The Anabolic Response of Bone to Mechanical Load Involves the Wnt/b–catenin Pathway

22 Pages·2006·0.36 MB·English

JBC Papers in Press. Published on August 14, 2006 as Manuscript M602308200
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M602308200
WNT/$-CATENIN SIGNALING IS A NORMAL PHYSIOLOGICAL RESPONSE TO
MECHANICAL LOADING IN BONE
John A Robinson*♦, Moitreyee Chatterjee-Kishore**♦, Paul J Yaworsky**, Diane M Cullen#,
Weiguang Zhao*φ, Christine Li**, Yogendra Kharode*, Linda Sauter**Q, Philip Babij**^, Eugene
L Brown**, Andrew A Hill**, Mohammed P Akhter#, Mark L Johnson#), Robert R Recker#, Barry
S Komm* and Frederick J Bex*
*Women’s Health & Musculoskeletal Biology, Wyeth Research, 500 Arcola Road,
Collegeville, PA 19426, ** Biological Technologies, Wyeth Research, 87 CambridgePark
Drive, Cambridge, MA 02140, and #Creighton University, Osteoporosis Research Center,
601 North 30th Street, Omaha, NE 68131 φ Current address: Aventis, 1041 Rt. 202/206,
N303A, Bridgewater, NJ 08807 ^ Current address: Amgen, Inc, 1 Amgen Center Drive,
Thousand Oaks, CA 91320 Q Current address: Department of Microbiology, School of Medicine,
University of Washington, Seattle, WA 98195 ) Current address: Department of Oral Biology,
Kansas City Dental School, Kansas City, MO 64108
Running Title: Wnt Signaling Enhances Sensitivity of Bone to Mechanical Load
Address correspondence to: John A Robinson, Department Women’s Health & Musculoskeletal Biology,
Wyeth Research, 500 Arcola Road, RN3247, Collegeville, PA 19426. Tel. [phone]; Fax 484-865-
D
9395; E-mail: [email protected]. ow
n
♦ These authors contributed equally to this manuscript lo
a
d
A preliminary expression profiling MC3T3-E1 cells were treated with either ed
analysis of osteoblasts derived from tibia GSK3βi or Wnt3A to activate Wnt signaling, fro
m
explants of the high bone mass (HBM) LRP5 and then subjected to load, a synergistic up- h
ttp
G171V transgenic mice demonstrated increased regulation of these genes was observed ://w
expression of canonical Wnt pathway and compared to vehicle treated cells. Collectively, w
w
Wnt/β-catenin target genes compared to non- the in vivo and in vitro mechanical loading .jb
c
transgenic explant derived osteoblasts. results support that Wnt/$-catenin signaling is .o
rg
Therefore, expression of Wnt/β-catenin target a normal physiological response to load and b/
y
genes were monitored following in vivo loading that activation of the Wnt/$-catenin pathway gu
e
of the tibia of LRP5 G171V transgenic mice enhances the sensitivity of st o
n
compared to non-transgenic mice. Loading osteoblasts/osteocytes to mechanical loading. A
p
resulted in the increase expression of Wnt ril 8
pathway and Wnt/β-catenin target genes Over the past few years the Wnt/β-catenin , 2
0
including Wnt10B, SFRP1, cyclin D1, FzD2, signaling pathway has been shown to be an 19
WISP2 and connexin 43 in both genotypes, important component of bone mass accrual,
however, there was a further increase in regulation and maintenance. Key to this
transcriptional response with the LRP5 G171V understanding were the identification of
transgenic mice. Similar increases in the inactivating mutations in LRP5 resulting in an
expression of these genes (except cyclin D1) osteoporosis pseudoglioma syndrome (1) and gain
were observed when non-transgenic mice were of function mutations in the LRP5 gene that give
pharmacologically treated with a canonical rise to a high bone mass phenotype in humans (2-
Wnt pathway activator, Glycogen Synthase 4). Both of these conditions can be mimicked in
Kinase 3-β inhibitor (GSK3βi) and then mice (5,6).
subjected to load. These in vivo results were LRP5 and LRP6 have been shown to be
further corroborated by in vitro mechanical co-receptors (7-10) with the frizzled family of
loading experiments in which MC3T3-E1 receptors and are involved in signaling through the
osteoblastic cells were subjected to 3,400 canonical Wnt/β-catenin pathway (11,12). Upon
µstrain alone for 5 hrs which increased the binding of extracellular Wnt to LRP5/6 and
expression of Wnt10B, SFRP1, cyclin D1, FzD2, frizzled co-receptor complex, dishevelled (Dsh) is
WISP2 and connexin 43. Furthermore, when activated causing axin recruitment to the
Copyright 2006 by The American Society for Biochemistry and Molecular Biology, Inc.
membrane (where it interacts with LRP5/6) and increases bone formation in LRP5 -/- mice.
