Co-ordination of TGF-b and FGF signaling pathways

in bone organ cultures

Aditi Mukherjeea,b,1, Sai Sai Dongc,1, Thomas Clemensa, Jesus Alvarezc,2, Rosa Serrac,*

aDepartment of Pathology, University of Alabama, Birmingham, AL, USAbDepartment of Cellular and Molecular Physiology, University of Cincinnati, Cincinnati, OH, USA

cDepartment of Cell Biology, University of Alabama at Birmingham, 1918 University Blvd, Birmingham, AL 35294-0005, USA

Received 3 June 2004; received in revised form 9 November 2004; accepted 9 November 2004

Available online 26 November 2004

Abstract

Transforming growth factor-b (TGF-b) is known to regulate chondrocyte proliferation and hypertrophic differentiation in embryonic bone

cultures by a perichondrium dependent mechanism. To begin to determine which factors in the perichondrium mediate the effects of TGF-b,

we studied the effect of Insulin-like Growth Factor-1 (IGF-I) and Fibroblast Growth Factors-2 and -18 (FGF2, FGF18) on metatarsal organ

cultures. An increase in chondrocyte proliferation and hypertrophic differentiation was observed after treatment with IGF-I. A similar effect

was seen after the perichondrium was stripped from the metatarsals suggesting IGF-I acts directly on the chondrocytes. Treatment with FGF-

2 or FGF-18 resulted in a decrease in bone elongation as well as hypertrophic differentiation. Treatment also resulted in a decrease in BrdU

incorporation into chondrocytes and an increase in BrdU incorporation in perichondrial cells, similar to what is seen after treatment with

TGF-b1. A similar effect was seen with FGF2 after the perichondrium was stripped suggesting that, unlike TGF-b, FGF2 acts directly on

chondrocytes to regulate proliferation and hypertrophic differentiation. To test the hypothesis that TGF-b regulates IGF or FGF signaling,

activation of the receptors was characterized after treatment with TGF-b. Activation was measured as the level of tyrosine phosphorylation

on the receptor. Treatment with TGF-b for 24 h did not alter the level of IGFR-I tyrosine phosphorylation. In contrast, treatment with TGF-bresulted in and increase in tyrosine phosphorylation on FGFR3 without alterations in total FGFR3 levels. TGF-b also stimulated expression

of FGF18 mRNA in the cultures and the effects of TGF-b on metatarsal development were blocked or partially blocked by pretreatment with

FGF signaling inhibitors. The results suggest a model in which FGF through FGFR3 mediates some of the effects of TGF-b on embryonic

bone formation.

q 2004 Elsevier Ireland Ltd. All rights reserved.

Keywords: Endochondral bone; Cartilage; Perichondrium; FGFR3; FGF18; IGF

1. Introduction

Endochondral bone formation involves the formation of

a cartilaginous template and its subsequent replacement by

osteoblasts to form bone (Cancedda et al., 1995). The

regulation of chondrocyte growth and differentiation during

this process is critical for proper skeletal development.

0925-4773/$ - see front matter q 2004 Elsevier Ireland Ltd. All rights reserved.

doi:10.1016/j.mod.2004.11.006

* Corresponding author. Tel.: C1 205 934 0842; fax: C1 205 975 5648.

E-mail address: rserra@cellbio.bhs.uab.edu (R. Serra).1 These authors contributed equally to the work.2 Current address: Departamento de Morfologı́a y Biologı́a Celular,

Instituto Universitario de Oncologı́a del Principado de Asturias (IUOPA),

Facultad de Medicina, Universidad de Oviedo, Oviedo 33006, Asturias,

Spain.

Defects in regulation of this process can lead to disorders in

skeletal morphogenesis. Endochondral bone formation is

regulated by multiple mechanisms including factors ema-

nating from the surrounding matrix, oxygen supply,

inflammatory mediators, cytokines, and various growth

factors. Transforming growth factor-beta (TGF-b), insulin-

like growth factor-1 (IGF-1) and fibroblast growth factors

(FGFs) are important growth factors in bone and their

functions and interactions at various stages of bone

development control the pace of chondrocyte

differentiation.

The members of the TGF-b super family are secreted

signaling molecules that regulate many aspects of growth

and differentiation (reviewed in Massague, 1998; Moses

Mechanisms of Development 122 (2005) 557–571

www.elsevier.com/locate/modo

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571558

and Serra, 1996; Roberts and Sporn, 1990; Serra and

Chang, 2003). Members of the TGF-b superfamily are

expressed in embryonic and adult skeletal tissue (Gatherer

et al., 1990; Heine et al., 1987; Millan et al., 1991; Pelton

et al., 1990, 1991; Sandberg et al., 1988). TGF-bpromotes chondrogenesis in early-undifferentiated

mesenchymal cells (Denker et al., 1994; Leonard et al.,

1991) but inhibits hypertrophic differentiation in mature

cultures (Ballock et al., 1993; Bohme et al., 1995; Kato

et al., 1988; Serra et al., 1999; Tschan et al., 1993).

TGF-b inhibits both growth and hypertrophic differen-

tiation in embryonic metatarsal cultures in a perichon-

drium dependent manner (Alvarez et al., 2001, 2002).

Previous findings using chick tibiotarsus cultures showed

that the perichondrium can elaborate signals that nega-

tively regulate both chondrocyte proliferation and differ-

entiation (Di Nino et al., 2001; Long and Linsenmayer,

1998). Parathyroid Hormone (PTH) was shown to prevent

increased differentiation resulting from removal of the

perichondrium and our previous results indicated that

TGF-b regulated hypertrophic differentiation through

PTHrP (Long and Linsenmayer, 1998; Serra et al.,

1999). The mechanism of TGF-b mediated growth

inhibition through perichondrium was not addressed. It

is likely one or more factors in the perichondrium are

involved.

Several factors have been shown to regulate chondrocyte

proliferation in vivo. The most well characterized effects are

those of FGF and IGF. Fibroblast Growth Factor Receptor 3

(FGFR3) is expressed in chondrocytes in the proliferating

zone (Deng et al., 1996; Peters et al., 1993). The association

of Fgfr3 mutations in human dwarfing conditions and

results from genetically altered mice clearly show a role for

FGF and specifically FGFR3 in negatively regulating

chondrocyte growth (reviewed in Coumoul and Deng,

2003; Naski and Ornitz, 1998). Fgfr3-null mice demonstrate

skeletal overgrowth (Colvin et al., 1996; Deng et al., 1996)

while mice with an activating mutation in Fgfr3

(Fgfr3G380R) demonstrate dwarfism as a result of reduced

chondrocyte proliferation (Naski et al., 1998). Other

activating mutations of Fgfr3 (Fgfr3G375C and Fgfr3K650E)

also cause inhibition of both proliferation and differentiation

of chondrocytes resulting in retarded bone growth (Chen

et al., 1999; Li et al., 1999). Mice with Fgfr3-K644E

mutation (analogues to human thanatophoric dysplasia

type II) and Fgfr3-K644M mutation (analogues to human

severe achondroplasia with developmental delay and

acanthosis nigricans, SADDAN mutation) also show

dwarfism due to defective long bone development (Iwata

et al., 2000, 2001).

FGFR1 and FGFR2 are expressed in the perichondrium

and FGFR1 is also expressed in the hypertrophic chon-

drocytes unlike FGFR3, which is mainly expressed in

proliferating chondrocytes (Delezoide et al., 1998; Orr-Ur-

treger et al., 1991; Peters et al., 1992, 1993). The exact role

of FGFR1 and FGFR2 in endochondral bone formation is

not clear. Mutations in Fgfr-1 and -2 are associated mainly

with the craniosynostosis syndromes, with some appendi-

cular skeleton defects (Burke et al., 1998; Naski and Ornitz,

1998; Wilke et al., 1997).

Misexpression of FGF2 or FGF9 in chondrocytes also

results in skeletal dysplasia primarily due to reduced

chondrocyte proliferation (Coffin et al., 1995; Garofalo

et al., 1999). In addition, the total length and hypertrophic

area is reduced in embryonic rodent bones in organ culture

treated with FGF2 (Chen et al., 2001; Mancilla et al., 1998).

