Cell culture and secretome preparation

The bone marrow MSCs were isolated as previously described [19, 20]. Briefly, cells were cultured in modified Eagle’s medium of Alpha (α-MEM; Invitrogen, USA) supplemented with 10 % fetal bovine serum (FBS; Gibco, USA) and 1 % penicillin/streptomycin (Gibco, USA) at 37 °C with 5 % CO2 and 95 % humidity. The MSCs from passages 3–6 were used in the experiments. Meanwhile, when MSCs reached 80 % confluence they were placed in serum-free α-MEM, which was used as the positive control in the animal studies, and incubated for 24 h in 5 % CO2 in a humidified condition, after which the conditioned medium was collected and centrifuged to purify for 10 min at 4 °C, 4000 g. Then 15 ml conditioned medium was re-centrifuged with Amicon Ultra Centrifugal Filters (Millipore, USA) for 60 min at 4 °C, 4000 g. Around 300–400 μl supernatant solution could be collected as cell-free secretome each time. For in vitro experiments, 100 μl secretome was added into 3 ml osteogenic induction medium (OIM), while for in vivo studies 100 μl secretome was locally injected to the regenerate zone. The protein content was measured using the BCSA kit (Thermo Scientific, Rockford, IL, USA) according to the manufacturer’s instruction. The concentrations of secretome used in vitro and vivo were 100 μg/μl and 3 mg/μl, respectively.

Cell viability assay

The rBMSCs were trypsinized and placed in a flat-bottomed 96-well plate at an initial density of 5000 cells per well. After 24 h of incubation, the medium was changed to hFMSC secretome containing medium at different doses (0, 10 μg/μl, 25 μg/μl, 50 μg/μl, 100 μg/μl, 200 μg/μl). Cells were incubated at 37 °C for 48 and 72 h. The proliferation was determined by methyl thiazolyl tetrazolium (MTT) reduction assay. After incubation, rBMSCs were treated using the MTT solution with a final concentration of 0.5 mg/ml for 4 h at 37 °C. The dark blue formazan crystals formed in intact cells were solubilized with 150 μl DMSO and the plate was shaken for 10 min. The absorbance was measured at 570 nm with a microplate reader.

Osteogenic differentiation of rBMSCs

Osteogenic differentiation induction was performed as previously described [21]. Briefly, the rBMSCs were trypsinized and placed in a 12-well plate at an initial density of 5000 cells/cm2. When over 80 % confluence was reached, the medium was replaced with osteogenic induction medium (OIM) including 1 nM dexamethasone, 50 μM L-ascorbic acid-2-phosphate and 20 mM βglycerophosphate. The osteogenic differentiation was evaluated by alkaline phosphatase (ALP) staining at day 3, Alizarin Red S staining at days 7 and 14, and quantitative real-time PCR examination of various osteogenic marker genes at days 3 and 10. Triplicate tests were conducted in each experiment.

ALP staining

After rBMSCs were treated with different kinds of secretome from rBMSCs, hFMSCs, and hAMSCs, as well as different doses of hFMSC secretome (0, 10 μg/μl, 25 μg/μl, 50 μg/μl, 100 μg/μl, and 200 μg/μl) for 3 days, the cells were washed with phosphate-buffered solution (PBS) twice and fixed with 70 % ethanol for 10 min. The cells were equilibrated by ALP buffer (0.1 M NaCl, 0.1 M Tris-HCl, 50 mM MgCl2.6H2O, pH 9.5) for 5 min twice, and incubated with ALP substrate solution (5 μl BCIP and 10 μl NBT in l ml ALP buffer) at 37 °C in the dark for 60 min, after which the reaction was stopped with distilled water and the plate was dried before taking photos.

Alizarin Red S staining

Alizarin Red S staining was performed to evaluate calcium deposit formation. After 7 days of osteogenic induction with different kinds of secretome (100 μg/μl) from rBMSCs, hFMSCs, and hAMSCs, and 7 and 14 days with different doses of hFMSC secretome (0, 10 μg/μl, 25 μg/μl, 50 μg/μl, 100 μg/μl, and 200 μg/μl), rBMSCs were washed with PBS and fixed with 70 % ethanol for 10 min, then the cells were stained with Alizarin Red S (pH 4.2) for 10 min at room temperature and washed three times with distilled water. To qualify the mineralization, the monolayer was eluted with 10 % cetylpyridinium chloride (CPC; Sigma), and the absorbance was measured at 570 nm.

