SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

Synovial mesenchymal stem cells (MSCs) are an attractive cell source for cartilage regeneration because of their high chondrogenic ability. In this study, we examined the synovium of patients with medial compartment knee osteoarthritis (OA) to determine the proportion of MSCs in relation to cellular compartmentalization, and to identify the culture parameters that could affect the chondrogenic potential of synovial MSCs.

Methods

Human synovium was collected from 4 different harvest sites in the knees of patients with medial compartment OA. Each synovial tissue sample was divided into 2 parts, one for histologic assessment and the other for analysis of the cell size, surface epitopes, and chondrogenic potential of colony-forming cells in vitro.

Results

The numbers of α-smooth muscle actin–positive vessels and CD31+ endothelial cells were higher in the medial outer region than in the other regions of OA synovial tissue. The numbers of these cells correlated with the number of colony-forming cells. In parallel with increasing duration of the preculture period, the size of the cells increased, while the chondrogenic potential decreased, and this was correlated with expression of CD90.

Conclusion

Medial compartment knee OA demonstrates variability in the distribution of vessels, which results in a varying distribution of MSCs. The preculture period should be utilized to assess both the potential for expansion and the chondrogenic potential of MSCs.

Mesenchymal stem cells (MSCs) are attractive as a cell source for regenerative medicine, particularly in the treatment of cartilage injuries (1) and diseases such as osteoarthritis (OA) (2). An increasing number of reports suggest that MSCs can be isolated from various types of adult mesenchymal tissue, such as synovium (3), periosteum (4), skeletal muscle (5), and adipose tissue (6), in addition to bone marrow (7). We previously compared MSCs derived from the various types of mesenchymal tissue in young patients during ligament reconstruction. In this prior study, we found that synovial MSCs retained both a higher capacity for proliferation and a greater chondrogenic potential than did MSCs from other sources (8). In addition, we previously analyzed cells harvested from the fibrous synovium and adipose synovium of both young and elderly donors. Fibrous synovium and adipose synovial cells demonstrated similar self-renewal and differentiation capacities, irrespective of donor age (9). Furthermore, we used synovial MSCs for the repair of cartilage defects in rabbit knees, and the results demonstrated that implanted MSCs differentiated into cartilage appropriate to the microenvironment in vivo (10).

OA is a condition of destruction of articular cartilage in a joint, and the disease progression is associated with varying degrees of synovitis. Mechanical factors will affect the synovium as well as the cartilage, possibly resulting in a different distribution in cellular compartmentalization. A successful outcome from the use of synovial MSCs for cartilage regeneration in the OA knee requires that a sufficient quantity of MSCs be obtained. Therefore, the distribution of MSCs in OA synovium must be determined. The first objective of the present study was to examine whether MSCs are distributed equally or unequally in the synovium of patients with knee OA. Determining the relationship between the number of MSCs and the histologic features of regions of the synovium would help clarify the progenitor, or niche, of synovial MSCs.

In addition, synovial MSCs may alter their properties during in vitro proliferation, and the conditions in which to expand the cells to optimize their potential to differentiate into multiple cell lineages has not yet been defined. Thus, our second objective was to examine the effects of 2 specific variables, preculture period and harvest site, on 2 cell parameters, cell size and surface epitopes, to determine how the chondrogenic differentiation potential of synovial MSCs can vary. This study is directed toward developing a clinically feasible strategy for use of synovial MSCs in the repair of cartilage defects. We thus set out to address this by identifying the optimal parameters for in vitro culture expansion and chondrogenesis of synovial MSCs.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Harvesting of synovium.

Human synovial tissue was harvested during total knee arthroplasty from the knee joints of 10 patients with medial compartment OA. The ages of the patients were between 63 years and 91 years. The femorotibial angle was between 185° and 195°. The articular cartilage of the medial femorotibial joint had deteriorated in all patients. The synovium was collected from 4 different sites in the knee: the medial capsule against the medial condyle of the femur (medial outer region), the suprapatellar pouch, the infrapatellar fat pad, and the medial condyle of the femur (medial inner region). Samples comprised the synovium with residual subsynovium, since separation of the layer of synovium from the subsynovial tissue was difficult. The study was approved by our Institutional Review Board, and informed consent was obtained from all study subjects.

