Synovial Stem Cells Are Regionally Specified According to Local Microenvironments After Implantation for Cartilage Regeneration

Authors


Abstract

We previously demonstrated that synovium-derived MSCs had greater in vitro chondrogenic ability than other mesenchymal tissues, suggesting a superior cell source for cartilage regeneration. Here, we transplanted undifferentiated synovium-derived MSCs into a full-thickness articular cartilage defect of adult rabbits and defined the cellular events to elucidate the mechanisms that govern multilineage differentiation of MSCs. Full-thickness osteochondral defects were created in the knee; the defects were filled with 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate-labeled MSCs and covered with periosteum. After 4 weeks, although the cell density decreased, transplanted MSCs produced a great amount of cartilage matrix extensively. The periosteum became thinner, and chondroprogenitors in the periosteum produced a small amount of cartilage matrix. In the deeper zone, transplanted MSCs progressed to the hypertrophic chondrocyte-like cells. In the deep zone, some transplanted cells differentiated into bone cells and were replaced with host cells thereafter. In the next phase, the border between bone and cartilage moved upwards. In addition, integrations between native cartilage and regenerated tissue were improved. Chondrocyte-like cells derived from the transplanted MSCs still remained at least after 24 weeks. Histological scores of the MSC group improved continuously and were always better than those of two other control groups. Immunohistological analyses and transmission electron microscopy confirmed that the MSCs produced abundant cartilage matrix. We demonstrated that transplanted synovium-derived MSCs were altered over a time course according to the microenvironments. Our results will advance MSC-based therapeutic strategies for cartilage injury and provide the clues for the mechanisms that govern multilineage differentiation of MSCs.

Introduction

MSCs are expected to play important roles in development, postnatal growth, repair, regeneration, and homeostasis of the body. They are easy to isolate, have extensive self-renewal potential, and have multilineage differentiation potential including chondrogenic differentiation [1, [2]–3]. Based on these attributes, MSCs have potential for use in regenerative medicine for cartilage injury.

Despite the fact that bone marrow is considered a well-accepted source of MSCs [1], various studies have been reported that MSCs can be isolated from various adult mesenchymal tissues, including synovium [4, [5], [6], [7]–8]. Synovium-derived MSCs were first reported by De Bari et al. [4]. They showed that synovium-derived MSCs had great proliferation potential and had multilineage differentiation potential in vitro. We previously compared human MSCs derived from bone marrow, synovium, periosteum, adipose tissue, and muscle and determined that synovium-derived MSCs had greater expansion and chondrogenic ability in vitro than MSCs from other tissues [9]. This suggests that synovium-derived MSCs are superior as a source for cartilage regeneration, although they have not been compared extensively in vivo.

In a large number of animal transplantation studies, MSCs expanded in vitro were able to differentiate into cells of the residing tissue and repair tissue damaged by trauma or disease [10, 11]. Despite diverse and growing information regarding MSCs and their use in cell-based strategies, the mechanisms that govern MSC self-renewal and multilineage differentiation are not well-understood and remain an active area of investigation.

For full-thickness articular cartilage defects, transplantation of MSCs in collagen gel with periosteum covering has been attempted. Although some studies have reported successful results [12, 13], a number of questions such as whether the donor cells differentiated into chondrocytes or how donor cells contributed to chondrogenesis still exist, limiting clinical applications for cartilage injury.

In this study, we isolated MSCs from the synovium of adult rabbits. After expansion in vitro, we transplanted 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiI)-labeled MSCs into a full-thickness articular cartilage defect. Intensive histological analyses demonstrated that transplanted MSCs were altered over a time course according to local microenvironments. This will advance the clinical application of MSCs for cartilage injury and provide the clues to clear the mechanisms that govern multilineage differentiation of MSCs.

Materials and Methods

Animals

Skeletally mature Japanese White Rabbits weighing about 3.2 kg (range 2.8–3.6 kg) were used in the experiments. Animal care was in strict accordance with the guidelines of the animal committee of Tokyo Medical and Dental University. Synovium was harvested or operation was performed under anesthesia induced by intramuscular injection of 25 mg/kg ketamine hydrochloride and i.v. injection of 45 mg/kg sodium pentobarbital.

