Page 1
University of Southern Denmark
PDGF Receptor Signaling in Osteoblast Lineage Cells Controls Bone Resorption ThroughUpregulation of Csf1 Expression
Brun, Julia; Andreasen, Christina Møller; Ejersted, Charlotte; Andersen, Thomas Levin;Caverzasio, Joseph; Thouverey, Cyril
Published in:Journal of Bone and Mineral Research
DOI:10.1002/jbmr.4150
Publication date:2020
Document version:Accepted manuscript
Citation for pulished version (APA):Brun, J., Andreasen, C. M., Ejersted, C., Andersen, T. L., Caverzasio, J., & Thouverey, C. (2020). PDGFReceptor Signaling in Osteoblast Lineage Cells Controls Bone Resorption Through Upregulation of Csf1Expression. Journal of Bone and Mineral Research, 35(12), 2458-2469. https://doi.org/10.1002/jbmr.4150
Go to publication entry in University of Southern Denmark's Research Portal
Terms of useThis work is brought to you by the University of Southern Denmark.Unless otherwise specified it has been shared according to the terms for self-archiving.If no other license is stated, these terms apply:
• You may download this work for personal use only. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying this open access versionIf you believe that this document breaches copyright please contact us providing details and we will investigate your claim.Please direct all enquiries to [email protected]
Download date: 10. Dec. 2021
Page 2
Andreasen Christina (Orcid ID: 0000-0002-2624-5677) Andersen Thomas (Orcid ID: 0000-0002-6981-7276) Thouverey Cyril (Orcid ID: 0000-0002-2741-1825) PDGF receptor signaling in osteoblast lineage cells controls bone resorption through up-
regulation of Csf1 expression
Julia Brun1, Christina Møller Andreasen2, Charlotte Ejersted3, Thomas Levin Andersen2,
Joseph Caverzasio1, Cyril Thouverey1
1 Service of Bone Diseases, Department of Medicine, University Hospital of Geneva, 1205
Geneva, Switzerland
2 Clinical Cell Biology, Pathology Research Unit, Odense University Hospital, DK-5000
Odense C, Denmark; Dept. of Molecular Medicine, University of Southern Denmark, DK-5000
Odense C, Denmark; Dept. of Clinical Research, University of Southern Denmark, DK-5000
Odense C, Denmark.
3 Dept. of Endocrinology, Odense University Hospital, DK-5000 Odense C, Denmark
Running title: PDGFRs in OB lineage controls bone resorption
Corresponding author: Cyril Thouverey - Service of Bone Diseases, Department of Medicine,
University Hospital of Geneva, 64 Avenue de la Roseraie, 1205 Geneva, Switzerland - E-mail
address: [email protected] - Tel: +41 22 379 46 76
Disclosures: The authors state that they have no conflicts of interest. This work was supported
by the Swiss National Science Foundation (310030-166341), by the Fondation pour la
This article is protected by copyright. All rights reserved.
This is the author manuscript accepted for publication and has undergone full peer review buthas not been through the copyediting, typesetting, pagination and proofreading process, whichmay lead to differences between this version and the Version of Record. Please cite this articleas doi: 10.1002/jbmr.4150
Page 3
2
Recherche sur l’Ostéoporose et les Maladies Osseuses, by the Novartis Foundation for medical-
biological research (Project number 16C212 - Basel, Switzerland), by the VELUX
FOUNDATION (VELUX25723), and by the Region of Southern Denmark Research Fund
(18/17871).
Abstract
The physiological functions of platelet-derived growth factor receptors (PDGFRs) α and β in
osteoblast biology and bone metabolism remain to be established. Here, we show that PDGFRA
and PDGFRB genes are expressed by osteoblast-lineage canopy and reversal cells in close
proximity to PDGFB-expressing osteoclasts within human trabecular bone remodeling units.
