Osteoblast proliferation or differentiation is regulated by relative strengths of opposing signaling pathways.

Abstract

Skeletal development requires the correct balance of osteoblast proliferation, survival, and differentiation which is modulated by a network of signaling pathways and transcription factors. We have examined the role of the AKT (PKB), and ERK1/2 signaling pathways in the osteoblast response to FGFs, which inhibit differentiation, and to IGF-1 and Wnt signaling, which promote it. Using osteoblastic cell lines as well as primary calvarial osteoblasts, we show that ERK1/2 and AKT have distinct effects in FGF-induced osteoblast proliferation and differentiation. ERK1/2 is a primary mediator of FGF-induced proliferation, but also contributes to osteoblast differentiation, while AKT is important for osteoblast survival. Signaling by IGF-1, that promotes osteoblast differentiation, strongly activates AKT and weakly ERK1/2, while the opposite results are obtained with FGF, which inhibits differentiation. By introducing a constitutively active form of AKT, we found that increased AKT activity drives osteoblasts to differentiation. Increasing the AKT signal in osteoblasts that harbor FGFR2 activating mutations, found in Crouzon (342Y) and Apert (S22W) syndromes, is also able to drive differentiation in these cells, that normally fail to differentiate. Wnt signals, that promotes differentiation, also induce AKT phosphorylation, and cells expressing active AKT have increased levels of stabilized beta-catenin, a central molecule in Wnt signaling. Our results indicate that the relative strengths of ERK and AKT signaling pathways determine whether osteoblasts are driven into proliferation or differentiation, and that the effects of AKT may be due, in part, to synergy with the Wnt pathway as well as with the Runx2 transcription factor.

Full text

Page 1

Osteoblast Proliferation or
Differentiation Is Regulated by
Relative Strengths of Opposing
Signaling Pathways
ANGELA RAUCCI, PAOLA BELLOSTA, ROBERTA GRASSI,
CLAUDIO BASILICO,*AND ALKA MANSUKHANI*
Department of Microbiology, New York University School of Medicine, New York, New York
Skeletal development requires the correct balance of osteoblast proliferation, survival, and differentiation which is modulated by a
networkof signaling pathwaysand transcription factors. Wehaveexamined therole of theAKT(PKB), and ERK1/2 signaling pathwaysin
the osteoblast response to FGFs, which inhibit differentiation, and to IGF-1 and Wnt signaling, which promote it. Using osteoblastic cell
lines as well as primary calvarial osteoblasts, we show that ERK1/2 and AKT have distinct effects in FGF-induced osteoblast proliferation
and differentiation. ERK1/2 is a primary mediator of FGF-induced proliferation, but also contributes to osteoblast differentiation, while
AKT is important for osteoblast survival. Signaling by IGF-1, that promotes osteoblast differentiation, strongly activates AKT and weakly
ERK1/2, while the opposite results are obtained with FGF, which inhibits differentiation. By introducing a constitutively active form of
AKT, we found that increased AKT activity drives osteoblasts to differentiation. Increasing the AKT signal in osteoblasts that harbor
FGFR2activatingmutations,foundinCrouzon(342Y)andApert(S22W)syndromes,isalsoabletodrivedifferentiationinthesecells,that
normallyfailtodifferentiate.Wntsignals,thatpromotesdifferentiation,alsoinduceAKTphosphorylation,andcellsexpressingactiveAKT
haveincreasedlevelsofstabilizedbeta-catenin,acentralmoleculeinWntsignaling.OurresultsindicatethattherelativestrengthsofERK
andAKTsignalingpathwaysdeterminewhetherosteoblastsaredrivenintoproliferationordifferentiation,andthattheeffectsofAKTmay
be due, in part, to synergy with the Wnt pathway as well as with the Runx2 transcription factor.
J. Cell. Physiol. 215: 442–451, 2008. ? 2007 Wiley-Liss, Inc.
The flat bones of the skull are formed by the process of
intramembranous ossification during which mesenchymal cells
condense directly to osteoprogenitors and undergo an orderly
spatial and temporal pattern of differentiation into mature,
bone forming osteoblasts. These cells secrete a matrix
(osteoid), which is eventually mineralized to form the calvarial
flat bones that are separated by flexible sutures (Morriss-Kay
and Wilkie, 2005). Activating mutations in FGF receptors are
the cause of several human autosomal dominant craniofacial
disorders whose unifying feature is craniosynostosis, or
premature fusion of the cranial sutures. In craniosynostosis
disorders, the regulated progression of osteoblast
proliferation, differentiation and apoptosis is perturbed.
While it is clear that excessive or unregulated FGF signaling
gives rise to defects in bone formation, the specific osteoblast
responsethatcausesthedefectisnotwelldefined(Marieetal.,
2005; Ornitz, 2005).
WeandothershavepreviouslyshownthatFGFsignaling has
a different effect on mature and immature osteoblasts.
Immature osteoblasts are induced to proliferate and the
proliferativeeffectislostasthecellsdifferentiate(Debiaisetal.,
1998; Mansukhani et al., 2000). Osteoblasts in differentiating
conditions continuously exposed to FGF undergo increased
apoptosisandtheirdifferentiationisblocked.Similarresultsare
produced by the constitutive expression of the
craniosynostosis-linked mutations FGFR2/S252W (Apert) or
FGFR2/C342Y (Crouzon) (Mansukhani et al., 2000). In this
study we have examined the contribution of the signal
transduction pathways ERK and AKT in the proliferation,
differentiation and apoptosis of osteoblasts, and in the FGF
response.
TheERKMAPkinasesignaltransductionpathwayisactivated
by growth factors and is associated with cell proliferation,
differentiation and survival. The PI3-kinases are important
signal transducers of responses to hormones and growth
factors. AKT (PKB), is a downstream target of
phosphphatidylinositol-3-kinase (PI3-K) and is a critical
mediator of cell proliferation and survival (Datta et al., 1999;
Brazil and Hemmings, 2001; Brazil et al., 2004). AKT can
phosphorylate a variety of targets leading to activation or
inhibition of their function. The anti-apoptotic effects of AKT
activation are linked to its ability to inhibit pro-apoptotic
proteases (caspase-9), inactivation of the Forkhead/FOXO
transcription factors that activate several pro-apoptotic
molecules, and by inhibiting the Bcl-2 family protein, BAD
(Brazil and Hemmings, 2001; Tran et al., 2003). Furthermore,
the AKT pathway is known to play an important role in bone
development (Liu et al., 2007; Peng et al., 2003).
Abbreviations: MAPK; mitogen-activated protein kinase; FBS; fetal
bovineserum;BrdU;bromodeoxyuridine;PBS;phosphate-buffered
saline; SD; standard deviation.
Contract grant sponsor: NIAMS;
Contract grant number: AR051358.
Angela Raucci’s present address is San Raffaele Scientific Institute,
DIBIT, Chromatin Dynamics Unit, via Olgettina 58, 20132 Milano,
Italy.
PaolaBellosta’spresentaddressis DepartmentofBiology,TheCity
College, Convent Avenue at 138th Street, New York, NY 10031.
*Correspondence to: Claudio Basilico or Alka Mansukhani,
Department of Microbiology, New York University School of
Medicine, New York, NY 10016. E-mail: basilc01@med.nyu.edu;
mansua01@med.nyu.edu
Received 21 May 2007; Accepted 11 September 2007
DOI: 10.1002/jcp.21323
ORIGINAL ARTICLE
442
Journal of
Journal of
Cellular
Physiology
Physiology
Cellular
? 2 0 0 7 W I L E Y - L I S S , I N C .