Immunohistochemical analysis.

Synovial tissue was fixed overnight at 4°C in 4% paraformaldehyde, embedded in paraffin, and sectioned at 5 μm. The sections were incubated with a mouse monoclonal antibody (clone A2547; Sigma-Aldrich, St. Louis, MO) against human α-smooth muscle actin (1:4,000 dilution with phosphate buffered saline [PBS] containing 1% bovine serum albumin [BSA]) for 1 hour or with a mouse monoclonal antibody (clone JC70A; Dako, Cambridge, UK) against human CD31 (1:50 dilution with PBS containing 1% BSA) overnight at 4°C. The sections were washed with PBS, and then incubated for 30 minutes with a biotinylated secondary antibody (1:200 dilution with PBS containing 1% BSA; Vector, Burlingame, CA).

After thorough washing with PBS, immunostaining was detected with an avidin–biotin–peroxidase complex (Elite ABC kit; Vector). The sections were counterstained with Mayer's hematoxylin (Vector). To determine the number of α-smooth muscle actin–positive vessels and the number of CD31+ cells, 3 slides of each tissue sample were analyzed by light microscopy (at 200× magnification). These were digitally converted to black and white images, and the numbers of positive cells were determined with Scion imaging software (Scion, Frederick, MD).

Isolation and cultures of MSCs.

Each synovial tissue sample was minced into small pieces with a surgical knife and then digested in a collagenase solution (3 mg/ml collagenase D; Roche Diagnostics, Mannheim, Germany) in α-minimum essential medium (α-MEM; Invitrogen, Carlsbad, CA) at 37°C. After 3 hours, the digested cells were filtered through a 70-μm nylon filter (Becton Dickinson, Franklin Lakes, NJ). Nucleated cells from the filtrate were cultured by plating 104 cells on three 60-cm2 dishes followed by incubation of the cells in complete medium (α-MEM containing 10% fetal bovine serum, 100 units/ml penicillin, 100 μg/ml streptomycin, and 250 ng/ml amphotericin B [all from Invitrogen]), with culturing at passage 0. After 14 days, the 3 dishes were stained with 0.5% crystal violet in methanol, and the number of colonies on each plate was counted. Colonies that were <2 mm in diameter or faintly stained colonies were ignored (11).

Flow cytometry assay.

Nucleated cells from the synovium of 5 donors were plated at 105 cells in three 60-cm2 dishes, and then cultured as passage 0 cells. The cells were trypsinized at 7, 14, and 21 days after plating, to count the cell number and to analyze surface epitopes. For analyses of cells at passage 1, the cells at passage 0 were cultured for 14 days and replated at 102 cells/cm2 in three 60-cm2 dishes, and again examined for surface epitopes.

One million cells were suspended in 500 μl PBS containing 20 μg/ml of a specific antibody. For determination of surface markers, we used fluorescein isothiocyanate (FITC)– or phycoerythrin (PE)–coupled antibodies against CD55 and CD90 (Becton Dickinson), CD54 and CD106 (eBioscience, San Diego, CA), CD68 (clone Ki-M6; Serotec, Kidlington, UK), and CD105 and CD166 (Ancell, Bayport, MN). CD68 was assessed after membrane permeabilization using Leucoperm (Serotec). 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 then suspended in 1 ml of PBS for analysis. For STRO-1 staining (Genzyme-Techne, Minneapolis, MN), the cells were incubated for 30 minutes with STRO-1 antibodies (mouse IgM). The cells were then incubated with a secondary antibody (FITC-conjugated goat anti-mouse IgM; Vector) for 30 minutes. As an isotype control for STRO-1, anti-mouse control IgM (eBioscience) was substituted. Cell fluorescence was evaluated by flow cytometry using a FACSCalibur instrument (Becton Dickinson), and the data were analyzed using CellQuest software (Becton Dickinson).

Differentiation assay.

