Embed Size (px)
Citation preview
OsteoArthritis and Cartilage (2007) 15, 1178e1189
ª 2007 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.joca.2007.03.015
Chondrogenic potential of human synovial mesenchymalstem cells in alginateT. Kurth M.Sc.y, E. Hedbom Ph.D.y, N. Shintani Ph.D.y, M. Sugimoto M.D.y, F. H. Chen Ph.D.z,M. Haspl M.D., Ph.D.x, S. Martinovic Ph.D.k and E. B. Hunziker M.D.y*y ITI Research Institute for Dental and Skeletal Biology, University of Bern, Bern, SwitzerlandzCartilage Biology and Orthopaedics Branch, National Institute of Arthritis andMusculoskeletal and Skin Diseases, National Institute of Health, Bethesda, MD, USAxDepartment of Orthopaedic Surgery, School of Medicine, Zagreb, CroatiakDepartment of Anatomy, School of Medicine, Zagreb, Croatia
Summary
Objective: In a recent study, we demonstrated that mesenchymal stem cells (MSCs) derived from the synovial membranes of bovine shoulderjoints could differentiate into chondrocytes when cultured in alginate. The purpose of the present study was to establish the conditions underwhich synovial MSCs derived from aging human donors can be induced to undergo chondrogenic differentiation using the same alginatesystem.
Methods: MSCs were obtained by digesting the knee-joint synovial membranes of osteoarthritic human donors (aged 59e76 years), and ex-panded in monolayer cultures. The cells were then seeded at a numerical density of 4� 106/ml within discs of 2% alginate, which were culturedin serum-containing or serum-free medium (the latter being supplemented with 1% insulin, transferrin, selenium (ITS). The chondrogenicdifferentiation capacity of the cells was tested by exposing them to the morphogens transforming growth factor-beta1 (TGF-b1), TGF-b2,TGF-b3, insulin-like growth factor-1 (IGF-1), bone morphogenetic protein-2 (BMP-2) and BMP-7, as well as to the synthetic glucocorticoid dexa-methasone. The relative mRNA levels of collagen types I and II, of aggrecan and of Sox9 were determined quantitatively by the real-timepolymerase chain reaction (PCR). The extracellular deposition of proteoglycans was evaluated histologically after staining with ToluidineBlue, and that of type-II collagen by immunohistochemistry.
Results: BMP-2 induced the chondrogenic differentiation of human synovial MSCs in a dose-dependent manner. The response elicited byBMP-7 was comparable. Both of these agents were more potent than TGF-b1. A higher level of BMP-2-induced chondrogenic differentiationwas achieved in the absence than in the presence of serum. In the presence of dexamethasone, the BMP-2-induced expression of mRNAs foraggrecan and type-II collagen was suppressed; the weaker TGF-b1-induced expression of these chondrogenic markers was not obviouslyaffected.
Conclusions: We have demonstrated that synovial MSCs derived from the knee joints of aging human donors possess chondrogenic potential.Under serum-free culturing conditions and in the absence of dexamethasone, BMP-2 and BMP-7 were the most potent inducers of this trans-formation process.ª 2007 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.
Key words: Synovial cells, Chondrogenesis, BMP-2, BMP-7, TGF-b1.
InternationalCartilageRepairSociety
Introduction
The destruction of articular cartilage tissue that is associ-ated with degenerative joint diseases and traumatic injuriesis not only painful and debilitating, but also a costly medicalproblem. Structural articular cartilage defects are commonlyencountered in osteoarthritic patients, and there existsa great clinical need to induce their healing, which wouldbring symptomatic relief, improve functionality and delayor even arrest the progression of the disease. Several treat-ment strategies have been developed to enhance the repair
*Address correspondence and reprint requests to: Prof. Ernst B.Hunziker, ITI Research Institute for Dental and Skeletal Biology,University of Bern, Murtenstrasse 35, P.O. Box 54, 3010 Bern,Switzerland. Tel: 41-31-632-8685; Fax: 41-31-632-4955; E-mail:[email protected]
Received 26 October 2006; revision accepted 16 March 2007.
11
of articular cartilage, such as the grafting of autologous os-teochondral tissue or the transplantation of autologouschondrocyte suspensions1,2. However, the biological andmechanical properties of the repair tissue formed are infe-rior to those of native articular cartilage. The difficulty stemsfrom the fact that articular cartilage, unlike other tissuessuch as bone or tendon, has a very poor capacity for self-regeneration.
A relatively new and promising therapeutic approach in-volves the use of mesenchymal stem cells (MSCs). Thesemultipotent progenitor cells are capable of differentiatinginto bone-marrow stromal cells, osteoblasts, chondrocytes,myocytes and adipocytes3. In some tissue types, such asbone-marrow stroma, fat, skeletal muscle and synovium,MSCs persist in adult life without losing their capacity to pro-liferate and differentiate. Under appropriate conditions, andirrespective of their topographic location in the body or thetissue of origin, they can multiply and transform into special-ized cell types4e7.
78
1179Osteoarthritis and Cartilage Vol. 15, No. 10
Hitherto, most of the published studies relating to chon-drogenic differentiation have focused on MSCs isolatedfrom the bone-marrow stroma8e12. But MSCs derivedfrom the synovium are now attracting considerable atten-tion, since they apparently have great chondrogenic poten-tial. Under pathological conditions, chondromatomas havebeen observed to develop within synovial tissue13. And inexperimental animal models, synovial cells with stem-cellpotential can be stimulated to migrate from the synovialmembrane into articular cartilage defects, where they prolif-erate and transform into chondrocytes which produce acartilage-like repair tissue14,15. In pellet-culture models ofchondrogenesis, stem cells of synovial origin have beenshown to manifest a greater chondrogenic potential thanthose derived from bone marrow6,16. However, the stimula-tion conditions need to be improved to optimize the forma-tion of a fully functional, hyaline type of articular cartilage.
It was the purpose of the present study to assess thechondrogenic potential of MSCs derived from the synovialtissue of osteoarthritic human patients. After monolayerexpansion, the synovial cells were cultured in alginate andexposed to various growth factors that were expected toinduce and sustain chondrogenic differentiation. Theseincluded several members of the TGF-b superfamily17,namely, TGF-b1, TGF-b2, TGF-b3, BMP-2 and BMP-7, aswell as IGF-1. The influence of culturing in the absence orpresence of 10% fetal bovine serum (FBS) and dexameth-asone was also examined. Chondrogenic transformationwas evaluated histologically (cell morphology and meta-chromatic staining of the extracellular matrix with ToluidineBlue), immunohistochemically (type-II collagen), and at thegene-expression level [quantification of mRNAs for Sox9,aggrecan and type-II collagen by the real-time polymerasechain reaction (PCR)].
