A Novel Strategy Incorporated the Power of Mesenchymal Stem Cells to Allografts for Segmental Bone Tissue Engineering

Abstract

Mesenchymal stem cells (MSCs) hold great promise for bone regeneration. However, the power of mesenchymal stem cells has not been applied to structural bone allografts in clinical practice. This study designed a new strategy to enhance the efficiency of allografts for segmental bone regeneration. Isolated MSCs were cultured to form a cell sheet. The MSC sheet was then wrapped onto structural allografts. The assembled structures were cultured in vitro to evaluate the differentiation potential of MSC sheet. The assembled structures were implanted subcutaneously into nude mice as well as into the segmental radius defect of rabbits to investigate the efficiency of MSC sheets to repopulate allografts for bone repair. MSC sheets, upon assembling on bone grafts, showed similar differentiation properties to the in situ periosteum in vitro. After implantation the MSC sheets accelerated the repopulation of bone grafts in nude mice. Moreover, MSC sheets induced thicker cortical bone formation and more efficient graft-to-bone end fusion at the segmental bone defects in rabbits. This study thus presented a novel, more efficient, and practical strategy for large weight-bearing bone reconstruction by using MSC sheets to deliver large number of MSCs to repopulate the bone allografts.

Introduction

At present, the most common biomaterials for segmental bone defects reconstruction are allogeneic cortical bone grafts (e.g., cryopreserved bone grafts). However, the slow healing of allografts often results in graft fracture and poor clinical outcome. It was generally recognized that slow healing of allograft is likely due to the lack of live periosteal cells (12,27).

Mesenchymal stem cells (MSCs) are multipotential cells capable of differentiating into various mesenchymal lineages such as bone, cartilage, adipocytes, tendon, and ligament (5,17,23,24). MSCs isolated from about 10 ml of bone marrow could be easily cultured and amplified in vitro to yield billions of cells (9). A number of previous studies successfully utilized MSCs as seed cells for cartilage and tendon repair in animal models (11,21,22). Recent advances in coupling MSCs to different types of porous scaffold have achieved very exciting and promising results for bone tissue engineering (2,19,30). Despite the bone tissue engineering progress, the weak porous scaffolds are not suitable in reconstruction of large weight-bearing skeletal defects. As a result, few of the mesenchymal stem cell research achievements can be adapted to the bedside practice (26,29).

The techniques of populating large numbers of progenitor cells to weight-bearing allografts are thus desired to cover this gap between basic research and clinical practice. Currently, cell-seeding techniques employ either cell–gel composites or cell suspension to deliver cells into the scaffold (1,7,32). However, high porosity is required by these cell-seeding techniques, whereas cortical bone grafts are too dense for cell distribution by physical infiltration.

The present study aims to help overcome the challenges of applying the power of MSCs to clinical large skeletal defects reconstruction. We designed a strategy that synergistically incorporated the power of MSCs by using cell sheet technique to the strong bony allografts. Three-stage experiments were carried out: 1) the formation of MSC cell sheets and their osteochondral differentiation on bone grafts; 2) hMSC cell sheets to repopulate bone grafts in nude mice model; 3) the addition of rabbit MSC cell sheets to bone grafts for engineering segmental radius bone in rabbit model. The results of this study indicated that it is an effective and practical strategy to repopulate dense bony allograft and may constitute a feasible therapy for large skeletal defects in orthopedic practice.

Materials and Methods

Ethic Statement for Human Cells and Animal Experiments

Human bone marrow (BM) used in this study was the spare samples of three patients who underwent BM transplantation for bone nonunion treatment. The procedure was agreed to by patients and approved by the hospital ethics committee. Animal experiments were carried out according to the ethic regulation of the university animal center.