We also report that, while removal of only one of the two PDGFRs in Osterix-positive cells
does not affect bone phenotype, suppression of both PDGFRs in those osteoblast lineage cells
increases trabecular bone volume in male mice as well as in female gonad-intact and
ovariectomized mice. Furthermore, osteoblast lineage-specific suppression of PDGFRs reduces
Csf1 expression, bone marrow level of M-CSF, number of osteoclasts and therefore, bone
resorption, but does not change bone formation. Finally, abrogation of PDGFR signaling in
osteoblasts blocks PDGF-induced ERK1/2-mediated Csf1 expression and M-CSF secretion in
osteoblast cultures, and calcitriol-mediated osteoclastogenesis in co-cultures. In conclusion, our
results indicate that PDGFR signaling in osteoblast lineage cells controls bone resorption
through ERK1/2-mediated Csf1 expression.
Key words: osteoblast lineage cells, PDGF receptors, ERK1/2, bone resorption, M-CSF
This article is protected by copyright. All rights reserved.
Page 4
3
Introduction
Bone development, repair and maintenance are regulated by a myriad of locally secreted
factors. Among them, platelet-derived growth factors (PDGFs) have emerged as important
regulators of multiple aspects of bone physiology (1). PDGF ligands consist in four distinct
polypeptide chains (PDGF-A, -B, -C and –D) that assemble in homodimers (PDGF-AA, -BB,
-CC and -DD) or heterodimers (PDGF-AB) (2,3). PDGFs act via two receptor tyrosine kinases,
PDGFRα and PDGFRβ. Ligand binding promotes receptor dimerization, autophosphorylation
and initiation of signaling (2,3). Depending on distinct ligand affinities and patterns of receptor
expression, different receptor dimers can form. Indeed, PDGFRαα is activated by PDGF-AA, -
AB, -BB and -CC, PDGFRαβ is activated by PDGF-AB, -BB and -CC, and PDGFRββ is
activated by PDGF-BB and -DD (2,3). PDGFRs play essential roles in embryonic development
as well as in different types of fibrosis and malignancies since it regulates proliferation,
chemotaxis and survival of cells of mesenchymal origin (3). In general, both PDGFRs exert
overlapping and distinct functions (3).
Numerous cells such as mesenchymal stem cells, osteoblasts, pre-osteoclasts, osteoclasts and
endothelial cells can produce PDGFs in the bone microenvironment (1). PDGF-AA, -AB and -
BB are most abundant PDGF members in bone and are mainly produced by osteoclast-lineage
cells in physiological condition (4-6). PDGFRs are expressed by perivascular mesenchymal
stem cells (MSCs) in metaphysis, endosteum and periosteum, and by
osteoprogenitors/osteoblasts located on trabecular, endosteal and periosteal surfaces (7-10).
Mechanistically, osteoclast-lineage cells couple angiogenesis to bone formation by releasing
PDGF-BB (6,9). In turn, elevated peri-osseous PDGF-BB concentration releases perivascular
This article is protected by copyright. All rights reserved.
Page 5
4
MSCs from vessel wall and enhances their proliferation and recruitment to bone surfaces, where
they are going to provide bone-forming osteoblasts (6,9,10).
PDGF-BB/PDGFRβ is a potent mitogenic and chemoattractant signal for skeletal MSCs and
committed osteoprogenitors in vitro (5,10,11). Accordingly, conditional suppression of
PDGFRβ in nestin-positive or leptin receptor-positive MSCs impairs periosteal and trabecular
bone formation due to defective self-renewal and migratory capacities (9). Conditional ablation
of PDGFRβ in osterix-positive cells does not affect bone formation in normal condition, but
decreases osteoprogenitor proliferation, trafficking and angiotropism during fracture healing
(10). On the other hand, PDGF-BB/PDGFRβ strongly inhibits osteoblast differentiation in vitro
(4,10-12). Although PDGFRα probably exerts redundant functions with PDGFRβ in bone
formation, it may play distinct roles since Pdgfra-deficient mice exhibit defective skeletal
patterning and maturation (13).
A growing body of evidence suggests that PDGF-BB/PDGFRβ also regulates
osteoclastogenesis and bone resorption. For instance, targeted Pdgfb overexpression in
endosteum and metaphyseal bone marrow stimulates bone resorption by elevating expressions
of Rankl (encoding receptor activator of nuclear factor κB ligand, RANKL) and Csf1 (encoding
macrophage colony-stimulating factor, M-CSF) (14). Moreover, PDGFRβ inhibition increases
expression of Opg (encoding osteoprotegerin, OPG, the RANKL decoy receptor) and reduces
that of Rankl by MSC and osteoblast cell lines in vitro (15). Finally, aberrant PDGF-
BB/PDGFRβ signaling due to absence of low-density lipoprotein receptor-related protein
(LRP)-1 in osteoblasts stimulates RANKL production and bone resorption (16).
