Efficient Engineering of Vascularized Ectopic Bone from Human Embryonic Stem Cell

Abstract

Human mesenchymal stem cells (hMSCs) can be derived from various adult and fetal tissues. However, the quality of tissues for the isolation of adult and fetal hMSCs is donor dependent with a nonreproducible yield. In addition, tissue engineering and cell therapy require large-scale production of a pure population of lineage-restricted stem cells that can be easily induced to differentiate into a specific cell type. Therefore, human embryonic stem cells (hESCs) can provide an alternative, plentiful source for generation of reproducible hMSCs. We have developed efficient differentiation protocols for derivation of hMSCs from hESCs, including coculture with murine OP9 stromal cells and feeder layer-free system. Our protocols have resulted in the generation of up to 49% of hMSCs, which expressed CD105, CD90, CD29, and CD44. The hMSCs exhibited high adipogenic, chondrocytic, and osteogenic differentiation in vitro. The latter correlated with osteocalcin secretion and vascular endothelial growth factor (VEGF) production by the differentiating hMSCs. hMSC-derived osteoblasts further differentiated and formed ectopic bone in vivo, and induced the formation of blood vessels in Matrigel implants. Our protocol enables generation of a purified population of hESC-derived MSCs, with the potential of differentiating into several mesodermal lineages, and particularly into vasculogenesis-inducing osteoblasts, which can contribute to the development of bone repair protocols.

Introduction

Human mesenchymal stem cells (hMSCs) are multipotent cells that give rise to mesodermal tissues, including bone, fat, and cartilage.^1–3 hMSCs have thus emerged as a promising cell type in regenerative medicine, with documented evidence of therapeutic efficacy in the field of bone repair.^4,5 In most preclinical and clinical studies, hMSCs were derived from bone marrow; however, their use has been hampered largely due to the invasive procedures that are mostly required for their isolation, their finite life span, and their instability in long-term cultures.^6,7 Although adipose tissue represents an alternative and accessible source of adult stem cells,⁸ their isolation still requires an invasive procedure, and their numbers, as well as differentiation capacities, decrease with age.^3,9 With their capacity for almost infinite expansion, pluripotent human embryonic stem cells (hESCs) provide an alternative source for the generation of hMSCs, circumventing the need for invasive techniques. The resultant hMSCs can further differentiate into specific lineages such as osteoblasts. Several protocols for the generation of hMSCs from hESCs (hESC-MSCs) have recently been described based on directed differentiation systems: coculture with the murine OP9 stromal cell line,^10,11 manual selection of cell populations,^12–15 and cell sorting.^16,17 In addition, different cell culture conditions have been used to induce differentiation of hESCs or hESC-MSCs through the osteoblastic lineage: by embryoid body (EB) formation, through outgrowth culture, and by coculture with a specific cell type. However, the osteogenic potential of the generated cells has generally been assessed by in vitro analysis, except for several studies where bone formation was demonstrated in vivo using osteoconductive scaffolds, which further stimulate osteogenic differentiation of implanted hMSCs post-transplantation.^9,14,18–20 In addition, the kinetics of bone maturation in vivo has not been described.

In the developing skeleton, the endochondral cartilage template is replaced by highly vascularized tissue.²¹ The formation and development of an active microvasculature are essential for bone remodeling and fracture healing.^22,23 Recent studies have suggested that vascular endothelial growth factor (VEGF), a potent angiogenic stimulator, may play a role in bone formation. VEGF is a subfamily of homodimeric proteins referred to as VEGF-A, B, C, and E, and placenta growth factor.^24,25 VEGF-A exists in five different isoforms. VEGF₁₆₅, the most abundant of these isoforms, is soluble and can bind to the extracellular matrix.²⁶ Maes et al. and Street et al. showed that administration of VEGF stimulates, and its antagonist FLT-1 suppresses, bone repair.^27,28 VEGF has also been shown to demonstrate a beneficial effect on bone healing elicited by bone morphogenic protein 4.^29,30 Further, the absence of VEGF isoforms in genetic models indicates a completely disrupted vascular pattern associated with impaired differentiation of hypertrophic chondrocytes and osteoblasts, suggesting that VEGF not only mediates bone vascularization but also affects the differentiation of progenitors to hypertrophic chondrocytes and osteoblasts.³¹

The current study implements considerable improvements in two protocols for derivation of hESC-MSCs: coculture with murine OP9 and feeder layer-free systems. The result is highly efficient and reproducible directed differentiation of hMSCs. After CD73-based sorting, a pure population of hMSCs was established in long-term cultures. Two lines of hESC-MSCs showed mesenchymal morphology, extended in vitro expandability, and robust multilineage potential, giving rise to chondrocytes, adipocytes, and osteoblasts in vitro. The hESC-MSCs also secreted osteocalcin and VEGF during osteoinduction. We further demonstrated the capability of the hESC-MSCs to differentiate into ectopic bone in vivo without the use of an osteoconductive carrier, and to induce endothelial cell sprouting in Collagen gels as well as invasion of blood vessels into hMSC-derived bone-like implants.

Materials and Methods

Cell culture and fluorescence-activated cell sorting

The undifferentiated hESC lines I-4 (P68-71) and H9.2 (P54) used in this study were derived from the preimplantation embryo as described previously.^32,33 The blastocysts that were produced by in vitro fertilization for clinical purposes were donated by individuals after informed consent and after the approval of the Rambam Medical Center Helsinki Committee and the Israeli Ministry of Health. The hESCs were cultured on mitotically inactivated mouse embryonic fibroblasts and maintained under growth conditions and passaging techniques as previously described.³⁴ In brief, hESCs were cultured with a medium consisting of 80% Dulbecco's modified Eagle's medium F12 (DMEM-F12; Biological Industries), 1 mM l-glutamine, 0.1 mM β-mercaptoethanol, and 1% nonessential amino acid stock (Gibco BRL), supplemented with 20% KnockOut SR, a serum replacer (Gibco BRL), and human recombinant basic fibroblast growth factor (bFGF) at a concentration of 4 ng/mL (R&D systems). hESCs were expanded by routine passage every 3–4 days with 1.2 mg/mL collagenase type IV (Gibco BRL). The OP9 cell line was purchased from the American Type Culture Collection and was maintained in an alpha-minimal essential medium (α-MEM; Beit Haemek) containing 20% fetal bovine serum (FBS; HyClone) and 2 mM l-glutamine. Mesenchymal differentiation was induced by plating hESCs at a density of 30,000 cells/cm², on a monolayer of 35,000 cells/cm² inactivated murine OP9 stromal cells (n=4), or alternatively on Matrigel-coated plates (BD Pharmingen) at a final dilution of 1:40 (n=6). First, the hESC colonies were removed from plates using 1.2 mg/mL collagenase type IV, and then the cells were treated with trypsin–ethylenediaminetetraacetic acid 0.25% for 5 min to obtain single cells. The cells were counted and seeded in the aforementioned concentration. The cells were grown for 20 days in a serum-free medium consisting of DMEM-F12 supplemented with insulin-transferrin and selenium (ITS; Gibco), and an additional 14 days using an α-MEM containing 20% FBS (hMSC growth medium). Fluorescence-activated cell sorting (FACS) for CD73-positive cells (CD73⁺ cells, CD73-PE; BD Pharmingen) was performed at day 35 on FACS Aria using DIVA software (Becton Dickinson Immunocytometry Systems). At this point, the cells were designated as hESC-MSCs, and were used in the analysis of the hMSC phenotypic characterization and differentiation potential.

