THE ANATOMICAL RECORD PART A 278A:428 436 (2004)

Chondrogenesis of Human Bone

Marrow-Derived Mesenchymal Stem
Cells in Agarose Culture
C.-Y. CHARLES HUANG,1,2 PAUL M. REUBEN,1,2 GIANLUCA DIPPOLITO,1
PAUL C. SCHILLER,1 AND HERMAN S. CHEUNG1,2*
1
Research Service and Geriatrics Research, Education, and Clinical Center,
Veterans Affairs Medical Center, Miami, Florida
2
Department of Biomedical Engineering, University of Miami, Coral Gables, Florida

ABSTRACT
Mesenchymal stem cells derived from human bone marrow (hBM-MSCs) can differentiate into chondrogenic cells for the potential treatment of injured articular cartilage. To
evaluate agarose gels as a supportive material for chondrogenesis of hBM-MSCs, this study
examined chondrogenesis of hBM-MSCs in the agarose cultures. Pellet cultures were employed to conrm the chondrogenic potential of the hBM-MSCs that were used in agarose
cultures. The hBM-MSCs were seeded in 2% agarose constructs at the initial cell-seeding
densities of 3, 6, and 9 106 cells/ml while each of pellets was formed using 2.5 105 cells.
Chondrogenesis of hBM-MSCs was induced by culturing cell-agarose constructs and pellets
for 21 days in the presence of a dened medium containing transforming growth factor 3
(TGF-3). The analysis of reverse transcription-polymerase chain reaction showed that
hBM-MSCs of agarose and pellet cultures expressed the chondrogenic markers of collagen
type II and aggrecan in the presence of TGF-3. The deposition of cartilage-specic macromolecules was detected in both agarose and pellet cultures by histological and immunohistochemical assessments. Chondrogenesis of hBM-MSCs in agarose gels directly correlated
with the initial cell-seeding density, with the cell-agarose constructs of higher initial cellseeding density exhibiting more cartilage-specic gene expressions. This study establishes a
basic model for future studies on chondrogenesis of hBM-MSCs using the agarose cultures.
Anat Rec Part A 278A:428 436, 2004. 2004 Wiley-Liss, Inc.

Key words: biomaterials; cartilage; adult stem cells; TGF-3; agarose; tissue
engineering

Due to the inaccessibility of reparative cells and low

cellularity, defects of articular cartilage, without the penetration of subchondral bone, result in insufcient healing
responses of adjacent chondrocytes. Such defects usually
progress to cartilage degeneration and nally advance to
osteoarthritis.
Genzyme Biosurgery (Cambridge, MA) was the rst to
market autologous cultured chondrocytes (Carticel) for
cell-based therapies such as cartilage repair. While the
procedure was used clinically, the protocol to harvest
chondrocytes from adjacent normal sites in the joint was
invasive and permanently damaged normal area of cartilage. It has been suggested that this invasive procedure
alone can initiate degenerative changes and eventually
lead to osteoarthritis in the joint (Erickson et al., 2002).
Mesenchymal stem cells derived from bone marrow
(BM-MSCs) have demonstrated the multipotential to differentiate into several cell lineages, including chondrocytes and osteocytes (Aydelotte and Kuettner, 1988;

2004 WILEY-LISS, INC.

Mackay et al., 1998; Yoo et al., 1998; Pittenger et al.,

1999). In addition to their multipotency, BM-MSCs can be
acquired by bone marrow aspiration without permanently
damaging tissues, efciently expanded in monolayers by
serial passages without altering their differentiation potential, and capable of repairing joint defects involving
both subchondral bone and articular cartilage. These

Grant sponsor: the National Institutes of Health; Grant number: AR 38421; Grant sponsor: VA Merit Review Grant.
*Correspondence to: Dr. Herman S. Cheung, Research Service,
Miami VA Medical Center, 1201 NW 16th Street, Miami, Fl
33125. Fax: 305-575-3365. E-mail: hcheung@med.miami.edu
Received 14 July 2003; Accepted 4 December 2003
DOI 10.1002/ar.a.20010