Isolation and Culture of Synovial Cells

Harvested synovium was digested in a 3 mg/ml collagenase D solution (Roche Diagnostics, Mannheim, Germany, http://www.roche-applied-science.com) in α-minimal essential medium (αMEM) (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) at 37°C. After 3 hours, digested cells were filtered through a 70-μm nylon filter (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com), and the remaining tissues were discarded.

The digested cells were plated at 5 × 104 cells/cm2 in 60-cm2 culture dishes (Nalge Nunc International, Rochester, NY, http://www.nalgenunc.com/) in complete culture medium: αMEM containing 10% fetal bovine serum (FBS) ([lot selected for rapid growth of bone marrow-derived MSCs; Invitrogen], 100 units/ml penicillin [Invitrogen], 100 μg/ml streptomycin [Invitrogen], and 250 ng/ml amphotericin B [Invitrogen]) and incubated at 37°C with 5% humidified CO2. After 3–4 days, the medium was changed to remove nonadherent cells and then cultured for 14 days as passage 0 without refeeding. Then, the cells were trypsinized, harvested, and replated as passage 1 at 50 cells/cm2 in 145-cm2 culture dishes [3]. After an additional 14 days of growth, the cells were harvested and cryopreserved. To cryopreserve the cells, they were resuspended at a concentration of 1 × 106 cells/ml in αMEM with 5% dimethylsulfoxide (Wako, Osaka, Japan, http://www.wako-chem.co.jp/english) and 20% FBS. Aliquots of 1 ml were slowly frozen and cryopreserved in liquid nitrogen (passage 1). To expand the cells, a frozen vial of the cells was thawed, plated at 50 cells/cm2 in 145-cm2 culture dishes, and incubated for 4 days in the recovery plate. These cells (passage 2) were used for further analyses.

Colony-Forming Unit Assay

One hundred cells were plated in 60-cm2 dishes, cultured in complete medium for 14 days, and stained with 0.5% crystal violet in methanol for 5 minutes.

In Vitro Differentiation Assay

One hundred cells were plated in 60-cm2 dishes and cultured in complete medium for 14 days. For adipogenesis, the medium was then switched to adipogenic medium that consisted of complete medium supplemented with 10−7 M dexamethasone (Sigma-Aldrich Corp., St. Louis, http://www.sigmaaldrich.com), 0.5 mM isobutylmethylxanthine (Sigma-Aldrich Corp.), and 50 μM indomethacin (Wako) for an additional 21 days. The adipogenic cultures were fixed in 4% paraformaldehyde and then stained with fresh oil red O solution [14]. For osteogenesis, the medium was then switched to calcification medium that consisted of complete medium supplemented with 10−9 M dexamethasone, 20 mM β-glycerol phosphate (Wako), and 50 μg/ml ascorbate-2-phosphate (Sigma-Aldrich Corp.) for an additional 21 days. These dishes were stained with 0.5% alizarin red solution [15].

For chondrogenesis, 250,000 cells were placed in a 15-ml polypropylene tube (Becton, Dickinson and Company) and centrifuged at 450g for 10 minutes. The pellet was cultured in chondrogenesis medium consisting of high-glucose Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 500 ng/ml bone morphogenetic protein-2 (Astellas Pharma Inc., Tokyo, http://www.astellas.com), 10 ng/ml transforming growth factor-β3 (R&D Systems, Minneapolis, http://www.rndsystems.com), 10−7 M dexamethasone (Sigma-Aldrich Corp.), 50 μg/ml ascorbate-2-phosphate, 40 μg/ml proline, 100 μg/ml pyruvate, and 1:100 diluted ITS+Premix (6.25 μg/ml insulin, 6.25 μg/ml transferrin, 6.25 ng/ml selenious acid, 1.25 mg/ml bovine serum albumin [BSA], and 5.35 mg/ml linoleic acid; BD Biosciences Discovery Labware, Bedford, MA, http://www.bdbiosciences.com/). For microscopy, the pellets were embedded in paraffin, cut into 5-μm sections, and stained with Toluidine blue [16, [17]–18].

