumor analysis, and Dave Waddington for statistical advice. This work was supported by the Biotechnology and Biological Sciences Research Council (BBSRC) and Geron.
REFERENCES Amit, M., Carpenter, M.K., Inokuma, M.S., et al. (2000). Clonally derived human embryonic stem cell lines maintain pluripotency and proliferative potential for prolonged periods of culture. Dev. Biol. 227, 271–278. Brimble, S.N., Zeng, X., Weiler, D.A., et al. (2004). Karyotypic stability, genotyping, differentiation, feeder-free maintenance, and gene expression sampling in three human embryonic stem cell lines derived prior to August 9, 2001. Stem Cells Dev. 13, 585–597. Buzzard, J.J., Gough, N.M., Crook, J.M., et al. (2004). Karyotype of human ES cells during extended culture. Nat. Biotechnol. 22, 381–382. Draper, J.S., Smith, K., Gokhale, P., et al. (2004). Recurrent gain of chromosomes 17q and 12 in cultured human embryonic stem cells. Nat. Biotechnol. 22, 53–54. Durkop, H., Latza, U., Hummel, M., et al. (1992). Molecular cloning and expression of a new member of the nerve growth factor receptor family that is characteristic for Hodgkin’s disease. Cell 68, 421–427. Durkop, H., Foss, H.D., Eitelbach, F., et al. (2000). Expression of the CD30 antigen in non-lymphoid tissues and cells. J. Pathol. 190, 613–618. Enver, T., Soneji, S., Joshi, C., et al. (2005). Cellular differentiation hierarchies in normal and culture-adapted human embryonic stem cells. Hum. Mol. Genet. 14, 3129–3140.
hESCs PASSAGED ENZYMATICALLY RETAIN A NORMAL KARYOTYPE Fletcher, J.M., Ferrier, P.M., Gardner, J.O., et al. (2006). Variations in humanized and defined culture conditions supporting derivation of new human embryonic stem cell lines. Cloning Stem Cells. 8, 319–334. Forsyth, N.R., Musio, A., Vezzoni, P., et al. (2006). Physiologic oxygen enhances human embryonic stem cell clonal recovery and reduces chromosomal abnormalities. Cloning Stem Cells 8, 16–23. Gallimore, P.H., and Richardson, C.R. (1973). An improved banding technique exemplified in the karyotype analysis of two strains of rat. Chromosoma 41, 259–263. Hasegawa, K., Fujioka, T., Nakamura, Y., et al. (2006). A method for the selection of human embryonic stem cell sub-lines with high replating efficiency after single cell dissociation. Stem Cells 24, 2649–2660. Herszfeld, D., Wolvvetang, E., Langton-Bunker, E., et al. (2006). CD30 is a survival factor and a biomarker for transformed human pluripotent stem cells. Nat. Biotechnol. 24, 351–357. Hewitt, Z., Forsyth, N., Waterfall, M., et al. (2006). Fluorescence-activated single cell sorting of human ES cells. Cloning Stem Cells 8, 225–234. Hewitt, Z., Priddle, H., Thomson, A.J., et al. (2007). Ablation of undifferentiated human embryonic stem cells by exploiting innate immunity against the -galactosyltransferase epitope. Stem Cells 25, 10–18. Inzunza, J., Sahlen, S., Holmberg, K., et al. (2004). Comparative genomic hybridization and karyotyping of human embryonic stem cells reveals the occurrence of an isodicentric chromosome after long-term cultivation. Mol. Hum. Genet. 10, 461–466. Kallioniemi, A., Kallioniemi, O.P., Sudar, D., et al. (1992). Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science 258, 818–821. Kallioniemi, O.P., Kallioniemi, A., Piper, J., et al. (1994). Optimizing comparative genomic hybridization for analysis of DNA sequence copy number changes in solid tumors. Genes Chromosome Cancer 10, 231–243. Ludwig, T.E., Levenstein, M.E., Jones, J.M., et al. (2006). Derivation of human embryonic stem cells in defined conditions. Nat. Biotechnol. 24, 185–187. McWhir, J., Schnieke, A.E., Ansell, R., et al. (1996). Selective ablation of differentiated cells permits isolation of embryonic stem cell lines from murine embryos with a non-permissive genetic background. Nat. Genet. 14, 223–226. McWhir, J., Wojtacha, D., and Thomson, A.J. (2006). Routine culture and differentiation of human embryonic
17
stem cells. In Human Embryonic Stem Cells Protocols: Methods in Molecular Biology, vol. 331. K. Turksen, ed. (Humana Press, Clifton, NJ) pp. 77–90. Maitra, A., Arking, D.E., Shivapurkar, N., et al. (2005). Genomic alterations in cultured human embryonic stem cells. Nat. Genet. 37, 1099–1103. Mitalipova, M.M., Rao, R.R., Hover, D.M., et al. (2005). Preserving the genetic integrity of human embryonic stem cells. Nat. Biotechnol. 23, 19–20. Pera, M.F., Bennett, W., and Cerretti, D.P. (1997). Expression of CD30 and CD30 ligand in cultured cell lines from human germ-cell tumors. Lab Invest. 76, 497–504. Priddle, H. (2004). Transfection of human embryonic stem cells. In Gene Targeting and Stem Cells: Advanced Methods. A.J. Thomson and J. McWhir, eds. (Garland Science/BIOS Scientific Publications, New York), pp. 171–202. Rosler, E.S., Fisk, G.J., Ares, X., et al. (2004). Long-term culture of human embryonic stem cells in feeder-free conditions. Dev. Dynam. 229, 259–274. Schuldiner, M., Itskovitz-Eldor, J., and Benvenisty, N. (2003). Selective ablation of human embryonic stem cells expressing a “suicide” gene. Stem Cells 21, 257–265. Siemen, H., Nix, M., Endl, E., et al. (2005). Nucleofection of human embryonic stem cells. Stem Cells Dev. 14, 378–383. Sottile, V., Halleux, C., Bassilana, F., et al. (2002). Stem cell characteristics of human trabecular bone derived cells. Bone 30, 699–704. Thomson, J.A., Itskovitz-Eldor, J., Shapiro, S.S., et al. (1998). Embryonic stem cell lines derived from human blastocysts. Science 282, 1145–1147. Xu, C., Inokuma, M.S., Denham, J., et al. (2001). Feederfree growth of undifferentiated human embryonic stem cells. Nat. Biotechnol. 19, 971–974. Zwaka, T.P., and Thomson, J.A. (2003). Homologous recombination in human embryonic stem cells. Nat. Biotechnol. 21, 319–321.
Address reprint requests to: Dr. Alison Thomson Division of Gene Function and Development Roslin Institute Roslin, Midlothian EH25 9PS, UK E-mail:
[email protected]