CHONDROGENESIS OF hBM-MSC

characteristics make BM-MSCs candidate cells for cell

transplantation (Wakitani et al., 1994, 2002; Diduch et al.,
2000) and the development of autologous cartilaginous
implants (Mackay et al., 1998; Caterson et al., 2001) for
the repair of injured articular cartilage that has very
limited capability for self-healing.
Biomaterials also play an important role in serving as a
delivery vehicle in cell transplantation as well as providing initial three-dimensional structures for developing tissues with required geometry. Generally, the suitability of
biomaterials for cell transplantation or engineering cartilaginous tissues for the repair of articular cartilage depends on certain criteria: the biomaterial is biocompatible,
can promote chondrogenic expressions of cells, can allow
the deposition of the extracellular matrix, and is mechanically stable. Recently, a number of biomaterials, such as
collagen gel (Wakitani et al., 1994, 2002), hyaluronic acidbased polymer (Angele et al., 1999; Solchaga et al., 1999),
polylactide/aliginate amalgam (Caterson et al., 2001), gelatin-based resorbable sponge (Ponticiello et al., 2000),
photopolymerized hydrogels (Williams et al., 2003), and
alginate (Diduch et al., 2000), have been proposed as cell
carriers of BM-MSCs for cell transplantation or cartilage
tissue engineering. These biomaterials have been shown
to support chondrogenesis of BM-MSCs under in vivo (Wakitani et al., 1994, 2002; Solchaga et al., 1999; Diduch et
al., 2000) or in vitro (Angele et al., 1999; Caterson et al.,
2001) conditions. However, other biomaterials such as
agarose, which may potentially promote chondrogenesis of
BM-MSCs (Diduch et al., 2000), still have not been explored.
Three-dimensional agarose cultures of chondrocytes
had been shown to maintain the stability of differentiated
phenotype (Benya and Shaffer, 1982; Aydelotte and
Kuettner, 1988) while promoting the chondrocyte biosynthesis of cartilage-specic aggrecan and collagen that
were deposited within agarose gels to develop a cartilagelike tissue (Buschmann et al., 1992; Mauck et al., 2000).
This suggests that the agarose culture of chondrocytes
may be able to synthesize a functional cartilaginous tissue
(Buschmann et al., 1992; Mauck et al., 2000). Due to their
mechanical stability, agarose cultures have also been
widely used as an in vitro model for studying the inuence
of compressive loading on the biosynthetic activities and
mechanotransduction events of chondrocytes, the deposition of pericellular matrix (Buschmann et al., 1995; Lee
and Bader, 1997; Knight et al., 1998; Mauck et al., 2000;
Roberts et al., 2001), and chondrogenic differentiation of
chick limb-bud cells (Elder et al., 2000). Furthermore, a
recent study has demonstrated promising results on the
repair of full-thickness cartilage defects in rabbit knee
joints by transplanting chondrocyte-agarose allografts
(Rahfoth et al., 1998). Therefore, agarose gels may serve
as a reasonable model for studying chondrogenic responses of BM-MSCs.
To evaluate agarose gels as a potential material for
cell-based therapies or cartilage tissue engineering using
BM-MSCs, the fundamental step is to examine whether
the three-dimensional agarose culture can support chondrogenesis of BM-MSCs under in vitro control conditions.
Since previous studies have shown that chondrogenic differentiation of human BM-MSCs (hBM-MSCs) can be induced in the suspension pellet culture (Yoo et al., 1998;
Pittenger et al., 1999) and in the polymer culture (Angele
et al., 1999; Ponticiello et al., 2000; Caterson et al., 2001)

429

with the treatment of TGF-, we hypothesize that the

three-dimensional agarose culture can provide a supportive environment for chondrogenesis of hBM-MSCs in the
presence of such growth factors. The objective of this study
was to examine the chondrogenesis of hBM-MSCs cultured in agarose gels with the treatment of TGF-3 by
analyzing the expressions of cartilage-specic macromolecules (collagen type II and aggrecan) as well as the deposition of the collagen type II proteins secreted by hBMMSCs within agarose gels.

MATERIALS AND METHODS

Culture of hBM-MSCs
The hBM-MSCs were provided by the Diabetes Research Institute at University of Miami School of Medicine
and the Geriatrics Research, Education, and Clinical Center at Miami Veterans Affairs Medical Center. The procedure for isolating hBM-MSCs from bone marrow of postmortem thoracolumbar (T1T5) vertical bodies had been
previously described (DIippolito et al., 1999). Briey, the
vertical columns were harvested from eight donors (age
range, 21 45) immediately after death from traumatic
injuries. The vertical bodies were separated along the
sagittal midline craniocaudal axis and cut into bone chips
( 5 mm3). The bone chips were placed at room temperature in 1,000 ml of processing medium (X-Vivo 10; Biowhittaker, Walkersville, MD) containing 2.5% human serum albumin (Calbiochem, La Jolla, CA), bacitracin (50
U/ml), heparin (10 U/ml), polymyxin B (500 U/ml), and 2
mM HEPES buffer (pH 7.2). The cell suspension was
separated from the bone chips by ltering through two
consecutive stainless steel screens (pore diameters 450
and 180 m) and then centrifuged at 300 g for 10 min at
4C. The cell pellet was suspended in -modied essential
medium (-MEM; Gibco-BRL, Grand Island, NY) with
10% heat-inactivated fetal bovine serum (FBS; GibcoBRL). The remaining bone marrow cells trapped within
bone trabeculae were released at room temperature by
two 30-min gentle rocking agitations ( 40 cycles/min) of
bone chips in 1,000 ml of RPMI-1640 (Gibco-BRL) containing 2.5% human serum albumin, heparin (10 U/ml), gentamicin (0.5 mg/ml), and 2 mM HEPES buffer (pH 7.2).
The suspension of bone marrow cells was ltered through
a Y-Type Blood Set (McGaw, Irvine, CA) to remove bone
fragments and clumps of cells. An aliquot of cells was
treated with 4% acetic acid to lyse erythrocyte and then
counted with a hemacytometer for trypan blue exclusion.
Cells were plated in culture asks with low-glucose Dulbeccos modied Eagles medium (DMEM; Gibco-BRL)
containing 10% FBS serum and 1% antibiotics and incubated at 37C with 5% CO2. After 4 days of primary
culture, nonadherent cells were removed by changing medium. After 14 days of primary culture, hBM-MSCs were
harvested with trypsin/EDTA and frozen in 1 ml aliquots
in liquid nitrogen.
The hBM-MSCs were thawed and expanded in 10 cm
culture dishes with low-glucose DMEM containing 10%
FBS serum and 1% antibiotics (passage 1). The cells of
each 10 cm culture disk were harvested and replated
equally into two 10 cm culture disks after 14 days (passage 2), 21 days (passage 3), and 28 days (passage 4).
Passage 4 cells were used in both agarose cultures and
pellet cultures.