Cell Labeling and Preparation of Collagen Implants

Passage 2 cells were resuspended at 1 × 106 cells/ml in αMEM, and a fluorescent lipophilic tracer, DiI (Molecular Probes, Eugene, OR, http://probes.invitrogen.com), was added at 5 μl/ml in αMEM. After incubation for 20 minutes at 37°C with 5% humidified CO2, the cells were centrifuged at 450g for 5 minutes and washed twice with PBS, and 5 × 106 DiI-labeled cells were resuspended in 50 μl of αMEM with 20% FBS. They were then mixed with an equal volume of collagen gel (Atelocollagen, 3% type I collagen; Koken, Tokyo, http://www.kokenmpc.co.jp/english/index.html). The mixture was placed into a six-well plate (Becton, Dickinson and Company), 3 ml of αMEM with 20% FBS was added to the plate, and the plate was incubated at 37°C for 1 day to allow for contraction [19].

In Vivo Transplantation

Sixty skeletally mature Japanese White Rabbits were used in this study. The rabbits were anesthetized, the right knee joint was approached through medial parapatellar incision, and the patella was dislocated laterally. Full-thickness osteochondral defects (5 × 5-mm wide, 3 mm deep) were created in the trochlear groove of the femur, and the animals were divided into three groups: “MSC,” “Gel,” and “Defect.” In the MSC group, the defects were filled with DiI-labeled MSCs embedded in collagen gel at 5 × 107 cells/ml. In the Gel group, the defects were filled with a mixture containing an equal volume of αMEM with 20% FBS and collagen gel without cells. In these two groups, the defects were covered with autologous periosteum that had been obtained from the medial proximal tibia. The periosteum was sutured to each corner of the defect with 6-0 nylon sutures with the cambium layer facing down. In the Defect group, the defects were left empty. All rabbits were returned to their cages after the operation and were allowed to move freely. Animals were sacrificed with an overdose of sodium pentobarbital at 1 day and 4, 8, 12, and 24 weeks after the operation (n = 4). Three samples in each group were examined histologically, and one sample in each group was examined by transmission electron microscopy (TEM).

Macroscopic Examination

Samples were examined macroscopically for color, integrity, and smoothness. Osteoarthritic changes and synovitis of the knee were also investigated. After the examination, the distal femurs were dissected and were photographed, and then the integration of donor with host was quantified by the ratio of the integrated margin to the total margin in Gel and MSC groups at 12 weeks. The length of total and integrated margin between the host cartilage and repaired tissue was measured, and then the percentage of the integrated margin to the total margin was calculated.

Histological Examination and Fluorescent Microscopic Examination

The dissected distal femurs were fixed in a 4% paraformaldehyde solution immediately. The specimen was decalcified in 4% EDTA solution, dehydrated with a gradient ethanol series, and embedded in paraffin blocks. Sagittal sections of 5-μm thickness were obtained from the center of each defect and stained with Toluidine blue. Sections dedicated for fluorescent microscopic visualization of DiI-labeled cells were not stained with Toluidine blue, and nuclei were counterstained by 4′, 6-diamidino-2-phenylindole dihydrochloride (DAPI).

Immunohistochemical Examination

Paraffin-embedded sections were deparaffinized using xylene and dehydrated through graded alcohols. The samples were pretreated with 0.4 mg/ml proteinase K (DAKO, Carpinteria, CA, http://www.dako.com) in Tris-HCl for 15 minutes at room temperature for antigen retrieval. Residual enzymatic activity was removed by washing in PBS and nonspecific staining was blocked with PBS containing 10% normal horse serum for 20 minutes at room temperature. For type I and type II collagen, primary antibodies (mouse anti-human anti-type I and type II collagen, 1:1,000 dilution; Daiichi Fine Chemical, Toyama, Japan, http://www.daiichi-fcj.co.jp/en/home.html) was placed on the sections for 1 hour at room temperature. For type X collagen, primary antibody (mouse anti-human anti-type X collagen, 1:10 dilution) [20, 21] was placed on the sections overnight at 4°C. After extensive washing with PBS, a secondary antibody of biotinylated horse anti-mouse IgG (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com) was placed on the sections for 30 minutes at room temperature. Immunostaining was detected with VECTASTAIN ABC reagent (Vector Laboratories), followed by 3,3′-diaminobenzidine staining. Counterstaining was performed with Mayer-hematoxylin.