For the chondrogenesis assay, 250,000 cells were placed in a 15-ml polypropylene tube (Becton Dickinson) and centrifuged for 10 minutes. The pellet was cultured in a chondrogenesis medium of high-glucose Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 500 ng/ml bone morphogenetic protein 2 (Astellas Pharmaceutical, Tokyo, Japan), 10 ng/ml transforming growth factor β3 (R&D Systems, Minneapolis, MN), 10−7M dexamethasone (Sigma-Aldrich), 50 μg/ml ascorbate-2-phosphate (Sigma), 40 μg/ml proline (Sigma), 100 μg/ml pyruvate (Sigma), and 50 mg/ml of an insulin–transferrin–selenium (ITS) mixture (ITS+Premix; Becton Dickinson) (12, 13). The medium was replaced every 3–4 days for 21 days.

For microscopy, the pellets were embedded in paraffin, cut into 5-μm sections, and stained with toluidine blue. For immunohistochemical staining, sections were treated with mouse monoclonal antibody against human type II collagen (1:100 dilution with PBS containing 1% BSA; Daiichi Fine Chemical, Toyama, Japan) for 1 hour at room temperature, and with a secondary antibody of biotinylated horse anti-mouse (1:200; Vector) for 30 minutes. Immunostaining was detected by Vectastain ABC reagent (Vector). Counterstaining was performed with Mayer's hematoxylin.

For assessment of adipogenesis, 100 cells were plated on 60-cm2 dishes and cultured in growth medium for 14 days. The medium was then changed to an adipogenesis medium that consisted of complete medium supplemented with 10−7M dexamethasone, 0.5 mM isobutylmethylxanthine (Sigma-Aldrich), and 50 μM indomethacin (Wako, Tokyo, Japan). The cells were cultured in adipogenesis medium for an additional 21 days. The adipogenic cultures were then fixed in 10% formalin for more than 1 hour, and stained with fresh oil red O solution for 2 hours. The oil red O solution was prepared by mixing 3 parts stock solution (0.5% in isopropanol; Sigma-Aldrich) with 2 parts water, and filtering through a 0.2-μm filter. Plates were washed 3 times with PBS, and the number of oil red O–positive colonies was determined. Colonies that were <2 mm in diameter or faintly stained colonies were ignored. The cultures were then stained with crystal violet, and the total number of cell colonies was counted (14).

For calcification, 100 cells were plated in 60-cm2 dishes and cultured for 14 days. The medium was then switched to calcification medium that consisted of complete medium supplemented with 10−9M dexamethasone (Sigma-Aldrich) and 20 mM β-glycerol phosphate (Wako), and then supplemented with 50 μg/ml ascorbate-2-phosphate (Sigma-Aldrich). The cells were cultured in calcification medium for an additional 21 days. The cells were then fixed in 10% formalin for more than 1 hour, and stained with 2% fresh alizarin red (pH 4.1, comprising 0.1 gram alizarin red, 5 ml distilled water, and 1 μl ammonium hydroxide) for 5 minutes. Plates were washed 3 times with PBS, and the number of alizarin red–positive colonies was counted. Colonies that were < 2 mm in diameter or colonies that had yellowish staining were ignored (15). The cultures were then stained with crystal violet, and the total number of cell colonies was counted.

Analysis of chondroitin sulfate synthesis.

The amount of chondroitin sulfate was quantified according to a previously described method (16). Each pellet (n = 4) was digested with 3 mg/ml collagenase in 0.4 ml DMEM for 3 hours at 37°C. The supernatant was digested with chondroitinase ABC (Chase ABC; Seikagaku, Tokyo, Japan) for 2 hours at 37°C. After ultrafiltration, the filtrate was analyzed by high-performance liquid chromatography. Unsaturated disaccharides derived from isomers of chondroitin 4-sulfate and chondroitin 6-sulfate were evaluated.

Statistical analysis.

Friedman's test was used for the analyses of flow cytometry data and findings in the cartilage pellets. Quantitative analyses of cell numbers and of MSC chondrogenesis in different areas were performed using the Kruskal-Wallis test. Spearman's rank correlation analysis was used to assess the relationship between histologic findings and cell colony numbers and also between chondrogenesis data and flow cytometry data.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Relationship between number of blood vessels and number of colony-forming cells in the synovium of patients with medial compartment OA.