Cell expansion and growth curves

Population-doubling time (PDT) was calculated by the formula t/PD; t is the number of hours in culture. PD was calculated by the formula ln(Xf/X0)/ln2, where X0 is the number of seeded cells, and Xf is the number of cells at confluence before splitting. The growth curve was established by using the XTT kit (Biological Industries) according to the manufacturer's instructions.

Flow cytometry analyses

hESC-MSCs were harvested using 0.25% trypsin (Biological Industries), and after neutralization, resuspended in a staining buffer, phosphate-buffer saline (PBS) +1% FBS. About 1.5×10⁵ cells were incubated with each of the following monoclonal conjugated antibodies: CD73, CD45 (IgG1–PE; BD Pharmingen), CD105, CD29, CD90, CD31 (IgG1–PE; eBioscience), CD44 (IgG2b-PE; eBioscience), and CD34 (IgG1–PE; IQ Products) and acquired using FACSCalibur. Nonspecific fluorescence was determined by incubation of similar cells with matched isotype control antibodies. Samples were analyzed using an FACSCalibur flow cytometer with CellQuest Acquisition software (Becton Dickinson Immunocytometry Systems).

Adipogenic differentiation

hESC-MSCs were seeded at a density of 20,000 cells/cm², and grown in a DMEM F-12 medium containing 10% FBS and 1 mM l-glutamine. The medium was supplemented with 0.5 mM 3-isobutyl-1-methylxanthine, 10 μg/mL insulin, 10⁻⁶ M dexamethasone, and 0.1 mM indomethacin (all from Sigma-Aldrich). The medium was replaced every 3–4 days for 4 weeks. Cells were rinsed three times with PBS, fixed with 4% paraformaldehyde (PFA; Electron Microscopy Sciences) for 20 min, rinsed again, and stained for 10 min at room temperature with a fresh Oil Red O solution (Sigma-Aldrich).

Osteogenic differentiation

hESC-MSCs were seeded at a density of 2000–3000 cells/cm² and grown in an osteogenic medium (OM): GMEM BHK-21 (Gibco BRL) medium supplemented with 10% FBS, 1 mM sodium pyruvate (Gibco BRL), 1% nonessential amino acids, 50 μg/mL l-ascorbic acid (Sigma), 0.1 mM β-mercaptoethanol, 10 mM β-glycerol-phosphate (Sigma), and 0.1 μM dexamethasone. The medium was replaced every 3–4 days for 3–4 weeks. The control medium did not include the osteogenic inducers β-glycerol-phosphate and dexamethasone. Cells were rinsed twice with PBS, fixed with 4% PFA for 20 min, rinsed again, and stained with 2% Alizarin red solution (pH 4.5; Sigma) for 15 min at room temperature.

Chondrogenic differentiation

For chondrogenic differentiation, 2×10⁵ hESC-MSCs were centrifuged at 300 g for 5 min in 15-mL polypropylene falcon tubes to form a cell pellet. The cells were grown for 9 weeks in the DMEM (Gibco BRL), supplemented with 10⁻⁷ M dexamethasone 1% ITS, 50 μg/mL l-ascorbic acid, 1 mM sodium pyruvate, 4 mM l-proline (Biological Industries), and 10 ng/mL TGFβ3 (PeproTech). The medium was replaced twice each week without disturbing the cell mass. Cell sections were made after fixing the cell pellets with 4% PFA and embedding in low melting point agarose (1.5%). Hematoxylin and eosin (H&E) and Alcian blue staining was conducted.

Myogenic differentiation

For myogenic differentiation, hMSCs (2000 cells/cm²) were cultured for 2 weeks with an α-MEM containing 20% serum. The cells were then tested for expression of NCAM/CD56, a marker known to be expressed on skeletal myoblasts, by FACS (CD56; eBioscience).

Reverse transcription polymerase chain reaction

Total RNAs were isolated using TRIzol/Tri-reagent (Invitrogen), and reverse transcribed by the iScript^™ cDNA synthesis kit (BIO-RAD). Reverse transcription polymerase chain reaction (RT-PCR) was performed by the DreamTaq^™ Green Master Mix (Fermentas). Primer sequences and product sizes are listed in Supplementary Table S1 (Supplementary Data are available online at www.liebertpub.com/tea).

Concentrations of VEGF₁₆₅ in hMSC-conditioned medium

hESC-MSCs were cultivated with an hMSC growth medium (see the section Cell culture and FACS) or with an OM (see the section Osteogenic differentiation) for 3 weeks. To measure the secretion of VEGF₁₆₅, the medium was replaced at the indicated time, and the cell number determined. Human VEGF₁₆₅ in a conditioned medium was detected by an enzyme-linked immunosorbent assay (ELISA) kit (PeproTech) according to the manufacturer's instructions.

Quantification of osteocalcin in hMSC-conditioned medium

Osteocalcin was quantified in a serum free-OM, according to the manufacturer's instructions, using the human osteocalcin instant ELISA kit (eBioscience). The assay was performed on culture days 7, 14, and 21.

