Osteogenesis from Human Induced Pluripotent Stem Cells: An In Vitro and In Vivo Comparison with Mesenchymal Stem Cells

Abstract

The purpose of this study was to examine the in vitro and in vivo osteogenic potential of human induced pluripotent stem cells (hiPSCs) against that of human bone marrow mesenchymal stem cells (hBMMSCs). Embryoid bodies (EBs), which were formed from undifferentiated hiPSCs, were dissociated into single cells and underwent osteogenic differentiation using the same medium as hBMMSCs for 14 days. Osteoinduced hiPSCs were implanted on the critical-size calvarial defects and long bone segmental defects in rats. The healing of defects was evaluated after 8 weeks and 12 weeks of implantation, respectively. Osteoinduced hiPSCs showed relatively lower and delayed in vitro expressions of the osteogenic marker COL1A1 and bone sialoprotein, as well as a weaker osteogenic differentiation through alkaline phosphatase staining and mineralization through Alizarin red staining compared with hBMMSCs. Calvarial defects treated with osteoinduced hiPSCs had comparable quality of new bone formation, including full restoration of bone width and robust formation of trabeculae, to those treated with hBMMSCs. Both osteoinduced hiPSCs and hBMMSCs persisted in regenerated bone after 8 weeks of implantation. In critical-size long bone segmental defects, osteoinduced hiPSC treatment also led to healing of segmental defects comparable to osteoinduced hBMMSC treatment after 12 weeks. In conclusion, despite delayed in vitro osteogenesis, hiPSCs have an in vivo osteogenic potential as good as hBMMSCs.

Introduction

Bone has the capacity for self-repair without scarring, which is a property uncommon in adult tissues. Most fractures heal spontaneously or with the help of surgical procedures [1]. However, despite the inherent ability of bone to regenerate, there are a number of clinical situations in which complete bone healing fails to occur [2]. Major trauma or malignant tumor resection results in critical-size bone defects (spanning>2 cm), which are not repaired without outside intervention [3].

The most commonly used surgical procedure to promote bone healing in these clinical situations is autogenous bone grafting [4]. Although this method has been the gold standard for treating bone defects or nonunion, there is a limit in the amount of available autologous bone. Moreover, the procedure is also associated with side effects, such as pain at the harvest site [5]. Therefore, allografts obtained from dead donors have been used to replace or supplement autografts. Whereas allografted bone is unlimited in quantity, it is far less osteogenic than an autograft [6]. Since it does not undergo remodeling like living bone, allografts frequently fail. Furthermore, they are associated with possible disease transmission and risk of infection [4].

Recently, the implantation of stem cells has risen as an alternative option to bone grafting procedures. Mesenchymal stem cells derived from bone marrow (BMMSCs) have excellent osteogenic potential and anti-inflammatory properties. However, they also require aspiration from the iliac crest, a painful procedure. Moreover, bone marrow aspiration yields only 10–40 mL of marrow and provides a small number of cells [7]. In addition, the number of hBMMSCs and their osteogenic potential decrease with aging. Adipose stem cells (ASCs) have properties similar to hMSCs from bone marrow (hBMMSCs). Whereas ASCs have the advantage of abundance and less morbidity in the acquirement procedure, their osteogenic potential is reportedly lower compared with BMMSCs [8].

Human embryonic stem cells (hESCs) have also drawn attention as a cell source for regenerative medicine [9]. The most compelling advantage of hESCs for bone regeneration is that they can potentially provide an unlimited number of osteoprogenitors for implantation. However, derivation of hESCs from early embryos raises ethical limitations for their use in clinical practice [10]. Induced pluripotent stem cells (iPSCs) generated from somatic cells by transduction of defined reprogramming transcription factors (typically OCT4, SOX2, KLF4, and c-MYC) offer a new path to avoid the controversial use of hESCs [11,12]. iPSCs express many of the markers associated with pluripotent cells and possess an epigenetic status similar to that of ESCs [11 –15]. iPSCs function in a manner indistinguishable from ESCs by differentiating in vitro and in vivo into cell types that are characteristic of the three germ layers. The high proliferation and differentiation capabilities of hiPSCs similar to those of hESCs also make them potential candidates for regenerative medicine [16].