430

HUANG ET AL.

TABLE 1. PCR primer sequences used in this study

Gene
Collagen type I
Sense
Antisense
Collagen type II
Sense
Antisense
Collagen X
Sense
Antisense
Aggrecan
Sense
Antisense
-actin
Sense
Antisense

Sequence

Size

Reference

5-CGTGGTGACAAGGGTGAGAC-3
5-TAGGTGATGTTCTGGGAGGC-3

827 bp

Genbank
Z74615

5-AACCAGATTGAGAGCATCCGC-3
5-CCTTCAGGGCAGTGTACGTGA-3

517 bp

(Su et al., 1989)

5-GCCCAAGAGGCGTGATGGCTTATTTGT-3
5-CCTGAGAAAGAGGAGTGGACATAC-3

703 bp

Genbank NM000493

5-ACCGTTGCAGACATTGACGAGTG-3
5-ACTCCTGCTCC-TCGGGGGTGACG-3

300 bp

(Freije et al., 1994)

5-GCTCGTCGTCGACAACGGCTC-3
5-CAAACATGATCTGGGTCATCTTCTC-3

353 bp

GibcoBRL

Pellet Culture of hBM-MSCs

Chondrogenic potential of the hBM-MSCs used in this
study was initially conrmed by pellet cultures that have
been shown to promote chondrogenesis of hBM-MSCs in
other studies (Yoo et al., 1998; Pittenger et al., 1999).
Approximately 2.5 105 cells were trypsinized and centrifuged in 15 ml polypropylene tube to form a pellet. The
pellets of the control group (n 30) were cultured in
serum-free medium consisting of high-glucose DMEM, 1%
insulin-transferrin-selenium supplements (nal concentrations, 10 g/ml bovine insulin, 5.5 g/ml transferrin,
6.7 ng/ml sodium selenite; Gibco-BRL), 1.25 mg/ml bovine
albumin, 5.33 g/ml linoleic acid, 40 g/ml proline, 50 g
/ml ascorbic acid, and 10-7 M dexamethasone (Sigma, St.
Louis, MO), while the serum-free medium supplemented
with 10 ng/ml of recombinant human TGF-3 (R&D Systems, Minneapolis, MN) was used for the treated group
(n 30). Three pellets in each group were used for histological and immunohistochemical evaluations while the
expression of the cartilage-specic markers were examined on the remaining pellets. All pellet cultures were
performed in humidied incubator maintained at 37C in
5% CO2 for 21 days. The culture medium was changed
every 23 days.

Agarose Culture of hBM-MSCs

This study examined chondrogenesis of hBM-MSCs cultured in 2% agarose gels that have been shown to maintain chondrocyte phenotype, promote the cartilage-specic
macromolecule biosynthesis of chondrocyte, and develop a
cartilage-like tissue (Buschmann et al., 1992; Mauck et
al., 2000). After trypsinizing and cell counting, hBMMSCs were suspended in the high-glucose DMEM solution and then mixed with an equal volume of 4% (wt/vol)
agarose solution at 37C to produce mixtures of 3 106,
6 106, and 9 106 cells/ml. The cell-agarose constructs
(8 mm in diameter and 1.5 mm thick) were formed by
casting the cell-agarose mixture in a custom-designed
mold and gelling for 15 min at room temperature. The
cell-agarose mixtures of 3 106, 6 106, and 9 106
cells/ml produced 20, 20, and 12 cell-agarose constructs,
respectively. For each cell-seeding density, the cell-agarose constructs were equally separated into two groups (the
control group and the TGF-3-treated group). Two cell-

agarose constructs in each group were analyzed histologically and immunohistochemically while the remaining
constructs were used for the analyses of cartilage-specic
gene expression. All cell-agarose constructs were cultured
in three 24-well culture plates in a humidied 5% CO2
atmosphere at 37C for 21 days with the culture media
used in the pellet culture study. The culture medium was
changed every 23 days.