This article is protected by copyright. All rights reserved.
Page 6
5
Osteoprogenitors and osteoblasts express PDGFRα and β. Although PDGF-BB/PDGFRβ
signaling stimulates their proliferation and migration, and inhibits their maturation in vitro, its
predominant in vivo function remains to be clarified. Suppression of PDGFRβ in Osterix-
positive cells does not affect osteoprogenitors or osteoblasts in physiological condition (10),
suggesting that PDGFRα plays overlapping functions with PDGFRβ. To investigate
physiological functions of PDGFRs in osteoblast biology and bone remodeling, we localized
expressions of PDGFRA and PDGFRB in human bone tissue by in situ hybridization, and we
analyzed mice in which PDGFRα, PDGFRβ or both receptors were deleted from osteoblast
lineage cells under the control of an inducible Osterix promoter.
Materials and Methods
Human bone specimens
3-mm needle biopsy specimens were obtained from the posterior iliac crest of four volunteers
without known musculoskeletal disease (two men, aged 27 and 41 years and two women, aged
33 and 49 years) using an 8G needle under local anesthesia. The biopsies were formalin-fixed,
decalcified and paraffin-embedded before serial sectioning. All volunteers gave their written
informed consent. The study was conducted according to the World Medical Association
Declaration of Helsinki and approved by the Danish National Committee on Biomedical
Research Ethics, journal no. S-20110112.
In situ hybridization
This article is protected by copyright. All rights reserved.
Page 7
6
3.5-mm-thick serial paraffin sections were in situ hybridized using an enhanced version of
RNAScope 2.5 high definition procedure (310035, ACD Bioscience) followed by tartrate-
resistant acid phosphatase (TRAP) immunostaining, as previously described (17). Sections
were rehydrated, deparaffinized, pretreated with pepsin digestion and heat treatment, and
hybridized overnight at 40°C with 20-ZZ-pair probes for human PDGFB mRNA (406701, ACD
Bioscience), PDGFRA mRNA (604481, ACD Bioscience) or PDGFRB mRNA (548991, ACD
Bioscience) diluted 1:1 in probe diluent (449819, ACD Bioscience) or only with probe diluent
(negative controls). The branch amplification was conducted according to the instructions
provided by the manufacturer, but enhanced with digoxigenin (DIG)-conjugated tyramide
(NEL748001KT, PerkinElmer) followed by alkaline phosphatase-conjugated sheep anti-DIG
FAB fragments (11093274910, Roche) and visualization with Liquid Permanent Red (Dako,
Denmark). The stained sections were then immunostained with mouse anti-TRAP antibodies
(clone 9C5, MABF96, Merck Millipore), which were labelled with horseradish peroxidase-
conjugated anti-mouse IgG polymers (BrightVision, Immunologic) and visualized with Deep
Space Black (Biocare Medical) and counterstained with Mayer’s hematoxylin.
Mice
Mice in which the expression of a tetracycline-Off-controllable GFP::Cre fusion protein is
transcriptionally regulated by an Osterix promoter (Osx-GFP::Cre, hereafter Osx-Cre) were on
a C57BL/6J genetic background and obtained from the Jackson Laboratory (#006361) (18).
Mice harboring floxed Pdgfra (#006492) (19) and floxed Pdgfrb (#010977) (20) were also
purchased from the Jackson Laboratory and bred to C57BL/6J mice for 3 and 6 generations,
This article is protected by copyright. All rights reserved.