Page 2

InhisreportweshowthatsignalingbyIGF-1,whichisknown
to promote osteoblast differentiation, strongly activates AKT
while FGF signaling,that inhibits differentiation, causes a strong
andsustainedERK1/2activation.WefoundthatERKisrequired
forFGF-inducedproliferation, while AKT,asinmany othercell
types, plays a role in osteoblast survival. Furthermore, we find
that ERK signaling negatively regulates AKT activity and that
ERK activation is reduced in cells expressing constitutively
active AKT. We examined the hypothesis that the potent
differentiation-inducing signals of AKT in osteoblasts act via
synergizewiththeWnt-betacateninpathwaywhichareknown
to enhance osteoblast differentiation and bone mass. Our
results suggest that the effects of AKT may be due, in part,
to synergy with differentiation-inducing signals of the Wnt
pathway, and that the relative strengths of opposing AKT and
ERK signaling pathways affect the osteoblast response.
Materials and Methods
Reagents and antibodies
Antibodiesagainstphospho-AKT(Ser473),AKT,phospho-p44/42
MAPK (ERK1/2), phospho-FOXO (Ser256), phospho-GSK3
(Ser 9) were from Cell Signaling Technology (Beverly, MA). The
antibody against ERK2 was purchased from Santa Cruz
Biotechnology (Santa Cruz, CA), and against active beta catenin
fromUBI(UpstateBiotechnology,LakePlacid,NY).Theanti-Cbfa1
polyclonal antibody was obtained from of Dr. G. Karsenty,
Columbia University. The horseradish peroxidase-conjugated
anti-mouse and anti-rabbit secondary antibodies were purchased
from Promega (Madison, WI). Anti-BrdU monoclonal antibody
was from Amersham/GE Gealthcare (Fairfield, CT)
(RPN-202). Anti-mouse secondary antibody Cy3-conjugated was
from Jackson Laboratories (West Grove, PA). MEK1/2 inhibitor
(U0126) was from Cell Signaling Technology and PI-3 kinase
inhibitor (LY294002) was from Calbiochem (La Jolla, CA).
Quantitative RT-PCR analysis was performed as previously
described (Mansukhani et al., 2005). Specific primers used:
OPN/Bsp1: DIR: 50AAGGAGTCCCTCGATG REV 50-CTGA-
TGTTCCAGGCTGG-30; osterix: DIR 50-GCTCAGGT-
ACAGTGAGC-30REV: 50-ACCATGACGACAAGGG-30; cbfa1:
DIR 50-AGAAGAGCCAGGCAGGTGCT-30REV: 50-TTCG-
TGGGTTGGAGAAGCG-30; beta–actin: DIR: 50-GGCCGCC-
CTAGGCACCAG-30; REV: 50-CTCTTGATGTCACGC-
ACGATTTCCC-30.
Cell culture and preparation of primary osteoblasts
OB1, OB5, primary osteoblasts, 293 gag-pol expressing (293GP)
cellsweregrowninDulbecco’sMinimalEssentialMedium(DMEM)
supplemented with 10% fetal calf serum (FCS). Primary calvarial
osteoblasts were isolated from newborn mice as previously
described (Mansukhani et al., 2000). For differentiation, cells
were cultured in DMEM containing 10% FCS, ascorbic acid
(AA; 100 mg/ml), and b-glycerophosphate (BGP, 4 mM) and the
medium was changed every 3 days.
Cloning and retroviral infections
MEK1(AF), was obtained from Dr. J. Han, Scripps Institute, CA.
cDNAs were digested with SspI and XbaI, and subcloned into
pSF1N retroviral expression vector encoding green fluorescence
protein,pSF1n-GFP,(fromRiccardoPriore,NewYorkUniversity),
digested with Hind III, generating pSF1n-GFP/MEK1(AF).
pBabe/Myr-AKT, (which contains the myristilation sequence of
c-srcfusedtotheNH2terminusofAKT),andpBabe/Myr(K179M),
the dominant negative AKT, (Myr-AKT-K179M), retroviral
expression vectors were a gift from Dr. M. Pagano (New York
University School of Medicine). Retrovirus construct FGFR2/Cr
contains mouse FGFR2/Crouzon(C342Y). Retrovirus production
was done in 293GP cells. Cells were plated in 100 mm plate and at
60% confluence, were cotransfected with 5 mg of the indicated
plasmid and 5 mg of pCMV-VSV-G env plasmid by calcium
phosphateprecipitation(ChenandOkayama,1988).Thefollowing
day the medium was changed and 48 h later the viral supernatant
was collected. About 2?105OB1, OB1/Cruzon, or OB5 cells
were infected with 1 ml of viral supernatant in presence of 8 mg/ml
of polybrene for 4–6 h. For pools and clones (e.g., AKTMyr#7),
selection was done for 10 days in complete mediumþ2 mg/ml
puromycin.
Proliferation and apoptotic assays
Cells(1?104)wereseededontoglasscoverslipsin24-wellsplates
(Costar) in DMEMþ10% FCS. After 24 h the cells were either
serum starved in 2% FCS, or induced to differentiate with AA and
BGP in the absence or presence of FGF1 (10 ng/ml) and heparin
(5 mg/ml). The cells were labeled with 1 mg/ml of
bromodeoxyuridine (BrdU; Sigma, St. Louis, MO) for 4–6 h at
378C, fixed in 3.7% paraformaldehyde for 10 min at room
temperature,permeabilizedfor10minwith0.5%TritoninPBSand
treatedfor15minwith1.5NHCl.Afterseveralwashes,analysisof
BrdU incorporation was performed using an anti-BrdU
monoclonal antibody (Amersham) followed by anti-mouse
secondary antibody conjugated with Cy3 (1:400). Cell nuclei were
stainedwith0.5mg/mlofHoechst(Sigma)inPBSfor20minpriorto
mounting on slides. The fluorescence was visualized using an
Axioplan 2 Zeiss microscope equipped with a digital camera. The
frequencyofSphasecellswascalculatedasaratioofBrdUpositive
nucleitothetotalHoechststainednuclei.Hoechststainingwasalso
used to detect apoptosis along with TUNEL assay staining kit
(Roche (Indianapolis, IN)). Apoptotic nuclei were counted as a
proportion of total nuclei.
Alkaline phosphatase and mineralization staining
Cells were fixed with citrate-acetone and histochemically stained
for alkaline phosphatase (ALP) according to the manufacturer’s
instructions (85L-2, Sigma). To visualize mineralized nodules, cells
were fixed and stained with Alzarin Red (75 mg/ml) overnight and
washed with H2O.
Cell lysate preparation and Western analysis
Cells were washed once with cold phosphate-buffered saline and
lysed either in RIPA buffer (10 mM Tris–HCl at pH 7.2, 150 mM
NaCl, 5 mM EDTA, 0.1% SDS, 1% Na-deoxycholate, 1% Triton
X-100) or HNTG buffer (50 mM HEPES at pH 7.5, 750 mM NaCl,
1.5 mM MgCl2, 1 mM EGTA, 105 glycerol, 15 Triton X-100) in
presence of protease and phosphatase inhibitors (1 mM
phenylmethylsulfonyl fluoride, 1 mg/ml aprotinin, leupeptin, and
pepstatin, 1 mM sodium orthovanadate, 10 mM sodium fluoride).
Lysates were clarified by centrifugation at 13,000g for 20 min at
48C. Protein concentration was determined by the Bradford
methods (Biorad, Hercules, CA). Total protein extracts
(10–20 mg) were resuspended in Laemmli sample buffer, boiled
for 5 min, separated by SDS–polyacrylamide gels (SDS–PAGE) as
previously described (Raucci et al., 2004).
Results
Activation of the ERK1/2 and AKT pathways
by FGF and IGF-1
The murine OB1 osteoblast cell line can be induced to
differentiate and express osteopontin, ALP, and bone
sialoprotein,markersofosteoblastdifferentiation(Mansukhani
et al., 2000). We have previously shown that FGF stimulation
induces proliferation of immature OB1 cells but that this
proliferative response is lost as the cells differentiate. With
long-term exposure to FGF, differentiation is inhibited and the
cells undergo increased apoptosis (Mansukhani et al., 2000).
JOURNAL OF CELLULAR PHYSIOLOGY
E R K A N D A K T I N F G F R E S P O N S E O F O S T E O B L A S T S
443