inactivation of glycogen synthase kinase-3$ Collectively, these data strongly suggest that
(GSK-3$) that results in stabilization and LRP5 and the Wnt/$-catenin signaling pathway
accumulation of $-catenin in the cytoplasm (13). play a critical role in the bone response to
Stabilized $-catenin translocates to the nucleus mechanical loading
and together with TCF/LEF transcription factors Although the molecular mechanisms by
activates transcription (12-14). Regulation of which mechanical loading affects BMD have not
GSK-3$ is a key step in the pathway and been fully elucidated, various in vitro and in vivo
inhibitors of this enzyme can lead to $-catenin models of mechanical loading have attributed it to
stabilization and initiation of target gene increased cell proliferation, activation of cell
expression, independent of Wnt binding. signaling and transcriptional activation of a
Signaling through this pathway is also regulated at number of genes. Rapid signaling responses have
the extracellular level by a number of modulator also been reported including changes in
protein (15-22) intracellular Ca+2, release of prostaglandins (PGE2
We have found that the increased peak and PGI2), nitric oxide release, increases in cAMP
bone mass that results in mice overexpressing the levels (30-35). The studies described in this report
HBM mutated gene product (LRP5 G171V) is due demonstrate that a number of target genes of the
primarily to osteoblast-induced positive bone Wnt/$-catenin signaling pathway are upregulated D
o
balance (5). The bone mineral density (BMD) in in response to mechanical loading and that the w
n
these animals plateaus and is maintained through LRP5 G171V mutation results in exaggerated loa
d
1.5 years of age without apparent abnormalities in increases in expression of several of these genes. ed
bone shape or size. Furthermore, the LRP5 We further show that activation of the pathway by from
G171V mutation in mice results in a skeleton that treatment with a GSK-3$ inhibitor results in an http
has enhanced structural strength (femur, femoral anabolic bone formation response and that use of ://w
neck, tibiae, and vertebral body), material this inhibitor in combination with mechanical ww
properties (vertebral body), and bone mass/ash loading produces a synergistic effect on the .jb
c
weight (ulnae) than their non-transgenic expression of Wnt/$-catenin pathway target genes. .org
littermates (23). These results strongly implicate the Wnt/$-catenin b/
y
The denser and stronger bones in LRP5 signaling pathway as being a critical component of gu
e
s
G171V transgenic mice have been suggested to be the bone response to mechanical loading. t o
n
due to greater sensitivity of bone to normal A
p
mechanical stimuli resulting in an altered response Experimental Procedures ril 8
to weight-related forces (23). This pattern is quite Reagents , 2
0
similar to that seen clinically in humans where Calcein, hematoxylin and eosin were 19
affected members have increased BMD and bones obtained from Sigma-Aldrich (Saint Louis, MO).
that are of normal shape and size. In the LRP5 Fetal bovine serum (FBS), heat inactivated FBS,
G171V high bone mass (HBM) kindred the most alpha Minimum Essential Medium ("MEM),
significant changes in bone density are associated Dulbecco’s Modified Eagle Medium (DMEM),
with the load bearing bones (24,25). Further, we penicillin, streptomycin, L-glutamine, glutamax
have shown that HBM transgenic mice (5) have and geneticin were purchased from Invitrogen
increased sensitivity to mechanical load (25,26) (Carlsbad, CA). Bovine serum albumin (BSA)
while Sawakami et al. (27) have shown in vivo that was purchased from Serologicals Proteins Inc.
bone formation response as a result of loading of (Kankakee, IL).
LRP5 -/- mice was nearly abolished compared to In Vivo Loading of Tibiae in LRP5 G171V
wildtype mice. In addition, Hens et al. (28) have Transgenic and Non-transgenic Littermates
shown the activation of the Wnt/$-catenin All animal protocols were conducted
signaling pathway in osteoblasts from TOPGAL with approval of the Wyeth and Creighton
mice containing a LacZ gene driven by a TCF University Institutional Animal Care and Use
promoter. Finally, Clement-Lacroix et al. (29)
Committees. The heterozygous LRP5 G171V
have shown that the GSK-3$ inhibitor LiCl
transgenic mice have been described and show
2
a statistically significant increase in bone RNA Isolation
density (5). Non-transgenic littermates were The mouse tibiae were dissected free of
used as controls. There were a total of 15 soft tissue, and the proximal and distal metaphysis
were removed leaving the diaphysis. The tibiae
animals/sex/genotype in each group. At 17
were then cut transversely with bone cutters to
weeks of age, all animals were anesthetized to
expose the trabecular bone and the bone marrow.
permit proper leg positioning prior to loading.
The trabecular bone and marrow cavities were
Using a four-point bending device (23), the
flushed with ice cold sterile PBS to remove the
mechanical loading regimen (~ 2,500 :strain)
marrow and the clean bone was placed in liquid
delivered to the right tibiae (left tibiae served nitrogen. A Bessman tissue pulverizer (Fisher
as the non-loaded controls) was 6N for Scientific, Pittsburgh, PA) rinsed in 100% ethanol
females and 7N for males (36 cycles, 2Hz) and pre-cooled in liquid nitrogen, was used to
which ensured that all mice were experiencing reduce the tibiae to a powder. Total RNA (2
similar levels of maximal compressive and :g/tibia) was isolated from non-loaded and loaded
tensile strains during bending loads. RNA bones using the ToTALLY RNA kit (Ambion,
Austin TX) as per manufacturer’s instructions.