To be regarded as a clinically viable alternative to MSCs, hiPSCs should demonstrate properties that are superior or comparable to those of MSCs. Although significant progress has recently been made in musculoskeletal regenerative medicine with respect to iPSCs, there is a relative paucity of translational research that demonstrates the in vivo regenerative potential of hiPSC-derived lineages in bone repair relative to that of hBMMSCs. Thus, the purpose of this study was to examine the in vitro and in vivo osteogenic potential of hiPSCs as compared with hBMMSCs.

Materials and Methods

hiPSC culture

To avoid potential safety issues associated with the use of viruses, in the present study, we used the hiPS cell line (SC802A-1; System Biosciences, Inc., Mountain View, CA) created by direct delivery of four proteins fused to a cell-penetrating peptide [17]. Undifferentiated hiPSCs were maintained as described previously [12,15]. Briefly, undifferentiated hiPSCs were maintained on mitotically inactivated mouse embryonic fibroblast (MEF) feeder layers in the ES medium [DMEM/F12 supplemented with 20% (v/v) knockout serum replacements (KSR; Gibco BRL, Grand Island, NY), 1% antibiotics (penicillin 100 IU/mL; streptomycin 100 μg/mL; Gibco BRL), 0.1 mM nonessential amino acids (Gibco BRL), 0.1 mM β-mercaptoethanol (Sigma-Aldrich, St. Louis, MO), and 10 ng/mL basic fibroblast growth factor (bFGF; R&D Systems, Minneapolis, MN)]. The medium was changed daily. For the maintenance of undifferentiated hiPSCs, cultures were passaged once every week by mechanically dissecting and transferring hiPSC colonies onto freshly prepared MEF feeders.

To prepare feeder-free hiPSCs for differentiation experiments, hiPSCs were passaged to Matrigel-coated polystyrene plates and cultured in the defined mTeSR1 medium (STEMCELL Technologies, Inc., Vancouver, BC, Canada). A combined use of the mTeSR1 medium and Matrigel-coated substrates has proven to support the feeder-independent maintenance of hiPSCs [18,19]. Cultures were passaged every week by mechanically dissecting hiPSC colonies onto freshly prepared Matrigel-coated polystyrene plates. hiPSCs were passaged to the 5th passage under feeder-free conditions to eliminate contaminating feeder cells. The hiPSCs used in this study exhibited typical morphological features similar to those of hESCs (eg, a large nucleus with prominent nucleoli) and expressed undifferentiated ESC markers, including NANOG, stage-specific embryonic antigen 4 (SSEA4), and OCT3/4.

Differentiation into embryoid bodies

In vitro differentiation of hiPSCs was performed using the standard embryoid body (EB) differentiation method with minor modifications [20]. For sphere formation, cells were dissociated with 0.05% trypsin-EDTA and plated onto six-well ultralow-attachment plates (Costar, Corning, NY). After 2 days of sphere formation, EBs were cultured in the ESC medium in the presence of 10⁻⁷ M all-trans retinoic acid (ATRA) for 10 days. The medium was changed every other day.

Osteogenic culture of hiPSCs

For osteogenic differentiation, hiPSC-EBs were dissociated to a single cell suspension by trypsinizing and then diluting to a final concentration of 3.0×10⁵ cells/mL. For subsequent differentiation, single cells were plated on gelatin-coated dishes and cultured with specific induction media (osteogenic medium [OM] consisting of α-MEM solution containing 10% FBS, 100 nM dexamethasone, 50 mM L-ascorbate-2-phosphate, 10 mM glycerophosphate, and 1% antibiotics). The cells were incubated in the OM for up to 2 weeks at 37°C in 5% CO₂ in a 6-cm dish at a density of 3×10⁵ cells. The medium was changed every third day. The analyses were performed on day 7 and 14 to test the osteogenic differentiation of hiPSCs (Fig. 1A).

FIG. 1.

Outline of the experimental procedure. (A) In vitro induction of osteogenic differentiation from human induced pluripotent stem cells (hiPSCs) and human bone marrow mesenchymal stem cells (hBMMSCs). (B) In vivo implantation in the calvarial defects. (C) In vivo implantation in the radial segmental defects.

