All animal experimental protocols were approved by the local Institutional Animal Care and Use Committee of Shandong University (Jinan, China) in compliance with the ‘Guide for the Care and Use of Laboratory Animals’ published by the National Academy Press (NIH Publication No. 85–23, revised 1996). Male beagles (n=10; 4–10 months of age), weighing between 8 and 10 kg, were obtained from the Experimental Animal Center of Shandong Province (Jinan, China) and used in the present study. Animals were kept in separate cages with an ambient temperature of 24°C and 45% humidity, with a 14 h light/dark cycle, were fed a standard diet with ad libitum access to food and water, and were allowed to move freely during the study.

Following the administration of 3% pentobarbital sodium (1 ml/kg; Solarbio Science & Technology Co., Ltd., Beijing, China), the tibia was shaved and treated with 1% povidone iodine solution (Shandong Luxi Pharmaceutical Co., Ltd., Heze, China). Bone marrow (5 ml) was aspirated with a sterile bone marrow aspiration needle attached to a 10 ml syringe containing 1 ml heparin solution (Shandong Lukang Pharmaceutical Group, Jining, China). Following the addition of an equal volume (5 ml) of phosphate-buffered saline (PBS) to the bone marrow serum, the mixture was homogenized. The aspirate was centrifuged at a speed of 168 × g for 5 min at 20°C. The supernatant was removed and the precipitate mixed with an equal volume of PBS (5 ml) and homogenized again. The mixture was then centrifuged at a speed of 1,048 × g for 20 min at 20°C and the cloudy coat in the centre of the centrifuge tube was harvested. This was combined with 5 ml PBS in a centrifuge tube and separated at a speed of 2,500 rpm for 5 min. Subsequently, BMSCs were harvested from the bottom of the centrifuge tube. The primary BMSCs were cultured in Dulbecco's Modified Eagle's Medium (DMEM; HyClone, Rockford, IL, USA) supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA) at 37°C, in a humidified incubator containing 5% CO₂. After three passages, the BMSCs were divided into osteoblast- and chondroblast-induced groups. The osteoblast-induced group was cultured in osteogenic differentiation medium containing DMEM supplemented with 10% fetal bovine serum, 10 nM dexamethasone, 50 mg/l ascorbic acid and 10 mM β-glycerophosphate (Wuhan Boster Biological Technology Co., Ltd., Wuhan, China). The chondroblast-induced group was cultured in chondrogenesis medium containing TGF-β1 (10 ng/ml), IGF-1 (50 g/l) and dexamethasone (40 g/l).

Cells were observed using a light microscope for morphology and growth status daily. In the chondroblast-induced group, type II collagen was detected by immunohistochemistry. Antibodies were purchased from Wuhan Boster Biological Technology Co., Ltd.. The cells were fixed with 3% paraformaldehyde for 10 min and rinsed with Tris buffer, pH 7.4, at room temperature for 5 min. The cells were then incubation with serum (1:20 in Tris buffer) at room temperature for 10 min, incubated with primary antibodies (anti-collagen II; cat. no. BA0533) in a moist chamber overnight at 4°C, washed with PBS and incubated with biotinylated goat anti-rabbit immunoglobulin G secondary antibody (1:50 in Tris buffer; cat. no. BA1003) at room temperature for 30 min, washed with PBS and incubated with dual system bridge antibodies (1:50 in Tris buffer; cat. no. BA1003) for 30 min at room temperature. The cells were then washed with PBS and fuscin (Wuhan Boster Biological Technology Co., Ltd.) was used for staining for 30 min at room temperature. Subsequently, the specimens were washed with PBS and dried, covered with glycerin/gelatin and examined under a light microscope. In the osteoblast-induced group, cells were identified with alkaline phosphatase (ALP) staining and Alizarin red staining (Wuhan Boster Biological Technology Co., Ltd.) to observe the activity of ALP and the number of mineral nodules.

β-TCP is a porous bio-ceramic 3D scaffold purchased from Shanghai Bio-lu Biomaterials Co., Ltd. (Shanghai, China). The diameter of the spherical pores was 500–600 µm and the porosity was 75±10% with >80% of the pores being spherical in shape. Using a mold, a cylinder of β-TCP 5 mm in diameter and 5 mm in height was constructed. The surface of the cylinder was smooth and was sterilized using ethylene oxide.

The perfusion bioreactor (Fig. 1) was designed by the East China University of Science and Technology (Shanghai, China). The system consists of a peristaltic pump, a device which pulls culture media through silicone tubes, a perfusion column in which the cell-scaffold composite can be implanted, two silicone tubes, an air filter and a flask filled with cell culture media. The function of the two silicone tubes is to enable the flow of culture media into a peristaltic pump via the silicone tubes when the peristaltic pump begins to rotate. This creates a circulatory flow to provide mechanical stimulation to the cells.