Recently, it was suggested that FGF18 is the endogenous

ligand of FGFR3 in prehypertrophic chondrocytes (Liu

et al., 2002; Ohbayashi et al., 2002). Mice with targeted

deletion of Fgf18 demonstrate increased proliferation,

similar to what is observed in Fgfr3-null mice. Furthermore,

FGF18 is expressed in the perichondrium adjacent to

FGFR3 in prehypertrophic chondrocytes suggesting para-

crine interactions regulate chondrocyte proliferation in vivo.

FGFR3 may also directly or indirectly regulate hypertrophic

differentiation (Chen et al., 2001; Naski et al., 1998). This

response is separate and independent of PTHrP (Chen et al.,

2001).

Insulin-like Growth Factors (IGF-I and IGF-II) stimulate

proliferation in chondrocytes (reviewed in Dupont and

Holzenberger, 2003; Schmid, 1995). IGFs stimulate longi-

tudinal growth in rat neonatal metatarsal cultures (Coxam

et al., 1995, 1996) and targeted disruption of Igf1, Igf2, or

Igf1r results in dwarfism (Liu et al., 1993). IGF-I and IGF-II

mRNA have been localized to the perichondrium in

developing long bones (Beck et al., 1988; Han et al.,

1987) and the receptors are concentrated in chondrocytes

and osteoblasts throughout skeletal development (Wang

et al., 1995). The activity of IGF is controlled by a family of

extracellular binding proteins, IGFBP1 through IGFBP6.

IGFBPs are differentially expressed in a variety of tissues

and cell types. All six IGFBPs are expressed in bone and

modulate IGF bioavailability. The expression pattern of

IGFBPs has been characterized in growth plate chondro-

cytes (Wang et al., 1995) but expression in the perichon-

drium has been largely overlooked and has only been

reported for IGFBP6 (Van Kleffens et al., 1998). TGF-b has

been shown to regulate expression of IGFs, IGF receptor,

and IGFBPs in many cell types including chondrocytes

(de Los Rios and Hill, 2000; Kveiborg et al., 2001;

Tsukazaki et al., 1994).

We propose that factors from the perichondrium mediate

the effects of TGF-b on endochondral bone development.

To begin to determine which factors are involved in TGF-bmediated inhibition of endochondral bone formation, we

test the effects of FGF2, which activates FGFR3 on

chondrocytes, and IGF-I on intact and perichondrium-free

metatarsal bones in organ culture. The results suggest FGF

and IGF act directly on chondrocytes to regulate growth and

hypertrophic differentiation. Next, we demonstrate that

treatment with TGF-b stimulates phosphorylation of

FGFR3 and general inhibitors of FGF signaling block or

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571 559

partially block the growth and differentiation effects of

TGF-b on the organ cultures. The results suggest that TGF-

b can activate FGF signaling pathways during endochondral

bone formation and that FGF signaling is at least partially

required for the effects of TGF-b on metatarsal develop-

ment. We then show that TGF-b stimulates expression of

FGF18 making it a candidate target for TGF-b in the

perichondrium.

2. Results

2.1. The effects of IGF-I and FGF2 on metatarsal

development are mediated directly through chondrocytes

TGF-b is known to regulate chondrocyte proliferation

and hypertrophic differentiation in metatarsal organ cultures

by a perichondrium dependent mechanism, suggesting the

involvement of additional factors (Alvarez et al., 2001,

2002). FGF and IGF are known mediators of chondrocyte

proliferation (Coumoul and Deng, 2003; Dupont and

Holzenberger, 2003). We proposed a model in which

TGF-b activates the FGF signaling pathway or down-

regulates the IGF pathway resulting in the reduction in total

bone length and chondrocyte proliferation observed in TGF-

b treated organ cultures. Since the effects of TGF-b are

dependent on the perichondrium, the first prediction of the

model is that the effects of FGF or IGF on endochondral

bone formation are due to direct action on chondrocytes.

First, the effect of IGF-I or FGF2 on development in our

organ culture model was determined. Metatarsal bones

isolated from embryonic day 15.5 mice were either

untreated or treated with 10, 30, and 100 ng/ml of IGF-I

or FGF2 for 5 days (Fig. 1). Treatment with 100 ng/ml or

IGF-I or FGF2 gave the maximal response. Treatment with

IGF-I resulted in an overall increase in the length of the

bone while treatment with FGF2 resulted in a decrease in

the total bone length as previously described (Fig. 1A–C,G;

Coxam et al., 1996; Mancilla et al., 1998).

Total bone length is a combination of changes in

hypertrophic differentiation and proliferation during the

development of the bone. As previously described,

hypertrophic cartilage is seen as the clear area in the center

of the organ culture (Dieudonne et al., 1994; Serra et al.,

1999). Hypertrophic differentiation was measured as the

length of the clear area over the total length of the bone. The

measurement normalizes for any differences in the total

length of the bone. Treatment with IGF-1 resulted in an

increase in the amount of hypertrophic cartilage in the

culture and treatment with FGF2 resulted in a dramatic

decrease (Fig. 1H). The level of hypertrophic differentiation

measured in this way correlated with the percentage of

histologically hypertrophic cartilage observed in hematox-

ylin and eosin stained sections from each of the conditions

tested (Fig. 1D,E) and in the area of cartilage expressing

Col10a mRNA (not shown).

Next, the effects of each factor on proliferation were

measured as the percentage of cells that incorporated BrdU

into their nucleus (Fig. 1I–L). Treatment with IGF-I for 24 h

resulted in an increase in labeled chondrocytes relative to

untreated controls (Fig. 1L). Treatment with FGF2 resulted

in a decrease in the number of labeled chondrocyte nuclei;

however, the number of labeled nuclei was increased in the

perichondrium (Fig. 1K). The effects of FGF2 on develop-

ment of the metatarsal cultures mimicked the effects of

TGF-b previously reported (Serra et al., 1999). That is both

TGF-b and FGF2 inhibit hypertrophic differentiation and

chondrocyte proliferation while promoting proliferation in

the perichondrium.

Next, the role of the perichondrium in IGF-I and FGF2

mediated effects on metatarsal development were deter-

mined (Fig. 2). FGF and IGF ligands as well as IGFBPs

have already been localized to the perichondrium (Beck

et al., 1988; Han et al., 1987; Liu et al., 2002; Ohbayashi

et al., 2002; Van Kleffens et al., 1998) while FGFR3 and

IGFR1 are localized on prehypertrophic chondrocytes

(Deng et al., 1996; Peters et al., 1993; Wang et al., 1995).

Metatarsal bones in which the perichondrium had been

enzymatically removed were either untreated or treated with

IGF-I or FGF2 for 5 days. The total length and hypertrophic

area of the cultures was then measured (Fig. 2). Treatment

with IGF-I again resulted in an increase in the total length

and hypertrophic area of the bone suggesting that the effects

of IGF-I on chondrocytes are direct and not mediated

through the perichondrium. Treatment with FGF2 resulted

in a decrease in total length and hypertrophic area in the

absence of the perichondrium suggesting that, unlike

TGF-b, the effects of FGF2 on chondrocytes are direct.

