First published online in STEM CELLSExpress January 8, 2009.
Author contributions: M.H.: conception and design, data analysis, collection of data, and manuscript writing; I.S.: conception and design, financial support, manuscript writing, final approval of manuscript; T. Muneta: conception and design, financial support, and administrative support; S.I.: electron microscopy; K.M.: microarray and real-time PCR; H.S.: microarray and real-time PCR; T. Murakami: provision of study material and manuscript writing; E.K.: conception and design, financial support, administrative support, provision of study material, manuscript writing.
Osteoarthritis in the knees, which can be caused by meniscal defect, constitutes an increasingly common medical problem. Repair for massive meniscal defect remains a challenge owing to a lack of cell kinetics for the menisci precursors in knee joint. The synovium plays pivotal roles during the natural course of meniscal healing and contains mesenchymal stem cells (MSCs) with high chondrogenic potential. Here, we investigated whether intra-articular injected synovium-MSCs enhanced meniscal regeneration in rat massive meniscal defect. To track the injected cells, we developed transgenic rats expressing dual luciferase (Luc) and LacZ. The cells derived from synovium of the rats demonstrated colony-forming ability and multipotentiality, both characteristics of MSCs. Hierarchical clustering analysis revealed that gene expression of meniscal cells was closer to that of synovium-MSCs than to that of bone marrow-MSCs. Two to 8 weeks after five million Luc/LacZ+ synovium-MSCs were injected into massive meniscectomized knee of wild-type rat, macroscopically, the menisci regenerated much better than it did in the control group. After 12 weeks, the regenerated menisci were LacZ positive, produced type 2 collagen, and showed meniscal features by transmission electron microscopy. In in-vivo luminescence analysis, photons increased in the meniscus-resected knee over a 3-day period, then decreased without detection in all other organs. LacZ gene derived from MSCs could not be detected in other organs except in synovium by real-time PCR. Synovium-MSCs injected into the massive meniscectomized knee adhered to the lesion, differentiated into meniscal cells directly, and promoted meniscal regeneration without mobilization to distant organs. STEM CELLS 2009;27:878–887
The meniscus is a wedge-shaped semilunar fibrocartilage that lies between the weight bearing joint surfaces of the femur and the tibia. For symptomatic meniscus injury, a meniscectomy is often performed. This, however, often leads to osteoarthritis . Meniscal suture to preserve its function is limited for its indication, and the result is not always satisfactory due to poor healing of the meniscus. Despite other therapeutic attempts , problems related to its effectiveness and invasion persist. A novel strategy for meniscus injury remains necessary.
Mesenchymal stem cells (MSCs) are postulated to participate in tissue homoeostasis, remodeling, and repair by ensuring the replacement of mature cells lost to physiological turnover, senescence, injury, or disease. Stem cell populations are found in most adult tissues, and in general, their differentiation potential may reflect the local cell population. Developmentally, intra-articular tissues are differentiated from common progenitors, referred to as common interzone cells . Synovium-MSCs have high chondrogenic potential [4, 5], clinically increase in number in synovial fluid after intra-articular tissue injury to contribute to its repair in part  and expand in the presence of pure synovial fluid in tissue cultures of the synovium . Synovial tissue may serve as a reservoir of stem cells that mobilize following injury and migrate to the wound site where, in cooperation with local cells, they participate in the repair response.
Thus, during the natural course of meniscal repair, synovium-MSCs are a potential cell source. Here, we investigated whether intra-articular injected synovium-MSCs enhanced meniscal regeneration in rat massive meniscal defect. Dual colored transgenic (Tg) rats expressing luciferase and LacZ (Luc/LacZ) were created for this study so that the fate of transplanted cells could be traced dynamically and precisely.
