Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

Activated b-catenin induces myogenesis andinhibits adipogenesis in BM-derived mesenchymal

stromal cells

YC Shang1, C Zhang1,2, SH Wang3, F Xiong2, CP Zhao1, FN Peng1, SW Feng1,

MJ Yu2, MS Li1, YN Zhang1 and Y Li2

1Department of Neurology, First Affiliated Hospital, Sun Yat-sen University, Guangzhou, PR China,2Center for Stem Cell Biology and Tissue Engineering, Sun Yat-sen University, Guangzhou, PR China, and

3Department of Neurology, Beijing Friendship Hospital, Capital Medical University, Beijing, PR China

Background

Mesenchymal stromal cells (MSC) have been thought to be attractive

candidates for the treatment of degenerative muscle diseases. However,

little is known about the molecular mechanisms governing the myogenic

differentiation in MSC. As the Wnt signaling pathway has been

associated with myogenesis in embryogenesis and post-natal muscle

regeneration, we hypothesized that the Wnt signaling pathway may be

involved in governing the myogenic differentiation in MSC.

Methods

Primary MSC were isolated from Sprague�Dawley rats and

expanded in proliferation medium. The rMSC were transfected with

a constitutively active hb-catenin (S37A) plasmid or control vector by

Lipofectamine followed by G418 selection. The transfected rMSC were

grown to 80% confluence and then cultured in myogenic or adipogenic

differentiation medium. Cells were characterized by light microscopy,

immunofluorescence and RT-PCR at different time points after

myogenic or adipogenic introduction.

Results

Ectopic expression of activated b-catenin located primarily in the

nucleus and activated transcription in rMSC. Overexpression of

stabilized b-catenin induced 27.193.91% rMSC forming long

multinucleated cells expressing MyoD, myogenin, desmin and myosin

heavy chain (MHC) via evoking the expression of skeletal muscle-

specific transcription factors. In addition, overexpression of activated

b-catenin inhibited the adipogenic differentiation in rMSC through

down-regulated expressions of C/EBPa and PPARg.

Discussion

To our knowledge, this is the first evidence that activated b-catenin

can induce myogenic differentiation in rMSC. The ability of stabilized

b-catenin to induce myogenic differentiation in rMSC may allow for

its therapeutic application.

Keywords

adipogenic differentiation, b-catenin, mesenchymal stromal cells,

myogenic differentiation, Wnt.

IntroductionDuchenne muscular dystrophy (DMD) is the most

common and lethal genetic muscle disorder in children.

The incidence is about 1/3500 live male births and there is

no efficient pharmacologic treatment at the moment [1].

Stem cell therapy provides a potential strategy for the

treatment of this disease. Multiple stem cell populations,

such as embryonic stem cells and stem cells derived from

muscle, have been assayed for their ability to treat

muscular dystrophy [2� 4]. However, there are many

ethical and immunologic problems associated with these

stem cells, restricting their use in the treatment of DMD

and other severe forms of muscular dystrophy [5,6].

Mesenchymal stromal cells (MSC), separated from

other cells in BM by virtue of their adherence to the

plastic walls of tissue culture containers, are capable of

Correspondence to: Cheng Zhang, Department of Neurology, First Affiliated Hospital, Sun Yat-sen University, Guangzhou 510080, PR China.

E-mail: zhangch6@mail.sysu.edu.cn

Cytotherapy (2007) Vol. 9, No. 7, 667� 681

– 2007 ISCT DOI: 10.1080/14653240701508437

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

differentiating into skeletal muscle cells, as well as

osteoblasts, chondrocytes and adipocytes, under appro-

priate culture conditions [7,8]. MSC are also easily

obtained from various sources and propagated manifold

ex vivo to clinically relevant numbers [9]. Furthermore,

MSC avoid ethical and immunologic hurdles associated

with embryonic stem cells, making them attractive candi-

dates for tissue repair and gene therapy [5,10].

Several studies have shown that transplanted MSC can

contribute to muscle cells and restore sarcolemmal

expression of dystrophin in the mdx mouse model of

DMD [11,12]. MSC as the source of cell therapy to treat

muscular diseases has been promising in principle, but the

process is rather inefficient when comparing the number of

cells implanted with the amount of formed muscle cells. A

key to improving current protocols lies in exploiting the

molecular mechanisms governing the distinct steps of

myogenic differentiation in MSC [13].

The Wnt family includes more than 20 cysteine-rich

secreted glycoproteins and plays an important role in

embryogenesis, covering generation of cell polarity, speci-

fication of cell fate and regulation of proliferation and

differentiation [14,15]. Wnts effect targeted gene expres-

sion by several different signaling pathways [16]. The

canonical pathway is best characterized at the present time.

In the absence of Wnts, b-catenin is phosphorylated by

glycogen synthase kinase-3b (GSK-3b), in association with

axin and adenomatous polyposis coli (APC). These

proteins target b-catenin for degradation by the ubiquitin

proteasome pathway [17]. When Wnts bind to the

Frizzled�LRP5�LRP6 receptor complex, Disheveled

(Dvl) is activated and inhibits the phosphorylation of

b-catenin. This results in b-catenin stabilization and

accumulation in cytoplasm. The stabilized b-catenin

enters the nucleus to bind with members of the T-cell

factor (TCF) and lymphoid enhancer factor (LEF)

transcription factor family and induces expression of target

genes [18].

The Wnt signaling pathway has been associated with

myogenesis in embryogenesis and postnatal muscle regen-

eration. During embryonic development, Wnt1 and

Wnt3a, expressed in the dorsal neural tube, and Wnt7a,

expressed in surface ectoderm, have been shown to activate

the expression of myogenic regulatory factor genes in the

paraxial mesoderm [19,20]. On the other hand, myogenesis

is severely reduced in pre-somitic mesoderm and newly

formed somites by soluble Frizzled-related proteins

(sFRP), which are the Wnt antagonists [21]. Furthermore,

defects in myogenesis have been observed in Wnt-1/Wnt-

3a double-knockout mouse embryos [22]. In post-natal

muscle regeneration, it has been shown that Wnt proteins

induce myogenic differentiation in CD45� stem cells [23].