2.2. TGF-b activates FGF signaling

If TGF-b acts through IGF or FGF we would expect that

treatment with TGF-b would result in alterations in the

activation state of the receptors. To test this hypothesis,

metatarsal cultures were either untreated or treated with

10 ng TGF-b/ml for 24 h. Bones treated with either FGF2 or

IGF-I were used as positive controls. IGFR1 phosphoryl-

ation was detected by Western blot using a phospho-IGFR1

specific antibody (Fig. 3). The total level of IGFR1 beta

subunit was also determined by Western blot. Alterations

were not detected in the total level or phosphorylation state

of the IGFR1 after treatment with TGF-b for 24 h (Fig. 3);

however, a reduction in the total level of IGFR1 was

detected after 5 days of treatment (not shown). FGFR3 was

immunoprecipitated from protein extracts made from

untreated and TGF-b treated metatarsals. Receptor

phosphorylation was detected by anti-phosphotyrosine

immunoblot. To control for the total receptor levels,

receptor protein was detected in blots that were stripped

of the anti-phosphotyrosine antibody and incubated with an

antibody for FGFR3. Alterations were not detected in the

total level of FGFR3 protein (Fig. 3). Little to no

Fig. 1. Effects of IGF-I and FGF2 on metatarsal bone development in organ culture. Metatarsal bones were cultured for 5 days in the absence (A) or

presence of 100 ng IGF-I/ml (B), or 100 ng FGF2/ml (C). The total length of bones was measured and the means and standard deviations were calculated

and graphed (G). Statistical significance was calculated using a Student’s t-test. P-values less than 0.0001 are indicated with an asterisk. Bones that were

untreated (D) or treated with IGF-I (E) or FGF2 (F) for 5 days were sectioned and stained with hematoxylin and eosin. The percent area of hypertrophic

cartilage was measured and the means and standard deviations were graphed (H). Statistical significance was calculated using a Student’s t-test. P-values

less than 0.0001 are indicated with an asterisk. Cultures that were untreated (I) or treated with IGF-I (J) or FGF2 (K) for 24 h were then incubated with

BrdU for 2.5 h. BrdU was detected by immunofluorescence. The percentage of BrdU positive cells was calculated and the means and standard deviations

were graphed (L). Treatment with IGF increased the total length, hypertrophic area, and incorporation of BrdU into the cultures. Treatment with FGF2

resulted in a reduction in the total length, percentage of hypertrophic cartilage, and incorpration of BrdU into chondrocytes.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571560

Fig. 2. The role of the perichondrium in IGF-I and FGF2 activity. The

perichondrium was enzymatically removed from metatarsal bones that

were then left untreated (A) or were treated with 100 ng IGF-I/ml (B) or

100 ng FGF2/ml (C). After 5 days in culture the total length (D) and the

percentage of the clear area representing hypertrophic cartilage (E) were

measured. The means and standard deviations were determined and

graphed. Statistical significance was calculated using a Student’s t-test.

P-values less than 0.0001 are indicated with an asterisk. Signaling by IGF-I

and FGF2 are not dependent on the perichondrium.

Fig. 3. Effects of TGF-b on IGF and FGF receptor activation. Proteins were

isolated from untreated bones (U) and bones treated with 10 ng TGF-b1/ml

(b) or 100 ng IGFI/ml or 100 ng FGF2/ml (C). Proteins were used in a

straight Western blot for phosphorylated IGF receptor 1 (PO4-IGFR) and

total IGF receptor 1 b subunit (IGFR). For FGFR3, proteins were

immunoprecipitated with an antibody to FGFR3, run on a gel, and Western

blot was performed with a phosphotyrosine specific antibody (PO4-

FGFR3). The blot was then stripped and incubated with an FGFR3 antibody

(FGFR3). Treatment with TGF-b resulted in an increase in the amount of

PO4-FGFR3 without and increase in total FGFR3 levels.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571 561

phosphorylated FGFR3 was detected in untreated cultures

whereas cultures treated with TGF-b or FGF2 demonstrated

significant phosphorylation of FGFR3 (Fig. 3). Increased

FGFR3 phosphorylation was also detected after 5 days of

treatment (not shown). The data together suggest that TGF-

b acts to regulate development of metatarsal bones at least

in part by regulating FGF signaling.

2.3. TGF-b regulates the expression of FGF

signaling molecules

To determine if TGF-b regulated expression of genes

that encode FGF signaling proteins, RNA was extracted

from metatarsal bones that were untreated or treated with

TGF-b for 24 h. Expression levels of FGF2 and FGF18

mRNA were determined using semi-quantitative RT-PCR

(Fig. 4). FGF18 has been detected in the perichondrium of

developing bones and is thought to act in a paracrine manner

through FGFR3 on chondrocytes (Liu et al., 2002;

Ohbayashi et al., 2002). FGF2 is also expressed in skeletal

elements. The specified gene and glyceraldehyde-3-phos-

phate dehydrogenase (GAPDH) were amplified from cDNA

generated from RNA extracted from untreated and TGF-btreated metatarsal cultures. GAPDH was used to control for

the amount and quality of RNA used in each reaction. PCR

amplification was performed for a varying number of cycles

Fig. 4. Gene expression in metatarsal bones treated with TGF-b. RNA was

isolated from metatarsal bones in culture that had been left untreated (0) or

were treated with 10 ng TGF-b1/ml (C) for 24 h. cDNA was generated

from the RNA and used in semi-quantitative PCR amplification. FGF2,

FGF18, and glyceraldehyde-3-phosphate dehydrogenase (G, Gapdh) were

amplified for varying numbers of cycles (20–45 cy). Gapdh was used to

control for the amount of RNA used. Samples run in the absence of reverse

transcriptase (RT-) were used to control for the presence of DNA

contamination in the samples. Molecular weight markers are shown (M).

To quantify changes in gene expression, band intensity measurements were

made using Kodak gel imaging software from two separate experiments.

After normalization, treatment with TGF-b resulted in FGF18 mRNA

levels that were 3.8- and 14.2-fold above the control. FGF2 was regulated

1.4- and 1.7-fold above the untreated control.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571562

to determine the linear range for PCR product formation.

Band intensities were determined using Kodak software.

Intensities were normalized to the intensity of the GAPDH

bands so that normalized fold-differences in gene

expression in untreated and treated tissue could be

determined. Treatment with TGF-b resulted in a dramatic

increase in the levels of FGF18 mRNA (3.8- and 14.2-fold

induction in two separate experiments). FGF2 was only

marginally regulated by TGF-b (1.4- and 1.7-fold induction

in two experiments). The data suggest that TGF-b can

modulate the FGF pathway by regulating expression of

important mediators of signaling.

2.4. FGF18 regulates proliferation and differentiation

in organ cultures

Since FGF18 was identified as being regulated by TGF-

b, the effects of FGF18 on development of the organ

cultures was investigated (Fig. 5). Two dose response

experiments were performed (Fig. 5 and data not shown).

Doses between 10 and 600 ng FGF18/ml were tested.

Treatment with 200 ng FGF18/ml for 5 days resulted in a

decrease in the total bone length that was significant

(P!0.05). Higher doses resulted in only marginal further

decreases in length (Fig. 5B). The area of hypertrophic

cartilage was also reduced after treatment for 5 days with

50 ng FGF18/ml. Higher doses resulted in significant further

reductions in hypertrophic differentiation (Fig. 5C and data

not shown). It has been shown that FGF18 binds with high

affinity to FGFR3 that is expressed on proliferating

chondrocytes (Xu et al., 2000). Treatment with 200 ng

FGF18/ml for 24 h resulted in decreased chondrocyte

proliferation as measured by the incorporation of BrdU

(Fig. 5D,E). The level of inhibition was comparable to that

observed with TGF-b in the same experiment (Fig. 5D,E).

In addition, like TGF-b, FGF18 increased proliferation in

the perichondrium (Fig. 5D). The results suggest that

FGF18 is sufficient to regulate the proliferation and

differentiation of chondrocytes in mouse metatarsal bones.

This observation, in addition to regulation of FGF18 mRNA

by TGF-b supports a model in which TGF-b may act

through FGF18 in the perichondrium to regulate chondro-

cyte proliferation and/ or differentiation.

2.5. FGF inhibitors alter TGF-b responsiveness

To determine if FGF signaling is required for some of the

effects of TGF-b on metatarsal development, cultures were

treated with inhibitors of FGF activity followed

by treatment with TGF-b. The inhibitors used were

PSS [poly(4-styrenesulfonic acid)] (Liekens et al., 1999)

and a 16mer inhibitory peptide that mimics the confor-

mation of the receptor binding domain of FGF2 but has no

sequence homology to FGF2 (Cosic et al., 1994). PSS binds

to FGF at the heparin binding domain and prevents

binding to the receptor and heparin sulfate proteoglycans.

This compound has been used to block FGF signaling in

lens organ cultures (Cerra et al., 2003). First, the efficacy

of the inhibitors to block signaling by FGF2 and FGF18

in organ cultures was tested (Figs. 6 and 7). Cultures

were either left untreated or treated with the indicated

inhibitors overnight at which time half of the cultures

were treated with FGF2 or FGF18 for 5 days. The total

length of the bones (Figs. 6B and 7B) and the area of

hypertrophic cartilage (Figs. 6C and 7C) were measured.