Isolation and Culture of MSCs

Bone marrow was aspirated from the iliac crest of humans or rabbits and collected into polypropylene tubes containing preservative-free heparin (1,000 U/ml). The bone marrow and heparin were well mixed. Bone marrow stroma cells were isolated by short-term adherence to plastic as described. Nucleated cells were isolated and plated in culture medium containing Dulbecco's modified Eagle's medium (DMEM; GIBCO, Grand Island, NY), fetal bovine serum (FBS, 15%, w/v; HyClone, Logan, UT), penicillin (100 U/ml), and streptomycin (100 mg/ml) (GIBCO). The nucleated cells were plated at a density of 5 million nucleated cells per 100-mm dish and incubated at 37°C with 5% humidified CO₂. After 24 h, nonadherent cells were discarded and adherent cells were cultured. The culture medium was changed every 3 days. When the culture dishes became nearly confluent after about 14 days, the cells were detached and serially subcultured. The second-passage (P2) cells were used in this study. Human MSCs were used for the experiments of in vitro and in nude mice models, whereas the rabbit MSCs were used for the experiment in a rabbit model.

According to a previous study (20), several monoclonal antibodies, including CD29, CD34, CD44, CD45, CD105, and CD166, were used to confirm the phenotype of MSCs. MSCs were directly labeled with monoclonal antibodies and were analyzed by a Coulter EPICS XL flow cytometer (Beckman-Coulter Corp., Hialeah, FL, USA)

Preparation of Bony Cylinder

Rabbit diaphyseal cortical bone cylinders, 1 cm in length, were placed in a jar under magnetic stirring. Bones were rinsed in distilled water for 2–3 h, in ethanol (70%, 96%, and 100%, consecutively) for 1 h, and in diethyl ether for 0.5 h, and then dried under a laminar flow hood. The bone grafts were completely demineralized in 0.6 M hydrochloride overnight or partially demineralized in hydrochloride for 0.5 h. The depth of partially demineralized grafts was calculated on the hard tissue sections. Then they were washed to remove the acid, dehydrated in ethanol and diethyl ether, and dried. All the procedures were performed at 4°C to prevent degradation of BMPs by endogenous proteolytic enzymes. The resultant completely or partially demineralized bone matrix (DBM) was stored at −20°C.

Fabrication of the MSCs Sheet and MSCs/Allograft Composites

After P2 MSCs were harvested, they were cultured in low glucose DMEM with 10% FBS and became confluence within 5–7 days. After confluence, MSCs were cultured in high glucose DMEM with 10% FBS and 50 μg/ml ascorbic acid. Within 2 weeks a MSC sheet was formed in a 75-cm² dish. It could be detached from the substratum by applying a small force. Then the completely or partially demineralized rabbit femur shaft was put on the cell sheet and rolled up to form MSC-allograft composit. The composites were cultured in complete medium with 50 μg/ml ascorbic acid and 10⁻⁷ M dexamethasone. The macromorphology and histology of the composite were evaluated at 3 week after culture.

Nude Mice Model

The efficiency of MSC cell sheet to repopulate the 1-cm-long demineralized bone grafts were tested in a nude mice model without in vitro culture of the composite of bone grafts and cell sheets. Six-week-old male athymic nude mice, weighing 20–25 g each, were used. Eight bone grafts with MSC cell sheets (group I) and eight bone grafts alone (group II) were implanted subcutaneously in the back of eight nude mice. Four implants from each group were retrieved at 3 or 6 weeks postoperatively and further processed by morphological examinations.

Rabbit Model

The potential of MSC sheet to repopulate completely demineralized bone cylinder was tested subcutaneously in a nude mice model. However, in most clinical cases, the allografts used are frozen bone grafts without demineralization. To mimic the clinical situation, allografts with surface demineralization were used in this study to evaluate whether MSC sheets can promote the efficacy of allografts for segmental weight-bearing bone reconstruction in vivo.

Fourteen front limbs of 4–5-month-old New Zealand White rabbits were shaved, prepped, and draped for aseptic surgery in the supine position. A 2-cm-long supermedial incision was made and the tissue overlying the diaphysis of the radius was dissected. A 1-cm segmental defect was prepared in the radius with a saw supplemented by copious sterile saline water irrigation. MSC-allograft was applied to the right defects, whereas the left defects were treated with allografts alone. Fixation of grafts was unnecessary because of the fibro-osseous union between the ulna and radius located distal and proximal to the surgical site. The soft tissue was approximated with interrupted No. 0 sutures and the skin was closed with No. 1 sutures. Immediately after surgery, the animals were returned to their individual cages. Water and food were supplied from the next day. A postoperative antibiotic gentamicin was administered intramuscularly at a dose of 6 mg/kg per day for 3 days.

