Higher chondrogenic potential of fibrous synovium and adipose synoviumderived cells compared with subcutaneous fatderived cellsDistinguishing properties of mesenchymal stem cells in humans.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 54, No. 3, March 2006, pp 843–853 DOI 10.1002/art.21651 © 2006, American College of Rheumatology Higher Chondrogenic Potential of Fibrous Synovium– and Adipose Synovium–Derived Cells Compared With Subcutaneous Fat–Derived Cells Distinguishing Properties of Mesenchymal Stem Cells in Humans Tomoyuki Mochizuki, Takeshi Muneta, Yusuke Sakaguchi, Akimoto Nimura, Akiko Yokoyama, Hideyuki Koga, and Ichiro Sekiya Results. Fibrous synovium– and adipose synovium–derived cells were higher in STRO-1 and CD106 and lower in CD10 compared with subcutaneous fat–derived cells. Cells derived from fibrous and adipose synovium had higher proliferative potential and colonyforming efficiency compared with subcutaneous fat– derived cells, both in mixed-population and in singlecell–derived cultures. In chondrogenic assays, pellets from fibrous synovium– and adipose synovium–derived cells produced more cartilage matrix than did cell pellets from subcutaneous fat. Osteogenic ability was also higher in fibrous synovium– and adipose synovium–derived cells, whereas adipogenic potential was nearly indistinguishable among the 3 populations. Differentiation potential of the cells was similar between young and elderly donors. Conclusion. Cells derived from the fibrous synovium and from the adipose synovium demonstrate comparable chondrogenic potential. Adipose synovium– derived cells are more similar to fibrous synovium– derived cells than to subcutaneous fat–derived cells. Objective. Mesenchymal stem cells from synovium have a greater proliferation and chondrogenic potential than do those from bone marrow, periosteum, fat, and muscle. This study was undertaken to compare fibrous synovium and adipose synovium (components of the synovium with subsynovium) to determine which is a more suitable source for mesenchymal stem cells, especially for cartilage regeneration, and to examine the features of adipose synovium–derived cells, fibrous synovium–derived cells, and subcutaneous fat–derived cells to determine their similarities. Methods. Human fibrous synovium, adipose synovium, and subcutaneous fat were harvested from 4 young donors and 4 elderly donors. After digestion, the nucleated cells were plated at a density considered proper to expand at a maximum rate without colony-to-colony contact. The surface epitopes, proliferative capacity, cloning efficiency, and chondrogenic, osteogenic, and adipogenic differentiation potentials of the cells were compared. Supported in part by the Japan Society for the Promotion of Science (grant 16591478), the Japan Orthopaedics and Traumatology Foundation, and the Japan Sports Medicine Foundation. Dr. Muneta’s work was supported by the Japan Society for the Promotion of Science (grant 16591477), the Japan Latest Osteoarthritis Society, and the Center of Excellence Program for Frontier Research on Molecular Destruction and Reconstruction of Tooth and Bone in Tokyo Medical and Dental University. Dr. Sekiya’s work was supported by the Nakatomi Foundation. Tomoyuki Mochizuki, MD, PhD, Takeshi Muneta, MD, PhD, Yusuke Sakaguchi, MD, PhD, Akimoto Nimura, MD, Akiko Yokoyama, MD, Hideyuki Koga, MD, Ichiro Sekiya, MD, PhD: Tokyo Medical and Dental University, Tokyo, Japan. Address correspondence and reprint requests to Ichiro Sekiya, MD, PhD, Section of Orthopedic Surgery, Graduate School, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8519, Japan. E-mail: firstname.lastname@example.org. Submitted for publication May 27, 2005; accepted in revised form November 21, 2005. For treatment of articular cartilage injury, one of the promising procedures is the transplantation of autologous cultured chondrocytes (1). However, surgical invasion of normal articular cartilage and limited ex vivo expansion of the chondrocytes lead to difficulties in repairing large defects. Mesenchymal stem cells (MSCs) have been a fascinating source for use in regenerative medicine because they can be harvested in a less invasive manner. Moreover, MSCs are easily isolated and expanded, with multipotential capabilities, including chondrogenesis (2,3). An MSC is defined as being derived from mesenchymal tissue and having the functional capacity for self843 844 MOCHIZUKI ET AL renewal, commonly identified by colony-forming unit fibroblast assay (4) and generation of a number of differentiated progeny (5). Increasing evidence suggests that postnatal stem cells are not exclusive to bone marrow, but also are present in various other tissues. We previously compared MSCs derived from bone marrow, synovium, periosteum, adipose tissue, and muscle, demonstrating that synovium was a better cell source for MSCs with regard to cartilage regeneration, in that synovium-derived MSCs had a greater proliferative capacity and chondrogenic potential (6). Synovium is a thin layer of tissue that lines the joint space and covers a subsynovium. Depending on its anatomic position, subsynovium comprises either a fibrous or an adipose connective tissue. There have been other reports describing human synovium–derived MSCs; however, details regarding the harvest site of synovium and the histologic characteristics were not mentioned (7,8). Most likely, synovium with subsynovium was used in these studies, because separation of only the synovium layer from the subsynovial tissue is difficult. Our first aim in the present study was to distinguish MSCs by their sources, that is, synovium with fibrous subsynovium, referred to as fibrous synovium, and synovium with adipose subsynovium, referred as adipose synovium. MSCs derived from adipose synovium, also commonly called the infrapatellar fat pad, have been reported to have multidifferentiation potential. MSCs of the adipose synovium have been regarded as similar to liposuctionderived cells (9), except that