Reverse Transcription-Polymerase Chain

Reaction (RT-PCR) Analysis
The expressions of cartilage-specic markers (collagen
type II and aggrecan) and collagen type I gene were examined to determine whether the chondrogenic differentiation of hBM-MSCs occurred or not. Because of the low
cell number in specimens and the low recovery rate of
RNA from agarose gels, the total RNA of hBM-MSCs was
extracted and combined from all pellets or all cell-agarose
constructs in the same group using the TRIzol reagent
(Gibco-BRL) according to the manufacturers instructions.
The RT-PCR was performed in GeneAmp PCR system
(9600, Perkin Elmer Ceyus, Norwalk, CT) using the ThermoScript RT-PCR system (Gibco-BRL). The cDNA synthesis was performed by a 60-min incubation at 50C with an
avian RNase H-minus reverse transcriptase and Oligo(dT)20 primer, followed by enzyme inactivation at 85C
for 5 min. PCR amplications for the resulting cDNA
samples were carried out for 30 cycles (for pellets) or 40
cycles (for cell-agarose constructs) by denaturing at 95C
for 30 sec, annealing at 58C for 45 sec, and extending at
72C for 45 sec, with nal extension at 72C for 10 min.
The primers used in the PCR amplications were listed in
Table 1. As an internal control, 353 bp of the constitutively
expressed housekeeping gene, -actin, was also synthesized and used to normalize the amount of mRNA in each
RT-PCR reaction. The PCR products were analyzed by
electrophoresis on 2% agarose gel containing ethidium
bromide and subsequently photographed using a low-light
image system (ChemiImager 4000, Alpha Innotech, San
Leandro, CA). Integrated density value (IDV) of each PCR
product on the image of electrophoresis was measured by
the AlphaEase software (Alpha Innotech) and normalized
by the IDV of the PCR product of -actin gene for compar-

431

CHONDROGENESIS OF hBM-MSC

ison between the groups of different initial cell-seeding

densities.

Histological and Immunohistochemical Analysis

Deposition of extracellular matrix proteins was examined by histological and immunohistochemical evaluations. Pellets and cell-agarose constructs were xed in
10% buffered formalin for 2 hr at 4C and 4% paraformaldehyde for 15 min at room temperature, respectively. After washing in PBS, the specimens were dehydrated in a
graded series of increasing concentrations of ethanol,
cleaned with xylene, embedded in parafn, and cut into 5
m sections. Proteoglycans were detected by staining sections with alcian blue solution. The deposition of collagen
type I and type II proteins was identied by immunohistochemical analysis. After deparafnizing, sections were
incubated in 0.3% hydrogen peroxide in distilled water for
30 min to remove endogenous peroxidase. After rinsing
with distilled water and PBS, 10% normal horse serum
was placed on the sections to block nonspecic background, then incubated with mouse monoclonal antihuman collagen antibody (type I: I-8H5, 500 g/ml; type II:
II-4CII, 500 g/ml; ICN Biomaterials, Aurora, OH) at 4C
overnight. Following extensive washing with PBS to remove the primary antibody, immunoactivity was detected
by incubating the sections with biotinylated horse antimouse antibody, followed by incubation with avidin-biotin-peroxidase complex (ImmunoPure ABC peroxidase
staining kits; Pierce, Rockford, IL). Peroxidase activity
was visualized using 3-3-diaminobenzidine (DAB) as substrate. The sections were incubated with 0.06% DAB in 0.1
M Tris-HCL, pH 7.5, containing 0.03% H2O2, followed by
counterstaining with hematoxyline. All incubations were
performed in a humidied chamber.

RESULTS
Chondrogenic Differentiation in Pellet Culture
After a 21-day culture, chondrogenesis of hBM-MSCs in
the TGF-3-treated pellet culture was detected by immunohistochemical and histological assessments, as illustrated in Figures 1 and 2. Only the specimens of the
TGF-3-treated group exhibited immunopositive reaction
for collagen type II, as visualized by strong brown products in Figure 1. The specimens of the TGF-3-treated
group exhibited stronger alcian blue staining as compared
with the specimens of the control group (Fig. 2), indicating
that more proteoglycans were deposited in the TGF-3treated specimens. The deposition of collagen type I proteins was also present in pellets of both groups (data not
shown). Chondrogenic differentiation of hBM-MSCs in
pellet cultures was further conrmed by the RT-PCR analysis. Figure 3 shows the results of the RT-PCR analysis for
both groups, demonstrating that the pellet culture of
hBM-MSCs in the presence of TGF-3 expressed the cartilage-specic genes (collagen type II and aggrecan),
whereas only weak expression of the aggrecan gene was
exhibited in pellets of the control group. However, pellets
of both groups still maintained a small expression of collagen type I gene after a 21-day culture (data not shown),
whereas collagen type X gene expression was not detected
in pellets of either groups.