Transmission Electron Microscopic Examination

Samples were fixed with 2.5% glutaraldehyde in 0.1 M PBS for 2 hours, washed overnight at 4°C in the same buffer, and postfixed with 1% OsO4 buffered with 0.1 M PBS for 2 hours. Then, the tissues were dehydrated in a graded series of ethanol and embedded in Epon 812. Ultrathin (90-nm) sections were collected on copper grids and double-stained with uranyl acetate and lead citrate and then examined by TEM (H-7100; HITACHI Ltd., Hitachinaka, Japan, http://www.hitachi.com/) [22].

Statistical Analysis

To assess differences, Mann-Whitney U tests were used. A value of p < .05 was considered significant.

Results

Characteristics of Synovium-Derived MSCs

Cells from rabbit synovium formed cell colonies and differentiated into adipocytes, osteoblasts, and chondrocytes when cultured in the appropriate medium. These results indicate that the rabbit synovium-derived cells had characteristics of MSCs (Fig. 1) [1, 23, 24].

Figure Figure 1..

Synovium-derived cells from rabbit have characteristics of MSCs. (A): Colony-forming potential. Colonies were stained with crystal violet. (B): Adipogenic potential. Adipocyte colonies were stained with oil red O. (C): Osteogenic potential. Calcified colonies were stained with alizarin red. (D, E): Chondrogenic potential. (D): Macropicture of a pellet with a 1-mm scale. (E): Histological section of the pellet stained with Toluidine blue. Bars = 50 μm.

Macroscopic Observation

In repeated experiments, we observed no technical failures macroscopically, such as periosteal delamination. Representative samples in each group are shown in Figure 2A. At 1 day, the cartilage defect was overlaid with blood clots in the Defect group. In the Gel and MSC groups, the defect was shown to be covered with periosteum. At 4 weeks, the center of the defect appeared a little whitish in the Defect group. The periosteum that covered over the defect became whitish in the Gel and MSC groups, whereas continuity between sutured periosteum and neighboring cartilage appeared better in the MSC group. At 12 weeks, the defect was still observed in the Defect group. In the Gel group, although the border between periosteum and the neighboring cartilage became smoother, periosteum was still observed distinctly. In the MSC group, cartilage defect appeared glossy, smooth, and similar with neighboring cartilage, and the margin of the repaired tissue seemed to integrate into the surrounding native cartilage. The integration rates were quantified in the Gel and MSC groups at 12 weeks (Fig. 2B). The margin of the repaired tissue in the MSC group was significantly better integrated than that in the Gel group at 12 weeks. In these three groups, there was no obvious synovitis and no severe osteoarthritic change, but mild spur formation was observed on the edge of the trochlear groove in some samples of the control group.

Figure Figure 2..

Macroscopic observation of cartilage defect after cell transplantation. (A): Femoral condyles at 1 day and 4 and 12 weeks after the surgery are shown. The corners of the margin between donor and host were indicated as arrowheads in the MSC group at 12 weeks. In the MSC group, cartilage defect appeared similar with host cartilage, and the margin of the repaired tissue seemed integrated into the surrounding native cartilage. (B): The integration rate in Gel and MSC groups at 12 weeks. The margin of the repaired tissue in the MSC group was significantly more integrated than that in the Gel group. The data are expressed as mean ± SD (∗, p < .05). Abbreviations: d, day; w, week(s).

Histological Observation

Representative samples in each group are shown in Figure 3. At 1 day in the Defect and Gel groups, there were blood clots in the defect. In the MSC group, the cartilage defect was covered with periosteum and filled with collagen gel, in which a great number of nucleated cells positive for DiI were observed (Fig. 3A).

Figure Figure 3..