the infrapatellar fat pad is covered with synovium. Our second aim in the present study was to identify whether infrapatellar fat pad–derived cells are more closely related to fibrous synovium–derived cells than to subcutaneous fat–derived cells. In this study, we collected fibrous synovium, adipose synovium, and subcutaneous fat and performed patient-matched quantitative comparisons of the properties of the 3 MSC populations. The properties examined were surface epitopes, proliferative capacity, cloning efficiency, and chondrogenic, osteogenic, and adipogenic differentiation potentials. The goal of this study was to characterize the suitability of fibrous synovium–derived MSCs as compared with adipose synovium– or subcutaneous fat–derived MSCs for cartilage regeneration, from the standpoint of the properties of each MSC population and the accessibility of the MSC sources. PATIENTS AND METHODS Harvest of fibrous synovium, adipose synovium, and subcutaneous fat. Tissues were harvested from 8 human donors during knee operations. These donors were 4 young patients with anterior cruciate ligament injury (mean ⫾ SD age 17.2 ⫾ 0.7 years) and 4 elderly patients with osteoarthritis (mean ⫾ SD age 70.5 ⫾ 9.2 years). Young patients underwent arthroscopic anterior cruciate ligament reconstructions, and elderly patients received total knee arthroplasties. Fibrous synovium was harvested from the inner side of the lateral joint capsule, which overlays the noncartilaginous areas of the lateral condyles of the femur. Adipose synovium was harvested from the inner side of the infrapatellar fat pad, which overlays the patellar tendon. Subcutaneous fat tissue under skin incisions over the tibiae was harvested. The study was approved by an institutional review board, and informed consent was obtained from all study subjects. Histologic analysis. A portion of the tissues was fixed overnight at 4°C in 10% formalin, embedded in paraffin, and sectioned at 5 m. Tissue sections were then stained with hematoxylin and eosin to identify histologic features. Isolation and culture of MSCs. The tissue was minced to pieces with a surgical knife, washed thoroughly with phosphate buffered saline (PBS) to remove hematopoietic cells, and digested in a collagenase solution (3 mg/ml collagenase D; Roche Diagnostics, Mannheim, Germany) in ␣–minimum essential medium (Invitrogen, Carlsbad, CA) at 37°C. After 3 hours, digested cells were filtered through a 70-m nylon filter (Becton Dickinson, Franklin Lakes, NJ). Nucleated cells from the tissues were plated at 103, 104, or 105 cells/cm2, each in 6 dishes, and cultured for 14 days at passage 0. After each of 3 dishes was stained with 0.5% crystal violet, optimal initial cell densities were determined; this was decided according to the supposition that colony size was not affected by colony-tocolony contact and that greater numbers of colonies were therefore obtained. Total cell yields were thus established, and the cells were used in further experiments (6). Flow cytometry. One million cells (at passage 3) were suspended in 500 l PBS containing 20 g/ml of antibody. After incubation for 30 minutes at 4°C, the cells were washed with PBS and suspended in 1 ml PBS for the analysis. Fluorescein isothiocyanate (FITC)– or phycoerythrin (PE)– coupled antibodies against CD34, CD45, CD90, and CD147 and anti–nerve growth factor receptor (anti-NGFR) antibody were from Becton Dickinson, CD31,CD44, CD54 (intracellular adhesion molecule 1 [ICAM-1]), CD106 (vascular cell adhesion molecule 1 [VCAM-1]), and CD117 were from eBioscience (San Diego, CA), CD105 and CD166 (activated leukocyte cell adhesion molecule [ALCAM]) were from Ancell (Bayport, MN), STRO-1 and vascular endothelial cell growth factor receptor 2 (VEGFR-2) were from Genzyme-Techne (Minneapolis, MN), and CD10 was from DakoCytomation (Copenhagen, Denmark). For isotype control, FITC- or PEcoupled nonspecific mouse IgG (Becton Dickinson) was substituted for the primary antibody. Cell fluorescence was evaluated by flow cytometry using a FACSCalibur instrument (Becton Dickinson), and data were analyzed using CellQuest software (Becton Dickinson). All analyses were performed on cells from 2 young donors and 2 elderly donors. For STRO-1 staining, the cells were incubated for 30 minutes with an antibody against STRO-1 (mouse IgM; Genzyme-Techne). The cells were then incubated with a secondary antibody (fluorescein-conjugated goat anti-mouse IgM; Vector, Burlingame, CA) for 30 minutes. For the isotype control against STRO-1, anti-mouse IgM (eBioscience) was substituted. Positive expression was defined as a level of PROPERTIES OF HUMAN SYNOVIAL MESENCHYMAL STEM CELLS fluorescence ⬎99% of that observed with the corresponding isotype-matched control antibodies (6,10). Colony-forming efficiency. The cells at passage 0 were replated at 100 cells per 60-cm2 dish, incubated for 14 days, and stained with 0.5% crystal violet in 4% paraformaldehyde for 5 minutes. The cells were washed twice with distilled water, and the number of colonies per dish was determined. Colonies ⬍2 mm in diameter and faintly stained colonies were ignored (11). In vitro chondrogenesis. Two hundred thousand cells were placed in a 15-ml polypropylene tube (Becton Dickinson) and centrifuged at 450g for 10 minutes. The pellet was cultured at 37°C with 5% CO2 in 400 l of chondrogenic medium that contained 500 ng/ml bone morphogenetic protein 2 (Yamanouchi Pharmaceutical, Tokyo, Japan) in high-glucose Dulbecco’s modified Eagle’s medium (DMEM; Sigma-Aldrich, St. Louis, MO) supplemented with 10 ng/ml transforming growth factor ␤3 (R&D Systems, Minneapolis, MN), 100 nM dexamethasone (Sigma-Aldrich), 50 g/ml ascorbate-2-phosphate, 40 g/ml proline, 100 g/ml pyruvate, and 1:100 diluted ITS⫹ Premix (6.25 g/ml insulin, 6.25 g/ml transferrin, 6.25 ng/ml selenious acid, 1.25 mg/ml bovine serum albumin, and 5.35 mg/ml linoleic acid; BD Biosciences, Bedford, MA). The medium was replaced every 3–4 days for 21 days. For microscopy, the