Although these data demonstrate the important roles of

Wnts in regulation of myogenic differentiation in embry-

ogenesis and muscle regeneration, little is known about

whether the Wnt signaling pathway induces myogenic

differentiation in BM-derived MSC.

Recent experiments have suggested that Wnt signaling

has the capacity to promote proliferation and regulate the

invasion of MSC [24]. Moreover, Wnt signaling has an

inhibitory effect on osteogenic and adipogenic differentia-

tion in MSC [25,26]. These studies suggest that Wnt

signaling plays a potentially important role in the control

of the stem cell properties of MSC.

In light of the established role of the canonical Wnt

pathway during embryo myogenesis, we hypothesized that

this pathway may regulate myogenic differentiation of

MSC. To test this hypothesis, we transfected rat MSC

(rMSC) with activated hb-catenin (S37A). Our results

show that the activated b-catenin induces rMSC differ-

entiation into multinuclear myotubes under appropriate

culture conditions and the induction is accomplished by

evoking the expressions of the myogenic regulatory factors

(MRF) in MSC. We also observed that activated b-catenin

blocked adipogenic differentiation in rMSC. These data

suggest that Wnt/b-catenin signaling plays important roles

in the differentiation of rMSC.

MethodsIsolation, culture and differentiation of rMSC

All animal experiments were approved by the animal care

and experimentation committee of Sun Yat-sen University

(Guangzhou, PR China). Adult male Sprague�Dawley rats

(60� 80 g) were obtained from Sun Yat-sen University

laboratory animal center. The rMSC were isolated from

Sprague�Dawley rats according to the method described

by Wakitani et al. [27], with some modification. In brief,

femora and tibiae of male Sprague�Dawley rats were

collected and the adherent soft tissues were carefully

removed. The BM plugs were hydrostatically expelled

from the bones by syringes filled with proliferation

medium. The BM cells were centrifuged and resuspended

twice in proliferation medium. After resuspension, the

cells were introduced into 25-cm2 tissue culture flasks.

668 YC Shang et al.

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

Approximately 7� 10 days after seeding, the culture flasks

became nearly confluent and the adherent cells were

released with 0.25% trypsin (Gibco Laboratories, USA),

split 1:2, and seeded into fresh culture flasks. The cells

were cultured in proliferation medium, comprising

DMEM (Gibco Laboratories) and 10% FCS (HyClone,

USA). The medium was changed every 3� 4 days. Cells

were grown at 378C in a humidified atmosphere with 5%

CO2.

Skeletal myogenic differentiation was induced by cul-

turing transfected rMSC in myogenic differentiation

medium consisting of DMEM and 2% horse serum (HS;

HyClone). The medium was changed every 3� 4 days and

immunofluorescence analysis for expression of the muscle-

specific markers MyoD, myogenin, desmin and myosin

heavy chain (MHC) was performed at 0, 2, 5, 10 and 15

days after myogenic induction. As a morphologic para-

meter of myogenic differentiation, the fusion index was

defined as the number of nuclei residing in cells containing

three or more nuclei divided by the total number of nuclei

counted. Six random fields at 20-fold magnification were

examined under the microscope at 0, 2, 5, 10 and 15 days

after myogenic induction. Each fusion index represented

the mean value of three independent cultures and was

expressed as mean9SD.

Osteogenic differentiation was studied after incubating

cells in proliferation medium supplemented with 0.01 mm

dexamethasone, 10 mM b-glycerophosphate and 50 mM

l- ascorbic acid-2-phosphate (Sigma, USA). Osteogenic

differentiation was revealed after 3 weeks by deposition of

a mineralized hydroxyapatite extracellular matrix, as

detected by microscopy after staining with 40 mM Alizarin

Red S (Sigma).

For adipogenic differentiation, cells were cultured in

proliferation medium supplemented with 1 mM dexametha-

sone, 0.5 mM methyl-isobutylxanthine and 10 mg/mL

insulin (Sigma). Terminal differentiation was confirmed

after 3 weeks by staining cytoplasmic lipid droplets with

0.5% Oil Red O (Sigma). Oil Red O stain was quantified by

extraction with 4% Igepal (Sigma) in isopropanol for

15 min and measurement of absorbance at 520 nm. Values

represent the mean9SD.

FACS analysis

Analysis of cell-surface molecules was performed on

passage 3 cultures of rMSC using flow cytometry and

the following procedure. Cells were collected and washed

in flow cytometry buffer consisting of 2% BSA and 0.1%

sodium azide in PBS. Then rMSC were incubated with

FITC-conjugated MAb, including anti-CD11b, anti-

CD29, anti-CD44, anti-CD45 (Chemicon, USA), anti-

CD34 and anti-c-Kit Ab (Santa Cruz Biotechnology,

USA). Non-specific fluorescence was determined using

equal aliquots of the cell preparation, which were

incubated with anti-mouse MAb. Data were acquired and

analyzed on a FACScalibur with CellQuest software

(Becton Dickinson and Co., USA).

Plasmids and cell transfections

Expression construct of mutant human b-catenin (S37A)

[hb-catenin (S37A)] was generously provided by Dr Z. Sun

(Departments of Surgery and Genetics, Stanford Univer-

sity School of Medicine, Stanford, CA, USA)[28]. The

mutant of hb-catenin (S37A) with a single point mutation

in the GSK-3b phosphorylation site was generated by a

PCR-based mutagenesis scheme. The appropriate frag-

ments with in-frame restriction enzyme sites were gener-

ated by PCR, cleaved with restriction enzymes and cloned

into the pcDNA3 vector (Invitrogen Corp., USA). Empty

pcDNA3 vector served as a control. The rMSC were

transfected with hb-catenin (S37A) expression plasmid or

empty pcDNA3 vector by Lipofectamine 2000 according

to the manufacturer’s instructions (Invitrogen Corp., USA).