RNA extraction and quantitative real-time PCR

After rBMSCs were treated with hFMSC secretome at a dose of 100 μg/μl for 3 and 10 days, total cellular RNA was extracted with RNA Mini Kit (Invitrogen), and reversely transcribed into cDNA with M-MLV reverse transcriptase (Invitrogen) according to the manufacturer’s instructions. Real-time PCR was performed using the Step One Plus Real-Time PCR System (Applied Biosystems, USA). The reaction conditions consisted of 10 μl reaction volumes with diluted cDNA template 1 μl, 5 μl SYBR-Green Master Mix (2×), 3.4 μl PCR-grade water, and 0.6 μl of each primer (10 μM). The amplification procedure was carried out as follows: first at 95 °C for 5 min, and then 40 cycles of 95 °C for 15 s and 60 °C for 60 s. Primer sequences were as follows: glyceraldehyde-3-phosphate dehydrogenase (GAPDH) forward: 5′GGCATGGACTGTGGTCATGAG3′, reverse: 5′TGCACCACCAACTGTTAGC3′; ALP forward: 5′ACCATTCCCACGTCTTCACATTT3′, reverse: 5′AGACATTCTCTCGTTCACCGCC3′; Runt-related transcription factor 2 (Runx2) forward: 5′ACTTCCTGTGCTCGGTGCT3′, reverse: 5′GACGGTTATGGTCAAGGTGAA3′; osteocalcin (OCN) forward: 5′CCTCACACTCCTCGCCCTATT3′, reverse: 5′CCCTCCTGCTTGGACACAAA3′; osteopontin (OPN) forward: 5′GTACCCTGATGCTACAGACG3′, reverse: 5′TTCATAACTGTCCTTCCCAC3′; Osterix (Osx) forward: 5′CCAGGCAACACTCCTACTCC3′, reverse: 5′GCCTTGCCATACACCTTGC3′. The relative quantification of gene expression was analyzed with the values of 2–ΔΔCT, normalized with GAPDH expression level.

Mixed rat peripheral blood lymphocyte reaction

For the mixed rat peripheral blood lymphocyte reaction, 1 × 105 peripheral blood lymphocytes (rPBLs) isolated from healthy rats were added into 100 μl α-MEM and plated on each well in 96-well plates. After 4 h of culture, different doses (0, 50 μg/μl, 100 μg/μl, and 200 μg/μl) of hFMSC secretome and hAMSC secetome in 100 μl α-MEM were added into each well. rPBSL culture with serum-free α-MEM served as the baseline control. After an additional 1, 3, and 5 days of culture, the proliferation of rPBLs was determined by bromodeoxyuridine (BrdU) incorporation assay according to the manufacturer’s manual (Cell Signaling Technology, USA). Optical density was measured at 450 nm.

Animal surgery and distraction osteogenesis protocol

Twenty-four 12-week-old SD male rats were used in our study. Before surgery, each rat was anesthetized with a solution of 0.2 % (vol/vol) xylazine and 1 % (vol/vol) ketamine in PBS. All animals were subjected to a right tibia transverse osteotomy procedure with a closed fracture at the midshaft near the fibula-tibia junction under sterile condition. Of note, the periosteum of the tibia should be retained as much as possible. A monolateral external distraction fixator (Tianjing Xinzhong Company, China) was placed to fix proximal and distal segments of the osteotomy site. Surgical incisions were then sutured sequentially. All rats were randomized equally into three groups with the following treatments: PBS group (n = 8); medium group (n = 8) and secretome group (n = 8).

The distraction protocol consisted of three phases: a latency phase of 5 days; a 10-day active lengthening phase (1 mm/day, in two steps, every 12 h); and a consolidation phase of 6 weeks. From the beginning of the consolidation phase, three groups received injection of PBS (100 μl), serum-free α-MEM (100 μl), and secretome (100 μl), respectively, into the distraction gap every 3 days until termination. All rats received subcutaneous injection of calcein (10 mg/kg; Sigma-Aldrich, St. Louis, MO, USA) at the beginning of the consolidation phase, and xylenol orange (30 mg/kg, Sigma-Aldrich) 3 days before termination (day 57 after surgery). Bilateral tibias were harvested, strapped free of muscle, and processed for further examinations.