PRP was prepared according to a two-step centrifugation method as previously described (13). The standard protocol involves the preparation of PRP from autologous blood by a two-step centrifugation process; a separation step and a concentration step (17,18). For the preparation of PRP, ~10 ml of blood was obtained from the femoral veins of 10 anesthetized (3% pentobarbital sodium) beagles. The blood was collected in sterilized tubes containing 1 ml of sodium citrate used as an anticoagulant. The blood was then centrifuged twice; initially at 20 × g for 16 min at 20°C to remove red blood cells, and then at 1,500 × g for 12 min at 20°C to obtain the platelet pellet (19). The pellet was resuspended in platelet-poor plasma and the PRP from the 10 beagles was mixed and stored at −80°C until required, and consequently allogenic PRP was used in the subsequent experiments. The PRP was combined with 1 ml coagulant (composed of 1 ml 10% CaCl₂ and 1,000 U thrombin) and homogenized. Platelet-rich plasma gel, a jelly-like substance (Fig. 2), was then harvested.

Once the cells had been passaged three times, BMSC cell suspensions were prepared by trypsin (Wuhan Boster Biological Technology Co., Ltd.) digestion prior to the addition to the β-TCP scaffolds. One scaffold was prepared with an osteoblast-induced suspension and another with a chondroblast-induced suspension. Subsequently, 3 ml cell culture media was added to each cultivation orifice plate. The BMSC-TCP composites were placed in an incubator for 2 days at 37°C in an atmosphere containing 5% CO₂ (Fig. 3). The composites were then placed in the culture chamber of a perfusion bioreactor with a flow rate of 3.5 ml/min for 3 weeks using culture media in the presence or absence of PRP. The culture medium was changed every 2–3 days. Adhesion, proliferation and growth of the BMSCs in the β-TCP scaffold were examined under a scanning electron microscope (SEM). The composites were divided into five groups as follows: i) Group A, composite cultured with PRP in the bioreactor; ii) Group B, composite cultured with PRP without the bioreactor; iii) Group C, composite cultured in the bioreactor without PRP; iv) Group D, composite cultured without PRP and the bioreactor and v) Group E, β-TCP scaffold only (negative control group).

Models of arthrodial cartilage defects were established in 20 joints of 10 beagles, with five defects in each joint. The beagles were anesthetized with an intravenous injection of 3% pentobarbital sodium. A median incision of 3 cm was made over the knee joints, soft tissues were dissected and the bone was exposed by gentle retraction of the muscles. Five defects with a depth of 10 mm and a width of 5 mm were created in the distal femoral articular surface with a hand brace.

Once the defects had been irrigated with sterile physiological saline solution, the BMSC-TCP composites were implanted into the defects and press-fitted. The osteoblast-induced BMSC-TCP composite (diameter, 5 mm; height, 5 mm) was implanted into the innermost region of the bone defect, and the chondroblast-induced BMSC-TCP composite of the same dimensions was implanted into the remaining shallow region of the bone defect. The osteoblast-induced BMSC-TCP composite was mechanically forced into contact with the chondroblast-induced composite. Following this, muscles, fascia and skin were separately closed over the regions of the defects using 4–0 sterile sutures. The animals were kept in standard conditions and were intramuscularly injected with 800,000 U penicillin (Wuhan Boster Biological Technology Co., Ltd.) each day for 3 consecutive days following the surgical procedure. The animals were then humanely sacrificed by anesthetic overdose (100 mg/kg) with pentobarbital sodium 3 months following the surgical procedure and the composites with tissue cells nearby were removed in order to perform reverse transcription-quantitative polymerase chain reaction (RT-qPCR) and histological analysis.

Template RNA used was from the total RNA of the BMSC-TCP composites. Total RNA was isolated using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. The concentration of RNA was determined using an ultraviolet spectrophotometer. QIAGEN RNase-Free DNAse was purchased from Cyagen Biosciences (Santa Clara, CA, USA). The cDNA sample was generated using the RevertAid First Strand cDNA Synthesis kit (Thermo Fisher Scientific, Inc.). Primers sequences used in RT-qPCR were as follows: Forward, 5′-TGTGCGGGGTCAAGGCTAAC-3′; reverse, 5′-GGCGTCCGAGTACCAGTTGC-3′ for ALP; forward, 5′-CTCCTTACCCGGATCCCCTG3′; reverse, 5′-GTAGAAGCGCTGGTAGGCGT3′ for BGLAP2. Reaction Buffer, RiboLock RNase Inhibitor, dNTP Mix, RevertAid M-MuLV Reverse Transcriptase were provided by Cyagen Biosciences Inc.. A negative control was included. Quantitative RT-PCR was performed using SYBR Green Realtime PCR Master mix (Toyobo, Osaka, Japan) in an Applied Biosystems AB7500 real-time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). Amplification was performed under the following thermal cycling conditions: 95°C for 60 sec followed by 40 cycles at 95°C for 15 sec and 60°C for 60 sec. GADPH was used as a reference gene, and the 2^−ΔΔCq method was used (20). RT-qPCR was performed in triplicate and the mean ± standard deviation was calculated.

