Co-culture of the bone and bone marrow: a novel way to obtain mesenchymal stem cells with enhanced osteogenic ability for fracture healing in SD rats

A 6-week-old male SD rat was sacrificed with an injection of 10% chloral hydrate, and the femur and tibia were removed and placed into a sterile Petri dish.

① To obtain B-BM-MSCs, the medullary cavity was washed with PBS mixed with heparin sodium (0.04 mg/ml, H8270, Beijing Solarbio Science & Technology Co., Ltd.) until it appeared clean of all periostea, and the total marrow isolate was collected by centrifugation. Then, the clean femur and tibia were cut into 3 mm × 3 mm bone pellets and placed in a Petri dish with collagenase I (3 mg/mL, c8150, Beijing Solarbio Science & Technology Co., Ltd.). The dish was incubated in a cell incubator (37 °C, 5% CO₂) for 45 min [19]. At the same time, the bone marrow sample with a cell concentration of 2 × 10⁸/mL to 1 × 10⁹/mL was resuspended in PBS, 5 mL rat mesenchymal cell separation fluid (LGS1072, Tianjin HaoYang HuaKe Biological Technology Co., Ltd.) was added to a 15 mL centrifuge tube, the cell suspension was placed onto the separation liquid surface, and the tube was centrifuged at 450×g for 30 min at room temperature. The second layer of the milky white cell layer was placed in another 15 mL centrifuge tube and washed twice with PBS. Then, the cells were filtered through a filter with a pore size of 74 μm, and the filtered liquid was collected into a 6-cm Petri dish containing 6 mL MSC complete medium [DMEM/High Glucose (HyClone, USA) + 10% FBS (No. 04-001-1A, Biological Industries, Israel) + 1% streptomycin/penicillin (100× SV30010, HyClone, USA) + 50 μmol/L β-mercaptoethanol (M8210, Beijing Solarbio Science & Technology Co., Ltd.)]. After a 45-min incubation, 3 pieces of the bone were moved to the same Petri dish, and the dish was placed into a cell incubator. The medium was changed every 48 h, and the cells were maintained in the same medium until they reached approximately 80% confluence. These cells were considered passage 0. Cells were then trypsinized (0.5% trypsin, T8150, Beijing Solarbio Science & Technology Co., Ltd.) and re-cultured for the next passage.

② To obtain BM-MSCs, the cells in the bone marrow were obtained as mentioned above and cultured using the same approach as that for B-BM-MSCs. The isolation and culture of BM-MSCs were guided by the methods reported by Blashki [22].

Two types of P3 MSCs were transfected with lentivirus carrying the green fluorescent protein (GFP) gene (multiplicity of infection = 100). After 12 h, the medium containing the virus solution was discarded, and MSC complete medium was added for 2 or 3 days. The cells were then photographed with a microscope. The sample size was 12.

The P3 cells were harvested and adjusted to a concentration of 1 × 10⁶/mL prior to staining with antibodies against CD90 (11-0900-81), CD44 (12-0444-80), CD29 (17-0291-80), CD45 (11-0461-80), CD31 (25-0310-80) (Thermo Fisher Scientific, Ebioscience, USA), and CD106 (lot 130-103-684, Miltenyi Biotec, Germany). Fluorescence-activated cell identification was performed with flow cytometry (Beckman USA), and the data were analyzed with CytExpert (Tree Star, Ashland, OR, USA). The sample size was 12.

After a 9-day cultivation, the concentration of the 2 types of MSCs was adjusted to 2 × 10⁴ cells/mL, and these cell solutions inoculated in a 96-well cell culture plate with 100 μL/well. After culture for 1, 2, 3, and 4 days, 10 μL CCK-8 reagent (FC101-03, TransGen Biotech, Beijing, China) was added to each well, and the plates were incubated at 37 °C with 5% CO₂ for 2 h. The OD values were determined at 450 nm with a microplate reader (Bio-Rad, USA). The cell proliferation rate was calculated as follows: rate (day X) = [OD450 (day X)-OD450 (day X-1)]/OD450 (day X-1). The sample size was 6.

