Extracted Text

Transforming Growth Factor-
-mediated Chondrogenesis of Human
Mesenchymal Progenitor Cells Involves N-cadherin and Mitogen-
activated Protein Kinase and Wnt Signaling Cross-talk*
Received for publication, May 20, 2003, and in revised form, July 30, 2003
Published, JBC Papers in Press, July 31, 2003, DOI 10.1074/jbc.M305312200
Richard Tuli‡§¶, Suraj Tuli‡, Sumon Nandi‡, Xiaoxue Huang‡, Paul A. Manner‡
,
William J. Hozack§, Keith G. Danielson§, David J. Hall‡, and Rocky S. Tuan‡§

**
From the‡Cartilage Biology and Orthopaedics Branch, NIAMS, National Institutes of Health, Department of Health and
Human Services, Bethesda, Maryland 20892, the§Department of Orthopaedic Surgery, Thomas Jefferson University,
Philadelphia, Pennsylvania 19107, and the

Department of Orthopaedic Surgery, George Washington University,
Washington, D. C. 20037
The multilineage differentiation potential of adult
tissue-derived mesenchymal progenitor cells (MPCs),
such as those from bone marrow and trabecular bone,
makes them a useful model to investigate mechanisms
regulating tissue development and regeneration, such
as cartilage. Treatment with transforming growth fac-
tor-

(TGF-
) superfamily members is a key require-
ment for thein vitrochondrogenic differentiation of
MPCs. Intracellular signaling cascades, particularly
those involving the mitogen-activated protein (MAP)
kinases, p38, ERK-1, and JNK, have been shown to be
activated by TGF-

s in promoting cartilage-specific
gene expression. MPC chondrogenesis in vitroalso re-
quires high cell seeding density, reminiscent of the
cellular condensation requirements for embryonic
mesenchymal chondrogenesis, suggesting common
chondro-regulatory mechanisms. Prompted by recent
findings of the crucial role of the cell adhesion protein,
N-cadherin, and Wnt signaling in condensation and
chondrogenesis, we have examined here their involve-
ment, as well as MAP kinase signaling, in TGF-

1-
induced chondrogenesis of trabecular bone-derived
MPCs. Our results showed that TGF-

1 treatment ini-
tiates and maintains chondrogenesis of MPCs through
the differential chondro-stimulatory activities of p38,
ERK-1, and to a lesser extent, JNK. This regulation of
MPC chondrogenic differentiation by the MAP kinases
involves the modulation of N-cadherin expression lev-
els, thereby likely controlling condensation-like cell-
cell interaction and progression to chondrogenic dif-
ferentiation, by the sequential up-regulation and
progressive down-regulation of N-cadherin. TGF-

1-
mediated MAP kinase activation also controls WNT-7A
gene expression and Wnt-mediated signaling through
the intracellular

-catenin-TCF pathway, which likely
regulates N-cadherin expression and subsequent N-
cadherin-mediated cell-adhesion complexes during
the early steps of MPC chondrogenesis.
Adult-derived mesenchymal progenitor cells (MPCs)
1
have
been considered a candidate cell source for tissue engineering
and reparative medicine by virtue of their potential to differ-
entiate into adipocytes, chondrocytes, fibroblasts, osteoblasts,
marrow stromal cells, and other tissues of mesenchymal origin
(1, 2). Numerous adult tissues have been identified that harbor
MPCs, including bone marrow (3–5), muscle (6 – 8), adipose
tissue (9, 10), periosteum (11, 12), and most recently in our
laboratory, human trabecular bone (13, 14). We have also de-
veloped improved techniques for the isolation and culture of
MPCs from trabecular bone to yield clinically significant num-
bers of such cells (15). That these cells retain their multilineage
differentiation potential through long term culture expansion
suggests they are a suitable cell source for potential therapeu-
tic and clinical treatment at least in osteogenic, adipogenic, and
chondrogenic applications.
Thein vitrochondrogenic differentiation of MPCs requires
the complex involvement of growth factors and cell-cell and
cell-matrix interactions, similar to developmental chondrogen-
esisin vivo(16). Expression of members of the transforming
growth factor-

(TGF-
) superfamily of growth factors has been
localized to sites of bone repair as well as sites of embryonic
bone and cartilage formationin vivo(17, 18), and the chondro-
inductive effects of the TGF-

superfamily members, particu-
larly the bone morphogenetic proteins and the TGF-

s, have
been well established in embryonic and adult mesenchymal
cells (13, 19 –24). Recent reports have demonstrated the critical
roles of intracellular signaling cascades activated by TGF-


family members in promoting cartilage-specific gene expres-
sion (25–27), including the mitogen-activated protein (MAP)
kinases, whose major subtypes include p38, extracellular sig-
nal-regulated kinase-1 (ERK-1), and c-Jun N-terminal kinase
(JNK or stress-activated protein kinase). These major subtypes
are activated by a variety of stimuli, and often are differentially
regulated by a single stimulus. The roles of the specific MAP
kinase subtypes, p38 and ERK-1, in regulating chondrogenesis
were elucidated to some degree by Ohet al.(28) who found that
p38 and ERK-1 have opposing roles during the chondrogenic
induction of chick limb bud cells. Specifically, p38 was an
* The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. Section 1734 solely to
indicate this fact.
¶Supported in part by a Percival E. and Ethel Brown Foerderer
Foundation fellowship from Thomas Jefferson University.
** To whom correspondence should be addressed: Cartilage Biology
and Orthopaedics Branch, NIAMS, National Institutes of Health, Bldg.
50, Rm. 1503, 50 South Dr., MSC 8022, Bethesda, MD 20892-8022. Tel.:
301-451-6854; Fax: 301-435-8017; E-mail: Tuanr@mail.nih.gov.
1
The abbreviations used are: MPCs, mesenchymal progenitor cells;
GAPDH, glyceraldehyde-3-phosphate dehydrogenase; COMP, cartilage oligomeric matrix protein; IGF I, insulin-like growth factor I; TGF-

,
transforming growth factor-

; JNK, c-Jun N-terminal kinase; MAP,
mitogen-activated protein; UTR, untranslated region; ERK-1, extracel- lular signal-regulated kinase-1; PBS, phosphate-buffered saline; TCF, T-cell factor; RT, reverse transcription; A-CAM, anti-A cell adhesion molecule; P, phosphorylated; ECM, extracellular matrix.
THEJOURNAL OFBIOLOGICALCHEMISTRY Vol. 278, No. 42, Issue of October 17, pp. 41227–41236, 2003
Printed in U.S.A.
This paper is available on line at http://www.jbc.org 41227
by guest on March 10, 2019http://www.jbc.org/Downloaded from

enhancer of chondrogenesis, whereas ERK-1 was a repressor of
chondrogenesis, and control was exerted, at least in part,
through the regulation of cell adhesion molecules, including
N-cadherin, fibronectin, and its receptor

5

1integrin, during
cellular condensation.
Precartilage mesenchymal condensation, a requisite for the
initiation of chondrogenesisin vivo
equally important inin vitrocultures (29–31). The spatiotem-
poral expression pattern of the Ca
2

-dependent, homotypic cell
adhesion molecule N-cadherin parallels its functional require-
ment for the initiation and subsequent progression of develop-
mental chondrogenesis (32, 33). As a transmembrane glycopro-
tein, N-cadherin is composed of extracellular domains that
mediate homophilic interactions between neighboring cells,
predominantly via a peptide domain containing the His-Ala-
Val (HAV) amino acid sequence, located near the N terminus of
the protein within the interface of two molecules and shown to
be critical for N-cadherin-mediated cell adhesion (34). Addi-
tionally, the cytoplasmic domain of N-cadherin is anchored to
the intracellular actin cytoskeleton through interactions with
the
-,
-, and-catenin complex. Besides its functional role,
cytoplasmic

-catenin has also been found to interact with
other proteins such as glycogen synthase kinase-3

, adenoma-
tous polyposis coli, the scaffolding components, axin and con-
ductin, as well as the transcriptional regulators, lymphoid
enhancing factor-1 (LEF-1)/T-cell factor (TCF), all of which
play critical roles in the canonical Wnt signal transduction
pathway (35), recently implicated in regulating chondrocyte
differentiation.
Wnts are a family of secreted glycoproteins that act in a
paracrine fashion, thereby mediating cellular interactions dur-
ing development (35, 36). Briefly, Wnt signaling proteins act by
binding to Frizzled receptors, the activation of which leads to
the stabilization of cytosolic

