Embryonic stem (ES) cells have the ability to self-replicate and differentiate into cells from all three germ layers, holding great promise for tissue regeneration applications. However, controlling the differentiation of ES cells and obtaining homogenous cell populations still remains a challenge. We hypothesize that a supportive three-dimensional (3D) environment provides ES cell-derived cells an environment that more closely mimics chondrogenesis in vivo. In the present study, the chondrogenic differentiation capability of ES cell-derived embryoid bodies (EBs) encapsulated in poly(ethylene glycol)-based (PEG) hy-drogels was examined and compared with the chondrogenic potential of EBs in conventional monolayer culture. PEG hydrogel-encapsulated EBs and EBs in monolayer were cultured in vitro for up to 17 days in chondrogenic differentiation medium in the presence of transforming growth factor (TGF)-β1 or bone morphogenic protein-2. Gene expression and protein analyses indicated that EB-PEG hydrogel culture upregulated cartilage-relevant markers compared with a monolayer environment and induction of chondrocytic phenotype was stimulated with TGF-β1. Histology of EBs in PEG hydrogel culture with TGF-β1 demonstrated basophilic extracellular matrix deposition characteristic of neocartilage. These findings suggest that EB-PEG hydrogel culture, with an appropriate growth factor, may provide a suitable environment for chondrogenic differentiation of intact ES cell-derived EBs.
Development of the vertebrate skeleton is a multistep process that involves lineage commitment of mesenchymal cells, migration of these cells to the site of skeletogenesis, mesenchymal-epithelial interactions that result in cell condensation, and differentiation of chondroblasts or osteoblasts [1, 2]. Tissue-specific progenitor cell microenvironments are required for chondrogenesis and cartilage formation in the developing limb. These environments may include specific cell-cell interactions, cell-matrix interactions, appropriate biological signals, and appropriate mechanical stresses to maintain morphological integrity of differentiating cells. Such environments can be mimicked using a tissue engineering approach by incorporation of necessary biological signals in suitable biomaterial scaffolds.
Embryonic stem (ES) cells, derived from the inner cell mass of the blastocyst, are versatile cells that are capable of differentiating into cells of all three germ layer lineages [3–5]. These cells are capable of unlimited symmetrical self-renewal, thus providing an unlimited cell source for tissue-engineering applications. However, one of the challenges in stem cell regenerative therapy is to control the differentiation process . Differentiation of ES cells through embryoid bodies (EBs) parallels embryonic development, EBs recapitulating early embryonic developmental phases . Recently, biological signals in the form of growth factors were shown to induce EB differentiation toward the chondrogenic lineage [8–10]. These attributes suggest that ES cells may represent a useful cell source for cartilage tissue engineering.
Recent tissue engineering studies have shown the potential use of ES cells for tissue regeneration in three-dimensional (3D) scaffold systems with appropriate growth factors [11–13]. Three-dimensional culture conditions more closely resemble the cell environment in developing tissue, and recent reports have indicated distinct cellular behavior in 3D culture that is not present in standard monolayer culture [14, 15]. Hydrogels are water-insoluble materials capable of absorbing large volumes of aqueous solution, creating 3D microenvironments for the encapsulated cells and tissue-like water content with efficient nutrient and waste exchange. Poly(ethylene glycol)-based (PEG) photopolymerizing hydrogels have demonstrated feasibility for cartilage tissue engineering using chondrocytes and mesenchymal stem cells [16–18].
In the present study, EBs encapsulated in PEG hydrogels and EBs plated on conventional monolayer were incubated in chemically defined chondrogenic medium in the presence or absence of transforming growth factor (TGF)-β1 or bone morphogenic protein (BMP)-2. Encapsulated EBs in PEG hydrogels demonstrated basophilic extracellular matrix depositions characteristic of neocartilage when supplemented with TGF-β1 and expressed numerous chondrocyte-specific genes. Moreover, changes in EB morphology in PEG hydrogel conditions modulated the ES cell response to growth factors and chondrogenic differentiation.