Page 3

In order to determine which of the known FGF-induced
signal transduction pathways may be responsible for the
FGF response of OB1 cells, we studied the activation of the
ERK1/2, and the PI3-kinase-AKT pathways. For all our
experiments we used FGF1, a universal ligand for all FGF
receptor isoforms. Osteoblasts were stimulated with FGF1 for
different times, and cellular lysates were analyzed by Western
blot using phospho-specific antibodies that recognize active
AKT or ERK1/2. We compared the kinetics of ERK and AKT
activation by FGF with IGF-1 in OB1 as well as in OB5 cells,
which are spontaneously immortalized murine calvarial
osteoblasts (Mansukhani et al., 2005). Although FGF and IGF
both signal through tyrosine kinase receptors, FGF signaling
inhibits osteoblast differentiation while IGF is known to
stimulate this process (Yeh et al., 1997; Zhang et al., 2002).
Figure 1A shows that in OB1 or OB5 cells, FGF induces strong
andsustainedERK1/2activationthatisdetectableat15minand
persists up to 2 h, while induction of AKT activation was
detectablebutveryweak.IncontrasttoFGF,IGF-1stimulation
leadstoarobustinductionofAKTphosphorylationwhichisstill
sustained at 1 h, while ERK phosphorylation is weak and
transient and no longer detectable by 30 min. Similar results
were obtained in mouse primary calvarial osteoblasts (data not
shown). Thus the relative signaling strength of these two
pathways differs, with IGF-1 preferentially activating the AKT
pathway, and FGF stimulation favoring the ERK pathway.
Highlighting the opposing effects of these pathways on
osteoblasts, Figure 1B shows that, unlike FGF, IGF-1 does not
stimulate proliferation of OB1 cells, and slightly reduces the
extent of BrdU incorporation by FGF (P¼0.05). While FGF
enhances apoptosis, IGF-1 treatment reduces the extent of
apoptosis in differentiating OB1 cells and reduces the extent of
apoptosis induced by FGF (P¼0.05).
The striking inverse relationship between AKT and
ERK1/2 phosphorylation observed in IGF or FGF treated cells
ledustoinvestigatewhetherthesepathwaysmayinterferewith
each other, such that inhibition of ERK signaling would
potentiate the activation of AKT by FGF, and inhibition of AKT
signals would enhance activation of ERK1/2. We treated OB1
cells with UO126, a MEK1/2 inhibitor, which prevents
ERK1/2 phosphorylation, and then subjected them to FGF
stimulation.TheresultsshowninFigure1Cclearlyindicatethat
inhibition of the ERK pathway results in stronger AKT
phosphorylation induce by FGF that is detectable up to 60 min.
In the converse experiment,, OB1 cells treated with the
LY294002, an inhibitor of the AKT pathway, showed that
inhibiting AKT activation by IGF produced no significant
enhancement of ERK activation by IGF.
Effect of ERK1/2 and AKT pathways on FGF-induced
proliferation of OB1 cells
ToassessthecontributionofERK1/2andAKTactivationtothe
proliferative effect of FGF on OB1 cells we used chemical
inhibitors that are known to specifically prevent the activation
of these pathways. We used the MEK1/2 inhibitor, U0126, and
thePI3-kinase inhibitor, LY294002,toprevent theactivation of
ERK1/2andAKTrespectively.Theinhibitorsatconcentrations
up to 30 mM blocked the activation of the respective pathways
while maintaining specificity (data not shown).
We performed BrdU incorporation assays to test the effect
of these inhibitors on FGF-induced proliferation of OB1 cells
that had been serum starved for 24 h. In accordance with our
previous results (Mansukhani et al., 2000), FGF1 induced a
modest increase in proliferation of OB1 cells Increasing
concentrations of the MEK1/2-ERK1/2 inhibitor, U0126,
Fig. 1.
(100 ng/ml) for the indicated times. B: Analysis of BrdU incorporation and apoptosis (TUNEL assay) in OB1 cells after 24-h treatment with
FGF,IGFor1:1FGFandIGFat100ng/mleach.Fortheapoptosisassay,cellsweredifferentiatedfor24hpriortotreatment.Thedatashownarethe
mean of three independent experimentWSD. Statistical differences between treatments were measured by ANOVA. P-values at or below
0.05wereconsideredsignificant.C:EffectofspecificpathwayinhibitorsonOB1cells.EffectoftheMEKinhibitor,U0126,onAKTactivationbyFGF
(10 ng/ml) for 15 and 60 min. D, DMSO vehicle (left part). Effect of PI3K-AKT inhibitor LY294002 (LY) on ERK11/2 activation by IGF (10 ng/ml)
for 15 and 60 min.
Signaling pathways in osteoblasts. A: Comparison of AKT and ERK activation in OB1 and OB5 osteoblasts treated with FGF or IGF
JOURNAL OF CELLULAR PHYSIOLOGY
444
R A U C C I E T A L .