-catenin. Interaction of
-catenin
with the high mobility group box transcription factors of the
LEF-1/TCF family allows translocation of the complex into the
nucleus to subsequently regulate the transcription of Wnt tar-
get genes (37). A number ofWNTgenes are expressed during
development, includingWNT3A(38) localized in mouse apical
ectodermal ridge,WNT4localized in developing joints (39),
WNT5Alocalized in distal mesenchyme (40), and WNT7A in
dorsal ectoderm (41). Wnt-7a has been shown to be chondro-
inhibitoryin vitro(42), and recently the misexpression of
WNT7Ain limb mesenchymal chondrogenic cultures directly
led to the prolonged expression of N-cadherin, the stabilization
of N-cadherin-mediated cell-cell adhesion, and the eventual
inhibition of chondrogenesis (43, 44). The involvement of Wnt
signaling has also been shown in the BMP-2 mediated chon-
drogenic effect on the mouse C3H10T1/2 mesenchymal cell line
(45, 46).
Recent evidence has increasingly suggested that signaling
cascades and pathways act in an interconnected manner within
the cell. Here we attempt to define the early cellular processes
and cross-regulatory signaling events that take place during
mesenchymal chondrogenesis, specifically in adult tissue-de-
rived, multipotential MPCs. A recent report (47) described the
cross-regulation and subsequent inhibition of the Wnt signal-
ing pathway by MAPKKK and the downstream MAP kinase,
NLK. Here we report that TGF-

1-stimulated chondrogenesis
of trabecular bone-derived MPCs initiates intracellular signal-
ing via activation of the chondro-stimulatory MAP kinases,
p38, ERK-1, and JNK, which differentially regulate cartilage-
specific gene and protein expression in a lineage-specific man-
ner. Additionally, our results suggest that mesenchymal cell
condensation initiated by TGF-

1 within the pellet culture is
mediated via N-cadherin and is critical for the progression of
chondrogenesis, similar to developmental chondrogenesisin
vivo
lated at the cellular condensation phase by the tight control of
WNT7Agene expression individually by the p38, ERK-1, and
JNK MAP kinases.
EXPERIMENTAL PROCEDURES
Reagents—All reagents were purchased from Sigma unless otherwise
stated.
Isolation and Culture of Human Trabecular Bone-derived Cells—
Normal human trabecular bone was obtained from the femoral heads of
patients undergoing total hip arthroplasty and processed using a high
efficiency and high yield protocol established previously in our labora-
tory (15) and approved by the Institutional Review Boards of Thomas
Jefferson University and George Washington University. Processed
trabecular bone fragments were subsequently cultured in Dulbeccos
modified Eagle’s medium (high glucose and
L-glutamine; Mediatech,
Inc., Herndon, VA) supplemented with 10% fetal bovine serum (Premi-
um Select, Atlanta Biologicals, Atlanta, GA), from selected lots (48), and
50
g/ml penicillin/streptomycin. Subconfluent cell monolayers were
dissociated and removed using 0.25% trypsin containing 1 m
MEDTA
(Invitrogen) and either passaged at a ratio of 1:3 or utilized for study.
In Vitro Chondrogenic Differentiation of MPCs—
induction of trabecular bone-derived MPCs was initiated using high
density pellet cultures (210
5
cells/pellet, 500gfor 5 min) in a
chemically defined medium containing Dulbeccos modified Eagle’s me-
dium supplemented with 50
g/ml ascorbate, 0.1Mdexamethasone, 40
g/mlL-proline, 100g/ml sodium pyruvate, and ITS-plus (Collabora-
tive Biomedical Products, Cambridge, MA) (4, 13, 19, 23, 49). Recom-
binant human transforming growth factor-

1 (R&D Systems, Minne-
apolis, MN) was added to the pellet cultures at a final concentration of
10 ng/ml. For MAP kinase inhibition studies, specific chemical inhibi-
tors of p38 (SB20350), ERK-1-specific MAP kinase kinase (MEK1)
inhibitor (PD98059), and JNK (SP600125, Calbiochem-Novabiochem)
were used at final concentrations of 5, 10, and 100 n
M, respectively,
representing concentrations well within the range of IC
50values deter-
mined for similar cell types.
Reverse Transcription (RT)-PCR Analysis—
extracted using Trizol reagent (Invitrogen) according to the manufac-
turer’s protocol. Chondrogenic pellet cultures were first briefly homog-
enized in Trizol reagent to increase yield efficiency. Equal amounts of
RNA samples were reverse-transcribed by using random hexamers and
the SuperScript First Strand Synthesis System (Invitrogen). PCR am-
plification of cDNA was carried out using AmpliTaq DNA polymerase
(PerkinElmer Life Sciences) and the gene-specific primer sets listed in
Table I. Amplification cycles consisted of 1-min denaturation at 95°C,
1-min annealing, 1-min polymerization at 72°C, and finally a 10-min
extension at 72°C. The housekeeping gene, glyceraldehyde 3-phos-
phate dehydrogenase (GAPDH), was used as a control for RNA loading
of samples. PCR products were analyzed electrophoretically using the
Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA).
Reporter Gene Assays—
harvested at 60–80% confluence for electroporation using the Human
Mesenchymal Stem Cell Nucleofector kit (Amaxa Biosystems, Cologne,
Germany) according to a modified protocol established in our laboratory
for the efficient transfection of human MPCs.
2
Promoter reporter con-
structs include the following: 1) TOPFLASH, containing multimeric
TCF-binding sites (Upstate Cell Signaling Solutions, Waltham, MA); 2)
FOPFLASH, containing multimeric mutated TCF-binding sites (Up-
state Cell Signaling Solutions, Waltham, MA); 3) a pGL3 basic vector
(Promega, Madison, WI) containing 4.0 kb of the 5-flanking sequences
of the human collagen type II
1 procollagen (COL2A1, 577/
3428)
gene, which encompasses the promoter, exon 1, and the putative en-
hancer sequence in the first intron, linked to a luciferase reporter (a
kind gift from Dr. M. Goldring); and 4) a plasmid pAGC1(2368)/5-
UTR containing 2368 bp of the human aggrecan promoter region along
with the entire exon 1 and 5-UTR, linked to a luciferase reporter (a
kind gift from Dr. W. B. Valhmu). Luciferase activity was determined
using the Luciferase Assay System kit (Promega, Madison, WI). A green
fluorescent protein expression vector under the control of the SV40
promoter (pCMS-EGFP; Clontech, Palo Alto, CA) was used to normalize
transfection efficiencies. Results were analyzed using Students ttest,
p0.05.
2
H. Haleem-Smith, A. Derfoul, C. Okafor, R. Tuli, D. Olsen, D. J.
Hall, and R. S. Tuan, manuscript in preparation.
Molecular Signaling in Mesenchymal Chondrogenesis41228
by guest on March 10, 2019http://www.jbc.org/Downloaded from

Protein Isolation and Western AnalysisCell pellets were washed
twice with ice-cold phosphate-buffered saline (PBS), lysed with immu-
noprecipitation buffer (50 m
MTris-HCl, pH 7.4; 1% Nonidet P-40, 0.25%
sodium deoxycholate, 150 m
MNaCl, 1 mMEDTA) and protease and
phosphatase inhibitor mixture, incubated on ice for 30 min, homoge-
nized, and centrifuged at 14,000gfor 15 min. The supernatant was
collected, and protein concentrations were determined by using the
BCA assay (Pierce). Equal amounts of protein extracts were fraction-
ated by 10% SDS-PAGE, electroblotted onto Hybond-P membrane (Am-
ersham Biosciences), probed with antibodies to p38, P-p38, ERK1,
P-ERK1/2, JNK1, P-JNK, N-cadherin,