Cells were selected by G418 (200 mg/mL; Sigma) for 14

days. Cells were allowed to recover to 80% confluency for

5� 7 days after G418 selection. To estimate the transfec-

tion efficiency, rMSC were transfected separately with the

pcDNA3-EGFP plasmid under the same conditions as

those used for hb-catenin (S37A) plasmid. The efficiency

of transfection was 82.490.3% after G418 selection. The

MSC transfected with hb-catenin (S37A) or empty

plasmid were termed rMSC (S37A) or rMSC (E). In order

to verify the expression of hb-catenin (S37A) in trans-

fected cells, total RNA and proteins were extracted from

rMSC (S37A) or control cells at 3 and 6 weeks after gene

transfection and used for RT-PCR and Western blot

analysis.

Immunofluorescence analysis

For immunofluorescence analysis, cells grown on glass

coverslips were fixed in methanol/acetone (1:1, v/v) for

10 min at room temperature. Blocking was carried out in

3% BSA in PBS for 30 min and the primary Ab was diluted

1:50 (b-catenin, Cell Signaling Technology, USA; Flag,

b-catenin induces myogenesis and inhibits adipogenesis 669

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

Sigma) and 1:200 (MyoD, myogenin, desmin, MHC, Santa

Cruz Biotechnology) in a solution containing 3% BSA.

Incubation was carried out at room temperature for 1 h or

at 48C overnight. After three washes in PBS, the secondary

Ab (CY3-conjugated goat� ant� rabbit, FITC-conjugated

goat� ant� rabbit, CY3-conjugated goat� ant�mouse Ab;

Santa Cruz Biotechnology) was used at a 1:200 dilution in

3% BSA, and the incubation lasted for 1 h at room

temperature. Nuclear localization of immunostaining

was confirmed by counterstaining with 4?,6-diamidino-

2-phenylindole (DAPI; Sigma) at a 1:5000 dilution in PBS.

The immunofluorescence images were recorded using an

Olympus immunofluorescence microscope. The percen-

tages of MyoD-, myogenin-, desmin- and MHC-positive

cells were calculated from the ratio of positive cells to total

cells counted. Six random fields at 20-fold magnification

were examined under the microscope at 0, 2, 5, 10 and 15

days after myogenic induction. Values represent the

mean9SD.

Reverse transcription�PCR

Total RNA was extracted from the cells at the indicated

time points using Trizol reagent (Invitrogen Corp.) and

treated with RNase-free DNase to remove any contam-

inating genomic DNA, according to the manufacturer’s

protocol. Complementary DNA (cDNA) was obtained by

reverse transcription of 1 mg total RNA using the Revert

AidTM H Minus First Strand Synthesis Kit (Fermentas,

USA). Each PCR reaction was carried out in 25 mL

mixture. The primer sequences were designed based on

published cDNA sequences: 5?-AAA GAC GAT GAC

GAC AAG-3?and 5?-CAT TAG TGG GAT GAG CAG-

3? for hb-catenin (S37A), generating a 391-bp fragment;

5?-GGC TTT CAA CCA TCT CAT TC-3? and 5?-GTT

GGT CAG AAG TCC CAT TAC-3? for Pax3, generating

a 343-bp fragment; 5?-TTC GGG AAG AAA GAG GAC

G-3?and 5?-ATG GTT GAT GGC GGA AGG-3? for

Pax7, generating a 512-bp fragment; 5?-CTA CAG CGG

CGA CTC AGA CG-3? and 5?-TTG GGG CCG GAT

GTA GGA-3? for MyoD, generating a 563-bp fragment

[29]; 5?-TTA GAA GTG GCA GAG GGC TC-3? and

5?-AGG TGC GCA GGA AAT CCG CA-3? for Myf4,

generating a 475-bp fragment [30]; 5?-GAG CCA AGA

GTA GCA GCC TTC G -3? and 5?-GTT CTT TCG

GGA CCA GAC AGG G-3? for Myf5, generating a 440-

bp fragment; 5?-ACT ACC CAC CGT CCA TTC AC-3?and 5?-TCG GGG CAC TCA CTG TCT CT-3? for

myogenin, generating a 233-bp fragment [29]; 5?-CTC

AGG CTT CAA GAT TTG GTG G-3? and 5?-TTG

TGC CTC TCT TCG GTC ATT C-3? for MHC,

generating a 265-bp fragment; 5?-GCC TTG CTG TGG

GGA TGT CT-3? and 5?-CGA AAC TGG CAC CCT

TGA AAA AT-3? for peroxisome proliferator activated

receptor gamma (PPARg), generating a 355-bp fragment;

5?-GGC GGG AAC GCA ACA ACA-3? and 5?-GAG

ATC CAG CGA CCC TAA ACC A-3? for CCAAT/

enhancer binding protein alpha (C/EBPa), generating a

292-bp fragment. To control the amount of RNA in

different samples, expression of glucose-6-phosphate de-

hydrogenase (Gapdh) was used, amplified with primers 5?-ACC ACA GTC CAT GCC ATC AC-3? and 5?-TCC

ACC ACC CTG TTG CTG TA-3?, generating a 451-bp

fragment. All the primers were synthesized by Saibaisheng

Tec. (China). Amplification products were electrophoresed

on 1.5% agarose gels and visualized by ethidium bromide

staining followed by UV light illumination.

Western blot analysis

Cell extracts were prepared by lysing 2�106 transfected

rMSC in 100 mL 1�PBS containing 1% Nonidet P-40,

0.5% sodium deoxycholate, 0.1% SDS, 1 mM phenyl-

methylsulfonyl fluoride, 10 mg/mL leupeptin and

1 mg/mL aprotinin for 10 min at 48C. Lysates were

centrifuged at 12 000 g for 15 min at 48C, and supernatants

were saved. The protein content was determined by the

Bradford assay (Bio-Rad, USA). Equal amounts of proteins

(10 mg) were loaded and separated by a 6% biphasic SDS�PAGE gel and electrotransferred to a PVDF membrane

(Millipore, USA). The membranes were blocked in TBST

containing 5% (w/v) powdered milk (Bio-Rad) for 1 h at

room temperature. For detection of b-catenin or FLAG,

blots were probed with a rabbit-derived anti-b-catenin Ab

(Cell Signaling Technology) or anti-FLAG M2 MAb

(Sigma) at a dilution of 1:1000 for 1 h at room temperature.