pellets were embedded in paraffin, cut into 5-m sections, and stained with toluidine blue (12,13). Reverse transcription–polymerase chain reaction (RTPCR). Cartilage pellets were digested with 3 mg/ml collagenase for 3 hours to collect cells, and total RNA was prepared using the RNAqueous kit (Ambion, Austin, TX). RNA was converted to complementary DNA (cDNA) and amplified with the Titan One Tube RT-PCR System (Roche Diagnostics) according to the manufacturer’s recommendations. RT was performed with a 30-minute incubation at 50°C, followed by a 2-minute incubation at 94°C to inactivate the RT. PCR amplification of the resulting cDNA was performed under the following conditions: 35 cycles at 94°C for 30 seconds, 58°C for 45 seconds, and 68°C for 45 seconds, the latter of which was increased by 5 seconds every cycle after 10 cycles. Forty cycles of these steps were performed for SOX5 and SOX9. The reaction products were resolved by electrophoresis on a 1.5% agarose gel and visualized with ethidium bromide (14). PCR primers were as follows: for COL2A1, 5⬘-TTCAGCTATGGAGATGACAATC-3⬘ (forward) and 5⬘AGAGTCCTAGAGTGACTGAG-3⬘ (reverse) (472 bp); for aggrecan, 5⬘-GCAGAGACGCATCTAGAAATT-3⬘ (forward) and 5⬘-GGTAATTGCAGGGAACATCAT-3⬘ (reverse) (505 bp); for decorin, 5⬘-CCTTTGGTGAAGTTGGAACG-3⬘ (forward) and 5⬘-AAGATGTAATTCCGTAAGGG-3⬘ (reverse) (300 bp); for biglycan, 5⬘-TGCAGAACAACGACATCTCC-3⬘ (forward) and 5⬘-AGCTTGGAGTAGCGAAGCAG-3⬘ (reverse) (475 bp); for link protein, 5⬘-CCTATGATGAAGCGGTGC-3⬘ (forward) and 5⬘-TTGTGCTTGTGGAACCTG-3⬘ (reverse) (618 bp); for SOX5, 5⬘-AGCCAGAGTTAGCACAATAGG-3⬘ (forward) and 5⬘-CATGATTGCCTTGTATTC-3⬘ (reverse) (619 bp); for SOX6, 5⬘-ACTGTGGCTGAAGCACGAGTC-3⬘ (forward) and 5⬘-TCCGCCATCTGTCTTCATACC-3⬘ (reverse) (562 bp); for SOX9, 5⬘-GAACGCACATCAAGACGGAG-3⬘ (forward) and 5⬘TCTCGTTGATTTCGCTGCTC-3⬘ (reverse) (631 bp); for chondroitin 4-sulfotransferase (C4ST), 5⬘-CATCTACTGCTACGTGCCCA-3⬘ (forward) and 5⬘-CTTCAGGTAGCTGCCCACTC-3⬘ (reverse) (547 bp); for C6ST, 5⬘-GACTTT- 845 GTGCACAGCCTGAA-3⬘ (forward) and 5⬘-CCCTGCTGGTTGAAGAACTC-3⬘ (reverse) (431 bp); and for ␤-actin, 5⬘-CCAAGGCCAACCGCGAGAAGATGAC-3⬘ (forward) and 5⬘-AGGGTACATGGTGGTGCCGCCAGAC-3⬘ (reverse) (587 bp). Analysis of glycosaminoglycans (GAGs). The GAG content was quantified according to a previously described method (15). Cartilage pellets were digested with 3 mg/ml collagenase in 0.25 ml of DMEM for 3 hours at 37°C, and a fraction of the cells was removed. The supernatant was digested with chondroitinase ABC (Chase ABC; Seikagaku, Tokyo, Japan) and hyaluronidase derived from Streptococcus dysgalactonase (HAase SD; Seikagaku) for 2 hours at 37°C. After ultrafiltration, the filtrate was analyzed by high-performance liquid chromatography. The levels of chondroitin 4-sulfate (C4S), chondroitin 6-sulfate (C6S), and hyaluronic acid were evaluated (14). Osteogenesis in a colony-forming assay. One hundred cells were plated in 60-cm2 dishes and cultured in complete medium for 14 days. The medium was switched to calcification medium that consisted of complete medium supplemented with 1 nM dexamethasone (Sigma-Aldrich), 20 mM glycerol phosphate (Wako, Osaka, Japan), and 50 g/ml ascorbate-2-phosphate for an additional 21 days. Dishes were subsequently stained with fresh 0.5% alizarin red solution, and the number of alizarin red–positive colonies was determined. Colonies ⬍2 mm in diameter and faintly stained colonies were ignored. The same calcification cultures were then stained with crystal violet, and the total number of cell colonies was determined (6,10). Adipogenesis in a colony-forming assay. One hundred cells were plated in 60-cm2 dishes and cultured in complete medium for 14 days. The medium was then switched to adipogenic medium that consisted of complete medium supplemented with 10 nM dexamethasone, 0.5 mM isobutylmethylxanthine (Sigma-Aldrich), and 50 M indomethacin (Wako) for an additional 21 days. The adipogenic cultures were fixed in 4% paraformaldehyde, stained with fresh 0.5% oil red O solution, and the number of oil red O–positive colonies was determined. Colonies ⬍2 mm in diameter and faintly stained colonies were ignored. The same adipogenic cultures were subsequently stained with crystal violet, and the total number of cell colonies was determined (16). Statistical analysis. Analysis of variance was used for assessing differences between cell populations. P values less than 0.05 were considered significant. RESULTS Macroscopic and histologic features of fibrous synovium, adipose synovium, and subcutaneous fat. On macroscopic analysis, fibrous synovium appeared relatively white, whereas both adipose synovium and subcutaneous fat were yellowish white (Figure 1A). Interestingly, fibrous synovium sunk, adipose synovium floated partially, and subcutaneous fat floated completely in PBS, indicating a difference in specific gravities (Figure 1B). Histologically, fibrous synovium was composed of mostly fibrous tissues, whereas adipose synovium consisted of both fibrous tissues and subsynovial fatty tissue. It appeared that samples from the elderly patients had more fibrous tissue than did those from the young donors, both 846 MOCHIZUKI ET AL Figure 1. Comparison of fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells. A and B, Macroscopic features of cells from an elderly donor were examined on a 1-mm scale (A) and in phosphate buffered saline (B). C, Histologic analysis was performed on cells (stained with hematoxylin and eosin) from both young and elderly donors. D, To determine the nucleated cell number per tissue weight, materials were digested with collagenase, nucleated cells were counted, and the mean (and SEM) number of nucleated cells per tissue weight was calculated (n ⫽ 4). E, Representative cell colonies from an elderly donor are shown. Nucleated cells were plated at 103, 104, and 105 cells/cm2 in 60-cm2 dishes, cultured for 14 days, and stained with crystal violet. Representative morphologic features of the cells shown in E are magnified (⫻200) in