Radiographic Evaluation

The radius-ulna complex containing the defect was taken A-P x-ray photo of a Kodak-FX in vivo imaging system (Kodak Inc., Rochester, NY).

Histology and Histomorphometry

Following euthanasia, experimental sites were recovered as a block and placed immediately into 10% phosphate-buffered formalin. After being fixed for 72?h, all samples were decalcified in 10% formic acid formalin solution for 14 days, followed by dehydration, paraffin embedding, and microsection with the direction parallel to the axis of the bone. Then sections were stained with hematoxylin and eosin (H&E) or Safranin O staining. Histological and histomorphometric observation was performed on an Olympus X71 microscope (Olympus, Tokyo, Japan). Three low magnification images (10×) were taken from each section for quantification. Image Pro plus 6.0 (Media Cybernetics, Inc, USA) software was used to quantify the thickness of periostea callus or the area of neotissue in the matrix. Five points were measured in each low magnification image, each point was between 0.5 mm. A xounting chamber was used for spatial calibration scale.

Statistic Analysis

Data are presented as mean ± SD. All data were evaluated by Kruskal-Wallis ANOVA with SPSS v13. A value of p ≤ 0.05 was considered statistically significant.

Results

Formation of the Cell Sheet and MSCs Differentiation In Vitro

MSCs revealed strong expression of CD29, CD44, and CD105 antigens. CD166 was only mildly expressed, while the expression of hematopoietic markers CD34 and CD45 ranged between very low to insignificant levels (data not shown). MSCs proliferated fast and formed coherent cellular sheets within 2 weeks after attaining confluence. The cell sheets could then be detached from the culture substratum (Fig. 1a). The total cell number of a MSC sheet from a 75-cm² dish was approximately 15 × 10⁶.

Figure 1.

Bone marrow-derived mesenchymal stem cells formed (a) cell sheets within 2 weeks of culture, which possess the differentiation potential to osteolineage and adipolineage exhibited by positive (b) Von Kossa staining and (c) Oil red staining. MSC sheet can easily wrap up with (d) allograft bone and differentiate into osteochondral lineage exhibited by (e) H&E staining and (f) Safarin O staining.

After 3 weeks of culture, the MSCs sheet on demineralized bone graft formed a periosteum tissue-like structure that had been induced into the osteochondral lineage (Fig. 1b) and adipose lineage (Fig. 1c). The cell sheet can be easily wrapped up with bone grafts (Fig. 1d). H&E staining (Fig. 1e) and Safarin O staining (Fig. 1f) of the cross and longitudinal sections showed the formation of osteochondro-precursor tissues around DBM. Moreover, MSCs around DBM had adopted hypertrophied chondrocyte morphology and calcium deposition was observed in the neotissues.

Repopulation of Allograft by hMSCs in a Nude Mice Model

MSCs were efficiently labeled with CFDA (Molecular Probes Inc., Eugene, OR) (Fig. 2a). After 3 weeks of implantation, MSCs migrated into the cavities at the surface (Fig. 2b) and inside of the bone matrix as deep as 0.7 mm (Figs. 2c and 3). This indicated that the MSCs from the cell sheets could survive and repopulate DBM.

Figure 2.

CFDA-labeled (a) MSCs can be observed at (b) the surface of allograft bone matrix and (c) the central area of allograft bone matrix after 3 weeks of implantation in nude mice model.

Figure 3.

Histology of (a, c) the allografts and (b, d) the MSC sheet–allograft at (a, b) 3 weeks and (c, d) 6 weeks after implantation in nude mice model (H&E staining, 100×).

From histological observation (Figs. 3 and Fig. 44), it was found that there was no inflammatory reaction around bone matrix with or without MSCs. Fibrous tissue was found in direct contact with the bone surface. All samples showed chondroneogenesis in crevices and cavities of the bone matrix. The lacunae of the matrix were enlarged by the cluster of chondrocytes. The chondrocytes were characterized by large, hypertrophied chondrocytes in H&E staining (Fig. 3) and positive for Safarin O staining (Fig. 4). The area of chondrocyte repopulation exhibited as positive Safarin O staining was caculated. Chondrocytes repopulated 2.5 ± 0.9% and 9.2 ± 1.2% of the demineralized allografts at 3 and 6 weeks, respectively, which was much less than the 5.6 ± 1.4% and 21.5 ± 2.3% in MSC-allograft (p < 0.05) (Fig. 4).