from the right tibiae was collected at 4 or 24
Ten :g of total RNA was DNased with 4 Units of
hr following application of load. Tibiae from
5 mice were pooled to comprise a single DNase I (Ambion) to remove any genomic DNA D
o
contamination. w
group. Three replicate groups for each n
To isolate RNA from MC3T3-E1 cells, lo
a
treatment/genotype were analyzed. d
the cultures were washed twice with 2ml each of ed
Cell Culture PBS and then the RNA was isolated using the fro
m
MC3T3-E1 osteoblastic cells, used in the QIAshredder and the RNeasy kit (Qiagen, h
in vitro mechanical loading experiments were ttp
cultured in "MEM supplemented with 10% heat VThaele nRcNiaA, CwAa)s atrse adteesdc rwibiethd 2b7y Uthnei tms aonf uDfaNctausree rI. ://ww
w
inactivated FBS, 1% glutamax and 1% (Qiagen) on the RNA isolation column provided in .jb
penicillin/streptomycin. Wnt3A conditioned the kit as described by the manufacturer. c.o
rg
media was obtained from an overexpressing Quantitative Real-Time RT-PCR (TaqMan®) b/
y
Wnt3A stable murine L-cell line (ATCC, A two-step TaqMan protocol was used. g
u
Manassas, VA) that was maintained in DMEM RNA was first converted to cDNA at 37 °C for 2hr est o
supplemented with 10% FBS, 1% L-glutamine and n
(High Capacity cDNA Archive kit, Applied A
0co.4n ditmiogn/emdl mGedeina,e tciceilnls. we reT soe edoebdt aiinnt o 1W00nmt3mA Biosystems). TaqMan PCR reactions were pril 8
performed on an ABI Prism 7700 DNA Sequence , 2
dishes and cultured for 4 days in growth media Detector (Applied Biosystems), using 20 ng 01
9
without geneticin, the media was removed and
cDNA per reaction. The conditions for TaqMan
sterile filtered and fresh media was added to the PCR were as follows: 2 min at 50 °C, 10 min at
plates and cultured for an additional 3 days. The 95 °C, then 40 cycles each of 15 sec at 95 °C and
media was then removed and sterile filtered and
1 min at 60 °C on MicroAmp Optical 96-well
combined with the initial batch of cultured media. plates, covered with MicroAmp Optical caps. Each
Control conditioned media was obtained in a plate contained triplicates of the test cDNA
similar fashion using the parental L-M(TK-) cell templates and no-template controls for each
line (ATCC, Manassas, VA). The Wnt3A
reaction mix. The expression for each mouse gene
conditioned media activated canonical Wnt
was normalized to murine GAPDH. A list of
signaling in MC3T3-E1 cells as determined using
TaqMan probe-primer pairs used in this study can
a TCF-luciferase reporter transiently transfected
be found in Table 1.
into these cells (10 :l of conditioned media Calvariae Treatment and Histochemistry and
showed a 10-fold induction of reporter activity Immunohistochemistry Analysis
compared to control media treated and untreated
GSK3βi [3-(3-chloro-4-
cells (data not shown).
hydroxyphenylamino)-4-(2-nitrophenyl)-
1Hpyrrole-2,5-dione)] (36) at 1mg/kg or vehicle
was injected subcutaneously (s.c.), daily for the
3
first 7 days over the right side of the calvaria in 4 TaqMan® on samples from the tibiae on selected
week-old male Swiss-Webster mice. For each load- and Wnt-response genes.
treatment group that contained 8 animals each, In Vitro Mechanical Loading
calcein (15mg/kg) was administered s.c. on day 1 MC3T3-E1 cells were plated at 80,000-
and again on day 8 for mineral apposition rate 100,000 cells per well in a type I collagen coated
(MAR) determination. On day 10, the calvariae Bioflex 6 well plate (Flexcell International Corp.,
were removed and the bones fixed in 70% ethanol McKeesport, PA) and then cultured for 3-4 days or
for 24 h. The anterior portion of the calvaria was until confluent. Twenty-four hours prior to
paraffin embedded while the posterior portion was loading, the cells were washed twice with 2 ml of
used for frozen sections. The paraffin sections basal "MEM prior to adding 2 ml fresh serum-free
were stained with hematoxylin and eosin for media containing "MEM, 0.25% BSA, glutamax
measurement of calvarial thickness. and Penicillin/streptomycin. Immediately prior to
To calculate mineral apposition rates, mechanical loading, the medium was removed and
linear measurements of single label surface (SLS), 1 ml of "MEM/BSA with or without GSK3βi,
double label surface (DLS) and bone surface (BS) Wnt3A conditioned media or control conditioned
were taken and the equation DLS + (1/2 SLS)/BS media was added to each well. The cells were
x 100 was used to calculate percent Mineralized subjected to mechanical distortion equivalent to
Surface / Bone Surface (MS/BS). Measurements 3,400 :strain (2 Hz, 7200 cycles/hr), for 5 hrs D
o
were made on unstained 6µm sections at 20X using a FX-3000 Flexercell® strain unit (Flexcell w
n
magnification. All measurements were made using International Corp). RNA was harvested loa
d
the Bioquant Image Analysis System (Bioquant, immediately or 24 hr post loading from both the ed
Nashville, TN). For immunohistochemical mechanical strained samples as well as the non- from
analysis of β-catenin, calvaria were decalcified in strained controls. http
Surgipath Decalcifier II (Surgipath, Richmond, IL) Statistical Analysis ://w
for 7-8 h, dehydrated in graded alcohol and The data are represented as the mean +/- w
w
sectioned. Non-phosphorylated β-Catenin was SD. For those data comparing non-loaded versus .jb
c
detected using a mouse monoclonal antibody loaded results, unpaired 2-tailed t-tests was .org
(Upstate Biotechnology, Lake Placid, NY). Signal performed. For those data were there are multiple b/
y
was detected using the avidin-linked AP system comparisons a 2-factor ANOVA analysis was gu
e
s
(Vector Laboratories, Burlingame, CA). performed to compare strain and GSK3 inhibitor t o
n
Endogenous alkaline phosphatase was detected or Wnt3A. A Tukey HSD multiple comparisons A
p
histochemically using the Vector Red Alkaline was then performed between each dose of the ril 8
Phosphatase Substrate Kit (Vector Laboratories) in GSK inhibitor or Wnt with strain versus the strain , 2
0
6µm frozen sections of the mouse parietal bone only treatment. 19
after fixation in 70% ethanol.