F. in fibrous synovium and in adipose synovium (Figure 1C). Few, if any, erythrocytes were observed. Nucleated cell number per tissue weight. The mean nucleated cell number per tissue weight is shown in Figure 1D. The greatest number of nucleated cells was found in fibrous synovium, an intermediate amount was found in adipose synovium, and the fewest number of nucleated cells was found in subcutaneous fat. The nucleated cell PROPERTIES OF HUMAN SYNOVIAL MESENCHYMAL STEM CELLS Figure 2. Flow cytometric analysis of fibrous synovium–derived cells (solid bars), adipose synovium–derived cells (shaded bars), and subcutaneous fat–derived cells (open bars). All analyses were performed on cells from 2 young donors and 2 elderly donors. Values are the mean and SD percentage expression of each cell-surface protein. Data were analyzed using one-way analysis of variance to assess the effect of cell sources. ⴱ ⫽ P ⬍ 0.05. VEGFR2 ⫽ vascular endothelial growth factor receptor 2; NGFR ⫽ nerve growth factor receptor. 847 number per tissue weight in the fibrous synovium of elderly donors was substantially larger than that in the fibrous synovium of young donors. Colony formation and morphologic features of the cells. In order to gain maximum yields per nucleated cells, we examined the effect of plating density on nucleated cells. Larger single-cell–derived colonies were observed when the nucleated cells were plated at 103 cells/cm2 in all populations (Figure 1E). When plated at 104 or 105 cells/cm2, colony size decreased or became indistinct, possibly due to colony-to-colony contact inhibition. These observations indicated an optimal initial cell density at 103 cells/cm2 to maximize cell yields per dish. Thus, we were able to identify the optimal initial cell density for nucleated cells, and this was used in further experiments. On morphologic analysis, fibrous synovium– Figure 3. A, Proliferation potential of fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells from 1 young donor. Cells at passage 1 were plated at 10, 50, and 100 cells/cm2 and the fold increase (n ⫽ 3) was calculated after 7 and 14 days. Data were analyzed using one-factor analysis of variance (ANOVA) to assess the effect of plating density on fold increase. B, The cells at passage 1 from 4 young and 4 elderly donors were plated in triplicate at 50 cells/cm2 and the fold increase was calculated after 14 days. Data were analyzed using two-factor ANOVA to assess the effect of cell sources and donor age. C, One hundred cells at passage 1 from 4 young and 4 elderly donors were plated in triplicate in 60-cm2 dishes and the colony-forming efficiency was calculated after 14 days. D, For analyses of single-cell–derived cultures, nucleated cells from 1 elderly donor and 1 young donor were plated at 103 cells/cm2 and cultured for 14 days, and 3 single-cell–derived colonies were collected. Three clones from fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells were replated at 50 cells/cm2, each in triplicate, and the fold increase was calculated after 14 days. E, One hundred cells from the clones were plated in triplicate in 60-cm2 dishes and the colony-forming efficiency was calculated after 14 days. Bars show the mean and SD. ⴱ ⫽ P ⬍ 0.05; ⴱⴱ ⴝ P ⬍ 0.01; ⴱⴱⴱ ⫽ P ⬍ 0.005. 848 MOCHIZUKI ET AL Figure 4. A and B, Chondrogenic potential of fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells. The cells were pelleted and cultured in chondrogenic medium for 21 days. Representative macroscopic findings from a young donor are shown on a 1-mm scale (A), and the histologic features were examined by staining with toluidine blue (B). C, The wet weight of the pellets from 3 young and 3 elderly donors was measured in triplicate; data are expressed as the mean and SD, analyzed using two-factor analysis of variance to assess the effect of cell sources and donor age. D, To compare glycosaminoglycans in the extracellular matrix, 3 cell pellets from 1 young donor were digested to remove the cells, and chondroitin 4-sulfate (C4S), chondroitin 6-sulfate (C6S), and hyaluronic acid (HA) in the supernatant were evaluated. E, To compare the mRNA expression for chondrogenic genes, total RNA was prepared from 3 cell pellets from 1 young donor, and reverse transcription–polymerase chain reaction was performed. C4ST ⫽ chondroitin 4-sulfotransferase. derived cells and adipose synovium–derived cells appeared similar, in that they were small and spindle-shaped. In contrast, subcutaneous fat–derived cells were larger and flatter, especially when cultured at higher densities (Figure 1F). Epitope profile. Of the 15 antibodies examined, the rate of positivity for CD45 (hematopoietic cell marker), CD31 (endothelial cell marker), CD117 (C-kit, stem cell factor receptor), and CD34 (hematopoietic progenitor cell antigen) (3,6,10,17–20) was ⬍2%, and PROPERTIES OF HUMAN SYNOVIAL MESENCHYMAL STEM CELLS that for VEGFR-2 (Flk-1) (6,10,18) and for NGFR (6,10,18) was ⬍4% in fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells (Figure 2). However, rates of the STRO-1 (6,10,18,21) and CD106 (VCAM-1) positivity (3,6,10,22) in fibrous synovium– and adipose synovium–derived cells were ⬍6%, which was significantly higher than that in subcutaneous fat–derived cells. In contrast, the rate of CD10 positivity (17,18) in subcutaneous fat–derived cells was 40%, which was higher than that in fibrous synovium– and adipose synovium–derived cells, both of which had ⬃10% positivity for CD10. The rates of positivity for CD54 (ICAM-1) (3,6,10,23) and CD166 (ALCAM, SB10) (23) were 20–40%, that for CD90 (Thy1) (17) was 40–70%, that for CD105 (SH-2) (3,24) was 60–80%, that for CD44 (hyaluronan receptor) (3) was 70–90%, and that for CD147 (neuroregulin) (18) was ⬎90% in all 3 MSC populations. Proliferative potential. Initial cell density affected increases in proliferation of fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells (Figure 3A). In young donors, fibrous synovium–derived cells had higher proliferative potential than