Page 4

reduced the FGF proliferative response. The PI3K inhibitor,
LY294002, did not appear to significantly affect the percentage
of cells incorporating BrdU in response to FGF. All the
inhibitors by themselves did not have any effect on the basal
proliferation of OB1 cells at 24 h (Figure 2A). Thus, these
results indicate that ERK1/2 signaling is required for the
FGF-induced proliferation of osteoblast cells.
WealsoinvestigatedthecontributionofsteadystateERK1/2
pathway on the proliferation in OB1 cells, grown in complete
medium in the absence of additional growth factors, by using
dominant negative MEK1 MEK1(AF) to specifically block ERK
signaling, (Wang et al., 2002; Li et al., 2006). OB1 cells were
infected with retroviruses carrying the vector expressing GFP
alone, or dominant negative plasmids GFP-MEK1(AF) for 24 h
and a BrdU incorporation assay was performed using GFP to
trackcellsexpressingthetransducedplasmids.Figure2Bshows
the percentage of GFP positive and GFP negative control cells
incorporating BrdU. The GFP expressing vector by itself does
not interfere with the incorporation of BrdU, while the
expression of a constitutively activated FGFR2 (FGFR2/Cr)
enhances the proliferative rate of the GFP-positive cells.
Expression of the dominant negative MEK1(AF) clearly
decreases the proliferation of OB1 cells. Thus, these results
suggest that, for OB1 growing in 10% serum, as well as for
FGF-induced proliferation of OB1, ERK1/2 is a major mediator
of FGF-induced osteoblast proliferation.
ERK1/2 and AKT contribute to osteoblast differentiation
Mouse calvarial osteoblasts are able to differentiate in culture.
Mature cells are characterized by increased expression of ALP,
anearlymarkerofdifferentiation,increasedmatrixproduction,
and eventual mineralization of the extracellular matrix. To
determine whether the activity of ERK1/2 and AKT is required
for osteoblasts differentiation, we used primary calvarial
osteoblasts that differentiate more rapidly than osteoblast cell
lines. Figure 3A shows that both ERK1/2 and AKT are
phosphorylated during primary osteoblasts differentiation.
Cells were then induced to differentiate in the presence and
absence of the MEK1/2-ERK/2 inhibitor, U0126 or the PI3K
inhibitor,LY294002.Theextentofdifferentiationwasassessed
by staining for ALP expression at days 3, 7, and 18.
Figure 3B shows that by 7 days primary osteoblasts stained
strongly positive for ALP and mineralization is detectable by
18 days of differentiation. When U0126 was included in the
medium, ALP staining was inhibited at 7 days and the inhibitory
effect on mineralization was evident at 18 days. Although it
appears that ALP staining is also inhibited in the presence of
LY294002, this is primarily due to an increase in apoptosis as a
result of blocking the AKT survival pathway and consequent
lossofthecellsduringtheexperiment.AsseeninFigure3C,the
inhibitor U0126 has a minor effect on apoptosis compared to
vehicle treated cells, while with the inhibitor LY294002 the
apoptosis rate is dramatically increased to 56% at 3 days and by
day 7, 90% of cells undergo apoptosis.
Together these results suggest that ERK1/2 activation
contributespositivelytoosteoblastdifferentiationwhileAKTis
required for osteoblasts survival during the differentiation
process. The block in differentiation was also seen in the
reduction of osteopontin/BSP1 (spp1) determined by RT-PCR
analysis of RNA from cells treated with the inhibitors (data not
shown).
To confirm the results obtained with chemical inhibitors on
differentiation,wealsoassesseddifferentiationofpoolsofOB1
infected with retroviruses expressing dominant negative
constructs that specifically blocked each pathway. Dominant
negative MEK1 MEK1(AF), was introduced into OB1 cells as
well as into OB5 cells. OB5 cells differentiate faster, and more
synchronously than OB1 cells and differences in changes in
differentiation are more apparent. We confirmed that pools of
cells expressing MEK1(AF) exhibited impaired differentiation
(data not shown).
Increased AKT signaling enhances
osteoblast differentiation
WhileitisclearthattheAKTpathwayisrequiredforosteoblast
survival during differentiation, we could not assess whether it
playsanadditionalroleinpromotingorblockingdifferentiation.
SinceFGFblocksdifferentiationandweaklystimulatestheAKT
pathway, while 1GF, a potent AKT activator, induces
differentiation,wedeterminedwhetherincreasingAKTactivity
affected osteoblast differentiation. A constitutively active form
of AKT (AKT-Myr), a wild-type AKT (AKT-wt) or a dominant
negative form of AKT (AKT-DN) have been previously
Fig. 2.
cellsoncoverslipswerestarvedovernightin2%serumandtreatedwith
the indicated concentrations of MEK1/2 inhibitor (U0126) or PI3K
inhibitor(LY294002).After1hFGF1(10ng/ml)andheparin(5mg/ml)
wereaddedfor24handthecellswerelabeledwithBrdUforthelast4h.
PercentageofcellsincorporatingBrdUwasdeterminedbycountingat
least 300 cells. The results shown are the mean of three
experimentWSD. B: Unstarved OB1 cells were infected with the
indicated retrovirus constructs tagged with GFP and BrdU
incorporationassaywasperformed48hlater.ThepercentageofBrdU-
positivecellswasdeterminedasaproportionofinfectedcells(GFP)and
of non-green uninfected cells (control) on the same coverslip. The
results shown are the mean of three experimentWSD. Dominant
negative constructs, MEK1(AF), FGFR2/Cr is the C342Y activated
FGFR2.
EffectofERK1/2andAKTonproliferationofOB1cells.A:OB1
JOURNAL OF CELLULAR PHYSIOLOGY
E R K A N D A K T I N F G F R E S P O N S E O F O S T E O B L A S T S
445