For detection of Gapdh, blots were probed with a mouse-

derived MAb at a dilution of 1 in 10 000 (Santa Cruz

Biotechnology). For secondary detection, goat anti-rabbit

(1:3000) or goat anti-mouse IgG (1:10 000) Ab coupled to

horseradish peroxidase was used (Santa Cruz Biotechnol-

ogy, USA) for 1 h. After washing, the blot was incubated in

peroxidase substrate for 5 min prior to exposure to

photographic films (Sigma).

670 YC Shang et al.

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

Statistical analysis

All data are presented as mean9SD. The Student’s t-test

was used to determine statistical significance. A value of

PB0.05 was considered statistically significant. The

experiments were repeated two to four times, and data

from representative experiments are shown.

ResultsMorphology and characterization of rMSC in

culture

Primary rMSC were allowed to attach to the surface of the

plates for 3 days. Adherent spindle-shaped rMSC colonies

were observed as sparsely distributed after removal of non-

adherent cells. Primary rMSC expanded to cover 90% of

the culture and were passaged at a 1:2 dilution after being

detached with trypsin after 7� 10 days. The rMSC used

in this study were from the third to the fifth passages

(Figure 1A).

To characterize the rMSC, we first performed FACS to

detect the immunophenotype of isolated rMSC at passage

3. The results showed that rMSC expressed CD29

(b1-integrin) and CD44. The positive ratios of CD29 and

CD44 were 97.8% and 95.1% separately. In contrast,

rMSC did not express CD11b, CD34, CD45 and c-kit

(Figure 1B). Furthermore, we induced cell differentiation

of isolated rMSC according to published protocols [8].

rMSC differentiated into osteoblastic cells after culture in

osteogenic differentiation medium for 3 weeks, as judged

by their ability to mineralize the extracellular matrix

(Figure 1C). After exposure to an adipogenic stimulus for 3

weeks, cells displayed an adipocyte phenotype, as shown

by the accumulation of neutral lipid droplets in the

cytoplasm (Figure 1D). These results confirmed that the

isolated cells exhibited the previously reported properties

of MSC [8].

Expression of activated hb-catenin in rMSC

To investigate the potential roles of Wnt signaling in the

differentiation of rMSC, we transfected rMSC with a

constitutively active hb-catenin (S37A) plasmid by Lipo-

fectamine followed by G418 selection. hb-catenin (S37A)

harboring a serine to alanine substitution at position 37

failed to be phosphorylated by GSK-3b. Consequently,

degradation via the ubiquitin pathway was blocked, leading

to an accumulation of unphosphorylated b-catenin, which

translocated to the nucleus and activated the transcription

Figure 1. Characterization of isolated rMSC. (A) Phase-contrast microscopy of the fibroblast-like morphology of rMSC in the 3rd passage. (B)

FACS analysis of rMSC: the result revealed rMSC expressed CD29 and CD45 but did not express CD11b, CD34, CD45 and c-kit. (C) Osteogenic

differentiation of rMSC: rMSC were cultured for 3 weeks with osteogenic induction medium, and calcium deposits were visualized by alizarin red

staining. (D) Adipogenic differentiation of rMSC: rMSC were cultured for 3 weeks in adipogenic induction medium. Lipid vacuoles were

visualized by Oil Red O staining. Results are representative graphs of three independent experiments. Scale bars: 100 mm.

b-catenin induces myogenesis and inhibits adipogenesis 671

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

of targeted genes. The expression of hb-catenin (S37A) in

transfected rMSC was detected by RT-PCR and Western

blot analysis. As shown in Figure 2A, a high level of hb-

catenin (S37A) expression in rMSC (S37A) was confirmed

by RT-PCR analysis at 3 weeks after gene transfection. We

also detected the expression of hb-catenin (S37A) in rMSC

(S37A) at 6 weeks after transfection, despite the fact that

the level of hb-catenin (S37A) expression decreased. On

the other hand, no expression of hb-catenin (S37A) was

detected in control cells at either 3 or 6 weeks after gene

transfection by RT-PCR. To confirm the data obtained by

RT-PCR analysis, Western blot analysis was performed.

Detection of hb-catenin (S37A) protein using a MAb

against FLAG confirmed the expression of exogenous

hb-catenin (S37A) gene at the protein level in rMSC

(S37A) (Figure 2B). We did not detect exogenous

hb-catenin (S37A) protein in control cells at either time

point. Both endogenous b-catenin and exogenous hb-

catenin (S37A) proteins were detected in total lysates using

an anti-b-catenin Ab. These data indicate that the

transfection of hb-catenin (S37A) successfully results in

ectopic expression of this gene in rMSC (S37A).

It has been shown that stabilized b-catenin (S37A)

localizes primarily in the nucleus and induces the expres-

sion of targeted genes after transfection into HCT116 cells

and chondrocytes [31,32]. Therefore we performed im-

munocytochemistry to determine the subcellular localiza-

tion of hb-catenin (S37A) in transfected rMSC. Exogenous

hb-catenin (S37A) was detected with anti-FLAG Ab and

the results demonstrated that exogenous hb-catenin

(S37A) localized primarily in the nucleus, with

weak staining in the cytoplasm (Figure 3A�C). No

FLAG-positive cells were observed in the control cells

(Figure 3D�F). On the other hand, anti-b-catenin Ab was

used to detect the distribution of endogenous and

exogenous b-catenin. We found that b-catenin localized

in the cytoplasm and at the cell surface of rMSC (S37A)

and control cells (Figure 3G�L). Moreover, there was a

clear nuclear staining of b-catenin in rMSC (S37A)

(Figure 3G� I) but not in control cells (Figure 3J�L).

These results suggested that transfected stabilized hb-

catenin (S37A) localizes primarily in the nucleus in rMSC

(S37A).

