Generation of new human embryonic stem cell lines with diploid and triploid karyotypes

Abstract

Human pluripotent embryonic stem cells (hESC) have great promise for research into human developmental biology and the development of cell therapies for the treatment of diseases. To meet the increased demand for characterized hESC lines, we present the derivation and characterization of five hESC lines on mouse embryonic fibroblast cells. Our stem cell lines are characterized by morphology, long-term expansion, and expression profiles of a number of specific markers, including TRA-1-60, TRA-1-81, alkaline phosphatase, connexin 43, OCT-4, NANOG, CXCR4, NODAL, LEFTY2, THY-1, TDGF1, PAX6, FOXD3, SOX2, EPHA2, FGF4, TAL1, AC133 and REX-1. The pluripotency of the cell line was confirmed by spontaneous differentiation under in vitro conditions. Whereas all of the cell lines expressed all the characteristics of undifferentiated pluripotent hESC, two of the cell lines carried a triploid karyotype.

Introduction

The ability to isolate and culture human embryonic stem cells (hESC), which was first reported in 1998, has opened a promising area in medical research (Thomson et al. 1998). Derived from the inner cell mass (ICM) of blastocyst-stage embryos, pluripotent hESC give rise to all cell types in the human body and are thought to be able to proliferate indefinitely in an undifferentiated state. Researchers predict that, if coaxed to differentiate in culture, hESC can be used to create specialized cells to treat a wide range of diseases and conditions, including Parkinson's disease, Alzheimer's disease, cancer, spinal cord injury, and juvenile-onset diabetes. Other envisioned uses of hESC include research to understand cell specialization and the development and testing of new drugs (Keller & Snodgrass 1999; Odorico et al. 2001; Edwards 2004).

So far, the available knowledge of conditions for deriving, characterizing, and culturing undifferentiated hESC is largely based on a relatively a few successfully isolated cell lines (Reubinoff et al. 2000; Lanzendorf et al. 2001; Mitalipova et al. 2003; Pickering et al. 2003; Baharvand et al. 2004; Cowan et al. 2004; Park et al. 2004; Stojkovic et al. 2004; Strelchenko et al. 2004; Lee et al. 2005). Furthermore, the underlying mechanisms that control the developmental decisions of hESC in culture remain essentially unknown. Due to the limited number of available hESC lines, there is an urgent need for the generation and characterization of more cell lines, as each line may have its own characteristics and advantages for different applications. Furthermore, the availability of more hESC lines for comparison will aid in defining criteria for bona fide hESC and the establishment of appropriate and robust methods for maintenance and expansion of hESC. Moreover, to overcome immune rejection of hESC after transplantation an approach proposed is to establish a ‘stem cell bank’ which contains hundreds of thousands of immunophenotyped stem cell lines in order to match the recipient's HLA profile for immunocompetency (Trounson 2002).

Here, we describe the successful establishment of hESC lines from normal and tripronuclei blastocysts. Three of the hESC lines with normal (46,XX and 46,XY) and two with triploid (69,XXY) karyotypes, continue to proliferate and maintain a morphology typical of hESC colonies in long-term culture, and express specific cellular and molecular markers characteristic of undifferentiated hESC lines. The pluripotency of these cell lines was also demonstrated by in vitro differentiation and expression of proteins indicative of cells of all three germ layers.

Materials and methods

Derivation and culture of human embryonic stem cells

Embryos were donated to the study following the approval of the institutional review board and after obtaining informed consent from the couple undergoing in vitro fertilization treatment. The embryos were cultured in sequential media under oil (Vitrolife, Goteborg, Sweden). Following routine assisted reproduction treatment procedures, pronucleate zygotes were transferred to fresh G1 (version 3, Vitrolife) microdrops on day 1, and then into G2 (version 3, Vitrolife) microdrops or co-cultured with Vero cells (green monkey cell line, Pasture institute, Tehran, Iran) late on day 2. Embryos that had been donated for stem cell research were cultured until day 5.5 postfertilization.

The blastocysts were graded at day 5.5 by the scale of Gardner et al. (2000). Briefly, blastocysts were given a numerical score from 1 to 6 on the basis of their degree of expansion and hatching status, as follows: 1, an early blastocyst with a blastocoele that is less than half of the volume of the embryo; 2, a blastocyst with a blastocoele that is half or greater than half of the volume of the embryo; 3, a full blastocyst with a blastocoele completely filling the embryo; 4, an expanded blastocyst with a blastocoele volume larger than that of the early embryo, with a thinning zona; 5, a hatching blastocyst with the trophectoderm (TE) starting to herniate though the zona; and 6, a hatched blastocyst, in which the blastocyst has completely escaped from the zona. For blastocysts graded as 3–6 (i.e. full blastocysts onward), the development of the ICM was assessed as follows: A, tightly packed, many cells; B, loosely grouped, several cells; or C, very few cells. The TE was assessed as follows: A, many cells forming a cohesive epithelium; B, few cells forming a loose epithelium; or C, very few large cells.

