Cell culture

Primary human bone marrow stromal cells (hBMSCs) and primary human umbilical vein endothelial cells (hUVECs) were purchased from ScienCell Research Laboratories and cultured in mesenchymal stem cell medium (MSCM; ScienCell Research Laboratories, USA) and endothelial cell medium (ECM; ScienCell Research Laboratories, USA) respectively, within a humidified 5% CO2 incubator at 37 °C. The cultured cells between passages 3 and 6 were then used in following experiments.

Preparation and characterization of CMVs

HBMSCs and hUVECs were utilized upon reaching 90% confluence within a 10 cm culture dish and were washed three times with phosphate buffered saline solution (PBS) (pH 7.4), followed by incubation in 3 mL of serum-free MSCM and ECM containing 10 μg/mL Cytochalasin B (Solarbio, beijing, China) for 30 min (37 °C, 5% CO2), respectively. At the end of the incubation, cells formed CMVs and were treated with 0.25% (w/v) trypsin and 0.01% (w/v) ethylenediaminetetraacetic acid (EDTA) at 37 °C for 1 min. Then, the trypsin was inactivated by adding the same volume of FBS and the detached cells were vigorously vortexed for 30 s to separate the cells and newly-formed CMVs. The CMVs were separated from cells by two sequential centrifugation steps (5 min at 200 g and 20 min at 2 000 g). The BMSC-CMVs and EC-CMVs were collected and fixed with 2.5% (v/v) glutaraldehyde for 1 h. After CMVs were fixed for 1.5 h in 1% (v/v) osmium tetroxide at room temperature, the CMVs were dehydrated in a graded series of ethanol solutions (30, 70, 95 and 100 vol %, respectively) and propylene oxide. Then the samples were embedded in Durcupan (SigmaAldrich) and ultrathin sections were mounted on nickel grids and stained with uranyl acetate. Then the morphology of the BMSC-CMVs and EC-CMVs were observed under transmission electron microscopy (TEM). Then, the hBMSCs were stained with 10 μg/mL DiI (Solarbio, Beijing, China) at 37 °C for 15 min, while the hUVECs were stained with 10 μg/mL CFDA-SE (Carboxyfluorescein diacetate, succinimidyl ester, Solarbio, Beijing, China) at 37 °C for 1 h. The fluorescent-labeled BMSC-CMVs and EC-CMVs were obtained from fluorescent-labeled hBMSCs and hUVECs respectively and observed under laser scanning confocal microscopy (CLSM, Leica). The size and ζ- potential measurement of CMVs were measured by dynamic light scattering (DLS) at room temperature using Malvern Zetasizer nano ZS.

Alkaline phosphatase (ALP) staining

In this study, CMVs could be regarded as a sphere, thus their volume was estimated by the diameter of the sphere and their density defaulted to 1 g/m3. To assess the pro-osteogenic effects of EC-CMVs on hBMSCs, the hBMSCs were co-cultured with EC-CMVs and re-suspended in MSCM in different quantities (2.5, 5, and 10 μg). HBMSCs co-cultured with conditioned medium of hUVECs were used as a control; while untreated hBMSCs were used as a blank. The medium in all groups were exchanged every 2 to 3 days. Then the hBMSCs cultured for 7 days were subjected to alkaline phosphatase (ALP) staining. Cells were fixed with 4% (w/v) paraformaldehyde for 15 min, then washed with PBS and stained by an ALP staining kit (Beyotime).

Alizarin red staining

To assess the mineralization, hBMSCs cultured for 14 days were stained with alizarin red S, followed by quantification of the staining intensity. Cells were fixed in 4% (w/v) paraformaldehyde and stained with 1% (w/v) AR-S (Sigma Aldrich) for 30 min. Then, the Alizarin Red-stained cells were dissolved with 10% (w/v) cetylpyridinium chloride solution at a concentration of 100 mmol/L (Sigma Aldrich) for 30 min, and the absorbance of the supernatant was then measured at a wavelength of 570 nm.

