| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
EMBRYONIC STEM CELLS: CHARACTERIZATION SERIES |
aDepartment of Comparative Medicine and
bDivision of Hematology, Department of Medicine, University of Washington, Seattle, Washington, USA
Key Words. Human embryonic stem cell • Doubling time • Cloning • Transfection
Correspondence: Carol B. Ware, Ph.D., University of Washington, Box 357190, Seattle, Washington 98195, USA. Telephone: 206-616-5143; Fax: 206-685-3006; e-mail: cware{at}u.washington.edu
Received September 16, 2005;
accepted for publication July 27, 2006.
First published online in STEM CELLS EXPRESS August 17, 2006.
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
In October 2003, the NIH established three extramural "Exploratory Centers for Human Embryonic Stem Cell Research," with the goal of broadening hESC expertise using the NIH-approved lines. NIH-approved hESC lines are those derived prior to August 9, 2001, and 22 eligible lines were listed on the NIH Stem Cell Registry website (http://stemcells.nih.gov/research/registry/) at the time of this study. An Exploratory Center shared by the University of Washington (UW) and the Fred Hutchinson Cancer Research Center generated an infrastructure for improving access to hESCs by local researchers, and our Core Laboratory has uncovered biological differences among the 15 NIH-approved lines acquired. These experiences should be useful for other institutions engaging in hESC research.
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
-irradiated with 3,000 rads and plated at 104 cells per cm2. All hESC lines were initially passaged according to the recommendations of their suppliers. To convert to enzymatic passage, cells were washed with phosphate-buffered saline (PBS) (Invitrogen) and exposed to 1.2 U/ml dispase (Invitrogen) dissolved in PBS supplemented with 10% fetal bovine serum (FBS) (Schenk Packing, Stanwood, WA, http://www.schenkpacking.com) or embryonic stem-qualified FBS (Invitrogen). Once colonies began to lift from the plate, the cell layer was washed off using a pipette, collected in a centrifuge tube containing culture medium, washed once by centrifugation, mechanically dispersed to 
100-µm diameter clusters using a 5-ml pipette, and washed again before plating. When converting ES Cell International (ESI) cell lines to enzyme passage, 1 mg/ml collagenase Type IV (Invitrogen) in hESC medium was used, and the cells were monitored closely for lifting of the colony edges. Once converted to enzyme passage, ESI lines were switched to dispase passage. For generating a single-cell suspension, trypsin-EDTA (Invitrogen) was used.
Immunohistochemistry
Maintainence of an undifferentiated phenotype was established by immunohistochemistry using antibodies for Oct-4 (1:200 dilution) (R&D Systems Inc., Minneapolis, http://www.rndsystems.com) and SSEA-4 (1:50 dilution) (Chemicon, Temecula, CA, http://www.chemicon.com) and by staining for alkaline phosphatase (AP) activity using a Black Alkaline Phosphatase Substrate Kit II (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com). For immunolabeling, the clusters were visualized using a biotinylated secondary antibody included in the Universal Quick Kit using Nova Red (both from Vector Laboratories).
