Cell culture and induction
All animal experiments were approved by the Animal Welfare and Ethics Committee of Southern Medical University. Rat SCs, thoracic aortic endothelial cells (AECs) and bone marrow-derived mesenchymal stem cells (BM-MSCs) were isolated according to previously described protocols [24-27]. BM-MSCs at passage 3 were differentiated into induced endothelial cells (IECs) or induced osteoblasts (IOBs) by corresponding induction medium for 21 days. The composition of culture medium used were shown in Table 1.
Cell identification
SCs were characterized by immunocytochemistry staining for S100. A mouse monoclonal antibody against S100 (1:200; Abcam, Cambridge, MA) was used, and the nuclei were counterstained with hematoxylin. AECs were characterized by immunofluorescence staining for von Willebrand factor (vWF). A rabbit polyclonal antibody against vWF (1:400; Abcam) and a DyLight 594-conjugated secondary antibody (1:200; Abcam) were used, and the nuclei were then stained with 4',6-diamidino-2-phenylindole (DAPI; Beyotime, Shanghai, China). Changes in the morphology and organelles of BM-MSCs and IECs were observed by light microscopy and transmission electron microscopy (TEM). IECs were characterized by immunofluorescence staining for vWF and Pecam1. The antibody against vWF (1:400) and a rabbit monoclonal antibody against Pecam1 (1:100; Abcam) were used, and the nuclei were stained with DAPI. IOBs were identified by staining with alkaline phosphatase (ALP; Beyotime) and alizarin red S (Sigma-Aldrich).
Conditioned medium
Confluent SCs in 10-cm dishes were rinsed three times with phosphate-buffered saline (PBS), and 10 ml of basal medium (high-glucose DMEM or M199) without FBS was added. This medium was harvested after 48 h and centrifuged for 5 min at 3000 x g. Then, the supernatant was collected as SC-conditioned medium (CM)-DMEM or SC-CM-M199 and stored at -70 °C until use.
Cell proliferation assay
The proliferation of IECs or AECs in SC-CM-DMEM or SC-CM-M199 supplemented with 2% FBS (for IECs) or 5% FBS (for AECs) without other supplements was detected by using the CyQUANT Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. In the control groups, the basal medium was DMEM or M199. The original number of cells in each well of a 96-well plate was 1.5×103 for IECs or 2.5×103 for AECs. The fluorescence intensity in triplicate wells each group was measured on days 1, 3, 5, 7, 9 and 11 at an excitation wavelength of 485 nm and an emission wavelength of 530 nm and converted to cell number according to a standard curve.
Cell migration assay
An endothelial cell migration assay was performed by using Transwell chambers (pore size of 8 μm, 6.5 mm in diameter; Millipore, Billerica, MA). In quadruplicate, IECs or AECs (1.2×104 cells/well) were seeded into the upper chambers in 120 µl of DMEM or M199 without FBS or other supplements, and 500 μl of SC-CM-DMEM or SC-CM-M199 (DMEM or M199 as a control) containing 0.5% bovine serum albumin (BSA; Beyotime) and 1% FBS was added to the lower chambers. The upper chambers were removed from the lower chambers after incubation at 37 °C for six hours and wiped with cotton swabs. The polycarbonate membranes were fixed using methanol and stained with crystal violet. The membranes were eluted by using 100 μl of 10% acetic acid. The optical density (OD) of the eluted fluid in triplicate wells was measured at an absorbance of 570 nm.
Capillary-like tube formation assay
Capillary-like tube formation was detected by using Matrigel (BD Biosciences, San Jose, CA). Liquid Matrigel was placed in a 96-well plate (60 μl/well). IECs or AECs (2×104 cells/well) were seeded on Matrigel in 40 μl of SC-CM-DMEM or SC-CM-M199. After incubation for 4 h at 37 °C, phase contrast micrographs were taken. Capillary-like meshes were quantified by ImageJ 1.4.3 software (National Institutes of Health, USA) in random fields in triplicate wells.