did adipose synovium–derived cells; however, the proliferative potential of the 2 populations was similar in elderly donors (Figure 3B). The proliferative potential was lowest in subcutaneous fat–derived cells, in young donors and in elderly donors. The colony-forming efficiency of subcutaneous fat–derived cells was significantly lower than that of fibrous synovium–derived cells, in young donors and in elderly donors (Figure 3C). Single-cell–derived cultures. To compare the cells and better define each population, clones of fibrous synovium–, adipose synovium–, and subcutaneous fat– derived cells were prepared. Both the magnitude of increase and the colony-forming efficiency of the fibrous synovium– and adipose synovium–derived cells were higher than those of the subcutaneous fat–derived cells in both young and elderly donors (Figures 3D and E). The results from single-cell–derived cultures generally appeared similar to those obtained in a mixed population of the cells. Chondrogenesis. To compare the chondrogenic potential of the MSC populations, cells were differentiated into cartilage in vitro. After 21 days of culture, the cell pellets became spherical (Figure 4A). During in vitro chondrogenesis, the pellets increased in size, which was attributable to production of extracellular matrix (12). Pellets from fibrous synovium– and adipose synovium–derived cells had greater amounts of cartilage matrix than did pellets from subcutaneous fat, as shown by staining with toluidine blue (Figure 4B). Pellets from 849 fibrous synovium– and adipose synovium–derived cells were also heavier than those from subcutaneous fat– derived cells (Figure 4C). Determination of GAGs in the extracellular matrix of the pellets showed that the amount of C4S, C6S, and hyaluronic acid differed among the cell pellets. Cells derived from fibrous synovium and adipose synovium exhibited greater amounts of GAGs than did those from subcutaneous fat (Figure 4D). RT-PCR demonstrated that expression of messenger RNA (mRNA) for the chondrogenic genes COL2A1, link protein, SOX6, SOX9, and C4ST in cell pellets derived from fibrous synovium and adipose synovium was higher than that in cells from subcutaneous fat (Figure 4E). Expression of mRNA for the other chondrogenic genes, aggrecan, decorin, biglycan, SOX5, and C6ST, appeared similar among the cell pellets from the 3 populations. These results indicate that fibrous synovium– and adipose synovium–derived cells had higher chondrogenic potential than did subcutaneous fat–derived cells. Osteogenesis. To evaluate the osteogenic potential of the MSC populations, cells were cultured in osteogenic medium. All cells were calcified and stained with alizarin red (Figure 5A). The ratios of alizarin red–positive (6,10) colonies in fibrous synovium– and adipose synovium–derived cells were higher than that in subcutaneous fat–derived cells (Figure 5B), indicating that there was a difference in osteogenic potential among these cells. Results in the 3 populations were similar between cells from young donors and those from elderly donors. Adipogenesis. To compare the adipogenic potential of the cells in the 3 populations, the ratio of oil red O–positive colonies to total colonies was evaluated (6,10,11,16). The oil red O–positive colony rate was similar among the 3 populations (Figures 6A and B). There were no differences between cells from young donors and those from elderly donors. DISCUSSION In this study, we compared fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells from the perspective of common properties of MSCs. In the 3 populations, epitope profiles of the cells were similar, in that the rate of positivity for CD34 and CD45 (hematopoietic cell markers) was low and the rate of positivity for CD44 (hyaluronan receptor) and CD105 (SH-2) was high. These results coincide with the phenotypic properties of bone marrow–derived MSCs (3,6,10,18,25). 850 MOCHIZUKI ET AL marrow (26) and has shown some promise for use in immunophenotyping MSCs (27). The rate of CD106 (VCAM-1) positivity in fibrous synovium– and adipose synovium–derived cells was 5%, which was higher than that in subcutaneous fat–derived cells (1%). VCAM-1 is Figure 5. Osteogenic potential of fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells. The cells were plated at 100 cells per 60-cm2 dish and cultured for 14 days to make cell colonies. The cells were then incubated in osteogenic medium for an additional 21 days. A, Calcified colonies stained with alizarin red are shown as red colonies. The same dishes were then stained with crystal violet to count total colony number. B, Analyses were performed in triplicate on cells from 3 young and 3 elderly donors, and the ratios of alizarin red–positive colonies to total colonies were calculated. Data were analyzed using two-factor analysis of variance to assess the effect of cell source and donor age; bars show the mean and SD. ⴱ ⫽ P ⬍ 0.05; ⴱⴱⴱ ⫽ P ⬍ 0.005. Interestingly, the rate of STRO-1 positivity in fibrous synovium– and adipose synovium–derived cells was higher than that in subcutaneous fat–derived cells. STRO-1 was originally reported to identify colonyforming osteogenic precursor cells isolated from bone Figure 6. Adipogenic potential of fibrous synovium–, adipose synovium–, and subcutaneous fat–derived cells. The cells were plated at 100 cells per 60-cm2 dish and cultured for 14 days to make cell colonies. The cells were then incubated in adipogenic medium for an additional 21 days. A, Adipocyte colonies stained with oil red O are shown as red colonies. The same dishes were then stained with crystal violet to count total colony number. B, Analyses were performed in triplicate on cells from 3 young and 3 elderly donors, and the ratios of oil red O–positive colonies to the total colonies were calculated. Data were analyzed using two-factor analysis of variance to assess the effect of cell source and donor age; bars show the mean and SD. PROPERTIES OF HUMAN SYNOVIAL MESENCHYMAL STEM CELLS a cell-surface