Isolation and cultivation and in vitro osteogenic differentiation of hBMMSCs

The bone marrow samples used to isolate hBMMSCs were obtained from three patients (mean age, 62 years; range, 50–82 years) undergoing total hip arthroplasty from osteoarthritis. Informed consent was obtained from all donors. Briefly, bone marrow samples (10 mL) were mixed with 0.3 mL of heparin to prevent coagulation, and then diluted with 20 mL of phosphate-buffered saline (PBS, Welgene, Daegu, Korea). The cells were then fractionated on a Lymphoprep1 density gradient (Axis-Shield, Oslo, Norway) through centrifugation at 600 g for 10 min. Interface mononuclear cells were isolated and washed with PBS. The erythrocyte (RBC) lysis buffer (0.154 M NH₄Cl, 10 mM KHCO₃, 0.1 mM EDTA; Sigma-Aldrich) was then added to destroy contaminating RBCs. The cells were then washed twice by centrifugation (600 g) in PBS, and 1×10⁶ cells were seeded in the α-MEM (Gibco BRL) containing 10% fetal bovine serum (FBS: Gibco BRL) and 1% penicillin/streptomycin at 37°C in a humidified 5% CO₂ atmosphere. For osteogenic differentiation, hBMMSCs were processed through the same methods and culture medium as hiPSCs.

Alkaline phosphatase staining

The alkaline phosphatase (ALP) activity was measured using a TRACP & ALP double-stain kit (Takara Bio, Inc., Tokyo, Japan) followed by incubation for 7 or 14 days. To fix the cultured cell samples, the culture supernatant was removed and discarded. The cells were washed once with sterilized PBS, and then fixed in a fixation solution at room temperature for 5 min. Approximately 2 mL of sterilized distilled water was added to each well to dilute the fixation solution, and the solution was then aspirated. Approximately 2 mL of sterilized distilled water was again added to wash the well, and all the liquid from the well was removed and discarded. For ALP staining, the substrate solution was added to a 6-cm dish onto which the cells had been fixed. The amount of substrate solution was 500 μL/well. The plate was incubated at 37°C for 45 min for the reaction. The solution was then removed and discarded. Subsequently, the dish was washed three times with sterilized distilled water to quench the reaction.

Alizarin red staining

To measure calcium deposition in the extracellular matrix, the cells were seeded on 6-cm cell culture dishes and cultured for 7 or 14 days under the OM. The differentiated cell cultures were washed twice with PBS and fixed in 10% formalin for 10 min. After three washes with PBS, cells were then stained with a 2% Alizarin red solution (Junsei Chemical, Tokyo, Japan) for 10 min.

Reverse transcription–quantitative polymerase chain reaction

Total RNA preparation, cDNA synthesis, and reverse transcription–quantitative polymerase chain reaction were performed as previously described [21]. Primer information is provided in Table 1. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal control for PCR amplification, and the relative normalization ratio of PCR products derived from each target gene was calculated using the LightCycler System software (Roche, Indianapolis, IN). All experiments were performed in triplicate.

Table 1.

Real-Time Polymerase Chain Reaction Primer Information

Gene	Sequence (5′-3′)	Accession No.
Primers for osteogenic differentiation markers
Runx-2	F- TTACTTACACCCCGCCAGTC	NM_001024630
	R- TATGGAGTGCTGCTGGTCTG
COL1A1	F- CCCCTGGAAAGA ATGGAGATG	NM_000088
	R- TCCAAA CCACTGAAACCTCTG
BSP	F- AGA ACCACTTCCCCACCTTT	BC_111920
	R- AGGTTCCCCGTTCTCACTTT
Housekeeping gene
G3PDH	F- GCTCAGACACCATGGGGAAGGT	NM_002046
	R- GTGGTGCAGGAGGCATTGCTGA

Western blot analysis

Proteins were extracted from cultures, electrophoresed with the sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) gel, and transferred to a nitrocellulose membrane, as previously reported [21]. Membrane-transferred proteins were incubated in 5% nonfat milk to block nonspecific binding. The blot was probed with anti-rabbit collagen type I (COL1A1; 1:500; Abcam, Cambridge, United Kingdom), bone sialoprotein (BSP; 1:500; Abcam), or anti-mouse Runx-2 (1:500; Abcam) followed by horseradish peroxidase-conjugated anti-rabbit IgG or anti-mouse IgG (1:2,000; Cell Signaling Technology, Beverly, MA). Signals were visualized using the enhanced chemiluminescence western blot analysis detection reagent (Amersham Biosciences, Piscataway, NJ). Membrane-transferred proteins were incubated in 5% nonfat milk to block nonspecific binding. Bands on immunoblots were analyzed using an LAS-3000 image reader (version 2.1; Fujifilm, Tokyo, Japan). This experiment was repeated in three samples, each from different individuals.