The P3 MSCs from the 2 different sources were seeded in 6-cm dishes with 2 × 10⁵ cells per dish. The cells were cultured at 37 °C with 5% CO₂ for 24 h. Then, the original medium was replaced with osteogenic induction medium containing 10⁻² mol/L sodium glycerophosphate (MB3195, Dalian Meilun Biotech Co., Ltd.), 10⁻⁷ mol/L dexamethasone (MB1434, Dalian Meilun Biotech Co., Ltd.), and 3 × 10⁻⁴ mol/L vitamin C (MB3195, Dalian Meilun Biotech Co., Ltd.). The induction medium was changed every 2 days, and the cells were induced for 14 days. Then, the cells were fixed, ALP and calcium nodules were stained with the modified Gomori Calcium-Cobalt method (DE0001, Beijing Leagene Biotech Co., Ltd.), and alizarin red staining (DS0002, Beijing Leagene Biotech Co., Ltd.) was performed according to the manufacturer’s instructions. The ratio of the positive area under each high-power field (RPA-HPT) was used to evaluate ALP expression and calcium nodules. The sample size was 6.

Cell samples were isolated with RIPA buffer (RIPA:PMSF = 100:1, R0020 and P8340, Beijing Solarbio Science & Technology Co., Ltd), and the total protein content was measured with the bicinchoninic acid protein assay kit (Thermo Scientific, Waltham, MA, USA). After the addition of loading buffer, the samples were boiled for 5 min for protein denaturation. Then, the samples were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) in a 12% gel under a constant voltage of 80 V for 30 min followed by a constant voltage of 110 V until the samples reached the bottom of the separation gel. Proteins were resolved by denaturing SDS-PAGE followed by transfer onto a nitrocellulose membrane. GAPDH (AB0037, Shanghai Abways Biotechnology Co., Ltd.) was used as the loading control. The membranes were incubated overnight at 4 °C with primary antibodies against TGF-β1 (1:1000, arg10002, Arigo Biolaboratories, Taiwan, China), BMP-2 (1:500, arg65980, Arigo Biolaboratories, Taiwan, China), and GAPDH (1:5000). The membranes were incubated with secondary antibodies (HS101-01 and HS201-01, Beijing TransGen Biotech Co., Ltd.) conjugated with horseradish peroxidase for 1 h. Target proteins were detected by an enhanced chemiluminescence system (4AWO12-050, Beijing 4A Biotech, Co., Ltd.) prior to development on X-ray film and photographic imaging to visualize the results. The sample size was 4.

Male SD rats weighing approximately 180 g to 220 g were anesthetized with 10% chloral hydrate. The right lower limbs were depilated with hair removal cream and disinfected with 2% iodophor and 75% alcohol. A longitudinal incision was made from the medial part 3 mm below the tibial platform to the medial malleolus at 10 mm, and the surface fascia at the incision site was cut. The tibia was dissected (Fig. 1a) and then cut in the middle with a wire clamp (Fig. 1b). The process required care to avoid damaging the fibula. The fracture area was washed with iodine and confirmed to be fully aligned (Fig. 1c). Then, 10 μL PBS or 3 × 10⁶ cells (dissolved in 10 μL PBS) were injected into the fracture sites of the control, BM-MSC, and B-BM-MSC groups with a 1-mL syringe. Then, the surface fascia and skin were sutured, the incision was wrapped with sterile gauze and fixed with plaster, and the treatment outcome was evaluated at subsequent time points. The sample size was 6.

At 2, 4, and 6 weeks after the operation, the fracture healing process was observed by computed tomography (CT) imaging, whereas visual inspection and hematoxylin and eosin (H&E) staining were used to evaluate the fracture healing process at weeks 4 and 6 after the operation. The sample size was 6.

The early-stage BM-MSCs (Fig. 2a, b) were polygon-shaped and round, while the early-stage B-BM-MSCs (Fig. 2e, f) were spindle-shaped, triangle-shaped, and polygon-shaped; some cells were clustered, and others were round and scattered. Both BM-MSCs and B-BM-MSCs tended to become fusiform or streamlined (Fig. 2c, g) over time. The primary BM-MSCs reached 80 to 90% confluence at days 14 to 16, while the primary B-BM-MSCs reached 80–90% confluence at days 9 to 11. However, mature BM-MSCs and B-BM-MSCs shared similar morphology (Fig. 2d, h) after GFP lentiviral transfection, and both adopted a fusiform or streamlined shape.

According to the flow cytometry results, CD90 (99.18% ± 0.15%), CD44 (98.47% ± 0.89%), and CD29 (99.39% ± 0.36%) were prominently expressed (> 97%) in B-BM-MSCs, whereas CD106 (0.11% ± 0.03%), CD45 (1.58% ± 0.31%), and CD31 (0.23% ± 0.02%) were barely expressed (< 2%) in B-BM-MSCs. These results were consistent with the results of BM-MSCs. These results indicated that B-BM-MSCs were successfully separated and cultured. It could be concluded that all the obtained MSCs were of high purity since MSC-associated markers were highly prevalent among these cells in every situation (Fig. 3).