glycoprotein that is produced by cytokineactivated endothelium, and is expressed primarily on lining layer cells in synovial tissue (28). In contrast, the rate of CD10 positivity in subcutaneous fat–derived cells was 40%, which was higher than that in fibrous synovium– and adipose synovium– derived cells (10%). CD10 is also known as common acute lymphocytic leukemia antigen or human membrane–associated neutral endopeptidase. Colter et al previously demonstrated that single-cell–derived colonies of bone marrow–derived MSCs contained 3 morphologically distinct cell types: large flat cells, small spindle-shaped cells, and extremely small, rapidly dividing cells, and showed that samples enriched for the small and extremely small cells had a greater ability for multipotential differentiation than did samples enriched for the large cells. Those authors found that CD10 was a negative marker for small and extremely small cells (18). In this study, fibrous synovium– and adipose synovium– derived cells expressed lower levels of CD10, showed increased proliferation, and had higher chondrogenic and osteogenic potential than did subcutaneous fat– derived cells, which seems to support the findings in bone marrow–derived MSCs reported by Colter et al. Several reports have described MSCs derived from human adipose tissue. To harvest the cells, great amounts of liposuction tissue were collected, digested with collagenase, separated by stromal–vascular fraction, and expanded (29–31). The processed lipoaspiratederived cells contaminate endothelial, smooth muscle, and pericyte cell populations (31). In our study, subcutaneous fat–derived cells lacked robust chondrogenic activity. We collected only an ⬃100-mg fat tissue, and after digestion, the cells were plated without gradient separation. The quantity of tissue and procedure for fractionation may explain the difference in properties observed in the adipose-derived cells in this study. MSCs derived from the infrapatellar fat pad (adipose synovium) have also been described (9,32). Dragoo et al (9) regarded infrapatellar fat pad–derived cells as adipose-derived cells; however, the infrapatellar fat pad is composed of synovium and subsynovial adipose tissues. In our study, the results from morphologic study of infrapatellar fat pad tissue, morphologic study of expanded cells, examination of surface epitopes, and studies of the proliferation, colony-forming efficiency, and chondrogenetic and osteogenetic potential indicate that adipose synovium–derived cells are more similar to fibrous synovium–derived cells than to subcutaneous fat–derived cells. We evaluated the properties of MSCs in the 3 populations in young and elderly donors separately, 851 because we had initial concerns that synovium-derived MSCs from elderly donors might lack the ability for expansion and differentiation. There are several reports describing the influence of aging on the properties of bone marrow–derived MSCs, and this topic remains controversial. Some studies have shown that aging does not affect colony-forming efficiency (33–35), adipogenesis (34), and calcification (33,34,36). In contrast, others have reported that aging affects the proliferative capacity at passage 1 (36) as well as the chondrogenic (36), osteogenic (35,37), and adipogenic (36) differentiation ability of bone marrow–derived MSCs. Our previous study indicated no obvious differences between bone marrow–derived MSCs from young and elderly donors in terms of the yields of cells at passage 0, the colony-forming efficiency at passage 0, surface-cell antigens, and chondrocyte, adipocyte, and osteoblast differentiation potentials. However, the proliferative ability of passage 1 cells decreased with age, with results observed as a decrease in cell numbers per colony (10). Apparently, this discrepancy can be attributable to the differences in donor status, site, differentiation protocol, and evaluation method. For example, D’Ippolito et al demonstrated that the number of MSCs with osteogenic potential decreased during aging. They collected bone marrow from vertebral bodies, switched to osteogenic medium 1 day after plating, and evaluated osteogenic potential as the ratios of alkaline phosphatase–positive colonies (35). We collected bone marrow from the proximal tibiae, switched osteogenic medium 14 days after plating, and evaluated osteogenic potential by the ratios of alizarin red–positive colonies (10). Oreffo et al demonstrated a significant decrease in the ratio of alkaline phosphatase–positive colonies in elderly donors with osteoporosis as compared with young donors, whereas there was no difference in this ratio between elderly donors with osteoarthritis and young donors (37). Murphy et al demonstrated a significant reduction in chondrogenic and adipogenic activity of bone marrow–derived MSCs from elderly osteoarthritis patients compared with those from younger donors (36). Their chondrogenic medium did not contain bone morphogenetic proteins, whereas ours included bone morphogenetic protein 2, which enhanced the in vitro chondrogenesis of MSCs (13). Furthermore, they evaluated chondrogenic potential by the amount of GAG standardized to DNA content, while we compared cartilage pellet weight. With regard to adipogenesis, Murphy et al plated MSCs at high density, differentiated the MSCs into adipocytes, and quantitated nile red fluorescence standardized to 4⬘,6-diamidino-2-phenylindole, 852 whereas we plated at low density, formed cell colonies, differentiated cells into adipocytes, and evaluated oil red O–positive colony rates. This study showed no remarkable differences between young and elderly donors in terms of the proliferative ability and colony-forming efficiency of the cells at passage 1, or the chondrocyte, osteoblast, and adipocyte differentiation potential in each MSC population derived from fibrous synovium, adipose synovium, and subcutaneous fat. Contrary to our initial predictions, the nucleated cell number per tissue weight in the fibrous