SEARCH

SEARCH BY CITATION

Keywords:

  • Human embryonic stem cells;
  • Serum replacement medium;
  • Foreskin fibroblasts;
  • Characterization;
  • Pluripotency

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Derivation and culture of human embryonic stem cells (hESCs) without animal-derived material would be optimal for cell transplantation. We derived two new hES (HS293 and HS306) and 10 early cell lines using serum replacement (SR) medium instead of conventional fetal calf serum and human foreskin fibroblasts as feeder cells. Line HS293 has been in continuous culture, with a passage time of 5–8 days, since October 2003 and is at passage level 56. Line HS306 has been cultured since February 2004, now at passage 41. The lines express markers of pluripotent hESCs (Oct-4, SSEA-4, TRA-1-60, TRA-1-81, GCTM-2, and alkaline phosphatase). The pluripotency has been shown in embryoid bodies in vitro, and the pluripotency of line 293 has also been shown in vivo by teratoma formation in severe combined immunodeficiency/beige mice. The karyotype of HS293 is 46,XY, and that of HS306 is 46,XX. Ten more early lines have been derived under similar conditions since September 2004. We conclude that hESC lines can be successfully derived using SR medium and postnatal human fibroblasts as feeder cells. This is a step toward xeno-free conditions and facilitates the use of these cells in transplantation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Derivation and culture of human embryonic stem cells (hESCs) from the inner cell mass (ICM) of blastocysts fertilized in vitro and establishment of permanent lines were originally carried out using media that contained fetal calf serum (FCS) and fetal mouse fibroblasts as feeder cells [1, 2]. The potential of transplanted hESCs in cases of severe degenerative diseases has been clearly recognized [35].

Animal-derived components, nonhuman sera, and feeder cells in the cultures bear a risk of transmitting animal pathogens to hESCs. Such materials are therefore not desirable in cell lines in regard to cell transplantation in humans.

Existing hESC lines, originally derived using fetal mouse fibroblasts and FCS, have been successfully cultured as nondifferentiated ESCs without feeder cells on Matrigel, using conditioned media from cultures of mouse feeder cells [6, 7]. These cultures were a step forward, although they were not entirely free of animal material. Human fetal fibroblasts and cells from fallopian tubes of women, obtained at surgery, have also enabled nondifferentiated growth of already existing hESCs, and the use of human serum in cultures has also been reported, with one line also derived in such conditions [8]. Recently, derivation of an hESC line on fetal mouse feeder cells but using serum replacement (SR) was reported [9]. We have derived and propagated new hESC lines using human foreskin fibroblasts as feeder cells [10].

Amit et al. [11] described serum-free culture conditions for a line that had originally been derived using the conventional FCS-mouse fibroblast system. They used SR medium and postnatal human fibroblasts as feeder cells. The same group also described a system involving culture of hESCs on fibronectin and adding transforming growth factor β1, basic fibroblast growth factor (bFGF), and leukemia inhibitory factor to the culture medium [12]. However, there are many differentiated cells at the margins of their colonies, as shown by their pictures. In these experiments, they used hESC lines that had been originally derived on mouse fibroblasts and in medium containing FCS.

We have also used SR medium for the culture of hESC lines that we originally derived in FCS-containing medium but with human foreskin fibroblasts as feeder cells [10, 13, 14]. In a systematic survey, we found that adding bFGF at 8 ng/ml to culture medium containing 20% SR medium gave the best support for nondifferentiated growth of these hESCs [15].

We now describe fully characterized hESC lines (HS293 and HS306) that had been derived from the beginning using SR medium in the culture and human foreskin fibroblasts as feeder cells.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The Ethics Committee of the Karolinska Institute approved the derivation, characterization, and early differentiation of hESC lines from donated blastocysts that could not be used in infertility treatment. Blastocysts were obtained as donations from infertile couples undergoing in vitro fertilization treatment at our fertility unit. Both partners signed an informed consent form after receiving oral and written information [16]. Only embryos that could not be used in infertility treatment were used in stem cell line derivation.

