Introduction
Human embryonic stem (ES) cells are derived from the inner cell mass (ICM). These cells can proliferate without differentiation and have the ability to give rise to many types of somatic cells as well as germ cells. Therefore, human embryonic stem cells (hESCs) can be used in studying the processes of early human development and provide potential applications in regenerative diseases. Importantly, hESCs therapy holds promise for injury/repair by transplantation therapy (
Thomson et al., 1998). However, the underlying molecular mechanisms that maintain the pluripotency of hESCs are still poorly understood. Experimental evidence demonstrates that mouse embryonic stem cells (ESCs) pluripotency maintenance requires leukemia inhibitory factor (LIF) and bone morphogenetic protein 4 (BMP4) (
Pease et al., 1990;
Ying et al., 2003); basic fibroblast growth factor (bFGF) and Activin can support self-renewal of human ESCs (
Amit et al., 2004;
Xiao et al., 2006); and recently, rat ESCs were derived and can be maintained in 3i culture medium (
Buehr et al., 2008;
Li et al., 2008). Consequently, there should be multiple regulatory molecules that play roles in ESCs stemness maintenance.
In recent years, several reports suggest that human ESCs develop aneuploidy when propagated
in vitro (
Draper et al., 2004;
Zeng et al., 2004;
Maitra et al., 2005;
Mitalipova et al., 2005;
Xiao et al., 2006). Here, we established an aneuploid hESC line (referred to as H1T; the T stands for translocation) that demonstrates a self-renewal advantage, indicating that probably the gene expression pattern of some key molecules was changed and these molecules may be momentous for hESCs pluripotency maintenance or differentiation. We decided to compare the aneuploid H1T hESC line versus the euploid H1 hESC line by a combination of mRNA and comparative genomic hybridization (CGH) microarray in order to screen for pluripotency maintenance related molecules. CGH is capable of detecting loss, gain and amplification of the copy number of a gene at the level of chromosomes. It can also reveal a characteristic pattern that includes mutations. Therefore, new information can be obtained about which genes might be involved in stem cell pluripotency maintenance at both mRNA and DNA levels by analyzing the data formed via a combination of these two approaches.
Materials and methods
Human ES culture
hES cell lines were maintained on feeders in ES medium, which contains 80% DMEM/F12 medium, 20% KNOCKOUT serum replacement (Invitrogen), 1 mmol/L L-glutamine, 0.1 mmol/L β-mercaptoethanol, 1% nonessential amino acids, and 4 ng/mL human basic fibroblast growth factor (hbFGF). Cultures were passaged as cells about once a week by incubation in 1 mg/mL collagenase IV for about 30 min at 37°C. Conditional medium was collected and feeder free culture was done according to Xu et al., 2001(
Xu et al., 2001).
Karyotype analysis
Karyotype analysis (G-banding) was performed on at least 20 cells from each sample. Karyotypes were analyzed and reported according to the International System for Human Cytogenetic Nomenclature. Multicolor spectral karyotyping (SKY) was performed according to Schröck et al. (
1996).
Cytogenetic methods
Cells were arrested in metaphase by exposure to Colcemid (Invitrogen) at 0.01 - 0.1 µg/mL for 2 to 18 h, treated with prewarmed cancer hypotonic solution (0.04 mol/L KCl, 0.02 mol/L Hepes, 0.5 mmol/L EGTA) for 40 min at 37°C, and fixed in 3∶1 methanol:acetic acid. Slides were banded with 0.5% Enzar-t (Intergen) and stained with Leishman stain. Twenty metaphases were analyzed and chromosome abnormalities were described according to ISCN (An International System for Human Cytogenetic Nomenclature) guidelines (
Mitelman, 1995).