Spheroid sprouting assay

The assay was performed as described elsewhere.³⁵ In brief, human umbilical vein endothelial cell (HUVEC) spheroids were generated overnight in a hanging-drop culture consisting of 400 cells in 80% endothelial cell medium (ScienCell) and 20% methylcellulose (Sigma). Spheroids were embedded in collagen (collagen extract of rat tails) and stimulated with 5 ng/mL bFGF (R&D systems) or with an osteogenic conditioned medium of hESC-MSCs in the presence or absence of 10 μM SU6668 inhibitor (Tocris).

Proliferation assay

HUVECs (1×10⁴ cells/well) were seeded in 24-well plates. The cells were incubated with 5 ng/mL bFGF or with a conditioned medium of hESC-MSCs in the presence or absence of 10 μM SU6668 inhibitor (Tocris). The cells were counted 3 and 5 days after seeding.

Osteogenic Matrigel implants

The formation of ectopic osteogenic structures in vivo was evaluated using a xenograft model of Matrigel implants. hESC-MSCs (1.5×10⁶) undergoing osteoblast differentiation in vitro for 3, 7, or 14 days and noninduced hESC-MSCs were mixed with 250 μL Matrigel on ice and injected subcutaneously to immune-deficient NOD/SCID 8–12-week-old mice. All implants were removed after 2 weeks and fixed in 4% PFA. Empty Matrigel implants served as control. For bone maturation, the hESC-MSCs (1.5×10⁶) underwent osteoblast differentiation in vitro for 14 days; the cells were mixed with 250 μL Matrigel on ice and injected subcutaneously to immune-deficient NOD/SCID 8–12-week-old mice. The implants were removed after 1, 2, or 4 weeks, fixed in 4% PFA, and stained with anti-osteocalcin. Placenta-derived MSCs were isolated as previously described³⁶ and served as a positive control. All procedures were approved by the committee for the Use and Care of Animals of the Technion-Israel Institute of Technology.

Teratoma formation

hESCs and hESC-MSC lines I-4 and H9.2 were injected into the rear leg muscle of 4-week-old, male SCID/beige mice (Harlan). The number of cells per injection ranged from 2.5×10⁶ to 5.0×10⁶. Ten weeks or more after the injection, the resulting teratomas were examined histologically.

Histology immunofluorescence and immunohistochemistry

Paraffin-embedded sections of the implants were stained with H&E and Alizarin red for evaluation of calcium deposit content. For fluorescence microscopy, the following uncoupled anti-human primary antibodies were used: alkaline phosphatase (ALP; BD Pharmingen; 1:100), rabbit-anti human MHC class I (Epitomics; 1:100), mouse-anti human osteocalcin (Abcam; 1:200), and goat anti-human osteopontin (1:50). The secondary conjugated antibodies included Alexa-488-conjugated donkey anti-goat (1:100; Invitrogen), goat anti-mouse Cy-3 (1:100; Jackson), and donkey anti-rabbit Cy-3 (1:100; Jackson). The nucleus was labeled with DAPI (4′,6-diamidino-2-phenylindole, dihydrochloride; Molecular Probes; 1:1000) for 5 min at room temperature.

Fluorescence in situ hybridization

Aquarius probes (human chromosome 1-red, human DAPI antifade; Cytocell) were used for whole-chromosome painting, according to the manufacturer's instructions.

Microvessel density analysis

Microvessels were quantified by evaluation of 10 randomly selected fields of H&E-stained sections obtained from various parts of the implants. Microvessels were identified as lumenal structures containing red blood cells. Microvessel density was reported as the average number of red blood cell-containing microvessels from the fields analyzed, and expressed as the number of lumens/mm².

Statistical analysis

Statistical significance (p<0.05; p<0.01) was determined using analyses of variances (one-way analysis of variances), followed by Tukey's post hoc testing.

Karyotype

Karyotype was performed as described elsewhere.³⁷

Results

Isolation of hESC-MSCs grown in coculture with OP9 stromal cells or on Matrigel-coated plates

We achieved mesenchymal differentiation by plating undifferentiated hESC lines I-4 and H9.2 at a high density of 30,000 cells/cm² on a monolayer of murine OP9 stromal cell–coculture system, or alternatively on Matrigel-coated plates as a feeder-free system. At day 35, the hESC-MSCs were harvested and sorted by FACS using the surface antigen CD73, which is known to be expressed in hMSCs. CD73-positive cells (CD73⁺) comprised 26±7% of the differentiated hESCs grown on OP9 cells (Fig. 1A, upper panel). The specificity of the CD73 antibody for detection of hMSCs was to the human protein only; cross-reactivity with the murine OP9 feeder was not shown (Fig. 1B). We also used a feeder-free system based on gelatin and Matrigel matrices to induce mesenchymal differentiation similar to that in the coculture system. Undifferentiated hESCs (lines I-4 and H9.2) were seeded on gelatin- or Matrigel-coated plates. Cells that were seeded on gelatin detached from the plates during the culture period, significantly reducing the number of cells available for CD73-based sorting. This phenomenon did not occur on Matrigel-coated plates. The feeder-free system yielded twice the percentage of CD73⁺ derivatives (49±12%) as the coculture system (Fig. 1A, lower panel). When hESCs were spontaneously differentiated in 3D culture to form EBs, only 4% of the cells were detected as CD73⁺ in 14-day-old EBs (Supplementary Fig. S1). Therefore, the feeder-free system seems to be the most efficient means of differentiating hMSCs (Fig. 1C). The procedure for isolating CD73⁺ cells appears reproducible, since a total of 10 independent experiments, four in the coculture system and six in the feeder-free system, yielded similar populations of cells. Sorted hESC-MSCs were further efficiently expanded (up to 10⁸ cell) in long-term culture for 20 passages, with normal karyotype retained (Supplementary Fig. S2). hESC-MSCs from the different systems exhibited different growth potential. hESC-MSCs from the feeder-free system had PDT of about 60 h in the first 4 weeks, whereas cells from the coculture system had PDT of 80 h (Fig. 1D, E). In addition, freeze and thaw cycles demonstrated good recovery of at least 90%.

FIG. 1.