Real-time PCR

The total RNA from cells was extracted using TRIzol reagent and the cDNA was synthesized with PrimeScript RT reagent kit (Takara Co. Japan), according to the manufacturer’s instructions. Then real-time quantitative polymerase chain (qRT-PCR) reaction was performed using SYBR Green qPCR Master Mix (Rox). Relative gene expression levels were calculated according to the cycle threshold (Ct) values relative to the endogenous housekeeping control gene (Gapdh) using QuantStudio Design & Analysis Desktop Software (Thermo Fisher Scientific). Glyceraldehyde-3-phosphate dehydrogenase (Gapdh) was used as the internal control. The primer sequences used for the experiment are listed in Table 1.

Table 1 Primer sequences utilized for quantitative real-time PCR analysisWestern blot analysis

The cells and CMVs were lysed by RIPA lysis buffer (Beyotime, China) containing a protease and phosphatase inhibitor cocktail (Thermo Fisher Scientific) on ice. After centrifugation at 12 000 g at 4 °C for 20 min, the total protein content was quantified by BCA protein assay kit (Beyotime, China). Then, the proteins were denatured and separated in SDS-polyacrylamide gels and were then transferred to a PVDF membrane by electrophoresis at 300 mA for 60 min. The membranes were then blocked with 5% (w/v) fat-free milk in TBST for 1 h and incubated at 4 °C overnight with the following primary antibodies: anti-Runx2 (1:1 000, Abcam), anti-BMP2 (1:1 000, Abcam), anti-VEGF (1:1 000, Cell signaling technology), anti-ephrinB2 (1:1 000, SAB), and anti-GAPDH (1:2 500, Abcam). Then, the membranes were incubated with secondary antibodies (1:5 000, Abcam) for 1 h at room temperature. The bands were then visualized by chemiluminescence using an ECL-PLUS kit (Pierce). The relative expression levels of proteins were compared through band density and the results were normalized to GAPDH.

Immunofluorescence

HBMSCs were fixed in 4% (w/v) paraformaldehyde solution for 20 min. After being washed by PBS for 3 times, the cells were permeabilized in 0.1% (w/v) Triton X-100 for 10 min and then blocked in 5% (w/v) bovine serum albumin (BSA) for 1 h at room temperature. Subsequently, the cells were incubated overnight at 4 °C with primary antibodies anti-ColI (1:500, Abcam) and anti-Runx2 (1:200, Abcam). After washing for 3 times in PBS, the cells were incubated with secondary antibodies (1:1 000, Abcam) for 1 h in the dark at room temperature and then stained with DAPI for 15 min. Images of hBMSCs were obtained using a laser scanning confocal microscope (CLSM, Leica).

Tube formation assays on Matrigel

The 24-well plates were coated with 250 μL Matrigel (Corning, America) per well and then placed at 37 °C for 30 min to allow Matrigel to form a gel. Then 10 × 104 cells/well of hUVECs (three replicates per group) were seeded on the Matrigel (Corning, America), co-cultured with different quantities of BMSC-CMVs (2.5, 5, and 10 μg) or without BMSC-CMVs, and cultured at 37 °C in 5% CO2. 10 × 104 hUVECs plated on the Matrigel were set as a blank control. 10 × 104 hUVECs co-cultured with conditioned medium of hBMSCs were set as a control. After incubation at 37 °C for 4 and 8 h, tube structures were observed using a phase contrast microscope and then quantified with ImageJ Pro Plus 6.0 Software.

Scratch wound assay

For the cell migration assay, hUVECs were seeded into 12-well plates and cultured to confluence. The scratch on the mono-layer was created using a P200 pipette tip, followed by washing three times with PBS to remove the detached cells. The cells were then co-cultured with BMSC-CMVs mixed in serum-free medium at different quantities. Cells cultured with serum-free medium were assigned as a blank group, while cells cultured with conditioned medium were assigned as a control. Then the cells were photographed at 6 h post-wounding and then analyzed using Image J pro plus 6.0 Software. The migration area (%) was calculated according to the formula: migration area = (M1-M0)/M0 x 100%, where M1 represents the wound area at 6 h and M0 represents the wound area at 0 h.

Flow cytometry (FCM) analysis

HBMSCs and hUVECs were co-cultured with 5 μg of DiI-labeled EC-CMVs and 5 μg of DiO-labeled BMSC-CMVs for 12 h, respectively. Then the cells were washed three times with PBS and analyzed by flow cytometry (BD Biosciences, San Jose, CA, USA). The untreated cells were set as a negative control. The laser line wavelengths used are 488 nm for DiO and 562 nm for DiI. A minimum of 10 000 events were acquired for each sample. The evaluation of the percentages of CMVs uptake by cells was analyzed using BD FACSDiva™ software version 8.0 (n = 3). The collected data were further evaluated using FlowJo software (TreeStar Inc., Ashland, OR, USA).

