Phenotypic and Differentiation Stability of Human Embryonic Stem Cell-Derived Osteoblasts

Abstract

Purpose

To ensure the efficiency and safety of transplanted human embryonic stem cell (hESC)-derived osteoblast-like cells (hESC-OS) for bone regeneration, this study was designed to determine the effects of continuous cell expansion on the osteoblastic differentiation stability, pluripotency, and tumorigenic potential of long-term expanded hESC-OS.

Methods

hESCs manually harvested as cell aggregates or enzymatically dissociated as single cells were directly incubated in osteogenic medium and serially passaged to passage 25. Expression of osteoblast-related genes, pluripotent regulator genes, and genes related to tumorigenesis were examined at the primary passage and every 5 passages thereafter. hESC-OS were subcutaneously transplanted into nude mice for 4–24 weeks to test for teratoma formation. hESC-OS were recultivated in hESC culture conditions to evaluate the extent to which reverse differentiation back to the undifferentiated stage may occur.

Results

hESC-OS derived from hESC aggregates and dissociated cells exhibited comparable osteoblast differentiation patterns. Expression levels of osteoblast-related genes reached plateau levels at passages 5–10 before declining in higher passages. Expression of tumor-associated genes was not significantly increased. Only hESC-OS at primary and first passages formed teratomas after 4 weeks in vivo. The hESC-OS were not able to revert to hESCs.

Conclusions

Expanded hESC-OS demonstrated lineage-specific differentiation stability, did not maintain the pluripotency of hES cells, and were genetically stable. Thus, hESC-OS may be considered for large animal preclinical studies.

Introduction

To generate the number of osteogenic cells derived from human embryonic stem cells (hESCs) required for cell-based regeneration, a number of challenges must be overcome. Among these challenges is the necessity to prolong the expansion of the differentiated cells before transplantation and to ensure transplantation of homogenous and lineage-restricted cells.

Long-term in vitro cultures of human mesenchymal stem cells (hMSCs) and osteoprogenitor cells are associated with reduced proliferative and differentiation potential due to cellular senescence [Bruder et al., 1997; Kim et al., 2009]. Furthermore, some reports have indicated that long-term culture may cause a spontaneous transformation of hMSCs [Rubio et al., 2005; Rosland et al., 2009]. This transformation potential is amplified when the differentiated cells are derived from pluripotent cells such as hES cells that are known to have a high capacity to form teratomas in vivo [Takahashi and Yamanaka, 2006].

Therefore, to ensure the efficiency and safety of hESC-derived osteoblast-like cells (hESC-OS), the goals of this study were: (1) to determine the effects of continuous cell expansion on osteoblastic differentiation stability and (2) to examine the pluripotency and tumorigenic potential of long-term expanded hESC-OS.

Materials and Methods

Human Embryonic Stem Cell Culture and Osteogenic Induction

Undifferentiated hESCs (BG01; Bresagen, Inc., USA) (passages 35–50) as (1) a suspension of single cells or (2) cell aggregates were regularly passaged in osteogenic medium as described [Arpornmaeklonget al., 2010]. A karyotype analysis was performed by cytogenetic analysis on 20 G-banded metaphase cells of hESCs-OS at passages 5, 10, 17, and 20 (Cell Line Genetics, USA).

Quantitative Real-Time Reverse Transcriptase Polymerase Chain Reaction

Quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR) was performed on hESCs and hESC-OS at the primary passage and every 5 passages thereafter as described [Arpornmaeklong et al., 2010]. The genes and primers used are listed in table 1.

Table 1.

List of primers used in the qRT-PCR analysis

Names/symbols	Gene description	Catalog numbers
ALPL	alkaline phosphatase, liver/bone/kidney	Hs01029141_g1
Beta-actin (ACTB)	human ACTB (actin, beta) endogenous control	4333762T
hRas	v-Ha-ras Harvey rat sarcoma viral oncogene homolog	Hs00610483_m1
Myc (c-Myc)	v-myc myelocytomatosis viral oncogene homolog	Hs99999003_m1
Nanog	Nanog homeobox	Hs02387400_g1
Oct4 (POU5F1)	POU class 5 homeobox 1	Hs03005111_g1
Osteocalcin (BGLAP)	bone gamma-carboxyglutamate (gla) protein	Hs01587814_g1
Runx2	Runt-related transcription factor 2	Hs00231692_m1
SNRPE (bRaf)	small nuclear ribonucleoprotein polypeptide E	Hs01635040_s1
Sox2	SRY (sex determining region Y)-box	Hs01053049_s1

Reversion of Differentiation

hESC-OS at passages 1, 5, and 10 were seeded in 6-well culture plates (5 × 10³ cells/cm²) and incubated for 21 days in hESC medium, human fetal fibroblast condition medium (Stemcell Technologies, Canada) on a Matrigel- (BD Biosciences, USA) coated surface for hESC culture, and osteogenic mediums (as described in the previous section) on a cell culture plate surface for differentiation culture. Cell morphology and undifferentiated cell colony formation were observed at the time of cell culture change. On day 21, hESC-OS were collected for expression of pluripotency related genes Oct4, Nanog, Sox2, and Lin28 (table 1) using qRT-PCR analysis [Takahashi and Yamanaka, 2006].