Isolation of hESC-MSCs. (A) Fluorescence-activated cell sorting of CD73⁺ cells grown in coculture with OP9 stromal cells (upper panel) and in feeder-free system (lower panel). An average of 27% of the differentiated hESCs grown in the coculture system (n=4) and 45% of those grown in the feeder-free system (n=6) expressed CD73. (Gate P5: CD73-positive sorted cells). (B) Specificity of the CD73 antibody for detection of MSCs was to the human protein only; cross-reactivity with the murine OP9 feeder was not observed (upper panel: isotype-control staining. Lower panel: OP9 cells stained with anti-human CD73). (C) Summary of the hESC-MSCs grown in directed differentiation (coculture or feeder-free system) compared with spontaneous differentiation (control embryoid bodies, n=2). *p<0.01 compared with the coculture system. (D) Expansion of hESC-MSCs during 8 weeks of culture. Population-doubling time of hESC-MSCs from either feeder-free and coculture system was 64 and 81 h, respectively, in the first 4 weeks of cultivation. (E) Representative growth curves of hESC-MSCs from coculture (left panel) and feeder-free (right panel) systems at the indicated passages. hESC-MSCs, human embryonic stem cell-derived mesenchymal stem cells. Color images available online at www.liebertpub.com/tea

Characteristics of hESC-MSCs

Sorted hESC-MSCs, from both coculture and feeder-free systems, exhibited heterogeneous morphology, depending on the cell density, ranging from fibroblastic shape to rectangular cells in confluent culture. The hESC-MSCs from both systems were analyzed by flow cytometry for the expression of a comprehensive set of markers known to be expressed in hMSCs. hESC-MSCs coexpressed CD105, CD29, CD44, and CD90 (Fig. 2A), and were devoid of the hematopoietic markers CD45 and CD34, as well as the endothelial/monocytic cell surface marker CD31/PECAM. Immunofluorescence staining and RT-PCR analyses demonstrated that long-term cultured CD73⁺ cells did not express the pluripotency markers Oct3/4 and Nanog, whereas their parental source H9.2 and I-4 hESC lines were positively stained for Oct3/4, Nanog, and TRA1-60, and also expressed Oct3/4 and Nanog, as demonstrated by RT-PCR analyses (Fig. 2B, C). Immunofluorescence staining and RT-PCR analyses revealed that similar to native tissue-derived hMSCs,³⁶ long-term cultured CD73⁺ cells expressed nestin, but did not express brachyury (Fig. 2B, C).

FIG. 2.

Characterization of hESC-MSCs. (A) Flow cytometry analyses of long-term cultured hESC-MSCs from coculture (left panel) and feeder-free systems (right panel), for a characteristic set of markers of hMSCs. The black histogram represents the isotype control staining. (B) Representative immunofluorescence stainings of hESCs (a, b) and hESC-MSCs (c, d). hESCs positively stained to TRA1-60 (a[i]), Nanog (a[ii]), and Oct3/4 (b[i]). hESC-MSCs negatively stained to Oct3/4 (c[i]), and positively stained to nestin (c[ii]). hESC-MSCs negatively stained to Nanog (d[i]). Nuclear staining by DAPI. (C) RT-PCR analyses of pluripotent genes (Oct3/4 and Nanog) and genes expressed by MSCs (nestin and brachuyry) in hESCs (lines H9.2 and I-4, lanes 1 and 2, respectively) and hESC-MSCs grown in coculture or in feeder-free system (lanes 3 and 4, respectively). GAPDH served as the housekeeping gene. (D) In vitro differentiation of hESC-MSCs into adipocytes. Left: more than 95% of the cells stained positive for Oil Red O staining. Right: RT-PCR analyses for adipocyte-associated genes in adipogenic hESC-MSCs (lane 1) and in placenta-derived MSCs (lane 2, positive control). (E) Chondrogenesis. Left: positive Alcian blue staining of cartilage-specific glycosaminoglycans was detected throughout the cell mass section. Right: RT-PCR analyses for chondrocyte-associated genes in chondrogenic hESC-MSCs (lane 1) and in placenta-derived (lane 2, positive control). DAPI, 4′,6-diamidino-2-phenylindole; RT-PCR, reverse transcription polymerase chain reaction; GAPDH, glyceraldehyde 3-phosphate dehydrogenase. Color images available online at www.liebertpub.com/tea

Differentiation of hESC-MSCs into adipocytes and chondrocytes in vitro

Adipogenic differentiation, first detected 5 days after culture initiation by the appearance of lipid vesicles, was completed within 3 weeks. The majority of the resultant cells were of the multilocular type, with polygonal morphology, and containing a large number of lipid droplets of various sizes in the cytoplasm. As differentiation progressed, the lipid-rich vacuoles within the adipocytes increased in size and became brighter. Adipogenic differentiation was highly efficient; after 3 weeks of induction, more than 95% of cells displayed Oil Red O-positive lipid droplets within the adipocytes (Fig. 2D). In addition, RT-PCR analyses revealed that upon adipogenic induction, hESC-MSCs expressed the early adipogenic transcription factor peroxisome proliferator-activated receptor gamma, as well as adipoprotein 2 (AP2) (Fig. 2D). Upon chondrogenic induction, mature lacunae appeared in the center of all pellets, and a perichondrium, a sheath of dense connective tissue surrounding the cartilage, was also detected. Positive Alcian blue staining of cartilage-specific glycosaminoglycans appeared throughout the cell section, with particular enhancement in the marginal area (Fig. 2E). Differentiated cells expressed aggrecan, Sox 9, and collagen type II, as demonstrated by RT-PCR analyses (Fig. 2E). hESC-MSCs had limited myogenic potential. After only a few of the trails, a limited number of cells expressing CD56/NCAM were observed. Sorting for these CD56-positive cells had failed. In addition, expression of MyoD was not observed with immunostaining (data not shown).

Differentiation of hESC-MSCs into osteoblasts is accompanied by secretion of osteocalcin

Osteogenic differentiation was detected by brown deposits first appearing 5 days after culture initiation. Robust osteogenic differentiation was detected at 2 weeks of culture, and remained stable until day 28, as demonstrated by an increased level of mineralized matrix stained positive by Alizarin red (Fig. 3A and Supplementary Fig. S3A). In addition, immunostaining of ALP was only identified in hESC-MSCs grown with an OM, and not in cells grown in the control medium in the absence of osteogenic inducers (Fig. 3B and Supplementary Fig. S3B). RT-PCR analyses revealed that hESC-MSCs expressed osteoblast-related genes, including osteocalcin, osteonectin, osteopontin, and collagen type I (Fig. 3C). Osteopontin expression was observed exclusively in hESC-MSCs grown in the OM. However, osteocalcin, osteonectin, and collagen type I were observed in hESC-MSCs grown either in an hMSC growth medium or in a control medium (without osteogenic inducers) as well (Fig. 3C). Yet, significant increases in osteonectin and collagen type I expression were detected in hESC-MSCs grown in the OM. Densitometry values demonstrated a twofold increase in osteonectin expression by hESC-MSCs grown in the OM as compared to cells grown in the control medium (without osteogenic inducers) (Supplementary Fig. S4). Levels of osteocalcin secretion from hESC-MSCs and placenta-derived MSCs that served as a positive control in OM were detected in vitro by ELISA. Osteocalcin levels increased during osteogenesis in both hESC-MSCs and placenta-derived MSCs. Levels of osteocalcin secreted by the placenta-derived MSCs were higher at each time point than the level secreted by the hESC-MSCs (Fig. 3D).