Chondrogenic Differentiation in Agarose

Culture
Figure 4a and b show the typical phase-contrast light
micrographs of hBM-MSCs cultured in agarose gel at the
initial seeding density of 9 106 cells/ml with and without
the treatment of TGF-3 for 21 days, demonstrating that
the agarose gel maintained the spherical morphology of
the hBM-MSCs; they were able to proliferate and deposit
extracellular matrix surrounding cells to form larger cellular aggregates within the agarose gel in the presence of
TGF-3. The cartilage-specic extracellular matrix macromolecues, collagen type II and proteoglycans, were detected in the cellular aggregate formed by hBM-MSCs
cultured in the agarose gel with the treatment of TGF-3
(Fig. 4d and f), whereas the deposition of collagen type II
and proteoglycans were not detected in the cell-agarose
constructs of the control group (Fig. 4c and e). The results
of the RT-PCR analysis showed that the hBM-MSCs of the
TGF-3-treated group with the initial seeding density of
9 106 cells/ml expressed both chondrogenic markers of
collagen type II and aggrecan (Fig. 5). However, the TGF3-treated specimens with the initial cell-seeding density
of 3 106 and 6 106 cells/ml only exhibited the expression of collagen type II gene (Fig. 6) but not aggrecan. By
doubling cell-seeding density from 9 106 to 1.8 107
cells/ml, the difference is less than 20% increase in both
cartilage-specic gene expressions (data not shown), suggesting that cell seeding density of 9 106 cells/ml is near
the optimum condition. The specimens of the control
group for the three initial cell-seeding densities expressed
neither collagen type II gene nor aggrecan gene. A diminished expression of collagen type I gene was still found in
the cell-agarose constructs of the TGF-3-treated group
and the control group after a 21-day culture (data not
shown).
Figure 6 shows a typical comparison on the normalized
IDVs of PCR products for the collagen type II gene between the TGF-3-treated groups of different initial cellseeding densities, demonstrating that the cell-agarose
constructs of higher cell-seeding density exhibited more
expression of collagen type II gene.

DISCUSSION
Due to the inaccessibility of reparative cells and low
cellularity, defects of articular cartilage, without the penetration of subchondral bone, result in insufcient healing
responses of adjacent chondrocytes. Such defects usually
progress to cartilage degeneration and nally advance to
osteoarthritis. When injury penetrates the subchondral
bone, mesenchymal stem cells can be recruited from bone
marrow to regenerate reparative tissues (Coletti et al.,
1972; Cheung et al., 1978, 1980; Furukawa et al., 1980;
Johnson, 1986). However, these reparative tissues were
usually identied as brocartilage that is histologically
similar yet mechanically inferior to hyaline cartilage (Coletti et al., 1972; Cheung et al., 1980; Furukawa et al.,
1980; Johnson, 1986). Since adult BM-MSCs exhibit a
promising potential for cartilage regeneration, a possible
strategy for the restoration of injured articular cartilage to
the normal status is the transplantation of BM-MSCs or
BM-MSC-generated cartilaginous implants using the biomaterials that potentially support the regeneration of new
functional cartilage.

432

HUANG ET AL.

Figure 1.

Figure 2.

CHONDROGENESIS OF hBM-MSC

Fig. 3. Typical RT-PCR analysis of cartilage-specic genes on pellet

specimens of hBM-MSCs after a 21-day culture.

Chondrogenic differentiation of hBM-MSCs used in this

study was initially conrmed by the ndings of immunohistochemical assessments and the RT-PCR analysis in
the pellet culture study, which is consistent with previous
studies (Yoo et al., 1998). This study provides supporting
evidence for the hypothesis that hBM-MSCs are able to
undergo chondrogenic differentiation in the three-dimensional agarose culture with the treatment of TGF-3. The
results of RT-PCR analysis showed that the TGF-3treated specimens expressed chondrogenic markers of collagen type II and aggrecan, wherein the deposition of
collagen type II proteins was detected by immunohistochemical assessments. These ndings suggest that the
agarose gel is able to provide a supportive environment for
chondrogenesis of hBM-MSCs. To the best of our knowledge, this study is the rst to demonstrate chondrogenesis
of hBM-MSCs in agarose cultures.
The study of Yoo et al. (1998) demonstrated that the
monolayer cultures of hBM-MSCs with the treatment of
TGF-1 did not exhibit the deposition of proteoglycans,
whereas the collagen type II proteins and proteoglycans
were accumulated in the TGF-1-treated pellet cultures of
hBM-MSCs. Other previous studies also demonstrated the

Fig. 1. Typical immunohistochemical analysis of collagen type II

proteins on the hBM-MSC pellet sections of the control group (a, 40;
c, 400) and the TGF-3-treated group (b, 40; d, 400) after a 21-day
culture (a and b, scale bar 150 m; c and d, scale bar 20 m). The
TGF-3-treated specimen exhibits the deposition of collagen type II that
is illustrated as dark brown staining.
Fig. 2. Typical histological analysis of proteoglycans on the hBMMSC pellet sections of the control group (a, 40; c, 400) and the
TGF-3-treated group (b, 40; d, 400) after a 21-day culture (a and b,
scale bar 150 m; c and d, scale bar 20 m). The TGF-3-treated
specimen deposits more proteoglycans, as manifested by dark blue
staining.