Materials and Methods
Mouse ES Cell Culture
ES-D3 GL cell line (American Type Culture Collection, Manassas, VA, http://www.atcc.org) was cultivated on a feeder layer of mitomcycin C-inactivated mouse embryonic fibroblasts with cultivation medium consisting of Dulbecco's modified Eagle's medium (DMEM) (Gibco, Gaithersburg, MD, http://www.invitrogen.com) supplemented with 15% fetal bovine serum (FBS) (HyClone, Logan, UT, http://www.hyclone.com), 2 mM L-glutamine (Gibco), 5 × 10−5 M β-mercaptoethanol (Sigma-Aldrich, Saint Louis, http://www.sigmaaldrich.com), nonessential amino acids (Gibco), and leukemia inhibitory factor (LIF) (Sigma-Aldrich), as described previously . EBs were formed in liquid suspension culture. In brief, ES cells were dissociated using 0.05% trypsin-EDTA (Gibco) for 5 minutes, and the cells were pipetted several times to obtain single-cell suspensions. ES cell solution (10 ml with ∼ 104 cells) was transferred to a Petri dish (Fisher Scientific, Hampton, NH, https://www1.fishersci.com) and cultured for 5 days for EB formation. By day 5 in suspension culture, the average cell number per EB was 3,239 cells, with a range of 1,436 to 4,524 cells per EB.
Photoencapsulation of EBs in Hydrogel
Poly(ethylene glycol)-diacrylate (PEGDA) (Nektar Therapeutics, Huntsville, AL, http://www.nektar.com) solution (10% wt/vol) was prepared with phosphate-buffered saline (PBS) according to previously described protocol . The photoinitiator, Igracure 2959 (Ciba Specialty Chemicals, Tarrytown, NY, http://www.cibasc.com), was added to the PEGDA solution and mixed thoroughly to make a final concentration of 0.05% (wt/vol). Immediately before photoencapsulation, EBs were resuspended in the polymer solution. One hundred microliters of EB-PEGDA solution containing 300 EBs or ∼ 106 cells was transferred to a cylindrical mold and exposed to 365-nm light with intensity of 4.5 mW/cm2 (Glowmark Systems, Upper Saddle River, NJ) for 5 minutes to complete the gelation. EB-PEG hydrogels were then removed from their molds and incubated for up to 17 days in 12-well plates with chondrogenic-conditioned medium containing 5% FBS (Hyclone) in addition to high-glucose DMEM supplemented with 10.7 M dexamethasone, 50 μg/ml ascorbate-2-phosphate, 40 μg/ml proline, 100 μg/ml sodium pyruvate, and 50 mg/ml ITS-Premix (Collaborative Biomedical [San Jose, CA, http://www.bdbiosciences.com]: 6.25 ng/ml insulin, 6.25 mg transferrin, 6.25 ng/ml selenious acid, 1.25 mg/ml bovine serum albumin, and 5.35 mg/ml linoleic acid), as described previously . To study the growth factor-induced chondrogenesis of EBs, BMP-2 (25 ng/ml) or TGF-β1 (10 ng/ml) (Research Diagnostics, Inc., Flanders, NJ, http://www.researchd.com) was added to the chondrogenic medium. Control cultures were maintained in mouse ES (mES) cell medium without LIF.
EBs in Monolayer
EBs (∼10 EBs/cm2) were plated onto gelatin (0.1%)-coated six-well plates and cultured up to 17 days in chondrogenic medium for comparison with encapsulated EBs in PEG hydrogel. In separate cultures, either BMP-2 (25 ng/ml) or TGF-β1 (10 ng/ml) was added to the chondrogenic medium for comparison with growth factor-treated EBs in PEG hydrogel culture. Control cultures were maintained in mES cell medium without LIF.
Cell Viability Assay
EB-PEG hydrogels were cut into 1-mm-thick samples to assess cell viability using Calcein AM/EthD-I Live/Dead Viability/Cytotoxicity Kit (Molecular Probes, Eugene, OR, http://probes.invitrogen.com). Calcein AM specifically stains live cells via their intracellular esterase activity, and EthD-I specifically stains dead cells that have lost plasma membrane integrity. Reagents were diluted according to the manufacturer's protocol and incubated for 30 minutes in serum-free medium.