Figure 4.

Safarin O staining of (a, c) the allografts and (b, d) the MSC sheet–allograft at (a, b) 3 weeks and (c, d) 6 weeks after implantation (100×). (e) The area of chondrocyte repopulation of MSC–allograft was much larger than that of allograft in nude mice model (p < 0.05).

Implantation of MSC-Coupled Allografts Into Segmental Bone Defects in Rabbit Model

Radiological analyses were conducted at 4 and 8 weeks postsurgery to compare the healing of the two different grafts. Radiographs (Figs. 5 and 6) demonstrated the host–graft nonunion at some of the allograft-treated defects. Because the allografts were only deminerized at their surface with the depth of 200 ± 35 μm, the new bone formation could not be evaluated on radiological examination.

Figure 5.

Radiographic healing of allografts (L) and MSC–allografts (R). The histology of (b) allografts exhibited less repopulation than that of (c) MSC–allografts at 4 weeks after implantation in a rabbit model (H&E staining, 100×).

Figure 6.

Radiographic healing of allografts (L) and MSC–allografts (R). The histology of (b) allografts and (c) MSC–allografts at 8 weeks after implantation (H&E staining, 100×) exhibited (d) the thickness of periosteal callus bone at the surface of MSC–allograft was much larger than that of allograft in a rabbit model (p < 0.05).

Histologic sections and staining were carried out to characterize bone formation in both types of grafts. Periosteal bone formation was observed on the surface of partially demineralized bone grafts. However, compared to MSC–allografts at 4 and 8 weeks postoperatively, the new bone formation was dramatically less on the surfaces of allografts alone (Figs. 5 and 6). Bone bridging occurred in all of the three MSC–allografts at 4 weeks. In contrast, only one of the allografts achieved healing at the host-graft junctions. At 8 weeks, three of the four allografts achieved host-graft healing. Histomorphometry was used to quantify new bone formation at the periosteal surface of the graft (Figs. 5 and 6). Compared to MSC-allograft, new bone formation on the side of the grafts (Figs. 5 and 6) was significantly less in allografts at 4 and 8 weeks. MSC-allografts healing was characterized by thicker periosteal callus bone formation (1161.5 ± 156 μm, 8 weeks), compared to cell-free allografts (318.7 ± 87.3 μm, p < 0.01).

At 4 weeks, a large amount of new bone formation was observed on the shaft of MSC-allograft (Fig. 5). By 8 weeks, new bone on the graft showed mature cortical bone structure after the procedure of remodeling (Fig. 6).

Discussion

The present study demonstrated that: 1) MSCs can form cell sheets with osteochondral differentiation potential; 2) MSC cell sheets help incorporate large number of MSCs with dense bone grafts; 3) MSC cell sheets can survive and accelerate the repopulation of bone matrix after implantation; 4) the structures of MSC cell sheets with allografts can efficiently reconstruct weight-bearing bone defect. These findings identified a novel strategy that efficiently adds the power of mesenchymal stem cells to allografts for segmental bone reconstruction.

Large Number of MSCs Were Efficiently Incorporated Into Dense Bone Grafts

The present study successfully incorporated about 15 × 10⁶ MSCs into the 1-cm-long nonporous cortical bone graft by assembling MSC sheets with dense allografts. This overcame the inherent disadvantages of current cell-seeding techniques in bone tissue engineering. It was known that the key factor in tissue repair is the availability of appropriate cells. The presence of cells is crucial because of their proliferation potential, cell-to-cell signaling, biomolecules production, and formation of extracellular matrix, all of which strongly influence the nature of skeletal tissue formation. It seems clear that a threshold quantity of cells is required at the repair sites for normal neotissue formation (4). With the MSC sheets, more than threshold amount of the quantity of seed cells can be delivered to the repair site.