Zona pellucida were removed by acidic Tyrod's solution (Hogan et al. 1994). TE were removed by immunosurgery (Solter & Knowles 1975) using antihuman whole serum (1:3, Sigma, Taufkirchen, Germany) and guinea pig complement (1:10, GPLCL 0502; IMVS Veterinary Services Division, Australia) in 50 µL droplets under oil. Cell lines were established and maintained on mitomycin-C (Sigma) mitotically inactivated mouse embryonic fibroblast (MEF) feeder layer (isolated from day 12.5–13.5 post-coitum fetuses of NMRI outbred strain used at 75 000 cell/cm²) in gelatin-coated tissue culture dish (Falcon, Franklin Lakes, NJ, USA), in hESC culture medium: knockout Dulbecco's modified Eagle's medium (DMEM; Gibco BRL, Gaithersburg, MD, USA) supplemented with 20% ES-qualified fetal calf serum (FBS; Gibco BRL), 2 mm L-glutamine (Gibco BRL), 1× minimal essential medium nonessential amino acids (Gibco BRL), 100 U/mL penicillin, 100 µg/mL streptomycin, 0.1 mmβ-mercaptoethanol (Gibco BRL).

During the isolation and early stages of hESC cultivation, the medium was supplemented with human recombinant leukemia inhibitory factor (hLIF) at 1000 units/mL (Chemicon, Temecula, CA, USA), and hLIF was replaced by insulin-transferrin-selenite (Gibco) in the next passages. Ten days after the initial plating, a hESC-like colony was dissociated with a combined approach of mechanical slicing with a pipette, followed by exposure to 10 mg/mL dispase (Gibco). The resulting colonies were grown in 5% CO₂ and 95% humidity, and they were further propagated in clumps of ∼200–500 stem cell-like cells on MEF approximately every 7 days. Colonies were also periodically selected and cryopreserved by open-pulled straw method (Reubinoff et al. 2001) and stored in liquid nitrogen.

Karyotype analysis

Karyotype analyses were carried out on the cells after passages 10–30 and approximately 3–4 days post subculture. hESC colonies were incubated with a final concentration of ∼0.1 µg/mL colcemid (Gibco) for 4 h at 37°C and in 5% CO₂ in air atmosphere. The cells were washed with phosphate-buffered saline (PBS), trypsinized, and neutralized with ES medium. Then, pellets were resuspended and incubated with 0.075 m KCl for 12 min at room temperature. Having been treated with hypotonic solution, the cells were fixed with 3:1 methanol : glacial acetic acid three times and dropped onto precleaned chilled slides. Chromosome spreads were Giemsa banded and photographed. At least 20 metaphase spreads and five banded karyotypes were evaluated for chromosomal rearrangements.

Alkaline phosphatase

Colonies were evaluated for alkaline phosphatase (AP) activity using the AP substrate kit (Sigma) in accordance with the manufacturer's instructions.

Relative reverse transcription–polymerase chain reaction analysis

Semiquantitative reverse transcription–polymerase chain reaction (RT–PCR) was performed to assess the expression of a set of genes that might produce pluripotency in embryonic and adult stem cells as shown in Table 1. After preparation for routine passage of undifferentiated colonies, several colony pieces were washed through PBS. Total RNA was collected from undifferentiated pooled pieces of each line using Nucleospin RNA II kit (Macherey-Nagel, Düren, Germany). Before reverse transcription, RNA samples were digested with DNase I (Fermentas, Ontario, Canada) to remove contaminating genomic DNA. Standard RT reactions were performed with 2 µg total RNA using oligo (dT)₁₈ as primer and RevertAid H Minus First Strand cDNA Synthesis Kit (Fermentas) according to the manufacture's instructions. Reaction mixtures for PCR included 2.5 µL cDNA, 1× PCR buffer (AMS), 200 µm dNTP, 0.5 µm of each primer pair and 1 unit Taq DNA polymerase (Fermentas). The sequence, expected fragment size, and annealing temperature of primers are listed in Table 1. Polymerase chain reactions were accomplished on a Mastercycler gradient machine (Eppendorf, Hamburg, Germany). Amplification conditions were as follows: initial denaturation at 94°C for 5 min followed by 30 cycles (for β-actin, 20 cycles) of denaturation at 94°C for 45 s, annealing at temperatures which are listed in Table 2 for 45 s, extension at 72°C for 30 s, and a final polymerization at 72°C for 10 min. Each PCR was performed under linear conditions and β-actin was used as an internal standard. Products were electrophoresed on 1.5% agarose gel. The gels were stained with ethidium bromide (10 µg/mL) and photographed on an ultraviolet transilluminator (UVIdoc; Uvitec, Cambridge, UK). The gel images were analyzed with the UVI bandmap program (Uvitec). Semiquantitative RT–PCR values were presented as a ratio of the specified gene's signal divided by the β-actin signal. RT–PCR signals were averaged from four separate experiments.