433

induction of chondrogenic differentiation of hBM-MSCs in

the polylactide/aliginate amalgam with the treatment of
TGF-1 (Caterson et al., 2001) and the gelatin-based
resorbable sponge with the treatment of TGF-3 (Ponticiello et al., 2000). These studies indicate that suspension
culture may play an important role in promoting chondrogenesis of hBM-MSCs. Indeed, suspension culture has
been shown to maintain the phenotypic expression of
chondrocytes. Benya and Shaffer (1982) found that rabbit
articular chondrocytes lost the differentiated phenotype
and expressed the broblastic phenotype during serial
monolayer culture, whereas dedifferentiated chondrocytes
reexpressed the differentiated phenotype during suspension cultures in agarose gels.
In this study, the initial cell-seeding density of cellagarose constructs was found to inuence chondrogenic
differentiation of hBM-MSCs, as demonstrated by stronger expression of collagen type II gene exhibited in the
cell-agarose constructs of higher initial cell-seeding density. The expression of aggrecan gene was only detected in
the cell constructs of 9 106 cells/ml. This nding is
consistent with the study of Ponticiello et al. (2000), who
found that hBM-MSCs cultured in the gelatin matrix with
a higher initial cell-seeding density produced more proteoglycans in the presence of TGF-3. Since a previous study
had noted that the interactions between cells were important for in vitro chondrogenic differentiation and maintenance of chondrocyte phenotype (Solursh, 1991), the high
cell-seeding density may improve the cell interactions that
promote chondrogenic differentiation of hBM-MSCs. Cells
are not uniformly dispersed when cast into the cell-agarose constructs. Microscopic examination revealed cell aggregates throughout the agarose gel. Not surprising, the
number of cell aggregates increases with higher cell numbers; perhaps this may account for higher efciency of
induction by high cellseeding density. Moreover, extracellular matrix proteins were deposited around individual
cells in agarose cultures, high cell-seeding density may
facilitate the integration of extracellular matrices secreted
by individual cells to form a cartilaginous tissue.
Collagen type I gene was found to be expressed in the
TGF-3-treated pellets and cell-agarose constructs, suggesting that hBM-MSCs have not been completely transformed into chondrocyte-like cells after the 21-day treatment of TGF-3. The TGF-3-treated pellets also
exhibited the deposition of collagen type I proteins after a
21-day culture, indicating that incomplete chondrogenic
transformation of hBM-MSCs may produce an extracellular matrix of different collagen bers. Indeed, the abundant deposition of collagen type I proteins was found in
the brocartilaginous tissues that are usually regenerated
at injury sites of articular cartilage by bone marrow cells
recruited by penetrating subchondral bone (Furukawa et
al., 1980; Johnson, 1986). These reparative brocartilaginous tissues were found to be functionally inferior to articular cartilage and degenerate over time (Coletti et al.,
1972; Furukawa et al., 1980). Therefore, the acceleration
and efciency of chondrogenic transformation of hBMMSCs are important subjects for the clinical applications
of hBM-MSCs on the repair of articular cartilage.
It has been demonstrated that compressive loading not
only stimulated chondrogenic differentiation of chick embryonic limb-bud cells (Elder et al., 2000), but also modulated the cartilage matrix biosynthesis of mature chondrocytes (Sah et al., 1989; Buschmann et al., 1995; Lee and

434

HUANG ET AL.

Fig. 4. Typical phase-contrast light micrographs of hBM-MSCs cultured in the agarose gel at the initial seeding density of 9 106 cells/ml
(a) without the treatment of TGF-3 and (b) with the treatment of TGF-3
for 21 days (400; scale bar 20 m). The hBM-MSCs are able to
proliferate and deposit extracellular matrix to form larger cellular aggregates (arrowhead) within the agarose gel with the treatment of TGF-3.
Typical immunohistochemical analysis of collagen type II on the typical
cellular aggregate formed by hBM-MSCs in the cell-agarose constructs
of (c) the control group and (d) the TGF-3-treated group after a 21-day
culture (400; scale bar 20 m). Immunopositive reaction for collagen

type II proteins (arrow) illustrated by dark brown staining is only seen

within the cellular aggregate in the TGF-3-treated specimen. Nuclei
(arrowhead) were counterstained using hematoxyline. Typical alcian blue
staining of proteoglycans on the typical cellular aggregate formed by
hBM-MSCs in the cell-agarose constructs of (e) the control group and (f)
the TGF-3-treated group after a 21-day culture (400; scale bar 20
m). The deposition of proteoglycans (arrow) illustrated by strong blue
staining was detected outside and inside the cellular aggregate in the
TGF-3-treated specimen. Nuclei (arrow head) were counterstained using nuclear fast red.