Immunohistochemistry

Cultured cells and cryosectioned calvarial bone slices were fixed with 4% paraformaldehyde in PBS. To detect human nuclear antigen (HN) in the transplanted calvarial bone, different antigen retrieval procedures, such as SDS treatment, were employed before the blocking procedure. Sectioned calvarial bone slices were treated with 1% SDS in PBS at room temperature for 5 min and then immersed in PBS [22]. The cells or sections were blocked with 5% normal goat serum and 0.1% Triton X-100 in PBS at room temperature for 1 h. The following primary antibodies were applied overnight at 4°C: rabbit polyclonal antibodies [including NANOG 1:200 (Santa Cruz Biotechnology, Santa Cruz, CA), brachyury/bry (1:200; Abcam), COL1A1 (1:200; Abcam), and osteocalcin (1:200; Abcam)] and mouse monoclonal antibodies [including Oct3/4 (1:200; Santa Cruz Biotechnology), Runx-2 (1:100; Abcam), and HN (1:100; Chemicon, Temecula, CA)]. Appropriate fluorescence-tagged secondary antibodies (Jackson Immunoresearch Laboratories, West Grove, PA) were used for visualization. The stained samples were mounted in VECTASHIELD^® with the 4′, 6-diamidino-2-phenylindole (DAPI) mounting medium (Vector Laboratories, Burlingame, CA) and photographed using an epifluorescence microscope (Leica, Wezlar, Germany). Measurement of fluorescence intensity was calculated using the Leica Application Suite (LAS) image analysis package.

Surgery and transplantation procedure

The animal experiments conducted in this study were approved by the Animal Research and Care Committee of our institution. Nine-week-old male Sprague Dawley rats were used in this study. The animals were anesthetized with zoletil (40 mg/kg) and xylazine (10 mg/kg).

Calvarial defect model

Two full-thickness calvarial bone defects 4 mm in diameter were created without dural perforation using a surgical microdrill fitted with a trephine burr. The wound was thoroughly irrigated with warmed saline to remove residual bone dust. A total of 18 animals were randomly assigned to three groups: group 1, the control defect; group 2, osteoinduced hiPSC implantation; and group 3, osteoinduced hBMMSC implantation (see Table 2). Two kinds of scaffolding materials were used, fibrin glue (Tisseel^® Darim, Seoul, Korea) and hydroxyapatite/β-tricalcium phosphate (HA/β-TCP, 7:3; Genoss, Suwon, Korea). hiPSCs or hBMMSCs underwent osteogenic differentiation for 7 days before implantation. The cells (1.5×10⁶ cells) were mixed with 30 μL fibrin glue or 30 mg HA/β-TCP and placed into the calvarial defect. After implantation of the cell/scaffold hybrid, the defect region was covered with the polycaprolactone membrane (Genoss), and the skin was closed with staples (Fig. 1B). The rats received daily injections of cyclosporin A to suppress immune responses. After 8 weeks, the rats were sacrificed using carbon dioxide. Calvarial bones were dissected out and underwent gross and histological analysis as well as microcomputed tomography (micro-CT: NFR Polaris-G90; Nano Focus Ray, Jeonju, Korea).

Table 2.

Treatment Groups of Experimental Animals

Group	n	Graft at left defect	Graft at right defect
Control (Group 1)	6	Fibrin only	HA/β-TCP
hiPSCs (Group 2)	6	Fibrin/osteoinduced hiPSCs	HA/β-TCP/osteoinduced hiPSCs
hBMMSCs (Group 3)	6	Fibrin/osteoinduced hBMMSCs	HA/β-TCP/osteoinduced hBMMSCs

hiPSC, human induced pluripotent stem cells; hBMMSC, human bone marrow mesenchymal stem cells.

Segmental defect model

Nine-week-old Sprague Dawley (SD) rats were used for critical-size long bone segmental defects. A total of 18 animals were assigned to three groups and fibrin was used as the scaffolding material. A 4-mm-long segmental defect was created on the right radius of rats using a surgical microdrill, and the wound was thoroughly irrigated with warmed saline to remove residual bone dust. The cells (1.5×10⁶ cells) were mixed with 30 μL fibrin glue and placed into the segmental defect. There were six animals in each of the three following groups: group 1, the control defect (fibrin glue only); group 2, osteoinduced hiPSC implantation; group 3, osteoinduced hBMMSC implantation. After cell implantation, the muscle and skin were closed with a surgical black silk suture (Fig. 1C). The rats received daily injections of cyclosporin A to suppress immune responses. Radiographs were taken every 3 weeks. After 12 weeks, the rats were sacrificed and analyzed through micro-CT.