Table 1.  Primers used for reverse transcription–polymerase chain reaction analysis of stemness gene expression of established human embryonic stem cell lines

Genes	Primer sequences (5′−3′)	Size (bp)	Annealing temperature (°C)	GenBank accession number
NANOG	CAG AAG GCC TCA GCA CCT AC	298	62	NM_024865
	GTC ACT GGC AGG AGA ATT TGG
CXCR4/HM89	GG TCA TGG GTT ACC AGA AGA	365	60	NM_003467
	GTC ATC TGC CTC ACT GAC GTT G
NODAL	GCG TAC ATG CTG AGC CTC TA	312	58	NM_018055
	GGT GAC CTG GGA CAA AGT G
LEFTY2/EBAF	GTT CAG CCA GAG CTT CCG AG	382	66	NM_003240
	CTC TGC ACC GAC ACC TGT AGC
THY1/CD90	CAT CGC TCT CCT GCT AAC AG	386	62	NM_006288
	GCT GAT GCC CTC ACA CTT G
TDGF1/CRIPTO	CGC TTC TCT TAC AGT GTG ATT TG	420	62	NM_003212
	GGT AGA AAT GCC TGA GGA AAG
ALP/GCAP	CCT AAA AGG GCA GAA GAA GGA C	445	65	NM_031313
	TCC ACC TAG GAT CAC ATC AAT G
PAX6/WAGR	TCC ATC AGT TCC AAC GGA GAA G	337	62	NM_001604
	GTG GAA TTG GTT GGT AGA CAC TG
FOXD3/GENESIS	GTG AAG CCG CCT TAC TCG TAC	353	62	NM_012183
	CCG AAG CTC TGC ATC ATG AG
REX1/ZFP42	GAC TGT TAC ATA GAA TGC GTC ATA AG	515	62	NM_174900
	CCA GTA TGA ACC AGG AAA TGT CAC
SOX2/ANOP3	GGC AGC TAC AGC ATG ATG CAG	396	65	NM_003106
	GCT CTG GTA GTG CTG GGA CAT G
EPHA2/ECK	TGC AGA ACA TCA TGA ATG ACA T	400	60	NM_004431
	AG GCA CCG ATA TCC TGG AAG
FGF4/HST	CAA CGT GGG CAT CGG CTT C	346	65	NM_002007
	GAG GAA GTG GGT GAC CTT CAT G
TAL1/SCL/TAL	CTC GGC AGC GGG TTC TTT G	411	65	NM_003189
	CGT CTT GCA GGA GGT CAT CTG
AC133/PROML1	GAA CAG TAT CAA TTC AGT GCT AG	514	60	NM_006017
	GGA AGA CGC TGA GTT ACA TTG
OCT4/POU5F1	CTT GCT GCA GAA GTG GGT GGA GGA A	187	67	NM_002701
	CTG CAG TGT GGG TTT CGG GCA
β-Actin	TGC GTG ACA TTA AGG AGA AG	213	60	NM_001101
	TGA AGG TAG TTT CGT GGA TG

Table 2.  Primers used for reverse transcription–polymerase chain reaction analysis of differentiation of established human embryonic stem cell lines