An additional slide was made for Spectral KaryotypingTM (SKY) (
Schröck et al., 1996) analysis according to the protocol supplied with the probe (Applied Spectral Imaging, Carlsbad, CA). The SKY probe is a mixture of whole chromosome paint probes for each chromosome, combinatorially labeled with one to five fluorochromes. Slides were pretreated with RNAse for 1 h and pepsin (1.2%) for 2 min, both at 37°C. The probe was denatured, pre-annealed with Cot-1 DNA for one hour, hybridized with the separately denatured slide for 72 h, washed and detected according to ASI protocol. Metaphase images were acquired using the ASI SpectraCube SD200 system. DAPI (4', 6-diamidino-2-phenylindole) counterstained images were captured and inverted by SkyView software (ASI, Carlsbad, CA) to permit enhanced banding. Eleven metaphases were captured and analyzed.
To confirm a small rearrangement of chromosomes 10 and 17, simultaneous fluorescence
in situ hybridization (FISH) paint was performed (probes generated from flow-sorted chromosomes and expanded by degenerate oligonucleotide primer-PCR (DOP-PCR) (
Schröck et al., 1996). Probes were differentially labeled with digoxygenin and biotin, respectively, also by DOP-PCR. Probe was mixed with Cot-1 DNA and hybridization mix (2X SSCP, 50% formamide, 10% dextran sulfate), denatured at 70°C for 5 min, preannealed at 37°C for 30 min, placed on separately denatured slides and hybridized at 37°C overnight. Slides were washed in 1X SSC/0.3% NP-40 at 72°C for 5 min, followed by 2 changes of 4XSSC/0.1%Tween 20 at room temperature for 5 min each. Hybridized probes were detected with rhodamine-anti-dig (Roche, Indianapolis, Indiana) and FITC-avidin (ID Laboratories, London, Ontario, Canada), following manufacturer's instructions.
The genomic DNA of H1 and H1T hESCs were labeled by Cy3 or Cy5 fluorescence dye, respectively, and hybridized on oligo-based microarray (Agilent Technology).
Results
Characterization of H1T hESC line at the chromosome level
We observed aneuploidy in the H1 hES cell line in our lab, after long-term passage with trypsin (
Xiao et al., 2006). G-banding analysis showed that chromosome 10 is deleted below band q24.3 (Fig. 1). Chromosome 17 has a large addition of unidentified material at the tip of the short arm (Fig. 1). All the other chromosomes appear normal by G-banding analysis. The net result is an unbalanced karyotype, thus parts of chromosome 10q and 17p might be monosomic, while part of some other unidentified chromosome might be trisomic. G-banding was not of sufficient resolution to identify the origin of the additional material on chromosome 17. Both copies of chromosome 22 appear to be normal by G-banding (Fig. 1).
In order to identify the origin of extra material on chromosome 17, we further analyzed the karyotype by SKY analysis. This analysis demonstrated that chromosome 10 is deleted (Fig. 1), and the additional material on the short arm of the chromosome 17 is mostly chromosome 17 material, with a small piece of chromosome 10 translocating to the tip, corresponding to the few bands missing from the shortened chromosome 10 (Fig. 1 and supplementary Table 1). SKY analysis does not show a balanced (reciprocal) translocation, i.e. no chromosome 17 material appears on the tip of the short chromosome 10. G-banding would predict it to be there. That is, the region of 10 on the “add (17)” corresponds to that portion missing from chromosome 10 and the G-banding would suggest there should be a balanced (reciprocal) translocation between chromosomes 10 and 17. SKY was unable to confirm that the very small region of chromosome 17 moved to chromosome 10. To confirm a small rearrangement of chromosomes 10 and 17, fluorescence
in situ hybridization (FISH) was performed using paints only for chromosomes 10 and 17, in contrasting colors. This confirmed the SKY result, and also indicated the small bit of chromosome 17 on the end of the chromosome 10 (
Xiao et al., 2006).