From 24 to 96 h, the proliferative capacity of B-BM-MSCs was higher than that of BM-MSCs (Fig. 4a); the corresponding cell proliferation rates on day 4 were 110.94% ± 17.02% and 79.95% ± 11.21% (p < 0.05). In addition, the proliferative capacity of B-BM-MSCs was greater than that of BM-MSCs (Fig. 4b). These cells all showed the greatest proliferation ability from 72 to 96 h.

The results of alizarin red staining and the modified Gomori Calcium-Cobalt method showed considerable ALP expression and numerous calcium nodules after osteogenic induction for 14 days (Fig. 5a). The results showed that B-BM-MSCs (RPA-HPT, 26.28% ± 1.11%) induced more ALP-stained black plaques than BM-MSCs (RPA-HPT, 19.08% ± 1.23%) (p < 0.05), and B-BM-MSCs generated more calcium nodules (RPA-HPT, 43.05% ± 2.62%) than BM-MSCs (RPA-HPT, 6.81% ± 0.72%) (p < 0.001) (Fig. 5b). These findings showed that the osteogenic differentiation ability of B-BM-MSCs was better than that of BM-MSCs in vitro.

Western blot results showed that the expression of both TGF-β1 and BMP-2 was higher in B-BM-MSCs than in BM-MSCs before osteogenic induction. After a 2-week osteogenic induction, the expression of TGF-β1 and BMP-2 remained higher in B-BM-MSCs compared to BM-MSCs (Fig. 6). This finding further confirmed that the osteogenic potential of B-BM-MSCs was greater than that of BM-MSCs in vitro.

In the 2nd week after the operation, the control, BM-MSC, and B-BM-MSC groups all showed little callus formation at the bone fracture sites, while the number of calluses increased in week 4 after the operation in all 3 groups. In the 6th week after the operation, there was substantial new bone formation in the B-BM-MSC group, and the medullary cavity was recanalized, indicating good regeneration, while remodeling was not prominent in the control group or the BM-MSC group (Fig. 7a). Lane-Sandhu score analysis (Fig. 7b) after the 4th and 6th weeks showed that B-BM-MSCs (3.00 ± 0.81 and 9.67 ± 0.94, respectively) scored higher than BM-MSCs (1.33 ± 0.47 and 6.67 ± 1.25, respectively) (p < 0.05). This finding indicated that B-BM-MSCs had a greater ability than BM-MSCs to promote fracture healing in vivo.

To confirm our CT imaging findings, we collected bone specimens 4 and 6 weeks after construction of the rat fracture model. In the control group, some new calluses formed around the fracture area 4 weeks after surgery (Fig. 8a). The fracture site showed on a large number of soft tissue connections, and the healing condition was not good enough. At the 6th week, the peripheral osteophytes had basically been absorbed, and the cortical bones were connected (Fig. 8d). The situation in the BM-MSC group (Fig. 8b, e) was slightly better than that in the control group, while the fracture healing in the B-BM-MSC group (Fig. 8c, f) at weeks 4 and 6 was significantly better than that in the control group. The sample size of this experiment was 6. Representative data from one animal per group are shown, and similar results were obtained for the other animals in each group. At the 4th and 6th weeks, fracture healing in the B-BM-MSC group was better than that in the BM-MSC group, while fracture healing in the BM-MSC group was better than that in the control group. These data indicated that both BM-MSCs and B-BM-MSCs can effectively promote fracture healing, although B-BM-MSCs produced superior results to BM-MSCs.

At the 4th week, large numbers of chondrocytes and osteogenic cells were observed in trabeculae, large populations of osteoblasts were observed around trabeculae, and hematopoietic cell proliferation in trabeculae was extremely active in the B-BM-MSC and BM-MSC groups (Fig. 9b, c). In the control group (Fig. 9a), some osteoblasts were also observed around trabeculae, whereas hyperplasia of chondrocytes and osteogenic cells was not obvious. At the 6th week, in the control (Fig. 9d) and BM-MSC groups (Fig. 9e), large numbers of osteoblasts were visible, and cells in trabeculae were tightly packed, whereas in the B-BM-MSC group (Fig. 9f), trabeculae were mainly filled with hematopoietic cells, cells in the medullary cavity were loosely packed, and the trabeculae were more mature. The sample size of this experiment was 6, and consistent results were obtained in other groups. At the 4th and 6th weeks, fracture healing and reconstruction were better in the B-BM-MSC group than in the BM-MSC group, which showed better results than the control group in these 2 aspects. This finding further proved that B-BM-MSCs had a greater ability than BM-MSCs to promote fracture healing.