synovium of elderly donors was larger than that in young donors. Revell et al reported that synovium from elderly patients with osteoarthritis is likely to be fibrous (38), and our results were consistent with this observation (as shown in Figure 1C). Osteoarthritis comprises a common, age-related heterogeneous group of disorders that are characterized pathologically by focal areas of loss of articular cartilage in synovial joints, associated with varying degrees of osteophyte formation, subchondral bone changes, and synovitis (39). The secondary inflammation in the synovium could alter its cellular composition and could be responsible for changes to the stem cell populations. Detailed pathologic investigation of synovium will be important for clarifying the role of the disease and the role of aging in the characteristics of stem cells derived from the synovium. This general information will be valuable for comparison with other studies. An age-related decrease in the chondrogenic differentiation potential has been reported in rabbit fibrous synovium (40) and periosteum (41) in an ex vivo organ culture (42). In contrast, De Bari et al demonstrated that the chondrogenic potential of synoviumderived cells was independent of donor age (7), which is similar to our result. These findings suggest that aging may affect the chondrogenic differentiation potential in organ culture but does not affect the cells expanded in vitro. We demonstrated that both fibrous synovium– and adipose synovium–derived MSCs had a better chondrogenic capacity than did subcutaneous fat–derived MSCs. Given this difference, important biologic questions are raised, and we have developed 2 hypotheses: 1) The observed differences between synovium and subcutaneous fat may be due to differences in the number of ancestral MSCs, or alternatively, 2) the observed differences in chondrogenesis could have arisen as a result of different MSC propensities to follow a chondrogenic pathway, suggesting that the local tissue microenvironment may be directing the “fate” of the MSCs toward a particular lineage. MOCHIZUKI ET AL We also demonstrated that synovium-derived MSCs had a higher osteogenetic ability than did adiposederived MSCs. Furthermore, single-cell–derived cultures as well as mixed synovial cells showed higher colonyforming efficiency and expansion ability than did those from adipose tissue. These findings may implicate the role of the local tissue microenvironment in directing the “fate” of the MSCs. The adipogenic capacity of MSCs derived from the synovium and those derived from the subcutaneous fat was similar, which would support the first hypothesis described above. However, MSCs derived from adipose tissue could have been preconditioned via the microenvironment, and thus slightly predisposed toward an adipocyte lineage, which would be consistent with the second hypothesis. The important consideration in tissue engineering is to harvest the greatest amount of MSCs with the highest potential while minimizing the amounts of mesenchymal tissues needed, resulting in less-invasive treatments. Fibrous synovium– and adipose synovium– derived MSCs were similar in terms of their cell morphologic features, epitope profiles, colony-forming efficiency, chondrogenesis, osteogenesis, and adipogenesis potentials. The nucleated cell number per tissue weight was higher in fibrous synovium than in adipose synovium, which may be an advantage of fibrous synovium. However, adipose synovium cells also have an advantage due to their high chondrogenic potential and accessibility, in that sufficient amounts of adipose synovium can be harvested with possibly fewer complications. We therefore conclude that both fibrous synovium and adipose synovium are suitable MSC sources for cartilage regeneration. ACKNOWLEDGMENTS We thank Kenichi Shinomiya, MD, PhD, for continuous support, Kazuyoshi Yagishita, MD, PhD, for sample collection, Izumi Nakagawa for excellent technical assistance, Miyoko Ojima for expert help with histology, Kyosuke Miyazaki for analysis of chondroitin sulfate, and Benjamin L. Larson for proofreading. REFERENCES 1. Brittberg M, Lindahl A, Nilsson A, Ohlsson C, Isaksson O, Peterson L. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med 1994;331: 889–95. 2. Prockop DJ. Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 1997;276:71–4. 3. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD, et al. Multilineage potential of adult human mesenchymal stem cells. Science 1999;284:143–7. PROPERTIES OF HUMAN SYNOVIAL MESENCHYMAL STEM CELLS 4. Friedenstein AJ. Precursor cells of mechanocytes. Int Rev Cytol 1976;47:327–59. 5. McKay R. Stem cells in the central nervous system. Science 1997;276:66–71. 6. Sakaguchi Y, Sekiya I, Yagishita K, Muneta T. Comparison of human stem cells derived from various mesenchymal tissues: superiority of synovium as a cell source. Arthritis Rheum 2005;52: 2521–9. 7. De Bari C, Dell’Accio F, Tylzanowski P, Luyten FP. Multipotent mesenchymal stem cells from adult human synovial membrane. Arthritis Rheum 2001;44:1928–42. 8. De Bari C, Dell’Accio F, Vandenabeele F, Vermeesch JR, Raymackers JM, Luyten FP. Skeletal muscle repair by adult human mesenchymal stem cells from synovial membrane. J Cell Biol 2003;160:909–18. 9. Dragoo JL, Samimi B, Zhu M, Hame SL, Thomas BJ, Lieberman JR, et al. Tissue-engineered cartilage and bone using stem cells from human infrapatellar fat pads. J Bone Joint Surg Br 2003;85: 740–7. 10. Sakaguchi Y, Sekiya I, Yagishita K, Ichinose S, Shinomiya K, Muneta T. Suspended cells from trabecular bone by collagenase digestion become virtually identical to mesenchymal stem cells obtained from marrow aspirates. Blood 2004;104:2728–35. 11. Sekiya I, Larson BL, Smith JR, Pochampally R, Cui JG, Prockop DJ. Expansion of human adult stem cells from bone marrow stroma: conditions that maximize the yields of early progenitors and evaluate their quality. Stem Cells 2002;20:530–41. 