Expression Analysis by Reverse Transcription- Polymerase Chain Reaction
To test differentiation ability, hESCs were either collected as undifferentiated cells or removed from MEFs and cultured on gelatinized 10-cm plates in the medium described above, except 10% FBS was substituted for SR. Cells were allowed to differentiate for 10 or 21 days and then washed twice with PBS and lysed on the plate, and whole RNA was isolated according to the manufacturer's recommendations (RNeasy; Qiagen, Valencia, CA, http://www1.qiagen.com). Contaminating genomic DNA was removed from the RNA preps using DNA-free (Ambion, Austin, TX, http://www.ambion.com). cDNA synthesis was performed using a kit (Protoscript; New England Biolabs, Ipswich, MA, http://www.neb.com), and the resultant cDNA was treated with RNase included in the kit. Reverse transcription-polymerase chain reaction (RT-PCR) was used to determine expression of 
-fetoprotein (AFP) (endoderm), neurofilament heavy chain (NFH) (ectoderm), renin (mesoderm), Oct-4, Nanog (undifferentiated), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) using the following primers: AFP forward, AGAACCTGTCACAAGCTGTG (680 bp); AFP reverse, GACAGCAAGCTGAGGATGTC; NFH forward, TGAACACAGA-CGCTATGCGCTCAG (400 bp); NFH reverse, CACC-TTTATGTGAGTGGACACAGAG; renin forward, AGTCG-TCTTTGACACTGGTTCGTCC (590 bp); renin reverse, GGTAGAACCTGAGATGTAGGATGC; Oct-4 forward, GA-GCAAAACCCGGAGGAGT (309 bp); Oct-4 reverse, TTC-TCTTTCGGGCCTGCAC; Nanog forward, GCTTGCGCTTTGAAGCA (255 bp); Nanog reverse, TTCTTGACTGGGAC-CTTGTC; GAPDH forward, TGATGACATCAAGAAG-CTGGTGAAG (240 bp); and GAPDH reverse, TCCTTGGAGGCCATGTGGGCCAT.
Mycoplasma Detection
A luminometer-based kit was used for detection of mycoplasma (MycoAlert; Cambrex, Walkersville, MD, http://www.cambrex.com). When mycoplasma was detected, it was verified using a polymerase chain reaction (PCR)-based test (Stratagene, La Jolla, CA, http://www.stratagene.com).
Karyotype
G-banded karyotype was analyzed by the Cytogenetics Laboratory at UW.
Growth Efficiency
Cell numbers were determined after dispase dispersal followed by trypsinization of an aliquot of the resulting cluster suspension for a hemacytometer count. Cells were plated in clusters at a density of 105 cells per 35-mm plate on feeders. Each trial was plated in duplicate. Cells were allowed to grow for 72 hours and then trypsinized and counted using a hemacytometer. Growth efficiency for all lines was measured three times. Growth efficiency reflects a combination of plating efficiency and doubling time and is a reflection of expected expansion time. This is measured as the amount of time in hours to double once.
Cloning Efficiency
Cells were dispersed to primarily single-cell suspension using trypsin, counted using a hemacytometer, and plated in duplicate 10-cm plates at 100, 1,000, and 10,000 cells per plate. The plates were fluid changed at least twice weekly, and colonies were counted when large enough to visualize macroscopically (15–18 days). Cloning efficiency for each line was measured at least three times.
Transfection Efficiency
All hESC lines were compared with an equivalent number of mouse ESCs (mESCs). hESCs were dispersed with dispase, and mESCs were dispersed with trypsin. Two million cells were placed into 0.4-cm cuvettes (Bio-Rad, Hercules, CA, http://www.bio-rad.com) in PBS containing 1 µg of a linearized PGK-neo construct. They were exposed to a pulse of 250 V, 500 µF (Gene Pulser II; Bio-Rad). Following electroporation, hESCs were plated onto single 10-cm plates seeded with neomycin-resistant MEF feeders. mESCs were plated without feeders on two gelatinized 10-cm plates in medium containing 103 U/ml leukemia inhibitory factor and 15% ES-qualified FBS rather than serum replacer. One-hundred eighty µg/ml neomycin (G418; Invitrogen) was added 24 hours later to the mESC cultures and 48 hours later to the hESC cultures. Plates were fluid changed at least twice weekly and counted when colonies could be visualized (10–13 days for mESCs and 15–18 days for hESCs). Transfections were performed on the Bio-Rad Gene Pulser II electroporator at least three times for each cell line. In addition, one million each of H1, BG02, and BG03 cells were transfected with linearized PGK-neo using a Nucleofector II device and the Mouse ES Cell Nucleofector Kit (Amaxa Biosystems, Koeln, Germany) using program A13. For stable expression comparison, mouse ESCs (105) were transfected using the same kit and a different program, A23. The nucleofection protocol recommended by the manufacturer was followed for both the mouse and human ES lines. Cells were selected as described above. Only one trial for three human lines was run on the Nucleofector II device as a means of determining whether relative differences seen between lines on the Bio-Rad device held true on the Amaxa device.