-actin (Santa Cruz Biotechnol-
ogy, Santa Cruz, CA), and

-catenin (Cell Signaling Technology,
Beverly, MA), immunoblotted using the ECF Western blotting kit ac-
cording to the manufacturers protocol (Amersham Biosciences), and
visualized using the Typhoon 9410 Imager (Amersham Biosciences).
Metabolic Sulfate Incorporation—Chondrogenic pellet cultures re-
ceived 1.0
Ci/ml sodium [
35
S]sulfate and 1.0Ci/ml [
3
H]leucine
(PerkinElmer Life Sciences) 24 h prior to the time point chosen for
measurement of newly synthesized proteoglycans and protein, respec-
tively. Incorporation of radioactivity was measured by liquid scintilla-
tion counting (51) and statistically analyzed by using Students ttest
(p0.05).
Histological and Immunohistochemical Analysis—Cell pellet cul-
tures, rinsed twice with PBS, were fixed fo
paraformaldehyde, dehydrated through a graded ethanol series, infil-
trated with isoamyl alcohol, embedded in paraffin, and sectioned at 8
m thickness for analysis. Histological staining with Alcian blue (pH
1.0) or hematoxylin and eosin was performed as described previously
(20, 31). Immunohistochemical localization of collagen type II (II-II6B3,
15
g/ml) and aggrecan (1-C-6, 10g/ml, Developmental Studies Hy-
bridoma Bank, Iowa City, IA) was performed by pre-digesting sections
with 300 units/ml hyaluronidase or 1.5 units/ml chondroitinase for 15
min at 37°C, respectively. Colorimetric detection of staining was per-
formed using Histostain-SP kit for DAB (Zymed Laboratories Inc., San
Francisco, CA). Cell pellets treated with the N-cadherin antibody, A-
CAM (see below), were fixed in 2% PBS-buffered paraformaldehyde and
stained directly using Alcian blue, pH 1.0.
Inhibition of N-cadherin-mediated Cell AdhesionA monoclonal an-
tibody to N-cadherin, anti-A cell adhesion molecule (A-CAM, Sigma),
was used to functionally inhibit homotypic interactions between N-
cadherin molecules during precartilage condensation. Trabecular bone-
derived MPCs were prepared for pellet culture as described above, and
A-CAM was added as a single dose at the beginning of culture at
concentrations varying from 0 to 240
g/ml. A nonspecific mouse anti-
body was used as a negative control. Chondrogenesis was assayed at
day 3 using the pAGC1(2368)/5-UTR promoter reporter construct,
and at day 21 by Alcian blue staining, as described above.
RESULTS
Activation of MAP Kinase Subtypes, p38, ERK-1, and JNK,
upon TGF-

1 StimulationTrabecular bone-derived cell pel-
lets treated with and without TGF-

1 were assayed for MAP
kinase activities and kinetics of activation. The addition of
TGF-

1 led to the rapid transient phosphorylation of p38,
ERK-1, and JNK, as determined by Western analysis (Fig. 1).
Phosphorylated p38 (P-p38) levels increased dramatically at
0.5 h, peaked at 1 h, and returned to basal levels by 2 h,
remaining constant through day 5 of chondrogenic culture.
This transient increase in protein levels upon TGF-

1 treat-
ment was kinetically mimicked by phosphorylated ERK-1 (P-
ERK-1), the major ERK isoform, as well as phosphorylated
JNK (P-JNK), which similarly increased at 0.5 h relative to
time 0 h, peaked at 1 h, and returned to basal levels by 2 h.
Unlike P-p38, however, P-ERK protein levels peaked again at
day 3, and P-JNK levels peaked again at day 1 as chondrogen-
esis proceeded. Reprobing the same blots with antibodies
against the unphosphorylated forms of p38, ERK-1, and JNK
MAP kinases revealed no change in total p38, ERK-1, or JNK
levels, respectively, during the chondrogenic culture period.
Activations of p38, ERK-1, and JNK were absent in untreated
control pellet cultures (data not shown).
p38, ERK-1, and JNK MAP Kinases Differentially Regulate
Chondrogenesis-associated Activities Stimulated by TGF-

1—
RT-PCR analysis of MPC pellets maintained in TGF-

1-sup-
plemented chondrogenic medium for 21 days showed signifi-
cant up-regulation of cartilage-specific gene expression (Fig.
2C), as compared with the untreated control (Fig. 2). The
addition of 5
Mp38 specific inhibitor (SB203580), 10M
MEK-1 inhibitor (PD98059), or 100 nMJNK-specific chemical
inhibitor (SP600125), to TGF-

1-treated pellet cultures led to
the lineage-specific down-regulation or complete abrogation of
chondrogenic gene expression levels (Fig. 2D). Inhibition of p38
with SB203580 completely abrogated TGF-

1-induced collagen
type II (COL2A1 SOX9gene expression, with significant
down-regulation of aggrecan expression, and reduction of col-
lagen type X (
factor I (IGF I) expression. Relative to the housekeeping gene,
GAPDH, collagen type IX (
trix protein (COMP) expression levels were unaffected by ad-
dition of the p38 inhibitor. ERK inhibition with PD98059 also
completely inhibited the TGF-

1-induced gene expression of
aggrecan, as well as collagen types II and IX andSOX9
regulated expression of collagen type X, dermatopontin, and
IGF I was also seen, whereas similar to the p38 inhibited
cultures, mRNA levels of GAPDH and COMP remained unal-
tered. Finally, JNK inhibition with SP600125 abolished TGF-

1-induced collagen type IX and Sox 9 mRNA expression, while
TABLEI
RT-PCR primers for differentiation-specific gene expression analysis, sequence and expected product size
Gene Primer sequences (5–3) Expected product size
bp
GAPDH
a
GGGCTGCTTTTAACTCTGGT 702
TGGCAGGTTTTTCTAGACGG
Aggrecan
b
TGAGGAGGGCTGGAACAAGTACCGGAGGTGGTAATTGCAGGGAACA 350
Col II
CAGGTCAAGATGGTCTTCAGCACCTGTCTCACCA 377
Col IX
GAAAATGAAGACCTGCTGGGAAAAGGCTGCTGTTTGGAGAC 516
Col X
GCCCAAGAGGTGCCCCTGGAATACCCTGAGAAAGAGGAGTGGACATAC 703
COMP
CAACTGTCCCCAGAAGAGCAATGGTAGCCAAAGATGAAGCCC 588
DMPN
TGGACCTCAGTCTTCTCTGGGTCCTAGCTAGCTTCAGAGCCG 547
IGF 1
TGTCCTCCTCGCATCTCTTCTGTACTTCCTTCTGGTCTTGGG 310
Sox 9
ATCTGAAGAAGGAGAGCGAGTCAGAAGTCTCCAGAGCTTG 264
WNT-3A
c
CAGGAACTACGTGGAGATCATGCCATCCCACCAAACTCGATGTC 326
WNT-5A
ACACCTCTTTCCAAACAGGCCGATTGTTAAACTCAACTCTC 392
WNT-7A
GCCGTTCACGTGGAGCCTGTGCGTGCAGCATCCTGCCAGGGAGCCCCGCAGCT 438
WNT-11
GTGAAGGACTCGGAACTCGTAGCGCTATGTCTGCAAGTGA 364
a
Control.
b
Cartilage-specific genes.
c
WNT-specific genes.
Molecular Signaling in Mesenchymal Chondrogenesis 41229
by guest on March 10, 2019http://www.jbc.org/Downloaded from

diminishing mRNA levels of aggrecan, collagen types II and X,
dermatopontin, and IGF I. GAPDH and COMP gene expression
were unaffected by addition of the JNK inhibitor, results that
were comparable with the p38 and ERK-1 MAP kinase studies,
suggesting their regulation by TGF-

1 through signaling cas-
cades other than p38, ERK-1, or JNK MAP kinase. Cell pellets
treated with 5
MSB203580, 10MPD98059, or 100 nM
SP600125 alone (Fig. 2) were unaffected similar to the nega-
tive control (Fig. 2). Experiments were repeated as described
above using additional inhibitors specific for the p38, ERK-1, or
JNK MAP kinase pathways; results obtained were similar to
those above (data not shown).
MPCs transiently transfected with either the pAGC1(2368)/
5-UTR-Luc promoter reporter (Fig. 2E) or the COL2a1-Luc
promoter reporter (Fig. 2) were cultured as pellets in TGF-

1
containing chondrogenic medium for 3 days, with or without
cotreatment with 5
MSB203580, 10MPD98059, or 100 nM
SP600125. Culture in medium without TGF-
1 was used as the
untreated control. TGF-

1 treatment of the pAGC1(2368)/5-
UTR-Luc transfected cell pellets resulted in an almost 2-fold
significant increase in relative luciferase activity as compared
with the control. Inhibition of p38 activation in TGF-