12. Sekiya I, Vuoristo JT, Larson BL, Prockop DJ. In vitro cartilage formation by human adult stem cells from bone marrow stroma defines the sequence of cellular and molecular events during chondrogenesis. Proc Natl Acad Sci U S A 2002;99:4397–402. 13. Sekiya IL, Vuoristo JT, Reger RL, Prockop DJ. Comparison of effect of BMP-2, -4, and -6 on in vitro cartilage formation of human adult stem cells from bone marrow stroma. Cell Tissue Res 2005;320:269–76. 14. Yokoyama A, Sekiya I, Miyazaki K, Ichinose S, Hata Y, Muneta T. In vitro cartilage formation of composites of synovium-derived mesenchymal stem cells with collagen gel. Cell Tissue Res 2005; 322:289–98. 15. Shinmei M, Miyauchi S, Machida A, Miyazaki K. Quantitation of chondroitin 4–sulfate and chondroitin 6–sulfate in pathologic joint fluid. Arthritis Rheum 1992;35:1304–8. 16. Sekiya I, Larson BL, Vuoristo JT, Cui JG, Prockop DJ. Adipogenic differentiation of human adult stem cells from bone marrow stroma (MSCs). J Bone Miner Res 2004;19:256–64. 17. Jones EA, Kinsey SE, English A, Jones RA, Straszynski L, Meredith DM, et al. Isolation and characterization of bone marrow multipotential mesenchymal progenitor cells. Arthritis Rheum 2002;46:3349–60. 18. Colter DC, Sekiya I, Prockop DJ. Identification of a subpopulation of rapidly self-renewing and multipotential adult stem cells in colonies of human marrow stromal cells. Proc Natl Acad Sci U S A 2001;98:7841–5. 19. Kawada H, Fujita J, Kinjo K, Matsuzaki Y, Tsuma M, Miyatake H, et al. Non-hematopoietic mesenchymal stem cells can be mobilized and differentiate into cardiomyocytes after myocardial infarction. Blood 2004;104:3581–7. 20. Okuno Y, Iwasaki H, Huettner CS, Radomska HS, Gonzalez DA, Tenen DG, et al. Differential regulation of the human and murine CD34 genes in hematopoietic stem cells. Proc Natl Acad Sci U S A 2002;99:6246–51. 21. Stewart K, Walsh S, Screen J, Jefferiss CM, Chainey J, Jordan GR, et al. Further characterization of cells expressing STRO-1 in cultures of adult human bone marrow stromal cells. J Bone Miner Res 1999;14:1345–56. 22. Gronthos S, Zannettino AC, Hay SJ, Shi S, Graves SE, Kortesidis A, et al. Molecular and cellular characterisation of highly purified stromal stem cells derived from human bone marrow. J Cell Sci 2003;116:1827–35. 853 23. Arai F, Ohneda O, Miyamoto T, Zhang XQ, Suda T. Mesenchymal stem cells in perichondrium express activated leukocyte cell adhesion molecule and participate in bone marrow formation. J Exp Med 2002;195:1549–63. 24. Barry FP, Boynton RE, Haynesworth S, Murphy JM, Zaia J. The monoclonal antibody SH-2, raised against human mesenchymal stem cells, recognizes an epitope on endoglin (CD105). Biochem Biophys Res Commun 1999;265:134–9. 25. Koc ON, Gerson SL, Cooper BW, Dyhouse SM, Haynesworth SE, Caplan AI, et al. Rapid hematopoietic recovery after coinfusion of autologous-blood stem cells and culture-expanded marrow mesenchymal stem cells in advanced breast cancer patients receiving high-dose chemotherapy. J Clin Oncol 2000;18:307–16. 26. Gronthos S, Zannettino AC, Graves SE, Ohta S, Hay SJ, Simmons PJ. Differential cell surface expression of the STRO-1 and alkaline phosphatase antigens on discrete developmental stages in primary cultures of human bone cells. J Bone Miner Res 1999;14:47–56. 27. Gronthos S, Graves SE, Ohta S, Simmons PJ. The STRO-1⫹ fraction of adult human bone marrow contains the osteogenic precursors. Blood 1994;84:4164–73. 28. Tak PP, Thurkow EW, Daha MR, Kluin PM, Smeets TJ, Meinders AE, et al. Expression of adhesion molecules in early rheumatoid synovial tissue. Clin Immunol Immunopathol 1995;77:236–42. 29. Gronthos S, Franklin DM, Leddy HA, Robey PG, Storms RW, Gimble JM. Surface protein characterization of human adipose tissue-derived stromal cells. J Cell Physiol 2001;189:54–63. 30. Zuk PA, Zhu M, Ashjian P, de Ugarte DA, Huang JI, Mizuno H, et al. Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell 2002;13:4279–95. 31. Zuk PA, Zhu M, Mizuno H, Huang J, Futrell JW, Katz AJ, et al. Multilineage cells from human adipose tissue: implications for cell-based therapies. Tissue Eng 2001;7:211–28. 32. Wickham MQ, Erickson GR, Gimble JM, Vail TP, Guilak F. Multipotent stromal cells derived from the infrapatellar fat pad of the knee. Clin Orthop Relat Res 2003;196–212. 33. Stenderup K, Justesen J, Eriksen EF, Rattan SI, Kassem M. Number and proliferative capacity of osteogenic stem cells are maintained during aging and in patients with osteoporosis. J Bone Miner Res 2001;16:1120–9. 34. Justesen J, Stenderup K, Eriksen EF, Kassem M. Maintenance of osteoblastic and adipocytic differentiation potential with age and osteoporosis in human marrow stromal cell cultures. Calcif Tissue Int 2002;71:36–44. 35. D’Ippolito G, Schiller PC, Ricordi C, Roos BA, Howard GA. Age-related osteogenic potential of mesenchymal stromal stem cells from human vertebral bone marrow. J Bone Miner Res 1999;14:1115–22. 36. Murphy JM, Dixon K, Beck S, Fabian D, Feldman A, Barry F. Reduced chondrogenic and adipogenic activity of mesenchymal stem cells from patients with advanced osteoarthritis. Arthritis Rheum 2002;46:704–13. 37. Oreffo RO, Bord S, Triffitt JT. Skeletal progenitor cells and ageing human populations. Clin Sci (Lond) 1998;94:549–55. 38. Revell PA, Mayston V, Lalor P, Mapp P. The synovial membrane in osteoarthritis: a histological study including the characterisation of the cellular infiltrate present in inflammatory osteoarthritis using monoclonal antibodies. Ann Rheum Dis 1988;47:300–7. 39. Dieppe PA, Lohmander LS. Pathogenesis and management of pain in osteoarthritis. Lancet 2005;365:965–73. 40. Nishimura K, Solchaga LA, Caplan AI, Yoo JU, Goldberg VM, Johnstone B. Chondroprogenitor cells of synovial tissue. Arthritis Rheum 1999;42:2631–7. 41. O’Driscoll SW, Saris DB, Ito Y, Fitzimmons JS. The chondrogenic potential of periosteum decreases with age. J Orthop Res 2001; 19:95–103. 42. O’Driscoll SW, Recklies AD, Poole AR. Chondrogenesis in periosteal explants: an organ culture model for in vitro study. J Bone Joint Surg Am 1994;76:1042–51.