MATERIALS AND METHODS
Establishment of Dual Colored Transgenic Rat
Dual colored Tg rats expressing luciferase and Lac-Z were created by cross-breeding ROSA/luciferase Tg Lewis rats  with ROSA/LacZ Lewis rats . The expression of luciferase was detected by an in vivo bioimaging system, and the expression of LacZ was detected by X-gal staining (detailed later). The F1 hybrids between ROSA/luciferase Tg and ROSA/LacZ Lewis rat neonate were imaged after intraperitoneal injection ofD-luciferin (30 mg/kg per body weight) (potassium salt; Biosynth, Postfach, Switzerland, http://www.biosynth.com), and then they were stained with X-gal. In the same manner, luciferase and LacZ expressions were examined in various tissues of these rats. Approximately one-fourth of these F1 hybrids expressed luciferase and LacZ in the whole body. We used these “dual colored” F1 hybrids expressing both luciferase and LacZ (Luc/LacZ) for the donor of MSCs.
All experiments were conducted in accordance with the institutional guidelines for the care and use of experimental animals of Tokyo Medical and Dental University and Jichi Medical University. The synovial membranes of bilateral knee joints were excised, minced, and digested for 3 hours at 37°C with type V collagenase (0.2%; Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com), and passed through a 40-μm filter (Becton Dickinson, Franklin Lakes, NJ, http://www.bd.com). Bone marrow was extruded by inserting a 22-gauge needle into the shaft of the femur and tibia bone and flushed out. Synovium and bone marrow cells were cultured in a complete medium (αMEM; Invitrogen, Carlsbad, CA, http://www.invitrogen.com; 20% FBS; Invitrogen; 100 units per milliliter penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine; Invitrogen) for 14 days. Then the cells were replated at 100 cells/cm2, cultured for 14 days, and frozen at −80°C as passage 1. The stocked cells were rapidly thawed in a water bath at 37°C, plated in a 150 cm2 dish, and harvested after 5 days. Then the cells were replated at 100 cells/cm2, cultured for 14 days, and collected for further analyses . For colony-forming assay, 100 cells were plated in 60 cm2 dishes and cultured for 14 days. The dishes were stained with X-gal, and the same dishes were then stained with 0.5% Crystal Violet.
In Vitro Differentiation Assay
For adipogenesis, the cells were cultured in the adipogenic medium that consisted of a complete medium supplemented with 0.5 μM dexamethasone, 0.5 mM isobutylmethylxanthine, and 50 μM indomethacin. After 4 days, the adipogenic cultures were stained with 0.3% Oil Red-O solution or X-gal solution .
For osteogenesis, the cells were cultured in the calcification medium in the presence of 100 nM dexamethasone, 10 mM β-glycerophosphate, and 50 μM ascorbic acid. After an additional 6 weeks, the dishes were stained with 0.5% Alizarin Red solution or X-gal solution.
For in vitro chondrogenesis, 8 × 105 cells were placed in a 15 ml polypropylene tube (BD Falcon, Bedford, MA, http://www.bdbiosciences.com) and pelleted by centrifugation at 450g for 10 minutes. The pellets were cultured for 21 days in chondrogenic media, which contained 500 ng/ml BMP-2 (Astellas Pharma Inc., Tokyo, Japan, http://www.astellas.com), in addition to high-glucose Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10 ng/ml transforming growth factor-β3 (TGF-β3) (R&D Systems Inc., Minneapolis, MN, http://www.rndsystems.com), 10−7 M dexamethasone, 50 μg/ml ascorbate-2-phosphate, 40 μg/ml proline, 100 μg/ml pyruvate, and 50 mg/ml ITS+TMPremix (Becton Dickinson). For histological analysis, the pellets were embedded in paraffin, cut into 5-μm sections, and stained with 1% Toluidine Blue .
Synovium-MSCs at passage 3 were harvested 14 days after plating. One million cells were suspended in 500 μl phosphate buffered saline (PBS) containing 20 ng/ml fluorescein isothiocyanate (FITC) or phycoerythrin (PE)-coupled antibodies against CD11b, CD45, CD90 (Becton Dickinson), CD34 (Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), and CD29 (BioLegend, San Diego, CA, http://www.biolegend.com) As an isotype control, FITC- or PE-coupled nonspecific mouse IgG (Becton Dickinson) was substituted for the primary antibody. After incubation for 30 minutes at 4°C, the cells were washed with PBS and resuspended in 1 ml PBS for analysis. Cell fluorescence was evaluated by flow cytometry in a FACSCalibur instrument (Becton Dickinson); data were analyzed by using CellQuest software (Becton Dickinson).