Genes	Primer sequences (5′−3′)	Annealing temperature (°C)	Size (bp)	GenBank accession number
αFP (α-Fetoprotein)	GCA GCC AAA GTG AAG AGG GAA GA	69	216	NM_001134
	GTC ATA GCG AGC AGC CCA AAG AAG
CK18 (Cytokeratin 18)	CCA TGC GCC AGT CTG TGG AG	65	322	X12881
	GTG GTG CTC TCC TCA ATC TGC T
CK19 (Cytokeratin 19)	TGA GGT CAT GGC CGA GCA GAA C	69	331	NM_002276
	CAT GAG CCG CTG GTA CTC CTG A
TTR (Transthyretin)	GGT GAA TCC AAG TGT CCT CTG AT	61	352	NM_000371
	GTG ACG ACA GCC GTG GTG GAA
AAT (α-1-Antitrypsin)	CCA TGT TTG TCA AAG AGC AAC T	61	345	NM_001085
	GGA AGT AAG GTA TAG TCA GGT GAT
CK8 (Cytokeratin 8)	CAG ATC AAG TAT GAG GAG CTG CA	63	503	NM_002273
	AGC TGG TGC GGC TGA AGG AT
G6P (Glucose 6 phosphatase)	GCT GAA TGT CTG TCT GTC ACG AA	60	494	NM_000151
	GCA GAA GGA CAA GAC GTA GAA GA
TAT (Tyrosine aminotransferase)	GCT AAG GAC GTC ATT CTG ACA AG	67	354	NM_000353
	GTC TCC ATA GAT CTC ATC AGC TAA G
ALB (Albumin)	CTG CTT GAA TGT GCT GAT GAC AG	60	365	NM_000477
	GGC ATA GCA TTC ATG AGG ATC TG
Glut-1 (Glucose transferase 1)	CCA CGA GCA TCT TCG AGA A	55	374	AY034633
	GCA CAT GCC CAC AAT GAA A
Glut-2 (Glucose transferase 2)	GGT TTG TAA CTT ATG CCT AAG	60	213	L09674
	GCC TAG TTA TGC ATT GCA G
Insulin	AGC CTT TGT GAA CCA ACA CC	60	245	NM_000207
	GCT GGT AGA GGG AGC AGA TG
Glucagon	CCA GAT CAT TCT CAG CTT CC	54	180	V01515
	GGC AAT GTT ATT CCT GTT CC
Nkx6.1	GTT CCT CCT CCT CCT CTT CCT C	58	381	NM_006168
	AAG ATC TGC TGT CCG GAA AAA G
PDX1	GGA TGA AGT CTA CCA AAG CTC AC	64	230	U35632
	CCA GAT CTT GAT GTG TCT CTC G
α-Cardiac actin	GGA GTT ATG GTG GGT ATG GGT C	65	486	BC009978
	AGT GGT GAC AAA GGA GTA GCC A

Grading scale of human embryonic stem cells

Colonies of hESC were evaluated 7 days after passaging when they should be 1.5–2 mm in diameter. The colonies were assessed using 10× magnification of an inverted microscope (Olympus, Tokyo, Japan). The quality of colonies was graded according to ES Cell International: (1) Grade A/excellent; colonies with even morphology and well defined edges. The cells in colonies were dense and not distinguish individual cells easily. The colonies were thick and multilayer, homogenous and exhibit 0–30% differentiation. The differentiated cells were migrating and passing from peripheral of colonies. (2) Grade B/good; 30–50% of peripheral area was differentiated. (3) Grade C/fair; colonies exhibit more than 50–80% differentiation. (4) Grade D/poor; colonies differentiated more than 80–100% which exhibit well differentiated morphology with inhomogeneous levels.

Assessment of growth of different human embryonic stem cells lines

The growth of hES cells on the MEF feeder layer was examined from day 2 to day 5 after plating the hESC clumps as described by Park et al. (2004). Cell doubling of several selected colonies (n = 10) was evaluated daily by counting along a major axis under the same magnification (10×) with a phase contrast inverted microscope (Olympus).

In vitro differentiation

To prove that established hESC are pluripotent, it is necessary to demonstrate their ability to contribute to all three germ layers, endoderm, ectoderm and mesoderm. This is normally achieved by the formation of embryoid bodies (EB) in vitro followed by histochemistry and morphological assessment. The hESC were cultured in the absence of MEF feeder layers in suspension in bacterial dishes to generate EB for 5 or 20 days, and subsequently grown in tissue culture plates for several days in hESC medium. The beatings of differentiating cardiomyocytes was observed under an inverted microscope. Differentiation of hESC into endodermal cells was evaluated by RT–PCR with primer sets shown in Table 2 as previously described.

Fluorescent immunostaining

The cells were rinsed twice with PBS, fixed with methanol : acetone (3:1) at −20°C, or 4% paraformaldehyde at room temperature, and incubated with primary antibody for 60 min at 37°C in a humid chamber. At the end of the incubation time, they were rinsed three times with PBS and incubated with the fluorescence isothiocyanate (FITC)-conjugated antimouse IgG (1:100; Sigma) for 60 min at 37°C. After rinsing with PBS, the cells were analyzed under fluorescent microscope (Nikon, Tokyo, Japan).