Page 8
7
respectively. Then, those mouse strains were used to generate Osx-Cre;Pdgfraf/f, Osx-
Cre;Pdgfrbf/f and Osx-Cre;Pdgfraf/f;Pdgfrbf/f (hereafter Pdgfra cKO, Pdgfrb cKO and Pdgfra
cKO;Pdgfrb cKO) mice. Osx-Cre mice were used as control animals. In a first experiment,
since Osx-Cre mice were shown to exhibit a slight growth delay by 1 month of age (21), Cre
expression and consequent Pdgfra or/and Pdgfrb inactivation was induced at 6 weeks of age by
stopping doxycycline treatment, and male mice were sacrificed at 18 weeks of age. In a second
experiment, 12-week-old Osx-Cre and Pdgfra cKO;Pdgfrb cKO female mice were randomly
subjected to sham-operation (Sham) or Ovx, and sacrificed 8 weeks after surgeries. Pdgfra and
Pdgfrb inactivations were induced one week prior to surgeries. Mice (3 to 6 animals per cage)
were maintained under standard non-barrier conditions, exposed to a 12-hour light/12-hour dark
cycle and had access to mouse diet RM3 containing 1.24 % calcium and 0.56 % available
phosphorus (SDS, Betchworth, UK) and water ad libitum. Experimental units were single
animals. Investigators were blinded during endpoint measurements. All performed experiments
were in compliance with the guiding principles of the Guide for the Care and Use of Laboratory
Animals (8th edition) and approved by the Ethical Committee of the University of Geneva
School of Medicine and the State of Geneva Veterinarian Office.
Bone phenotyping
Mice were sacrificed and their bones were excised for micro-computed tomography (µCT)
analyses. Trabecular bone microarchitecture of proximal tibiae and fifth lumbar vertebral
bodies (100 slices from the beginning of secondary spongiosa), and cortical bone geometry of
This article is protected by copyright. All rights reserved.
Page 9
8
tibial midshafts (50 slices) were assessed using µCT (Viva-CT40, Scanco Medical,
Switzerland) employing a 12-μm isotropic voxel size.
To measure dynamic indices of bone formation, mice received subcutaneous injections of
calcein (10 mg/kg body weight; Sigma) at 9 and 2 days before euthanasia. Formalin-fixed
undecalcified femurs were embedded in methylmethacrylate (Merck). 8-µm transversal
sections of midshafts and 8-µm sagittal sections of distal femurs were cut and mounted
unstained for fluorescence visualization. Additional sagittal sections were stained with Goldner
trichrome for osteoblast counting or with tartrate-resistant acid phosphatase (TRAP) substrate
for osteoclast counting. Histomorphometric measurements were carried out using a Nikon
Eclipse microscope and the BioQuant software. The nomenclature for µCT and
histomorphometric analyses followed recommendations of published ASBMR guidelines
(22,23).
Immunohistochemistry
Femurs were fixed in 4% paraformaldehyde overnight, demineralized in 17% EDTA for 2
weeks, washed in demineralized water, and then incubated in 30% Sucrose for 48 hours at 4 °C
under agitation. Femurs were then frozen and embedded in NEG50 medium (Thermo Fischer
Scientific) and 10-µm frozen sections were cut using a Leica cryostat CM3050S (Leica
Biosystems). Cryosections were blocked with 3% bovine serum albumin at room temperature
for 30 minutes and then incubated at 4 °C overnight with primary antibodies: anti-PDGFRα
(#3174) and anti-PDGFRβ (#3169) (from Cell Signaling Technology), and anti-Endomucin
(Ab106100 from Abcam). Cryosections were washed, incubated with appropriate secondary
This article is protected by copyright. All rights reserved.
Page 10
9
antibodies: anti-rabbit (Ab150075) or anti-rat (Ab150155) conjugated with Alexa Fluor 647.
Samples were visualized using a Zeiss Axio Imager M2 microscope.