Materials and methods
TISSUE PREPARATION AND CELL CULTURING
Biopsies of synovial tissue and articular cartilage werederived from human donors (aged 59e76 years) who wereundergoing knee-joint surgery at the Department of Ortho-paedic Surgery, University of Zagreb, Croatia. The materialwas obtained with the informed consent of the patients andwith the permission of the local ethical commission. Initially,the synovial tissue was coarsely minced with a scalpel.Thereafter, it was exposed first to 1% pronase (Roche Diag-nostics AG, Rotkreuz, Switzerland) for 1 h at 37�C, and thento overnight digestion with 0.14% collagenase (Invitrogen,Basel, Switzerland), likewise at 37�C. The tissue digestwas filtered sequentially through nylon meshes with porediameters of 120 mm and 20 mm, and the cells were pelletedby centrifugation at 2000 rpm for 5 min. After resuspensionin high-glucose Dulbecco’s Modified Eagle’s Medium[DMEM (Invitrogen)] containing 10% heat-inactivated FBS(SigmaeAldrich, Buchs, Switzerland), 0.35 mM proline andgentamycin (50 mg/ml), the synovial cells were plated on100-mm-diameter cell-culture dishes (Falcon, Bedford,MA, USA) for monolayer expansion of the mesenchymalpopulation18. The cells were passaged once, trypsinized(0.25% trypsin in 1 mM ethylenediaminetetraacetic acid(EDTA), and then seeded in alginate, as previously de-scribed19. Briefly, the pelleted cells were resuspended ata numerical density of 4� 106/ml within a 2% solution ofalginate in 0.9% NaCl. The alginate was a 1:1 mixture ofKeltone LV alginate (Kelko Co, San Diego, CA, USA) andPronova LVG (Pronova Biocare, Oslo, Norway). The
alginate/cell suspension (2 ml) was transferred to a cus-tom-designed mold (22 mm in diameter and 2 mm in height),which was then submerged in 50 mM CaCl2 for 30 min. Thegel formed was rinsed twice for 5-min periods in 1.8 mMCaCl2. The cast samples were stamped with a sterile dermalpunch to yield alginate discs that were 5.5 mm in diameterand 2 mm in height. These discs were cultured for 2e6weeks at 37�C in a humidified atmosphere containing 5%CO2 under the conditions defined below.
Alginate constructs containing chondrocytes that wereisolated from the articular cartilage biopsies were also pre-pared. However, this material could not be obtained from allof the donors. Hence, it could be used as a reference only insome of the experiments. Freshly-isolated chondrocyteswere not subjected to monolayer expansion. Alginate discscontaining these cells were cultured in high-glucose DMEMcontaining 10% FBS, 0.35 mM proline and gentamycin(50 mg/ml) for 2 weeks at 37�C in a humidified atmospherecontaining 5% CO2.
According to the experimental design, alginate discs con-taining synovial cells (n¼ 3e6 donors per experimentalgroup) were cultured in high-glucose DMEM, which was sup-plemented with 0.35 mM proline and gentamycin (50 mg/ml),in the presence either of 10% FBS or of 1% insulin, transferrin,selenium, linoleic acid and bovine serum albumin (ITSþ�Premix) (Becton Dickinson, Bedford, MA, USA). Dexametha-sone (10�7 M) [SigmaeAldrich] was included in some exper-iments. The growth factors tested included TGF-b1 (10 or50 ng/ml), TGF-b2 (10 ng/ml), TGF-b3 (10 ng/ml), IGF-1(25 ng/ml) [all from SigmaeAldrich], BMP-2 (10, 50 or200 ng/ml) [Wyeth, Cambridge, MA, USA] and BMP-7(200 ng/ml) [Stryker Biotech, Hopkinton, MA, USA].
The medium was changed every 2e3 days. Ascorbic acid(25 mg/ml) was administered on a daily basis.
ISOLATION OF RNA, REVERSE TRANSCRIPTION
AND REAL-TIME PCR
The alginate discs were dissolved in 55 mM sodium citratecontaining 0.9% NaCl for 30 min at 37�C. The cells were pel-leted by centrifugation at 2000 rpm for 5 min, lysed in 350 mlof lysis buffer RLT (QIAGEN, Basel, Switzerland), vortexedbriefly and frozen overnight at �80�C. The lysates werethawed for 30 min at 37�C and then applied to QIAshreddercolumns (QIAGEN). The RNA was isolated using theRNeasy Micro Kit (QIAGEN) in accordance with the manu-facturer’s protocol, which included a 15-min DNase diges-tion step. The purity and the yield of the isolated RNAwere monitored using a NanoDrop spectrophotometer ND1000 (NanoDrop Technologies, Wilmington, DE, USA).The samples were stored at �80�C until they were requiredfor reverse transcription.
Total RNA (200 ng) was reverse transcribed using the Im-Prom-II� Reverse Transcription System (Promega, Madi-son, WI, USA). Reactions were conducted in duplicate tocompensate for variations in efficiency20. cDNA was syn-thesized at a final MgCl2 concentration of 7 mM using ran-dom hexamers, in accordance with the manufacturer’sprotocol. The samples were diluted 1:10 in nuclease-freewater and stored at �20�C prior to real-time PCR.
Each cDNA sample (1 ng) was utilized for the relativequantification of mRNA by real-time PCR using an ABI Prism7700 Sequence Detection System (PerkineElmer AppliedBiosystems, Foster City, CA, USA). The primers andprobes that were utilized are listed in Table I. They weredesigned using Primer Express software (PerkineElmer)and purchased as qPCR MasterMix Plus (Eurogentec,
1180 T. Kurth et al.: Chondrogenic potential of human synovial MSCs in alginate
Table IPrimers and probes used for real-time PCR
Gene Sequence Final concentration (nM) Amplicon size Accession no.
Type-I collagenForward CATGCCGTGACTTGAGACTCA 50 86 AF004877Probe* CCACCCAGAATGGAGCAG(T/C)GGTTACTACT 100Reverse GGCTTCCATAGTGCATCCTTGGT 900
Type-II collagenForward GGCAATAGCAGGTTCACGTACA 900 79 BC007252Probe* CCGGT(A/G)TGTTTCGTGCAGCCATCCT 100Reverse CGATAACAGTCTTGCCCCACTT 300
AggrecanForward CTACCGCTGCGAGGTGATG 900 74 NM_001135Probe* ATGGAACACGATGCCTTT(C/T)ACCACGA 100Reverse TCGAGGGTGTAGCGTGTAGAGA 900
Sox9Forward ACGCCGAGCTCAGCAAGA 900 71 NM_000346Probe* CGTTCAG(T/G)AGTCTCCAGAGCTTGCCCA 100Reverse CACGAAGGGCCGCTTCT 300
18s rRNAForward CGGCTACCACATCCAAGGAA 900 187 X03205Probe* TGCTGGCACCAGACTTGCCCTC 100Reverse GCTGGAATTACCGCGGCT 50
*Sequences selected on the basis of mismatches between humans and cows.