The embryos were donated on day 2 after fertilization after all embryos with good quality (a minimum score of 2.0 out of 3.5) had been transferred or frozen. The scoring was performed according to Mohr and Trounson [17]. From the original score of 3.5, 0.5 was reduced at a time based on the following features: greater than 20% of cellular fragmentation, unequal size of the blastomeres, multinuclear blastomeres, or the embryo did not fill the zona. During the period of the derivation of the lines 293 and 306, 10 blastocysts were received. Three had been cultured from frozen and seven from fresh embryos.

The blastocysts used for the present lines had been cultured in medium designed for blastocyst culture (MediCult, Ronnehavn, Denmark). They were transferred to the research laboratory on day 6 after fertilization. The blastocyst that gave origin to line HS293 was an expanded blastocyst with a moderately large ICM (7,876 μm) and a moderate number of cells in the trophectoderm. The blastocysts for line HS306 had not expanded, the ICM was moderately large (6,041 μm), and there was a medium number of cells in the trophectoderm. Both were fresh embryos. Recently, from September to October 2004, 27 more blastocysts were obtained, three of them from frozen-thawed embryos. Ten attached to the feeder cells and began to grow out and have formed early lines being now at passage levels three through eight. One of the new early lines (HS346) was derived from a blastocyst that had been cryopreserved for 5 years.

Derivation of the present lines was carried out in principle as described previously [10, 18]. Separation of the ICM from the trophectoderm cells was carried out by first removing the zona pellucida using 0.5% Pronase (Sigma, St. Louis). The trophectoderm was removed by immunosurgery as described earlier by Solter and Knowles [19] using rabbit antihuman whole serum (Sigma) and guinea pig complement serum (Sigma). We placed two blastocysts on the feeder layer after removing the zona pellucida only [20], but these ICMs disappeared within a few days from the plate.

As feeder cells, human foreskin fibroblasts (CRL-2429; ATCC, Manassas, VA) were used. They were mitotically inactivated using irradiation (35 Gy) and plated onto 2.84-cm2 dishes to form a confluent monolayer to be used as substrate cells on the following day. For derivation of a new line, 150,000 fibroblasts were plated, and for passaging later on, 350,000 fibroblasts were plated. The medium used in culture of these feeder cells was Iscove's medium (Gibco Invitrogen Corporation, Paisley, Scotland) supplemented (10%) with FCS (Gibco Invitrogen Corporation).

The ICM was transferred to plastic dishes tested for use in embryo culture (Falcon; Becton, Dickinson, Franklin Lakes, NJ) onto the feeder cell layer.

The culture medium used for derivation and culture of the hESCs consisted of Knockout Dulbecco's modified Eagle's medium (Gibco Invitrogen Corporation) supplemented (20%) with SR medium (Knockout SR, Gibco Invitrogen Corporation), 2 mM L-glutamine (Gibco Invitrogen Corporation), 1% penicillin-streptomycin (Gibco Invitrogen Corporation), 1% nonessential amino acids (Gibco Invitrogen Corporation), 0.5 mM 2-mer-captoethanol, 1% insulin-selenium-transferrin (Sigma-Aldrich Co.), and bFGF (8 ng/ml; R&H Systems, Oxon, U.K.).

After an initial growth period of 12 days (Fig. 1), the cell aggregates were removed mechanically from the original plate and transferred to fresh feeder cells. Mechanical passaging was performed by cutting the colony, which is approximately 2,000 μm in diameter, to eight pieces using a scalpel under the stereo microscope. Mechanical splitting was then carried out at 5- to 8-day intervals (mean, 7 days). Nondifferentiated cells, as judged by morphology, were chosen for each further passage. The doubling time of the hESCs was approximately 24 hours.

thumbnail image

Figure Figure 1.. A colony of human embryonic stem cells growing from the inner cell mass of a blastocyst (line HS351) 12 days after derivation, immediately before the first mechanical splitting. Magnification ×200.