Cell Metabolic Activity Assay
Culture medium was removed, and WST-1 (Roche Molecular Biochemicals, Hannheim, Germany, https://www.roche-applied-science.com) was added. WST-1-derived precipitate, produced by metabolically active cells in the culture, was quantified by spectrophotometry at A450.
Histology and Immunostaining
EB-PEG hydrogels were fixed overnight in 4% paraformaldehyde in PBS (pH 7.4) at 4°C and transferred to 70% ethanol until processing. Hydrogels were embedded in paraffin and cut into 5-μm sections that were stained with hematoxylin and eosin, Safraonin-O/fast green, and Masson's trichrome. Immunohistochemistry was performed with a Histostain-SP kit (Zymed Laboratories, San Francisco, http://www.zymed.com). Polycolonal rabbit antibodies against mouse types I, II, X collagen (Research Diagnostics), and osteocalcin (Biotrend, Destin, FL, http://www.biotrend.com) were used with 1:40 to 1:100 dilutions.
EB-PEG hydrogels were collected at different time points, lyophilized, and digested with papain. DNA assay was performed as previously described . Determination of proteoglycan content was characterized by chondroitin sulfate using dimethylmethylene blue (DMMB) spectrophotometric assay at A525, as previously described . Chondroitin sulfate C (Sigma-Aldrich), dissolved in deionized water, was used to generate the standard curve.
Reverse Transcription-Polymerase Chain Reaction and Real-Time Polymerase Chain Reaction
Total RNAs were extracted from three EB-PEG hydrogels (combined) for encapsulated EBs and two wells from a six-well plate for monolayer-cultured EBs using TRIzol (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) following the manufacturer's instruction. Two micrograms of total RNA per 20 μl of reaction volume was reverse transcribed into cDNA using the SuperScript First-Strand Synthesis System (Invitrogen). Real-time polymerase chain reactions (PCRs) were performed and monitored using the SYBR Green PCR Mastermix and the ABI Prism 7700 Sequence Detection System (Perkin Elmer/Applied Biosystems, Rotkreuz, Switzerland, http://www.perkinelmer.com). cDNA samples (2 μl for total volume of 25 μl per reaction) were analyzed for gene of interest and for the reference gene glyceraldehyde-3-phosphate-dehydrogenase. The level of expression of each target gene was then calculated as −2ΔΔCt as previously described . Each sample was repeated at least three times for each gene of interest. Reverse transcription-PCR was performed at 95°C for 2 minutes followed by 34 cycles of 30-second denaturation at 95°C, 30 seconds of annealing at the primer-specific temperature, and 1 minute of elongation at 72°C. PCR products were verified by electrophoresis. The PCR primers are listed in Table 1.
Data are expressed as mean ± standard deviation. Statistical significance was determined by analysis of variance (single factor) with p < .05.
EBs were successfully encapsulated in PEG-based hydrogels. Approximately 300 EBs (averaging 3,239 cells/EB) were photoencapsulated in PEG-based polymer, resulting in approximately 1 million cells per construct (Fig. 1A). Initial gel clarity allowed visualization of the EBs inside the hydrogel (Fig. 1B). EBs remained intact and were homogenously distributed throughout the hydrogel after the photopolymerization. Most cells within the EBs were viable immediately after encapsulation, whereas some dead cells were sparsely distributed throughout the EBs (Fig. 1C).
PEG Hydrogel Culture Upregulates Chondrogenic Gene Expression in EBs
Expression of cartilage-specific genes varied significantly in monolayer and EB-PEG hydrogel cultures. EBs expressed cartilage-specific genes in both monolayer and PEG hydrogel cultures in chondrogenic conditions, but the expression of cartilage-relevant markers was dramatically induced in PEG hydrogel culture. A significant increase in gene expression for aggrecan, link protein, and Sox9 was observed during PEG hydrogel culture of EBs compared with monolayer culture (Figs. 2B–2D). TGF-β1 treatment resulted in the greatest increase of aggrecan and link protein expression compared with other medium conditions. In addition to chondrogenic genes, the hypertrophic chondrocyte marker type X collagen and osteogenic markers cbfa1 and osteocalcin were significantly upregulated in PEG hydrogel culture compared with the monolayer counterpart in all medium conditions (Figs. 2E–2G). However, PEG hydrogel culture with TGF-β1 treatment resulted in minimal or no increase in expression of osteogenic markers compared with the monolayer counterpart.