Cryopreservation
The method used for cryopreservation has been described previously [5].
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
Preliminary Characterization
All cell lines except three from Bresagen were received frozen. Note that in some cases, provider codes are used for the lines and in others the NIH code was the accepted name (Table 1). The number of cells provided and their viability following cryopreservation by the provider differed significantly between suppliers. Fewer than 10 clusters were recovered from most of the frozen lines. HSF-6 was a notable exception. Receipt of actively growing lines (Bresagen) facilitated establishment and subsequent cryostorage of the lines. Survival of rare cells following freeze that take 2–3 weeks to appear as clusters suggests that the survivors may represent subclones of the parent culture. Conversely, all the hESC lines are able to survive freezing with an efficiency of >50% if frozen using embryo methods. Following a slow-cooling protocol [2], the cells have shown no long-term degradation in liquid nitrogen storage after 2 years, and there appear to be no striking differences between cell lines with regard to surviving cryopreservation both short term and long term.
|
|
Maintenance of an Undifferentiated Phenotype. Immunohistochemistry for SSEA-4 and Oct-4 and detection of AP activity showed the expected patterns, indicating an undifferentiated state in all nine lines. Within healthy cultures, loss of these markers can be seen in whole or partial colonies (Fig. 1, arrows at SSEA-4 [H13], Oct-4 [H9], and AP[BG03]), indicating an inherent propensity to differentiate. Oct-4 patterns vary among the cell lines, in that not all cells label with Oct-4 in H1, H9, and BG01 clusters, whereas the other lines appear to label uniformly. Because this pattern is not consistently reproducible, this may be an artifact of both quality of passage of the sample and density of Oct-4 nuclear staining. The downregulation of AP activity that accompanies the initial stages of hESC differentiation makes this a rapid and unambiguous method of culture assessment.
|
Differentiation Ability. To confirm the ability of an hESC line to form the three embryonic germ lineages, in vitro differentiation was performed, followed by RT-PCR using primers to detect transcripts associated with endoderm (AFP), ectoderm (NFH), and mesoderm (renin) formation. The protocol relied upon undirected differentiation. These results were compared with those of equivalent cells cultured to prevent differentiation using primers for Oct-4 and Nanog. For all 10 lines examined, cells allowed to differentiate for 10 or 21 days expressed RNA for the three germ lineages (data not shown). Some lines expressed differentiation markers for renin (BG01, BG02, and ES04) and NFH (H7, H13, H14, BG01, BG02, BG03, and hSF6) prior to induction of differentiation, indicating that they were carrying detectable differentiation when cultured under self-renewal conditions. RNA for Oct-4 and Nanog was assessed among the lines prior to differentiation and following 21 days of differentiation (Fig. 2). Oct-4 and Nanog expression persisted in H7 and BG03, suggesting that they retain undifferentiated cells despite the differentiation culture conditions, whereas H9, BG02, and hSF6 had Oct-4-positive but Nanog-negative cells following differentiation. A faint Nanog band was seen in the differentiated H13 culture with no apparent Oct-4. This needs to be explored further and may be an artifact. Overall, other means must be used to confirm comparison of lines for the persistence of stem cells, whereas robustness of contribution to specific lineages will need to be compared using protocols directing the cells down specific pathways.
|
Mycoplasma Detection. Mycoplasma testing was carried out on all 15 lines using a luminometer-based test. Although 14 lines were free from contamination, one line harbored mycoplasma upon receipt, as determined with the cooperation of the supplier and confirmed using PCR. This provided an opportunity to examine the microscopic appearance of mycoplasma contamination (Fig. 3). This contamination, representing either Mycoplasma hyorhinis or Mycoplasma orale, excluded this line from further evaluation.
|
G-Banded Karyotype. All lines were karyotypically stable to the passage number shown in Table 2. However, within roughly 20 passages from receipt, two of five chromosomal spreads from H13 (passage 36) cells contained a trisomy 17. By passage 39, all five chromosomal spreads of the H9 line contained a trisomy X. Thus, both abnormalities likely confer a growth advantage [7]. Table 3 provides a summary of all karyotype data collected.
|
Growth Efficiency.