Comparison of HIT with H1 hES cell lines by combination of mRNA and CGH microarray
Although they are aneuploid, H1T hESC cell lines express high levels of undifferentiated hESC markers (Oct4, Nanog, SSEA-3, SSEA-4, Tra-1-60, Tra-1-81), are able to form teratomas, and show ability to differentiate into all three germ layers as do the euploid H1 human ES cells (
Xiao et al., 2006). These activities of the H1T line suggest that H1T human embryonic stem cells maintain pluripotency and show a self-renewal advantage compared with the parental line (H1). We speculated that this self-renewal advantage might be associated with the upregulation of genes that maintain pluripotency or downregulation of genes that facilitate differentiation. In addition, a high level of DNA amplification always associates with an elevated mRNA, so the combination of mRNA and CGH could confirm the results that some key molecules play important roles in pluripotency maintenance. Therefore, we performed both mRNA microarray and CGH microarray to identify genes which were located at the translocation chromosome region using the mRNA and genomic DNA of H1T human ESC and H1 human ESC.
As expected, the CGH microarray could not detect the reciprocal translocation between chromosomes 10 and 17 due to neither gain or loss of genomic DNA (Fig. 2). Surprisingly, we detected additional chromosome abnormalities which were not revealed by karyotyping. A small hemizygous deletion on chromosome 17p spanning about 1.4 Mb (from 0.1 Mb to 1.5 Mb) was revealed (Fig. 2 and supplementary Table 2). Both copies of chromosome 22 appear to be normal by karyotype analysis (Fig. 1). However, a small hemizygous deletion of about 6.1 Mb on chromosome 22 (from 29.9 Mb-36.0 Mb) was demonstrated by CGH microarray analysis (Fig. 2 and supplementary Table 3). It is very difficult to detect alterations of chromosome 22 by low resolution karyotyping since chromosome 22 is one of the smallest chromosomes in the human genome. The deletion of a chromosome fragment will not only cause the loss of one allele of the genes located in that region, but also possibly cause the fusion of two genes. In many cases fusion genes such as the Philadelphia (Ph) chromosome in chronic myeloid leukemia (CML), the result of a reciprocal translocation between chromosomes 22 and 9 fusing the
abl and
bcr genes(
Shtivelman et al., 1985), cause malignant transformation of cells. Interestingly, the
bcr gene located on chromosome 22q (from 21.8 Mb to 21.9 Mb) (
Prakash and Yunis, 1984) is only about 8 Mb away from the deleted region of the chromosome 22 in H1T human ESC.
Among the genes located in the trisomic region, there are several genes which have been reported to be effective in anti-apoptosis and pluripotency maintenance, e.g. TBX2, Survivin, Col1a1, Tex14 and Wnt3(Table 1). On the other hand, as there is a hemizygous deletion on chromosome 22, we suspect that the gene localized on the monosome may have effects on facilitating differentiation or inhibiting self-renewal, these include Fbxo7, HSPC117, YWHAE, and HMG2L1(Table 2) . The results of mRNA microarray are analogous to that of CGH microarray; thus, the combination of the two approaches stabilize the data at both the RNA and DNA levels.
Discussion
Among current analytical methods for chromosome structure and content G-band karyotyping is the conventional approach to identify genomic alterations. However, small DNA fragment deletions or additions might have taken place before detectable alterations are revealed by G-banding. Thus CGH microarray appears to provide the best resolution for detection of chromosomal deletions or additions. CGH microarray shows gene expression changes at the DNA level, and essentially it ought to be in accord with mRNA microarray. Therefore, the combination of the two approaches provides additional solid evidence of gene expression alteration.
As the H1T hESC line demonstrates self-renewal advantage (
Xiao et al., 2006), the H1T hESC line provides a feasible tool to study the function of specific genes by comparing it with the normal H1 hESC line. The gene up-regulated in the H1T hESC line might function to induce anti-apoptosis, proliferation and self-renewal. Conversely, the gene down-regulated might function to stimulate apoptosis and differentiation.