Type II collagen is found specifically in articular cartilage and is synthesized as a procollagen in two forms (IIA and IIB), generated by differential splicing of the gene primary transcript. Type IIA collagen is expressed in juvenile or prechondrogenic cells, whereas type IIB collagen is expressed in adult or differentiated chondrocytes. The juvenile splice variant of type IIA collagen was predominantly expressed in monolayer culture after 17 days of culture, with enhanced expression in the presence of TGF-β1. However, the adult splice variant of type IIB collagen was predominantly expressed in PEG hydrogel culture (Fig. 2A). Treatment of hydrogels with BMP-2 potentiated the expression of type IIB collagen transcript but to a lesser degree than TGF-β1-treated hydrogels. Hydrogels treated with TGF-β1 were the only conditions where the type IIB collagen was solely expressed without the juvenile splice variant.
TGF-β1-Dependent Chondrogenic Differentiation of EBs in PEG Hydrogel Culture
TGF-β1 treatment upregulated several cartilage-specific markers after 7 and 17 days of culture, suggesting that PEG hydrogel culture with TGF-β1 is an efficient culture condition to induce chondrogenesis of EBs. TGF-β1 treatment resulted in upregulation of Sox9 and almost an eightfold increase of aggrecan and link protein expressions compared with chondrogenic medium and chondrogenic medium with BMP-2 after 17 days of culture (Fig. 3A). The expression of the transcription factor scleraxis increased 15-fold in the presence of TGF-β1 (Fig. 3A). In contrast, osteocalcin expression only moderately increased after TGF-β1 treatment of 17 days compared with the other culture conditions (p < .05). TGF-β1 treatment in PEG hydrogel culture resulted in reduced or no change in the expression of type X collagen and cbfa1, respectively (Fig. 3A).
The time-course analysis of chondrogenic marker gene expression further validated the chondrogenic tissue lineage progression of EBs treated with TGF-β1.
Type IIA collagen was expressed as early as day 1 of culture and reached the highest levels between days 5 and 7. After day 10, the expression of collagen IIA by EBs decreased, reaching minimal levels by day 15. However, collagen IIB, which was undetected until day 7 of culture, increased over the culture period (Fig. 3B). Sox9, a transcription factor involved in regulation of type II collagen, and the cartilage proteoglycan link protein were also minimally expressed at earlier time points but increased over time in PEG hydrogel culture with TGF-β1.
Increase in chondrogenic (a mesodermal tissue) gene expression was specific, with no increase in expression of markers for ectodermal or endodermal lineages observed. Exposure to TGF-β1 in PEG hydrogel culture promoted loss of phenotypic markers relevant to ectodermal and endodermal lineages. AFP, an endodermal cell marker, and nestin, an ectodermal cell marker, were minimally expressed during the culture period. Expression of cytokeratin K18 gradually decreased over the first 10 days in culture and was completely absent after day 10.
Morphological analysis confirmed the results of chondrogenic gene expression and formation of tissue structures characteristic of cartilage (Fig. 3C). Histology revealed that cartilaginous EBs were larger in size and contained round cells surrounded by large amounts of extracellular matrix. In contrast, smaller EBs demonstrated a higher cell density due to low amount of matrix (Fig. 3C, arrow). Smaller EBs often contained large numbers of dead cells (Fig. 3D, arrow).
Tissue Morphology and Matrix Synthesis of EBs Encapsulated in Hydrogels
Metabolic activity and matrix secretion in hydrogels were stimulated by growth factors. TGF-β1 treatment resulted in the highest rate of GAG production per cell and resulted in a significant increase in GAG accumulation over time (0.7% dry weight [dw]; Fig. 4A). Measurement of metabolic activity indicated that EBs exposed to TGF-β1 ncreased metabolic activity, whereas the other three groups maintained similar levels (Fig. 4B). There was no significant difference in DNA contents among the three culture medium conditions, indicating that the increase in metabolic activity in TGF-β1-treated hydrogels was not due to an increase in cell number but due to higher metabolic activity per cell (Fig. 4C).