Download figure to PowerPoint

For cryopreservation of the cells, vitrification in pulled open straws, using ethylene glycol, dimethylsulphoxide (20% each), and 1 M sucrose as cryoprotectants, was carried out as described by Reubinoff et al. [21].

The growing hESCs were characterized immunohistochemically using antibodies against markers of undifferentiated ESCs and a marker of differentiated hESCs (SSEA-1). The primary antibodies were specific for TRA-1-60, TRA-1-81, SSEA-4, SSEA-1 (Chemicon, Temecula, CA), Oct-4 (Santa Cruz Biotechnology Inc., Santa Cruz, CA), and GCTM-2 (kindly provided by Martin Pera, Monash University, Clayton, Australia). The TERA-2 cell line and human foreskin fibroblasts were used as controls. Nonimmune serum was used as isotype control. The expression of alkaline phosphatase was shown using a Vector Blue/Red substrate kit (Vector Laboratories, Burlingame, CA).

In vivo pluripotency was tested as previously described [10, 14]. In brief, exponentially growing HS293 cells from passage 16 were harvested from the culture plate using dispase plus mechanical treatment. Five colonies (103 to 104 hESCs) were washed twice in phosphate-buffered saline and subsequently implanted beneath the testicular capsule of a young (6-week) severe combined immunodeficiency (SCID)/beige male mouse (C.B.-17/GbmsTac-scid-bgDF N7, M&B, Ry, Denmark). Teratoma growth was determined by palpation every week, and the mice were killed (cervical dislocation) 11 weeks after implantation. The teratoma was fixed, and sections were stained with hematoxylin and eosin. The presence of tissue components of all three embryonic germ cell layers was shown, as analyzed from the stained sections.

The pluripotency of both lines was also shown in vitro using reverse transcription–polymerase chain reaction (RT-PCR). Embryoid bodies were formed from hESCs growing in 20% SR medium by aggregation and replating of cells in 25-μl drops of 20% SR medium without bFGF. Arrays were plated as hanging drops as described by Mountford et al. [22]. Embryoid bodies were moved to fresh medium every second day and cultured for 1 month. Total RNA for RT-PCR analyses was extracted using RNeasy mini kits (Qiagen, Valencia, CA). In addition, RNA from embryoid bodies was Dnase-treated (Ambion, Austin, TX) to avoid DNA contamination in RT-PCR. Complementary DNA was synthesized from 50 ng of total RNA using Sensiscript reverse transcriptase, and PCR was performed using HotStarTaq DNA polymerase (Qiagen). The expression of tissue component markers characteristic of ectodermal (ND-1), mesodermal (alpha-cardiac actin), and endodermal (alpha-fetoprotein [AFP]) development in embryoid bodies was determined using RT-PCR primers and conditions reported by Amit et al. [11] and Henderson et al. [23], with GAPDH as the housekeeping control. The presence of markers of the three embryonic germ cell layers was also shown immunohistochemically. Antibodies against Nestin (DSHB, Iowa/Chemicon, Temecula, CA), bone morphogenetic protein-4 (BMP-4) (Novocastra Laboratories Ltd., Newcastle upon Tyne, U.K.), and AFP (Sigma), representing ectoderm, mesoderm, and endoderm, respectively, were used for this.

Karyotyping of cell line HS293 was performed at passage levels 9 and 49, and that of HS306 was performed at passage level 32. Samples of cells were treated with colcemid KaryoMAX (0.07 μg/ml; GIBCO, Paisley, U.K.) overnight. After washing, the cells were incubated in 0.4% trypsin solution (GIBCO) for 2–3 minutes. They were treated with collagenase (1,400 IU/ml; Worthington, Lakewood, NJ) at 37°C for 20 minutes and harvested using standard procedures. The metaphases were analyzed after Q-banding.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The isolated ICM of the donated blastocyst received in October 2003 became attached to the feeder cells and grew to form a colony of typical hESCs (Fig. 2A). After 12 days, the first splitting onto new feeder cells was carried out, and thereafter the cells were passaged at 5- to 8-day intervals. This line, HS293, has been in continuous culture for 56 passages, with a doubling time of 24–36 hours. It was mechanically split and transferred to new dishes when the aggregate reached 2.5 to 6.4 × 103 cells. The size of the transferred aggregates has been 1 to 2 × 103 cells. Spontaneous differentiation was observed in 10%–20% of the cells in 1 out of 20 colonies during the 5- to 8-day culture period. The line has been frozen and thawed, and it grows well after thawing.