Samples of OA synovium were harvested during total knee arthroplasty from 4 different regions of the medial compartment (Figure 1A, top). Each synovial tissue sample was divided into 2 parts, one for histologic evaluation, and the other for analysis of colony-forming cells. According to immunohistologic analysis of α-smooth muscle actin in the synovial tissue, a method involving staining for vascular pericytes (17, 18), the vascularity of the tissue appeared to be more concentrated at the medial outer region than at the other 3 regions (Figure 1A, middle). Quantitative analysis by Kruskal-Wallis test demonstrated a significantly different distribution of α-smooth muscle actin–positive vessels between the 4 regions analyzed (Figure 1B). Similar results were obtained in assessing the number of CD31+ vascular endothelial cells.

thumbnail image

Figure 1. Relationship between number of blood vessels and number of colony-forming cells in the synovium of patients with medial compartment osteoarthritis. A, Macroscopic (macro) features of the harvest sites (top), and histologic staining for α-smooth muscle actin (αSMA)–positive vessels (middle) and CD31+ cells (bottom) in a representative synovial sample. Bar = 100 μm. B, Number of αSMA-positive vessels and CD31+ cells in the different harvest sites. Values are the mean ± SD of 3 slides per tissue from 4 donors. C, Relationship between histologic findings and cell colony number, by Spearman's rank correlation. Synovial cells from the 4 harvest sites were plated at 105 cells per 60-cm2 dish, cultured for 14 days, and stained with crystal violet. Plotted values are the mean number of colonies per 103 nucleated cells of 3 slides per tissue from 4 donors. D, Colony numbers in the different harvest sites after 14 days in culture. Values are the mean ± SD number of colonies per 103 nucleated cells of 3 slides per tissue from 4 donors. E, Cell number per colony after 7, 14, and 21 days in culture. Values are the mean ± SD of 3 slides per tissue from 3 donors. P values in B and C were determined by Kruskal-Wallis test; P values in D were determined by Friedman's test.

Download figure to PowerPoint

Interestingly, the number of α-smooth muscle actin–positive vessels and the number of CD31+ endothelial cells were correlated with the colony-forming capabilities of the synovial MSCs (Figure 1C). Spearman's rank correlation analysis showed that the number of colony-forming cells had a higher correlation with the number of CD31+ endothelial cells than with the number of α-smooth muscle actin–positive vessels. Furthermore, the α-smooth muscle actin–positive vessel number was correlated with the CD31+ endothelial cell number (results not shown).

As expected from these results, the colony number per 103 nucleated cells was dependent on the anatomic location of the harvested tissue from the medial femorotibial joints. The medial outer region had the most colony-forming potential (Figure 1D). However, the cell number per colony was not affected by the harvest site, regardless of the duration of the culture period (Figure 1E).

Increase in synovial cell size during culture.

To determine the growth curve of the cells during culture, nucleated cells from OA synovium were plated at 105 nucleated cells/60-cm2 dish and cultured for 21 days. Passage 0 cells from most donors expanded rapidly between day 4 and day 7, and then grew moderately up to day 21 (Figure 2A). Upon staining with crystal violet, the cell colonies appeared faintly stained on day 7, showed distinctive staining on day 14, and overlapped by day 21 (Figure 2B). The cells in the colonies were spindle-shaped, as determined by microscopic observation, and the size of the cells increased during the 21 days of culture, according to the results of flow cytometry.

thumbnail image

Figure 2. Growth curve and morphologic features of synovial cells at passage 0. Nucleated cells from the suprapatellar pouch were plated at 105 cells per 60-cm2 dish and cultured. A, Growth curves of synovial cells at 4, 7, 11, 14, and 21 days of culture. Values are the mean of 3 slides per tissue from 5 donors. B, Sequential analyses of the cell colony numbers (left) and cell morphologic features (middle) by crystal violet staining, and analysis of cell size (right) by flow cytometry in the synovium from a representative donor (same donor as indicated by the solid black dot in A). The broken vertical lines in the flow cytometry analyses (right) indicate the median value of forward scatter height over 20 counts.