For rat meniscal cells, menisci were minced, digested for 3 hours at 37°C with type II collagenase (0.2%; Sigma), and passed through a 40 μm filter (Becton Dickinson). Nucleated cells were plated at 100 cells/cm2 and cultured in a complete medium. Total RNA was isolated from passage 1 colony-formed cells derived from the synovium, bone marrow , and meniscus with the RNeasy Total RNA Mini Kit (Qiagen, Valencia, CA, http://www1.qiagen.com).
A comprehensive microarray analysis was performed using 3 μg of total RNA from each sample and GeneChip Rat 230 2.0 probe arrays (Affymetrix, Santa Clara, CA, http://www.affymetrix.com) . Data analysis was performed with GeneSpring software version 7.2 (Agilent Technologies, Palo Alto, CA, http://www.agilent.com). To normalize the variations in staining intensity among chips, the “Signal” values for all genes on a given chip were divided by the median value for the expression of all genes on the chip. To eliminate genes containing only a background signal, genes were selected only if the raw values of the “Signal” were more than 30, and expression of the gene was judged to be “Present” by the GeneChip Operating Software version 1.4 (Affymetrix). After elimination, expression data of a total of 14,882 probe sets were employed for further analysis. A hierarchical-clustering analysis was performed using a minimum distance value of 0.001, a separation ratio of 0.5, and the standard definition of the correlation distance. A dendrogram was obtained from a hierarchically clustering analysis using average linkage and distance metric equal to one minus the Pearson correlation applied to the microarray data .
Meniscectomy and MSCs Injection
Wild-type male Lewis rats at 12 weeks of age (Charles River, Yokohama, Japan, http://www.crj.co.jp) were used (n = 27). Under anesthesia, a straight incision was made on the anterior side of bilateral knee, the anteromedial side of the joint capsule was cut, and the anterior horn of the medial meniscus was dislocated anteriorly with a forceps. The meniscus was then cut vertically at the level of medial collateral ligament, and the anterior half of medial meniscus was excised. The dislocated meniscus was removed and the wound was closed in layers. Immediately after the skin incision was closed, a 27-gauge needle was inserted at the center of the triangle formed by the medial side of the patellar ligament, the medial femoral condyle, and the medial tibial condyle, toward the intercondylar space of the femur. Then 5 × 106 Luc/LacZ+ synovium-MSCs (n = 14) or bone marrow-MSCs (n = 9) in 50 μl PBS were injected into the right knee joint. For the control, the same volume of PBS was injected into the left knee. The rats were allowed to walk freely in the cage.
For control of in vivo Imaging analysis, 5 × 106 Luc/LacZ+ synovium-MSCs were injected into the normal right knee of the wild-type Lewis rats (n = 4).
Histology and Detection of LacZ Expression
The whole medial meniscus was collected at 2, 4, 8, and 12 weeks after MSCs injection (n = 3 each time point). The samples were fixed with a fixative solution (0.2% glutaraldehyde, 2 mM MgCl2, and 5 mM EGTA) in PBS for 10-30 minutes at room temperature and washed three times in a washing solution (2 mM MgCl2, 0.01% sodium deoxycholate, and 0.02% Nonidet P40) in PBS. Then they were treated with an X-gal staining solution (1 mg/ml of 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside, 2 mM MgCl2, and 5 mM potassium hexacyanoferrate [III], 5 mM potassium hexacyanoferrate [II] trihydrate) at 37°C for 3 hours. They were subsequently fixed again in 4% paraformaldehyde and decalcified with 0.5 M EDTA (pH 7.5) for 3 days at 4°C, followed by a gradient replacement with 20% sucrose for 24 hours at 4°C, and then evaluated by Toluidine Blue or Eosin staining of paraffin sections.