High mechanical strength and large cell attachment surface are two antagonistic goals in scaffold design. Scaffolds often need to sacrifice strength by increasing porosity for cell attachment (28). However, the strength of tissue grafts is the priority in most clinical conditions. Hence, the tissue grafts have to be preserved intact, which, unfortunately, largely hampered the cell seeding. With MSCs sheet technique, their synthesized matrixes that facilitated the assembly of cells within dense grafts, without compromising the strength of tissue grafts, connected cells. This study created the possibility of engineering strong and living tissue grafts for large segmental skeletal defect repair in clinical practice.

The MSC Sheet Differentiated Into Osteochondral Lineage on the Surface of Completely and Partially Demineralized Bone Allograft

In the present study, MSCs sheet technique not only served as a platform to connect cells, but also maintain MSCs differentiation potential. It was evidenced by the cartilage-like layers around demineralized bone graft after 3 weeks of culture in vitro. This suggested that MSC sheets possess osteochondrogenic differentiation potential (8). The in vivo findings from the nude mice model also confirmed the present novel strategy improved the repopulation of allografts as well as initiated the stage of chondral differentiation in the procedure of endochondral ossification.

There are a few factors that may influence the induction of MSC sheet into osteochondral lineage. The demineralized bone (DBM) is the most reported (18). Probably due to the presence of bone morphogenetic proteins within it, DBM displayed osteogenic potential, which could efficiently induce cells towards bone differentiation, as evidenced by numerous reports based on both in vitro and in vivo models (6,18).

Although allogeneic demineralized bone matrix can induce osteogenesis via endochondral ossification (14), it cannot work well alone on treating large bone defects. A common complication of using allografts is allograft fracture. There are currently no effective ways to improve it. We therefore used the partially demineralized bone for the in vivo experiments in the present study. As expected, the results indicated that partially demineralized bone not only displayed good mechanical strength but also exhibited inductive potential of osteochondrogenesis, which was verified by periosteal bone formation on the surface of allografts in the rabbit model. However, a study could be proposed to systemically evaluate the relationship between partial bone demineralization and graft strength in order to shape a better strategy for future clinical applications.

MSC Sheets Acted as Fresh Tissue-Engineered Periosteum to Improve Segmental Bone Defect Repair

The cartilage-like layers around demineralized bone graft observed in vitro suggested that MSC sheets could act as fresh periosteum. More importantly, we demonstrated an abundant periosteal bone formation on MSC–allograft, which was much less in allograft alone. The difference of bone formation between the groups was very similar to what had been reported in the previous studies. For example, one study compared the use of fresh bone graft and frozen allograft for bone regeneration in a rabbit model and demonstrated that free autogenous periosteum wrapped around frozen allografts dramatically stimulated the healing and repair (16). Another study transplanted live bone graft harvested from Rosa 26A mice into murine segmental femoral bone defects and showed that approximately 70% of osteogenesis on the graft was attributed to the expansion and differentiation of donor periosteal progenitor cells (31). Moreover, another study using a free subcutaneous isograft demonstrated that marrow cells and osteocytes have little or no contribution to early osteogenesis, while live cells in periosteum and endosteum plus stromal cells are responsible for 90% of the early osteogenesis (13,15). Also it is known from clinical experience that preservation of the periosteum or use of a periosteum tube graft significantly improves cortical bone graft incorporation and remodeling (10,25). Based on the findings in a rabbit model together with those in vitro and in nude mice model of this study, it seems that MSC sheets acted as fresh tissue-engineered perioteum to repopulate allograft and improved the bone formation.

Conclusion

In summary, MSCs possess great proliferative capacity and at least three mesodermal lineage differentiations potential. The proliferation potential enables bMSCs to form cell sheets much faster compared to terminally differentiated cell types. The differentiation potential of MSCs enhances the application of the cell sheet technique, providing a wider scope of application for connective tissue engineering. More importantly, assembly of MSC sheets and large allografts provides a convenient and practical tissue engineering strategy for clinical regeneration of large musculoskeletal defects. However, a number of questions have to be answered before this technique goes to clinical practice, such as the time of MSC culture and MSC cell sheet formation, the immunogenicity of MSC cell sheets, etc. Taken together, our results, combined with others, strongly indicate the future directions to enhance allograft healing and repair via tissue engineering and stem cell engraftment (3,17,24).