Cell growth efficiency was determined following plating of 105 cells on a MEF feeder, growth for 3 days, and a hemacytometer count of trypsinized cells. Using this method, differences between lines can be caused by a number of different parameters, including plating efficiency, tendency to differentiate, and relative growth response between the lines to culture density, whereas true doubling time takes into account that growth rate follows a curve. This method measures relative growth between lines. Other means are required to measure the individual factors that cause the comparative growth differences. Doubling times ranged from 31 hours for HSF-6 and BG03 to 56 hours for H13 and BG02 (Table 2). While making these measurements, other cell line-specific differences were noted. For example, on several occasions during routine culture, we observed an abrupt increase in the rate of cell proliferation. H9 grows very slowly when plated at low density and noticeably more rapidly when plated at high cell density. For instance, H9 was karyotyped following an increased growth rate with reduced differentiation and was found to be abnormal (passages 49 and 60; Table 3). Cell growth efficiency is influenced by a number of parameters, including medium composition, the presence of MEFs, karyotype integrity, survival following passage, and cell density. Values reported in Table 2 reflect growth as measured by the time it took to double when cells were seeded at uniform and optimal density on MEFs. A relatively high SEM was observed between trials. Because of this, a comparison of results between growth efficiency for HSF-6 (31.3 ± 3.7 hours) versus H13 (57.0 ± 12.0 hours) approached significance following three trials (p = .12; t test) and can only be regarded as a trend.
Cloning Efficiency. The frequency with which single cells form colonies ranged from 0.8% ± 0.1% for BG02 to 9.2% ± 4.7% for H13 (p = .22, t test). By comparison, standard cloning efficiencies for the mouse cell line R1 are typically in the range of 10%. Again, variability between trials prevents ability to reach significance and suggests that improvements in culture conditions will likely affect reproducibility. hESC clones frequently acquired karyotypic abnormalities. For example, of three H7 clones analyzed, two had a normal 46XX karyotype, whereas the other acquired a trisomy 17; only two of three clones of H14 were normal, and two of two BG03 clones carried a trisomy 17 (Table 3).
Transfection Efficiency. Developing mechanistic insight into hESC function requires methods for genetic manipulation. We compared nine hESC lines for their ability to be stably transfected via electroporation using the Bio-Rad electroporation system. Relative to the murine ESC line, R1, hESC lines displayed a widely variable efficiency of stable transfection, ranging from 0% for BG02 to 53% ±23 for BG03 (p = .11; t-test) (Table 2). This relative ranking of transfection efficiency changed in a trial transfection using the Amaxa nucleofection system. Using the Amaxa, the first two BG02 neo-resistant colonies were created, BG03 efficiency dropped from 53% to 2%, and the transfection efficiency of H1 improved from 7% to 17%. Since transfection efficiency increases approximately 10-fold in mouse ESCs using Amaxa-based nucleofection (A.M. Nelson and C.B. Ware, unpublished observation), the H1 line is very efficiently transfected by nucleoporation. There is variability between experiments, and the nucleoporation comparison was run only once, but it reveals differences between the transfection systems that highlight the need to devise optimal conditions for each line. The mouse line used as a standard can also show transfection variability from run to run. We generally expect 100 mouse ES colonies when two million cells are transfected. There are likely many variables that can affect efficiency that hinge on the quality of the culture, but the number of times the construct is frozen and thawed before use will also make a clear difference. Standardizing the human line to a mouse for each trial should remove this variable.
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
During routine culture, embryonic stem cells are in dynamic balance between undifferentiated expansion versus a drive toward differentiation. In addition, they are exquisitely sensitive to environmental conditions. This sensitivity could be either an inherent attribute or a reflection of inadequate culture conditions. The combination of inherent attributes and currently uncontrollable culture variables is the likely cause of the relatively large standard errors observed in this study. For instance, the tendency for a culture to differentiate could dramatically change the overall culture ability in these trials. Interestingly, the longer it took the cells to plate and grow, the higher the cloning efficiency; and the higher the transfection efficiency, the larger the variability between runs. It is not apparent that slow growth is preferable in an hESC, but increased cloning and transfection efficiencies are desirable and not reliably achieved, according to the results presented here.