Sections were pretreated with 0.4 mg/ml proteinase K (DAKO, Carpinteria, CA, http://www.dakousa.com) in Tris-HCl for 15 minutes at room temperature for optimal antigen retrieval. Residual enzymatic activity was removed by washes in PBS, and nonspecific staining was blocked with PBS containing 10% normal horse serum for 20 minutes at room temperature. A primary anti-rat monoclonal antibody against human type II collagen (1 : 200 dilution with PBS containing 1% BSA; Daiichi Fine Chemical, Toyama, Japan, http://www.daiichi-fcj.co.jp) was applied to the section which was incubated at room temperature for 1 hour and rinsed again with PBS. Immunostaining was detected by Vectastain ABC reagent (Vector Laboratories, Burlingame, CA; http://www.vectorlabs.com), followed by diaminobenzidine staining.
In Vivo Bioluminescent Imaging
A noninvasive bioimaging system IVIS (Xenogen, Alameda, CA, http://www.caliperls.com) was used for analysis using IGOR (WaveMetrics, Lake Oswego, OR, http://www.wavemetrics.com) and IVIS Living Image (Xenoxgen) software packages . To detect photons from Luc+ cells, undifferentiated MSCs or chondrocyte pellets were suspended in PBS and imaged immediately after the addition of 0.15 mg D-luciferin (potassium salt; Biosynth). Also, for transplanted cell tracking in vivo, D-luciferin was injected into the penile vein of anesthetized rats (30 mg/kg per body weight) under anesthesia with isoflurane. The signal intensity was quantified as photon flux in units of photons per seconds cm2 per steradian in the region of interest.
Transmission Electron Microscopy
The regenerated tissues at 12 weeks in the synovium-MSCs treated group and control groups were selected and fixed with 2.5% glutaraldehyde in 0.1 M PBS for 5 hours, washed overnight at 4°C in the same buffer, postfixed with 1% OsO4 buffered with 0.1 M PBS for 2 hours, dehydrated in a graded series of ethanol, and embedded in Epon 812. Ultrathin sections at 90 nm were collected on copper grids, double-stained with uranyl acetate and lead citrate, and then examined with a transmission electron microscope (H-7100, Hitachi, Hitachinaka, Japan, http://www.hitachi.co.jp) .
Quantitative Real-Time PCR
Total RNAs were prepared from brain, lung, liver, spleen, kidney, and knee synovium at 3 days after the synovium-MSCs injected rat by RNAqueous Kit (Ambion, Austin, TX;http://www.ambion.com) according to the manufacturer's instructions. For the positive control, total RNAs from various organs of Luc/LacZ Tg rat and expanded MSCs including Luc/LacZ positive cell rates ranged from 0.001 to 100% were used. Also total RNAs from various organs of wild-type rat were prepared as a negative control. They were subjected to real-time PCR to measure the level of LacZ.
The primer sets for LacZ, (sense, 5′-GGTGCAGAGAGACAGGAACCAC-3′; antisense, 5′-CCTTCATACTGCACAGGTCTGCT-3′) and β-actin (sense, 5′-CCGAGCGTGGCTACAGCTT-3′; antisense, 5′-GGCAGTGGCCATCTCTTGC-3′) were synthesized at FASMAC (Kanagawa, Japan, http://www.fasmac.co.jp).
First-strand cDNA was synthesized using the I Script cDNA synthesis kit (Bio-Rad Laboratories, Hercules, CA, http://www.bio-rad.com) with an Oligo (dT) (12-18mers) primer. Real-time quantitative PCR analyses were performed with the ABI Prism 7700 Sequence Detection System (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com) using SYBR Green I PCR reagents (TOYOBO, Osaka, Japan, http://www.toyobo.co.jp/e/) as described previously . To determine the exact copy numbers of the target genes, the quantified concentrations of subcloned PCR fragments of LacZ and β-actin were serially diluted and used as standards in each experiment. Aliquots of cDNA equivalent to 5 ng of total RNA samples were used for each real-time PCR. Data were normalized with β-actin levels in each sample. The copy number is expressed as the number of transcripts per nanogram total RNA.