1-treated
cultures led to a significant 133% decrease in aggrecan pro-
moter activation as compared with TGF-

1 treated alone, and
a 71% decrease in activity as compared with the control. Sim-
ilarly, cotreatment of pellets with PD98059 resulted in a 94 and
32% decrease in TGF-

1-mediated transcriptional activation of
aggrecan as compared with TGF-

1 alone and the control,
respectively. Finally, the addition of JNK inhibitor, SP600125,
also suppressed TGF-

1-induced transcriptional activation by
127%, which was also 65% less than the control. These data
clearly indicate the lineage-specific positive control of aggrecan
promoter activation by the MAP kinases.
Analysis using theCOL2A1
sults (Fig. 2). As seen in the RT-PCR study, treatment of
pellets with TGF-

1 led to a 3.5-fold significant increase in
COL2a1transcriptional activation as compared with the un-
treated control. Addition of 5
MSB203580 to TGF-
1-treated
cultures resulted in a 172% decrease in relative luciferase
activity when compared with TGF-

1 treatment alone. The
addition of 10
MPD98059 and 100 nMSP600125 to TGF-
1-
treated cultures also repressedCOL2A1transcriptional activa-
tion by 264 and 203%, respectively. There were no significant
differences inCOL2a1transcriptional activation between TGF-

1-treated MAP kinase inhibited cultures and the untreated
control or pellets treated with inhibitors alone. The pCMS-
EGFP expression plasmid was used to normalize transfection
efficiencies in each experiment.
MAP Kinase Subtypes Individually Mediate Chondrogenic
Induction by TGF-

1—To elucidate further the functional in-
volvement of p38, ERK-1, and JNK MAP kinases in the TGF-

1-stimulated chondrogenic differentiation of trabecular bone-
derived MPCs, the effects of SB203580, PD98059, and
SP600125 added individually to TGF-

1-treated and untreated
pellet cultures were assessed by [
35
S]sulfate incorporation, as
an estimate of sulfated proteoglycan synthesis as a function of
time (Fig. 3). In cell pellets treated with TGF-

1, a significant
increase in the levels of [
35
S]sulfate incorporation was seen at
days 1, 14, and 21 as compared with all other cultures. The
level of sulfate incorporation in TGF-

1 cultures peaked at day
14 and remained stable thereafter, indicating a relatively
steady rate of incorporation from days 14 to 21. Inhibition of
MAP kinase activation in TGF-

1-stimulated pellet cultures
led to a significant decrease in the rates of sulfate incorporation
beginning at day 7 and continuing through day 21 of culture.
Levels of inhibition ranged from 35 44, 5758, and 39 to 25%,
from days 14 to 21 in p38, ERK-1, and JNK-inhibited cultures,
respectively. This inhibition was considered significant because
in the absence of TGF-

1 treatment, cell pellets treated with
individual MAP kinase inhibitors alone exhibited no differ-
ences as compared with the untreated control (data not shown).
Following 21 days of chondrogenic culture, the extent of
chondrogenesis was also assessed histologically using Alcian
blue and hematoxylin and eosin staining, as well as by immu-
nostaining for collagen type II and aggrecan (Fig. 4). Alcian
blue staining of cultures confirmed the [
35
S]sulfate incorpora-
tion results, showing an abundant sulfated proteoglycan-rich
cartilage-like matrix in the TGF-

1-treated cultures (Fig. 4),
which also contributed significantly to the dramatic increase in
pellet size as compared with control untreated cultures (Fig.
4A1). Inhibition of p38 MAP kinase in TGF-

1-treated cultures
using 5
MSB203580 led to a marked down-regulation in the
level of sulfated proteoglycan staining intensity (Fig. 4A3),
with resulting pellet size comparable with that of the untreated
control. Cotreatment with PD98059 cultures also resulted in
reduced staining intensity and pellet size (Fig. 4A4), with an
apparently distinct pattern of inhibition and staining relative
to other treatments. JNK-inhibited, TGF-

1-treated cultures
also exhibited lower levels of Alcian blue staining, with a less
elaborate proteoglycan matrix (Fig. 4); levels of staining
intensity in JNK-inhibited cultures, similar to the sulfate in-
corporation data of 21 days, were most comparable with TGF-

1-treated cultures, indicative of their less crucial role in reg-
ulating cartilage-specific proteoglycan production. Higher
magnification of Alcian blue-stained, TGF-

1-treated cultures
(Fig. 4B2) revealed a considerably more organized and struc-
tured matrix as compared with control (B1) and inhibitor co-
treated cultures (–B5). Hematoxylin and eosin-stained pel-
lets showed morphologically distinct, chondrocyte-like round
cells evenly distributed throughout pellets treated with
TGF-

1(C2), in contrast to the more fibroblast-like, elongated
cells in control (Fig. 4) and in MAP kinase inhibitor-treated
cultures. Additionally, eosin staining of acidophilic collagen
fibers demonstrates the abundant ECM elaborated by TGF-

1
cultures (Fig. 4) as compared with the untreated ( C1) and
MAP kinase-inhibited cultures (–C5).
Immunocytochemical detection of the cartilage-specific colla-
gen type II and aggrecan ECM molecules (Fig. 4,D1–D5 and
E1–E5) supported the RT-PCR gene expression data of 21 days
(Fig. 2,A–D). Collagen type II and aggrecan staining appear
more intense and extensively distributed throughout TGF-

1-
treated cultures (Fig. 4,D2andE2), as compared with their
respective controls ( andE1). Cotreatment with 5
MFIG.1.Temporal profiles of TGF-
1 activation of p38, ERK-1,
and JNK MAP kinases.Western analysis of fractionated lysates from
trabecular bone-derived MPC pellets treated with TGF-

1 and probed
for the phosphorylated forms of p38 (), ERK-1 (P-ERK-1), and
JNK (P-JNK). Blots were stripped and reprobed for total amounts of
p38, ERK-1, and JNK, included as controls d, day.
Molecular Signaling in Mesenchymal Chondrogenesis41230
by guest on March 10, 2019http://www.jbc.org/Downloaded from

SB203580, 10MPD98059, or 100 nMSP600125 in TGF-
1-
treated pellets led to a reduction of both collagen type II (Fig. 4,
D3–D5) and aggrecan (–E5) staining, indicative of the con-
trol exerted by the individual p38, ERK-1, and JNK MAP
kinase pathways in regulating TGF-
1-induced cartilage-spe-
cific ECM molecule production.
Effect of N-cadherin Inhibition on TGF-

1-stimulated Chon-
drogenesisA-CAM, a monoclonal antibody reactive with the
FIG.2.Effect of MAP kinase inhibitors on cartilage-specific gene expression.RT-PCR analysis of genes associated with chondrogenesis
(see Table I) and GAPDH was performed on day 21 cell pellets that were either left untreated (A), treated with MAP kinase inhibitors specific for
p38 (p38), ERK (ERK), and JNK (JNK), respectively (B), treated with TGF-

1(C), or exposed to concurrent administration of TGF-
1 and
individual MAP kinase inhibitors (D). The addition of TGF-

1 significantly up-regulated the expression of cartilage-specific genes as compared
with the control and inhibitor treatments alone, whereas simultaneous treatment with TGF-

1 and MAP kinase inhibitors either significantly
down-regulated or completely abrogated TGF-

1-induced gene expression in a lineage-specific manner. Regulation of aggrecan () and collagen
type II ()( F) gene expression in control, TGF-

1, and TGF-
1 plus MAP kinase inhibitor-treated cell pellets analyzed using promoter-
luciferase constructs. Significant up-regulation of aggrecan and collagen type II promoter activity was seen upon treatment with TGF-

1as
compared with respective controls, which was significantly inhibited upon cotreatment with MAP kinase inhibitors. *,p0.05, relative to control
cultures.COMP, cartilage oligomeric matrix protein.
FIG.3.Effect of MAP kinase inhibi-
tors on MPC pellet proteoglycan syn-
thesis as a function of time.Cultures
are designated as in Fig. 2. Chondrogen-
esis was assayed on days 1, 7, 14, and 21
for [
35
S]sulfate incorporation. Treatment
of cell pellets with TGF-

1 led to a signif-
icant increase in levels of sulfate incorpo-
ration at days 1, 14, and 21. Inhibition of
individual MAP kinases in the TGF-