Page 5

described (Kohn et al., 1996; Aoki et al., 1998). These
constructs were introduced into OB1 or into OB5 cells using
retroviral vectors. Cells were transferred to differentiation
medium, and 7 days later they were fixed and stained for the
expressionofALP. AsshowninFigure4A,only?30%ofa pool
ofOB5expressingvectoraloneorwild-typeAKT(wtAKT)are
positive for ALP, while cells expressing the constitutively
activated form of AKT (AKT-Myr) clearly show increased ALP.
Thepotentdifferentiationinducingability ofMyr-AKTwas also
evidentinOB1cellsthatdifferentiatelesswellthanOB5,aswell
asinOB1/CrouzoncellsthatexpressactivatedFGFR2,andthat
usually do not differentiate. This result suggests that increasing
AKT signaling can overcome the inhibitory effect of FGF on
differentiation. The increased differentiation of OB1 or
OB5 cells expressing AKT-Myr was confirmed by assessing the
mRNA expression of differentiation markers ALP and
bonesialoprotein (Bsp1) by Northern analysis at 0 and 5 days
of differentiation. Both BSP1 and ALP were increased in
OB5-AKT-Myr cells relative to the control cells expressing
vector alone, AKT-wt or AKT-DN (Fig. 4B).
We also created stable clones of OB5 cells expressing
AKT-Myr and confirmed that high-expressing clones
differentiatedbetterthantheparentalorlow-expressingclones
(data not shown). Figure 4C shows a comparative time
course of differentiation of OB5 cells, and of OB5 cells
expressing AKT-Myr or the dominant negative form of AKT.
The accelerated differentiation exhibited by the AKT-Myr cells
is in stark contrast to the poor differentiation of the AKT-DN
cells. We then examined the expression of two transcription
factors, Cbfa1/Runx2 and osterix that are required for
osteoblast differentiation (Nakashima and de Crombrugghe,
2003). It has been reported that signaling through AKT can
potentiate osteoblast differentiation by cooperation with the
transcription factor Runx2/CbfaI (Fujita et al., 2004).
Figure 4D shows that the mRNA level of cbfa1 is elevated
threefold in AKT-Myr cells by 5 days of differentiation
(P¼0.01),whilethelevelsofosterix,whichliesdownstreamof
cbfa1 in the osteoblast differentiation program, is strongly
induced in AKT-Myr expressing cells prior to differentiation
and remains elevated at day 5. Thus AKT activation could
Fig. 3.
up to 14 days. Cellular lysates were collected on the indicated days and analyzed by Western blot using the indicated antibodies. Cells were
stained for ALP expression at days 0 and 14 and percent of ALP-positive cells was determined. B: Effect of ERK1/2 and AKT inhibitors on ALP
expression during osteoblast differentiation. Equal numbers of cells were seeded on 6-well plate and maintained in differentiation medium
containing 20 mM U0126, LY294002, or DMSO 0.1%. Plates were fixed and stained for the expression of ALP (purple) at 7 and 18 days.
Mineralization (red-brown) nodules were detected by Alizarin Red stain. C: Effect of ERK1/2 and AKT inhibitors on apoptosis during osteoblast
differentiation. Cells were treated as in B, and coverslips were fixed and stained with Hoechst. Two hundred nuclei from at least five
different microscope fields were counted and the apoptotic nuclei were calculated as percentage of total cells. Experiment was repeated
twice with similar results. Results of one experiment are shown.
ERK1/2 and AKT contribute to osteoblast differentiation. A: Primary calvarial osteoblasts were maintained in differentiation medium
JOURNAL OF CELLULAR PHYSIOLOGY
446
R A U C C I E T A L .

Page 6

accelerate osteoblast differentiation by inducing factors
required for the process.
We also examined the effect of AKT-Myr on a known
downstream effector of the AKT pathway (Tran et al., 2003).
Figure 5A shows that the AKT target FOXO3, a pro-apoptotic
transcription factor known to be inactivated by AKT, is
phosphorylated in OB1, OB5, or OB1/Crouzon cells
expressing AKT-Myr. Interestingly, the basal level of FOXO3
phosphorylation is higher in OB5 cells, which differentiate
better than OB1 cells. The effect of AKT-Myr and AKT-DN on
apoptosiswasmonitoredduringdifferentiation.At3and7days
of differentiation, AKT-Myr expressing cells have a lower rate
of apoptosis than control cells, while AKT-DN cells have
increased apoptosis (Fig. 5B). This trend is also evident in
OB1/Crouzon cells as well which we have previously shown to
havehigher apoptosis. ThusAKT signaling increases osteoblast
differentiation and survival during differentiation.
AKT signaling affects molecules in the Wnt pathway
Another downstream target of AKT is the enzyme GSK3b
whichisinactivatedbyphosphorylationonserine9(Braziletal.,
2004;JopeandJohnson,2004).GSK3bisanegativeregulatorof
the canonical Wnt b-catenin pathway, and Wnt signaling has
been shown to promote differentiation in osteoblasts
(Westendorf et al., 2004). Active GSK3b phosphorylates
b-catenin on conserved serine/threonine residues in the
N-terminus, that target it for degradation, and Wnt pathway
activation leads to the inactivation of GSK3b. The stabilized
b-cateninistranslocatedtothenucleusandcomplexeswiththe
LEF/TCF family of transcription factors to regulate Wnt target
genes (Clevers, 2006; Nelson and Nusse, 2004). Although
GSK3 inactivation by Wnt and AKT are thought to occur via
distinct mechanisms, we found that Wnt3A treatment of OB1
osteoblasts also induced GSK3b phosphorylation on serine 9,
andalsocausedaverystrongphosphorylationofAKT(Fig.6A).
This finding is in line with the notion that, like Wnt3A, AKT
activity promotes differentiation. We therefore examined
whether these pathways cooperate and whether increased
AKT activity in osteoblasts and consequent inactivation of
GSK3 would result in enhancement of b-catenin/Wnt activity.
We examined whether GSK3b was phosphorylated and
whether b-catenin was stabilized and increased in AKT-Myr
cells. Figure 6B shows that p-GSK3 is increased in AKT-Myr
Fig. 4.
thepBabevector,pBabe/AKT-wt,pBabe/AKT-Myr,orpBAbe/AKT-dominantnegative(DN).After48h,atconfluence,thecellsweremaintained
indifferentiationmediumfor7days,fixedandstainedforALPexpression(purplestain).B:Expressionofdifferentiationmarkerbonesialoprotein
(Bsp1) and ALP by northern analysis of OB5 cells infected with the indicated constructs at day 0 or 5 of differentiation. C: OB5 cells were stably
infected with AKT-Myr or AKT-DN construct and a pool (AKT-Myr or AKT-DN) or clone (AKT-Myr # 7) was selected in puromycin.
EqualnumbersofcellswereplatedfordifferentiationandstainedforALPontheindicateddays(purplestain).D:QuantitativeRT-PCRanalysisof
cbfa1 andosterix mRNA levels from stably infected OB5cells as indicated atdays0and5of differentiation. Actinwas usedas aninternal control
tonormalizeforRNAlevelswithineachsample.DataarepresentedasfoldchangeoverthelevelinOB5cellswhichwassetat1.Eachtimepointis
the average of 2–3 samplesWSD. Statistical differences between treatments were measured by ANOVA. P-values at or below 0.05 were
considered significant.
Increased AKT signaling enhances osteoblast differentiation. A: OB5, OB1, or OB1/Crouzon cells were infected with virus carrying
JOURNAL OF CELLULAR PHYSIOLOGY
E R K A N D A K T I N F G F R E S P O N S E O F O S T E O B L A S T S
447