Macroscopic observation and histology

Following macroscopic examination, the calvarial bone was dissected and embedded in an Optimal Cutting Temperature (OCT) compound (an aqueous embedding medium within a mold) and then frozen in a metal pan over a bath of liquid nitrogen. All frozen tissue blocks were cryosectioned to a nominal thickness of 10 μm. Goldner's trichrome staining was performed for all specimens.

Statistical analysis

All quantitative data are expressed as mean±SEM. Statistical comparisons were made using the Mann–Whitney U test and ANOVA with Tukey post hoc analysis (SPSS 15.0; SPSS, Inc., Chicago, IL). Statistical significance was set at a P<0.05.

Results

Disappearance of ESC markers and appearance of osteogenic markers in osteoinduced hiPSCs

EBs were formed from undifferentiated hiPSCs through the addition of retinoic acid. After EBs were dissociated into single cells, osteogenic cultures were performed using the same medium as used for hBMMSCs (Fig. 2A). ESC markers, NANOG and OCT3/4, were observed in the nuclei of undifferentiated hiPSCs. They stopped appearing in EBs and osteoinduced cells. On the other hand, brachyury the mesodermal marker, which was not observed in undifferentiated hiPSCs, appeared in abundance in EBs and then disappeared from osteoinduced hiPSCs. Osteocalcin, the marker of osteogenic differentiation, appeared in osteoinduced hiPSCs (Fig. 2B). These findings suggest that osteoinduced plated cells lost the properties of undifferentiated hiPSCs and acquired those of osteoblasts.

FIG. 2.

Disappearance of embryonic stem cell (ESC) markers and appearance of mesodermal and osteogenic markers in osteoinduced hiPSCs. (A) General scheme of ostegenic differentiation from hiPSCs. (B) Expression of ESC markers NANOG, OCT3/4, mesodermal marker brachyury, and osteogenic marker osteocalcin.

In vitro osteogenic differentiation of hiPSCs versus hBMMSCs

hiPSCs dissociated from EBs underwent osteogenic induction under a standard OM, and then were compared with hBMMSCs. At a basal state without induction, hiPSCs had a much lower gene expression of COL1A1 and BSP (P<0.05) than hBMMSCs, while Runx-2 levels were comparable to those of hBMMSCs. After 7 days of osteoinduction, the expression of osteogenic markers Runx-2, COL1A1, and BSP increased by 5- to 60-fold (P<0.05) compared with that of hiPSCs at day 0 under a standard OM. However, the level of COL1A1 and BSP gene expressions was significantly lower in hiPSCs than in hBMMSCs at day 7 of osteoinduction (P<0.05). The level of gene expressions further increased at day 14 in osteoinduced hiPSCs. The expression of Runx-2 was significantly higher than hBMMSCs (P<0.05). COL1A1 and BSP expression caught up that of hBMMSCs, while BSP expression was still significantly lower compared with hBMMSCs (P<0.05) (Fig. 3A). The protein expression of Runx-2, COL1A1, and BSP largely paralleled that of gene expression (Fig. 3B). Immunohistochemistry (IHC) also showed corresponding expression patterns of osteogenic markers Runx-2 and COL1A1. The COL1A1 expression in hiPSCs at 14 days of osteoinduction was comparable to that of hBMMSCs at 7 days of osteoinduction. Runx-2, a transcription factor, was localized in both the nucleus and cytoplasm, whereas COL1A1, a matrix component, was located in the cytoplasm and also in the extracellular area (Fig. 3C). ALP staining and Alizarin red staining showed that osteoblastic differentiation and mineral deposition were significantly weaker than simultaneously cultured hBMMSCs on day 7 and 14 of osteoinduction (Fig. 3D).

FIG. 3.

In vitro osteogenic differention from hiPSCs and hBMMSCs. (A) Gene expression. (B) Western blotting. (C) Immunohistochemistry (IHC) for osteogenic markers. (D) Alkaline phosphatase (ALP) staining for osteoblastic differentiation and Alizarin red staining for mineralization. Data are presented as mean±SE. n=3, *P<0.05 versus day 0, ^# P<0.05 versus hiPSCs.