Seraing, Belgium). The probes were labeled with 6-carboxy-fluorescein (FAM) at the 50 end and with Eclipse� DarkQuencher at the 30 end. Oligonucleotides were used at thegiven final concentrations (Table I) in a volume of 25 ml.The PCR entailed an initial 10-min denaturation at 95�C,followed by 45 cycles of a 15-s denaturation at 95�Cand a 60-s extension at 60�C. The data were analyzedusing Sequence Detection Software (PerkineElmer). ThemRNA levels were quantified relative to 18s rRNA usingthe cycle threshold values (Ct) and by applying the compar-ative Ct method (DDCt), where x¼ 2�DDCt, DDCt¼DE�DC, DE¼Ct experimental�Ct 18s rRNA (experimental) andDC¼Ct control�Ct 18s rRNA (control)
21. The data were statisti-cally analyzed on the basis of a Pair Wise Fixed ReallocationRandomisation Testª using the Relative Expression Soft-ware Tool22. Statistical significance was set at P< 0.05.
HISTOCHEMISTRY AND IMMUNOHISTOCHEMISTRY
Alginate discs were chemically fixed in 4% paraformalde-hyde solution containing 10 mM CaCl2 and 0.1 M sodium ca-codylate (pH 7.4) for 4 h at ambient temperature. They werethen washed overnight at 4�C in 0.1 M sodium cacodylate(pH 7.4) containing 50 mM BaCl2 before dehydrating inethanol, embedding in paraffin, and sectioning.
To demonstrate the synthesis of a proteoglycan-rich ex-tracellular matrix, 5-mm-thick sections were deparaffinized,rehydrated, and stained with 0.1% Toluidine Blue O (pH2.5) for 30 s. After dehydration, the sections were mountedin Entellan (Merck, Darmstadt, Germany).
Likewise, 5-mm-thick sections were used for the immuno-histochemical analysis. Endogenous peroxidase activitywas quenched by treatment with 0.3% hydrogen peroxidein phosphate-buffered saline. Antigen retrieval wasachieved by incubating the sections with hyaluronidase(SigmaeAldrich) [1 mg/ml of phosphate-buffered saline(pH 5.0)] for 30 min at 37�C. Non-specific backgroundreactivity was blocked with 10% normal goat serum inTris-buffered saline (pH 7.4). The sections were incubatedwith a primary antibody against type-II collagen [(II-II6B3;
Developmental Studies Hybridoma Bank, Iowa City, IA,USA) at a dilution of 1:10 (in Tris-buffered saline containing10% normal goat serum and 0.1% Tween 20)] for 1 h at37�C in a humid chamber. After rinsing, the sections wereexposed first to a broad-spectrum secondary antibody[(Histostain-SP kit, Zymed Laboratories, South San Fran-cisco, CA) for 1 h at ambient temperature, and then toa streptavidineperoxidase conjugate for 15 min. Diamino-benzidine (Zymed Laboratories, South San Francisco,CA) was used as the chromogen for visualization. The sec-tions were dehydrated and mounted in Permount prior toinspection in a Vanox AH-2 microscope (Olympus,Hamburg, Germany).
Results
EFFECTS OF THE CULTURING SYSTEM
Compared to articular cartilage chondrocytes that hadbeen cultured in alginate for 2 weeks, freshly-isolated syno-vial cells expressed very low levels of the mRNAs for type-IIcollagen and aggrecan. After monolayer expansion of thesynovial cells for 2 weeks, expression of the type-II-collagengene remained low, whereas the mRNA level of aggrecanincreased to about 12% of that in chondrocytes. Three-dimensional culturing of synovial cells in alginate for 2weeks resulted in an up-regulation of both genes: themRNA levels of type-II collagen and aggrecan increasedto 7% and 64%, respectively, of those in chondrocytes(Fig. 1). Expression of the type-II-collagen gene increasedas a function of culturing time, with a dramatic rise betweenthe third and the fourth weeks. Differences in the kineticsand in the level of expression of the type-II-collagen genewere observed between the individual donors (Fig. 2).
GROWTH FACTORS
In an endeavor to up-regulate the expression of the chon-drogenic differentiation markers type-II collagen and aggre-can, particularly the former, alginate-cultured synovial cells
1181Osteoarthritis and Cartilage Vol. 15, No. 10
grown in the presence of 10% FBS for 3 weeks were ex-posed during the third week to growth factors that areknown to induce the chondrogenic differentiation of MSCsin various systems. Neither TGF-b1, TGF-b2, TGF-b3(each at 10 ng/ml) nor IGF-1 (25 ng/ml) significantly influ-enced the mRNA level of aggrecan [Fig. 3(a)]. IGF-1 likewisehad no effect on the mRNA level of type-II collagen. But theTGF-bs induced a dramatic down-regulation of the lattergene. An up-regulation of both markers above the serum con-trol level was achieved only with BMP-2 (200 ng/ml). The con-centration dependency of the effects elicited by TGF-b1 andBMP-2 was then investigated. Increasing the concentrationof TGF-b1 to 50 ng/ml did not heighten the gene-expressionlevels of either type-II collagen or aggrecan above thoseobserved at 10 ng/ml [Fig. 3(b)]. On the other hand, BMP-2stimulated the expression of each gene in a dose-dependentmanner between 50 and 200 ng/ml [Fig. 3(b)]. Hence, BMP-2can consistently stimulate the chondrogenic differentiation ofhuman synovial cells.
Using BMP-2, we then investigated the effects of differentmodes of exposure to this morphogen on the chondrogenicdifferentiation of synovial cells during a 4-week culturing pe-riod. In the first experiment, BMP-2 (200 ng/ml) was presentduring only the first and the second weeks of culturing; inthe second experiment, it was present during only the thirdand the fourth weeks. Compared to control cultures (totalabsence of BMP-2), the type-II-collagen gene was ex-pressed at higher levels when BMP-2 was present during
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
In
du
ctio
n fo
ld
(m
RN
A level)
Type-II collagenAggrecan
Freshlyisolated
2 weeks asmonolayers
2 weeksin alginate
Chondrocytes2 weeks
in alginateSynovial cells
Fig. 1. Expression levels of the type-II-collagen and the aggrecangenes in synovial cells and articular cartilage chondrocytes derivedfrom the knee joints of osteoarthritic human patients. Data pertain-ing to synovial cells were gleaned immediately after their isolation,after 2 weeks of culturing as monolayers, and after 2 weeks ofgrowth in alginate. Data pertaining to articular cartilage chondro-cytes were gleaned 2 weeks after culturing in alginate. ThemRNA levels for synovial cells are expressed relative to those inchondrocytes that had been cultured in alginate for 2 weeks (valueset at ‘‘1’’ for each chondrogenic differentiation marker). In eachinstance, the culture medium was supplemented with 10% FBS.Data are expressed as mean values together with the standard
error of the mean (n¼ 4 donors).
only the first and the second weeks of culturing thanwhen it was present during only the third and the fourthweeks (Fig. 4).