Page 7

OB5 cells and reduced in OB5 cells expressing the dominant
negative, AKT-DN. We used an antibody that specifically
recognizes an epitope containing the unphosphorylated
serine residues in the N-terminus of b-catenin, that is the
target of GSK3 phosphorylation, and thus identifies active
b-catenin(vanNoortetal.,2002).InOB5cells,activeb-catenin
is detectable in untreated cells, is increased by Wnt3A
treatment, and decreased in FGF-treated cells. In
OB5/AKT–Myr cells, the levels of active B-catenin were
higher in untreated and Wnt-treated cells (Fig. 6B).
Figure 6C shows that in transiently transfected OB1 or
OB1/Crouzon cells, GSK3 phosphorylation is clearly increased
in cells expressing Myr-AKT, in which AKT is constitutively
phosphorylated. Active b-catenin, which is essentially
undetectable in untransfected cells, is visible in cells
expressing Myr-AKT. These data suggest that increasing the
amplitude of AKT signaling may increase differentiation of
osteoblasts by contributing to the activation of Wnt signaling
pathway via activation of b-catenin. We also found that
constitutive activation of AKT results in decreased
phosphorylation of ERK1/2 in osteoblasts expressing
AKT-Myr (Fig. 6C), in line with the notion that these pathways
have opposing effects.
Wntsignalinghasbeenreportedtopromoteosteogenesisby
stimulating Runx2 expression (Gaur et al., 2005). To confirm
whether Runx2protein expression was elevated inosteoblasts
expressing Myr-AKT in the transient assay, we examined
protein expression and found that Runx2/Cbfa1 protein is
indeedincreasedinthesecells(Fig.6D).Thusthepositiveeffect
of AKT on differentiation is likely to involve increased Runx2
expression as well.
Discussion
In this report we have examined the contribution of the
ERK1/2 and AKT (PKB) signaling pathways to the proliferation,
differentiation, and apoptosis of osteoblasts. This complex
process is controlled by a network of temporally and spatially
regulated signals, that when disrupted result in abnormal bone
formation and craniofacial deformities.
While it is known that several transcription factors, such as
Runx2 and Osterix are essential for osteoblast commitment
and differentiation/function, the process by which signaling
molecules and specific signaling pathways regulate osteoblast
function are still unclear (Nakashima and de Crombrugghe,
2003). FGFs, IGFs and Wnt signaling are known to regulate
osteoblast differentiation. In vitro, FGF signaling increases
osteoblast proliferation and inhibits differentiation, while IGF
and Wnt promote differentiation both in vitro and in vivo
(Yeh et al., 1997; Zhang et al., 2002; Krishnan et al., 2006). FGF
signalinginvitroantagonizestheWnteffects(Mansukhanietal.,
2005).
Although they utilize similar tyrosine kinase receptors, FGF
and IGF-1 have opposing biological effects on osteoblast
proliferationanddifferentiations.WeshowthatwhileFGFacts
preferentially through the ERK pathway and only weakly
stimulates AKT activation, IGF-1 treatment of the same cells
leadstoa stronginductionofAKT phosphorylationanda weak
ERK response. AKT is thought to be the main effector of IGF-1
signaling and AKT null mice and IGFR-1 null mice have a similar
skeletal phenotype of delayed ossification (Peng et al., 2003).
More recently, an osteoblast-specific knockout of the AKT
inhibitor, PTEN resulted in mice with increased bone mineral
Fig. 4.(Continued)
JOURNAL OF CELLULAR PHYSIOLOGY
448
R A U C C I E T A L .

Page 8

density (Liu et al., 2007). Thus, AKT activation correlates with
osteoblast differentiation while ERK activation correlates with
proliferation. Several reports suggest that the ERK and AKT
pathways have opposing effects in other cell types. A strikingly
similar situation has been observed in myoblasts treated with
HGF and IGF-1. In these cells HGF, that inhibits differentiation,
induces strong ERK1/2 activation and a weak AKT response,
while IGF-1, which promotes differentiation, strongly activates
the P13-K/AKT pathway and only weakly the MAP/ERK
kinases (Halevy and Cantley, 2004). Furthermore, ERK and
AKT are activated downstream of the VEGF receptor in
angioblasts-cells, and have opposing roles in artery/vein
specification (Hong et al., 2006).
We show that blocking the ERK pathway with chemical
inhibitors leads to increased activation of AKT by FGF in OB1
cells, suggesting that active ERK signaling blocks the AKT
pathway. While this observation is intriguing, its mechanism is
still unclear. ERK1 and 2 exert a negative feedback on the
function of FRS2, an essential FGF signaling intermediate, by
phosphorylating FRS2 on threonine residues (Lax et al., 2002).
FRS2 is also involved in PI3-K activation (Hadari et al., 2001),
and it is possible that inhibiting threonine phosphorylation of
FRS2favors activationorrecruitmentoftheregulatory subunit
of P13K, or simply potentiates FRS2 activity. Although in the
reciprocal experiment, in the presence of AKT pathway
inhibitors, we did not observe an appreciable induction of
IGF-inducedERKactivity,wedidfindthatERKphosphorylation
is suppressed in osteoblasts stably expressing constitutively
activeAKT(AKT-Myr).Thesedatasuggestthatthesepathways
cross-regulate and may oppose each other in osteoblasts.
Using chemical inhibitors, we show that blocking the ERK
pathway impairs osteoblast proliferation while blocking AKT
has no effect. We found that the ERK pathway contributes to
osteoblast differentiation as well, but has little effect on
osteoblast survival. The requirement for ERK activity during
differentiation was somewhat unexpected since ERK signals
induce proliferation and inhibit the differentiation-inducing
effects of BMP (Osyczka and Leboy, 2005). However, ERK
activity may induce factors required early in the differentiation
program and is induced by differentiation-inducing signals such
as Wnt. On the other hand, inhibitors of AKT pathway applied
during differentiation resulted in increased apoptosis,
highlighting the primary role for the AKT-pathway in survival
duringosteoblastdifferentiation.Inosteoblasts,AKTactivation
occurs downstream of strong differentiation inducing signals
such as BMP, and it has been shown to be essential for
BMP-induced expression of ALP and differentiation
(Ghosh-Choudhury et al., 2002). Our results from two
osteoblastlines,OB1andOB5,clearlyindicatethatconstitutive
activationoftheAKTpathwaybyintroductionofamyristylated
form of AKT, enhances osteoblast differentiation. This effect is
not only due to prevention of apoptosis, which is not
particularly high in untreated cells, but is likely to involve a
direct or indirect effect of AKT on the expression of
differentiation genes such as cbfa1 and osterix. These findings
are also in line with the hypothesis that the opposite effects of
Fig. 5.
infected with the viruses carrying the indicated cDNAs, or uninfected, (?), were analyzed by Western blot using antibodies against AKT,
phospho-FOXO3orERK2asaloadingcontrol.AKTexpressedbythecDNAsmigrateslowerthantheendogenousAKT.B:Afterinfection,OB5and
OB1/Crouzon expressing AKT-DN, AKT-Myr or just the vector pBabe were maintained for 3 and 7 days in differentiation medium. Cells were
then fixed and stained for TUNEL and Hoechst. Nuclei from at least ten different fields were counted and the apoptotic nuclei were
calculated as percentage of total cells. Data are represented as the meanWSD.
AKT signaling increases osteoblast survival during differentiation. A: Cellular lysates from OB5, OB1, or OB1/Crouzon, previously
JOURNAL OF CELLULAR PHYSIOLOGY
E R K A N D A K T I N F G F R E S P O N S E O F O S T E O B L A S T S
449