Preparation of CMVs with downregulated expression of ephrinB2

HBMSCs were seeded within 6-well culture dishes and transfected with the lentivirus vectors (Gnenchem, Shanghai, China) for knockdown of EFNB2 at a multiplicity of infection (MOI) of 50. The sequence of the shRNA was as follows: 5′-CCGGCGACAACAAGTCCCTTTGTAACTCGAGTTACAAAGGGACTTGTTGTCGTTTTTG-3′. Then, the transfected cells were expanded for the preparation of BMSC-CMVs using the previous method. The shRNA transfection efficiency was determined by Real-time PCR and western blot analysis.

Animal surgery and treatment

Nine 4 weeks old male BALB/c nude mice were purchased from the Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China) and housed in a pathogen-free facility. Nude mice were injected subcutaneously with 5 × 106 hUVECs mixed with BMSC-CMVs or BMSC-CMVs (EFNB2-shRNA) produced by 5 × 106 hBMSCs and hBMSCs (EFNB2-shRNA) mixed in Matrigel, respectively. Mice injected with 5 × 106 hUVECs were assigned as the blank control. Each group included 6 mice: (i) hUVECs; (ii) hUVECs + BMSC-CMVs in; (iii) hUVECs + BMSC-CMVs (EFNB2-shRNA).

Twenty-four 6 to 8 weeks old male SD rats were purchased from the Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China) and maintained in a pathogen-free facility. After being anesthetized by intraperitoneal administration of 100 mg/kg pentobarbital sodium (Sigma-Aldrich), a full thickness bone defect of 5 mm in diameter was prepared in each rat cranium. The rats were randomly allocated into 4 groups: (i) Matrigel; (ii) BMSC-CMVs produced by 5 × 106 hBMSCs mixed in Matrigel; (iii) EC-CMVs produced by 5 × 106 hUVECs mixed in Matrigel; (iv) BMSC-CMVs + BMSC-CMVs produced by 5 × 106 hBMSCs and 5 × 106 hUVECs respectively mixed in Matrigel. After 4 and 8 weeks, the skull samples were collected and fixed with 4% (w/v) paraformaldehyde for subsequent micro-CT scanning and histological analysis.

Micro-computed tomography (CT)

Micro-CT was performed to analyze new bone formation within the defects. The harvested calvaria samples were collected and examined using micro-CT as previously described.37 The samples were scanned using a micro-computed tomographic system (GANTRY- STD CT 3121; Siemens, Knoxville, TN). After three-dimensional (3D) visualization, bone volume analyses of the samples were carried out on the region of interest (ROI) using a microtomographic analysis software (Tomo NT; Skyscan, Belgium).

Hematoxylin and eosin (H&E) and Masson’s trichrome staining

After scanning by micro-CT, the samples were collected for histological analysis. The samples were fixed in 4% (w/v) paraformaldehyde solution, decalcified with 5% (w/v) EDTA and embedded in paraffin. Then the embedded samples were cut into 4-μm-thick sections (Leica Instruments GmbH, Hubloch, Germany) and sections from the defect region were stained with hematoxylin and eosin (H&E) according to the manufacturer’s protocol. Then histological sections of the tissues were processed with Masson’s trichrome staining to evaluate the degree of collagen maturity within the defect sites. The stained sections were imaged under an optical microscope (Olympus, Tokyo, Japan).

Immunohistochemistry

The paraffin-embedded samples were cut into 5 μm thick sections for immunohistochemistry analysis. The sections were incubated with antibodies against CD31 (Abcam) overnight at 4 °C, followed by incubation with secondary antibody (Servicebio). Immunoreactivity for CD31 was assessed by evaluating the positive area percentage values with Image J 2.0.0 software.

Statistical analysis

All experiments were performed with at least 3 replicates per group and all data were presented as mean ± SD. Comparisons between groups were evaluated by one-way ANOVA. Statistical analysis was conducted using PRISM software, version 7 and the threshold of statistical significance was set at P < 0.05.