Colonies were evaluated by anti-connexin 43 (1:200; Sigma), anti-TRA-1-60 (1:20) and anti-TRA-1-81 (1:20; gifts from Peter Andrews, University of Sheffield, UK), and anti-Oct-4 (1:25; R&D systems, Minneapolis, MN, USA).

Neurons were detected by anti-microtubule-associated protein (MAP2; 1:200; Sigma), and neurofilament protein (1:50; Sigma), synaptophysin (1:250; Sigma), neuron-specific tubulin-III (1:250; Sigma). Glial fibrillary acidic protein (Sigma) used in accordance with manufacturer's protocol. Cardiomyocytes were recognized by anti-α-actinin (1:800; Sigma).

Evaluation of human chorionic gonadotropin and α-fetoprotein production

Colonies at passages 10–15 were allowed to grow on mitotically inactivated MEF to confluency (approximately 6 weeks). The medium was replaced every day after 2 weeks of differentiation, and medium in triplicate wells, conditioned for 24 h, was assayed by ELISA for α-fetoprotein and hCG, markers of endoderm lineage and trophoblast differentiation, respectively. Conditioned medium of MEF was used as control group.

Statistics analysis

Grading of hESC lines were compared using an anova test by SPSS software (SPSS, Chicago, IL, USA). The RT–PCR ratio values were analyzed using GLM frequency and correlation procedures. Results are expressed as the mean ± SEM, and P < 0.05 was considered to be statistically significant.

Results

Human embryonic stem cells line characterization

Characterization of each of the derived cell lines was conducted according to the following parameters.

Morphology.  Each of the lines continued to proliferate and maintained a morphology typical of hESC colonies grown on MEF, that is, round colonies with sharp edges, in which the spaces between cells are clear (Fig. 1a). Single cell morphology showed a high nucleus to cytoplasm ratio with the presence of at least two nucleoli. The lines were named Royan H2 to Royan H6 (‘Royan’ in Persian means ‘embryo’). When colonies of small compact cells were observed 7 days after seeding, they were manually dissociated into clumps of ∼200–500 cells and replated on fresh feeder layers. The cell lines were grown for several passages in vitro, and the cell lines still consisted mainly of cells with the morphology of undifferentiated ES cells (Fig. 1a). The cell lines were isolated from embryos graded in Table 3. The cell lines were successfully frozen and thawed.

Figure 1. Morphology of colonies of five establishment human embryonic stem cell lines (a) and the percent of different qualities of the colonies based on their morphology (b).

Table 3.  Embryo culture, quality of blastocysts and establishment of human embryonic stem cell (ESC) lines along with their karyotype and doubling time

Treatment procedure	Zygote (PN)	Embryo grade (blastocyst, ICM, TE)	Blastocyst culture method	Cell line	Karyotype	Doubling time (h)
—, no hESC line produced. ICM, inner cell mass; ICSI, intracytoplasmic sperm injection; IVF, in vitro fertiliziation; PN, pronuclei; TE, trophectoderm.
IVF	2PN	4BA	Sequential media	Royan H2	46,XX	31.5 ± 0.8
IVF	3PN	4AA	Sequential media	Royan H3	69,XXY	23.8 ± 0.7
IVF	2PN	4BB	Sequential media	—	—	—
ICSI	2PN	4AB	Sequential media and co-culture with Vero cells	Royan H4	Mosaic 69,XXY	23.2 ± 0.4
ICSI	2PN	4AA	Sequential media and co-culture with Vero cells	Royan H5	46,XX	32.9 ± 0.6
ICSI	2PN	4AA	Sequential media	Royan H6	46,XY	32.1 ± 0.8

Comparison of established human embryonic stem cells lines by morphology.  Morphological evaluation was carried out on the colonies during passages 5–15 and 7 days post subculture. The number of evaluated colonies were 1211, 1222, 679, 763, and 701 in Royan H2 to Royan H6, respectively. In accordance with morphology grading, Royan H5 and Royan H6 had better quality (A and B grade) than Royan H2 and/or other triploid hESC lines (P < 0.05) and Royan H4 was better than other triploid line (Royan H3; P < 0.05; Fig. 1b).

Cell markers.  Whereas each of the putative hESC lines expressed cell surface markers that characterize undifferentiated non-human primate and hESC, including AP, TRA-1-60, and TRA-1-81 (Fig. 2), they were all negative for SSEA-1 immunoreactivity (not shown). Cell–cell communication was examined by assessing the presence of gap junctions. The cells appear tightly adhered to each other, and we have demonstrated that the cells have gap junctions by the presence of positive immunoreactivity to connexin 43 cells (Fig. 2). Furthermore, all five lines were positively labeled with antibody against Oct-4 (Fig. 2).