FIG. 3.

Osteogenesis in vitro and osteocalcin secretion. (A) An increased level of the mineralized matrix stained positive by Alizarin red was detected after 2 weeks in cells grown in an OM (×20). (B) Immunostaining for bone-specific alkaline phosphatase in hESC-MSCs grown in an OM (×10). (C) Changes in osteogenic gene expression during osteogenic differentiation of hESC-MSCs by RT-PCR analyses. Left lane: hESC-MSCs grown in an hMSC GM. Middle lane: hESC-MSCs grown for 28 days in a CM (w/o osteogenic inducers). Right lane: hESC-MSCs grown for 28 days in an OM. Osteopontin expression was observed exclusively in hESC-MSCs grown in an OM. A significant increase in osteonectin and collagen type I expression was detected in hESC-MSCs grown in an OM. (D) Levels of osteocalcin secretion in a conditioned medium of hESC-MSCs grown in an OM. Placenta-derived MSCs served as positive control. GM, growth medium; CM, control medium; OM, osteogenic medium. Color images available online at www.liebertpub.com/tea

hESC-MSCs form ectopic vascularized bone in vivo

The ability of the cells to produce a mineralized matrix in vitro, which although indicates osteogenesis, may nevertheless not predict the capacity of the cells to develop bone tissue in vivo. Therefore, the hESC-MSCs were tested for their capacity to survive and to produce human bone tissue after ectopic subcutaneous transplantation in NOD/SCID mice without the use of an osteoconductive carrier or scaffold. The hESC-MSCs were subjected to osteogenic induction of 3, 7, or 14 days in vitro. The injected cells remained within a small well-limited area at the injection site, and the developed tissue showed clearly defined boundaries toward the murine tissue (Fig. 4A), while no teratoma formation was observed (Fig. 4B). As we have previously demonstrated, the parental lines of the hESC-MSCs (I-4 and H9.2) did form teratoma after intramuscular injection into SCID/beige mice.³⁸ After 2 weeks, the comprehensive area of a newly formed mineralized bone matrix was observed in the central region of all implants. The calcified matrix was identified by positive Alizarin red staining of the engineered tissue and relevant controls (Fig. 4A, inset). Furthermore, an extensive expression of osteopontin and osteocalcin was detected by immunofluorescence labeling in sectioned implants (Fig. 4C, D). To verify the human cell identity in 2-week-old bone-like Matrigel implants, implanted cells were positively labeled with antibodies that specifically recognize human MHC class I (Fig. 4C). In addition, human cells were identified by red fluorescence staining of chromosome 1 in the cell nuclei within the newly formed tissue, by in situ hybridization using the whole-chromosome painting technique (Fig. 4C, inset). Empty Matrigel was implanted as well, and did not induce bone-like tissue formation or the formation of blood vessels (Fig. 4E). Vascularization of engineered tissue is critical for cell survival and remodeling at initial stages of implant integration to the host tissue. In line with this demand, an extensive network of blood vessels (266±86 lumens/mm²) was formed when the hESC-MSCs of the feeder-free system were subjected to osteogenic induction for 7 days in vitro (Fig. 4F, G). When the cells were exposed to osteogenic induction for 3 days, the number of blood vessels was reduced significantly to 38±16 lumens/mm² (Fig. 4G). All blood vessels contained non-nucleated erythrocytes, indicating that the blood vessels within the implant anatomized the murine circulation (Fig. 4F). Cells from the coculture system contained significantly less blood vessels (89±35 lumens/mm², p<0.01) when the hESC-MSCs were subjected to osteogenic induction for 7 days in vitro; when exposed to osteogenic induction for 3 days, the implants did not contain any blood vessels (Fig. 4G).

FIG. 4.

Ectopic bone formation in vivo. hESC-MSCs were subjected in vitro to osteogenic differentiation for 3, 7, and 14 days (A–F), and then implanted subcutaneously into NOD/SCID mice using Matrigel. Implants were harvested 2 weeks after transplantation. (A) Ectopic bone formation (black circle) in vivo. Inset: massive mineralization of the bone matrix positively stained for Alizarin red. (B) Undifferentiated hMSCs in a Matrigel implant with no teratoma formation. (C) Immunofluorescence staining of implants revealed extensive expression of osteopontin (green). Positive labeling with human specific MHC class I antibodies (red) as well as in situ hybridization for human chromosome 1 (inset, red), confirming their human identity. Nuclear staining by DAPI (blue). (D) Immunofluorescence staining of implants revealed extensive expression of osteocalcin (red) by the induced hESC-MSCs. Nuclear staining by DAPI (blue). (E) Negative control. Matrigel implant without cells. (F) Extensive network of blood vessels containing erythrocytes (arrows) invaded into the Matrigel implant. (G) Density of erythrocyte-containing blood vessels in feeder-free and coculture systems. **Comparison of matched kinetic point (day 7) between systems (p<0.01). ^#,##Comparison of day 7 versus day 14 within each system (p<0.05, p<0.01, respectively). ^+,++Comparison of day 7 versus day 3 within each system (p<0.05, p<0.01, respectively). Color images available online at www.liebertpub.com/tea