Digital radiographs

At the end of lengthening, a weekly anterior-posterior x-ray including the distraction zone was taken until termination using a digital x-ray machine (MX-20, Faxitron X-Ray Corp., Wheeling, IL, USA) under an exposure time of 6000 ms and a voltage of 32 kv.

Micro-computed tomography (μCT)

The structural change within the distraction zone in the rat DO model was quantitatively assessed with μCT as previously described [22]. Briefly, all the specimens were imaged using a high-solution μCT (μCT40, Scanco Medical, Bassersdorf, Switzerland) at a custom isotropic resolution of 8 μm isometric voxel size with a voltage of 70 kV and a current of 114 μA. Three dimensional (3D) reconstructions of the mineralized callus were performed using a global threshold (165 mg hydroxyapatite/cm3), and a Gaussian filter (sigma = 0.8, support = 2) was applied to suppress noise. Sagittal images of the distraction zone were used to perform 3D histomorphometric analysis. The region of interest was defined as the distraction zone (regenerate) between the two closest proximal and distal half-pins. Low- and high-density mineralized tissues were reconstructed using different thresholds (low attenuation = 158, high attenuation = 211) using our established evaluation protocol with a small modification [23]. We selected the volume of interest to cover the distraction zone. The high-density tissues represented the newly formed highly mineralized bone, while the low ones represented the newly formed callus. Bone volume/total tissue volume (BV/TV) of each specimen was recorded for analysis.

Four-point bending mechanical testing

A mechanical test was performed within 24 h after termination at room temperature. The contralateral tibia was tested as an internal control. A four-point bending device (H25KS; Hounsfield Test Equipment Ltd., UK) with a 250 N load cell was used to test the tibia to failure. The tibias were loaded in the anterior-posterior direction with the inner and outer span of the blades set as 8 and 18 mm, respectively. The long axis of the tibia was placed perpendicular to the blades during the test [23]. The modulus of elasticity (E-modulus), ultimate load, and energy to failure were obtained and analyzed with built-in software (QMAT Professional; Tinius Olsen, Inc., Horsham, PA, USA). The biomechanical properties of the new bone were expressed as percentages of the contralateral intact bone properties.

Histology and immunohistochemistry

All tibias were initially fixed in 10 % formalin for 48 h. Half of them were followed by decalcification in 10 % EDTA solution for 3 weeks and embedded into paraffin. Thin sections (5 μm) were cut by a rotary microtome (HM 355S, Thermo Fisher Scientific, Inc., Germany) along the long axis of each tibia in the sagittal plane. After deparaffinization, immunohistochemistry staining was performed. The other half of the specimens were managed by gradient alcohol dehydration, xylene defatting, and undecalcification embedded in methyl methacrylate. Thin (5 μm) and thick (10 μm) sections were cut with the RM2155 hard tissue microtome (Leica, Wetzlar, Germany) along the long axis of the tibia, respectively. The 5-μm sections were stained with Trichrome Goldner and Von Kossa for static histomorphometric analysis, while the unstained 10-μm ones were used for dynamic histomorphometric measurements, which contained singled labeled surface (sL.S), double-labeled surface (dL.S), ratio of mineralizing surface to bone surface (MS/BS, calculated as double plus half of single-labeled surfaces (sL.S)), mineral apposition rate (MAR), bone formation rate per unit of bone surface (BFR/BS), bone formation rate of bone volume (BFR/BV), and bone formation rate of tissue volume (BFR/TV).

Immunohistochemistry staining was performed using a standard protocol as previously reported [24]. Secretions were treated with primary antibodies to rabbit osterix (Osx; Abcam, 1:100, ab22552) and osteocalcin (OCN; Santa Cruz, 1:100, sc30045) overnight at 4 °C; a horseradish peroxidase-streptavidin detection system (Dako, USA) was used, followed by counterstaining with hematoxylin. The positive stained cell numbers in the whole distraction zone per specimen in three sequential sections (50 μm, 150 μm, and 250 μm) per rat in each group were counted and compared, and were expressed as the percentages of the bone volume.

Statistical analysis

All quantitative data were analyzed using SPSS 18.0 software for windows (SPSS, Chicago, IL, USA). Non-parametric test was used for comparison of mean values with p < 0.05 considered as statistically significant.