The Mann-Whitney U test was used to compare two groups at each period. p values less than .05 were considered significant.
Synovium-MSCs Expressing Luc/LacZ Genes
The F1 hybrids between ROSA/luciferase Tg and ROSA/LacZ Lewis rat neonate produced luminescence after intraperitoneal injection of D-luciferin (Fig. 1A), and then they were positive for X-gal staining (Fig. 1B). In the same manner, we next examined luciferase and LacZ expression in various tissues of these rats. Approximately one-fourth of these F1 hybrids expressed luciferase and LacZ in whole body including the synovium (Fig. 1C). We termed these “dual colored” rats as Luc/LacZ Tg rats. MSCs were isolated from the synovium of Luc/LacZ Tg rats. In vitro imaging of luciferase activity showed that as few as one thousand MSCs were detected over the background in the linear dose-dependent output of luminescence (Fig. 1D, 1E). Synovium-MSCs from Luc/LacZ Tg rats formed LacZ+ single cell-derived colonies consisting of spindle cells (Fig. 1F). They could differentiate into adipocytes (Fig. 1G) and calcified with LacZ expression in vitro (Fig. 1H). The cells could also form cartilage with expressions of both LacZ and luciferase (Fig. 1I). The cells derived from synovium of Luc/LacZ Tg rat demonstrated characteristics of MSCs, and the dual markers were maintained after the differentiation. Flow cytometric analysis demonstrated that the majority of synovium-MSCs expressed CD29 and CD90, and were negative for CD11b, CD34, and CD45 (Fig. 1J).
Hierarchical clustering analysis revealed that the gene expression of meniscal cells was closer to that of synovium-MSCs than that of bone marrow-MSCs (Fig. 2). This indicates that synovium-MSCs may retain a more advantageous character as a MSC source for meniscal regeneration than bone marrow-MSCs.
Meniscal Regeneration After Intra-articular Injection of Synovium-MSCs
To obtain in vivo evidence to support the synovium-MSCs potential, we performed massive meniscectomy in both sides of the knee of wild rats (Fig. 3A), and Luc/LacZ+ synovium-MSCs suspension (5 × 106 in 50 μl PBS) were injected into the right knee joint immediately after the incision was closed. For the control, the same volume of PBS was injected into the left knee. At 2 weeks, dark blue areas for LacZ were observed around the meniscal defect (Fig. 3B, yellow arrow) and sutured capsule (black arrow), indicating that injected synovium-MSCs intensively adhered to injured sites. Dark blue areas for LacZ were still observed around the meniscal defect and sutured capsule, but not observed in intact synovium, cartilage surface, or cruciate ligaments even at 12 weeks (data not shown). At 4 weeks, the anterior part of the meniscal defect of MSC injection side exhibited better regeneration than the control side (Fig. 3C).
All menisci for the experiment of synovium-MSCs are shown in Figure 3D. At 2 weeks, the regenerated part of the menisci appeared blue after X-gal staining in the MSC injection group, indicating that injected synovium-MSCs contributed to the repair. Square measures of the meniscus in synovium-MSCs injection groups were significantly larger than those in the control groups at 2, 4, and 8 weeks (Fig. 3E). We also evaluated cartilage and observed fibrillation, ones of the degenerative changes on the surface of the medial femoral condyle and medial tibial plateau in the control group at 12 weeks (Fig. 3F).
When we injected the same amount of bone marrow-MSCs, the meniscal defect was more rapidly regenerated than that on the control side at 4 weeks (Fig. 3D). Macroscopically, there were no remarkable differences between the synovium-MSC injection group and the bone marrow-MSC injection group.