thumbnail image

Figure Figure 2.. Colonies of lines (A) HS293 (passage level 12) and (B) HS306 (passage level 15) growing on the feeder layer. Magnification ×40.

Download figure to PowerPoint

The other ESC line (HS306) from a blastocyst was received in February 2004. It has been similarly processed and is now at passage level 42 (Fig. 2B). It has similar characteristics to the line HS293 (Fig. 3). Its pluripotency in vitro has been analyzed, and the same markers were expressed as in line HS293.

thumbnail image

Figure Figure 3.. Immunofluorescence staining for markers characterizing undifferentiated human embryonic stem cells and for markers of the three embryonic germ cell layers. Undifferentiated HS293 (passage 49) and HS306 (passage 40) expressed (A, F) Oct-4, (B, G) TRA-1-60, (C, H) TRA-1-81, and (D, I) SSEA-4 and did not express (E, J) SSEA-1. Embryoid bodies of HS293 (passages 49–51) and HS306 (passages 33–37) contained (K, N) ectoderm shown by Nestin expression, (L, O) mesoderm shown by bone morphogenetic protein-4 (BMP-4) expression, and (M, P) endoderm shown by alpha fetoprotein (AFP) expression. (A–E, K–M): HS293; (F–J, N–P): HS306. Specific staining is green (Alexa 488), and nuclear background staining is blue (DAPI). Original magnification (C–E, J, M–O) ×10 and (A, B, F–I, K, L, P) ×40.

Download figure to PowerPoint

The growing cells of both lines expressed markers characteristic of hESCs. Colonies of different sizes stained positively for Oct-4 (Figs. 3A, 3F), TRA-1-60 (Figs. 3B, 3G), TRA-1-81 (Figs. 3C, 3H), and SSEA-4 (Fig. 3D, 3I). In addition, undifferentiated colonies showed positive staining for GCTM-2 and alkaline phosphatase activity (data not shown). Staining for SSEA-1 was negative (Fig. 3E, 3J). Pluripotency in vivo of the line HS293 was shown by teratoma formation in beige/SCID mice. Tissue components could be identified representing ectodermal (Fig. 4A), mesodermal (Figs. 4B, 4C), and endodermal (Fig. 4D) tissue.

thumbnail image

Figure Figure 4.. Teratomas formed after injection of the human embryonic stem cells to beige/severe combined immunodeficiency mice with tissue components of all three embryonic germ cell layers. Ectoderm [(A) epithelium], mesoderm [(B) cartilage; (C) muscle], and endoderm [(D) intestinal mucosa]. Original magnification ×100.

Download figure to PowerPoint

The embryoid bodies formed from lines HS293 and HS306 expressed SOX-1, ND-1, alpha-cardiac actin, and AFP as constituents of the three germ cell layers (Fig. 5), as revealed by RT-PCR. The three germ layers were also confirmed using immunofluorescent staining against Nestin, BMP-4, and AFP, as seen in Figures 3K–3M, which correspond to HS293, and Figures 3N–3P, which correspond to HS306.

thumbnail image

Figure Figure 5.. Embryoid bodies from lines HS293 and HS306 expressed markers for the three embryonic germ cell layers, SOX-1, ND-1, α-cardiac actin and α-fetoprotein, with GAPDH as the housekeeping control as revealed by reverse transcription–polymerase chain reaction.

Download figure to PowerPoint

The karyotype of line HS293 was normal male 46,XY at passage 9 and at passage 49. The karyotype of the line HS306 was normal female 46,XX at passage 32.