Bader, 1997; Knight et al., 1998; Mauck et al., 2000;

Roberts et al., 2001). In addition, a recent study transplanted BM-MSCs to large full-thickness cartilage defects
of rabbit knee joints and showed that 6 months after the

implantation, the defects were repaired by regenerated

hyaline-like cartilage, which exhibited different mechanical properties on posterior and anterior aspects of the joint
(Wakitani et al., 1994). It suggested that mechanical dif-

435

CHONDROGENESIS OF hBM-MSC

chanical loading on chondrogenesis of hBM-MSCs using

agarose cultures.
In summary, this study has demonstrated that chondrogenesis of hBM-MSCs can be induced in agarose gels with
the treatment of TGF-3, as attested by the results of gene
expression analyses and histological and immunohistochemical assessments. The agarose culture study of hBMMSCs also demonstrated the effect of initial cell-seeding
density on chondrogenesis of hBM-MSCs. Since agarose
gels are mechanically stable, this study has established a
basic agarose culture system for further investigations of
the effects of mechanical loading on chondrogenesis of
hBM-MSCs. Furthermore, the pellet culture study conrms the chondrogenic potential of hBM-MSCs and shows
the mixing deposition of collagen type I and II proteins
during chondrogenic differentiation of hBM-MSCs. It explains that collagen type I and II proteins are found in
reparative brocartilaginous tissues that are usually
formed by bone marrow cells when injury on articular
cartilage penetrates the subchondral bone. Therefore,
complete chondrogenic transformation of hBM-MSCs
could be an important factor for regeneration of functional
articular cartilage using hBM-MSCs.

ACKNOWLEDGMENTS
Fig. 5. Typical RT-PCR analysis of cartilage-specic genes and
-actin gene on cell-agarose constructs of hBM-MSCs with the initial
seeding density of 9 106 cells/ml after a 21-day culture.

We thank Ms. Leonor Wenger and Mr. Felix Soto for

their technical assistance with cell cultures and histological and immunohistochemical analyses and Ms. Kristen
Hagar, Lauren Frost, and Marella Kurikose for their critical reading of the article.

LITERATURE CITED

Fig. 6. Typical comparison on the normalized IDVs of PCR products

for the collagen type II gene between the TGF-3-treated groups of
different initial cell-seeding densities. Cells were obtained from a 24year-old female donor.

ferences might result from substantially different mechanical environments at two locations (Wakitani et al.,
1994). The cartilage repair studies of autogenous periosteal grafts in which MSCs reside have demonstrated that
continuous passive motion enhanced the repair of fullthickness defect of articular cartilage (ODriscoll et al.,
1986). Based on these previous studies, mechanical stimuli may inuence chondrogenesis of BM-MSCs as well as
the deposition of extracellular matrix secreted by transformed cells. Agarose cultures have been widely used to
study the effects of mechanical loading on the responses of
chondrocytes due to their mechanical stability. Thus, this
study establishes a basic model of agarose culture of hBMMSCs for future studies to investigate the effects of me-

Angele P, Kujat R, Nerlich M, Yoo J, Goldberg V, Johnstone B. 1999.

Engineering of osteochondral tissue with bone marrow mesenchymal progenitor cells in a derivatized hyaluronan-gelatin composite
sponge. Tissue Eng 5:545554.
Aydelotte MB, Kuettner KE. 1988. Differences between sub-populations of cultured bovine articular chondrocytes: I, morphology and
cartilage matrix production. Connect Tissue Res 18:205222.
Benya PD, Shaffer JD. 1982. Dedifferentiated chondrocytes reexpress
the differentiated collagen phenotype when cultured in agarose
gels. Cell 30:215224.
Buschmann MD, Gluzband YA, Grodzinsky AJ, Kimura JH, Hunziker
EB. 1992. Chondrocytes in agarose culture synthesize a mechanically functional extracellular matrix. J Orthop Res 10:745758.
Buschmann MD, Gluzband YA, Grodzinsky AJ, Hunziker EB. 1995.
Mechanical compression modulates matrix biosynthesis in chondrocyte/agarose culture. J Cell Sci 108:14971508.
Caterson EJ, Nesti LJ, Li WJ, Danielson, KG, Albert TJ, Vaccaro AR,
Tuan RS. 2001. Three-dimensional cartilage formation by bone
marrow-derived cells seeded in polylactide/alginate amalgam.
J Biomed Mater Res 57:394 403.
Cheung HS, Cottrell WH, Stephenson K, Nimni ME. 1978. In vitro
collagen biosynthesis in healing and normal rabbit articular cartilage. J Bone Joint Surg Am 60:1076 1081.
Cheung HS, Lynch KL, Johnson RP, Brewer BJ. 1980. In vitro synthesis of tissue-specic type II collagen by healing cartilage: I,
short-term repair of cartilage by mature rabbits. Arthrit Rheum
23:211219.
Coletti JM Jr, Akeson WH, Woo SL. 1972. A comparison of the
physical behavior of normal articular cartilage and the arthroplasty
surface. J Bone Joint Surg Am 54:147160.
Diduch DR, Jordan LC, Mierisch CM, Balian G. 2000. Marrow stromal cells embedded in alginate for repair of osteochondral defects.
Arthroscopy 16:571577.
DIppolito G, Schiller PC, Ricordi C, Roos BA, Howard GA. 1999.
Age-related osteogenic potential of mesenchymal stromal stem cells

436

HUANG ET AL.