Histologically, the contour of the regenerated menisci sharpened and the ultimate forms were closer to the normal meniscal shape (Fig. 4A). Expression of type II collagen increased in a time-dependent manner. LacZ+ MSCs still existed at 12 weeks. In the control group at 12 weeks, regenerated tissue was occupied with less metachromasia, and type II collagen expression was hardly detected. In the bone marrow-MSC injection group, we observed similar features as seen in the synovium-MSC injection group (Fig. 4B).
Electron microscopic analysis of the regenerated meniscus 12 weeks after synovium-MSCs injection demonstrated that round cells with short processes were surrounded by a pericellular matrix, suggesting that meniscal cells were morphologically equivalent to those of the normal meniscus. In contrast, the cell feature in control groups remained to be fibroblastic (Fig. 4C).
Injected MSCs Stay in Knee Joint
The distribution of topically injected synovium-MSCs was evaluated using luciferase-based in vivo imaging. When MSCs were injected into the normal knee (n = 4), MSC-derived photons were detected around the right knee, to then decrease within 14 days. When injected into menisectomized knee (n = 7), the photons increased in 3 days, then moderately decreased, but could be still observed for 28 days. Substantial luminescence light could not be detected in any other organs of either group (Fig. 5A). Sequential quantification demonstrated that luciferase activities were significantly higher in the meniscectomy group than those in the control group at each time point up to 21 days (Fig. 5B).
To further evaluate whether injected MSCs could migrate to distant organs or not, quantitative real-time PCR was performed. Total RNAs were isolated from the brain, lung, liver, spleen, kidney, and knee synovium of meniscectomized rat at 3 days after the synovium-MSC injection, and were subjected to real-time PCR to measure the level of LacZ expression. Injected MSC-derived LacZ gene could be detected only at the injected knee synovium and was not detected in any other organs (Fig. 6). These data confirm the in vivo imaging results and indicate that knee-injected MSCs stayed only in the knee joint.
A number of reports have previously described the injection of bone marrow-MSCs into the joint for meniscus injury; however, the kinetics and role of injected MSCs remain unknown. In a goat study, a previously removed medial meniscus regenerated 6 months after MSC injection . However, Caplan et al. suggested in their review article that there were too few prelabeled cells to account for the massive regeneration of the meniscus and inferred that the MSCs trophically enhanced the regeneration of the meniscus . In two other articles, cartilage matrix was present around injected bone marrow MSCs in only a small portion of the injured meniscus, but the roles of injected MSCs were not fully demonstrated [18, 19] To refine our analysis, we created Tg rats expressing dual Luc/LacZ genes.
There are three main factors involved in the interaction between injected synovium-MSCs and meniscal defect: (a) an increase and decrease in the number of the cells, (b) an adherence to the meniscal defect, and (c) differentiation and maintenance.
In in vivo imaging analysis, synovium-MSCs injected into the menisectomized knee transiently increased in number unlike the situation in which the same population of the cells was injected into the intact knee. This indicates that meniscus injury or incision of the articular capsule can produce some cytokines/chemokines to proliferate synovium-MSCs. In a rabbit model, meniscal lesions expressed TGF-β, interleukin-1α, and platelet derived growth factor (PDGF) . Human synovium-MSCs have PDGF receptor α and β, and neutralizing PDGF decreases the proliferation of synovium-MSCs in vitro . PDGF may affect the number of transplanted synovium-MSCs.
We could detect MSC-derived photons in the knee joint up to 28 days after the injection but we could not detect it at a longer time point. We have two speculations about this result. One is that a luciferase-based in vivo imaging system cannot detect a small number of Luc+ cells, as shown in Figure 1D. More than 1,000 cells are needed to detect the light emission. The other is due to blood circulation. The photons are produced only when luciferase is exposed to luciferin substrate and we injected luciferin intravenously. If the Luc+ cells had existed in the hypovascular area, they could not have been detected. However, in vivo imaging analysis has a great advantage in tracking cells in vivo because we did not have to sacrifice animals, and we could observe the same recipient throughout the observation periods.
Those synovium-MSCs which seem unnecessary for meniscal repair decreased in number and finally fell below measurable limits based on an in vivo imaging system. Possibly, synovium-MSCs participate in intra-articular tissue homoeostasis and repair as do MSCs in mesenchymal tissues throughout the body. This phenomenon differs markedly from ES cells, which form teratomas in the mouse knee joint and subsequently destroy the joint . Induced pluripotent stem cells may hold attraction for future applications, but teratoma formation cannot be overlooked .
In vivo imaging analysis suggests that the injected cells did not mobilize out of the injected joint. To confirm this result, we performed real-timePCR to detect Lac Z transcripts. We determined that LacZ transcripts were not detected in brain, lung, liver, spleen, or kidney 3 days after the injection, although the minimum detection limit was 50 copies per nanogram RNA which corresponded to 1% LacZ+ cells. Detection of LacZ+ cells in all sections thorough whole tissues will provide data from another point of view; however, it will be very arduous work.
Injected synovium-MSCs intensively adhered to injured sites of meniscus and synovium. We previously reported that injected synovium-MSCs efficiently adhered to the defect of articular cartilage  and anterior cruciate ligament . The mechanisms that guide the homing of injected cells are not well-understood, but stromal cell-derived factor-1 and monocyte chemotactic protein-1 are candidates to explain them .
Undifferentiated synovium-MSCs, attached around the meniscal defect, differentiate into meniscal cells. The environment of the meniscal defect is surrounded by femur and tibia cartilage, synovial tissue, the remaining meniscus, and synovial fluid. The space is also influenced by mechanical stress. This environment itself will provide sufficient signals to induce and maintain meniscal differentiation of synovium-MSCs. Similarly, undifferentiated synovium-MSCs implanted onto an articular cartilage defect differentiate into cartilage cells .
In our massive meniscectomized model, bone marrow-MSCs also promoted meniscus regeneration. There were no notable differences of regenerated meniscus in morphology between the synovium-MSC group and the bone marrow-MSC group, although gene profile of synovium-MSCs is closer to that of meniscal cells than that of bone marrow-MSCs. The situation seems to be similar in chondrogenesis as was the case in our previous studies. The gene profile of synovium-MSCs is closer to that of chondrocytes than that of bone marrow-MSCs , transplanted bone marrow-MSCs onto the cartilage defect differentiated into chondrocytes at a similar level to that synovium-MSCs which were transplanted in the same way . In contrast, in vivo chondrogenic assay demonstrated that synovium-MSCs produced more cartilage matrix than bone marrow-MSCs [4, 5]. An in vivo model with more sensitivity may distinguish the difference.
Although reparative potential of synovium- and bone marrow-MSCs is similar, synovium-MSCs have an advantage in that they have a higher proliferation potential. In rats, the colony number per nucleated cell was approximately 1/100 in synovium, whereas it was 4/105 in bone marrow. Rat synovium-MSCs expanded much faster than bone marrow-MSCs . Also, human synovium-MSCs proliferated much faster than bone marrow-MSCs when cultured with autologous human serum .
For adipogenesis, it usually takes 3 weeks for human bone marrow MSCs to differentiate into adipocytes . In this study, rat bone marrow- and synovium-MSCs differentiated into adipocytes in only 4 days, which was similar to our previous report . Although the content of the adipogenic differentiation medium is similar, the duration to induce sufficient accumulation of lipid vesicles is totally different between MSCs in humans and rats. This indicates the species specificity of MSCs.
For calcification, we found that rat MSCs already calcified in 3 weeks, which was similar to our previous report . In this study, we expected that the calcified area would increase after an additional 3 weeks; therefore, we observed calcification for a total of 6 weeks. Seemingly, the calcified area did not increase during the last 3 weeks (data not shown).
We previously created a 1 mm diameter cylindrical defect in the anterior part of medial meniscus in rats to examine the effect of synovium-MSCs injected intra-articularly. Contrary to our expectation, the cylindrical defect was filled spontaneously, and there was no effect on injected synovium-MSCs through 2-12 weeks . To avoid spontaneous healing, we resected the anterior half of medial meniscus for this study. Meniscal size also increased in the control group, and the difference of meniscal size between the two groups disappeared at 12 weeks; however, the synovium-MSC injected groups showed better results from the standpoint of type II collagen expression and electron microscopic features.
For clinical application, interspecies differences have to be considered. The inherent healing capacity of the meniscus has been shown to be lacking in the inner third and is very limited in the middle third of this poorly vascularized tissue in humans  and dogs . We used a rat model, and rat meniscus had a greater spontaneous healing potential. To demonstrate the effectiveness of intra-articular injection of synovium-MSCs for meniscus regeneration, further experimental studies in larger animals are needed.
Native meniscus play an important role in knee stability and shock absorption , and this property is linked to the biphasic microstructure of the meniscus . The extracellular matrix of the meniscus is composed mainly of collagen, with smaller quantities of proteoglycans, matrix glycoproteins, and elastin . In this study, we showed that in the synovium-MSC injection group, the menisci regenerated much better than it did in the control group morphologically, and synovium-MSC injection prevented cartilage degeneration at 12 weeks after the meniscectomy as shown in Figure 3F. However, our data lack the details about biomechanical and biochemical properties of regenerated tissues, and it is still uncertain whether the regenerated menisci function in a normal manner and prevent secondary osteoarthritic change in the long term. Therefore, future studies should include biomechanical and biochemical analysis of the regenerated menisci.
Recently, we reported that human synovium-MSCs increased in synovial fluid after intra-articular ligament injury and that exogenous synovium-MSCs adhered to the injured ligament in a rabbit model [6, 34]. We also demonstrated that autologous synovial fluid enhanced migration of MSCs from synovium of osteoarthritis patients in a tissue culture system seemingly to delay the progression of the cartilage degeneration . We speculate that synovial tissue may serve as a reservoir of stem cells that mobilize following intra-articular tissue injury and migrate to the site to participate in the repair response.
According to our speculation, in the case of meniscus injury, MSCs are mobilized from synovium into synovial fluid, and these cells adhere to the injured meniscus. However, the number of MSCs suspended in the synovial fluid and attached to the site is too low to repair or regenerate the injured meniscus, explaining poor spontaneous healing potential of meniscus. Intra-articular injection of synovium-MSCs can boost natural healing ability for meniscal regeneration.
Intravenous infusion of synovium-MSCs may be another route for administration. Although promising results were reported with i.v. infusion of bone marrow-MSCs in animal disease models , contrary views are also reported showing that a large fraction of intravenously infused bone marrow-MSCs are trapped in the lung . Cell therapy for meniscal injury has an advantage in that intra-articular injection is possible instead of systemic injection.
The meniscal-deficient knee is a common problem faced by orthopedic surgeons. Although repair of meniscal lesions is possible in selected cases, the poor healing capacity of the tissue often dictates meniscectomy, the most common treatment for meniscal injury. Although meniscectomy provides pain relief and return to function, the loss of meniscal tissue results in long-term dysfunction and secondary osteoarthritis. Currently, there are multiple strategies for addressing this objective, including meniscal allografts, biologic scaffolds for tissue-engineered replacement tissue, and biologic stimuli for meniscal tissue regeneration. In this study, we focused on meniscal regeneration by using synovium-MSCs. Our method has a possibility of regenerating meniscus with less invasion in comparison with meniscal transplantation.
Synovium-MSCs injected into the massive meniscectomized knee adhered to the lesion, differentiated into meniscal cells directly, and promoted meniscal regeneration without mobilization to distant organs.
We thank Miyoko Ojima for expert help with histology, Yasuko Sakuma for help with animal experiments, and Noriko Hashimoto for skillful technical assistance with microarray and real-time PCR analyses. This study was supported by grants from “the Japan Society for the Promotion of Science (16591478)” to I.S. and “the Japanese Ministry of Education Global Center of Excellence (GCOE) Program, International Research Center for Molecular Science in Tooth and Bone Diseases” to T.M.
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
The authors indicate no potential conflicts of interest.