We also wished to ascertain whether the presence of se-rum (FBS) influenced the chondrogenic differentiation of al-ginate-cultured synovial cells. In these experiments, theserum-free medium was supplemented with 1% ITS, andthe synovial cells were cultured for 2 weeks. Articular carti-lage chondrocytes, likewise cultured in alginate for 2 weeks(in the presence of 10% FBS and in the absence of BMP-2),served as a reference standard. Synovial cells cultured for 2weeks in alginate expressed higher basal levels of bothtype-II collagen and aggrecan in the presence than in theabsence of serum. Stimulation with BMP-2 led to an up-regulation of each chondrogenic differentiation marker,both in the absence and in the presence of serum, but themRNA levels were significantly higher in the formerinstance (Fig. 5). A similar trend was observed for the chon-drogenic transcription factor Sox9, although the differenceswere less pronounced than for aggrecan and type-II colla-gen. The gene for type-I collagen was marginally up-regu-lated under serum-free conditions, both in resting and inBMP-2-stimulated cells. In all synovial cultures, the mRNAlevels of type-I collagen were much higher than in articularcartilage chondrocytes (Fig. 5).
Morphologically, BMP-2-stimulated synovial cells that hadbeen cultured for 2 weeks in the absence of serum [Fig. 6(a)]were still small and round. At this stage, only a few larger,chondrocyte-like cells were encountered. These were sur-rounded by a ring of metachromatically-stained extracellularmatrix [Fig. 6(d)], which was immunoreactive for type-II col-lagen [Fig. 6(g)]. After 4 weeks of culturing, clusters of cellswere observed [Fig. 6(b)]. The cells within these clusterswere surrounded by an abundant, metachromatically-stained matrix [Fig. 6(e)], which was immunoreactive fortype-II collagen [Fig. 6(h)]. By the sixth week, the number
0
5
10
15
20
25
30
35
40
2w 3w 4w
In
du
ctio
n fo
ld
(m
RN
A level)
Type-II collagen
Culturing time
(in weeks)
Fig. 2. mRNA levels of the type-II-collagen gene in synovial cellsthat were derived from six patients, represented individually foreach donor as a function of culturing time in alginate. The expres-sion levels at 3 and 4 weeks are represented relative to those at 2
weeks. The culture medium was supplemented with 10% FBS.
1182 T. Kurth et al.: Chondrogenic potential of human synovial MSCs in alginate
b
a
0
0.5
1
1.5
2
2.5
3
3.5
4
In
du
ctio
n fo
ld
(m
RN
A level)
Aggrecan
TGF-
β1
Con
trol
BMP-
220
0 ng
/ml
IGF-
125
ng/
ml
TGF-
β2
TGF-
β3
00.5
11.5
22.5
33.5
4
In
du
ctio
n fo
ld
(m
RN
A level)
TGF-
β1
Con
trol
BMP-
220
0 ng
/ml
IGF-
125
ng/
ml
TGF-
β2
TGF-
β3
10 ng/ml
4.55 Type-II collagen
10 ng/ml
0
1
2
3
4
5
6
In
du
ctio
n fo
ld
(m
RN
A level)
Aggrecan
0
5
10
15
20
25
30
35 Type-II collagen
Con
trol
10 n
g/m
l
50 n
g/m
l
10 n
g/m
l
50 n
g/m
l
200
ng/m
l
BMP-2
In
du
ctio
n fo
ld
(m
RN
A level)
Con
trol
10 n
g/m
l
50 n
g/m
l
10 n
g/m
l
50 n
g/m
l
200
ng/m
l
TGF-β1 TGF-β1 BMP-2
∗
∗ ∗ ∗
∗
∗
∗ ∗
∗
∗
Fig. 3. Effects of various growth factors at a defined concentration (a), and of TGF-b1 and BMP-2 at different concentrations (b), on the relativegene-expression levels of aggrecan and type-II collagen in synovial cells that were derived from three patients. Data pertain to synovial cellsthat were cultured for 3 weeks in alginate. The culture medium was supplemented with 10% FBS. In each case, the growth factor was intro-duced at the onset of the third week of culturing. Data are represented as mean values together with the standard error of the mean (n¼ 3
donors). )P< 0.05 compared to the control.
and the size of these cell clusters had increased [Fig. 6(c)].The quantity of proteoglycans [Fig. 6(f)] and of type-II colla-gen [Fig. 6(i)] deposited around the cells had likewiseincreased.
At the same concentration (200 ng/ml), BMP-7 had a sim-ilar effect to BMP-2 on the mRNA levels of type-II collagenand aggrecan that were expressed after 2 weeks of cultur-ing in the absence of serum (Fig. 7). The morphologicalchanges observed after staining with Toluidine Blue werelikewise similar (data not shown).
EFFECTS OF DEXAMETHASONE
In a final set of experiments, we evaluated the effect of thesynthetic glucocorticoid dexamethasone on the chondro-genic differentiation of human synovial cells that were cul-tured in alginate under serum-free conditions in thepresence of TGF-b1 or BMP-2. Under serum-free condi-tions, TGF-b1 induced the expression of the type-II collagenand the aggrecan genes irrespective of the absence or pres-ence of dexamethasone. However, dexamethasone hada profound influence on the gene expression stimulated byBMP-2. In the presence of dexamethasone, BMP-2 was un-able to induce the expression of the type-II-collagen gene;
the mRNA level did not rise above the control value[Fig. 8(a)]. The aggrecan gene was up-regulated in the pres-ence of dexamethasone, but to a lower level than in its ab-sence [Fig. 8(b)]. For the transcription factor Sox9, thesituation was different [Fig. 8(c)]. In the absence of dexa-methasone, TGF-b1 and BMP-2 stimulated the expressionof the Sox9 gene, although not to a level that was signifi-cantly higher than the control value in the case of the lattermorphogen. In the presence of dexamethasone, Sox9 wasup-regulated to higher levels by both growth factors. In theabsence of this agent, expression of the type-I-collagengene was unchanged either by TGF-b1 or by BMP-2[Fig. 8(d)]. The presence of dexamethasone did not influ-ence the TGF-b1-induced expression of type-I collagen,but it elicited a decrease in the BMP-2-induced one.