Effect of Systemic GSK3βi Administration on the RESULTS
In Vivo Response to Mechanical Load
17-week old female, wild type C57Bl6 Application of In Vivo Mechanical Load Induces
mice were injected with GSK3βi 50 mg/kg/BID or Transcription of Wnt/β-Catenin Pathway Genes
vehicle (control). The injections were It has been previously reported that the
administered subcutaneously, twice daily for 14 LRP5 G171V transgenic mice have increased bone
days. There were 15 animals per group. The right formation compared to non-transgenic mice (5)
tibiae were loaded at 6 N for 36 cycles at 2 Hz. and mechanical loading increases bone formation.
The left tibiae served as unloaded controls. The Given the fact that this mutation leads to altered
animals were sacrificed at 4 hours post-load and regulation of the Wnt/$-catenin pathway, we
the tibia was isolated and flash frozen in liquid tested the hypothesis that downstream targets of
nitrogen. Tibiae (five) were pooled from each the pathway would have exaggerated changes in
group to provide 3 replicates per group. The RNA expression in response to loading compared to
was purified from tibiae (loaded and unloaded). normal mice. We performed a preliminary
Transcriptional analyses were performed by analysis of gene expression profiles from tibial
4
osteoblast explant cultures from LRP5 G171V in the skeleton, a small molecule inhibitor of
versus non-transgenic mice and identified several GSK3β (GSK3βi) was used to stabilize and
differentially expressed genes including Wnt10B, promote nuclear accumulation of β-catenin and
secreted frizzled related protein 1 (SFRP1), activation of the Wnt canonical pathway (36).
SFRP2, Dickkopf-3 (Dkk3), cyclin D1 (CCND1) GSK3βi (1mg/kg/day) was administered to
and Wnt1 inducible signaling pathway protein 2 wildtype C57BL/6 mice via local subcutaneous
(WISP2) associated with the mutation (data not injection over the right side of the calvaria. Strong
shown). We next examined whether the
β-catenin expression was observed in pre-
expression of these Wnt/$-catenin target genes osteoblasts and osteoblastic cells lining the
were regulated by mechanical loading.
periosteum following 7 days of GSK3βi treatment
Transcriptional analysis of tibia samples from
indicating that the agent was active in this model
LRP5 G171V mice following in vivo mechanical
(Fig. 3). We also observed significant
loading by four-point bending resulted in expected
enhancement of several indices of osteoblast
increases in the mRNA expression of known stress
activity including an increase in staining for
responsive genes including prostaglandin synthase
alkaline phosphatase, in mineral apposition rate
(COX-2), prostacyclin synthase (Ptgis) and
(MAR) and in calvarial thickness (Fig. 4A, 4B).
endothelial nitric oxide synthase (eNOS) (37,38)
These observations suggest that activation of β-
(Fig. 1). The expression of these genes was
catenin signaling through inhibition of GSK3β D
o
increased 4 hours post-loading in wild type and w
produces a bone anabolic effect. n
LRP5 G171V transgenic mice tibiae, however, the lo
Increased Wnt/$-Catenin Pathway Gene ad
response was 4-10 fold greater in the transgenics ed
(Fig. 1). This response was attenuated 24 hours Expression in Response to Mechanical Loading in fro
Wild Type Mice Treated with GSK3βi m
post-loading (data not shown). h
Having demonstrated that targets of the ttp
In non-transgenic mice, transcription of Wnt/$-catenin signaling pathway are upregulated ://w
Wnt10b, SFRP1, CCND1, connexin 43 (Cxn43) w
by mechanical loading and that an inhibitor of w
and WISP2 genes were significantly increased (~ .jb
2-4 fold, p<.01) at both 4 hr and 24 hr following in GSK3β increased the expression of $-catenin in c.o
vivo mechanical loading (Fig. 2A, 2B). RNA osteoblasts, and promotes bone formation, we brg/
y
levels of frizzled 2 (FzD2) was increased 2.8 fold wanted to determine the role of Wnt/$-catenin g
u
(p<.01) at 4hr post-load but returned to baseline by pathway activation in the bone formation response est o
24 hr. Expression of LRP5 was not affected by to mechanical loading. In vivo loading n A
lmoaardg iinna ltlhyi s inmcordeaesl edw hiilne LmRaPle6 eLxRprPe5s sioGn1 7w1aVs esyxspteermimiceanltlsy wwietrhe GpSeKrf3oβrmi e(5d0 wmitgh/ kgm/dicaey , trseca tfeodr pril 8, 2
animals at 4 hr post-load (Fig. 2A, 2B). In LRP5 14 days) to activate the canonical Wnt pathway, 01
9
G171V transgenic mice, a more significant and and gene expression analysis was performed on
sustained increase (3-30 fold, p<.01) in the RNA harvested from the loaded and unloaded
transcription of the Wnt/β-catenin target genes tibiae. In these experiments, the expression of
analyzed was observed, including SFRP4, which representative Wnt pathway and Wnt/β-catenin
was not regulated by mechanical loading in wild target genes was measured, as well as the
type mice (Fig. 2A, 2B). Unlike the pattern expression of the stress responsive genes, COX-2
demonstrated in the non-transgenic mice, there and eNOS, which served as positive controls to
was an increased expression of FzD2 at both 4 and confirm that the tibiae were responding to the
24 hr post-load. Thus, application of mechanical loading regimen.
load increased expression of Wnt/β-catenin In these in vivo experiments, load
regulated genes and the response was significantly increased the expression of COX-2 and eNOS 10-
greater in the LRP5 G171V transgenic mice. fold and 3-fold, respectively (Fig. 5).