from human vertebral bone marrow. J Bone Miner Res 14:1115

1122.
Elder SH, Kimura JH, Soslowsky LJ, Lavagnino M, Goldstein SA.
2000. Effect of compressive loading on chondrocyte differentiation
in agarose cultures of chick limb-bud cells. J Orthop Res 18:78 86.
Erickson GR, Gimble JM, Franklin DM, Rice HE, Awad H, Guilak F.
2002. Chondrogenic potential of adipose tissue-derived stromal cells
in vitro and in vivo. Biochem Biophys Res Commun 290:763769.
Freije JM, Diez-Itza I, Balbin M, Sanchez LM, Blasco R, Tolivia J,
Lopez-Otin C. 1994. Molecular cloning and expression of collagenase-3, a novel human matrix metalloproteinase produced by
breast carcinomas. J Biol Chem 269:16766 16773.
Furukawa T, Eyre DR, Koide S, Glimcher MJ. 1980. Biochemical
studies on repair cartilage resurfacing experimental defects in the
rabbit knee. J Bone Joint Surg Am 62:79 89.
Johnson LL. 1986. Arthroscopic abrasion arthroplasty historical and
pathologic perspective: present status. Arthroscopy 2:54 69.
Knight MM, Lee DA, Bader DL. 1998. The inuence of elaborated
pericellular matrix on the deformation of isolated articular chondrocytes cultured in agarose. Biochim Biophys Acta 1405:6777.
Lee DA, Bader DL. 1997. Compressive strains at physiological frequencies inuence the metabolism of chondrocytes seeded in agarose. J Orthop Res 15:181188.
Mackay AM, Beck SC, Murphy JM, Barry FP, Chichester CO, Pittenger MF. 1998. Chondrogenic differentiation of cultured human
mesenchymal stem cells from marrow. Tissue Eng 4:415 428.
Mauck RL, Soltz MA, Wang CC, Wong DD, Chao PH, Valhmu WB,
Hung CT, Ateshian GA. 2000. Functional tissue engineering of
articular cartilage through dynamic loading of chondrocyte-seeded
agarose gels. J Biomech Eng 122:252260.
ODriscoll SW, Keeley FW, Salter RB. 1986. The chondrogenic potential of free autogenous periosteal grafts for biological resurfacing of
major full-thickness defects in joint surfaces under the inuence of
continuous passive motion: an experimental investigation in the
rabbit. J Bone Joint Surg Am 68:10171035.
Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca
JD, Moorman MA, Simonetti DW, Craig S, Marshak DR. 1999.
Multilineage potential of adult human mesenchymal stem cells.
Science 284:143147.

Ponticiello MS, Schinagl RM, Kadiyala S, Barry FP. 2000. Gelatinbased resorbable sponge as a carrier matrix for human mesenchymal stem cells in cartilage regeneration therapy. J Biomed Mater
Res 52:246 255.
Rahfoth B, Weisser J, Sternkopf F, Aigner T, Von der MK, Brauer R.
1998. Transplantation of allograft chondrocytes embedded in agarose gel into cartilage defects of rabbits. Osteoarthr Cartil 6:50 65.
Roberts SR, Knight MM, Lee DA, Bader DL. 2001. Mechanical compression inuences intracellular Ca2 signaling in chondrocytes
seeded in agarose constructs. J Appl Physiol 90:13851391.
Sah RL, KimYJ, Doong JY, Grodzinsky AJ, Plaas AH, Sandy JD.
1989. Biosynthetic response of cartilage explants to dynamic compression. J Orthop Res 7:619 636.
Solchaga LA, Dennis JE, Goldberg VM, Caplan AI. 1999. Hyaluronic
acid-based polymers as cell carriers for tissue-engineered repair of
bone and cartilage. J Orthop Res 17:205213.
Solursh M. 1991. Formation of cartilage tissue in vitro. J Cell Biochem
45:258 260.
Su MW, Lee B, Ramirezf F, Machado M, Horton M. 1989. Nucleotide
sequence of the full length cDNA encoding for human type II procollagen. Nucl Acids Res 17:9473.
Wakitani S, Goto T, Pineda SJ, Young RG, Mansour JM, Caplan AI,
Goldberg VM. 1994. Mesenchymal cell-based repair of large, fullthickness defects of articular cartilage. J Bone Joint Surg Am
76:579 592.
Wakitani S, Imoto K, Yamamoto T, Saito M, Murata N, Yoneda M.
2002. Human autologous culture expanded bone marrow mesenchymal cell transplantation for repair of cartilage defects in osteoarthritic knees. Osteoarthr Cartil 10:199 206.
Williams CG, Kim TK, Taboas A, Malik A, Manson P, Elisseeff J.
2003. In vitro chondrogenesis of bone marrow-derived mesenchymal stem cells in a photopolymerizing hydrogel. Tissue Eng 9:679
688.
Yoo JU, Barthel TS, Nishimura K, Solchaga L, Caplan AI, Goldberg
VM, Johnstone B. 1998. The chondrogenic potential of human bonemarrow-derived mesenchymal progenitor cells. J Bone Joint Surg
Am 80:17451757.