We analyzed VEGF₁₆₅ concentrations in a conditioned medium of hESC-MSCs 7, 14, and 21 days after osteogenic induction in vitro by ELISA. VEGF₁₆₅ concentrations were found to be higher in cells grown in an hMSC osteogenic growth medium than in those grown in the control medium (w/o osteogenic inducers). VEGF₁₆₅ concentrations increased during osteogenesis (Fig. 5A). These findings correlate with the observed mineralization appearance and indicate that VEGF₁₆₅ is maximally secreted at the late phase of osteogenic differentiation, when the matrix is mineralized (Fig. 5B). We further tested whether hESC-MSCs are involved in the regulation of angiogenesis using an in vitro endothelial cell-sprouting assay, wherein spheroids of HUVECs were incubated with either the HUVEC medium in the presence or absence of bFGF, or the conditioned medium of osteogenic differentiated hESC-MSCs in the presence or absence of 10 μM of the broad inhibitor SU6668 (ATP-competitive platelet-derived growth factor receptor [PDGFR], VEGF, and fibroblast growth factor receptor [FGFR] inhibitor). In the presence of bFGF or in the osteogenic hESC-MSC-conditioned medium, the spheroids invaded the gel and formed elongated capillary-like structures 24 h after stimulation (Fig. 5C). In the presence of SU6668, the number and length of sprouts were reduced significantly by ∼30% and 55%, respectively (p<0.01, Fig. 5D[i, ii]). In addition, the osteogenic conditioned medium of hESC-MSCs induced proliferation of HUVECs, whereas the presence of the inhibitor reduced proliferation (Fig. 5E).

FIG. 5.

VEGF secretion, spheroid sprouting assay, and proliferation. (A) VEGF concentration was detected after osteogenic induction by enzyme-linked immunosorbent assay. VEGF was measured in a conditioned medium of hESC-MSCs grown in an hMSC GM or in CM (without osteogenic inducers), or in an OM. VEGF₁₆₅ concentration increased during osteogenesis and was found to be higher in cells grown in an hMSC OM than in those grown in a CM. (B) Increased mineralization during osteogenesis in vitro. (C) Spheroid sprouting assay. HUVEC spheroids were incubated in the absence (negative control) or presence (positive control) of bFGF (5 ng/mL), or with osteogenic conditioned medium of hESC-MSCs in the absence (CM) or presence (CM+SU66668) of 10 μM su6668 inhibitor. (D) Quantitative analysis of HUVEC cell sprouting. (D[i]) Cumulative sprout length (*comparison of cumulative sprout length of spheroids incubated with CM vs. CM+SU6668 or vs. negative control p<0.01; ^#comparison between negative and positive control, p<0.01) (D[ii]) Average number of sprouts per spheroid. (E) HUVEC proliferation assay. HUVECs were seeded and incubated in the presence or absence of bFGF (5 ng/mL, positive and negative control, respectively), or with a conditioned medium of hESC-MSCs in the presence or absence of 10 μM SU6668 inhibitor. The cells were counted 3 and 5 days after initial seeding (*comparison of positive control vs. all other groups, p<0.01). VEGF, vascular endothelial growth factor; HUVEC, human umbilical vein endothelial cells; bFGF, basic fibroblast growth factor.

The kinetics of bone maturation was assessed by immunofluorescence staining of osteocalcin. Implants of hESC-MSCs, after osteogenic induction of 14 days in vitro, were removed after 1, 2, or 4 weeks in vivo. Placenta-derived MSCs served as positive control. After 1 week in vivo, low expression of osteocalcin was detected in sectioned implants, both in hESC-MSCs and placenta-derived MSCs (Fig. 6A, B). However, after 2 weeks in vivo, a significant increase in the expression of osteocalcin was observed. The expression of osteocalcin remained high thereafter, with no additional increase in either hESC-MSCs or placenta-derived MSCs (Fig. 6A, B). Altogether, these findings demonstrate that the kinetics of osteogenic maturation of hESC-MSCs is similar to that of their native tissue MSC counterparts, further highlighting their equivalent MSC characteristics.

FIG. 6.

Kinetics of bone maturation–osteocalcin expression. hESC-MSCs were subjected in vitro to osteogenic differentiation for 14 days and then implanted subcutaneously into NOD/SCID mice using Matrigel. Implants were harvested 1, 2, or 4 weeks after transplantation. Placenta-derived MSCs served as positive control. (A) hESC-MSCs. (B) Placenta-derived MSCs. The expression of osteocalcin was low after 1 week in vivo, both in hESC-MSCs and in placenta-derived MSCs. After 2 weeks in vivo, the expression of osteocalcin increased significantly. The expression of osteocalcin remained high thereafter with no additional increases in either hESC-MSCs or placenta-derived MSCs. Color images available online at www.liebertpub.com/tea

Discussion

In this study, we displayed two simple methods, based on cell sorting, to derive from hESCs a pure population of hMSCs (hESC-MSCs) with high osteogenic potential. The hESC-MSCs exhibited characteristics of hMSCs, including expression of known surface markers and the capability of differentiating into adipocytes, chondrocytes, and osteoblasts in vitro. In addition, the osteogenic differentiated hESC-MSCs secreted osteocalcin and VEGF, induced capillary sprout formation in collagen gels, and formed an ectopic vascular bone when implanted subcutaneously into immunodeficient mice. Previous protocols for the establishment of hESC-MSCs include coculturing with OP9 stromal cells,^10,11 isolation based on manual/selective selection, consecutive passaging,^12–15 or cell sorting.^16,17 However, in some of the protocols, only a small proportion of the cells differentiated into hMSCs, thereby requiring further enrichment of the target population. In others, the population obtained was heterogeneous. In contrast, both systems that were used for derivation of hMSCs from hESCs in the present study, namely feeder free and coculture, with OP9 stromal cells, are easy to perform, do not require expensive growth factors, and exhibit a high degree of reproducibility and homogeneity in the resultant cells. Nevertheless, the feeder-free system demonstrated a greater efficiency in inducing mesenchymal differentiation. Moreover, since differentiation through EBs results in a heterogeneous cell population, and coculture using cells from a different species can lead to contamination from animal products, the feeder-free system is highly relevant to the development of techniques for clinical applications.

The differentiation potential of hESC-MSCs toward the osteogenic lineage was demonstrated by secretion of a mineralized matrix and by the expression of osteoblast-associated genes in vitro. The observation that the hESC-MSCs did not express the pluripotent genes Oct3/4 and Nanog implies the derivation of the osteoblasts from a restricted cell lineage that lacks undifferentiated cells.

In the few studies that have demonstrated the ability of hESC-MSCs to form bone in vivo, the cells were mixed and implanted with an osteoconductive scaffold, which further stimulates osteogenic differentiation of implanted hMSC post-transplantation.^9,14,18–20 The hESC-MSCs generated in the current study formed new vascular bone-like tissue when implanted in vivo, without the need of any osteo-inducing scaffold. The implants exhibited bone-like morphology and mineralized matrix. The cells also secreted noncollagenous bone proteins such as osteopontin and osteocalcin. Osteocalcin secretion by the differentiated hESC-MSCs was detected in vitro as well.