Histological analysis demonstrated TGF-β1-dependent cartilage-like tissue formation of EBs encapsulated in the hydrogels (Fig. 5). After 1 week of culture, clusters of cells were present with little extracellular matrix (Figs. 5A–5C). After 17 days of culture, the chondrogenic and BMP-2-treated hydrogels demonstrated little change in size or morphology whereas the TGF-β1-treated hydrogels contained cells that had a round morphology and were surrounded by significant amounts of extracellular matrix, with a typical cartilage-like tissue structure (Figs. 5D–5I). Approximately 10% of EBs in the TGF-β1-treated hydrogels formed large cartilaginous tissue structures. The cartilage tissue structure was further elaborated by confirmation of proteoglycans and collagen in the extracellular matrix (Figs. 5J–5M). The extracellular matrix of EBs cultured in chondrogenic, BMP-2, and TGF-β1 medium stained positive for types I, II, and X collagen to varying degrees. EBs incubated in chondrogenic conditions (Figs. 6A, 6D) or with BMP-2 (Figs. 6B, 6E) stained weakly for types I and II collagen. TGF-β1-treated EBs showed strong immunoreactivity for types I and II collagen, predominantly within the pericellular matrix of chondrocytes in the cartilaginous tissue structures (Figs. 6C, 6F). Hydrogels in TGF-β1, BMP-2, and chondrogenic culture produced 61.1%, 33.9%, and 27.8% EBs that stained positive for type II collagen (greater than 50% of the EB area). Type X collagen stained strongest in EBs cultured in chondrogenic medium but less intensely in hydrogels cultured with BMP-2 or TGF-β1 (Figs. 6G–6I). Osteocalcin was not detected in any of the samples (Figs. 6J–6L).
PEG-based hydrogels provide a 3D, nonadhesive environment for encapsulating cells, enabling the maintenance of intact EBs and round cell morphology. We previously demonstrated the PEG-based hydrogels to be supportive for cartilage formation by chondrocytes and bone marrow-derived MSCs [16, 19]. This study extends the application of hydrogels to chondrogenic differentiation of ES cells.
Chondrogenesis is a multistep process, and chondrogenic commitment of mesenchymal progenitor is under the control of numerous transcription factors [23–26]. Musculoskeletal-related transcription factors were expressed during EB differentiation in the hydrogels, including those for cartilage (Sox9) and bone (cbfa1). Compared with conventional monolayer culture, PEG hydrogel culture upregulated the expression of Sox9 and cbfa1 in EBs, regardless of culture medium. This suggests that PEG hydrogel culture provides an environment that promotes osteogenic and chondrogenic differentiation of ES cell-derived EBs. Upregulation of aggrecan in PEG hydrogel compared with monolayer conditions was due to differences in culture systems because we did not observe cell density-dependent upregulation of aggrecan in TGF-β1-treated monolayer culture (data not shown). Several differences between monolayer and PEG hydrogel cultures may account for the observations, including cellular morphology as well as the microenvironment surrounding the EBs. Three-dimensional culture of intact EBs may promote closer cell-cell interactions, entrapment of secreted extracellular matrix, and maintenance of spherical cellular morphologies. The 3D microenvironment surrounding the EBs, due to its limited space, may also inhibit the proliferation and differentiation of endothelial cells or other migratory cells from EBs, which are predominant in monolayer culture of EBs. In the present study, TGF-β1 exposure upregulated type IIA collagen expression in monolayer culture, whereas type IIB collagen expression was upregulated in PEG hydrogel culture, indicating that the microenvironment of encapsulated EBs may influence the EBs' response to growth factors and expression of alternative splicing of type II collagen (Fig. 2A).
Biomaterial choice may also influence ES cell differentiation. Recently, Tanaka et al.  found no significant chondrogenic differentiation of EBs encapsulated in alginate over conventional EB plating. Several factors, including differences in physical properties and chemistry of the PEG-based and alginate hydrogels, may account for the differences in EB differentiation and tissue development in the different materials. Furthermore, it remains to be elucidated how physical cues such as stiffness of a hydrogel, in combination with biological cues, may influence chondrogenic differentiation and tissue formation .