Effect of activated b-catenin on myogenic

differentiation in rMSC

It has been shown that the Wnt canonical pathway plays an

important role in induction of embryonic myogenesis. We

hypothesized that the canonical pathway would also

contribute to the muscle differentiation of MSC. To verify

this hypothesis, we determined whether stimulation of

Wnt signaling by constitutive expression of hb-catenin

(S37A) influenced the myogenic differentiation in

rMSC. When the rMSC (S37A) (Figure 4A) or rMSC

(E) (Figure 4E) were at approximately 80% of confluence,

the cells were induced by using myogenic differentiation

medium. A few multinucleated cells were observed after

3� 5 days of treatment with differentiation medium in

rMSC (S37A) (Figure 4B). The number and length of

multinucleated cells increased gradually along with the

induction in rMSC (S37A) (Figure 4C, D).The fusion

index was 8.6592.58% at 5 days after administration of

myogenic induction medium, and reached 27.193.91% 15

days after induction (Figure 4G). Spontaneous contrac-

tions were observed in a few of these multinucleated cells

when viewing the living cultures. Fifteen days was selected

as the latest time point examined because culture beyond

Figure 2. Expression of hb-catenin (S37A) in rMSC (S37A). (A)

Total cellular RNA was extracted and then RT-PCR was conducted to

estimate the level of hb-catenin (S37A) mRNA expression in rMSC

(S37A) and control cells at 3 and 6 weeks after gene transfection.

Equal loading of RNA samples was checked by amplification of Gapdh

cDNA. (B) Detection of expression of exogenous and endogenous

b-catenin protein in rMSC by Western blots at 3 and 6 weeks after

gene transfection. Exogenous hb-catenin (S37A) protein was detected

with anti-FLAG Ab. Total b-catenin levels were detected with an Ab

directed against b-catenin. Relative amounts of protein present in

aliquots of each cell extract were revealed by detection expression of

Gapdh protein. Results are representative graphs of three independent

experiments.

672 YC Shang et al.

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

this led to lifting of the myotubes from the plate as the

result of spontaneous contractions of the cells. On the

other hand, culture of control rMSC (E) under identical

induction conditions for up to 15 days failed to result in

muscle differentiation (Figure 4F).

To confirm further the myogenic nature of the multi-

nucleated cells observed by phase-contrast microscopy, a

time� course immunofluorescence analysis was examined.

Prior to myogenic induction, both rMSC (S37A) and

control cells were negative for markers of the muscle

Figure 3. Immunofluorescence analysis of subcellular localization of exogenous and endogenous b-catenin in rMSC (S37A) (A�C, G� I) and

control cells (D� F, J�L). Exogenous hb-catenin (S37A) was detected with an anti-FLAG Ab (A, C, D, F). Total b-catenin levels were detected

with an anti-b-catenin Ab (G, I, J, L). The nucleus was stained with DAPI (B, E, H, K). Results are representative graphs of three independent

experiments. Scale bars: 50 mm.

b-catenin induces myogenesis and inhibits adipogenesis 673

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

lineage, including MyoD, myogenin, desmin and MHC.

After exposure to myogenic induction medium, the rMSC

(S37A) gradually expressed these markers (Figure 5A�D).

The percentage of MyoD-positive cells at day 2 was

12.191.1% in rMSC (S37A). The percentage peaked at

day 5 (20.593.2%) and declined thereafter (Figure 6A).

The expression of myogenin increased sequentially, with

0%, 5.1%, 9.9%, 17.8% and 26.7% of cells being positive

on days 0, 2, 5, 10 and 15 after myogenic induction,

respectively (Figure 6B). The expression of desmin and

MHC increased in a similar fashion. The percentages of

desmin- (Figure 6C) and MHC- (Figure 6D) positive cells

reached a maximum on day 15 (28.794.4% and 29.59

3.8%, respectively). These results suggested that the

multinucleated cells induced from rMSC (S37A) were

myogenic. In contrast, the control cells cultured under

identical conditions for up to 15 days were negative for

these markers (Figure 5E�H).

To analyze further the initiation of the skeletal mus-

cle program in rMSC, the expressions of the skeletal

Figure 4. Phase� contrast microscopy observations on the morphology of rMSC during myogenic induction. rMSC (S37A) were cultured in

myogenic differentiation medium for 15 days. Phase� contrast photographs were taken at 0 (A), 5 (B), 10 (C) and 15 (D) days of incubation in

rMSC (S37A). (E, F) rMSC (E) cultured at 0 and 15 days, respectively, in myogenic differentiation medium. The myogenic fusion index, defined as

the number of nuclei residing in cells containing three or more nuclei divided by the total number of nuclei counted, was determined at 0, 2, 5, 10, 15

days in rMSC (S37A) after induction of myogenic differentiation (G). Results are representative graphs of three independent experiments. Scale

bars: 100 mm.

Figure 5. Expression of myogenic markers in rMSC (S37A) and control cells after myogenic introduction. rMSC (S37A) and control cells were

cultured in myogenic induction medium for 15 days and stained for MyoD (A, E), myogenin (B, F), desmin (C, G) and MHC (D, H). Results are

representative graphs of three independent experiments. Scale bars: 100 mm.

674 YC Shang et al.

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

Figure 5 (Continued )

b-catenin induces myogenesis and inhibits adipogenesis 675

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

muscle-specific transcription factors, including Pax3, Pax7,

MyoD, Myf5, Myf4 and myogenin, were examined by RT-

PCR 15 days after the initiation of myogenic induction (5

weeks after gene transfection). As shown in Figure 7, hb-

catenin (S37A) was expressed as expected in differentiated

rMSC (S37A) cells and not in control cells. Pax3 expres-

sion was detected in both control and rMSC (S37A) cells.

However, compared with control cells, the expression of

Pax3 was down-regulated in rMSC (S37A) cells. The

expression of Pax7, MyoD, Myf5, Myf4 and myogenin was

detected only in rMSC (S37A) cells. Control cells cultured

for 15 days in myogenic differentiation media failed to

express these transcription factors. These data demon-

strated that activated b-catenin induced myogenic differ-

entiation in rMSC by triggering the expressions of the

skeletal muscle-specific transcription factors.

Figure 6. Percentages of MyoD- (A), myogenin- (B), desmin- (C) and MHC- (D) positive cells at 0, 5, 10, 15 days after myogenic introduction in

rMSC (S37A). Error bars represent the SD. Results are representative of three independent experiments.