1-
stimulated cultures resulted in a signifi-
cant decrease in the rates of sulfate incor-
poration beginning at day 7 and
continuing through day 21. *,p0.05,
relative to control and cotreated cultures.
Molecular Signaling in Mesenchymal Chondrogenesis 41231
by guest on March 10, 2019http://www.jbc.org/Downloaded from

N-terminal extracellular domain of N-cadherin, was used to
test the functional involvement of N-cadherin-mediated cell-
cell junction formation in TGF-

1-stimulated chondrogenesis
of MPCs (Fig. 5). pAGC1(2368)/5-UTR-transfected cells were
cultured as a cell pellet under chondrogenic conditions for 3
days with and without TGF-

1 and treated with a single dose
of A-CAM, at concentrations varying from 80 to 240
g/ml. The
results showed that TGF-

1-induced transcriptional activation
of the aggrecan promoter was significantly down-regulated by
the addition of 80
g/ml A-CAM (Fig. 5). Similar levels of
inhibition of luciferase activity were seen using 160 and 240
g/ml (Fig. 5), suggesting that TGF-
1 activation of the car-
tilage-specific aggrecan promoter and subsequent chondrogen-
esis is dependent on cell adhesion activities mediated by N-
cadherin. Histologically, as seen in Fig. 5, day 21 pellets
maintained in the presence of TGF-

1 and treated with 80
g/ml A-CAM revealed a significantly smaller pellet size as
well as less intense Alcian blue staining as compared with
culture treated with TGF-

1 alone, suggesting their inability to
elaborate a proteoglycan-rich matrix. At 240
g/ml A-CAM,
high density cell pellet formation was unsuccessful, presum-
ably as a result of extensive inhibition of cell-cell adhesion and
inability to form pre-cartilage condensation. Controls of TGF-

1-treated pellets incubated with nonspecific antibodies
showed no effect on chondrocytic phenotype following 21 days
of culture, strongly suggesting a specific role for N-cadherin-
mediated cell-cell interactions during precartilage condensa-
tion for the chondrogenic differentiation of MPCs.
MAP Kinase Regulation of N-cadherin in TGF-

1-stimulated
ChondrogenesisWe next investigated whether N-cadherin is
involved in the TGF-

1-mediated regulation of chondrogenesis.
After 1 day of pellet culture, N-cadherin protein levels were
markedly up-regulated in TGF-

1-treated cultures as deter-
mined by Western analysis (Fig. 6). In a temporal profile sug-
gestive of active involvement in precartilage cellular condensa-
tion, N-cadherin levels decreased slightly in these TGF-

1-
treated cultures by day 3, and returned to basal levels by day 5,
at the onset of overt chondrogenic differentiation. Inhibition of
p38, ERK-1, or JNK MAP kinases in TGF-

1-treated cultures
did not affect the elevated N-cadherin protein levels on day 1.
Interestingly, the N-cadherin levels in these MAP kinase-in-
hibited cultures remained high throughout the entire culture
period, suggesting the cell adhesion junctions were stabilized
in these cultures.

-Actin protein levels, included as internal
controls, remained constant through 5 days of pellet culture.
TGF-

1-stimulated Chondrogenesis Involves Regulation of
Wnt Signaling by MAP Kinases —We next investigated
whether TGF-

1-stimulated endogenous MAP kinase activa-
tion and signaling involves Wnt signal transduction. We ana-
lyzed the effects of 5
MSB203580, 10MPD98059, and 100 nM
SP600125 MAP kinase inhibitors on the transcriptional acti-
vation of a

-catenin-TCF-responsive luciferase reporter con-
struct (TOPFLASH) following 3 days of MPC pellet culture
(Fig. 7). Treatment of cell pellets with TGF-

1 resulted in a
significant 2-fold increase in luciferase activity relative to un-
treated, TOPFLASH-transfected pellet cultures. Cotreatment
with p38 inhibitor enhanced the TGF-

1-induced TCF-depend-
ent transcriptional activation by 3-fold relative to the un-
treated control. Similarly, the individual addition of ERK and
JNK inhibitors to TGF-

1 treated cultures led to a 4.5- and
4.6-fold increase, respectively, in luciferase activity relative to
the untreated control, levels that were also significantly higher
than TGF-

1-treated cultures alone. As expected, cells trans-
fected with the mutated reporter construct FOPFLASH showed
FIG.4.Inhibition of MAP kinases disrupts TGF-
1-induced chondrogenic phenotype. All cultures were treated as described and
examined histologically on day 21. Cultures are designated as in Fig. 2.AandB, Alcian blue staining;C, hematoxylin and eosin staining;D,
collagen type II immunostaining; andE, aggrecan immunostaining.A, Alcian blue staining reveals an abundant proteoglycan-rich cartilage-like
matrix in TGF-

1-treated pellets (A2), which contributes significantly to the dramatic increase in size as compared with control cultures ().
Cotreatment with MAP kinase inhibitors (A3–A5) drastically reduces the quantity of matrix proteoglycans as evidenced by the decrease in staining
intensity, as well as overall pellet sizes. Higher magnification of Alcian blue-stained, TGF-

1-treated cultures () reveals a much more organized
and ordered matrix, with higher staining intensity as compared with control and cotreated cultures (B1andB3–B5). Hematoxylin and
eosin-stained pellets show morphologically distinct, chondrocyte-like round cells throughout TGF-

1-treated cultures (), as compared with more
fibroblast-like, elongated cells in control (C1) and MAP kinase-treated cultures (C3–C5). Immunohistochemistry of cartilage matrix components
showed that collagen type II staining is significantly higher in cell pellets cultured with TGF-

1(D2) as compared with the respective control ().
Cotreatment of pellet cultures with MAP kinase inhibitors leads to a differential decrease in level of staining (–D5). Immunocytochemical
detection of aggrecan follows a similar pattern of staining (–E5). A,bar200
m;B–E, bar75 m.
Molecular Signaling in Mesenchymal Chondrogenesis41232
by guest on March 10, 2019http://www.jbc.org/Downloaded from

no differences in luciferase activity between control and treated
cultures (Fig. 7). Western analysis of

-catenin protein levels
following 3 days of pellet culture (Fig. 7) revealed a corre-
sponding increase in the nuclear pool of

-catenin in TGF-
1
cultures relative to the untreated control, which was consistent
with TOPFLASH activation. When both cytoplasmic and nu-
clear pools are considered, an even greater increase of both

-catenin levels was seen upon individual inhibition of p38,
ERK-1, or JNK MAP kinase in TGF-

1-induced cultures, con-
sistent with the significant increase in

-catenin-dependent
TCF activation seen in Fig. 7. These results suggest that
TGF-

1-stimulated chondrogenesis of MPCs is likely to involve
the negative regulation of Wnt signaling by the p38, ERK-1,
and JNK MAP kinases.
TGF-

1-stimulated Chondrogenesis Involves MAP Kinase
Regulation of WNT7A Gene Expression—To assess further the
involvement of Wnt, the effect of TGF-

1onWNT expression
was examined. RT-PCR analysis of MPC pellets maintained in
chondrogenic culture showed the up-regulated gene expression
ofWNT7Afollowing 1 day of TGF-

1 treatment (Fig. 8), as
compared with the untreated control (Fig. 8A). Specifically, the
increase in Wnt-7a mRNA levels was transient and was tem-
porally coincident with that of N-cadherin protein levels; by
day 3,WNT7Aexpression returned to basal levels where they
remained until at least day 5. The addition of MAP kinase
inhibitors (5
MSB203580, 10MPD98059, or 100 nM
SP600125) individually enhanced TGF-
1-stimulatedWNT7A
gene expression levels at day 1, and sustained these elevated
levels through day 5 of pellet culture (Fig. 8,C–E, respectively).
Similar analysis showed thatWNT3AandWNT11were not
expressed in the chondrogenic cultures and thatWNT5Agene
expression, although present, was not regulated by the addi-
tion of TGF-

1 or MAP kinase inhibitors. Expression of
WNT5Awas also unchanged upon treatment of cell pellets with
SB203580, PD98059, or SP600125 (data not shown). GAPDH
mRNA levels, included as internal controls, remained constant
through the 5-day culture period.
DISCUSSION
MPCs derived from human trabecular bone serve as a useful
model for the investigation of mechanisms responsible for the
generation, maintenance, and particularly the regeneration of
cartilage tissue (13, 15). By using this model system, in the
present study we have examined the mechanisms of TGF-