Furthermore, load significantly increased the
Activation of Wnt/β-catenin Signaling and Bone
expression of Wnt10B, SFRP1, Cxn43, CCND1,
formation by an Inhibitor of GSK3β
FzD2 and WISP2 ranging from 2-4 fold compared
To directly establish that β-catenin
to the non-loaded limb. When mice were treated
signaling is associated with an anabolic response
with the GSK3βi for 14 days and then subjected to
5
mechanical loading, there was a further increase (Fig 6). The GSK-3β inhibitor alone had no effect
(p<.01 or p<.001) in the expression of stress on CCND1, SFRP1 or c-jun expression. We next
responsive genes COX-2 and eNOS as well as tested whether a natural canonical Wnt pathway
Wnt10B, SFRP1, Cxn43 and FzD2 when ligand (Wnt3A) that acts through the LRP5
compared to the loaded tibiae RNA samples from receptor would have similar effects on the strain
the control mice . The maximal increase in gene induced gene expression as inhibition of GSK-3β.
expression for COX-2, eNOS, Wnt10B, SFRP1, Treatment of MC3T3-E1 cells with increasing
Cxn43 and FzD2 was 28, 5, 7, 3, 5 and 7 fold amounts of Wnt3A conditioned media in the
respectively (Fig. 5). Changes in CCND1 did not presence of strain resulted in a biphasic, dose
reach statistical significance in response to load in dependent, synergistic increase in COX-2, c-fos,
mice treated with GSK3βi. Furthermore, the c-jun Wnt10B, CCND1, Cxn43, WISP2 and
GSK3β inhibitor had no effect alone on the SFRP1 expression compared to mechanical strain
expression of the Wnt/β-catenin responsive genes (Fig. 7). The magnitude of these responses was
because the transcriptional responses are likely 1.8-2.6 fold (p<.05) above strain alone. Wnt alone
dependent on dose, which may not have been had no effect on the expression of these genes
sufficient to elucidate a response alone. (Fig. 7).
Activation of the Wnt/$-Catenin Pathway
Enhances the Sensitivity of Bone Cells to DISCUSSION D
o
Mechanical Loading wn
To further investigate these load induced The skeletal phenotype of individuals loa
d
gexenpee rirmesepnotns seuss iinng at hcee llFulleaxre mrcoedlle li,n wveit rpoe rlfooardminedg duinsipqluaey ingam tohneg LRgePn5e tGic 17d1iVso rmdeurtsa tiolena diisn gq uittoe ed from
system (39). MC3T3-E1 cells were subjected to increased bone density since the bones are http
3,400 :strain for 5 hours resulting in the increased predominately of normal shape and size. There ://w
expression of COX-2 (2.5 fold), eNOS (2.5 fold), have been individuals in one of the kindreds with ww
as well as c-fos (3.5 fold) and c-jun (3.5 fold) this mutation presenting with a mandibular .jb
c
(p<.05) indicating that the cells were responsive to phenotype (2), torus palatinus, or more extreme .org
the applied strain (Fig. 6). Consistent with our in phenotypes (40), but for most affected individuals by/
vivo loading studies we also observed increased the increased density is benign. What is gu
e
expression of Wnt10B, CCND1, Cxn43, SFRP1, particularly striking about the skeletal phenotype st o
n
FzD2 (3-3.5-fold, p<.05)) and WISP2 (which was is the fact that the increase in bone density appears A
p
minimal but statistically significant p<.05) in to be most evident in load-bearing bones, which ril 8
strain versus the control cultures. has led to the hypothesis that the genetic mutation , 2
0
We next tested whether mechanical strain is acting to augment the skeleton’s response to 19
would synergize with independent activation of mechanical loading (24,25).
Once achieved, this high bone density
the Wnt/$-catenin pathway by treatment of the
phenotype is maintained as the normal new
cells with either the GSK-3β inhibitor or addition
homeostatic level that is supported by the
of exogenous Wnt3A. When cells were subjected
relatively normal index of mineral metabolism and
to mechanical strain in the presence of 5 :M
bone turnover markers (2). The skeletal pattern
GSK3βi, a synergistic transcriptional response of
seen in affected humans is recapitulated in the
COX-2, eNOS, c-fos, c-jun as well as Wnt10B,
LRP5 G171V transgenic mice where bone density
CCND1, Cxn43, SFRP1, WISP2 and FzD2 was
levels reach a dramatically elevated plateau that
observed (Fig. 6). The maximum increase in
appears to be maintained throughout life. Also,
expression for these genes with GSK3βi was 2-3
the increased trabecular BMD and trabecular
fold (p<.05) above mechanical strain alone. FzD2
number as well as the slight increase in cortical
and WISP2 demonstrated a 2-fold (p<.05) increase
thickness of LRP5 G171V mice compared to non-
in gene expression in the presence of the GSK-3β
transgenic mice is directly associated with
inhibitor alone, and although there was little effect
enhanced structural and mechanical properties in
on COX-2, Cxn43, Wnt10B, c-fos and eNOS
the femurs and vertebra (5). We have previously
expression, it was statistically significant (p<.05)
shown applying physiological levels of strain
6
(>5,000 :strain is considered pathological (31,39)) expression of Wnt pathway target genes in
using an established loading paradigm, that less response to mechanical loading, they differ in
mechanical strain is required to stimulate similar some respects. For instance, we did not observe a
bone responses in LRP5 G171V transgenic mice significant increase in LRP5 gene expression in
than in non-transgenic littermates (25,26). our in vivo loading bones, whereas they reported
Furthermore, we have previously demonstrated large increases in isolated ostoblasts from the two
that the LRP5 G171V transgenic mice experience mouse lines. Several possible explanations may
significantly less bone loss than non-transgenic account for this particular difference. First Lau et
mice in a disuse bone un-loading model (41). al used higher loads in their in vivo loading
Collectively, these observations suggest that the compared to what we applied (9N versus 6N).