676 YC Shang et al.

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

Effect of activated b-catenin on adipogenic

differentiation in rMSC

The Wnt signaling pathway has been shown to play

important roles in mesodermal cell fate determination in

pre-adipocytes and myoblasts [33]. rMSC are multipotent

stem cells that are capable of differentiating into both

myoblasts and adipocytes. We speculated that activation of

Wnt signaling by transfected stabilized b-catenin might

direct the MSC fate towards myogenesis while preventing

commitment to the adipogenesis in rMSC. To investigate

the effects of activated b-catenin on adipogenic differ-

entiation of rMSC, both control and rMSC (S37A) cells

were induced by using adipogenic differentiation medium

for 3 weeks. The adipogenic differentiation was evidenced

by positive Oil Red O staining of lipid droplets present in

the cytoplasm of the cells. As shown in Figure 8B, control

cells accumulated a lot of lipid droplets after 3 weeks

of adipogenic induction. In contrast, accumulation of

lipid droplets was blocked dramatically in rMSC (S37A)

(Figure 8A). Lipid formation in control cells and rMSC

(S37A) was quantified and the results showed that the

accumulation of lipid in rMSC (S37A) was significantly

lower than that in control cells (0.6390.042 vs. 0.949

0.101, PB0.001; Figure 8C).

C/EBPa and PPARg have been thought to be the

essential transcriptional activators and markers of terminal

differentiation of adipogenesis. To investigate further the

mechanism involved in the inhibition of adipogenesis, we

examined the effect of activated b-catenin on expres-

sions of C/EBPa and PPARg by RT-PCR. As shown in

Figure 8D, the expressions of both C/EBPa and PPARgwere down-regulated in rMSC (S37A) at 3 weeks after

adipogenic induction (6 weeks after gene transfection)

compared with control cells (Figure 8D). These data

demonstrated that activated b-catenin could inhibit adipo-

genic differentiation by suppressing the expressions of

C/EBPa and PPARg in rMSC.

DiscussionWhile evidence for the significance of Wnt signaling in

embryonic development and differentiation has accumu-

lated in many model organisms, such as drosophila,

zebrafish, chicken and mouse, the specific roles of

b-catenin in myogenic and adipogenic differentiation in

adult stem cells have not been well studied. Only a few

studies have indirectly revealed that b-catenin may

contribute to myogenic specification and inhibit adipo-

genic differentiation in adult stem cells [23,26,34]. In the

current paper, we have used rat BM MSC to study the

effects of b-catenin signaling on myogenic and adipogenic

differentiation. To our knowledge, this is the first study

demonstrating that activated b-catenin can induce myo-

genic differentiation in rMSC. We have also shown that

stabilized b-catenin has an inhibitory effect on adipogenic

differentiation in rMSC. All these data implicate b-catenin

as a key signal in controlling myogenic and adipogenic

differentiation in adult stem cells.

Figure 7. RT-PCR analysis of the expression of skeletal muscle-

specific transcription factors. RNA was isolated from rMSC (S37A)

and control cells, which were cultured in myogenic induction medium

for 15 days (about 5 weeks after gene transfection). Gapdh was used as

a loading control. Results are representative graphs of three

independent experiments.

b-catenin induces myogenesis and inhibits adipogenesis 677

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

b-catenin is a multifunctional protein that involves the

cadherin-based cell adhesion and Wnt signaling pathway

[35]. b-catenin located in the cell membrane is associated

with the cytoplasmic region of E-cadherin and involved in

cell� cell adhesion [36]. b-catenin located in both the

nucleus and cytoplasm is a key component in the Wnt

signaling pathway. The role in cell adhesion of b-catenin is

not associated with the one in the Wnt signal pathway.

Mutants of b-catenin cannot function in cell adhesion

because they fail to interact with a-catenin. However,

mutant b-catenin still can transduce Wnt signals [37]. The

amino terminus of b-catenin, which contains a series of

serine and threonine residues, is involved in regulating its

stability [38]. These residues can be phosphorylated by

GSK-3b. Once these residues are phosphorylated, b-

catenin is degraded rapidly by the ubiquitin�proteasome

pathway [17]. Mutations in these residues means b-catenin

avoids being degraded and becomes stable. b-catenin with

these mutations is thought to be constitutively active.

Previous studies have shown that ectopic expression of

stabilized b-catenin (S37A) can enter the nucleus and

regulate the target gene expression in HCT116 cells

and chondrocytes [31,32]. In the present study, we also

applied activated hb-catenin (S37A) to mimic the function

of the Wnt signal and explore its possible role in myogenic

differentiation of rMSC.

A myriad of evidence has demonstrated that MSC can

differentiate into a variety of other cell types in vitro and in

Figure 8. Effect of hb-catenin (S37A) on adipogenic differentiation of rMSC. The rMSC (S37A) (A) and control cells (B) were cultured in

adipogenic induction medium for 3 weeks and stained with Oil Red O cells. Scale bars: 100 mm. (C) Quantification of lipid formation in rMSC

(S37A) and control cells. The asterisk represents a statistically significant difference in the accumulation of lipid between rMSC (S37A) and

control cells; error bars represent the SD. Semi-quantitative RT-PCR analysis of the expression of C/EBPa and PPARg (D). RNA was isolated

from rMSC (S37A) and control cells, which were cultured in adipogenic induction medium for 3 weeks (6 weeks after gene transfection). Gapdh

was used as a loading control. Results are representative graphs of three independent experiments.