1-
mediated MPC chondrogenesis, specifically the involvement of
MAP kinase and Wnt signaling cascades. Our results show the
FIG.5.Effects of N-cadherin function-blocking antibody (A-CAM) on TGF-
1-induced chondrogenesis as analyzed by aggrecan
promoter luciferase activity (A) and Alcian blue staining (). A-CAM was added at varying concentrations to pellets of MPCs transfected
with an aggrecan promoter-luciferase construct with or without TGF-

1 treatment.A,treatment of cell pellets with 80 g/ml A-CAM for 3 days
significantly down-regulated TGF-

1-induced transcriptional activation of the aggrecan promoter. Similar inhibition was seen with both 160 and
240
g/ml A-CAM.B,Alcian blue staining of day 21 cell pellets maintained in the presence of TGF-
1 and treated with A-CAM led to significant
inhibition of proteoglycan production and reduction of pellet size at 80
g/ml A-CAM and complete inhibition of pellet formation and abrogation
of the chondrocytic phenotype at 240
g/ml.B,bar300 m.
Molecular Signaling in Mesenchymal Chondrogenesis 41233
by guest on March 10, 2019http://www.jbc.org/Downloaded from

requisite involvement of the p38, ERK-1, and JNK MAP kinase
cascades and their positive regulation of mesenchymal chon-
drogenesis induced by TGF-

1. Additionally, we demonstrate
the functional role of TGF-

1-stimulated N-cadherin expres-
sion in the chondrogenic differentiation of MPCs. Interestingly,
this control of N-cadherin expression by TGF-

1 occurs via
MAP kinase regulation with coincidentWNT7Agene expres-
sion and signaling, reminiscent of our recent findings in which
Wnt-7a signaling has been shown to inhibit chondrogenesis in
embryonic limb mesenchymal cultures by modulating N-cad-
herin expression and cell adhesion complexes (43).
MAP kinase signaling activities have been implicated in
many forms of cellular differentiation, including chondrogene-
sis of mesenchymal cells (28, 52). Our results indicate that the
transient activations of p38, ERK-1, and JNK are independ-
ently essential for the chondrogenic differentiation of adult
multipotent MPCs. The requirement of each of these MAP
kinase subtypes for chondro-stimulation is most clearly evident
by using inhibition studies. In the presence of individual MAP
kinase inhibitors, cartilage-specific gene expression induced by
TGF-

1 is differentially down-regulated or completely abro-
gated in a manner specific to each signaling pathway. Thus,
our results suggest that a more complete inhibition of TGF-

1-
induced chondrogenesis and gene expression in MPC pellet
cultures results from inhibition of p38 or ERK-1, compared
with JNK. In this manner, activation of each MAP kinase
pathway by TGF-

1 is required for and, as such, contributes
significantly to the thorough induction of chondrogenic differ-
entiation. It is also interesting to note that although the three
subtypes are activated by a single ligand (TGF-

1), their down-
stream transcriptional effects are markedly different. How-
ever, that the gene expression of the transcription factor Sox 9,
a critically important transcriptional regulator of cartilage-
specific genes and a potent inducer of the chondrocytic pheno-
type (53), is exclusively regulated by the p38, ERK-1, and JNK
subtypes suggests the importance of MAP kinase signaling in
chondrogenesis and the regulation of expression of chondro-
regulatory and chondrocyte-specific genes.
The regulated expression of cell adhesion molecules such as
N-cadherin, which functions during precartilage mesenchymal
condensation leading to subsequent progression to overt chon-
drogenic differentiation, has been well established in the chick
limb bud (29, 30, 32, 33) and murine C3H10T1/2 micromass
systems (31). We report here that the chondrogenic induction of
adult MPCs also requires the precise control of N-cadherin
expression, dependent on TGF-

1-initiated MAP kinase signal-
ing cascades, most likely to mediate the appropriate cell-cell
adhesion required for precartilage mesenchymal condensation
and ensuing differentiation. Thus, the rapid transient up-reg-
ulation of N-cadherin protein levels by day 1 in TGF-

1-treated
cultures initiates the requisite cell-cell interactions in precar-
tilage condensation. The subsequent down-regulation in N-
cadherin expression by day 5 is consistent with the increase in
production of ECM components and changes in cellular mor-
phology from fibroblast-like to round, morphologically distinct
chondrocytes, and a concomitant termination of N-cadherin-
mediated cell-cell interaction. The functional involvement of
N-cadherin-mediated activities in TGF-

1-induced chondro-
genesis is verified by the effect of treating the pellet cultures
with the N-cadherin-specific A-CAM antibody, resulting in a
significant reduction in aggrecan promoter activation following
3 days of culture, and decreased Alcian blue positive staining
and pellet size of 21 day cell pellets, as compared with pellets
treated with TGF-

1 alone or cotreated with a nonspecific
antibody.
Of the possible signaling mechanisms activated by TGF-

1
ligand binding that regulate N-cadherin expression during pre-
cartilage mesenchymal condensation, the p38, ERK-1, and
JNK MAP kinases were individually shown to be involved.
Based upon our results, we conclude that inhibition of MAP
kinase signaling in TGF-

1-treated cultures retarded the pro-
gression from precartilage condensation to overt chondrogenic
differentiation by sustaining N-cadherin expression and pre-
sumably stabilizing the cell-cell adhesion complexes. This was
evident upon microscopic examination of TGF-

1-treated, MAP
kinase-inhibited pellet cultures, which appeared to condense
normally from days 1 through 3, similar to pellets treated with
TGF-

1 alone. However, the cotreated pellets remained at this
condensed size and appeared not to proceed further along the
differentiation pathway, whereas the TGF-

1 cultures began to
elaborate a cartilage-specific ECM (data not shown). The addi-
tion of p38, ERK-1, and JNK inhibitors individually to TGF-

1-treated pellet cultures led to persistently high levels of
N-cadherin protein that failed to return to basal levels even
following 5 days of chondrogenic culture, thereby blocking
chondrogenic differentiation. Interestingly, De Lise and Tuan
(29) have found that transfection-mediated overexpression of
wild-type N-cadherin in primary chick limb mesenchymal cul-
tures allows cells to condense normally but inhibits subsequent
differentiation due to the persistence of increased cell-cell in-
teraction. Thus, our results here strongly implicate TGF-

1-
induced MAP kinase signaling through the p38, ERK-1, and
JNK cascades in the regulation of N-cadherin during the pro-
gression of prechondrogenic mesenchymal cells to differenti-
ated cartilage. However, the mechanisms by which this is ac-
complished are not understood.
We therefore investigated the possible regulation of the Wnt
signaling pathway by the MAP kinases. TOPFLASH-trans-
fected cell pellets treated with TGF-

1 showed significantly
increased transcriptional activation of the

-catenin-TCF-re-
sponsive reporter, as well as corresponding nuclear

-catenin
protein levels, indicating the activation of the canonical Wnt
signaling pathway in response to TGF-

1 ligand binding, and
suggesting a probable role for the Wnt cascade in mediating
chondrogenesis. Cotreatment with MAP kinase-specific inhib-
itors significantly enhanced the TGF-

1 stimulation of TCF-
dependent transcriptional activation, at least in part by further
elevating total

-catenin protein levels, whereas no significant
effects were noted using the control, mutated reporter con-
struct, FOPFLASH. These results indicate that the individual
FIG.6.Temporal profile of the effects of TGF-
1 and MAP
kinase inhibitor treatment on N-cadherin protein levels in MPC
pellet cultures as determined by Western analysis.Cultures are
designated as in Fig. 2. Compared with untreated controls, TGF-

1
treatment elicited a rapid transient increase in N-cadherin protein
expression with 24 h of TGF-

1 treatment, peaking at day 1, and
returning to basal levels by day 5 of pellet culture. Upon cotreatment
with inhibitors to p38, ERK-1, and JNK MAP kinases, the elevated
level of N-cadherin protein was sustained continuously through 5 days
of culture, as compared with TGF-