LRP5 G171V mutation alters the response of the Also, they used younger mice (8 weeks versus 17
skeleton to normal mechanical load. weeks of age) and hence the skeletons of the mice
It is interesting to note that we observed in their study were still growing. Finally, it is
no difference in the expression of the osteoblast possible that this difference reflects differences in
differentiation transcription factor RUNX2 in the the C57BL/6 mice from different vendors.
osteoblasts of the LRP5 G171V transgenic Regardless, our results are in general agreement
compared to non-transgenic mice (data not with those of Lau et al. in that mechanical loading
shown). This suggests that like the low bone mass does result in the upregulation of the Wnt/β- D
o
phenotype in the LRP5 null mice (6), the LRP5 catenin signaling pathway. Thus, it appears that w
n
G171V mutation induced increase in anabolic the regulation of individual components of the lo
a
d
activities in bone are also RUNX2 independent. Wnt/β-catenin signaling pathway may vary ed
In fact, our preliminary findings from osteoblasts depending upon the model and as such, it will be fro
m
of LRP5 G171V transgenic animals, showing important to sort out both similarities and h
ttp
increased expression of Wnt/β-catenin pathway differences in terms of understanding the ://w
components and target genes, implicates activated mechanism leading to this response. This is w
w
canonical Wnt signaling as a key pathway particularly true when attempting to interpret data .jb
c
involved in the high bone mass phenotype. from cell culture studies using cell lines in .org
Our data further confirm a growing body comparison with in vivo results where the b/
y
of evidence that the upregulation of the Wnt/β- interactions of multiple cell types are involved in gu
e
catenin signaling pathway is required for bone the response to mechanical loading. st o
n
formation in response to mechanical loading. We There is accumulating evidence from the A
p
have previously shown that the LRP5 G171V mice cancer literature that crosstalk between various ril 8
are more sensitive to mechanical loading (23). signaling pathways such as the prostaglandin (44), , 2
0
Sawakami et al. (27) demonstrated that LRP5-/- BMP’s (45) and the Pten tumor suppressor (46) 19
mice fail to form bone in response to in vivo may regulate β-catenin through inhibition of
mechanical loading. Clement-LaCroix et al. GSK-3β. If similar mechanisms function in bone
showed that LiCL, which inhibits GSK-3 is bone then interpreting the cellular response to
anabolic in LRP5-/- mice (29). Hens et al. (28) mechanical loading requires an integration of
demonstrated that the Wnt/β-catenin signaling multiple signals and understanding the temporal
pathway is activated by in vitro mechanical sequence of events that result.
loading of primary osteoblasts isolated from the In light of our studies and those of other
TOPGAL mouse. Also, Lau et al. (42) recently investigators, how does Wnt/β-catenin signaling
demonstrated that osteoblasts isolated from fit into the mechanism by which bone responds to
C57BL/6 mice activate the expression of Wnt mechanical loading? We have shown that known
pathway genes more robustly in response to stress response genes such as COX-2, eNOS, c-fos
mechanical loading compared to C3H/HeJ mice, and c-jun are upregulated by mechanical loading
which have been shown to respond less to loading consistent with the work of others. PGE and NO
2
(43) release are some of the earliest changes measured
Although our results are similar to those in response to mechanical loading of bone. We
reported by Lau et al. in that we find increased believe that these events are upstream of the
7
activation of the Wnt/β-catenin signaling pathway. GSK-3$, then sustained activation requires the
Recent work suggested that PGE can lead to a expression of genes whose protein products (such
2
Wnt independent activation of the canonical as Wnts, Frizzles, etc.) will amplify the initial load
β−catenin signaling pathway by inhibition of activation and commit the cells to new bone
GSK-3β and binding of G"s to Axin (44). formation. This could explain why LRP5-/- mice
Furthermore, Sawahami et al. showed that PGE fail to form bone in response to loading because
2
release from LRP5-/- osteoblasts by fluid sheer the feedback amplification cannot occur when
stress was not effected by loss of LRP5 expression LRP5 is absent.
(27). This may suggest that cross talk between Collectively, the in vivo and in vitro
PGE2 signaling and Wnt/β-catenin signaling could mechanical loading results suggest that Wnt/β-
occur in response to mechanical loading catenin signaling is a normal physiological
independent of LRP5. So we directly tested the response to load and that activation of the Wnt/β-
effects of activating Wnt/β-catenin signaling on in catenin pathway enhances the sensitivity of bone
vivo bone formation using GSK3βi and found a cells to mechanical loading. The HBM phenotype
bone anabolic response that was equivalent to that shows an enhanced anabolic response to loading
generated by exogenous administration of PTH on bone that may be due to an increase in
(data not shown) osteoblast differentiation and function and/or
Having shown a bone anabolic effect with reduced inhibition of LRP5 (2-4) enabling a longer D
o
sustained activation of the pathway in response to w
GSK3βi, studies were designed to determine n
whether the effects of mechanical loading on gene load. In fact, we find genes important for such load
e
functions including c-fos, c-jun (AP-1), eNOS, d
ewxopurleds smioinm iinc twhailtd i nty LpRe Pm5i cGe1 t7re1aVte tdr awnsitghe nGicS Km3icβei COX2 (involved in PGE2 production) and Cxn43 from
where the mutation is a canonical Wnt pathway (47-52) all up-regulated to a greater extent with http
activating mutation. Interestingly, we observed mechanical loading when the canonical Wnt ://w
pathway was activated. Furthermore, it is also w
similar increases in the expression of the Wnt/β- w
possible that the cooperative actions of mechanical .jb
ctraatnesnginen tiacr gmeitc eg esnuebsj ecinte dth teo GloSaKd 3cβoim tpreaaretedd t on othne- lpoaathdwinagy andar ea ctipvraotimoont inogf thoes tecoabnloanstiocgale neWsinst, bc.org/
LRP5 G171V transgenic mice. These data because an induction in expression of Wnt10B an y gu
strongly suggest that activation of β-catenin es
signaling rather than other non-canonical signaling inhibitor of adipogenesis (53,54) was also t on
observed. A
pathways enhance the effects of mechanical p
loading on bone. These observations are The fact that a combination of ril 8
pharmacologic treatment with a GSK-3 inhibitor , 2
supported by the mechanical loading experiments 0
and mechanical loading produce a synergistic 19
where MC3T3-E1 cells treated with increasing
effect has important clinical implications.