The mean, standard deviation, and significance were

calculated and graphed. The inhibitors alone had an

effect on metatarsal development. PSS treatment resulted

in increased length and hypertrophy. The 16mer peptide

resulted in a slight decrease in length and hypertrophy.

These effects may be due to blocking different endogen-

ous FGF signaling pathways. Treatment with FGF2 or

FGF18 alone resulted in a statistically significant

decrease in the total length and the percentage of the

bone that contained hypertrophic cartilage, as expected

from our previous experiments. In the presence of PSS,

treatment with FGF2 or FGF18 resulted in a similar

statistically significant decrease in length suggesting PSS

does not block the signaling pathway used by FGF2 or

FGF18 that affects total bone length (Figs. 6B and 7B).

In contrast, the presence of PSS blocked the ability of

FGF2 and FGF18 to inhibit hypertrophic differentiation

(Figs. 6C and 7C). In the presence of the 16mer peptide,

total length was not significantly inhibited by FGF2 or

FGF18 suggesting that this inhibitor blocks the signaling

pathways used by FGF to regulated total bone length

(Figs. 6B and 7B). The presence of the 16mer peptide

partially blocked or blocked the effects of FGF2

and FGF18, respectively, on hypertrophic differentiation.

In the absence of the inhibitor, the area of hypertrophic

cartilage was reduced to 37 or 67% of control after

treatment with FGF2 and FGF18, respectively. In the

presence of the inhibitor, the area of hypertrophic

cartilage was reduced to 63 or 79% of the control after

FGF2 or FGF18 treatment, respectively. The results

suggest that the 16mer inhibitor can block or at least

partially block the signaling pathways used by FGF2 and

FGF18 that regulate both growth and hypertrophic

differentiation.

Next, cultures were either untreated or pretreated with

PSS or the 16mer peptide followed by treatment with

TGF-b (Fig. 8). As expected, treatment with TGFb in the

absence of inhibitors resulted in a decrease in the area of

hypertrophic cartilage. Treatment with PSS blocked most

of the effects of TGF-b on hypertrophic differentiation.

In the absence of the inhibitor, the area of hypertrophic

cartilage after treatment with TGF-b was 58% the level

of the untreated control (Fig. 8C). In the presence of

PSS, the area of hypertrophic cartilage was reduced after

treatment with TGF-b to 88% the level of the untreated

control. The presence of the 16mer peptide partially

blocked the effects of TGF-b on hypertrophic

Fig. 5. Effects of FGF18 on metatarsal development. (A) Whole mount appearance of bones untreated (control), treated with 200 ng FGF18/ml, or 600 ng

FGF18/ml, for 5 days. The black bars along the bones mark the central clear area of hypertrophic cartilage. The total length (B) and percent hypertrophic area

(C) of the bones was calculated and graphed. Statistical significance was determined using Student’s t-test. Statistical significance relative to the untreated

control is indicated with an asterisk. P-values are indicated on the graph. FGF18 inhibited total length and hypertrophic differentiation in the metatarsal

cultures. (D) Bones that were either untreated (control) or treated with 200 ng/ml FGF18 or 10 ng/ ml TGF-b1 for 24 h were incubated with BrdU for 2.5 h.

BrdU incorporation was measured by immunofluourescence. The percentage of BrdU positive chondrocytes was calculated and graphed (E). The standard

deviation and statistical significance are shown. FGF18 inhibited chondrocyte proliferation.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571 563

Fig. 6. Effects of FGF inhibitors on the FGF2 response. (A) Bones were untreated or pretreated with PSS or the 16mer peptide inhibitor followed by treatment

with 0 or 100 ng FGF2/ml. After 5 days, total bone length (B) and hypertrophic area (C) were measured. The means and standard deviations were calculated

and graphed. Statistical significance was calculated using a Student’s t-test. P-values are indicated on the graph. A single asterisk represents significance

relative to the 0 FGF2 control of the sample treated with the same drug. A double asterisk indicates significance relative to the control without drug and without

FGF2. The dark lines next to the bones marks the clear area representing the hypertrophic zone.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571564

differentiation (Fig. 8C). In the absence of the inhibitor,

TGF-b inhibited hypertrophic differentiation to 49% the

level seen in the untreated control. In the presence of the

16mer peptide hypertrophic differentiation was reduced to

only 75% the level of control without TGF-b. The results

suggest that the FGF inhibitors, PSS and the 16mer

peptide, block or partially block the signaling pathway

used by TGF-b to regulate hypertrophic differentiation.

Since the drugs also block the effects of FGF2

and FGF18 on hypertrophic differentiation, we propose

that part of the effects of TGF-b on hypertrophic

differentiation are mediated through FGF.

Treatment with TGF-b in the absence of inhibitors

resulted in a decrease in the overall length of the bone as

previously reported (Fig. 8B). Total length was not

significantly altered after treatment with TGF-b in the

presence of the PSS inhibitor suggesting PSS blocked the

effects of TGF-b on length. However, since PSS did not

affect the signaling pathways used by FGF2 or FGF18

that regulate total bone length, it is not clear if this effect

of TGF-b is mediated through FGF2 or FGF18.

Furthermore, in the presence of the 16mer peptide,

TGF-b significantly inhibited bone length albeit to a

lesser degree than that seen in the absence of the peptide

(Fig. 8B). The 16mer peptide was able to more

completely block the effects of FGF2 and FGF18 on

bone length suggesting FGF may only partially mediate

the effects of TGF-b on bone length.

3. Discussion

TGF-b1 inhibits hypertrophic differentiation and chon-

drocyte proliferation by a perichondrium dependent mech-

anism in embryonic metatarsal cultures (Alvarez et al.,

2001; Alvarez et al., 2002). We proposed IGF and/or the

Fig. 7. Effects of FGF inhibitors on the FGF18 response. (A) Bones were untreated or pretreated with PSS or the 16mer peptide inhibitor followed by treatment

with 0 or 200 ng FGF18/ml. After 5 days, total bone length (B) and hypertrophic area (C) were measured. The means and standard deviations were calculated

and graphed. Statistical significance was calculated using a Student’s t-test. P-values are indicated on the graph. A single asterisk represents significance

relative to the 0 FGF18 control of the sample treated with the same drug. A double asterisk indicates significance relative to the control without drug and

without FGF18. The dark lines next to the bones marks the clear area representing the hypertrophic zone.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571 565

FGF systems as the possible targets of TGF-b1 in the

perichondrium. We established that IGF-I and FGF2 can

regulate chondrocyte growth and differentiation in the

embryonic metatarsal cultures independent of the perichon-

drium. To test the hypotheses that TGF-b1 either down

regulates the IGF system and/or up regulates the FGF

system, we looked at IGFR1 and FGFR3 phosphorylation

after treatment with TGF-b. FGFR3 phosphorylation was

increased after TGF-b treatment but no change in IGFR1

phosphorylation was observed. We then focused on FGF

signaling. The inhibition of FGF signaling by pretreatment

with FGF inhibitors was used to access the requirement for

FGF signaling in the TGF-b response. The results suggested

that TGF-b1 regulates endochondral bone formation in

culture at least in part through activation of FGF signaling.

We also showed that TGF-b stimulates expression of FGF18

mRNA making it a likely target for TGF-b in the

perichondrium.

During endochondral bone formation, the perichon-

drium is involved in signaling pathways which negatively

regulate chondrocyte proliferation and differentiation (Di

Nino et al., 2001; Long and Linsenmayer, 1998),

including Ihh and TGF-b signaling (Alvarez et al.,

2001, 2002; Long et al., 2001). In contrast, we found that

both IGF-I and FGF2 can regulate chondrocyte growth

and differentiation even in the absence of the perichon-

drium. The treatment of metatarsal organ cultures with

IGF-I resulted in an increase in total bone length and

hypertrophic differentiation while treatment with FGF2

caused a decrease in bone length and hypertrophic

differentiation, both in the presence and the absence of

the perichondrium. Both IGF-I and IGF-II, and many

FGF ligands, most notably FGF18, (Beck et al., 1988;

Han et al., 1987; Liu et al., 2002; Ohbayashi et al., 2002;

Van Kleffens et al., 1998) are known to be expressed in

the perichondrium during endochondral bone formation

Fig. 8. Effects of FGF inhibitors on TGF-b response. Metatarsal bones were treated with PSS or with two concentrations of a 16mer inhibitor. They were then

treated with TGF-b1 and the bones were photographed after 5 days and total bone length (B) and hypertrophic area (C) were measured. The means and standard

deviations were calculated and graphed. Statistical significance was calculated using a Student’s t-test. P-values are indicated on the graph. A single asterisk

represents significance relative to the 0 TGF-b control of the sample treated with the same drug. A double asterisk indicates significance relative to the control

without drug and without TGF-b. PSS and, to a lesser extent, the higher concentration of the 16mer partially blocked the inhibitory effects of TGF-b on the total

length and hypertrophic area observed in the cultures. The dark lines next to the bones marks the clear area representing the hypertrophic zone. PSS and the

40 nM concentration of the 16mer peptide were tested in separate experiments.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571566

and it is possible that they are mediating the perichon-

drium dependent effects of TGF-b.