Page 9

FGF and IGF could be due to the reciprocal strengths of
activation of the ERK and AKT pathways.
We have previously described that FGF can inhibit Wnt
signaling and differentiation in osteoblasts, and represses
several genes identified as Wnt targets (Mansukhani et al.,
2005).ItwasthereforeconceivablethatincreasedAKTactivity,
that enhances differentiation, could have potentiated the
activation of Wnt pathways in osteoblasts, since AKT is known
to phosphorylate and inactivate GSK3b, an inhibitor of the
canonical Wnt pathway. In line with the hypothesis that AKT
can cooperate with the Wnt pathway, we could show that in
osteoblasts expressing AKT-Myr, GSK3b is phosphorylated
and the amount of active b-catenin enhanced. We also found
that in addition to activating the Wnt-b-catenin pathway,
Wnt3A treatment of OB1 osteoblasts also activated AKT.
Activation of the PI3K-AKT pathway has also been shown to
playaroleintheanti-apoptoticeffectofWnt3A(Almeidaetal.,
2005).
Toprove moredirectly thatincreasedAKT activitytargeted
b-catenin/Wnt target genes, we cultured calvarial osteoblasts
from transgenic mice harboring a TOPGAL plasmid, where the
b-galactosidase gene is controlled by a b-catenin/TCF
responsive promoter (DasGupta and Fuchs, 1999). Cells were
infected with a retroviral vector expressing AKT-Myr, or
treatedwithIGF-1orWnt3A,andb-galactivity wasmeasured.
Wecouldshowthatcellswhichhadbeenkeptindifferentiation
medium for 14 days had high b-gal activity (about 15? that of
controls) and that IGF-1 and Wnt3A also stimulated b-gal
activity about five- to sixfold, but infection with a retrovirus
expressingMyr-AKThadnosignificanteffect(datanotshown).
Twopossible explanationsforthisnegativeresultarethat(i)as
recentlydiscussed(Barolo,2006),ithastobeconsideredthata
transgenic b-catenin/TCF reporter may not give a complete
readout of Wnt target gene activation. (ii) It is possible that in
osteoblastsasinothercelltypes,inactivationofGSK3byAKTis
distinctfrominactivationofGSK3byWnt(Dingetal.,2000).By
quantitative RT-PCR analysis of the mRNA of Wnt-induced
secreted protein (WISP-1), a Wnt-responsive gene that plays a
role in bone development (French et al., 2004), we found that
WISP-1 was elevated in AKT-Myr expressing osteoblasts but
WISP-2 was not (data not shown). Further experiments
defining Wnt target genes in osteoblasts, that are in progress,
will be needed to clarify this issue.
An additional mechanism by which AKT activation leads to
differentiation may be via transcription factors such as osterix
and runx2 that are essential for osteoblast differentiation and
bone development. AKT signaling has been reported to
enhance DNA binding of Runx2 and Runx2 upregulates PI3-K
subunits suggesting that Runx2 and AKT may be mutually
dependent in osteoblasts (Fujita et al., 2004). Runx2 has also
been shown to be a direct target of Wnt-beta catenin signaling
(Gaur et al., 2005). The increase in Runx2 mRNA and protein
Fig. 6.
withWnt3A(100ng/ml)for5or15min.B:Steady-statelevelsofp-GSK3binOB5control,OB5expressingAKT-Myr(Myr)orAKT-DN,(DN)cells
(upper part). Active beta catenin in OB5 control or OB5 expressing AKT-Myr in cells left untreated in complete medium (?) or treated
with FGF1 (F) or Wnt3A (W) for 24 h. C: p-AKT, p-GSK3, p-ERK and active beta catenin in OB1 or OB1/Crouzon cells expressing AKT-Myr
(Myr), AKT-DN (DN), or AKT-wt (wt). D: OB1 osteoblasts expressing the indicated constructs were analyzed by Western blot for Cbfa1
expression.
Signaling molecules in the Wnt-AKT pathway. A: Western analysis of AKT and GSK3b activation in OB1 cells untreated (0) or treated
JOURNAL OF CELLULAR PHYSIOLOGY
450
R A U C C I E T A L .