Histological analyses. (A): Histological analyses at 1 day. Sagittal sections of the cartilage defect stained with Toluidine blue in Defect group (a), Gel group (b), and MSC group (c) at 1 day. Serial section of MSC group under epifluorescent microscopy (d). Higher magnifications of the framed area in (c) stained with Toluidine blue (e) and under epifluorescent microscopy (f). The nuclei were counterstained by 4,6-diamidino-2-phenylindole (DAPI) in (f). Distal side was located at right side. In the MSC group, a great number of transplanted cells were observed (c–f). Bars = 1 mm (a–d); 50 μm (e, f). (B): Histological analyses at 4 weeks. Sagittal sections of the defects stained with Toluidine blue in Defect group (a) and Gel group (b) at 4 weeks. (c, d): Higher magnification of the framed area in (b). Section in MSC group stained with Toluidine blue (e), under fluorescence (f), higher magnification of the framed area in (e) stained with Toluidine blue (g, i, k, m), and under epifluorescent microscopy (h, j, l, n). The nuclei were counterstained by DAPI in (h, j, l, n). The distal side is located at the right side. In the MSC group, transplanted MSCs differentiated into chondrocyte-like cells and produced cartilage matrix, although the cell density decreased (e, f, i, j). The periosteum became thinner, and chondroprogenitors in the periosteum produced a small amount of cartilage matrix (g, h). In the deeper zone, transplanted MSCs progressed to the hypertrophic chondrocyte-like cells (k, l). In the deep zone, some transplanted cells also differentiated into bone cells (m, n). Bars = 1 mm (a, b, e, f); 50 μm (c, d, i, j, m, n); 25 μm (g, h, k, l). (C): Histological analyses at 12 weeks. Sagittal sections of the defects stained with Toluidine blue in the Defect group (a), Gel group (b), and MSC group (c) at 12 weeks. Serial section from the MSC group under epifluorescent microscopy (d). Higher magnifications of the framed area in (c) stained with Toluidine blue (e) and under epifluorescent microscopy (f). Higher magnifications of the framed area in (c) stained for type I collagen (g) and type II collagen (h). The nuclei were counterstained by DAPI in (f). Distal side was located at right side. In the MSC group, the border between regenerated cartilage-like tissue and subchondral bone moved upward. A number of DiI-positive cells still remained at the regenerated tissue, although DiI-positive cells disappeared at subchondral bone (c–f). Regenerated tissue was negative for type I collagen (g) and positive for type II collagen (h). Bars = 1 mm (a–d); 50 μm (e, f); 250 μm (g, h). (D): Histological analyses at 24 weeks. Sagittal sections of the defects stained with Toluidine blue in Defect group (a), Gel group (b), and MSC group (c) at 24 weeks. Serial section of MSC group under epifluorescent microscopy (d). Higher magnification of the framed area in (c) with Toluidine blue (e). The arrow in (e) indicates the boundary between the original cartilage (left side) and the regenerate cartilage (right side). Higher magnifications of the framed area in (c) stained with Toluidine blue (f) and under epifluorescent microscopy (g). The nuclei were counterstained by DAPI in (g). Higher magnifications of the framed area in (c) stained for type X collagen (h). Hypertrophic cartilage of the growth plate in immature rabbit stained for type X collagen was shown as a positive control (i). The distal side is located at right side. In the MSC group, the border between the regenerated cartilage-like tissue and the bone moved upwards (c). The integration between native cartilage and regenerated cartilage-like tissue was improved (e). DiI-positive cells still remained at the cartilage zone (d, g). Regenerated tissue was negative for type X collagen (h), suggesting regenerated tissue was stable at least for 24 weeks. Bars = 1 mm (a–d); 100 μm (e, i); 50 μm (f–h). Abbreviation: DiI, 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate.