Histologically, BMP-2-stimulated synovial cells that werecultured for 4 weeks in the presence of dexamethasonewere smaller and less inclined to form clusters than thosethat were cultured in the absence of this agent [Fig. 9(aand b)]. In the presence of dexamethasone, the extracellularmatrix manifested only a weak metachromatic signal afterstaining with Toluidine Blue [Fig. 9(b)]. TGF-b1-stimulatedsynovial cells that were cultured for 4 weeks in the absenceor presence of dexamethasone manifested similar degrees
1183Osteoarthritis and Cartilage Vol. 15, No. 10
0
10
20
30
40
50
60
70
80
In
du
ctio
n fo
ld
(m
RN
A level)
Type-II collagen
Total culturing time (in weeks):
Weeks during which BMP-2 (200 ng/ml) waspresent in the culture medium:Weeks during which BMP-2 (200 ng/ml) wasabsent in the culture medium:
4
3 & 41 & 2
4
1 & 23 & 4
Fig. 4. Effect of the mode of exposure to BMP-2 on the relative gene-expression levels of type-II collagen in synovial cells that were derivedfrom four patients. Data pertain to synovial cells that were cultured for 4 weeks in alginate. The culture medium was supplemented with 10%FBS. BMP-2 (200 ng/ml) was present either during the first and the second weeks (third and fourth weeks: morphogen-free), or during the thirdand the fourth weeks (first and second weeks: morphogen-free) of culturing. Data are represented as mean values relative to the controls
(synovial cells cultured for 4 weeks in the complete absence of BMP-2), together with the standard error of the mean (n¼ 4 donors).
of proliferation and chondrogenic differentiation. The meta-chromatic signal after staining with Toluidine Blue wasweak, both in the absence and in the presence of dexameth-asone [Fig. 9(c and d)].
Discussion
Due to its relative abundance in the knee joint and itsdemonstrated capacity to undergo chondrogenic differentia-tion, synovial tissue is a potentially very attractive source ofcells for the repair of articular cartilage. Such a therapeuticapproach would have the advantage of utilizing portions ofautologous tissue derived from a site that readily self-regen-erates after surgery23. In the present study, we investigatedthe culturing and stimulation conditions that were requiredto induce the chondrogenic differentiation of human syno-vial cells derived from the knee joints of osteoarthriticpatients. The cells were grown in a three-dimensional algi-nate system and were stimulated with various growthfactors under various culturing conditions.
When grown in monolayers, the synovial cells exhibiteda fibroblast-like phenotype, as reported by other investiga-tors24. When cultured in a three-dimensional alginate sys-tem, they underwent an immediate chondrogenic switch,even in the absence of a growth factor. This behavior isanalogous to that of dedifferentiated chondrocytes, whichregain their chondrocyte-specific morphology and gene-expression profile after seeding within a three-dimensionalmatrix25. MSCs derived from both the synovium16 and thebone-marrow stroma26 have been previously shown toundergo spontaneous chondrogenic differentiation whencultured under three-dimensional conditions.
In our study, the capacity of the synovial cells to undergochondrogenic differentiation varied between the individualdonors. This variability was not unexpected, since the syno-vial tissue of osteoarthritic patients is known to contain
a heterogeneous population of progenitor cells27. This fea-ture is indeed characteristic of many tissue sources ofMSCs, and the different subpopulations of progenitor cellscan be expected to manifest different chondrogenicpotentials.
As a cell-carrier material, alginate is well suited to supportchondrogenesis, since the cellularity of this hydrogel can bereadily adjusted to optimize the gene-expression pattern.When cultured within alginate beads, MSCs express higherlevels of the type-II-collagen gene than when theyare grown in a pellet system12. In contrast, the type-X-colla-gen gene, which is a marker of late hypertrophic chondro-cytes, is expressed at a higher level in pellet cultures12.Alginate-cultured bovine synovial cells can be stimulatedto express the mRNAs for Sox9, aggrecan and type-IIcollagen at levels that are comparable to those inchondrocytes18.
Apart from spatial and microenvironmental considerations,the chondrogenesis of MSCs depends upon an adequatesupply of several diffusible molecules. An appropriate choiceof culture-medium supplements is of great importance, asevidenced by our data. The BMP-2-induced gene-expressionlevels of all three chondrogenic differentiation markers (ag-grecan, type-II collagen and Sox9) were lower in the pres-ence than in the absence of 10% FBS. These data suggestthat FBS contains components that suppress BMP-2-induced chondrogenesis. FBS contains unknown concentra-tions of numerous biologically-active components which areknown to interfere with the growth-factor-induced redifferen-tiation of dedifferentiated chondrocytes28 and with the chon-drogenic transformation of periosteal explants29. Forexample, the mRNA levels of aggrecan and type-II collagenexpressed by dedifferentiated chondrocytes have beenshown to vary according to the batch of FBS used30.
Members of the TGF-b superfamily are known stimulatorsof chondrogenic differentiation. The binding of TGF-b to itsreceptors causes their dimerization, which triggers
1184 T. Kurth et al.: Chondrogenic potential of human synovial MSCs in alginate
0
0.5
1
1.5
2
2.5
3
3.5
0
0.5
1
1.5
2
2.5
3
3.5
In
du
ctio
n fo
ld
(m
RN
A level)
Type-I collagenSox9
∗
0
200
400
600
800
1000
1200
1400
1600
1800
In
du
ctio
n fo
ld
(m
RN
A level)
Con
trol
BMP-
220
0 ng
/ml
Con
trol
BMP-
220
0 ng
/ml
serum — Cho
ndro
cyte
s
Type-II collagen
∗
∗
∗
∗∗
∗
∗
∗
Synovial cells
0
5
10
15
20
25
30
35
In
du
ctio
n fo
ld
(m
RN
A level)
Con
trol
BMP-
220
0 ng
/ml
Con
trol
BMP-
220
0 ng
/ml
serum Cho
ndro
cyte
s
Aggrecan
#
Synovial cells
4
4.5
In
du
ctio
n fo
ld
(m
RN
A level)
—
Con
trol
Con
trol
Con
trol
Con
trol
BMP-
220
0 ng
/ml
BMP-
220
0 ng
/ml
serum — Cho
ndro
cyte
s
Synovial cells
BMP-
220
0 ng
/ml
BMP-
220
0 ng
/ml
serum Cho
ndro
cyte
s
Synovial cells
—
Fig. 5. BMP-2-induced expression of the genes for type-II collagen, aggrecan, Sox9 and type-I collagen in synovial cells that were derived fromthree patients. Data pertain to alginate-cultured synovial cells that were grown for 2 weeks in the absence or presence of serum (10% FBS).The serum-free culture medium was supplemented with 1% ITS. Data pertaining to articular cartilage chondrocytes, which were derived fromthe same donors and cultured in alginate for 2 weeks in the presence of 10% FBS, are included as a reference standard. The mRNA levels inBMP-2-stimulated synovial cells are expressed relative to those in unstimulated synovial cells cultured in the presence of 10% FBS (value setat ‘‘1’’ for each marker). Data are presented as mean values together with the standard error of the mean (n¼ 3 donors). )P< 0.05 compared
to control synovial cells cultured in the presence of 10% FBS; #P< 0.05 in the indicated group.