678 YC Shang et al.

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

vivo [39]. The specific differentiation is regulated by a

series of signaling pathways. The canonical Wnt signaling

is thought to be a key regulator [40,41]. Recent studies

have demonstrated that Wnt signaling controls the balance

between myogenic and adipogenic potential in vivo and in

vitro [33,42,43]. Myoblasts isolated from Wnt10b� /�mice show increased adipogenic potential, probably con-

tributing to excessive lipid accumulation in actively

regenerating myofibers in vivo in Wnt10b� /� mice. The

alteration in Wnt signaling in myoblasts with age may

contribute to the impairment of muscle regenerative

capacity and increased muscle adiposity [42]. Inhibition

of the Wnt signaling pathway by overexpression of Axin or

Dickkopf-1 (Dkk-1) in pre-adipocytes promotes these cells

to differentiate into adipocytes [33,43]. On the other hand,

activation of the Wnt signaling pathway by overexpression

of Wnt-1 or a b-catenin mutant (b-catenin S33Y) inhibits

adipogenesis in vitro [33]. Furthermore, disruption of Wnt

signaling by expression of dominant-negative TCF4

(dnTCF4) causes transdifferentiation of myoblasts into

adipocytes in vitro . This implies that Wnt signaling plays

an important role in mesodermal cell fate determination

[33]. Other reports have demonstrated that activated

b-catenin prevents adipogenic differentiation and induces

early osteoblast differentiation in C3H10T1/2 MSC

[44,45]. As our results demonstrate that activation of the

Wnt signaling pathway by overexpression of a stabilized

b-catenin promotes myogenesis and inhibits adipogenesis

in rMSC, we hypothesize that, through activation of Wnt

signaling, b-catenin might direct a mesenchymal stromal

cell fate towards myogenesis while preventing commit-

ment to the adipogenesis.

Pax7 belongs to a family of genes that encode paired-

box transcription factors involved in the control of

developmental process. Pax7 is specifically expressed in

satellite cells and is required for the specification of the

satellite cell lineage [46]. The result that stable b-catenin

activates the expression of Pax7 is supported by other

researches. One recent study has shown that the expression

of Pax7 in isolated CD45� adult stem cells is rapidly

induced following Wnt treatment [23]. Another study has

demonstrated that direct expression of the constitutively

active form of LEF (CA-LEF) results in the activation of

Pax7 in mBM-MASC [34]. These observations provide

compelling evidence that Wnt signaling induces adult

stem cells to form myogenic progenitors by activating the

expression of Pax7. Our results show that the transfected

activated form of b-catenin induces the expression of Pax7

in rMSC. So these data suggest that Pax7 represents the

target gene of Wnt signaling that directs the myogenic

specification of adult stem cells.

At the present time, there are no effective treatments for

degenerative muscle diseases. There is a large interest in

transplanting adult stem cells to treat these disorders. The

ability of activated b-catenin to induce myogenic differ-

entiation in rMSC may allow for its therapeutic applica-

tion. Besides regulating the differentiation of adult stem

cells, Wnt signaling could influence the proliferation,

survival and migration of adult stem cells. Therefore the

biologic roles of Wnt signaling in adult stem cells should

be clarified further to avoid possible detrimental adverse

effects in stem cell-based treatment [47]. Future studies

will provide a rational foundation for stem cell-based

tissue repair in humans.

AcknowledgementsWe thank Dr Zijie Sun for providing the mutant human

b-catenin (S37A) plasmid. This work was supported by

grants from the National Natural Science Foundation of

China (30370510, 30170337), the Key Project of the State

Ministry of Public Health (2001321).

References

1 Chakkalakal JV, Thompson J, Parks RJ et al . Molecular, cellular,

and pharmacological therapies for Duchenne/Becker muscular

dystrophies. FASEB J 2005;19:880� 91.

2 Bhagavati S, Xu W. Generation of skeletal muscle from

transplanted embryonic stem cells in dystrophic mice. Biochem

Biophys Res Commun 2005;333:644� 9.

3 Torrente Y, Tremblay JP, Pisati F et al . Intraarterial injection of

muscle-derived CD34 (�) Sca-1(�) stem cells restores

dystrophin in mdx mice. J Cell Biol 2001;152:335� 48.

4 Payne TR, Oshima H, Sakai T et al . Regeneration of dystrophin-

expressing myocytes in the mdx heart by skeletal muscle stem

cells. Gene Ther 2005;12:1264� 74.

5 Price FD, Kuroda K, Rudnicki MA. Stem cell based therapies to

treat muscular dystrophy. Biochim Biophys Acta 2007;1772:272�83.

6 Cossu G, Mavilio F. Myogenic stem cells for the therapy of

primary myopathies: wishful thinking or therapeutic perspec-

tive? J Clin Invest 2000;105:1669� 74.

7 Friedenstein AJ, Gorskaja JF, Kulagina NN. Fibroblast precur-

sors in normal and irradiated mouse hematopoietic organs. Exp

Hematol 1976;4:267� 74.

8 Pittenger MF, Mackay AM, Beck SC et al . Multilineage

potential of adult human mesenchymal stem cells. Science

1999;284:143� 7.

b-catenin induces myogenesis and inhibits adipogenesis 679

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

9 Reyes M, Lund T, Lenvik T et al . Purification and ex vivo

expansion of postnatal human marrow mesodermal progenitor

cells. Blood 2001;98:2615� 25.

10 Le Blanc K, Pittenger M. Mesenchymal stem cells: progress

toward promise. Cytotherapy 2005;7:36� 45.

11 Dezawa M, Ishikawa H, Itokazu Y et al . Bone marrow stromal

cells generate muscle cells and repair muscle degeneration.

Science 2005;309:314� 7.

12 De Bari C, Dell’Accio F, Vandenabeele F et al . Skeletal muscle

repair by adult human mesenchymal stem cells from synovial

membrane. J Cell Biol 2003;160:909� 18.

13 Parker MH, Seale P, Rudnicki MA. Looking back to the embryo:

defining transcriptional networks in adult myogenesis. Nat Rev

Genet 2003;4:497� 507.

14 Moon RT, Kohn AD, De Ferrari GV et al . WNT and beta-

catenin signalling: diseases and therapies. Nat Rev Genet

2004;5:691� 701.

15 Yu JM, Kim JH, Song GS et al . Increase in proliferation and

differentiation of neural progenitor cells isolated from postnatal

and adult mice brain by Wnt-3a and Wnt-5a. Mol Cell Biochem

2006;288:17� 28.

16 Gordon MD, Nusse R. Wnt signaling: multiple pathways,

multiple receptors, and multiple transcription factors. J Biol

Chem 2006;281:22429� 33.

17 Aberle H, Bauer A, Stappert J et al . Beta-catenin is a target for

the ubiquitin� proteasome pathway. Embo J 1997;16:3797� 804.

18 Li L, Yuan H, Xie W et al . Dishevelled proteins lead to two

signaling pathways. Regulation of LEF-1 and c-Jun N-terminal

kinase in mammalian cells. J Biol Chem 1999;274:129� 34.