At 4 weeks in the Defect and Gel groups, cartilage matrix formation appeared poor, although chondrocyte-like cells were partially detected at the peripheral area of the defect (Fig. 3Ba–3Bd). In the MSC group, although the number of DiI-positive cells decreased, they differentiated into chondrocyte-like cells, and the defect was filled with cartilage matrix. The remnant of periosteum became thinner, and the amount of the cartilage matrix at the remnant of periosteum was less than that at the center of regenerated cartilage (Fig. 3Be, 3Bf). Cells at the remnant of periosteum were DiI-negative (Fig. 3Bg, 3Bh); however, a number of chondrocyte-like cells adjacent to the remnant of periosteum were DiI-positive (Fig. 3Bi, 3Bj). DiI-positive hypertrophic chondrocyte-like cells were observed at the deep area of the cartilage zone (Fig. 3Bk, 3Bl). The deep area of the defect was partially replaced with newly formed trabecular bone, and some cells composing bone were also DiI-positive; on the other hand, cells in the medullary cavity were DiI-negative (Fig. 3Bm, 3Bn).

At 12 weeks, in the Defect and Gel groups, the defect had still healed poorly (Fig. 3Ca, 3Cb). In the MSC group, the border between regenerated cartilage-like tissue and subchondral bone moved upward (Fig. 3Cc). A number of DiI-positive cells still remained at the regenerated tissue (Fig. 3Cd–3Cf), which was negative for type I collagen (Fig. 3Cg) and positive for type II collagen (Fig. 3Ch). DiI-positive cells disappeared at subchondral bone (Fig. 3Cd).

At 24 weeks in the Defect and Gel groups, the cartilage was not regenerated (Fig. 3Da, 3Db). In the MSC group, the subchondral bone moved upward (Fig. 3Dc), and the integration between native cartilage and regenerated cartilage-like tissue seemed to be improved (Fig. 3De). However, the thickness of the regenerated cartilage-like tissue became thinner over the majority of the repair zone. Regenerated tissue was negative for type X collagen, suggesting that the regenerated tissue was stable and did not become hypertrophic at least for 24 weeks (Fig. 3Dh). DiI-positive cells remained at the cartilage zone, although they decreased in number (Fig. 3Dd, 3Dg).

Histological Score

Histological sections of the repaired tissue were analyzed in a blinded manner by two observers who were not informed of the group assignment and were quantified using a histological grading system for cartilage defects described by Wakitani et al. [25]. This system consists of five categories (cell morphology, matrix staining, surface regularity, cartilage thickness, and integration of donor with host) scored on a 0- to 14-point scale, in which 0 stands for complete regeneration and 14 for no regeneration. There was no significant difference in the scoring between two observers. The scores of the MSC group improved continuously through 24 weeks and were always better than those of the Gel group and the Defect group at each point (Fig. 4).

Figure Figure 4..

Histological score for cartilage defect after cell transplantation. Histological findings were quantitated using the scoring system described by Wakitani et al. [25] in which a full score was 14 and a lower score indicated improvement. The scores of the MSC group improved continuously through 24 weeks and were better than those of control groups at each point. The data are expressed as mean ± SD (∗, p < .05). Abbreviation: w, week(s).

Transmission Electron Microscopic Examination

The centers of the cartilage defects in the MSC group were examined with TEM. At 1 day, implanted MSCs were round with a euchromatic and notched nucleus. A large number of cell processes were present at the cell surface. The cell contained well-developed organelles, including mitochondria, endoplasmic reticulum, Golgi apparatus, and large quantities of free ribosomes (Fig. 5A, 5B). At 4 weeks, spindle-shaped or polygonal cells with an ovoid nucleus and well-developed endoplasmic reticulum appeared to produce large amounts of extracellular matrix. Their morphology resembled that of normal chondrocytes at uninjured articular cartilage morphologically (Fig. 5C, 5D, 5F). At 8 weeks, the cells were surrounded with higher levels of collagen matrix (Fig. 5G, 5H). Throughout 1 day to 8 weeks, the cells appeared not to divide because no features of cell mitosis were observed. On the other hand, a number of apoptotic bodies were observed especially at 4 weeks, and these cells contained characteristic condensations of the chromatin (Fig. 5E).

Figure Figure 5..

Transmission electron microscopy (TEM) findings in the MSC group. TEM findings in the MSC group at 1 day (A, B), 4 weeks (C–F), and 8 weeks (G, H). Lower (A, C, G) and higher (B, D, H) magnification of transplanted cells at each period. An apoptotic body at the injured site, which had condensation of the chromatin (E). A normal chondrocyte at uninjured cartilage (F). Bars = 5 μm (A–H). Abbreviations: d, day; w, week(s).