a cascade of phosphorylation-dependent signalingevents31. The signals are transduced via smad and non-smad pathways to the nucleus, where they activate tran-scription factors, such as Sox9. These transcription factorsthen induce the expression of genes involved in chondrogen-esis32. Members of each TGF-b subgroup (the TGF-b/acti-vin/nodal-subfamily and the BMP/GDF/MIS-subfamily)have been demonstrated to stimulate the chondrogenictransformation of MSCs in various systems. When exposedto TGF-b1, alginate-cultured lapine synovial cells expresstype-II collagen33. The same morphogen also stimulatesthe expression of the type-II-collagen gene and the produc-tion of proteoglycans by alginate-cultured human bone-marrow stromal cells11,34. In periosteal MSCs, BMP-2 is
more potent than TGF-b1 in inducing the expression ofthe aggrecan core-protein and the type-II-collagen geneand protein35. BMP-2 can up-regulate the expression ofthe Sox9 gene by C3H10T1/2 cells even in monolayer cul-tures36. And when these cells are grown in alginate, thismorphogen promotes the transcription and the translationof the type-II-collagen gene37. Similarly, BMP-2 stimulatesthe expression of type-II collagen, aggrecan and Cartilagealigomeric matrix protein (COMP) by alginate-cultured hu-man bone-marrow stromal cells8. Our data accord withthese findings. They indicate that the choice of morphogen,as well as its concentration, are crucial. In the presence ofserum, a low dose of TGF-b1 (10 ng/ml) did not raise themRNA level of aggrecan above the control value, and it
1185Osteoarthritis and Cartilage Vol. 15, No. 10
2 w 4 w 6 w
Toluidine
Blue
Type-II
collagen
a
fed
b c
ihg
Fig. 6. Light micrographs of 5-mm-thick sections through alginate discs that had been seeded with synovial cells and grown under serum-freeconditions for 2 weeks (a,d,g), 4 weeks (b,e,h) or 6 weeks (c,f,i) in the presence of BMP-2 (200 ng/ml). aec: Low-magnification views of almostthe entire depth of the alginate discs after staining with Toluidine Blue, illustrating the relative abundances of cell clusters, which are repre-sented at higher magnification in def. def: Undifferentiated synovial cells are small and round. Synovial cells that are undergoing chondro-genic transformation are larger and more irregular in shape. The presence of a cartilage-specific extracellular matrix is indicated by themetachromatic staining of proteoglycans with Toluidine Blue (purple coloration). gei: Immunohistochemistry for type-II collagen. A positive
reaction within the extracellular matrix is indicated by a brown coloration. Bars: (aec): 500 mm; (dei): 50 mm.
inhibited the expression of the type-II-collagen gene. Ata higher concentration (50 ng/ml), it inhibited the expres-sion of the aggrecan gene, but did not further suppressthe mRNA level of type-II collagen. Such an inhibitory ef-fect of TGF-b1 has been observed also by Bosnakovskiet al.34. But in most other reports, a stimulatory effecthas been documented. The discrepancy probably relatesto the presence of serum in the culture medium, which,
as aforementioned, contains components that could inter-fere unpredictably with a growth-factor-induced chondro-genic response. In contrast to these findings, BMP-2exerted a dose-dependent stimulatory effect on the gene-expression levels of aggrecan and type-II collagen between50 and 200 ng/ml.
Glucocorticoids, especially synthetic dexamethasone,have been shown in some cases to enhance the
0
500
1000
1500
2000
2500
3000
3500
4000
4500
In
du
ctio
n fo
ld
(m
RN
A level)
Type-II collagen
0
100
200
300
400
500
600
700
In
du
ctio
n fo
ld
(m
RN
A level)
Aggrecan
BMP-2200 ng/ml
Control BMP-7200 ng/ml
BMP-2200 ng/ml
Control BMP-7200 ng/ml
∗ ∗
∗
∗
Fig. 7. Effects of BMP-2 and BMP-7 (both at 200 ng/ml) on the gene-expression levels of type-II collagen and aggrecan in synovial cells thatwere derived from four patients. Data pertain to synovial cells that were cultured in alginate for 2 weeks under serum-free conditions. Theserum-free medium was supplemented with 1% ITS. Data are presented as mean values together with the standard error of the mean
(n¼ 4 donors). )P< 0.05 compared to the control.
1186 T. Kurth et al.: Chondrogenic potential of human synovial MSCs in alginate
0
2
4
6
8
10
12
14
In
du
ctio
n fo
ld
(m
RN
A level)
Sox 9
0
50
100
150
200
250
300
350
400
In
du
ctio
n fo
ld
(m
RN
A level)
Aggrecan
00.20.40.60.8
11.21.41.61.8
2
In
du
ctio
n fo
ld
(m
RN
A level)
Type-I collagen
0
20
40
60
80
100
120
140In
du
ctio
n fo
ld
in
th
ou
san
ds
(m
RN
A level)
Type-II collagen
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
— Dexamethasone
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
— Dexamethasone
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
— Dexamethasone
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
— Dexamethasone
BMP-
220
0 ng
/ml
TGF-
β110
ng/
ml
Con
trol
a
c
b
d
∗
∗
∗
∗∗∗
∗
∗
∗∗
∗∗
#
#
Fig. 8. Effect of dexamethasone on the expression of the genes for type-II collagen (a), aggrecan (b), Sox9 (c) and type-I collagen (d) insynovial cells that were derived from three patients. Data pertain to synovial cells that were cultured for 2 weeks under serum-free conditionsin the absence or presence of dexamethasone (10�7 M), and in the absence or presence of either TGF-b1 (10 ng/ml) or BMP-2 (200 ng/ml).The mRNA levels are expressed relative to those for control cells that were cultured in the absence of dexamethasone and a morphogen(value set at ‘‘1’’ for each marker). Data are presented as mean values together with the standard error of the mean (n¼ 3 donors)
)P< 0.05 compared to the control; #P< 0.05 between the indicated groups.
chondrogenic response induced by several growth factors.In combination with TGF-b3, dexamethasone augmentsthe gene expression of aggrecan, type-II collagen andCOMP by pelleted human bone-marrow stromal MSCs38.And in combination with TGF-b1, it enhances the expres-sion of glycosaminoglycans, aggrecan, type-II collagenand Sox9 by monolayers of lapine and human bone-marrowstromal MSCs39. A comparison of human MSCs harvestedfrom different sources (bone marrow, skeletal muscle, syno-vium, adipose tissue and periosteum) has revealed those ofsynovial origin to have a greater potential for chondrogene-sis when treated with a combination of dexamethasone,TGF-b3 and BMP-26. In our study, the TGF-b1-mediatedexpression of the aggrecan and the type-II-collagen geneswas not enhanced in the presence of dexamethasone,whereas the BMP-2-induced expression of these markerswas markedly suppressed. These findings support those re-ported for alginate-cultured bovine synovial cells18. As indi-cated above, TGF-b1 and BMP-2 belong to separatesubgroups of the TGF-b superfamily. Hence, their differenteffects could reflect the activation of different signaling
pathways. Interestingly, dexamethasone down-regulatedthe expression of the type-I-collagen gene both in the ab-sence and in the presence of BMP-2. It also enhancedthe TGF-b1- and BMP-2-induced expression of the Sox9gene to varying degrees. The chondrocytes of newbornmice have been shown to express elevated levels of theSox9 gene within 24 h of introducing dexamethasone intothe culture medium40. In contrast, murine embryonic cellsgrown in alginate for 5 days express lower levels of theSox9 gene when dexamethasone is present41. In the pres-ence of dexamethasone, BMP-2 has been reported to in-duce the osteogenesis of human42 and rat43 bone-marrowstromal cells. Conceivably, such a dexamethasone-depen-dent differentiation along the osteoblast lineage could ac-count for the suppressed expression of the aggrecan andthe type-II-collagen genes in our study.