Healing of critical-size calvarial defect by the implantation of osteoinduced hiPSCs

After confirming in vitro osteogenic differentiation from hiPSCs relative to hBMMSCs, we proceed to assess the in vivo osteogenic capability of hiPSCs in the calvarial defect model, which represents bone repair through intramembranous ossification. Two critical-size bone defects were created on the parietal bone of rats. The defects were filled with fibrin glue with or without cells (hiPSCs or hBMMSCs) on the left side, and HA/β-TCP with or without cells on the right side (Fig. 4A). Eight weeks after implantation, rats were sacrificed and evaluated.

FIG. 4.

In vivo healing of critical-size calvarial defect by osteoinduced hiPSCs and osteoinduced hBMMSCs after 8 weeks of implantation. (A) Creation of calvarial defects, gross and micro-CT findings of the defect. (B) Area repaired with bone (calculated from micro-CT). (C) Histological finding from Goldner's trichrome staining. (D) Average thickness of healed calvarial defects. (E) IHC for human nuclear antigen to detect implanted osteoinduced hiPSCs and osteoinduced hBMMSCs. Bar represents mean±SE. n=6, *P<0.05, **P<0.01 compared with control.

In left calvarial defects treated with fibrin glue, the addition of osteoinduced hiPSCs (group 2, P<0.05) or osteoinduced hBMMSCs (group 3) significantly increased the area healed with bone (P<0.01) and the quality of regenerated bone relative to the control defect that was filled with fibrin glue only (group 1). There was no significant difference between the defects treated with osteoinduced hiPSCs and those treated with osteoinduced hBMMSCs (P=0.90). The histological results observed through trichrome staining also demonstrated better quality of new bone formation, including the full restoration of bone width and the robust formation of trabeculae, in defects treated with osteoinduced hiPSCs or osteoinduced hBMMSCs compared with control. The thickness of restored bone did not have a significant difference between the osteoinduced hiPSC group and the osteoinduced hBMMSC group (P=0.35, Fig. 4B–D).

Due to the osteoconductive properties of HA/β-TCP, calvarial defects treated with HA/β-TCP showed better bone regeneration (greater filling area and restoration of bone width) relative to those treated with fibrin glue in general. There was no statistical difference in the area filled with bone among the different treatment groups (P>0.05). However, histological findings definitely showed a better quality of new bone formation and a greater thickness of bone in groups treated with osteoinduced hiPSCs (group 2, P<0.05) or osteoinduced hBMMSCs (group 3, P<0.05) to the control defects. There was no significant difference between group 2 and 3 (P=0.51, Fig. 4B–D).

Human nuclear antigens were detected to see if the implanted human cells persisted in the newly regenerated bone. Human cells were observed abundantly in defects treated with osteoinduced hiPSCs or osteoinduced hBMMSCs (Fig. 4E).

Healing of critical-size long bone segmental defect by hiPSCs

After confirming that hiPSCs have an osteogenic potential comparable to hBMMSCs in healing calvarial defects, their osteogenic potential in long bone segmental defects was tested (Fig. 5A). After 12 weeks, the volume and quality of regenerated bone were compared among the different groups. Near-complete bony regeneration of defects was achieved in four out of six rats in group 2 (osteoinduced hiPSCs) and group 3 (osteoinduced BMMSCs), while it was achieved in two out of six rats in group 1 (P=0.24, Fig. 5B).

FIG. 5.

Healing of critical-size long bone segmental defect by osteoinduced hiPSCs and osteoinduced hBMMSCs. (A) Creation of segmental defects from radial shaft of rats. (B) Findings from plain radiographs immediately and 12 weeks after implantation. (C) CT findings after 12 weeks of implantation.

Discussion

The overall results demonstrated successful in vitro induction of osteogenic differentiation from hiPSCs. One notable point was a relatively lower and delayed in vitro expression of osteogenic markers except Runx-2. On the other hand, the in vivo results demonstrated successful repair of the critical-size calvarial defect and segmental long bone defect that were comparable to those achieved by hBMMSCs. hiPSCs persisted and were differentiated into functional osteoblasts on the created defect.

The rationale for using hiPSCs comes from the idea of having an unlimited number of cells for regeneration. These cells can be expanded indefinitely in an undifferentiated state, enabling the derivation of an adequate number of differentiated progeny in contrast to adult stem cells that show changes in differentiation potential after ∼4 passages in culture [23]. Implanting a large number of cells is very important to repair critical-size large bone defects, which do not heal spontaneously and need clinical intervention. Derivation of hiPSCs from minimally invasive sources such as skin fibroblasts allows for patient-matched tissue engineering even in elderly patients [24,25]. However, before hiPSCs are applied for clinical treatment, it should be proved that the cell differentiation process results in therapeutically relevant cell types, and that the differentiated phenotypes are maintained in vivo.

Osteogenic differentiation of iPSCs can be induced with or without the intervening stage of EBs [26 –28]. Whereas the direct differentiation method holds promise as a simpler way to induce osteogenic differentiation from hESCs, we used the differentiation protocol using the intermediate step of EB formation, a better-known method that yields predictable results. We used standard osteogenic culture, including ascorbate and β-glycerophosphate, without adding osteogenic factors such as BMPs and vitamin Ds. Whereas the in vitro osteogenesis was delayed compared with hBMMSCs, reasonable osteogenic differentiation was achieved from hiPSCs after 14 days of culture.

Since the ultimate purpose of this study was to see if the hiPSCs can stay on the defects and regenerate bone there, in vivo implantation studies were performed. Although our in vitro study demonstrated that osteogenic differentiation was delayed and lower in hiPSCs than in hBMMSCs, which are cells primed for osteogenesis, in vivo bone repair by osteoinduced hiPSCs was comparable to that of osteoinduced hBMMSCs. Calvarial defect models are the gold standard model to assess in vivo osteogenesis. However, the model evaluates intramembranous bone formation in a nonweight-bearing area and does not provide information on whether endochondral bone formation in long bone defects is also promoted. Hence, the segmental long bone defect model was used to assess the bone-forming effect of osteoinduced hiPSCs in weight-bearing bone. The defects were created in the ulna of rats because external fixation was not necessary, unlike in tibial or femoral defects. We used both the fibrin and ceramic scaffold in the calvarial defect model of this study. Because fibrin does not contribute to bone formation, the pure effect of cell implantation can be deduced from the results in fibrin groups. On the other hand, bioactive ceramic scaffolds have osteoconductive effects. HA/β-TCP is frequently used for the bone regeneration experiments using stem cells. Therefore, the results from combined bioceramic/stem cell treatment can be extrapolated to predict the osteogenic effects of iPSCs in possible clinical applications.

Because of its teratoma formation, long-term observation is necessary to verify the safety of iPSCs in in vivo implantation. Although there was no occurrence of teratoma from the implanted osteoinduced iPSCs during the period of 12 weeks in immunosuppressed rats, further follow-up data will be necessary for the consideration of iPSC implantations as a cell therapy for bone regeneration. In addition, while this study demonstrated successful bone regeneration using hiPSCs in rodents, further experimentation in large animals will be necessary to explore the possibility of clinical applications.

Footnotes

Acknowledgment

This research was supported by the Bio & Medical Technology Development Program of the National Research Foundation (NRF) funded by the Korean government (MEST) (2012M3A9B4028566).

Author Disclosure Statement

No competing financial interests exist.

References

Einhorn

. (2005). The science of fracture healing. J Orthop Trauma, 19:S4–S6.

Hollinger

and Kleinschmidt

. (1990). The critical size defect as an experimental model to test bone repair materials. J Craniofac Surg, 1:60–68.

McKee

. (2006). Management of segmental bony defects: the role of osteoconductive orthobiologics. J Am Acad Orthop Surg, 14:S163–S167.

Evans

. (2010). Gene therapy for bone healing. Expert Rev Mol Med, 12:e18.

Kim

, Rhim

, Li

, Martha

, Swaim

, Banco

, Jenis

and Tromanhauser

. (2009). Prospective study of iliac crest bone graft harvest site pain and morbidity. Spine J, 9:886–892.

Delloye

, Cornu

, Druez

and Barbier

. (2007). Bone allografts: What they can offer and what they cannot. J Bone Joint Surg Br, 89:574–579.

Pittenger

, Mackay

, Beck

, Jaiswal

, Douglas

, Mosca

, Moorman

, Simonetti

, Craig

and Marshak

. (1999). Multilineage potential of adult human mesenchymal stem cells. Science, 284:143–147.

, Shin

and Lee

. (2005). Do adipose tissue-derived mesenchymal stem cells have the same osteogenic and chondrogenic potential as bone marrow-derived cells?. Osteoarthritis Cartilage, 13:845–853.

Lerou

and Daley

. (2005). Therapeutic potential of embryonic stem cells. Blood Rev, 19:321–331.

10.

Gaspar-Maia

, Alajem

, Meshorer

and Ramalho-Santos

. (2011). Open chromatin in pluripotency and reprogramming. Nat Rev Mol Cell Biol, 12:36–47.

11.

Takahashi

and Yamanaka

. (2006). Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell, 126:663–676.

12.

Takahashi

, Tanabe

, Ohnuki

, Narita

, Ichisaka

, Tomoda

and Yamanaka

. (2007). Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell, 131:861–872.

13.

Wernig

, Meissner

, Foreman

, Brambrink

, Ku

, Hochedlinger

, Bernstein

and Jaenisch

. (2007). In vitro reprogramming of fibroblasts into a pluripotent ES-cell-like state. Nature, 448:318–324.

14.

Yamanaka

. (2007). Strategies and new developments in the generation of patient-specific pluripotent stem cells. Cell Stem Cell, 1:39–49.

15.

, Vodyanik

, Smuga-Otto

, Antosiewicz-Bourget

, Frane

, Tian

, Nie

, Jonsdottir

, Ruotti

, et al. (2007). Induced pluripotent stem cell lines derived from human somatic cells. Science, 318:1917–1920.

16.

Itoh

, Kiuru

, Cairo

and Christiano

. (2011). Generation of keratinocytes from normal and recessive dystrophic epidermolysis bullosa-induced pluripotent stem cells. Proc Natl Acad Sci U S A, 108:8797–8802.

17.

Kim

, Kim

, Moon

, Chung

, Chang

, Han

, Ko

, Yang

, Cha

, Lanza

and Kim

. (2009). Generation of human induced pluripotent stem cells by direct delivery of reprogramming proteins. Cell Stem Cell, 4:472–476.

18.

, Inokuma

, Denham

, Golds

, Kundu

, Gold

and Carpenter

. (2001). Feeder-free growth of undifferentiated human embryonic stem cells. Nat Biotechnol, 19:971–974.

19.

Akopian

, Andrews

, Beil

, Benvenisty

, Brehm

, Christie

, Ford

, Fox

, Gokhale

, et al. (2010). Comparison of defined culture systems for feeder cell free propagation of human embryonic stem cells. In Vitro Cell Dev Biol Anim, 46:247–258.

20.

, Zhou

, Abujarour

, Zhu

, Young Joo

, Lin

, Hao

, Schöler

, Hayek

and Ding

. (2009). Generation of human-induced pluripotent stem cells in the absence of exogenous Sox2. Stem Cells, 27:2992–3000.

21.

, Kim

and Lee

. (2011). Chondrogenesis of adipose stem cells in a porous PLGA scaffold impregnated with plasmid DNA containing SOX trio (SOX-5,-6 and −9) genes. Biomaterials, 32:4385–4392.

22.

Brown

, Lydon

, McLaughlin

, Stuart-Tilley

, Tyszkowski

and Alper

. (1996). Antigen retrieval in cryostat tissue sections and cultured cells by treatment with sodium dodecyl sulfate (SDS). Histochem Cell Biol, 105:261–267.

23.

Solchaga

, Penick

, Goldberg

, Caplan

and Welter

. (2010). Fibroblast growth factor-2 enhances proliferation and delays loss of chondrogenic potential in human adult bone-marrow-derived mesenchymal stem cells. Tissue Eng Part A, 16:1009–1019.

24.

Kim

, Son

, Seol

, Kim

, Park

, Kim

, Park

, et al. (2011). Generation of human induced pluripotent stem cells from osteoarthritis patient-derived synovial cells. Arthritis Rheum, 63:3010–3021.

25.

Wei

, Zeng

, Wan

, Wang

, Zhou

, Qiu

and Singh

. (2012). Chondrogenic differentiation of induced pluripotent stem cells from osteoarthritic chondrocytes in alginate matrix. Eur Cell Mater, 23:1–12.

26.

Cao

, Heng

, Ye

, Liu

, Toh

, Robson

, Li

, Hong

and Stanton

. (2005). Osteogenic differentiation within intact human embryoid bodies result in a marked increase in osteocalcin secretion after 12 days of in vitro culture, and formation of morphologically distinct nodule-like structures. Tissue Cell, 37:325–334.

27.

Karp

, Ferreira

, Khademhosseini

, Kwon

, Yeh

and Langer

. (2006). Cultivation of human embryonic stem cells without the embryoid body step enhances osteogenesis in vitro . Stem Cells, 24:835–843.

28.

Karner

, Unger

, Sloan

, Ahrlund-Richter

, Sugars

and Wendel

. (2007). Bone matrix formation in osteogenic cultures derived from human embryonic stem cells in vitro . Stem Cells Dev, 16:39–52.