Download figure to PowerPoint

Sequential epitope profiles during proliferation.

Surface marker expression by passage 0 cells was investigated using synovial tissue samples from 5 donors (Figure 3). CD105 (endoglin, SH2) expression was observed in almost 100% of the cells on day 7 of the culture period, and decreased thereafter. Expression of CD166 (activated leukocyte cell adhesion molecule) was observed in 50–80% of the cells on day 7, and then decreased rapidly. Expression of CD90 (Thy-1) was evident in 40–60% of the cells on day 7, and then decreased gradually. Expression of CD55 (decay-accelerating factor), reported to be a marker of fibroblast-like synoviocytes (19), was evident in 20–30% of the cells on day 7, and subsequently decreased. Expression of STRO-1 was observed in ∼50% of the cells after 7 days, and thereafter remained stable. Furthermore, throughout the duration of culture, 20–40% of the cells were positive for CD54 (intercellular adhesion molecule 1), and 10–20% were positive for CD106 (vascular cell adhesion molecule 1). Finally, CD68, a marker of macrophages, was not detected after 7 days in culture.

thumbnail image

Figure 3. Sequential analyses of surface epitope expression during expansion of synovial cells at passage 0. Nucleated cells from the suprapatellar pouch were plated at 105 cells per 60-cm2 dish, cultured, and harvested at 7, 14, and 21 days. Individual results in samples from 5 patients are shown by culture period. P values were determined by Friedman's test. NS = not significant.

Download figure to PowerPoint

Chondrogenesis.

To determine the effect of in vitro expansion on the chondrogenic potential of synovial cells, the cells derived from the suprapatellar pouch were precultured for 7, 14, or 21 days, and 250,000 cells were then pelleted and placed in chondrogenesis medium. After 21 days in chondrogenesis medium, all of the pellets had increased in size, and this was attributed to the production of extracellular cartilage matrix (12).

The in vitro chondrogenesis assay of cells from the suprapatellar pouch showed that the size of the pellets decreased as the preculture period increased (Figure 4A, panel a). The weight of the cartilage pellets derived from passage 0 cells also significantly decreased with increasing preculture period (Figure 4A, panel b). In addition, the amount of chondroitin 4-sulfate per pellet and chondroitin 6-sulfate per pellet decreased in parallel with increasing preculture period, and these findings were related to the weight of the cartilage pellets (Figure 4A, panel c). The maximum chondrogenic potential of the MSCs was also observed at 7 days of preculture in the 3 other harvest sites (Figure 4B, panel a).

thumbnail image

Figure 4. Chondrogenic potential of mesenchymal stem cells (MSCs). A, Effects of preculture on chondrogenesis of MSCs in synovial tissue derived from the suprapatellar pouch. Nucleated cells from the suprapatellar pouch were plated at 105 cells per 60-cm2 dish, cultured, and harvested at 7, 14, and 21 days. The cells were differentiated into chondrocytes using a pellet culture system. a, Macroscopic (macro) features of the cartilage pellets (scale = 1 mm) (top), histologic staining with toluidine blue (TB) (middle), and immunohistologic staining for type II collagen (COL2) (bottom) in a representative synovial sample. b, Wet weight of the cartilage pellets. Values are the mean ± SD of 4 pellets from 4 donors. P values were determined by Friedman's test. c, Expression of chondroitin 4-sulfate and chondroitin 6-sulfate per pellet. Four pellets were pooled and digested with collagenase, and the supernatant was analyzed. B, Chondrogenesis of MSCs in synovial tissue derived from the different harvest sites. a, Pellet wet weight by culture period in the 4 harvest sites. Values are the mean and SD of 3 pellets from 3 donors. b, Relationship of surface epitope expression to MSC chondrogenesis in tissue derived from the suprapatellar pouch of knees from 3 donors. c, Relationship of CD90 expression to MSC chondrogenesis in tissue derived from 3 harvest sites in knees from 3 donors. P values were determined by Spearman's rank correlation.