Discussion

Articular cartilage defect has very limited intrinsic healing capacity. Generally, partial thickness defects that do not penetrate the subchondral bone never repair spontaneously [26]. Repair of full thickness defects that penetrate the subchondral bone depends on the defect size and location [27]. Usually, small defects can repair spontaneously with hyaline cartilage, whereas larger defects only repair with fibrous tissue. Although various methods have been attempted to regenerate articular cartilage defects, there are some problems with each of the previously described methods. By abrasion [28] or microfracture [29], defects only repair with fibrous tissue. Mosaicplasty [30] or autologous chondrocyte implantation [31] are often performed to regenerate cartilage defects with hyaline cartilage; however, there are often issues with obtaining enough chondrocytes for repair, thereby limiting such applications for the repair of large defects.

Here, we have demonstrated that synovium-derived mesenchymal stems cells are well suited to repair full thickness articular cartilage defects in rabbits with hyaline cartilage within 12–24 weeks when applied in collagen gel and covered with periosteum. Within this time span, the repair tissue integrates well into the host cartilage and does not further differentiate into hypertrophic cartilage.

A considerable amount of retrospective data is available that describes putative MSCs, however, there is still very little knowledge available that documents the properties of a MSC, especially in its native environment. Although the precise identity of MSCs remains a challenge [32], we herein define an MSC as being derived from mesenchymal tissue and by its functional capacity both to self-renew and to generate a number of differentiated progeny [24]. Since the earliest work by Friedenstein [23], the standard assay used to identify MSCs is the colony-forming unit-fibroblast assay, which identifies adherent, spindle-shaped cells that proliferate to form colonies. By this method, we isolated colony-forming cells from rabbit synovium. We also demonstrated the in vitro multipotentiality of the cells. All of the cells studied in this article are called MSCs; however, their biological properties varied considerably.

Transplanting autologous chondrocytes cultured in collagen gel had been reported for the treatment of full-thickness defects of cartilage [33], and the chondrocyte density in collagen gel applied to the clinical study was 106 cells/ml. Our previous in vitro study demonstrated that MSCs/gel composites with 5 × 107 cells/ml produced a greater amount of cartilage matrix than composites containing 106 cells/ml [19]. Here, MSCs/gel composites with 5 × 107 cells/ml were transplanted successfully, whereas transplantation of MSCs/gel composites containing only 106 cells/ml resulted in failure (data not shown). Therefore, the cell density required for graft success is likely to be different for mature chondrocytes and MSCs.

Before this report, the fate of transplanted MSCs during cartilage regeneration of full-thickness articular cartilage defects was still unknown. Although there are several cartilage repair studies with MSCs, transplanted MSCs did not fully differentiate into chondrocytes, or transplanted MSCs were not tracked completely in most reports [25, 34]. The use of DiI, a membrane-bound fluorescent dye, is one of the ways to track the fate of implanted cells [35]. Previous studies showed that DiI typically exhibits very low cell toxicity and does not compromise cell viability and differentiation potential [36, 37]. DiI also retains its fluorescence for a long time, at least 26 weeks after the surgery [38]. DiI is weakly fluorescent in aqueous solutions but is highly fluorescent and photostable when incorporated into lipid membrane. Transfer of these probes between intact membranes is usually negligible (Molecular Probes Product Data Sheet). Because of this, even if the dye were to leach out of a dead cell, it would not emit significant fluorescence in an aqueous environment such as the cartilage matrix.

When DiI-labeled MSCs proliferated rapidly, their fluorescence decreased along with the culture period. On the other hand, DiI-labeled MSCs maintained their fluorescence during in vitro chondrogenesis (data not shown). According to our previous reports, MSCs did not divide during in vitro chondrogenesis, whereas viable MSCs decreased by apoptosis [39, 40]. In this study, during in vivo chondrogenesis, the cells also maintained their fluorescence for at least 24 weeks, and we could not observe features of cell division, whereas we could observe apoptosis of transplanted chondrocyte-like cells. DiI-labeled cells decreased mainly by apoptosis, in contrast, they maintained their fluorescence for long time possibly because MSCs did not divide during both in vitro and in vivo chondrogenesis.