Figure 2. Characterization of human embryonic stem cell (hESC) lines. Photomicrographs show alkaline phosphatase (AP) and the expression of Tra-1-60, Tra-1-81, Oct-4, and connexin in the five hESC lines.

Karyotyping.  Three of the five cell lines, Royan H2, Royan H5 and Royan H6, have been cytogenetically analyzed and had a normal karyotype (Royan H2 and Royan H5, 46,XX; and Royan H6, 46,XY; Fig. 3), whereas Royan H3 had a triploid (69,XXY) karyotype (Fig. 3). However, Royan H4 had numerical aberration from a triploid karyotype in the cells (13.5%, 53%, 37.5%, 6%, and 3%, for 69, 68, 67, 65 and 70 chromosomes, respectively).

Figure 3. Karyotype analysis of human embryonic stem cell lines using the G-band method showing a normal 46,XX karyotype for Royan H2 and Royan H5, a 46,XY karyotype for Royan H6, a 69,XXY karyotype for H3, and a triploid karyotype with a numerical aberration (67,XXY) for Royan H4.

Genetic analysis.  In order to assess the biological relevance of the variable gene expression found in the five hESC lines, we focused more closely on the expression profiles of selected genes implicated in other stem cell systems and on the expression of a set of genes that might function in regulating self renewal and pluripotency in embryonic and adult stem cells, according to published reports (Ivanova et al. 2002; Ramalho-Santos et al. 2002; Brivanlou et al. 2003) by RT–PCR. We found that a number of genes important to other stem cell populations, including POU5F1 (OCT4), SOX2, ZFP42 (REX1), TDGF1 (CRIPTO), NODAL, EBAF (LEFTY), EPHA2, THY1, PROML1 (AC133), NANOG, PAX6, FOXD3, FGF4, CXCR4 and TAL1 (SCL/TAL), were expressed at detectable levels in all five of the undifferentiated hESC lines that we examined, except TAL-1 that was not expressed in Royan H3 (Fig. 4A). Semiquantitative RT–PCR showed that OCT4, REX1, CRIPTO, NODAL, EPHA2, AC133, NANOG, and ALP were approximately equally expressed in all hESC lines. TAL-1, FGF-4, and FOXD3 showed a higher level of expression in H5 than H2 (diploid lines) and LEFTY2, SOX2 and FGF-4 were more expressed in Royan H5 versus Royan H6. In contrast THY1 and FOXD3 showed lower expression in Royan H2 and Royan H5 versus Royan H6. Moreover, FGF-4, PAX6, and CXCR4 were expressed at higher levels in H4 (mosaic triploid line) than diploid line H3 (P < 0.05; Fig. 4B).

Figure 4. (A) reverse transcription–polymerase chain reaction (RT–PCR) analysis of the pluripotency-related genes of established lines. Total RNA was extracted from Royan H2, Royan H3, Royan H4, Royan H5 colonies. RT⁻ indicates no cDNA. (A) PCR was performed for 30 cycles for all of genes, except β-actin which was used as an internal standard and amplifed for 20 cycles. (B) Semiquantitative RT–PCR analysis. Gene transcript quantity was measured by relative RT–PCR using the internal standard β-actin. The figure represents data from four separate experiments. For all of presented genes P < 0.05.

Doubling time.  The growth range of hESC on the MEF feeder layer, which, determined by counting along a major axis under the same magnification (Fig. 5), was 2.0–2.5, 1.7–2.3, 1.9–2.3, 2.0–2.4, and 2.2–2.5 doublings over 3 days for Royan H2 to Royan H6, respectively (Table 3). Thus, doubling time of hESC lines was 31.5 ± 0.8 h for Royan H2, 23.8 ± 0.7 h for Royan H3, 23.2 ± 0.4 h for Royan H4, 32.9 h ± 0.6 h for Royan H5, and 32.1 ± 0.8 h for Royan H6 (e.g. 72 h/2.3 = 31.3 h), which is similar to the doubling time observed with the more commonly used fresh MEF for the culture of hESC.

Figure 5. Phase contrast imaging of a cultured human embryonic stem cell (hESC) on a mitotically inactivated mouse embryonic fibroblast feeder layer. (A) Mechanically dissociated human embryonic stem cell clumps subcultured and stably positioned on the feeder layer on day 2 after plating. (B–D) A gradually growing hESC colony. Growing speed is indicated as doubling over the dotted line from day 2 to day 5 (B–D) after plating of hESC clumps. Bars, 500 µm.