TGF-β1 and BMP-2 have significant effects on chondrogenic differentiation of human and mouse embryonic and adult stem cells [8, 20, 29, 30] and may act by distinct mechanisms to regulate chondrogenic cell fate [31–35]. Kramer et al.  demonstrated that chondrogenesis can be stimulated in monolayer culture by BMP-2 and BMP-4, whereas TGF-β1 treatment results in unaltered or slightly reduced chondrogenesis of ES-derived EBs. In contrast, a strong chondrogenic response to TGF-β1 was observed in embryonic cells cultured in 3D micromass [20, 30] and on polymer scaffolds . These results are consistent with our observations that BMP-2 promoted aggrecan expression more effectively in monolayer culture (supplemental Fig. 1), whereas TGF-β1 stimulated the highest aggrecan expression in PEG hydrogels. TGF-β1 treatment alone in hydrogels promoted the formation of extracellular matrix with morphological similarities to developing neocartilage, such as round cells embedded in lacunae. TGF-β1 also induced gene expression of extracellular matrix proteins such as aggrecan and type IIB collagen more effectively than BMP-2 in PEG hydrogel culture. Combination of TGF-β1 and BMP-2 did not result in enhanced cartilaginous matrix production (supplementary Fig. 2). Therefore, TGF-β1 alone was sufficient to promote chondrogenesis of EBs in a PEG hydrogel culture system.
Despite the fact that TGF-β1 induced early chondrogenic activation in the PEG hydrogel condition, real-time PCR of cbfa1 and type X collagen expression indicated that hypertrophic differentiation and further differentiation into osteogenic lineages was limited with TGF-β1 treatment. Results from Nakayama et al.  indicated that TGF-βhad an early positive function and a later negative function on chondrogenesis, and early treatment of BMP negatively influenced the chondrogenesis of ES cell-derived mesodermal progenitor cells . We observed that BMP-2 or TGF-β1 treatment resulted in steep downregulation of type X collagen and no significant upregulation of cbfa1. Strong staining of type X collagen was seen only when growth factors were not present, confirming that BMP-2 and TGF-β1 might play an inhibitory role in hypertrophic differentiation. On the contrary, osteocalcin expression was upregulated in growth factor-treated constructs. However, no significant osteocalcin protein was detected in cartilaginous EBs.
Proteoglycan (GAG) content in engineered cartilage varies depending on cell density and may also be related to the time required for cell differentiation. In this study, hydrogels that contained approximately 1 × 106 cells per construct produced approximately 0.7% GAG (dw) after treatment with TGF-β1 for 17 days. This is significantly lower than the amount reported by Freed et al.  and Williams et al.  using bovine chondrocytes in PGA (10% GAG, dw) and goat mesenchymal stem cells in hydrogels (3.5% GAG, dw). The difference in GAG production may be due to species variation as well as the time required for differentiation of the ES cells into chondrocytes. EBs did not consistently undergo chondrogenesis, as indicated by heterogeneous type II collagen staining and number of cartilaginous EBs. However, it is possible that the regions of chondrogenesis produced similar amounts of GAG compared with chondrocytes or MSCs.
We hypothesize that both cell selection and differentiation processes were present in the encapsulated EBs exposed to TGF-β1. Culture conditions may have selected for the growth of a subpopulation of ES cells with the capability for mesenchymal differentiation. We observed an increase in chondrogenic markers such as aggrecan and type II collagen that are present in cartilaginous cells. Low or decreasing levels of expressions of ectodermal and endodermal cell lineage markers also suggest that the capability of cells in EBs to differentiate into different lineages may be limited after chondrogenic differentiation.
Table Table 1.. Sequences of primers and conditions used in reverse transcription-polymerase chain reaction (RT-PCR) and real-time PCR
This study was supported by the Johns Hopkins University (JHU)-Technion Joint Program and the Whitaker Foundation. The authors are grateful to Dr. Shyni Varghes (JHU) and Angela Ferran (JHU) for their critical review and technical assistance.
The authors indicate no potential conflicts of interest.