Download figure to PowerPoint

We investigated the relationship between surface epitope expression and the chondrogenic potential of MSCs derived from the suprapatellar pouch. There was no relationship between the chondrogenic potential and expression of CD105, CD166, or STRO-1. Interestingly, the expression of CD90 was correlated with pellet weight (Figure 4B, panel b). A relationship between CD90 expression and the chondrogenic potential of the MSCs was also observed in MSCs derived from the medial outer region (Figure 4B, panel c).

Adipogenesis and calcification.

To examine the influence of the preculture period on the differentiation potential of MSCs, passage 0 cells were precultured for 7, 14, or 21 days. The cells were then replated at 100 cells/60-cm2 dish, and either an in vitro adipogenesis assay or a calcification assay was performed. In the adipogenesis assay, we found that the ratio of oil red O–positive colonies to total colonies was similar, at ∼60%, regardless of the duration of the preculture period (Figure 5A).

thumbnail image

Figure 5. Adipogenic and calcification potentials. Synovial cells from the suprapatellar pouch at passage 0 were harvested 7, 14, and 21 days after plating. The cells were replated and cultured to generate colonies, and then differentiated into adipocytes or calcified. A, Adipogenesis of synovial cells. a, Colonies staining positive for adipogenesis in cultures with oil red O and crystal violet. b, Proportion of oil red O–positive colonies to total number of colonies. B, Calcification of synovial cells. a, Colonies staining positive for calcification in cultures with alizarin red and crystal violet. b, Proportion of alizarin red–positive colonies to total number of colonies. Values are the mean and SD of 3 slides per tissue from 3 donors.

Download figure to PowerPoint

In addition, the calcification potential of the MSCs was evaluated. Similar to the results of the adipogenesis assay, the calcification assay showed that the ratio of alizarin red–positive colonies to total colonies was similar in the 3 populations, being ∼30% in each (Figure 5B). Therefore, the preculture period appears to have no effect on either the adipogenic potential or the calcification potential of MSCs.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Although the identity of MSCs has yet to be fully defined (20), we recognize that MSCs are derived from mesenchymal tissue and have the functional capacity both to self-renew and to generate a number of differentiated progeny (21). Since the earliest studies by Friedenstein et al (22), the standard assay used to identify the self-renewal ability of MSCs is the colony-forming–unit fibroblast assay. This assay measures the percent of cells with high replicative capabilities in a culture. Although clonal colonies of MSCs are readily prepared, the colonies rapidly become heterogeneous as they expand (23–25). The synovial cells used in the present study showed both colony-forming ability and multipotentiality, the hallmarks used to define cells as MSCs.

We recently reported that fibrous synovium, harvested from the inner side of the lateral joint capsule, and adipose synovium, harvested from the infrapatellar fat pad, had similar growth and differentiation potentials in young patients with anterior cruciate ligament injury and in elderly patients with OA (9). These findings suggest that adherent colony-forming cells derived from the synovium have a similar proliferation ability and multipotentiality, independent of the location in the knee. In this study, we also demonstrated that the capacity for proliferation and potential for chondrogenesis of synovial MSCs were similar among the different harvest sites in the OA synovium.

CD31, recognized as platelet endothelial cell adhesion molecule 1 (26), has been used as a candidate marker of endothelial cells (26, 27). It has been reported that both von Willebrand factor and CD31 are expressed on the same endothelial cells (28). However, CD31 is strongly expressed on vascular endothelial cells, whereas von Willebrand factor is strongly produced on other types of endothelial cells. In addition, previous studies have used CD31 as a marker of vascular endothelial cells (29), and α-smooth muscle actin has been considered a marker of vascular pericytes, specifically in the synovium (18). Therefore, in our study, we used CD31 to label vascular endothelial cells.