FGF signaling has been shown to play a critical role in

the regulation of bone growth. FGFR3 is expressed in

proliferating and prehypertrophic chondrocytes in the

epiphyseal growth plates (Naski et al., 1998; Ohbayashi

et al., 2002; Peters et al., 1993). It has been shown that

expression of activating Fgfr3 mutants in mice results in

dwarfism (Coumoul and Deng, 2003). In contrast, lack of

Fgfr3 in mice causes skeletal overgrowth, indicating that

FGFR3 signaling inhibits endochondral bone growth

(Colvin et al., 1996; Deng et al., 1996), making FGF

signaling a likely candidate for mediating the inhibitory

effects of TGF-b during endochondral bone formation. The

relationship between TGF-b and FGF signaling during

endochondral bone formation was explored in this study.

The results show that TGF-b increases FGF18 mRNA levels

and FGFR3 phosphorylation in the organ cultures,

suggesting that TGF-b1 may be targeting perichondrial

FGF ligands that in turn activate FGFR3, which is known to

inhibit chondrocyte proliferation and differentiation.

Recent work has suggested that FGF18 is the endogenous

ligand of FGFR3 in prehypertrophic chondrocytes (Liu

et al., 2002; Ohbayashi et al., 2002), and the expression of

FGF18 in the perichondrium has also been reported (Liu

et al., 2002; Ohbayashi et al., 2002) making it a possible

perichondrial target for TGF-b1. In our studies, TGF-b1

increases expression of FGF18 mRNA. Recent genetic

studies have identified a defect in chondrogenesis and

osteogenesis in mice lacking Fgf18 similar to that seen in

the Fgfr3 mutant mice (Liu et al., 2002; Ohbayashi et al.,

2002). Our study demonstrated that treatment of the

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571 567

metatarsal organ cultures with FGF18 also results in

inhibition of chondrocyte proliferation and hypertrophic

differentiation, similar to treatment with FGF2, another

ligand for FGFR3.

We used two inhibitors of FGF signaling to determine if

there was a functional connection between TGF-b and FGF

signaling in endochondral bone formation. The first inhibitor,

PSS, is a heparin-mimicking polysulfonated compound

(Liekens et al., 1999). It was shown to inhibit binding of

FGF2 to heparin sulfate proteoglycans, abrogating the FGF/

proteogylcan and receptor complex. PSS has been used in

cell culture and organ culture models to block the effects of

FGF2 on specific biological processes (Cerra et al., 2003;

Liekens et al., 1999). The second inhibitor used was a novel

16mer peptide that was developed through resonant

recognition modeling (Cosic et al., 1994). The peptide has

conformational similarity to the receptor-binding domain of

FGF2 but has no sequence homology. It has been shown to

block FGF2 binding to FGFR1 and inhibit FGF2 induced

proliferation in fibroblasts and human glioma cells (Cosic

et al., 1994; Kono et al., 2003). Both inhibitors have been

characterized with regards to FGF2 and it’s high affinity

receptor FGFR1. It was not clear if the inhibitors would have

the same efficacy against FGF2 and FGF18 signaling in

metatarsal organ cultures. In the present study, we showed

that both inhibitors are able to block specific effects of FGF2

and FGF18. PSS blocked the effects of both FGF2 and FGF18

on hypertrophic differentiation but did not block the effects

on total bone length. The result suggests there may be

differences in the way FGF regulates proliferation versus

differentiation. It has been suggested in a recent study that

during FGFR3 signaling in bone growth, the MAPK pathway

mediates inhibition of hypertrophic chondrocyte differen-

tiation, whereas Stat1 mediates inhibition of chondrocyte

proliferation (Murakami et al., 2004). In contrast to PSS, the

16mer peptide blocked the effects of FGF2 and FGF18 on

total bone length but only partially blocked the effects on

hypertrophic differentiation. In addition, treatment with the

inhibitors alone had and effect on development of the

metatarsals. Treatment with PSS resulted in longer bones

with more hypertrophic differentiation while treatment with

the 16mer peptide resulted in shorter bones with less

hypertrophic differentiation. The effects of the inhibitors

alone likely represent the blocking of various endogenous

FGF signaling pathways. It also illustrates one of the caveats

of drug studies, which is that the exact specificity of the drug

in a particular system is difficult to determine.

In the present study PSS and the 16mer peptide blocked

and partially blocked, respectively, the effects of TGF-b on

hypertrophic differentiation. The drugs had similar effects

on TGF-b and FGF-mediated inhibition of hypertrophic

differentiation. We interpret this result to mean that that

FGF is at least partially required to mediate the effects of

TGFb on hypertrophic differentiation. As mentioned above,

the specificity of PSS has not been clearly established and

we cannot rule out the requirement of additional heparin-

binding growth factors required for the TGF-b response.

Such a factor could be required for TGF-b mediated effects

on bone length which, were blocked by PSS under

conditions in which the effects of FGF on bone length

were not affected. Likewise, the partial block of TGF-beffects on bone length with the 16mer peptide suggest that

additional factors may be required for the TGF-b mediated

reduction in bone length.

As TGF-b has been shown to regulate expression of IGFs,

IGF receptor, and IGFBPs in many cell types including

chondrocytes (de Los Rios and Hill, 2000; Kveiborg et al.,

2001; Tsukazaki et al., 1994), we also investigated whether

TGF-b1 regulates activation of the IGF receptor in

embryonic growth plate. We did not detect alterations in

IGFR phosphorylation after treatment with TGF-b. A lack of

change in receptor phosphorylation does not necessarily

mean a lack of an effect. Receptors in cells that are not targets

of TGF-b could drown out any effects of TGF-b on a subset of

receptors, thus, any co-ordination between TGF-b and IGF

signaling needs to be investigated further.

In conclusion, the study shows that both IGF-I and FGF2

have perichondrium independent effects on chondrocyte

proliferation and differentiation and that TGF-b interacts

with components of FGF signaling during endochondral

bone formation. Our results suggest that FGF ligands,

perhaps FGF18, modulate some of the perichondrium-

dependent inhibitory effects of TGF-b1 during endochon-

dral bone formation. FGF 18 and/or FGFR-3 knock out

mouse models can be used to enable further studies aimed at

understanding co-ordination of TGF-b and FGF18 signaling

specifically. Investigating the complex interaction between

TGF-b and FGF systems, which are key regulators of

skeletal development, will enhance our understanding of the

process of endochondral bone formation immensely.

4. Materials and methods

4.1. Embryonic metatarsal rudiment organ culture

The three central metatarsals from the hind limbs of

day E15.5 mouse embryos were removed and placed in

culture as previously described (Dieudonne et al., 1994;

Serra et al., 1999). For experiments to remove the

perichondrium, metatarsals from one limb were stripped

of the perichondrium while those from the contralateral

limb were kept intact. Collagenase type 2, 1 mg/ml in

PBS (Worthington biochemical Corp.), was used to

remove the perichondrium, as previously reported (Haaij-

man et al., 1999; Thesingh and Burger, 1983). Enzymatic

activity was stopped by transferring the bones to 10%

Fetal Bovine Serum in PBS and the remaining perichon-

drium was removed mechanically by rolling on a plastic

surface. All rudiments were cultured in 24-well plates in

1 ml of medium containing alpha-MEM (Gibco-BRL)

supplemented with 0.05 mg/ml ascorbic acid, 0.3 mg/ml

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571568

L-glutamine, 0.05 mg/ml gentamycin, 1 mM b-glycero-

phosphate, and 0.2% Bovine Serum Albumin (BSA).