When using FCS in the derivations in 2002 to 2003, a total of 67 blastocysts resulted in four permanent lines (HS181, HS207, HS235, and HS237), and at the same time, nine additional lines grew from 2 to 9 passages and then faded off. When using 10 blastocysts in derivations with the SR medium, we obtained the present two permanent lines (HS293 and HS306), and from September to October 2004, a total of 27 blastocysts resulted in 10 early lines (HS346, HS351, HS356, HS360, HS361, HS362, HS363, HS364, HS366, and HS368) when using the same human foreskin feeder cells and SR medium. The line HS346 was obtained from an embryo that had been frozen 5 years ago, and it is presently at passage nine. The other early lines are at passages three through seven.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We derived two characterized hESC lines (HS293 and HS306) using SR culture medium and postnatal human skin fibroblasts as feeder cells. These are the first described hES lines derived and propagated under such conditions. Previously, SR medium and human feeder cells have been used in propagating hESCs established using FCS and mouse fetal fibroblasts [9, 11, 12]. Derivation of one line has been reported using fetal human feeder cells and human serum [8]. Derivation of a line was recently also reported using SR on mouse fetal fibroblasts [9]. Our earlier lines, HS181, HS235, and HS237 [9, 12], derived using the same foreskin fibroblasts as feeders, but in FCS-containing culture medium, can be cultured in SR medium with a better nondifferentiated growth rate than in serum-containing medium [15].

Deriving the lines using SR medium is an important step forward, but the optimum situation is to use a completely animal material–free system. One source of factors of animal origin in customary culture systems is the enzymes often used for passaging the cells. We have used mechanical isolation to keep the line free in this respect, even though enzymes can also be used. Human recombinant collagenase is a safe alternative. The feeder cells, human foreskin fibroblasts, were originally cultured in FCS-containing medium. Even though good-quality FCS was used, it would be better for human cell transplantation in the future to avoid it. Serum-free conditions for establishing and propagating primary feeder cell cultures may be problematic, but the use of human serum might be an alternative. In the present study we wanted to show the feasibility of the culture system with foreskin fibroblasts and SR medium, and the use of completely animal material–free feeder cells remains a study for the near future. The SR medium used obviously also contains some animal-derived purified proteins, which in an optimal culture system should be replaced by human recombinant proteins. Immunosurgery using animal sera is another complication. Mechanical isolation would be better in this respect, and we are considering using it in our future derivations. There are reports in which the ICM has not been isolated at all [19]. The hatched blastocyst has been placed directly on the feeder cells. We were not successful when testing a similar procedure with two blastocysts. We did not want to change several factors in our derivations at the same time. Hence, comparison between mechanical isolation and isolation by only removing the zona remain subjects that have to be systematically studied in the future.

The efficacy of derivation in SR, two permanent lines out of 10 blastocysts, seemed to be at least as good as or better than that during the period when FCS was used (four permanent lines from 67 blastocysts, plus nine early lines that faded off). The latest period, September through October 2004, with 10 early lines growing from 27 blastocysts, is promising. A prospective comparative study using embryos in derivation in either FCS or SR has not been feasible. Our existing lines originally derived in FCS containing one medium grew better in SR medium [15]. At this stage we do not see a reason for going back to FCS containing derivations or cultures.

Feeder-free serum-free derivation and culture have proved successful in mouse cell lines [24], and this is, of course, also important for human cells. For this, the factors regulating self-renewal of hESCs should be identified.

We encourage derivation and culture of hESCs in SR medium and on human skin fibroblasts as feeder cells as an alternative that can be applied in the laboratories of in vitro fertilization units. Many more hESC lines are needed, and embryologists used to working with human embryos have excellent possibilities to be successful in this work.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This study was supported by grants from the Swedish Research Council, the Juvenile Diabetes Research Foundation, the Finnish Academy, NorFa, and the Karolinska Institute. We thank Nicholas Bolton for revising the language of this manuscript, Marta Imreh for her culture work, and the laboratory personnel of the IVF Unit for their embryo cultures.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References