1 treatment alone. Loading was
normalized on the basis of

-actin level.
Molecular Signaling in Mesenchymal Chondrogenesis41234
by guest on March 10, 2019http://www.jbc.org/Downloaded from

activation of the p38, ERK-1, and JNK pathways by TGF-
1in
chondrogenic pellet cultures differentially represses the

-cate-
nin-mediated canonical Wnt signal to levels that allow the
pathway to remain active under TGF-

1 induction to promote
mesenchymal chondrogenesis.
Interestingly, our recent studies (43, 54) have shown that
Wnt-7a, a member of the Wnt signaling glycoprotein family, is
capable of interfering with the progression of limb mesenchy-
mal cells from precartilage condensation to overt chondrogenic
differentiationin vitroby modulating the expression of N-
cadherin mRNA and protein levels; this effect likely prolongs
the stabilization of N-cadherin-dependent intercellular junc-
tions, also seen when N-cadherin is overexpressed (29), and
misexpressed, similar to the effects of MAP kinase inhibition
reported here (Fig. 6). We therefore investigated the potential
regulation of gene expression of representativeWNTfamily
members from the two major functional groups, theWNT1and
WNT5Aclass, by TGF-
1 and MAP kinases. Fortuitously, we
found thatWNT7Agene expression was up-regulated in TGF-

1-treated cell pellets following 24 h of culture, diminishing to
basal levels by day 3, and remaining there through day 5 (Fig.
8); this transient up-regulation parallels N-cadherin expres-
sion levels in similarly treated cultures and, taken together,
suggest the regulation of N-cadherin by Wnt-7a. In this man-
ner, the process of cell-cell adhesion mediated by N-cadherin is
likely to be regulated by the action of Wnt-7a. This is consistent
with the role of Wnt-1 class signaling molecules shown to
induce the stabilization of

-and-catenin, proteins critical to
FIG.7.Regulation of TGF-
1-medi-
ated Wnt signaling by MAP kinases
on the basis of

-catenin-TCF pro-
moter activity and

-catenin stabili-
zation and nuclearization. Cultures
are designated as in Fig. 2.A, pellet cul-
tures treated with and without TGF-

1
for 3 days were assayed for Wnt signal-
mediated

-catenin-TCF-regulated tran-
scription using luciferase reporter plas-
mids containing intact (TOPFLASH) and
mutated (FOPFLASH) multimeric TCF-
binding sites. TGF-

1-induced transcrip-
tional activation of TOPFLASH is further
enhanced by inhibition of individual MAP
kinase signaling pathways, whereas no
effects are seen using the mutated re-
sponse element. Concomitant with TOP-
FLASH activation, an increase in the nu-
clear localization of

-catenin protein
levels is seen in TGF-

1-treated pellet
cultured for 3 days as compared with con-
trol cultures (B). Similarly, inhibition of
MAP kinases in TGF-

1-stimulated cul-
tures results in even higher levels of cy-
toplasmic and nuclear

-catenin protein.
FIG.8.Effects of p38, ERK, and JNK MAP kinase inhibitors on TGF-
1-mediated regulation ofWNTgene expression.Cultures are
designated as in Fig. 2. RT-PCR analysis was performed on cell pellets maintained with and without TGF-

1 for 5 days. Complementary DNA was
primed for Wnt-3a, -5a, -7a, and -11, as well as the internal control GAPDH. A transient increase inWNT-7Agene expression is seen following 1
day of culture in TGF-

1-treated cultures (B) returning to basal levels by day 3, as compared with the untreated control (). Following inhibition
of individual MAP kinases in TGF-

1-stimulated cultures (–E), Wnt-7a mRNA expression levels are further up-regulated at day 1 and sustained
through 5 days of pellet culture.WNT-5Aappears to be constitutively expressed in a manner unaffected by TGF-

1 treatment or MAP kinase
inhibition.
Molecular Signaling in Mesenchymal Chondrogenesis 41235
by guest on March 10, 2019http://www.jbc.org/Downloaded from

the formation of N-cadherin intercellular junctions, thereby
prolonging cell-cell adhesion (50). Moreover, the inhibition of
MAP kinases in TGF-

1-treated cultures significantly up-reg-
ulatedWNT7Agene expression levels through day 5 of culture
as compared with TGF-

1-induced levels (Fig. 8); in addition,
the effect of MAP kinases onWNTgene expression is totally
dependent on the presence of TGF-

1. These sustained levels of
WNT7Aexpression in MAP kinase inhibited TGF-

1-treated
pellet cultures, again analogous to N-cadherin expression lev-
els in similarly treated cultures, suggest the involvement of
MAP kinase signaling in the Wnt-7a regulation of N-cadherin-
mediated cell-cell junctions, required for cells to progress into
overt chondrogenic differentiation. These findings suggest that
a strictly regulated, tonically inhibited level of Wnt-7a, con-
trolled by the individual activation of the p38, ERK-1, and JNK
MAP kinase cascades by TGF-

1, is required for the progres-
sion of cells from mesenchymal condensation to overt differen-
tiation. Although Wnt-5a mRNA was expressed in all treat-
ment groups throughout the chondrogenic culture period, the
level of gene expression did not appear to be regulated by
TGF-

1 and/or the MAP kinases. Moreover, it is unlikely that
Wnt-5a contributed to the increase in

-catenin-TCF-depend-
ent Wnt signaling or the regulation of N-cadherin expression,
especially because the TOPFLASH response elements were
differentially regulated by both TGF-

1 as well as the MAP
kinases, unlikeWNT5Agene expression, and our previous
studies (43) have shown misexpression ofWNT5Ato have no
effect on N-cadherin expression during mesenchymal chondro-
genesis in chick limb bud micromass cultures.
In conclusion, we have demonstrated in this study that
TGF-

1 initiates and maintains chondrogenesis of trabecular
bone-derived MPCs through the differential yet well coordi-
nated chondro-stimulatory activities of p38, ERK-1/2, and to a
lesser extent JNK. This regulation of MPC differentiation by
the MAP kinases involves the indirect modulation of N-cad-
herin expression levels to control precartilage condensation
events and the progression to chondrogenic differentiation. A
target of MAP kinase activation in regulating the events of
cell-cell adhesion via N-cadherin is the control ofWNT7Agene
expression levels as well as subsequent Wnt-mediated signal-
ing through the intracellular