The possibility of teratoma formation has been a major obstacle to the clinical use of hESCs. However, hESC-MSCs, unlike their parental line, did not induce teratoma formation, indicating phenotypic stability and highlighting their safety. Moreover, detection of human cells in the newly formed bone matrix, both by antibodies that specifically recognize human MHC class I and by in situ hybridization, confirmed the fate of the transplanted human cells as the bone-forming cells, which are able to further survive and remodel after transplantation. However, their capacity to regenerate impaired bone should be further examined in bone repair models. VEGF and bFGF, potent angiogenic stimulators, have been shown to play important roles in stimulating the restoration of blood flow at the site of a fracture, thus promoting the bone repair process.³¹ In accordance with a previous study,³¹ we have demonstrated that in the process of osteogenic differentiation, hESC-MSCs first secrete low levels of VEGF, and subsequently increased levels during terminal differentiation. Mineralization appearance increased during osteogenesis as well indicates that VEGF₁₆₅ is maximally secreted at the late phase of osteogenic differentiation, when the matrix is mineralized. The correlation between secreted levels of VEGF and the degree of matrix mineralization suggests a direct role of VEGF-A in the mineralization process.

All blood vessels observed in the implants and adjacent to the newly formed bone contained un-nucleated erythrocytes, indicating that the vessels that invaded the implant anatomized the murine circulation, similarly to another Matrigel implant model, in which human bone marrow-derived MSCs were shown to secrete VEGF and to promote remodeling of newly formed human blood vessels in the implant.³⁹ In addition, we also observed reduction in the number of blood vessels similarly to this study, in which the durability of blood vessels as examined in vivo was found to decrease over time.³⁹ These findings suggest different kinetics of VEGF and other angiogenic factors in vivo over culture conditions.

The capability of MSCs to recruit murine blood vessels into the implant can be attributed to the secretion of angiogenic factors, including VEGF, bFGF, and platelet-derived growth factor, as was previously reported.^36,39,40 In agreement, we have demonstrated in vitro VEGF secretion by hESC-MSCs, and capillary sprout formation originating from HUVEC spheroids that were stimulated with an osteogenic differentiated hESC-MSC-conditioned medium. The cell sprouting was significantly reduced by the use of ATP-competitive PDGFR, VEGF, and FGFR inhibitor, implying that implanted osteogenic hESC-MSCs participate in regulation of neovascularization in the newly formed bone via secretion of paracrine angiogenic factors.

To apply differentiated hESCs in clinically relevant scenarios, a large number of cells are required. The high expansion rate of hESC-MSCs and the fact that freeze and thaw cycles do not affect cell features enable high survival rates. In addition, it has been shown that hESC-MSCs may be more immunoprivileged than adult hMSCs, which is another advantage of their clinical application.^13,16 The generation of induced pluripotent stem cells (iPSCs) from adult somatic cells by reprogramming is a significant scientific breakthrough. iPSCs hold a great promise, particularly because they provide an alternative source of functional MSCs for tissue repair, without raising the ethical and immunological issues integral to the use of hESCs. Graft rejection, another problem concerning hESCs, can be overcome with the use of iPSC-derived MSCs for autologous transplantations. Nevertheless, there are several challenges that need to be overcome before iPSCs and their derivatives can be used clinically, such as the low efficiency of iPSC generation and the safety of viral integration of transgenes.

In conclusion, the cell culture protocol presented herein affords a high yield of pure populations of hESC-MSCs with a high osteogenic capability. The osteoblasts derived from these hESC-MSCs were able to form a new mineralized and vascularized bone matrix in vivo. Once the functionality of these derived cells will be thoroughly examined in a bone repair model, the potential of the described method as a source of osteoblasts for bone tissue engineering may be realized.

Footnotes

Acknowledgments

This work was performed at the Berlin Family Laboratory for Stem Cell and Tissue Regeneration Research at the Sohnis and Forman Families Center for Stem Cell and Tissue Regeneration Research. We thank Dr. Shoham Shivtiel for assistance with in vivo experiments. This study was partially supported by the Technion R&D Foundation. J.I.-E. holds the Sylvia and Stanley Shirvan Chair in Cell and Tissue Regeneration Research at the Technion-Israel Institute of Technology.

Disclosure Statement

No competing financial interests exist.

References

Friedenstein

A.J.

et al. Stromal cells responsible for transferring the microenvironment of the hemopoietic tissues. Cloning in vitro and retransplantation in vivo. Transplantation, 17:331. 1974.

Javazon

E.H.

, Beggs

K.J.

, Flake

A.W.

Mesenchymal stem cells: paradoxes of passaging. Exp Hematol, 32:414. 2004.

Pittenger

M.F.

et al. Multilineage potential of adult human mesenchymal stem cells. Science, 284:143. 1999.

Horwitz

E.M.

et al. Isolated allogeneic bone marrow-derived mesenchymal cells engraft and stimulate growth in children with osteogenesis imperfecta: implications for cell therapy of bone. Proc Natl Acad Sci U S A, 99:8932. 2002.

Horwitz

E.M.

et al. Clinical responses to bone marrow transplantation in children with severe osteogenesis imperfecta. Blood, 97:1227. 2001.

Bigdeli

et al. Superior osteogenic capacity of human embryonic stem cells adapted to matrix-free growth compared to human mesenchymal stem cells. Tissue Eng Part A, 16:3427. 2010.

Kassem

, Abdallah

B.M.

Human bone-marrow-derived mesenchymal stem cells: biological characteristics and potential role in therapy of degenerative diseases. Cell Tissue Res, 331:157. 2008.

Zuk

P.A.

et al. Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng, 7:211. 2001.

Arpornmaeklong

et al. Expansion and characterization of human embryonic stem cell-derived osteoblast-like cells. Cell Reprogram, 12:377. 2010.

10.

Trivedi

, Hematti

Simultaneous generation of CD34+ primitive hematopoietic cells and CD73+ mesenchymal stem cells from human embryonic stem cells cocultured with murine OP9 stromal cells. Exp Hematol, 35:146. 2007.

11.