The sustained expression of type-I collagen by BMP-2-and TGF-b1-stimulated synovial cells raises a question asto the quality of the cartilage tissue produced. In a futurestudy, this issue will be addressed by extending the panelof chondrogenic differentiation markers monitored. Most
1187Osteoarthritis and Cartilage Vol. 15, No. 10
-Dexamethasone +Dexamethasone
BMP-2
TGF-β1
a b
c d
Fig. 9. Light micrographs of Toluidine-Blue-stained, 5-mm-thick sections through cell-containing alginate discs that had been grown for 4 weeksunder serum-free conditions in the absence (a,c) or presence (b,d) of dexamethasone (10�7 M), and in the presence of either BMP-2 [200 ng/ml] (a,b) or TGF-b1 [10 ng/ml] (c,d). a,b: BMP-2-stimulated cells were more inclined to form clusters in the absence (a) than in the presence (b)of dexamethasone. The cell clusters elaborated a substantial quantity of proteoglycan-rich extracellular matrix (purple coloration). c,d: TGF-b1-stimulated cells cultured in the absence (c) or presence (d) of dexamethasone manifested similar degrees of chondrogenic transformation. In
both instances, metachromatic staining of the extracellular matrix for proteoglycans was weak. Bars (aed): 50 mm.
importantly, we will quantify the expression of the type-X-collagen gene. Type-X collagen is a marker of advancedchondrocytic hypertrophy. Its expression precedes enchon-dral ossification, and has been correlated with the vascular-ization and mineralization of pelleted MSC preparationsdeposited at an ectopic site in severe combined immunode-ficiency (SCID) mice44. For the development of a functional,hyaline type of articular cartilage that would be useful in thecontext of repair, chondrogenic differentiation must be ar-rested before it progresses to this advanced stage of hyper-trophy by optimizing the stimulation conditions.
In conclusion, our data reveal that synovial cells originatingfrom elderly human osteoarthritic patients can be induced toundergo chondrogenic differentiation when cultured in algi-nate. This finding agrees with previous reports that the chon-drogenic potential of human synovial MSCs is not influencedby the age of the donor4. In our study, chondrogenesis wasmost pronounced when the cells were stimulated with eitherBMP-2 or BMP-7 under serum- and dexamethasone-freeconditions.
Acknowledgments
We would like to thank Prof. Slobodan Vukicevic (Departmentof Anatomy, University of Zagreb, Croatia) for his support,and Mark Siegrist (ITI Research Institute, University ofBern, Switzerland) for his excellent technical assistance.BMP-2 was generously provided by Wyeth (Cambridge,MA, USA), and BMP-7 was a gift from Stryker Biotech(Hopkinton, MA, USA). This project was supported bythe Novartis Foundation (Basel, Switzerland) and by the
Swiss National Science Foundation [Bern, Switzerland (toEBH)]. FHC is supported by an intramural research grantfrom the National Institute of Health (Bethesda, Maryland,USA).
References
1. Hunziker EB. Articular cartilage repair: basic scienceand clinical progress. A review of the current statusand prospects. Osteoarthritis Cartilage 2002;10:432e63.
2. Redman SN, Oldfield SF, Archer CW. Current strate-gies for articular cartilage repair. Eur Cell Mater2005;9:23e32 (discussion 23e32).
3. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK,Douglas R, Mosca JD, et al. Multilineage potential ofadult human mesenchymal stem cells. Science1999;284:143e7.
4. De Bari C, Dell’Accio F, Tylzanowski P, Luyten FP. Mul-tipotent mesenchymal stem cells from adult human sy-novial membrane. Arthritis Rheum 2001;44:1928e42.
5. De Bari C, Dell’Accio F, Vandenabeele F,Vermeesch JR, Raymackers JM, Luyten FP. Skeletalmuscle repair by adult human mesenchymal stemcells from synovial membrane. J Cell Biol 2003;160:909e18.
6. Sakaguchi Y, Sekiya I, Yagishita K, Muneta T. Compar-ison of human stem cells derived from various mesen-chymal tissues: superiority of synovium as a cellsource. Arthritis Rheum 2005;52:2521e9.
1188 T. Kurth et al.: Chondrogenic potential of human synovial MSCs in alginate
7. Zuk PA, Zhu M, Ashjian P, De Ugarte DA, Huang JI,Mizuno H, et al. Human adipose tissue is a sourceof multipotent stem cells. Mol Biol Cell 2002;13:4279e95.
8. Majumdar MK, Wang E, Morris EA. BMP-2 and BMP-9promote chondrogenic differentiation of human multi-potential mesenchymal cells and overcome the inhibi-tory effect of IL-1. J Cell Physiol 2001;189:275e84.
9. Sekiya I, Colter DC, Prockop DJ. BMP-6 enhanceschondrogenesis in a subpopulation of human marrowstromal cells. Biochem Biophys Res Commun 2001;284:411e8.
10. Schmitt B, Ringe J, Haupl T, Notter M, Manz R,Burmester GR, et al. BMP2 initiates chondrogenic lin-eage development of adult human mesenchymal stemcells in high-density culture. Differentiation 2003;71:567e77.
11. Ma HL, Hung SC, Lin SY, Chen YL, Lo WH. Chondro-genesis of human mesenchymal stem cells encapsu-lated in alginate beads. J Biomed Mater Res A 2003;64:273e81.
12. Yang IH, Kim SH, Kim YH, Sun HJ, Kim SJ, Lee JW.Comparison of phenotypic characterization between‘‘alginate bead’’ and ‘‘pellet’’ culture systems as chon-drogenic differentiation models for human mesenchy-mal stem cells. Yonsei Med J 2004;45:891e900.
13. O’Connell JX. Pathology of the synovium. Am J ClinPathol 2000;114:773e84.
14. Hunziker EB, Rosenberg LC. Repair of partial-thick-ness defects in articular cartilage: cell recruitmentfrom the synovial membrane. J Bone Joint Surg Am1996;78:721e33.
15. Hunziker EB. Growth-factor-induced healing of partial-thickness defects in adult articular cartilage. Osteoar-thritis Cartilage 2001;9:22e32.
16. Shirasawa S, Sekiya I, Sakaguchi Y, Yagishita K,Ichinose S, Muneta T. In vitro chondrogenesis of humansynovium-derived mesenchymal stem cells: optimalcondition and comparison with bone marrow-derivedcells. J Cell Biochem 2006;97:84e97.