-catenin-TCF pathway, which is
likely to translate into the strict regulation of N-cadherin ex-
pression during condensation. Efforts are currently underway
to assess further the specificity of such a mechanistic pathway
by means of targeted perturbation of gene expression using
small interfering RNA technology, as well as to identify addi-
tional candidate regulatory genes by means of microarray-
based gene expression profiling.
Acknowledgments—We thank Dr. M. Goldring for the plasmid
(pCAT-B/4.0) and Dr. W. B. Valhmu for the plasmid pAGC1(23680/5-
UTR). The II-II6B3 and 1-C-6 monoclonal antibodies were obtained
from the Developmental Studies Hybridoma Bank developed under the
auspices of the NICHD of the National Institutes of Health and main-
tained by the University of Iowa, Department of Biological Sciences,
Iowa City, IA.
REFERENCES
1. Tuan, R., Boland, G., and Tuli, R. (2003)Arthritis Res.5,32–45
2. Caplan, A. I. (1994)Clin. Plast. Surg.21,429–435
3. Osyczka, A. M., Noth, U., OConnor, J., Caterson, E. J., Yoon, K., Danielson,
K. G., and Tuan, R. S. (2002)Calcif. Tissue Int.71,447 458
4. Pittenger, M. F., Mackay, A. M., Beck, S. C., Jaiswal, R. K., Douglas, R., Mosca,
J. D., Moorman, M. A., Simonetti, D. W., Craig, S., and Marshak, D. R.
(1999)Science284,143147
5. Jiang, Y., Jahagirdar, B. N., Reinhardt, R. L., Schwartz, R. E., Keene, C. D.,
Ortiz-Gonzalez, X. R., Reyes, M., Lenvik, T., Lund, T., Blackstad, M., Du, J.,
Aldrich, S., Lisberg, A., Low, W. C., Largaespada, D. A., and Verfaillie,
C. M. (2002)Nature418,41–49
6. Adachi, N., Sato, K., Usas, A., Fu, F. H., Ochi, M., Han, C. W., Niyibizi, C., and
Huard, J. (2002)J. Rheumatol.29,1920–1930
7. Bosch, P., Musgrave, D. S., Lee, J. Y., Cummins, J., Shuler, T., Ghivizzani,
T. C., Evans, T., Robbins, T. D., and Huard, J. (2000)J. Orthop. Res.18,
933944
8. Wada, M. R., Inagawa-Ogashiwa, M., Shimizu, S., Yasumoto, S., and Hashi-
moto, N. (2002)Development129,29872995
9. Zuk, P. A., Zhu, M., Mizuno, H., Huang, J., Futrell, J. W., Katz, A. J., Benhaim,
P., Lorenz, H. P., and Hedrick, M. H. (2001)Tissue Eng.7,211228
10. Gronthos, S., Franklin, D. M., Leddy, H. A., Robey, P. G., Storms, R. W., and
Gimble, J. M. (2001)J. Cell. Physiol.189,54–63
11. De Bari, C., DellAccio, F., and Luyten, F. P. (2001) Arthritis Rheum.44,85–95
12. Nakahara, H., Goldberg, V. M., and Caplan, A. I. (1991)J. Orthop. Res.9,
465 476
13. Noth, U., Osyczka, A. M., Tuli, R., Hickok, N. J., Danielson, K. G., and Tuan,
R. S. (2002)J. Orthop. Res.20,1060–1069
14. Osyczka, A. M., Noth, U., Danielson, K. G., and Tuan, R. S. (2002)Ann. N. Y.
Acad. Sci.961,73–77
15. Tuli, R., Seghatoleslami, M. R., Tuli, S., Wang, M. L., Hozack, W. J., Manner,
P. A., Danielson, K. G., and Tuan, R. S. (2003)Mol. Biotechnol.23,37–49
16. DeLise, A. M., Fischer, L., and Tuan, R. S. (2000)Osteoarthritis Cartilage8,
309–334
17. Horner, A., Kemp, P., Summers, C., Bord, S., Bishop, N. J., Kelsall, A. W.,
Coleman, N., and Compston, J. E. (1998)Bone23,95–102
18. Hogan, B. L. (1996)Genes Dev.10,1580–1594
19. Johnstone, B., Hering, T. M., Caplan, A. I., Goldberg, V. M., and Yoo, J. U.
(1998)Exp. Cell Res.238,265272
20. Denker, A. E., Haas, A. R., Nicoll, S. B., and Tuan, R. S. (1999)Differentiation
64,67–76
21. Sekiya, I., Colter, D. C., and Prockop, D. J. (2001)Biochem. Biophys. Res.
Commun.284,411 418
22. Majumdar, M. K., Wang, E., and Morris, E. A. (2001)J. Cell. Physiol.189,
275284
23. Mackay, A. M., Beck, S. C., Murphy, J. M., Barry, F. P., Chichester, C. O., and
Pittenger, M. F. (1998)Tissue Eng.4,415 428
24. Barry, F., Boynton, R. E., Liu, B., and Murphy, J. M. (2001)Exp. Cell Res.268,
189–200
25. Watanabe, H., de Caestecker, M. P., and Yamada, Y. (2001)J. Biol. Chem.276,
14466–14473
26. Tuli, R., Seghatoleslami, M. R., Tuli, S., Howard, M. S., Danielson, K. G., and
Tuan, R. S. (2002)Ann. N. Y. Acad. Sci.961,172177
27. Nakamura, K., Shirai, T., Morishita, S., Uchida, S., Saeki-Miura, K., and
Makishima, F. (1999)Exp. Cell Res.250,351363
28. Oh, C. D., Chang, S. H., Yoon, Y. M., Lee, S. J., Lee, Y. S., Kang, S. S., and
Chun, J. S. (2000)J. Biol. Chem.275,56135619
29. DeLise, A. M., and Tuan, R. S. (2002)J. Cell. Biochem.87,342359
30. Delise, A. M., and Tuan, R. S. (2002)Dev. Dyn.225,195204
31. Haas, A. R., and Tuan, R. S. (1999)Differentiation64,77–89
32. Oberlender, S. A., and Tuan, R. S. (1994)Cell Adhes. Commun.2,521537
33. Oberlender, S. A., and Tuan, R. S. (1994)Development120,177187
34. Blaschuk, O. W., Sullivan, R., David, S., and Pouliot, Y. (1990)Dev. Biol.139,
227229
35. Behrens, J. (2000)J. Cell Sci.113,911919
36. Cadigan, K. M., and Nusse, R. (1997)Genes Dev.11,3286–3305
37. Behrens, J., von Kries, J. P., Kuhl, M., Bruhn, L., Wedlich, D., Grosschedl, R.,
and Birchmeier, W. (1996)Nature382,638–642
38. Kengaku, M., Capdevila, J., Rodriguez-Esteban, C., De La Pena, J., Johnson,
R. L., Belmonte, J. C., and Tabin, C. J. (1998)Science280,1274–1277
39. Kawakami, Y., Wada, N., Nishimatsu, S. I., Ishikawa, T., Noji, S., and Nohno,
T. (1999)Dev. Growth Differ.41,29–40
40. Parr, B. A., Shea, M. J., Vassileva, G., and McMahon, A. P. (1993)Development
119,247261
41. Parr, B. A., and McMahon, A. P. (1995)Nature374,350–353
42. Rudnicki, J. A., and Brown, A. M. (1997)Dev. Biol.185,104–118
43. Tufan, A. C., and Tuan, R. S. (2001)FASEB J.15,1436–1438
44. Tufan, A. C., Daumer, K. M., DeLise, A. M., and Tuan, R. S. (2002)Exp. Cell
Res.273,197203
45. Fischer, L., Boland, G., and Tuan, R. S. (2002)J. Cell. Biochem.84,816–831
46. Fischer, L., Boland, G., and Tuan, R. S. (2002) J. Biol. Chem.277,
30870–30878
47. Ishitani, T., Ninomiya-Tsuji, J., Nagai, S., Nishita, M., Meneghini, M., Barker,
N., Waterman, M., Bowerman, B., Clevers, H., Shibuya, H., and Mat-
sumoto, K. (1999)Nature399,798–802
48. Caterson, E. J., Nesti, L. J., Danielson, K. G., and Tuan, R. S. (2002)Mol.
Biotechnol.20,245256
49. Yoo, J. U., Barthel, T. S., Nishimura, K., Solchaga, L., Caplan, A. I., Goldberg,
V. M., and Johnstone, B. (1998)J. Bone Joint Surg. Am.80,17451757
50. Shibamoto, S., Hayakawa, M., Takeuchi, K., Hori, T., Oku, N., Miyazawa, K.,
Kitamura, N., Takeichi, M., and Ito, F. (1994)Cell Adhes. Commun.1,
295305
51. San Antonio, J. D., and Tuan, R. S. (1986)Dev. Biol.115,313324
52. Yoon, Y. M., Oh, C. D., Kim, D. Y., Lee, Y. S., Park, J. W., Huh, T. L., Kang,
S. S., and Chun, J. S. (2000)J. Biol. Chem.275,1235312359
53. Bi, W., Deng, J. M., Zhang, Z., Behringer, R. R., and de Crombrugghe, B. (1999)
Nat. Genet.22,85–89
54. Stott, N. S., Jiang, T. X., and Chuong, C. M. (1999)J. Cell. Physiol.180,
314–324
Molecular Signaling in Mesenchymal Chondrogenesis41236
by guest on March 10, 2019http://www.jbc.org/Downloaded from

Hozack, Keith G. Danielson, David J. Hall and Rocky S. Tuan
Richard Tuli, Suraj Tuli, Sumon Nandi, Xiaoxue Huang, Paul A. Manner, William J.
Wnt Signaling Cross-talk
Progenitor Cells Involves N-cadherin and Mitogen-activated Protein Kinase and
-mediated Chondrogenesis of Human MesenchymalbTransforming Growth Factor-
doi: 10.1074/jbc.M305312200 originally published online July 31, 2003
2003, 278:41227-41236.J. Biol. Chem. 
 
10.1074/jbc.M305312200Access the most updated version of this article at doi:
 Alerts:
 
When a correction for this article is posted• 
When this article is cited• 
to choose from all of JBC's e-mail alertsClick here
 
http://www.jbc.org/content/278/42/41227.full.html#ref-list-1
This article cites 54 references, 13 of which can be accessed free at
by guest on March 10, 2019http://www.jbc.org/Downloaded from