Explants were grown at 37 8C in a humidified 5% CO2

incubator. IGF-1 (GroPep, Australia), FGF2 or FGF18

(R&D Systems, MN or US Biologicals), or TGF-b1

(R&D Systems, MN) were added to cultures 12–16 h

after dissection. Inhibitors PSS [poly(4-styrenesulfonic

acid)], MWZ70,000, (PolySciences,) and a 16mer

inhibitory peptide (Bachem, #H2176) were added within

a few hours after the dissection and TGF-b was added

the next morning. Cultures were observed and photo-

graphed with an Olympus SZH12 dissecting microscope

after 5 days of treatment. Cultures were fixed in fresh

4% paraformaldehyde in PBS overnight at 4 8C then the

cultures were dehydrated through a series of ethanols and

xylene and embedded in paraffin. Sections (5 mm) were

stained with hematoxylin and eosin or used for

immunodetection of BrdU incorporation and for in situ

hybridization.

The length of bones was calculated after 5 days of

treatment by measuring the length of each metatarsal in

photographs taken with an Olympus SZH12 dissecting

microscope with a Magnafire digital camera. The total

length was measured in photographs that were all taken

at the same magnification (25!). The actual bone length

was calculated by dividing the length measured on the

images by the magnification indicated on the microscope

and camera. The mean, standard deviation, and signifi-

cance were determined using a Student’s t-test calculated

with Microsoft Excel. Groups with a probability value

less than 0.05% (P!0.05) were considered significantly

different.

The percent of the bone rudiment containing hypertrophic

cartilage was calculated after 5 days of treatment by three

separate methods: (1) Measuring the total length of the bone

rudiment and the length of the ‘clear’ area on whole bones

(Alvarez et al., 2001; Dieudonne et al., 1994). (2) Measuring

the histologically hypertrophic area in hematoxylin and eosin

stained sections and/or (3) Measuring the area of the bone

culture containing Type X collagen (Col10a1) mRNA after

in situ hybridization. Measurements were taken from

photographs from a SZH12 dissecting microscope for

pictures of whole bones or an Olympus BX-51 upright

microscope for pictures of sections. The percent of the total

cartilage that was hypertrophic was calculated as: (length of

hypertrophic zone/total length)!100. The mean and stan-

dard deviation was calculated for each group. Significant

Table 1

RT-PCR conditions

Gene Forward Reverse

GAPDH ACCACAGTCCATGCCATCAC ACCACC

FGF2 AAGCGGCTCTACTGCAAGAACG TTCTGTC

FGF18 CAGATACCTTCGGGAGTCAAGTCC GCAGTT

differences were determined using the Student’s t-test.

Significantly different results were indicated by P!0.05.

4.2. BrdU labeling

Metatarsal bones were untreated or treated with IGF-1,

FGF2, or FGF18 for 24 h followed by treatment with

10 mM BrdU (Boehringer Mannheim) for 2.5 h as

previously described (Alvarez et al., 2001). Metatarsal

rudiments were then washed twice in PBS at 37 8C, fixed

in paraformaldehyde at 4 8C overnight, embedded in

paraffin, and cut into 5-mm sections. Sections were

deparaffinized, denatured in 2 N HCl for 20 min at 37 8C,

and neutralized in 1% boric acid/0.0285% sodium borate,

pH 7.6. Next, the sections were treated with 0.005 mg

trypsin/ml 0.05 M Tris, pH 7.6, for 3 min at 37 8C and

washed three times in PBS. Immunostaining was then

performed using Vectastain Elite staining kit (Vector

Laboratories) as described by the manufacturer. A rat

mAb directed to BrdU (Harlan) was used as the primary

antibody at a 1:200 dilution. Cy3-conjugated avidin

(Vector Laboratories) was substituted for the avidin–

biotin-peroxidase complex in the Vector Elite kit. Excess

Cy3-conjugated avidin was removed from the sections by

washing three times for 10 min each in PBS at room

temperature. The nuclei were counterstained with YoPro

(Molecular Probes), washed, and immediately mounted

with Aquapoly mount (Poly Sciences). Fluorescence was

observed and imaged using an Olympus BX-51 upright

microscope with a Magnafire digital camera and Photo-

shop software. BrdU postive and negative cells were

counted in a 400! field of cells about halfway between

the hypertrophic area and the end of the bone at both

ends of each bone. Thus, two fields were counted in each

of three to four bones under each condition. Perichon-

drium and hypertrophic cells were excluded. The

percentage of labeled nuclei for each field was calculated

as the number of red, BrdU stained cells over the total

number of cells (redCgreen) in the field multiplied by

100. The mean, standard deviation, and P-values were

determined using Microsoft Excel.

4.3. Semi-quantitative RT-PCR

RNA was extracted from cartilage cultures by lysis in

guanidine thiocyanate using the Ambion RNaqueous kit

(Cat# 1912; (Austin, TX)) and the manufacturers instructions

with the modification that the cartilage rudiments were first

Product size (bp) Anneal (8C)

CTGTTGCTGTAGCC 448 57

CAGGTCCCGTTTTGG 344 50

TCCTCGTTCAAGTCCTCC 647 57

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571 569

homogenized in the guanidine solution in a microcentrifuge

tube with a small pestle. RNA was treated with RNase free

DNase (Promega, Madison, WI) for 1 h at 37 8C, phenol:-

chloroform extracted, and ethanol precipitated. Precipitation

of RNA was facilitated with 20 mg glycogen per sample

(Boehringer Mannheim). RNA concentration was deter-

mined spectrophotometrically. For RT-PCR analysis, cDNA

was synthesized from 1 mg of total RNA using oligo dT

primers using standard protocols (Curent Protocols in

Molecular Biology, Wiley). For each sample 5 ml cDNA

was amplified with 0.2 mM primers and 0.2 mM nucleotides

for varying cycles (15–45 cycles). Samples incubated

without reverse transcriptase were used to determine if

there was DNA contamination in the RNA samples. GAPDH

was used as an internal control for the amount of cDNA used

in each reaction. The marker is the 100 bp ladder from In

Vitrogen. Band intensity was determined using Kodak 1D

software and the Kodak gel imaging system. Band intensities

were normalized to GAPDH and the fold-change in gene

expression with and without TGF-b was determined. Primer

sequences and conditions used are shown in Table 1.

4.4. Immunoprecipitation and immunoblotting

After the indicated time in culture the bones were

transferred to standard RIPA buffer and homogenized. The

amount of protein was quantified using the Biorad protein

assay and 1 mg of protein was immunoprecipitated using

an anti-FGFR3 antibody (Santa Cruz Biologicals, CA,

#SC-123) or anti-IGFR1 antibody (Oncogene Science,

#GR11T) and protein G conjugated agarose (Sigma, MO)

overnight at 4 8C. The extracts were centrifuged at

15,000 rpm for 15 min and the pellets were washed with

the lysis buffer three times. The pellet was resuspended in

standard denaturing sample buffer, heated in a boiling

water bath for 3 min then run on a 7% SDS-polyacryl-

amide gel. The protein was transferred to nitrocellulose

membranes (BIORAD) and incubated with anti-phospho-

tyrosine antibody (PY20, Oncogene Science), anti-IGF

receptor1 beta antibody (Santa Cruz Biologicals, #sc-713)

or anti-FGR3 antibody (Santa Cruz Biologicals #SC-123).

The membrane was incubated with the appropriate

HRP-conjugated secondary antibody then developed

using the ECL kit from Amersham. For Western blots of

activated IGFR1, a phospho-specific IGFR1 antibody was

used (Cell Signaling, #3021). For straight Western blots,

equal amount of protein was loaded into each well. The

blots were processed as described (Current Protocols in

Molecular Biology, Wiley).

Acknowledgements

We thank Philip Sohn for excellent technical assistance.

This work is supported by AR45605 and AR46982 from

NIAMS/ NIH to Dr Serra.

References

Alvarez, J., Horton, J., Sohn, P., Serra, R., 2001. The perichondrium plays

an important role in mediating the effects of TGF-beta1 on

endochondral bone formation. Dev. Dyn. 221, 311–321.