The recommended hESC culture conditions vary widely among the different suppliers. After attending three different training courses (WiCell, University of California San Francisco [UCSF], and ESI), we established our ability to generate stocks of cells that had been cultured as recommended by the suppliers before attempting to convert each line to a uniform set of culture conditions. Our effort to generate a uniform, simplified method for culturing hESCs is pragmatically based. Mechanical passage of individual colonies is impractical for experiments requiring large numbers of cells. Variability in adapting to our simplified culture conditions further highlights the biological differences between the lines. Although we are not yet able to establish methods for enzymatic passage of 2 of the 14 cell lines tested, there remains the expectation that we will eventually be successful.
Another key difference among the various cell lines tested here is an apparent propensity of some cell lines to acquire karyotype abnormalities over time. An increase in doubling time with a decrease in background differentiation can be a warning sign that the karyotype is abnormal. It has been noted that when plates are so dense that colonies begin to grow together, growth characteristics alter (information learned the hESC training courses at UCSF and ESI). Another means of inducing increased rate of growth is by subcloning from a single-cell suspension [8] (our observations). Transgenic clones appear to be more karyotypically stable if they are allowed to divide before applying antibiotic selection pressure, presumably because they are never in a single-cell state. Clonal growth from single cells is likely to be a means of selecting for growth advantage and potential abnormality. Some lines are inherently less karyotypically stable than others [8, 9]. Using our culture conditions, H9 and H13 have a tendency to develop problems even when handled by an experienced technician, whereas ES04, hSF6, BG01, BG02, H1, H7, and H14 appear relatively stable. Overgrowth of cells with a growth advantage is rapid and can be a serious problem when studying self-renewal culture conditions. Some suppliers send frozen cultures with only a few viable cells. When cryopreserved cells survive poorly, there must be an advantage for those that do survive, which may represent another inadvertent means of selecting for karyotype change. Poor survival on thaw means starting a line with only a few clusters that are unlikely to adequately represent the prefreeze culture. Initiating a culture with only a few cells also increases the passage number before cells can be stored, making it difficult to establish a margin of safety for the number of frozen samples at low passage of karyotypically normal cells. Consequently, distributors of frozen hESCs should use freezing protocols to maximize consistency.
The differences among the various hESC lines described here indicate that some cell lines are likely better suited for particular applications than others. It is therefore important that researchers be able to choose the appropriate cell lines for their application. In this regard, intellectual property issues pose a significant roadblock to NIH-funded hESC research [10], with a cost of $500–$6,000 per line and MTA provisions that prevent the transfer of hESC lines between laboratories. We have attempted to address these obstacles in two ways. First, we reached agreements with all five hESC suppliers that allow local investigators access to all hESC lines purchased by the Core Laboratory to perform preliminary studies before purchasing the optimal line directly from the supplier. Our agreement requires that preliminary studies be conducted in the Core Laboratory. However, one beneficial consequence of this arrangement is that it places investigators new to hESC research in prolonged contact with Core Laboratory personnel, and prolonged training appears necessary for gaining expertise in hESC culture techniques. A second barrier to hESC research is the prohibition against transferring hESC lines between laboratories. For example, our laboratory has recently described an improved technique for freezing hESCs [2] but is prevented from distributing hESCs frozen in this manner to outside laboratories. To date, only Bresagen has agreed to allow our Core Laboratory to distribute their hESCs to local investigators, provided that the investigators then make a substantially reduced payment directly to Bresagen.
In summary, NIH-approved hESCs exhibit differences in cell culture. The development of uniform culture conditions will facilitate comparisons between these and other hESC lines. A fundamental restructuring of MTA provisions is necessary to broaden the accessibility of NIH-funded hESC research.
|
DISCLOSURES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDisclosuresAcknowledgmentsReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