Preparation of Composite Scaffolds and Subcutaneous Transplantation

All procedures were approved by the University of Michigan Committee on the Use and Care of Animals. On culture days 6 and 7, undifferentiated hESCs (BGO1) at passages 50–60 were dissociated and hESC-OS at passages 1, 5, 10, 15, 20, and 25 were trypsinized. Cell- (2 × 10⁶ cells/pellet) gelatin sponge- (7 × 3 mm; Pharmacia & Upjohn Co., USA) fibrin gel [40 mg/ml fibrinogen with 200 units/ml human thrombin (Sigma, USA)] composite scaffolds were prepared as described [Arpornmaeklong et al., 2010]. Six composite scaffolds were prepared for the transplantation of hESCs at passages 50–60 and hESC-OS at passage 1. A total of 18 composite scaffolds were prepared for the transplantation of dissociates and aggregates of hESC-OS at passages 5–25. Composite scaffolds were subcutaneously transplanted in 30 immunocompromised mice (N:NIHbg-nu-xid; Harlan Laboratories, Inc., Indianapolis, Ind., USA), with 4 pockets in 1 mouse and 1 sample in 1 pocket. Each pocket on 1 animal contained a composite scaffold with cells from different passages. The transplants were harvested 4, 12, and 24 weeks after transplantation. Histological sections were examined for tumor formation [Takahashi and Yamanaka, 2006].

Statistical Analysis

Results were derived from 3 strains each of hESCs cultured as dissociated cells and cell aggregates. Data were analyzed by 1-way analysis of variance (ANOVA) at p < 0.05.

Results and Discussion

The decrease in osteoblastic differentiation potential after the extensive expansion of hESC-OS in vitro (fig. 1a-c) is similar to those of osteoblast-like cells derived from hBMSCs [Bruder et al., 1997] and hMSCs [Kim et al., 2009] in aging or presenescent stages.

Fig. 1.

hESC-OS showed osteogenic differentiation potential and genotypic stability by exhibiting an absence of spontaneous tumorigenic differentiation. Expression profiles of osteoblast-related genes (n = 3) (a–c) and oncogenes (d–f) over 20 cell passages in osteogenic medium (n = 3). g–i Decrease in the expression levels of pluripotency regulator genes at passages 5 and 10 following a reverse differentiation study (n = 6). g Oct4. h Nanog. i Sox2. ∗ = Significant differences between groups at the same passages; ‡ = lower levels than passage 10; ++ = significantly lower levels than the primary passage (passage 0); + = highest expression levels (mean ± SEM, p < 0.05).

Tumorigenic transformation of the differentiated cells is another major concern for safe clinical application when using differentiated cells derived from stem cells. Transformed mesenchymal stem cells have a significantly increased proliferation rate, altered morphology and phenotype, and form tumors in vivo [Rubio et al., 2005; Rosland et al., 2009]. In the current study, the phenotypic and genotypic stability of hESC-OS in long-term expansion was indicated by the consistent absence of chromosomal abnormalities of hESC-OS at passages 5–25 revealed by normal karyotypes (data not shown), decreased expression levels of oncogenes (fig. 1d-f), the lack of dedifferentiation (fig. 1g-i), and the absence of tumor formation after long-term transplantation (fig. 2). Moreover, an absence of undifferentiated colonies in the reversion-of-differentiation study further supported a lack of pluripotent cells and lack of undifferentiated cells in the hESC-OS (data not show) [Takahashi and Yamanaka, 2006].

Fig. 2.

Absence of tumorigenicity and undifferentiated cells from long-term transplanted hESC-OS. Transplantation of hESCs (a–c) and hESC-OS (d–f) at passage 1 for 4 weeks and a representative of expanded hESC-OS (g–i) at passage 10 for 24 weeks. The insets indicate selected regions of interest. The magnified images from the selected regions of interest are shown in the subsequent images (c, f, i). a Solid tumor from transplanted hESCs. b, c Histology of solid tumor revealing predominantly ductal tissue from the endodermal lineage, ductal tissue (arrow), and cartilage (arrowhead). d Cystic lesion containing clear fluid from the transplantation of hESC-OS at passage 1. e Revealing cystic lesion with multiple compartments of thin fibrous tissue wall (arrow) and remnants of the gelatin sponge (∗). f Cystic wall comprised of collagenous stroma (∗∗), blood vessel (arrow) and epithelium cells (arrowhead). g Remnant of an hESC-OS-gelatin sponge construct 24 weeks after transplantation. h Approximate margins (arrows) of a remnant gelatin sponge (∗) embedded in host connective and fat tissues. i Gelatin sponge (∗) infiltrated with connective tissue (arrowheads) with no evidence of teratoma formation 24 weeks after transplantation.

The absence of tumor formation from long-term transplantation of expanded hESC-OS up to passage 25 further substantiated genomic stability and the absence of undifferentiated cells in the expanded hESC-OS population. Solid tumor and cystic lesion formation were found 4–6 weeks after transplantation in all samples of hESCs (fig. 2a-c) and hESC-OS at passage 1, respectively (fig. 2d-f). All expanded hESC-OS at passages 5 and higher (10, 15, 20 and 25) did not form teratomas even after 6 months in vivo. Sixth months after implantation, the transplants were extensively resorbed and residual scaffolds were infiltrated with connective tissue (fig. 2g-i).

Despite these findings, the alloimmunogenic reaction and integration of differentiated cells into a local environment and continuous growth and function as differentiated cells are the principle challenges requiring further studies before hESC-OS can be used for clinical applications [Motaln et al., 2010].

Taken together, expanded hESC-OS acted as normal somatic tissue with a limited life span. Our data suggest that the application of hESC-OS at passages 5–10 should be used for developmental studies and transplantation. The current study supports the safe use of hESC-OS in in vivo applications and thus hESC-OS may be considered for large animal preclinical studies.

Glossary

Abbreviations used in this paper

hESC-OS	osteoblast-like cells derived from hESCs in osteogenic medium
hESCs	human embryonic stem cells
hMSCs	human mesenchymal stem cells
qRT-PCR	quantitative real-time reverse transcriptase polymerase chain reaction