amounts of either GSK3$i (data not shown) or
Concerns over the safety of therapies designed to
Wnt3A (a canonical Wnt pathway ligand)
focus on regulating the Wnt signaling pathway has
conditioned media caused a dose dependent
been debated in the literature (11,55,56). If a
synergistic up-regulation of Wnt/$−catenin gene
combination of exercise (loading) and a otherwise
expression. Interestingly, high concentrations of
sub-efficacious dose of pharmaceutical could be
exogenous Wnt3A resulted in a down regulation
used to generate a bone anabolic response, this
of these genes. This biphasic transcriptional
novel approach might offer great hope for
response observed with Wnt3A (that was not seen
individuals afflicted with diseases of low bone
with GSK3$i) may have been due to negative
mass such as osteoporosis. Much more work in
feedback regulation at the level of the LRP5/FzD
this regard is needed, but our data suggest this as a
co-receptor that may be different with GSK3$i
real possibility.
which acts downstream of this co-receptor
In summary we have used both genetic
complex. Implicit in our model is a feedback
and pharmacological models to show the
regulation of the Wnt/$−catenin signaling involvement of the canonical Wnt pathway on
pathway. If initial activation of $−catenin results mechanical loading responses in the skeleton. The
from crosstalk with other pathways at the level of in vitro and in vivo data corroborate the conclusion
8
that mechanical loading regulates Wnt pathway signaling in adult bone mass regulation. It is
and Wnt/β-catenin target gene expression and that conceivable that mimicking the positive effects of
Wnt signaling enhances the sensitivity of bone the HBM LRP5 G171V mutation on the skeleton
cells to mechanical loading. These findings pharmacologically, may provide novel therapies
demonstrate a novel role for canonical Wnt for the treatment of osteoporosis.

REFERENCES
1. Gong, Y., Slee, R. B., Fukai, N., Rawadi, G., Roman-Roman, S., Reginato, A. M., Wang,
H., Cundy, T., Glorieux, F. H., Lev, D., Zacharin, M., Oexle, K., Marcelino, J., Suwairi,
W., Heeger, S., Sabatakos, G., Apte, S., Adkins, W. N., Allgrove, J., Arslan-Kirchner,
M., Batch, J. A., Beighton, P., Black, G. C., Boles, R. G., Boon, L. M., Borrone, C.,
Brunner, H. G., Carle, G. F., Dallapiccola, B., De Paepe, A., Floege, B., Halfhide, M. L.,
Hall, B., Hennekam, R. C., Hirose, T., Jans, A., Juppner, H., Kim, C. A., Keppler-
Noreuil, K., Kohlschuetter, A., LaCombe, D., Lambert, M., Lemyre, E., Letteboer, T.,
Peltonen, L., Ramesar, R. S., Romanengo, M., Somer, H., Steichen-Gersdorf, E.,
D
Steinmann, B., Sullivan, B., Superti-Furga, A., Swoboda, W., van den Boogaard, M. J., o
w
n
Van Hul, W., Vikkula, M., Votruba, M., Zabel, B., Garcia, T., Baron, R., Olsen, B. R., lo
a
d
and Warman, M. L. (2001) Cell 107, 513-523 ed
2. Boyden, L. M., Mao, J., Belsky, J., Mitzner, L., Farhi, A., Mitnick, M. A., Wu, D., fro
m
Insogna, K., and Lifton, R. P. (2002) N Engl J Med 346, 1513-1521 http
3. Little, R. D., Carulli, J. P., Del Mastro, R. G., Dupuis, J., Osborne, M., Folz, C., ://w
w
Manning, S. P., Swain, P. M., Zhao, S. C., Eustace, B., Lappe, M. M., Spitzer, L., w
Zweier, S., Braunschweiger, K., Benchekroun, Y., Hu, X., Adair, R., Chee, L., .jbc
.o
FitzGerald, M. G., Tulig, C., Caruso, A., Tzellas, N., Bawa, A., Franklin, B., McGuire, rg
b/
S., Nogues, X., Gong, G., Allen, K. M., Anisowicz, A., Morales, A. J., Lomedico, P. T., y g
u
Recker, S. M., Van Eerdewegh, P., Recker, R. R., and Johnson, M. L. (2002) Am J Hum es
t o
Genet 70, 11-19 n
A
4. KVeayn, WL.e, sRenebnetoenc,k T, L., .B, Carlteeilrse,n C, E., .G, Gonragm, Y, J.,. ,W Baeramlsa, nR, .M K.. ,L B.,e Dniec hVoeur,n Oej.o, uSlc, oMp.e lCli.t,i , D., pril 8, 2