Among the genes up-regulated in H1T hESC line,
TBX2 is associated with many carcinomas such as melanoma (
Vance et al., 2005) and breast cancer (
Jacobs et al., 2000) and was identified as a potent immortalizing gene that can suppress senescence (
Jacobs et al., 2000).
Survivin encodes an inhibitor of apoptosis (
Ambrosini et al., 1997). It is mainly expressed in the embryo, fetus and carcinoma, but infrequently expressed in normal adult tissues and organs (
O’Connor et al., 2000). It is reported that Survivin suppression by RNAi in mouse ESC will reduce cell proliferation and colony formation (
Coumoul et al., 2004). Besides, Survivin is involved in anti-apoptosis in ESC through regulating Oct4 indirectly and STAT3 directly (
Guo et al., 2008). This evidence indicates that Survivin plays an important role in pluripotency maintenance.
Col1a1 encodes part of the procollagen I which is a precursor of type I collagen, and type I collagen mRNA is down-regulated during mesenchymal stem cell differentiation (
Chen et al., 2004). What’s more, it is shown that Col1al plays a role in maintaining type A spermatogonia in an undifferentiated state in testis (
He et al., 2005).
Tex14 encodes a predicted protein with 2 protein kinase domains which is expressed mainly in male germ cells. Tex14 exhibits an expression profile in untreated differentiating ESCs: initially decreased and then increased, and high expression of Tex14 is detected in undifferentiated ESC (
Silva et al., 2009). Therefore, Tex14 is important for male germ cell maturity, and might be involved in pluripotency maintenance. Moreover, it is predictable that members of the Wnt family ought to have a higher expression in the H1T hESC line because it is reported that the Wnt signal pathway can be activated by inhibition of the glycogen synthase kinase (GSK) signaling pathway, processing the expression of stemness associated transcription factors such as Oct4, Rex-1 and Nanog (
Sato et al., 2004). It can also stimulate the proliferation of human ESC although it is only able to maintain the pluripotency of human ESC partially and temporarily (
Sato et al., 2004;
Dravid et al., 2005). To sum up, the genes up-regulated in H1T hESC line probably positively regulates self-renewal of hESC and are supportive for pluripotency maintenance.
As there is a deletion on chromosome 22, it is not difficult to imagine that the genes located on this region might support differentiation or inhibit self-renewal. Among these down-regulated genes, Fbxo7 interacts with cIAP1 which is a member of the inhibitor of apoptosis proteins, mediates the ubiquitination of cIAP1, thereby regulates the function of cIAP1 (
Chang et al., 2006). For that reason, down-regulation of Fbxo7 might contribute to anti-apoptosis. Though little is known about the function of HSPC117 in embryo development, it is reported that HSPC117 is essential to facilitate placenta development in post-implantation stages (
Wang et al., 2010). This implicates HSPC117 having functions contributing to ESC differentiation. YWHAE may interact with the hepatitis C virus (HCV) core which in turn causes the release of Bax from the Bax/YWHAE complex, resulting in apoptosis (
Lee et al., 2007). In addition, YWHAE plays a role in neuronal development, which indicates that it may process ESC differentiation (
Ikeda et al., 2008). Recent evidence suggests that HMG2L1 may negatively regulate the Wnt signaling pathway, inhibiting β-catenin-stimulated transcriptional activity in mammalian cells (
Yamada et al., 2003), and HMG2L1 can regulate smooth-muscle specific gene expression, stimulate smooth muscle differentiation (
Zhou et al., 2010). Therefore HMG2L1 may act by inhibiting pluripotency maintenance and facilitating differentiation.
In conclusion, genes down-regulated in the H1T hESC line are associated with the cell cycle, apoptosis or differentiation which ultimately results in stimulating ESC differentiation. The combination of both mRNA and CGH microarray provides us a high resolution screen for factors which maintain pluripotency and it is feasible to study the underlying mechanisms of hESCs self-renewal.
Higher Education Press and Springer-Verlag Berlin Heidelberg