There is vast research-related interest in determining the anatomic location of MSCs within the tissue. In this study, we demonstrated that the number of α-smooth muscle actin–positive vessels and the number of CD31+ cells were strongly correlated with the number of colony-forming cells derived from the synovium. It was reported that the endothelial cells in both rheumatoid arthritis synovium and OA synovium are surrounded with cells positive for the STRO-1 surface protein, one marker of MSCs (29). In bone marrow, the perivascular region has been suggested as the niche for MSCs (30). In addition, microvascular pericytes derived from the retina and aorta were reported to have multipotential differentiation capabilities (31). This suggests that synovial MSCs may also exist in the perivascular niche. Our results, as well as those from other studies, indicate the possibility that MSCs in the synovium arise from committed progenitors that belong to the distinct lineage of vascular pericytes.

We collected synovium from 4 different sites in the knee, and our analyses showed that synovium at the medial outer region contained more CD31+ vascular endothelial cells and more α-smooth muscle actin–positive vessels than were found in the synovium at the suprapatellar pouch, infrapatellar fat pad, and medial inner regions. The femorotibial angle in all patients was >185°, and all patients were diagnosed as having medial compartment OA. The pathogenesis of OA is described as a process of destruction of the articular cartilage by both mechanical and biochemical factors, which is associated with varying degrees of synovitis (32–34). Saito and Koshino compared the distribution of neuropeptides in the synovium of knees with medial compartment OA and showed a higher perivascular neural network in the medial compartment than in the lateral compartment or the suprapatellar pouch (34). In our study, the medial compartment of the OA knees demonstrated this variability in the distribution of vascular endothelial cells and vascular pericytes, which also accounted for the varying distribution of colony-forming cells.

We previously reported that bone marrow–derived MSCs can undergo a time-dependent transition from small cells to large cells, and that the preculture period affects the chondrogenic differentiation potential (11, 12). In this study, we examined 2 variables, preculture period and harvest site, for their effects on 2 parameters, cell size and surface epitopes, to determine how the chondrogenic differentiation potential of synovial MSCs can vary. Similar to previous findings, the preculture period affected chondrogenesis, with an increase in size of the synovium-derived MSCs with increasing preculture period. Anatomic location did not affect chondrogenesis. Of the surface epitopes tested, no relationship was observed between chondrogenic potential and CD31/α-smooth muscle actin–positive staining or expression of CD105, CD166, or STRO-1. In contrast, CD90 expression was predictive of chondrogenesis (Figure 6).

thumbnail image

Figure 6. Summary of cell size, surface antigen expression, and chondrogenic potential during the expansion of synovial cells. During clonal expansion of synovial cells over the 21 days of preculture, the cell size increased, while the rate of positive expression of CD90 and the chondrogenic potential of the cells decreased.

Download figure to PowerPoint

CD90 (Thy-1), which was first recognized as a marker of thymus-derived lymphocytes (35), was detected through a screening of heterologous antisera against mouse leukemia cells. CD90 has also been considered a marker antigen of MSCs (36), although the precise biologic function is not yet clear (37). Fickert et al reported that initial sorting for CD9/CD90/CD166 triple-positive synovial cells revealed the multipotency of these cells (38), which supports our results. In our study, we found that CD90 is an important indicator of the chondrogenic differentiation potential of synovial MSCs.

We have previously quantitatively evaluated the chondrogenic potential of MSCs by pellet wet weight. According to our previous studies, during in vitro chondrogenesis of MSCs, the pellet increased in size, weight, and cartilage matrix synthesis. Conversely, the DNA yield per pellet decreased. Radioactivity per DNA in the cells, assessed by prelabeling with 3H-thymidine, was found to be stable during in vitro chondrogenesis of MSCs (39). These results indicate that the increase in pellet size can be attributed to production of extracellular matrix, and not the proliferation of the cells. We believe that both the size and the weight of the pellet are quantitative indicators of the ability of MSCs to produce chondrogenesis in vitro.

In our study, the pellet weight decreased in conjunction with increasing preculture period. It is indeed possible that beyond a certain number of population doublings, the chondrogenic potential would tend to plateau to a minimum level, so that differences would not be detectable, at least not with the assays used in this study. A similar mechanism can apply to adipogenesis and osteogenesis. For these assays, passage 0 cells were precultured for 7, 14, or 21 days and then replated at a very low density, in order to obtain colonies for the differentiation assays. Thus, cells underwent additional population doublings prior to differentiation. Also, in this case, it is possible that with the increase in population doublings, there was a decrease in the osteogenic and adipogenic differentiation potentials, and this would tend to plateau to a minimum, so that a difference, if any, would be difficult to detect. In other words, minimally expanded cells could have overall greater differentiation potential, which would decrease quickly during culture expansion.