In vitro differentiation.  Similar to other published hESC lines, once removed from its feeder layer and cultured in suspension, all of the Royan cell lines formed EB, including cystic ones. Stem cells within these EB differentiated into various cell types. The developmental potential of lines was examined by expression and staining with antibodies of representatives of the three germ layers: ectoderm (Fig. 6-I) (NF, GFAP, synaptophysin, MAP-2 and β-tubulin III), mesoderm (cardiac actin and α-actinin (Fig. 4)), and endoderm (Fig. 6-II) Langerhans islet related genes ((insulin, glucagon, Glut-1, Glut-2, somatostatin, Pax4, glucokinase, Pdx-1, Nkx6.1) and hepatocyte related genes (α-FP, CK18, CK1, TTR, AAT, CK8, G6P, TAT, Albumin)). After each of the cell lines were allowed to differentiate for 5–6 weeks, both α-fetoprotein (>1000 IU/mL) and hCG (>35 mIU/mL) were detected in conditioned cultured medium, indicating endoderm and trophoblast differentiation (Thomson et al. 1998). No hCG or α-fetoprotein was detected in conditioned medium from MEF. In summary, all five of the Royan hESC lines were able to spontaneously form derivatives of the three embryonic germ layers under in vitro conditions. However, theses genes, except CK19, are not expressed in hESC lines.

Figure 6. Spontaneous differentiation of established human embryonic stem cell (hESC) lines into different cell lineages. (I) Ectodermal (neuronal) cells. Neurons assessed by phase contrast microscopy (A) and for MAP2 (B), neurofilament (C), synaptophysin (D), β-tubulin III (D), and glial cells for GFAP (F) immunoreactivity. (II) Analysis of endodermal (pancreatic and hepatic) and mesodermal (cardiomyocytes) cells derived from spontaneous-differentiated hESC by reverse transcription–polymerase chain reaction (RT–PCR) and immunocytochemistry with anti-α-actinin. RT⁻ (containing no cDNA) PCR were used as a negative control. RH, Royan H; SC, embryonic stem cell; SD, spontaneous differentiation.

Discussion

In this study, we report the establishment of five new hESC lines using immunosurgery. Characterization of the hESC lines revealed that all had the common features of hESC based on morphological assessment, extended culturing, expression of cell surface carbohydrate epitopes (TRA-1-60, and TRA-1-81 AP), expression of characteristic hESC genes and the ability to differentiate into all three germ layers, even though two cell lines, Royan H3 and Royan H4, were karyotypically abnormal (69,XXY and 69,XXY with chromosomal mosaicism, respectively). Such chromosomal abnormality is not surprising, taking into account that Hardarson et al. (2003) recently found that only 42% of surplus assisted reproductive technology embryos were chromosomally normal at the blastocyst stage. Heins et al. (2004) reported a triploid hESC line as well. Of tripronuclei zygotes, 10 ± 30% can develop into blastocysts. Therefore, hESC derived from zygotes and/or blastocysts with genetic defects could provide new opportunities for developing pharmaceuticals and toxicological screening technologies. Recently, Verlinsky et al. (2005) also reported derivation of several hESC lines with genetic anomalies. Moreover, our results indicate that the doubling time of diploid lines was ∼32 h for diploid lines which is similar to the doubling time observed with the more commonly used fresh MEF for the culture of hESC. However, the doubling time of triploid lines was lower (∼23 h).