Page 10

expression that we observe in osteoblasts expressing
myristylated AKT, suggests that the AKT pathway may
cooperate with Wnt signaling to promote osteoblast
differentiation.
Acknowledgments
We thank Lizbeth Cornivelli for excellent technical assistance,
and M. Pagano and R. Priore for plasmid constructs. We also
thank Dr. J. Han for the dominant negative MEK1. This
investigation was supported by PSH Grant AR051358 from the
NIAMS.
Literature Cited
Almeida M, Han L, Bellido T, Manolagas SC, Kousteni S. 2005. Wnt proteins prevent
apoptosis of both uncommitted osteoblast progenitors and differentiated osteoblasts by
beta-catenin-dependent and -independent signaling cascades involving Src/ERK and
phosphatidylinositol 3-kinase/AKT. J Biol Chem 280:41342–41351.
Aoki M, Batista O, Bellacosa A, Tsichlis P, Vogt PK. 1998. The akt kinase: Molecular
determinants of oncogenicity. Proc Natl Acad Sci USA 95:14950–14955.
Barolo S. 2006. Transgenic Wnt/TCF pathway reporters: All you need is Lef? Oncogene
25:7505–7511.
BrazilDP,HemmingsBA.2001.TenyearsofproteinkinaseBsignalling:AhardAkttofollow.
Trends Biochem Sci 26:657–664.
BrazilDP,YangZZ,HemmingsBA.2004.Advancesinprotein kinaseBsignalling: AKTionon
multiple fronts. Trends Biochem Sci 29:233–242.
Chen CA, Okayama H. 1988. Calcium phosphate-mediated gene transfer: A highly efficient
transfection system for stably transforming cells with plasmid DNA. Biotechniques
6:632–638.
Clevers H. 2006. Wnt/beta-catenin signaling in development and disease. Cell 127:469–480.
DasGupta R, Fuchs E. 1999. Multiple roles for activated LEF/TCF transcription complexes
during hair follicle development and differentiation. Development 126:4557–4568.
Datta SR, Brunet A, Greenberg ME. 1999. Cellular survival: A play in three Akts. Genes Dev
13:2905–2927.
Debiais F, Hott M, Graulet AM, Marie PJ. 1998. The effects of fibroblast growth factor-2 on
human neonatal calvaria osteoblastic cells are differentiation stage specific. J Bone Miner
Res 13:645–654.
Ding VW, Chen RH, McCormick F. 2000. Differential regulation of glycogen synthase kinase
3beta by insulin and Wnt signaling. J Biol Chem 275:32475–32481.
FrenchDM,KaulRJ,D’SouzaAL,CrowleyCW,BaoM,FrantzGD,FilvaroffEH,DesnoyersL.
2004. WISP-1 is an osteoblastic regulator expressed during skeletal development and
fracture repair. Am J Pathol 165:855–867.
Fujita T, Azuma Y, Fukuyama R, Hattori Y, Yoshida C, Koida M, Ogita K, Komori T. 2004.
Runx2inducesosteoblastandchondrocytedifferentiationandenhancestheirmigrationby
coupling with PI3K-Akt signaling. J Cell Biol 166:85–95.
GaurT,LengnerCJ,HovhannisyanH,BhatRA,BodinePV,KommBS,JavedA,vanWijnenAJ,
Stein JL, Stein GS, Lian JB. 2005. Canonical WNT signaling promotes osteogenesis by
directly stimulating Runx2 gene expression. J Biol Chem 280:33132–33140.
Ghosh-Choudhury N, Abboud SL, Nishimura R, Celeste A, Mahimainathan L, Choudhury
GG. 2002. Requirement of BMP-2-induced phosphatidylinositol 3-kinase and Akt serine/
threonine kinase in osteoblast differentiation and Smad-dependent BMP-2 gene
transcription. J Biol Chem 277:33361–33368.
Hadari YR, Gotoh N, Kouhara H, Lax I, Schlessinger J. 2001. Critical role for the
docking-protein FRS2 alpha in FGF receptor-mediated signal transduction pathways. Proc
Natl Acad Sci USA 98:8578–8583.
Halevy O, Cantley LC. 2004. Differential regulation of the phosphoinositide 3-kinase and
MAP kinase pathways by hepatocyte growth factor vs. insulin-like growth factor-I in
myogenic cells. Exp Cell Res 297:224–234.
HongCC,PetersonQP,HongJY,PetersonRT.2006.Artery/veinspecificationisgovernedby
opposingphosphatidylinositol-3kinaseand MAPkinase/ERKsignaling.CurrBiol 16:1366–
1372.
Jope RS, Johnson GV. 2004. The glamour and gloom of glycogen synthase kinase-3. Trends
Biochem Sci 29:95–102.
KohnAD,TakeuchiF,RothRA.1996.Akt,apleckstrinhomologydomaincontainingkinase,is
activated primarily by phosphorylation. J Biol Chem 271:21920–21926.
Krishnan V, Bryant HU, Macdougald OA. 2006. Regulation of bone mass by Wnt signaling.
J Clin Invest 116:1202–1209.
LaxI,WongA,LamotheB,LeeA,FrostA,HawesJ,SchlessingerJ.2002.Thedockingprotein
FRS2alpha controls a MAP kinase-mediated negative feedback mechanism for signaling by
FGF receptors. Mol Cell 10:709–719.
Li Z, Zhang G, Feil R, Han J, Du X. 2006. Sequential activation of p38 and ERK pathways by
cGMP-dependentproteinkinaseleadingtoactivationoftheplateletintegrinalphaIIbbeta3.
Blood 107:965–972.
Liu X, Bruxvoort KJ, Zylstra CR, Liu J, Cichowski R, Faugere MC, Bouxsein ML, Wan C,
Williams BO, Clemens TL. 2007. Lifelong accumulation of bone in mice lacking Pten in
osteoblasts. Proc Natl Acad Sci USA 104:2259–2264.
Mansukhani A, Bellosta P, Sahni M, Basilico C. 2000. Signaling by fibroblast growth
factors (FGF) and fibroblast growth factor receptor 2 (FGFR2)-activating mutations
blocks mineralization and induces apoptosis in osteoblasts. J Cell Biol
149:1297–1308.
Mansukhani A, Ambrosetti D, Holmes G, Cornivelli L, Basilico C. 2005. Sox2 induction by
FGF and FGFR2 activating mutations inhibits Wnt signaling and osteoblast differentiation.
J Cell Biol 168:1065–1076.
Marie PJ, Coffin JD, Hurley MM. 2005. FGF and FGFR signaling in chondrodysplasias and
craniosynostosis. J Cell Biochem 96:888–896.
Morriss-Kay GM, Wilkie AO. 2005. Growth of the normal skull vault and its alteration in
craniosynostosis: Insightsfrom human geneticsand experimentalstudies. J Anat 207:637–
653.
Nakashima K, de Crombrugghe B. 2003. Transcriptional mechanisms in osteoblast
differentiation and bone formation. Trends Genet 19:458–466.
Nelson WJ, Nusse R. 2004. Convergence of Wnt, beta-catenin, and cadherin pathways.
Science 303:1483–1487.
Ornitz DM. 2005. FGF signaling in the developing endochondral skeleton. Cytokine Growth
Factor Rev 16:205–213.
Osyczka AM, Leboy PS. 2005. Bone morphogenetic protein regulation of early
osteoblast genes in human marrow stromal cells is mediated by extracellular signal-
regulated kinase and phosphatidylinositol 3-kinase signaling. Endocrinology
146:3428–3437.
PengXD,XuPZ,ChenML,Hahn-WindgassenA,SkeenJ,JacobsJ,SundararajanD,ChenWS,
Crawford SE, Coleman KG, Hay N. 2003. Dwarfism, impaired skin development, skeletal
muscle atrophy, delayed bone development, and impeded adipogenesis in mice lacking
Akt1 and Akt2. Genes Dev 17:1352–1365.
Raucci A, Laplantine E, Mansukhani A, Basilico C. 2004. Activation of the ERK1/2 and p38
mitogen-activated protein kinase pathways mediates fibroblast growth factor-induced
growth arrest of chondrocytes. J Biol Chem 279:1747–1756.
Tran H, Brunet A, Griffith EC, Greenberg ME. 2003. The many forks in FOXO’s road. Sci
STKE 2003:RE5.
van Noort M, Meeldijk J, van der Zee R, Destree O, Clevers H. 2002.Wnt signalingcontrols
the phosphorylation status of beta-catenin. J Biol Chem 277:17901–17905.
WangW,ChenJX,LiaoR,DengQ,ZhouJJ,HuangS,SunP.2002.Sequentialactivationofthe
MEK-extracellular signal-regulated kinase and MKK3/6-p38 mitogen-activated protein
kinase pathways mediates oncogenic ras-induced premature senescence. Mol Cell Biol
22:3389–3403.
Westendorf JJ, Kahler RA, Schroeder TM. 2004. Wnt signaling in osteoblasts and bone
diseases. Gene 341:19–39.
Yeh LC, Adamo ML, Olson MS, Lee JC. 1997. Osteogenic protein-1 and insulin-like growth
factor I synergistically stimulate rat osteoblastic cell differentiation and proliferation.
Endocrinology 138:4181–4190.
ZhangM,XuanS,BouxseinML,vonStechowD,AkenoN,FaugereMC,MallucheH,ZhaoG,
Rosen CJ, Efstratiadis A, Clemens TL. 2002. Osteoblast-specific knockout of the insulin-
like growth factor (IGF) receptor gene reveals an essential role of IGF signaling in bone
matrix mineralization. J Biol Chem 277:44005–44012.
JOURNAL OF CELLULAR PHYSIOLOGY
E R K A N D A K T I N F G F R E S P O N S E O F O S T E O B L A S T S
451