We found variability in pellet weight between donors after in vitro chondrogenesis (Figure 4A, panel b). The harvest site, digestion of the synovium, expansion of the cells, and findings on in vitro chondrogenesis assay were similar between donors. Variability between donors depends on the number of chondroprogenitors and their potential for cartilage matrix synthesis.

Our findings demonstrate that the in vitro chondrogenic potential of MSCs can be affected by preculture conditions. It will be intriguing to examine whether the in vitro results would reflect those obtained in vivo. We are currently investigating the relationship between in vitro and in vivo chondrogenesis. For assessment of in vivo chondrogenesis, undifferentiated MSCs were transplanted into the joints of rabbits with cartilage defects (10), and the chondrogenic potential of the MSCs was evaluated histologically. The results of that study showed that when the difference in pellet weights between the 2 populations was large, this was reflected in the in vivo results.

In the present study, the adipogenic colony-forming efficiency of the synovial MSCs (∼60%) was higher than the calcification colony-forming efficiency (∼30%). We have previously compared MSCs derived from several mesenchymal tissue types and found that the adipogenic colony-forming efficiency was higher than the calcification colony-forming efficiency in MSCs derived from the synovium and adipose tissue. In contrast, the calcification colony-forming efficiency was higher than the adipogenic colony-forming efficiency of MSCs derived from the bone marrow and periosteum (8). These results suggest that the local tissue microenvironment may be directing the “fate” of the MSCs toward a particular lineage (9).

In this study, the preculture period did not affect the adipogenic potential or osteogenic potential of synovial MSCs. We previously reported that the ability of bone marrow–derived MSCs to undergo adipogenesis was higher when the cells were plated at lower densities and cultured for shorter preculture periods (11). We also demonstrated that the adipogenic potential was much higher in synovial MSCs than in bone marrow–derived MSCs (8). Evaluation of adipogenic potential through assessment of the oil red O–positive colony-forming efficiency is a simple method, but its sensitivity may be too low to detect any difference in synovial MSCs. For determination of osteogenic ability, more sensitive methods may be able to detect differences induced by preculture conditions.

The number of MSCs was correlated with the number of CD31/α-smooth muscle actin–positive cells, and these cells were located in different distributions in different areas of the synovium of patients with medial compartment OA. However, the proliferation and differentiation potentials of the MSCs were not affected by harvest site.

MSCs in a shorter preculture period display higher chondrogenic potential, which suggests that the whole synovium in patients with medial compartment knee OA can be used to obtain a high number of MSCs with high chondrogenic potential, provided that the cells are expanded in a minimal culture period. However, since the amount of synovium needed to obtain a sufficient number of MSCs is limited, a longer culture period is required to obtain a higher number of MSCs. Therefore, 2 issues need to be addressed with regard to assessment of synovial MSCs in the preculture period: how to expand the synovial cells to sufficient numbers for clinical relevance, and how to retain the chondrogenic potential of the synovial MSCs to achieve efficacy. In addition, analysis of CD90 expression on synovial MSCs may serve as a useful tool for optimizing chondrogenesis of these cells in cartilage repair.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Dr. Sekiya had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Muneta, Sekiya.

Acquisition of data. Nagase, Ju, Hara, Morito, Koga, Nimura, Mochizuki.

Analysis and interpretation of data. Nagase, Sekiya.

Manuscript preparation. Nagase, Sekiya.

Statistical analysis. Nagase, Sekiya.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

We thank Kenichi Shinomiya for supporting our studies, Kazuyoshi Yagishita and Shinichi Shirasawa for harvesting the synovium, Izumi Nakagawa for excellent technical assistance, Miyoko Ojima for expert help with the histology, and Alexandra Peister for proofreading.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES