Isolation and culture of human BMSCs and ATSCs

ATSCs and BMSCs were prepared as described previously [20, 21]. Briefly, the BMSCs were fractionated on a Ficoll density gradient (Ficoll-Paque™-PLUS; Amersham Pharmacia, Sweden) and the MSC-enriched fraction was washed, seeded in culture flasks, and maintained at 37 °C in a humidified atmosphere. The adipose tissue was washed extensively with equal volumes of phosphate-buffered saline (PBS), and the extracellular matrix was digested with 0.075% collagenase (type I; Sigma-Aldrich, St Louis, MO, USA) at 37 °C for 30 min. With α-MEM containing 10% FBS and antibiotics (100 U/ml penicillin G and 100 μg/ml streptomycin), the sample was centrifuged at 1200 × g for 10 min. The cell pellet was resuspended in 160 mM NH4Cl and incubated at room temperature for 10 min. After removing cellular remains through a 100-μm Nylon mesh (Cell Strainer; Becton Dickinson and Company, Franklin Lakes, NJ, USA), the cells were incubated in the culture medium. The adhered ATSCs were cultured for about 2 weeks, and nearly all cells transformed into fibroblast-like cells, which are morphologically similar to BMSCs.

Phenotypic characterization of hMSCs

After reaching 80% confluence, the cells were rinsed twice with PBS and treated with 0.05% trypsin–EDTA for 2 min. Serum-containing medium was then immediately added to the culture to end trypsinization. The fluid was then collected and centrifuged (800 × g for 5 min). After discarding the supernatant, the precipitate was resuspended in staining buffer and incubated with fluorochrome-conjugated primary antibodies against CD34, CD44, CD45, CD73, CD90, and CD105 or corresponding isotype control (BD Biosciences, USA) at 4 °C for 30 min. The stained cells were immediately detected using flow cytometry (BD Biosciences, USA).

Osteogenic differentiation

MSCs were plated at 4 × 103 cells/cm2 in a 12-well plate and cultured in the basal medium until the cells reached confluence. The cells were then incubated in osteogenic induction medium (OIM), which is basal medium supplemented with 1 nM dexamethasone, 50 μM ascorbic acid, and 20 mM β-glycerolphosphate (all from Sigma-Aldrich), at 37 °C, 5% CO2 as described previously [20, 21]. At day 14, the mineralization of MSCs was assessed by Alizarin Red S staining. Briefly, to evaluate the mineralized nodule formation in vitro, the cell/matrix layer was washed with PBS, fixed with 70% ethanol for 10 min, and stained with 0.5% Alizarin Red S (pH 4.1; Sigma, St Louis, MO, USA) for 5 min.

Adipogenic differentiation

MSCs were plated at 4 × 103 cells/cm2 in a six-well culture plate and cultured until the cells reached confluence. The medium was then replaced with adipogenic induction medium (AIM), which is basal medium supplemented with 500 nM dexamethasone, 0.5 mM isobutylmethylxanthine, 50 mM indomethacin, and 10 mg/ml of insulin (all from Sigma-Aldrich). The cells were cultured for another 21 days, and then the cells were fixed with 70% ethanol for 10 min and stained with 0.3% fresh Oil Red O solution (Sigma-Aldrich) for 10 min. The wells were rinsed three times with distilled water and viewed using a LEICA Q500MC microscope (Leica Cambridge Ltd).

Chondrogenic differentiation

For chondrogenic differentiation, a micromass culture system was used. MSCs (in 5 μl) at a centration of 1.6 × 107 cells/ml were dropped in the centers of 24-well plates. The plates were placed in incubator at 37 °C, 5% CO2 without culture medium for 2 hours. These cells were then cultured in chondrogenic induction medium (CIM), which is basal medium supplemented with 10 ng/ml transforming growth factor-β3 (R&D Systems), 500 ng/ml bone morphogenetic protein-2 (R&D Systems), 10–7 M dexamethasone, 50 mg/ml ascorbate-2-phosphate, 40 mg/ml proline, 100 mg/ml pyruvate (all from Sigma-Aldrich), and 1:100 diluted ITS + Premix (6.25 mg/ml insulin, 6.25 mg/ml transferrin, 6.25 mg/ml selenous acid, 1.25 mg/ml bovine serum albumin, and 5.35 mg/ml linoleic acid) (Becton Dickinson). The chondrogenic medium was changed every 3 days.

Quantitative real-time RT-PCR

The cells were harvested and homogenized for RNA extraction with the RNeasy mini kit (Qiagen, Hilden, Germany). The mRNA was reverse-transcribed to cDNA by the PrimeScript First Strand cDNA Synthesis Kit (TaKaRa). Then 5 μl of total cDNA from each sample was amplified in a final volume of 25 μl of reaction mixture containing Platinum SYBR Green, qPCR SuperMix-UDG ready-to-use reaction cocktail, and specific primers using the ABI StepOne Plus system (all from Applied Biosystems, CA, USA). The expression level of the target gene was normalized to that of the β-actin gene, which was shown to be stable in this study. Relative gene expression was calculated with the 2CT formula. The sequences of the primers were presented in Additional file 1: Table S1.

DNA isolation and bisulfite treatment

Genomic DNA was isolated from MSCs using the PureLink® Genomic DNA isolation kit following the manufacturer’s instructions (Invitrogen). Bisulfite modification was done as described previously [22]. Briefly, about 2 μg of genomic DNA was denatured by NaOH (final concentration 0.2 mol/L) for 10 min at 37 °C. Hydroquinone and sodium hydroxide were added, and samples were incubated at 50 °C for 16 hours. Modified DNA was purified using the Wizard DNA Clean-Up System following the manufacturer’s instructions (Promega) and eluted into 50 μl of water. DNA was treated with NaOH (final concentration 0.3 mol/l) for 5 min at room temperature, ethanol precipitated, and resuspended in 20 μl of water. Modified DNA was used immediately or stored at –20 °C.

Bisulfite sequencing

Bisulfite-modified genomic DNA was amplified by PCR. All PCRs were carried out using KAPA2G™ Fast HotStart DNA Polymerase Polymerase. The sequences of primers used for the bisulfite sequencing analysis are presented in Additional file 1: Table S2. PCR products were run on 1.5% agarose gels and bands were excised using the TaKaRa MiniBEST Agarose Gel DNA Extraction Kit following the manufacturer’s instructions (TaKaRa). Purified bands were cloned using the pMD™19-T Vector Cloning Kit following the manufacturer’s instructions (TaKaRa). Colonies were selected and grown overnight in Luria-Bertani medium containing ampicillin (100 μg/ml) with shaking at 37 °C. Plasmid DNA was isolated using the TaKaRa MiniBEST Agarose Gel DNA Extraction Kit following the manufacturer’s instructions (TaKaRa). Plasmids were sequenced using the M13 universal reverse primer (BGI).

Ectopic bone formation

In-vivo studies were performed with the approval of the Animal Experimentation Ethics Committee of The Chinese University of Hong Kong. After anesthesia, an incision was made on the dorsum and a subcutaneous pocket was created. 2.5 × 106 ATSCs or BMSCs were seeded onto sterilized Skelite® resorbable Si-TCP bone graft substitute, and Si/TCP cubes with PBS served as the control group. The cells were then transplanted into the same mice. The wound was then closed in layers. At week 8, the scaffolds with cells were harvested for HE staining, as well as immohistochemical staining of osteocalcin (OCN). The osteoid matrix areas were measured using ImageJ software, and five microscopic fields were chosen from each sample and measured.

Calvarial bone critical defect model

Six nude mice (6 weeks old, body weight 50 g) were used. All animals were placed under general anesthesia with a dosage of 0.2 ml/100 g body weight via intraperitoneal injection of a combination of ketamine, xylazine, and saline at a ratio of 3:2:3. The dorsal part of animal’s cranium was shaved and disinfected with iodine solution. The skin and underlying tissues including the periosteum were detached to expose the parietal bones on both sides. One piece of circular bone was removed in the middle region of the cranium using a hollow trephine bur with a 5-mm outer diameter. Continuous irrigation with sterile PBS was used to prevent overheating of the bone margins and to maintain moisture in the tissue. Any animal with evidence of meninges injury or continuous hemorrhaging was excluded. Then 50 μl of 2% hyaluronic acid hydrogel (5-mm-diameter cylinder) with 1 × 105 human ATSCs or BMSCs was immediately implanted into the defect cavity. The periosteum and scalp were closed by suture. Animals were allowed to move following recovery from the anesthesia and were sacrificed by overdose of pentobarbital 6 weeks after surgery. The defect sites were removed, including sufficient parietal bone and soft connective tissues surrounding the defect areas.

Micro-computed tomography imaging analysis

Micro-computed tomography (microCT) was used for quantitative evaluation of the bone formation. The samples were imaged using a high-resolution 70-kVp scan by microCT machine (VivaCT; Scanco Medical, Bassersdorf, Switzerland). The 3D reconstruction was performed using standardized segmentation parameters (sigma 0.8, threshold 160–1000), which were kept constant through the scan. Circular contour lines were drawn around the defect area (diameter = 5 mm) excluding the neighboring native bone. The 3D reconstructive images of samples were generated from 2D slices by machine built-in software. The bone volume within the selected circular defect was calculated using the quantitative 3D evaluation program included in the microCT software package.

Histology and immunohistochemistry

Immunohistochemical staining was performed as described previously [23]. The samples were washed in PBS, fixed in 4% paraformaldehyde, decalcified, dehydrated, and embedded in paraffin. Sections were cut at a thickness of 5 μm and were stained with HE after deparaffination. Endogenous peroxidase activity was quenched with 3% hydrogen peroxide for 20 min at room temperature. Antigen retrieval was then performed with citrate buffer at 80 °C for 10 min for immunohistochemistry detection. Primary antibody against osteocalcin (1:100, sc-365797; Santa Cruz, CA, USA) was used. Donkey anti-goat IgG horseradish peroxidase (HRP)-conjugated secondary antibody was then added for 1 hour, followed by 3,3′-diaminobenzidine tetrahydrochloride (DAKO, Glostrup, Denmark) in the presence of H2O2 for signal detection of osteocalcin. Afterward, the sections were rinsed, counterstained in hematoxylin, dehydrated with graded ethanol and xylene, and mounted with p-xylene-bis-pyridinium bromide (DPX) permount (Sigma-Aldrich). Primary antibody was replaced with blocking solution in the negative controls. All incubation times and conditions were strictly controlled. The sections were examined under light microscopy (DMRXA2; Leica Microsystems Wetzlar GmbH, Germany).

Data analysis

All experiments were performed at least three times. All data were expressed as the mean ± SD. The data were analyzed by independent two-tailed Student’s t test using SPSS (version16.0; Chicago, IL, USA). p < 0.05 was regarded as statistically significant.