Despite the use of allogenic cells in this study, macroscopically, we could not observe features of immune reactions such as hydrarthrosis or synovial inflammation. Immune tolerance against MSCs is still controversial [41, 42]; however, when comparing allografts with autografts, there was no difference in either synovial inflammation or cartilage regeneration (data not shown).

For clinical application, species differences could be an issue. We used a rabbit model in which osteochondral defects did not repair spontaneously [43]; however, rabbits are known for their animal variability and higher spontaneous repair ability of osteochondral defects than other species. To solve this issue, further experimental studies in large animals will be needed.

According to our observations, we summarized the sequence of cellular events during in vivo cartilage regeneration by synovium-derived MSCs in Figure 6. First, the cartilage defect was filled with a great number of transplanted MSCs and a few donor-derived MSCs at the peripheral site of the defect, which were covered with periosteum (Fig. 6A). In the next phase, although the cell density decreased mainly by apoptosis, transplanted and donor-derived MSCs differentiated into chondrocyte-like cells and produced cartilage matrix. The periosteum became thinner, and chondroprogenitors in the periosteum produced a small amount of cartilage matrix. In the deeper zone, transplanted and donor-derived MSCs progressed to the hypertrophic chondrocyte-like cells. In the deep zone, some transplanted cells also differentiated into bone cells, which were replaced with host cells thereafter (Fig. 6B, 6B′). Finally, the border between bone and cartilage moved up, and integrations between native cartilage and regenerated cartilage-like tissue were improved. Chondrocyte-like cells derived from the transplanted MSCs decreased gradually in number but still remained at 24 weeks after implantation (Fig. 6C).

Figure Figure 6..

Scheme for the sequence of events during in vivo cartilage regeneration by synovium-derived MSCs. Schematic drawings showing the sequential events during cartilage regeneration in the MSC group. The amount of the cartilage matrix is represented as violet-colored.

One concern from the standpoint of treatment for cartilage injury is that the thickness of the cartilage became thinner at 24 weeks, and the tidemark moved upwards over the majority of the repair zone. This is a general phenomenon seen in cartilage repair of osteochondral defects [13, 25]. Violation of the tidemark and the subchondral plate will involve a problem with regard to restoration of the original tidemark. This phenomenon is also seen in humans and has been described in microfracture [44]. In this rabbit study, this upward movement of the bone front was shown at 24 weeks and may result in long-term failures of the repair procedure.

In past studies of the rabbit cartilage defect model using MSCs, regenerated tissue seemed to fail after 24 weeks. Although there is a lack of molecular events, it is clear which events should not occur in regenerate cartilage, such as dedifferentiation to fibrocartilage, marked by the expression of type I collagen, or maturation to hypertrophic cartilage, as indicated by the expression of type X collagen. In the repair cartilage induced by synovium-derived MSCs, we could not detect type I or type X collagen, indicating a stable hyaline phenotype of the repair cartilage. The surrounding circumstance may be suitable for synovium-derived MSCs because of stimulation from synovial fluid or trophic factors secreted by host-derived synovial tissues.

Conclusion

Synovium-derived MSCs altered over a time course according to local microenvironments. The cells differentiated into chondrocyte-like cells at the place where the original cartilage located. This system will advance the clinical application of MSC-based therapeutic strategies for the repair of cartilage injury. In addition, this will elucidate the mechanisms that govern multilineage differentiation of MSCs.

Disclosures

The authors indicate no potential conflicts of interest.

Acknowledgements

We thank Kenichi Shinomiya, M.D., Ph.D., for continuous support; Izumi Nakagawa for excellent technical assistance; Miyoko Ojima for expert help with histology; and Jeffrey L. Spees for proofreading. This study was supported by grants from the Center of Excellence Program for Frontier Research on Molecular Destruction and Reconstruction of Tooth and Bone in Tokyo Medical and Dental University to T.M. and the Japan Society for the Promotion of Science (16591478) to I.S. Recombinant human bone morphogenetic protein-2 was provided by Astellas Pharma Inc.

Ancillary