17. Hunziker EB, Driesang IM, Morris EA. Chondrogenesisin cartilage repair is induced by members of the trans-forming growth factor-beta superfamily. Clin Orthop2001;S171e81.
18. Park Y, Sugimoto M, Watrin A, Chiquet M, Hunziker EB.BMP-2 induces the expression of chondrocyte-specificgenes in bovine synovium-derived progenitor cells cul-tured in three-dimensional alginate hydrogel. Osteoar-thritis Cartilage 2005;13:527e36.
19. Wong M, Siegrist M, Goodwin K. Cyclic tensile strainand cyclic hydrostatic pressure differentially regulateexpression of hypertrophic markers in primary chon-drocytes. Bone 2003;33:685e93.
20. Stahlberg A, Hakansson J, Xian X, Semb H, Kubista M.Properties of the reverse transcription reaction inmRNA quantification. Clin Chem 2004;50:509e15.
21. Livak KJ, Schmittgen TD. Analysis of relative gene ex-pression data using real-time quantitative PCR andthe 2(-Delta Delta C(T)) method. Methods 2001;25:402e8.
22. Pfaffl MW, Horgan GW, Dempfle L. Relative expressionsoftware tool (REST) for group-wise comparison andstatistical analysis of relative expression results inreal-time PCR. Nucleic Acids Res 2002;30:e36.
23. Ostergaard M, Ejbjerg B, Stoltenberg M, Gideon P,Volck B, Skov K, et al. Quantitative magnetic reso-nance imaging as marker of synovial membrane
regeneration and recurrence of synovitis after arthro-scopic knee joint synovectomy: a one year follow upstudy. Ann Rheum Dis 2001;60:233e6.
24. Vandenabeele F, De Bari C, Moreels M, Lambrichts I,Dell’Accio F, Lippens PL, et al. Morphological andimmunocytochemical characterization of culturedfibroblast-like cells derived from adult human synovialmembrane. Arch Histol Cytol 2003;66:145e53.
25. Schulze-Tanzil G, de Souza P, Villegas Castrejon H,John T, Merker HJ, Scheid A, et al. Redifferentiationof dedifferentiated human chondrocytes in high-density cultures. Cell Tissue Res 2002;308:371e9.
26. Worster AA, Brower-Toland BD, Fortier LA, Bent SJ,Williams J, Nixon AJ. Chondrocytic differentiation ofmesenchymal stem cells sequentially exposed totransforming growth factor-beta1 in monolayer andinsulin-like growth factor-I in a three-dimensionalmatrix. J Orthop Res 2001;19:738e49.
27. Fickert S, Fiedler J, Brenner RE. Identification, quantifi-cation and isolation of mesenchymal progenitor cellsfrom osteoarthritic synovium by fluorescence auto-mated cell sorting. Osteoarthritis Cartilage 2003;11:790e800.
28. Fortier LA, Nixon AJ, Mohammed HO, Lust G. Alteredbiological activity of equine chondrocytes cultured ina three-dimensional fibrin matrix and supplementedwith transforming growth factor beta-1. Am J Vet Res1997;58:66e70.
29. Fitzsimmons JS, Sanyal A, Gonzalez C, Fukumoto T,Clemens VR, O’Driscoll SW, et al. Serum-free mediafor periosteal chondrogenesis in vitro. J Orthop Res2004;22:716e25.
30. Yaeger PC, Masi TL, de Ortiz JL, Binette F, Tubo R,McPherson JM. Synergistic action of transforminggrowth factor-beta and insulin-like growth factor-I in-duces expression of type II collagen and aggrecangenes in adult human articular chondrocytes. ExpCell Res 1997;237:318e25.
31. Shi Y, Massague J. Mechanisms of TGF-beta signalingfrom cell membrane to the nucleus. Cell 2003;113:685e700.
32. Wan M, Cao X. BMP signaling in skeletal development.Biochem Biophys Res Commun 2005;328:651e7.
33. Nishimura K, Solchaga LA, Caplan AI, Yoo JU,Goldberg VM, Johnstone B. Chondroprogenitor cellsof synovial tissue. Arthritis Rheum 1999;42:2631e7.
34. Bosnakovski D, Mizuno M, Kim G, Ishiguro T,Okumura M, Iwanaga T, et al. Chondrogenic differen-tiation of bovine bone marrow mesenchymal stemcells in pellet cultural system. Exp Hematol 2004;32:502e9.
35. Hanada K, Solchaga LA, Caplan AI, Hering TM,Goldberg VM, Yoo JU, et al. BMP-2 induction andTGF-beta 1 modulation of rat periosteal cell chondro-genesis. J Cell Biochem 2001;81:284e94.
36. Zehentner BK, Dony C, Burtscher H. The transcriptionfactor Sox9 is involved in BMP-2 signaling. J BoneMiner Res 1999;14:1734e41.
37. Steinert A, Weber M, Dimmler A, Julius C, Schutze N,Noth U, et al. Chondrogenic differentiation of mesen-chymal progenitor cells encapsulated in ultrahigh-viscosity alginate. J Orthop Res 2003;21:1090e7.
38. Derfoul A, Perkins GL, Hall DJ, Tuan RS. Glucocorti-coids promote chondrogenic differentiation of adulthuman mesenchymal stem cells by enhancing expres-sion of cartilage extracellular matrix genes. Stem Cells2006.
1189Osteoarthritis and Cartilage Vol. 15, No. 10
39. Chen CW, Tsai YH, Deng WP, Shih SN, Fang CL,Burch JG, et al. Type I and II collagen regulation ofchondrogenic differentiation by mesenchymal progen-itor cells. J Orthop Res 2005;23:446e53.
40. Sekiya I, Koopman P, Tsuji K, Mertin S, Harley V,Yamada Y, et al. Dexamethasone enhances SOX9expression in chondrocytes. J Endocrinol 2001;169:573e9.
41. Tanaka H, Murphy CL, Murphy C, Kimura M, Kawai S,Polak JM. Chondrogenic differentiation of murine em-bryonic stem cells: effects of culture conditions anddexamethasone. J Cell Biochem 2004;93:454e62.
42. Diefenderfer DL, Osyczka AM, Garino JP, Leboy PS.Regulation of BMP-induced transcription in cultured
human bone marrow stromal cells. J Bone Joint SurgAm 2003;85-A(Suppl 3):19e28.
43. Rickard DJ, Sullivan TA, Shenker BJ, Leboy PS,Kazhdan I. Induction of rapid osteoblast differentia-tion in rat bone marrow stromal cell cultures bydexamethasone and BMP-2. Dev Biol 1994;161:218e28.
44. Pelttari K, Winter A, Steck E, Goetzke K, Hennig T,Ochs BG, et al. Premature induction of hyper-trophy during in vitro chondrogenesis of humanmesenchymal stem cells correlates with calcifica-tion and vascular invasion after ectopic transplan-tation in SCID mice. Arthritis Rheum 2006;54:3254e66.