Corneal angiogenesis assay
IECs were resuspended in concentrated SC-CM-DMEM or DMEM at a density of 1×107 cells/ml. A total of 2 μl of medium containing cells was homogeneously mixed with 2 μl of 1% gelatin (Aladdin, Shanghai, China). The mixture was injected into the corneas of eight anesthetized eight-week-old female Sprague-Dawley (SD) rats according to the groups (4 rats per group), which were purchased from the Experimental Center of Southern Medical University (Guangzhou, China). After 12 days, the rats were sacrificed and perfused with waterproof drawing ink (Zhonghua, Shanghai, China) by intracardiac injection. The eyes were fixed with 10% neutralized buffered formalin, and then the corneas were excised to observe angiogenesis. Vascular segments were quantified by ImageJ 1.4.3 software in random fields in quadruplicate samples.
Indirect coculture of SCs and IECs
SCs and IECs were indirectly cocultured by using Transwell chambers (pore size of 0.4 μm, 24 mm in diameter; Millipore). SCs (1×104 cells/well) were seeded into the upper chambers in 1.5 ml of DMEM supplemented with 10% FBS without EGM-2 SingleQuots. IECs (1×104 cells/well) were seeded into the lower chambers in 2.6 ml of the same medium. IECs not cocultured with SCs were plated in the control wells. IECs were collected on days 3 and 7 to examine the effect of SCs on them.
Likewise, IECs (1×104 cells/well) were seeded into the upper chambers in 1.5 ml of DMEM supplemented with 10% FBS without other supplements. SCs (1×104 cells/well) were seeded into the lower chambers in 2.6 ml of the same medium. SCs not cocultured with IECs were plated in the control wells. SCs were collected on days 3 and 7 to examine the effect of IECs on them.
Quantitative reverse-transcription polymerase chain reaction (qPCR)
Total RNA was extracted from IECs, SCs and BM-MSCs by using an RNeasy Mini Kit (Qiagen, Valencia, CA) with on-column DNA digestion to eliminate genomic contamination. cDNA was synthesized by using the iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Hercules, CA). Real-time PCR was conducted by using SYBR Green PCR Master Mix (Applied Biosystems, Thermo Fisher Scientific). The primer sequences used were shown in Table 2. The fold changes in expression were calculated using the 2-ΔΔCt method [28].
Enzyme-linked immunosorbent assay (ELISA)
Secreted vascular endothelial growth factor (VEGF) levels were detected with a Quantikine® rat VEGF assay (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions. VEGF concentration in the SC-CM-DMEM was detected firstly. IECs were cultured in SC-CM-DMEM or DMEM with 2% FBS. The VEGF concentration in the supernatants of the wells was detected by ELISA on days 1, 3, 5, and 7. The medium in all the detected wells was fully replaced 24 h before each time point at which the supernatant was collected and analyzed.
Western blot (WB) analysis
Cells were harvested and lysed in radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime) to extract protein. The supernatants collected from the coculture system or the wells without coculture on days 3 and 7 were concentrated after constant volume, and the protein of supernatants was obtained. Equivalent amounts of protein were separated on a sodium dodecyl sulfate (SDS) gel (Beyotime) and electrotransferred to PVDF membranes (Millipore). WB analysis was performed using the appropriate diluted antibodies. Signals were detected using an ImageQuant LAS System (GE Healthcare Life Sciences, Marlborough, MA). Primary antibodies specific for Flt1, Kdr, nestin (Rat-401) (all from Santa Cruz Biotechnology, Santa Cruz, CA), and TIMP-2 (Thermo Fisher Scientific) and a control antibody specific for β-actin (Sigma-Aldrich) were used in this experiment.
Immunofluorescence staining
IECs or AECs were fixed and incubated with a mouse monoclonal antibody against nestin (1:200). Then, they were rinsed and incubated with a secondary antibody conjugated to DyLight 488 (1:200; Abcam). The nuclei were then stained with DAPI.
IECs cultured in SC-CM-DMEM for 7 days were first immunostained for nestin (1:200) as described above and then immunostained for Flt1 (1:200) following the same procedure. Different secondary antibodies, including the DyLight 488-conjugated antibody (1:200) for nestin and a DyLight 405-conjugated antibody (1:200, Abcam) for Flt1, were used. The cells were finally visualized under a TCS-SP5 confocal laser scanning microscope (Leica, Buffalo Grove, IL).