Alvarez, J., Sohn, P., Zeng, X., Doetschman, T., Robbins, D.J., Serra, R.,

2002. TGFbeta2 mediates the effects of hedgehog on hypertrophic

differentiation and PTHrP expression. Development 129, 1913–1924.

Ballock, R.T., Heydemann, A., Wakefield, L.M., Flanders, K.C.,

Roberts, A.B., Sporn, M.B., 1993. TGF-b1 prevents hypertrophy of

epipheseal chondrocytes: Regulation of gene expression for cartilage

matrix proteins and metalloproteases. Dev. Biol. 158, 414–429.

Beck, F., Samani, N.J., Byrne, S., Morgan, K., Gebhard, R., Brammar, W.J.,

1988. Histochemical localization of IGF-I and IGF-II mRNA in the rat

between birth and adulthood. Development 104, 29–39.

Bohme, K., Winterhalter, K.H., Bruckner, P., 1995. Terminal differen-

tiation of chondrocytes in culture is a spontaneous process and is

arrested by TGF-b2 and basic fibroblast growth factoe in synergy. Exp.

Cell Res. 216, 191–198.

Burke, D., Wilkes, D., Blundell, T.L., Malcolm, S., 1998. Fibroblast growth

factor receptors: lessons from the genes. Trends Biochem. Sci. 23, 59–

62.

Cancedda, R., Cancedda, F.D., Castagnola, P., 1995. Chondrocyte

differentiation. Int. Rev. Cytol. 159, 265–358.

Cerra, A., Mansfield, K.J., Chamberlain, C.G., 2003. Exacerbation of TGF-

beta-induced cataract by FGF-2 in cultured rat lenses. Mol. Vis. 9, 689–

700.

Chen, L., Adar, R., Yang, X., Monsonego, E.O., Li, C., Hauschka, P.V.,

et al., 1999. Gly369Cys mutation in mouse FGFR3 causes achondro-

plasia by affecting both chondrogenesis and osteogenesis. J. Clin.

Invest. 104, 1517–1525.

Chen, L., Li, C., Qiao, W., Xu, X., Deng, C., 2001. A Ser(365)/Cys

mutation of fibroblast growth factor receptor 3 in mouse downregulates

Ihh/PTHrP signals and causes severe achondroplasia. Hum. Mol. Genet.

10, 457–465.

Coffin, J.D., Florkiewicz, R.Z., Neumann, J., Mort-Hopkins, T., Dorn

2nd., G.W., Lightfoot, P., et al., 1995. Abnormal bone growth and

selective translational regulation in basic fibroblast growth factor (FGF-

2) transgenic mice. Mol. Biol. Cell 6, 1861–1873.

Colvin, J.S., Bohne, B.A., Harding, G.W., McEwen, D.G., Ornitz, D.M.,

1996. Skeletal overgrowth and deafness in mice lacking fibroblast

growth factor receptor 3. Nat. Genet. 12, 390–397.

Cosic, I., Drummond, A.E., Underwood, J.R., Hearn, M.T., 1994. In vitro

inhibition of the actions of basic FGF by a novel 16 amino acid peptide.

Mol. Cell Biochem. 130, 1–9.

Coumoul, X., Deng, C., 2003. Roles of FGFRs in mammalian development

and congenital disease. Birth Defects Res. (Part C) 69, 286–304.

Coxam, V., Miller, M.A., Bowman, B.M., Qi, D., Miller, S.C., 1995.

Insulin-like growth factor 1 and parathyroid hormone effects on the

growth of fetal rat metatarsal bones cultured in serum-free medium.

Biol. Neonate 68, 368–376.

Coxam, V., Miller, M.A., Bowman, M.B., Miller, S.C., 1996. Ontogenesis

of IGF regulation of longitudinal bone growth in rat metatarsal

rudiments cultured in serum-free medium. Arch. Physiol. Biochem.

104, 173–179.

de Los Rios, P., Hill, D.J., 2000. Expression and release of insulin-like

growth factor binding proteins in isolated epiphyseal growth plate

chondrocytes from the ovine fetus. J. Cell Physiol. 183, 172–181.

Delezoide, A.L., Benoist-Lasselin, C., Legeai-Mallet, L., le Merrer, M.,

Munnich, A., Vekemans, M., Bonaventure, J., 1998. Spatio-temporal

expression of FGFR 1, 2 and 3 genes during human embryo-fetal

ossification. Mech. Dev. 77, 19–30.

Deng, C., Wynshaw-Boris, A., Zhou, F., Kuo, A., Leder, P., 1996.

Fibroblast growth factor receptor 3 is a negative regulator of bone

growth. Cell 84, 911–921.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571570

Denker, A.E., Nicoll, S.B., Tuan, R.S., 1994. Formation of cartilage-like

spheroids by micromass cultures of murine C3H10T1/2 cells upon

treatment with TGF-b1. Differentiation 59, 25–34.

Di Nino, D.L., Long, F., Linsenmayer, T.F., 2001. Regulation of

endochondral cartilage growth in the developing avian limb: coopera-

tive involvement of perichondrium and periosteum. Dev. Biol. 240,

433–442.

Dieudonne, S.C., Semeins, C.M., Goei, S.W., Vukicevic, S., Nulend, J.K.,

Sampath, T.K., et al., 1994. Opposite effects of osteogenic protein and

TGF-b on chondrogenesis in cultured long bone rudiments. J. Bone

Miner. Res. 9, 771–780.

Dupont, J., Holzenberger, M., 2003. Biology of insulin-like growth factors

in development. Birth Defects Res. (Part C) 69, 257–271.

Garofalo, S., Kliger-Spatz, M., Cooke, J.L., Wolstin, O.,

Lunstrum, G.P., Moshkovitz, S.M., et al., 1999. Skeletal dysplasia

and defective chondrocyte differentiation by targeted overexpression

of fibroblast growth factor 9 in transgenic mice. J. Bone Miner. Res.

14, 1909–1915.

Gatherer, D., ten Dijke, P., Baird, D.T., Akhurst, R.J., 1990. Expression of

TGF-b isoforms during first trimester human embryogenesis. Devel-

opment 110, 445–460.

Haaijman, A., Karperien, M., Lanske, B., Hendriks, J., Lowik, C.,

Bronckers, A., Burger, E., 1999. Inhibition of terminal chondrocyte

differentiation by bone morphogenetic protein 7 (OP-1) in vitro

depends on the periarticular region but is independent of parathyroid

hormone-related peptide. Bone 25, 397–404.

Han, V.K.M., D’Ercole, A.J., Lund, P.K., 1987. Cellular localization of

somatomedin (insulin-like growth factor) messenger RNA in the human

fetus. Science 236, 193–197.

Heine, U.I., Munoz, E.F., Flanders, K.C., Ellingsworth, L.R., Lam, H.-

Y.P., Thompson, N.L., et al., 1987. Role of transforming growth

factor-beta in the development of the mouse embryo. J. Cell Biol.

105, 2861–2876.

Iwata, T., Chen, L., Li, C., Ovchinnikov, D.A., Behringer, R.R.,

Francomano, C.A., Deng, C.X., 2000. A neonatal lethal mutation in

FGFR3 uncouples proliferation and differentiation of growth plate

chondrocytes in embryos. Hum. Mol. Genet. 9, 1603–1613.

Iwata, T., Li, C.L., Deng, C.X., Francomano, C.A., 2001. Highly activated

Fgfr3 with the K644M mutation causes prolonged survival in severe

dwarf mice. Hum. Mol. Genet. 10, 1255–1264.

Kato, Y., Iwamoto, M., Koike, T., Suzuki, F., Takano, Y., 1988.

Terminal differentiation and calcification in rabbit chondrocyte

cultures grown in centrifuge tubes: regulation by transforming

growth factor beta and serum factors. Proc. Natl Acad. Sci. USA 85,

9552–9556.

Kono, K., Ueba, T., Takahashi, J.A., Murai, N., Hashimoto, N.,

Myoumoto, A., Itoh, N., Fukumoto, M., 2003. In vitro growth

suppression of human glioma cells by a 16-mer oligopeptide: a

potential new treatment modality for malignant glioma. J. Neurooncol.