Barberi

et al. Derivation of multipotent mesenchymal precursors from human embryonic stem cells. PLoS Med, 2:e161. 2005.

12.

Olivier

E.N.

, Rybicki

A.C.

, Bouhassira

E.E.

Differentiation of human embryonic stem cells into bipotent mesenchymal stem cells. Stem Cells, 24:1914. 2006.

13.

Trivedi

, Hematti

Derivation and immunological characterization of mesenchymal stromal cells from human embryonic stem cells. Exp Hematol, 36:350. 2008.

14.

Harkness

et al. Selective isolation and differentiation of a stromal population of human embryonic stem cells with osteogenic potential. Bone, 48:231. 2010.

15.

Karlsson

et al. Human embryonic stem cell-derived mesenchymal progenitors-potential in regenerative medicine. Stem Cell Res, 3:39. 2009.

16.

Lian

et al. Derivation of clinically compliant MSCs from CD105+, CD24-differentiated human ESCs. Stem Cells, 25:425. 2007.

17.

Barberi

et al. Derivation of engraftable skeletal myoblasts from human embryonic stem cells. Nat Med, 13:642. 2007.

18.

Arpornmaeklong

et al. Phenotypic characterization, osteoblastic differentiation, and bone regeneration capacity of human embryonic stem cell-derived mesenchymal stem cells. Stem Cells Dev, 18:955. 2009.

19.

Bielby

R.C.

et al. In vitro differentiation and in vivo mineralization of osteogenic cells derived from human embryonic stem cells. Tissue Eng, 10:1518. 2004.

20.

Koob

et al. Bone formation and neovascularization mediated by mesenchymal stem cells and endothelial cells in critical-sized calvarial defects. Tissue Eng Part A, 17:311. 2011.

21.

Carlevaro

M.F.

et al. Vascular endothelial growth factor (VEGF) in cartilage neovascularization and chondrocyte differentiation: auto-paracrine role during endochondral bone formation. J Cell Sci, 113,Pt 1:59. 2000.

22.

Carano

R.A.

, Filvaroff

E.H.

Angiogenesis and bone repair. Drug Discov Today, 8:980. 2003.

23.

Gerber

H.P.

, Ferrara

Angiogenesis and bone growth. Trends Cardiovasc Med, 10:223. 2000.

24.

Cross

M.J.

et al. VEGF-receptor signal transduction. Trends Biochem Sci, 28:488. 2003.

25.

Ferrara

, Gerber

H.P.

, LeCouter

The biology of VEGF and its receptors. Nat Med, 9:669. 2003.

26.

Neufeld

et al. Vascular endothelial growth factor (VEGF) and its receptors. FASEB J, 13:9. 1999.

27.

Maes

et al. Impaired angiogenesis and endochondral bone formation in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Mech Dev, 111:61. 2002.

28.

Street

et al. Vascular endothelial growth factor stimulates bone repair by promoting angiogenesis and bone turnover. Proc Natl Acad Sci U S A, 99:9656. 2002.

29.

Boiret

et al. Characterization of nonexpanded mesenchymal progenitor cells from normal adult human bone marrow. Exp Hematol, 33:219. 2005.

30.

Peng

et al. Synergistic enhancement of bone formation and healing by stem cell-expressed VEGF and bone morphogenetic protein-4. J Clin Invest, 110:751. 2002.

31.

Mayer

et al. Vascular endothelial growth factor (VEGF-A) expression in human mesenchymal stem cells: autocrine and paracrine role on osteoblastic and endothelial differentiation. J Cell Biochem, 95:827. 2005.

32.

Amit

, Itskovitz-Eldor

Derivation and spontaneous differentiation of human embryonic stem cells. J Anat, 200,Pt 3:225. 2002.

33.

Thomson

J.A.

et al. Embryonic stem cell lines derived from human blastocysts. Science, 282:1145. 1998.

34.

Itskovitz-Eldor

et al. Differentiation of human embryonic stem cells into embryoid bodies compromising the three embryonic germ layers. Mol Med, 6:88. 2000.

35.

Kern

et al. Vasohibin inhibits angiogenic sprouting in vitro and supports vascular maturation processes in vivo. BMC Cancer, 9:284. 2009.

36.

Dar

et al. Multipotent vasculogenic pericytes from human pluripotent stem cells promote recovery of murine ischemic limb. Circulation, 125:87. 2012.

37.

Amit

et al. Human feeder layers for human embryonic stem cells. Biol Reprod, 68:2150. 2003.

38.

Amit

et al. Clonally derived human embryonic stem cell lines maintain pluripotency and proliferative potential for prolonged periods of culture. Dev Biol, 227:271. 2000.

39.

Melero-Martin

J.M.

et al. Engineering robust and functional vascular networks in vivo with human adult and cord blood-derived progenitor cells. Circ Res, 103:194. 2008.

40.

Lian

et al. Functional mesenchymal stem cells derived from human induced pluripotent stem cells attenuate limb ischemia in mice. Circulation, 121:1113. 2010.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.11 MB

0.16 MB

0.09 MB

0.02 MB

0.00 MB

Efficient Engineering of Vascularized Ectopic Bone from Human Embryonic Stem Cell–Derived Mesenchymal Stem Cells

Abstract

Introduction

Materials and Methods

Cell culture and fluorescence-activated cell sorting

Cell expansion and growth curves

Flow cytometry analyses

Adipogenic differentiation

Osteogenic differentiation

Chondrogenic differentiation

Myogenic differentiation

Reverse transcription polymerase chain reaction

Concentrations of VEGF165 in hMSC-conditioned medium

Quantification of osteocalcin in hMSC-conditioned medium

Spheroid sprouting assay

Proliferation assay

Osteogenic Matrigel implants

Teratoma formation

Histology immunofluorescence and immunohistochemistry

Fluorescence in situ hybridization

Microvessel density analysis

Statistical analysis

Karyotype

Results

Isolation of hESC-MSCs grown in coculture with OP9 stromal cells or on Matrigel-coated plates

Characteristics of hESC-MSCs

Differentiation of hESC-MSCs into adipocytes and chondrocytes in vitro

Differentiation of hESC-MSCs into osteoblasts is accompanied by secretion of osteocalcin

hESC-MSCs form ectopic vascularized bone in vivo

Discussion

Footnotes

Acknowledgments

Disclosure Statement

References

Supplementary Material

Concentrations of VEGF₁₆₅ in hMSC-conditioned medium