Cell transfection and labeling
BM-MSCs at passage 3 were infected with lentiviral medium (GeneChem, Shanghai, China) containing enhanced green fluorescent protein (EGFP) or red fluorescent protein (RFP) according to the manufacturer’s instructions to obtain BM-MSCs expressing EGFP or RFP, which were called BM-MSCs-EGFP or BM-MSCs-RFP, respectively. Stably infected BM-MSCs were selected with 1.5 μg/ml puromycin (Solarbio, Beijing, China). BM-MSCs-EGFP were differentiated into IOBs, and BM-MSCs-RFP were differentiated into IECs for scaffold loading. SC nuclei were stained with 100 μg/ml Hoechst 33342 (Beyotime) to obtain labeled SCs.
Preparation of scaffolds
Sterile cylindrical β-tricalcium phosphate (β-TCP) scaffolds (6 mm in length and 4 mm in diameter; pore size of 200~300 μm, porosity volume of >85%) were obtained from Bio-lu Biomaterials Company (Shanghai, China) and presoaked in DMEM overnight. In group I, labeled SCs and IECs-RFP in 200 µl of DMEM (1:1 ratio of cell number) were loaded into scaffolds at a concentration of 2×107 cells/ml using the negative pressure suction method. After 4 h, an additional 10 ml of DMEM was added to each well to completely submerge the scaffolds in medium. The same number of IOBs-EGFP in 200 µl of DMEM were loaded into scaffolds that had been incubated for 3 days. After 4 h, the scaffolds were incubated in DMEM for another 3 days. In group II, only IECs-RFP were loaded in the first three days, and other procedures were the same as those in group I. In group III, only IOBs-EGFP were loaded and incubated for 3 days.
Animal models
Thirty-six ten-week-old male SD rats were purchased from the Experimental Center of Southern Medical University and divided into three groups according to the scaffold groups described above (12 rats per group). The rats were anesthetized and sterilized. The left middle femur was exposed in the intermuscular space through a lateral longitudinal incision. Four holes were drilled in the outer side of the femoral shaft to allow two screws to be placed on either side of the proposed osteotomy site. After periosteal dissection, a 6-mm-long defect was created by using a scroll saw. A scaffold loaded with cells was implanted into the defect. A stainless steel linear reconstruction plate with four holes was then placed along the femur and fixed with four cortical screws. A 20-gauge steel wire was looped between the two screws on either side of the plate to strengthen fixation. The scaffold was fixed to the steel plate with two sutures to keep it stable. The subcutaneous and skin layers were closed in a routine manner.
Radiological observations
We randomly selected six rats from each group for radiological examination at 6 weeks and 12 weeks. The conditions of X-ray examination included a voltage of 40 kV, a current of 50 mA, and an exposure time of 200 msec. After the steel plate and screws were removed, micro-CT scans were performed at 12 weeks with an X-ray tube voltage of 80 kV, a current of 500 μA, and an exposure time of 200 msec. The average bone volume/total volume (BV/TV %) values of the femoral defect areas were analyzed for each group.
Histological assessments
After removing the steel plate and screws at 12 weeks, we obtained femur tissue samples from each group, embedded the samples in paraffin and sectioned scaffold segments. The sections were deparaffinized and blocked, and immunohistochemistry (IHC) staining was performed at 6 weeks with anti-nestin (1:500) and anti-TIMP-2 (1:200) primary antibodies and corresponding secondary antibodies. H&E and Masson staining were performed at postoperative 12 weeks according to routine procedures.
Statistical analysis
The data obtained are presented as the means±SDs. The data from the cell and animal model experiments were analyzed using Student’s t-test (parametric data) or the Mann-Whitney test (nonparametric data). GraphPad Prism 8.0.1 software (GraphPad Software Inc., San Diego, CA) was used to perform calculations and generate graphs. P <0.05 was considered statistically significant (***P<0.001, **P <0.01 and *P<0.05; ns=nonsignificant).