63, 163–171.

Kveiborg, M., Flyvbjerg, A., Eriksen, E.F., Kassem, M., 2001.

Transforming growth factor-beta1 stimulates the production of

insulin-like growth factor-I and insulin-like growth factor-binding

protein-3 in human bone marrow stromal osteoblast progenitors.

J. Endocrinol. 169, 549–561.

Leonard, C.M., Fuld, H.M., Frenz, D.A., Downie, S.A., Massague, J.,

Newman, S.A., 1991. Role of transforming growth factorb in

chondrogenic pattern formation in the embryonic limb: stimulation of

mesenchymal condensation and fibronectin gene expression by

exogenenous TGF-b and evidence for endogenous TGF-b-like activity.

Dev. Biol. 145, 99–109.

Li, C., Chen, L., Iwata, T., Kitagawa, M., Fu, X.Y., Deng, C.X., 1999. A

Lys644Glu substitution in fibroblast growth factor receptor 3 (FGFR3)

causes dwarfism in mice by activation of STATs and ink4 cell cycle

inhibitors. Hum. Mol. Genet. 8, 35–44.

Liekens, S., Leali, D., Neyts, J., Esnouf, R., Rusnati, M., Dell’Era, P., et al.,

1999. Modulation of fibroblast growth factor-2 receptor binding,

signaling, and mitogenic activity by heparin-mimicking polysulfonated

compounds. Mol. Pharmacol. 56, 204–213.

Liu, J.P., Baker, J., Perkins, A.S., Robertson, E.J., Efstratiadis, A., 1993.

Mice carrying null mutations of the genes encoding insulin-like growth

factor I (Igf-1) and type 1 IGF receptor (Igf1r). Cell 75, 59–72.

Liu, Z., Xu, J., Colvin, J.S., Ornitz, D.M., 2002. Coordination of

chondrogenesis and osteogenesis by fibroblast growth factor 18.

Genes Dev. 16, 859–869.

Long, F., Linsenmayer, T.F., 1998. Regulation of growth region cartilage

proliferation and differentiation by perichondrium. Development 125,

1067–1073.

Long, F., Zhang, X.M., Karp, S., Yang, Y., McMahon, A.P., 2001. Genetic

manipulation of hedgehog signaling in the endochondral skeleton

reveals a direct role in the regulation of chondrocyte proliferation.

Development 128, 5099–5108.

Mancilla, E.E., de Luca, F., Uyeda, J.A., Czerwiec, F.S., Baron, J., 1998.

Effects of fibroblast growth factor-2 on longitudinal bone growth.

Endocrinology 139, 2900–2904.

Massague, J., 1998. TGF-b signal transduction. Annu. Rev. Biochem. 67,

753–791.

Millan, F.A., Denhez, F., Kondaiah, P., Akhurst, R., 1991. Embryonic gene

expression patterns of TGF beta 1, beta 2, and beta 3 suggest different

developmental functions in vivo. Development 111, 131–143.

Moses, H.L., Serra, R., 1996. Regulation of Differentiation by TGF-b. Curr.

Opin. Genet. Dev. 6, 581–586.

Murakami, S., Balmes, G., McKinney, S., Zhang, Z., Givol, D., de

Crombrugghe, B., 2004. Constitutive activation of MEK1 in

chondrocytes causes Stat1-independent achondroplasia-like dwarfism

and rescues the Fgfr3-deficient mouse phenotype. Genes Dev. 18,

290–305.

Naski, M.C., Ornitz, D.M., 1998. FGF signaling in skeletal development.

Front Biosci. 3, D781–D794.

Naski, M., Colvin, J., Coffin, J., Ornitz, D., 1998. Repression of hedgehog

signaling and BMP4 expression in growth plate cartilage by fibroblast

growth factor receptor 3. Development 125, 4977–4988.

Ohbayashi, N., Shibayama, M., Kurotaki, Y., Imanishi, M., Fujimori, T.,

Itoh, N., Takada, S., 2002. FGF18 is required for normal cell

proliferation and differentiation during osteogenesis and chondrogen-

esis. Genes Dev. 16, 870–879.

Orr-Urtreger, A., Givol, D., Yayon, A., Yarden, Y., Lonai, P., 1991.

Developmental expression of two murine fibroblast growth factor

receptors, flg and bek. Development 113, 1419–1434.

Pelton, R.W., Dickinson, M.E., Moses, H.L., Hogan, B.L.M., 1990. In situ

hybridization analysis of TGF-b3 RNA expression during mouse

development: comparative studies with TGF-b1 and b2. Development

110, 600–620.

Pelton, R.W., Saxena, B., Jones, M., Moses, H.L., Gold, L.I., 1991.

Immunohistochemical localization of TGF-b1, TGF-b2, and TGF-b3 in

the mouse embryo: expression patterns suggest multiple roles during

embryonic development. J. Cell Biol. 115, 1091–1105.

Peters, K.G., Werner, S., Chen, G., Williams, L.T., 1992. FGF Two

receptor genes are differentially expressed in epithelial and mesench-

ymal tissues during limb formation and organogenesis in the mouse.

Development 114, 233–243.

Peters, K., Ornitz, D., Werner, S., Williams, L., 1993. Unique expression

pattern of the FGF receptor 3 gene during mouse organogenesis. Dev.

Biol. 155, 423–430.

Roberts, A.B., Sporn, M.B., 1990. The transforming growth factor-bs, in:

Sporn, M.B., Roberts, A.B. (Eds.), Peptide Growth Factors and their

Receptors. Springer, Heidelberg, pp. 419–472.

Sandberg, M., Vurio, T., Hirrovan, H., Alitalo, K., Vuorio, E., 1988.

Enhanced expression of TGF-beta and c-fos mRNAs in the

growth plates of developing human long bones. Development 102,

461–470.

Schmid, C., 1995. Insulin-like growth factors. Cell Biol. Int. 19, 445–457.

Serra, R., Chang, C., 2003. TGF-beta signaling in human skeletal and

patterning disorders. Birth Defects Res. (Part C) 69, 333–351.

A. Mukherjee et al. / Mechanisms of Development 122 (2005) 557–571 571

Serra, R., Karapalis, A., Sohn, P., 1999. PTHrP-dependent and -

independent effects of TGF-b on endochondral bone formation.

J. Cell Biol. 145, 783–794.

Thesingh, C., Burger, E., 1983. The role of mesenchyme in embryonic long

bones as early deposition site for osteoclast progenitor cells. Dev. Biol.

95 (2), 429–438.

Tschan, T., Bohme, K., Conscience, E.M., Zenke, G.,

Winterhalter, K.H., Bruckner, P., 1993. Autocrine or paracrine

TGF-b modulates the phenotype of chick embryo sternal

chondrocytes in serum-free agarose culture. J. Biol. Chem. 5,

5156–5161.

Tsukazaki, T., Usa, T., Matsumoto, T., Enomoto, H., Ohtsuru, A.,

Namba, H., et al., 1994. Effect of transforming growth factor-beta on

the insulin-like growth factor-I autocrine/paracrine axis in cultured rat

articular chondrocytes. Exp. Cell Res. 215, 9–16.

Van Kleffens, M., Groffen, C., Rosato, R.R., Van den Eijnde, S.M., Van

Neck, J.W., Lindenbergh-Kortleve, D.J., et al., 1998. mRNA expression

patterns of the IGF system during mouse limb bud development,

determined by whole mount in situ hybridization. Mol. Cell Endocrinol.

138, 151–161.

Wang, E., Wang, J., Chin, E., Zhou, J., Bondy, C.A., 1995. Cellular

patterns of insulin-like growth factor system gene expression

in murine chondrogenesis and osteogenesis. Endocrinology 136,

2741–2751.

Wilke, T.A., Gubbels, S., Schwartz, J., Richman, J.M., 1997. Expression of

fibroblast growth factor receptors (FGFR1, FGFR2, FGFR3) in the

developing head and face. Dev. Dyn. 210, 41–52.

Xu, J., Liu, Z., Ornitz, D.M., 2000. Temporal and spatial gradients of Fgf8

and Fgf17 regulate proliferation and differentiation of midline

cerebellar structures. Development 127, 1833–1843.