19 Munsterberg AE, Kitajewski J, Bumcrot DA et al . Combinatorial

signaling by Sonic hedgehog and Wnt family members induces

myogenic bHLH gene expression in the somite. Genes Dev

1995;9:2911� 22.

20 Tajbakhsh S, Borello U, Vivarelli E et al . Differential activation

of Myf5 and MyoD by different Wnts in explants of mouse

paraxial mesoderm and the later activation of myogenesis in the

absence of Myf5. Development 1998;125:4155� 62.

21 Borello U, Coletta M, Tajbakhsh S et al . Transplacental delivery

of the Wnt antagonist Frzb1 inhibits development of caudal

paraxial mesoderm and skeletal myogenesis in mouse embryos.

Development 1999;126:4247� 55.

22 Ikeya M, Takada S. Wnt signaling from the dorsal neural tube is

required for the formation of the medial dermomyotome.

Development 1998;125:4969� 76.

23 Polesskaya A, Seale P, Rudnicki MA. Wnt signaling induces the

myogenic specification of resident CD45� adult stem cells

during muscle regeneration. Cell 2003;113:841� 52.

24 Neth P, Ciccarella M, Egea V et al . Wnt signaling regulates the

invasion capacity of human mesenchymal stem cells. Stem Cells

2006;24:1892� 903.

25 Jian H, Shen X, Liu I et al . Smad3-dependent nuclear

translocation of beta-catenin is required for TGF-beta1-induced

proliferation of bone marrow-derived adult human mesenchy-

mal stem cells. Genes Dev 2006;20:666� 74.

26 De Boer J, Wang HJ, Van Blitterswijk C. Effects of Wnt signaling

on proliferation and differentiation of human mesenchymal stem

cells. Tissue Eng 2004;10:393� 401.

27 Wakitani S, Saito T, Caplan AI. Myogenic cells derived from rat

bone marrow mesenchymal stem cells exposed to 5-azacytidine.

Muscle Nerve 1995;18:1417� 26.

28 Sharma M, Chuang WW, Sun Z. Phosphatidylinositol 3-kinase/

Akt stimulates androgen pathway through GSK3beta inhibition

and nuclear beta-catenin accumulation. J Biol Chem

2002;277:30935� 41.

29 Hyatt JP, Roy RR, Baldwin KM et al . Nerve activity-indepen-

dent regulation of skeletal muscle atrophy: role of MyoD and

myogenin in satellite cells and myonuclei. Am J Physiol Cell

Physiol 2003;285:C1161� 73.

30 Armand AS, Pariset C, Laziz I et al . FGF6 regulates muscle

differentiation through a calcineurin-dependent pathway in

regenerating soleus of adult mice. J Cell Physiol 2005;204:297�

308.

31 Lee HK, Choi YS, Park YA et al . Modulation of oncogenic

transcription and alternative splicing by beta-catenin and an

RNA aptamer in colon cancer cells. Cancer Res 2006;66:10560� 6.

32 Ryu JH, Kim SJ, Kim SH et al . Regulation of the chondrocyte

phenotype by beta-catenin. Development 2002;129:5541� 50.

33 Ross SE, Hemati N, Longo KA et al . Inhibition of adipogenesis

by Wnt signaling. Science 2000;289:950� 3.

34 Belema Bedada F, Technau A et al . Activation of myogenic

differentiation pathways in adult bone marrow-derived stem

cells. Mol Cell Biol 2005;25:9509� 19.

35 Willert K, Nusse R. Beta-catenin: a key mediator of Wnt

signaling. Curr Opin Genet Dev 1998;8:95� 102.

36 Hinck L, Nathke IS, Papkoff J et al . Dynamics of cadherin/

catenin complex formation: novel protein interactions and

pathways of complex assembly. J Cell Biol 1994;125:1327� 40.

37 Orsulic S, Peifer M. An in vivo structure� function study of

armadillo, the beta-catenin homologue, reveals both separate

and overlapping regions of the protein required for cell adhesion

and for wingless signaling. J Cell Biol 1996;134:1283� 300.

38 Yost C, Torres M, Miller JR et al . The axis-inducing activity,

stability, and subcellular distribution of beta-catenin is regulated

in Xenopus embryos by glycogen synthase kinase 3. Genes Dev

1996;10:1443� 54.

39 Herzog EL, Chai L, Krause DS. Plasticity of marrow-derived

stem cells. Blood 2003;102:3483� 93.

40 Kleber M, Sommer L. Wnt signaling and the regulation of stem

cell function. Curr Opin Cell Biol 2004;16:681� 7.

41 Reya T, Clevers H. Wnt signalling in stem cells and cancer.

Nature 2005;434:843� 50.

42 Vertino AM, Taylor-Jones JM, Longo KA et al . Wnt10b

deficiency promotes coexpression of myogenic and adipogenic

programs in myoblasts. Mol Biol Cell 2005;16:2039� 48.

43 Christodoulides C, Laudes M, Cawthorn WP et al . The Wnt

antagonist Dickkopf-1 and its receptors are coordinately

regulated during early human adipogenesis. J Cell Sci

2006;119:2613� 20.

680 YC Shang et al.

Dow

nloa

ded

By:

[Yal

e U

nive

rsity

Lib

rary

] At:

13:5

2 21

Oct

ober

200

7

44 Bain G, Muller T, Wang X et al . Activated beta-catenin induces

osteoblast differentiation of C3H10T1/2 cells and participates

in BMP2 mediated signal transduction. Biochem Biophys Res

Commun 2003;301:84� 91.

45 Mbalaviele G, Sheikh S, Stains JP et al . Beta-catenin and BMP-2

synergize to promote osteoblast differentiation and new bone

formation. J Cell Biochem 2005;94:403� 18.

46 Seale P, Sabourin LA, Girgis-Gabardo A et al . Pax7 is required

for the specification of myogenic satellite cells. Cell

2000;102:777� 86.

47 Luo J, Chen J, Deng ZL et al . Wnt signaling and human diseases:

what are the therapeutic implications? Lab Invest 2007;87:97�

103.

b-catenin induces myogenesis and inhibits adipogenesis 681