0
Bollerslev, J., and Van Hul, W. (2003) Am J Hum Genet 72, 763-771 1
9
5. Babij, P., Zhao, W., Small, C., Kharode, Y., Yaworsky, P. J., Bouxsein, M. L., Reddy, P.
S., Bodine, P. V., Robinson, J. A., Bhat, B., Marzolf, J., Moran, R. A., and Bex, F. (2003)
J Bone Miner Res 18, 960-974
6. Kato, M., Patel, M. S., Levasseur, R., Lobov, I., Chang, B. H., Glass, D. A., 2nd,
Hartmann, C., Li, L., Hwang, T. H., Brayton, C. F., Lang, R. A., Karsenty, G., and Chan,
L. (2002) J Cell Biol 157, 303-314
7. Wehrli, M., Dougan, S. T., Caldwell, K., O'Keefe, L., Schwartz, S., Vaizel-Ohayon, D.,
Schejter, E., Tomlinson, A., and DiNardo, S. (2000) Nature 407, 527-530
8. Mao, J., Wang, J., Liu, B., Pan, W., Farr, G. H., 3rd, Flynn, C., Yuan, H., Takada, S.,
Kimelman, D., Li, L., and Wu, D. (2001) Mol Cell 7, 801-809
9. Tamai, K., Semenov, M., Kato, Y., Spokony, R., Liu, C., Katsuyama, Y., Hess, F., Saint-
Jeannet, J. P., and He, X. (2000) Nature 407, 530-535
10. Pinson, K. I., Brennan, J., Monkley, S., Avery, B. J., and Skarnes, W. C. (2000) Nature
407, 535-538
9
11. Moon, R. T., Bowerman, B., Boutros, M., and Perrimon, N. (2002) Science 296, 1644-
1646
12. Wodarz, A., and Nusse, R. (1998) Annu Rev Cell Dev Biol 14, 59-88
13. Huelsken, J., and Birchmeier, W. (2001) Curr Opin Genet Dev 11, 547-553
14. Willert, J., Epping, M., Pollack, J. R., Brown, P. O., and Nusse, R. (2002) BMC Dev Biol
2, 8
15. Mao, B., and Niehrs, C. (2003) Gene 302, 179-183
16. Mao, B., Wu, W., Davidson, G., Marhold, J., Li, M., Mechler, B. M., Delius, H., Hoppe,
D., Stannek, P., Walter, C., Glinka, A., and Niehrs, C. (2002) Nature 417, 664-667
17. Jones, S. E., and Jomary, C. (2002) Bioessays 24, 811-820
18. Mao, B., Wu, W., Li, Y., Hoppe, D., Stannek, P., Glinka, A., and Niehrs, C. (2001)
Nature 411, 321-325
19. Bafico, A., Liu, G., Yaniv, A., Gazit, A., and Aaronson, S. A. (2001) Nat Cell Biol 3,
683-686
20. Semenov, M. V., Tamai, K., Brott, B. K., Kuhl, M., Sokol, S., and He, X. (2001) Curr
Biol 11, 951-961
21. Itasaki, N., Jones, C. M., Mercurio, S., Rowe, A., Domingos, P. M., Smith, J. C., and D
o
w
Krumlauf, R. (2003) Development 130, 4295-4305 n
lo
a
22. Li, X., Zhang, Y., Kang, H., Liu, W., Liu, P., Zhang, J., Harris, S. E., and Wu, D. (2005) d
e
d
J Biol Chem 280, 19883-19887 fro
m
23. Akhter, M. P., Wells, D. J., Short, S. J., Cullen, D. M., Johnson, M. L., Haynatzki, G. R., h
Babij, P., Allen, K. M., Yaworsky, P. J., Bex, F., and Recker, R. R. (2004) Bone 35, 162- ttp://w
169 w
w
24. Johnson, M. L. (2004) J Musculoskelet Neuronal Interact 4, 135-138 .jb
c
25. Johnson, M. L., Picconi, J. L., and Recker, R. R. (2002) Endocrinologist 12, 445-453 .org
26. Cullen, D. M., Akhter, M. P., Mace, D., Johnson, M. L., Babij, P., and Recker, R. R. b/
y
g
(2002) Journal of Bone and Mineral Research 17, S332 u
e
s
27. Sawakami, K., Robling, A. G., Ai, M., Pitner, N. D., Liu, D., Warden, S. J., Li, J., Maye, t o
n
A
P., Rowe, D. W., Duncan, R. L., Warman, M. L., and Turner, C. H. (2006) Journal p
Biological Chemistry doi:10.1074/jbc.M601000200 ril 8
, 2
28. Hens, J. R., Wilson, K. M., Dann, P., Chen, X., Horowitz, M. C., and Wysolmerski, J. J. 0
1
9
(2005) J Bone Miner Res 20, 1103-1113
29. Clement-Lacroix, P., Ai, M., Morvan, F., Roman-Roman, S., Vayssiere, B., Belleville,
C., Estrera, K., Warman, M. L., Baron, R., and Rawadi, G. (2005) Proc Natl Acad Sci U
S A 102, 17406-17411
30. Pavalko, F. M., Gerard, R. L., Ponik, S. M., Gallagher, P. J., Jin, Y., and Norvell, S. M.
(2003) J Cell Physiol 194, 194-205
31. Hughes-Fulford, M. (2004) Sci STKE 2004, RE12
32. Danciu, T. E., Adam, R. M., Naruse, K., Freeman, M. R., and Hauschka, P. V. (2003)
FEBS Lett 536, 193-197
33. Nomura, S., and Takano-Yamamoto, T. (2000) Matrix Biol 19, 91-96
34. Pommerenke, H., Schmidt, C., Durr, F., Nebe, B., Luthen, F., Muller, P., and Rychly, J.
(2002) J Bone Miner Res 17, 603-611
35. Ziros, P. G., Gil, A. P., Georgakopoulos, T., Habeos, I., Kletsas, D., Basdra, E. K., and
Papavassiliou, A. G. (2002) J Biol Chem 277, 23934-23941
10