Data are presented as the mean ± standard deviation and were analyzed using the SPSS, version 20.0 statistical software package (IBM SPSS, Armonk, NY, USA). One-way analysis of variance or paired Student's t-tests were used for comparisons between groups. P<0.05 was considered to indicate a statistically significant difference.

BMSCs from beagle bone marrow were harvested and characterized microscopically in a cell culture. Following a culture period of 3 days (Fig. 4A), a small number of spindle-shaped primary cells were observed growing against the wall of the cell culture flask. Cell colonies formed after 7 days of culture (Fig. 4B). After 7 days of culture, the cells gathered into swirl-shaped colonies (Fig. 4C).

BMSCs were induced to differentiate into chondroblasts for 7 days. The cells displayed brownish yellow cytoplasmic granules following type II collagen immunohistochemical staining (Fig. 5A). Subsequent to 7 days of osteoblastic differentiation, alkaline phosphatase and Alizarin red staining revealed a large quantity of gray-black granules or black deposits (Fig. 5B). The aforementioned observations are consistent with the initial stem cell phenotype, which was further confirmed by the expression of CD29 and CD44 following differentiation. Alizarin red staining revealed a large number of mineral nodes in the osteoblast-induced group (Fig. 5C and D).

Differentiated BMSCs were deposited on β-TCP scaffolds and allowed to grow for 21 days. SEM revealed that cells adhered to the scaffold and were widely distributed, secreting abundant extracellular matrix onto the scaffold (Fig. 6). Cell proliferation and a stretched, extended morphology was also observed, suggesting that the β-TCP scaffold is a good substrate for BMSC growth (Fig. 6).

Once the BMSCs had been cultured for 3 weeks in the bioreactor, BMSC scaffolds (Group A) or controls (Groups B-E) were implanted into an arthrodial cartilage defect model in beagles as described in the Materials and methods section. Histological examination demonstrated that several of the implanted BMSC scaffolds regenerated cartilage tissue, and filled the defective area. Numerous newly differentiated chondrocytes, in addition to abundant bone formation, were observed with many chondrocytes and osteoblasts present. Osteotylus was also observed and this was regularly arranged and dispersed throughout the site of the defect (Fig. 7A).

Several controls (bioreactor, PRP and BMSCs) were performed to investigate the importance of each component of the culture system. Implantation of BMSCs cultured on scaffolds without the bioreactor led to the formation of high levels of hyaline chondrocytes and cartilage lacunae, with the new cartilage integrating with the existing articular cartilage. Conversely, in Group A, new bone was not evenly distributed and the quantity of hyaline chondrocytes was reduced (Fig. 7B). In Group C, in which BMSCs were cultured in the absence of PRP, irregular, newly-formed osteotylus was present in the defect, and chondrocytes and osteoblasts were randomly and loosely arranged (Fig. 7C). In the absence of both PRP and the bioreactor (Group D), a small number of newly-generated chondrocytes and osteoblasts were detected in the defect. New bone formation was minimal, with abundant fibrous tissues infiltrating the damaged area (Fig. 7D). In the negative control with no BMSCs (Group E), several fibroblasts were observed to adhere to adjacent tissue on the defect border, resulting in the proliferation of fibrous and scar tissue (Fig. 7E).

The expression levels of osteogenic differentiation markers, ALP and BGLAP, in BMSC implants was determined by RT-qPCR (Fig. 8). In experimental Groups A, B, C and D, the expression levels of ALP and BGLAP were significantly higher compared with the negative control group (Group E; P<0.05). The expression levels of ALP and BGLAP were significantly higher in Group A compared with Group C (P<0.05) and in Group B compared with Group D (P<0.05). These results therefore suggested that the presence of PRP promotes bone regeneration. The expression levels of ALP were not significantly higher in Group A compared with Group B, or in Group C compared with Group D, (Group A compared with Group B, t=0.169, P>0.05; Group C compared with Group D, t=0.646, P>0.05). The present data supports the hypothesis of an osteogenic promoting effect of the bioreactor. Consequently, a novel method for the efficient generation of tissue-engineered cells for the repair of bone defects has been developed in the present study.