We also tested the expression of another cell surface marker reported to be present on blastocysts or other stem cell populations, such as gap junction proteins connexin-43 (Duval et al. 2002). Genetic analysis demonstrated that the putative hESC lines expressed a number of genes shown to be specifically expressed within the preimplantation embryo/ICM of the developing blastocyst, adult stem cells or in pluripotent hESC lines by genetic profiling of different stem cell populations (Ramalho-Santos et al. 2002). This included expression of OCT-4, a POU transcription factor required for the generation of pluripotent hESC from the ICM (Nichols et al. 1998; Niwa et al. 2000), Rex-1, an acidic zinc finger transcription factor shown to be a specific regulatory target of Oct-4 in mouse ESC (Ben-Shushan et al. 1998), and FGF-4, a fibroblast growth factor present within the ICM (Rappolee et al. 1994) and another transcriptional regulatory target of Oct-4 in ES cells (Yuan et al. 1995). SOX-2, and Cripto/TDGF1 were also expressed in all the Royan hESC lines. These genes are expressed in the ICM and trophoblast cells of the mouse blastocyst, mouse ES cells, and in human and mouse embryonic carcinoma cells (reviewed in Adamson et al. 2002). Another surprising result was expression of FOXD3 in the hESC lines, while Ginis et al. (2004) reported that the H1 hESC line shows an absence of FOXD3 expression. However it should be noted that this might be line or growth conditions-specific since some reports (reviewed in Carpenter et al. 2003) have suggested expression of FOXD3 in some hESC lines. In contrast to human ES cells, FOXD3 is readily detected in mouse ES cell cultures (Ginis et al. 2004). FOXD3 or GENESIS is expressed early in mouse embryonic development, and abrogation of FOXD3 function leads to a failure of the blastocyst to develop beyond the 4-cell stage (Hanna et al. 2002). In addition, FOXD3 is critical for endodermal differentiation when it antagonizes the activity of OCT-4 (Guo et al. 2002). The putative hESC lines generated in the laboratory also expresses Nanog, a homeodomain protein obligatory for maintaining pluripotency in mouse ES cells (Chambers et al. 2003; Mitsui et al. 2003). Therefore, these cell lines generated many of the molecular and cellular characteristics of a hESC population. We also noted that CXCR4, LEFTY 2, NODAL, THY-1, ALP, PAX6, EPHA2, TAL1, AC133 were expressed in all lines while TAL1 was not expressed in triploid Royan H3. However, we observed differences in the expression of a subset of these genes (TAL-1, FGF-4 and FOXD3, LEFTY2, SOX2, THY1) between diploid lines (Royan H2, Royan H5, and Royan H6). Moreover, the genes FGF-4, PAX6 and CXCR4 were expressed at higher levels in the triploid mosaic line (Royan H4) in comparison to the triploid Royan H3, while TAL-1 was absent in Royan H3. Based on these data, a correlation seems to be observed between stemness genetic profile and quality of morphology and/or source of new of the hESC lines.

These data indicate that the variability in gene expression in hESC lines extends to several potentially important stem cell genes. Although these hESC lines are pluripotent, these results suggest a likelihood of a functional significance associated with the unique gene expression signatures of independently derived hESC lines. Given the considerations noted above, both the shared and unique gene expression signatures that we uncovered among the five independent hESC lines have important implications for further studies.

It seems that an important step in the establishment of a hESC line is to culture the human embryo until it forms a good quality blastocyst. It has been reported that hESC lines may be derived more efficiently from frozen blastocysts than from frozen cleavage-stage embryos, since the poor quality of many of the latter prevents their development and division after thawing (Cowan et al. 2004). Therefore, securing high-quality embryos and improving embryo culture conditions are important for improving the efficiency of establishment of hESC. In this study, culturing fresh pronuclei-stage human embryos in sequential G1.3/2.3 media and/or in combination with Vero cell co-culture to the blastocyst stage yielded high-quality blastocysts. Although the limited number of embryos and stemness gene expression involved makes it difficult to compare the efficiency of the quality of the blastocyst, and/or the morphology of hESC colonies, it is clear that the quality of blastocyst is a major factor in the successful establishment of a hESC line. However, the quality and quantity of stemness gene expression are important factors for the morphology of the hESC lines (see Material and methods for quality of hESC colonies based on morphology), while there are numerous lines derived from blastocysts of poor quality (Mitalipova et al. 2003).

In summary, the human embryo-derived cell lines described here meet standard criteria used to define hESC. These include expression of markers commonly used to identify hESC, normal and stable karyotype, and demonstrated ability to differentiate in vitro into a variety of cell types. Interestingly, Royan H3 and Royan H4 which are triploid cell lines can differentiate in vitro to a variety of cell types. Continued characterization of the transcriptional profile of other hESC along with functional testing of the candidate genes will be extremely valuable for deciphering mechanisms related to pluripotency and also for developing directed differentiation strategies for producing unique cells of interest. The hESC lines with chromosomal aberration may be used as a model for in vitro studies.

Acknowledgements

We wish to thank all our colleagues for the assistance of the staff in the Department of Stem Cells, especially Mohammad Pakzad and Sara Soudi. We gratefully acknowledge the support of Dr Abdoulhossein Shahverdi and Dr Ahmad Vosough, Dr Hamid Gourabi from the Department of Genetics at the Royan Institute for karyotyping, and Dr Stephen L. Minger (King's College London, UK) and Dr Alice Pebay (Monash Institute of Medical Research, Australia) for critically reading the manuscript.