Osteoblast cultures
Primary osteoblast were isolated from long bones of Pdgfraf/f;Pdgfrbf/f mice as previously
described (24). Briefly, bone chips were prepared from cleaned long bones and digested in 1
mg/mL collagenase II (Sigma) for 90 minutes at 37 °C. Bone pieces were washed several times
and incubated in α-MEM (Amimed, Bioconcept) containing 10% FBS (Gibco) for 9 days to
allow cell migration from bone fragments. At that point, cells and bone chips were trypsinized
(with trypsin/EDTA from Sigma) and passaged at a split ratio of 1:3. At the second passage,
bone chips were removed. Medium was changed every 2-3 days. Osteoblasts at passages 3-4
were used for in vitro experiments. Pdgfraf/f;Pdgfrbf/f osteoblasts were infected with 400 moi
of empty or Cre-expressing adenoviruses (Vector Biolabs) to obtain Pdgfraf/f;Pdgfrbf/f (control)
and ΔPdgfra;ΔPdgfrb osteoblasts. Cell proliferation was assessed using a cell counter. Cell
migration was evaluated by wound scratch assays. Osteoblast differentiation was determined
by incubating confluent osteoblast cultures in osteogenic medium containing α-MEM, 10%
FBS, 0.05 mM L-ascorbate-2-phosphate (Sigma) and 10 mM β-glycerphosphate (AppliChem
GmbH) for 6 days for the measurement of alkaline phosphatase (ALP) activity and for 14 days
for the measurement of matrix mineralization by Alizarin Red-S staining (Sigma).
Co-cultures
This article is protected by copyright. All rights reserved.
Page 11
10
For co-culture experiments, primary Pdgfraf/f;Pdgfrbf/f osteoblasts infected with 400 moi of
empty or Cre-expressing adenoviruses were seeded at 30000 cells per well in 24-well plates.
The day after, non-adherent bone marrow cells isolated from wildtype mice were seeded over
osteoblasts at 300000 cells per well in α-MEM supplemented with 10% FBS and treated with
vehicle (Veh), 10 ng/mL PDGF-BB (Peprotech) or/and 10-8 M 1,25-dihydroxyvitamin D3
(calcitriol, Vit.D3). After 8 days, co-cultures were fixed and stained, and multinucleated TRAP-
positive cells were counted.
Western Blot
Pdgfraf/f;Pdgfrbf/f and ΔPdgfra;ΔPdgfrb osteoblasts were treated with Veh or 25 ng/mL PDGF-
BB at indicated times, rapidly frozen in liquid nitrogen and stored at -80 °C until their use for
analysis. Cell lysates were prepared by incubating cell cultures in RIPA buffer containing
phosphatase and protease inhibitors at 4 °C for 30 minutes. Lysates were then centrifuged at
6000g for 30 minutes. Lysate supernatants were diluted with equal volumes of 2-fold
concentrated reducing sample buffer. Those mixtures were then heated at 70 °C for 30 minutes
and subjected to gel electrophoresis on 6% to 15% gels. Proteins were electro-transferred to
Immobilon P membranes and immunoblotted with specific antibodies: anti-PDGFRα (#3174),
anti-PDGFRβ (#3169), anti-p-PDGFRs (#3170), anti-p-AKT (#4060), anti-p-SRC (Tyr416)
(#6943), anti-p-ERK1/2 (#9106), anti-ERK1/2 (#4695), anti-p-PLCγ2 (#3874) and anti- PLCγ2
(#3872) (all from Cell Signaling Technology). Detection was performed using peroxidase-
coupled secondary antibody, enhanced chemiluminescence reaction, and visualization by using
This article is protected by copyright. All rights reserved.
Page 12
11
G:Box gel analysis system (Syngene). Reprobed membranes were stripped according to the
manufacturer’s protocol.
Biochemistry
Serum levels of C-terminal telopeptide fragments of type I collagen (CTX), type I procollagen
N-terminal propeptide (PINP) and osteocalcin (OCN) were determined by using RatLaps
immunoassay kits (Immunodiagnostic Systems Ltd). RANKL, OPG and M-CSF protein levels
were determined in bone marrow supernatants (obtained by centrifugation at 1000 g for 10
minutes) or in cell culture medium using Quantikine ELISA kits (R&D Systems).
RNA isolation and real-time PCR
Total RNA was extracted from proximal tibial metaphyses (without bone marrow) or primary
osteoblast cultures using Tri Reagent® (Molecular Research Center) and purified using a
RNeasy Mini Kit (Qiagen). Single-stranded cDNA was synthesized from 2 µg of total RNA
using a High-Capacity cDNA Archive Kit (Applied Biosystems) according to the
manufacturer’s instructions. Real-time PCR was performed to measure the relative mRNA
levels using the QuantStudio™ 5 Real-Time PCR System with SYBR Green Master Mix
(Applied Biosystems). The mean mRNA levels were calculated from triplicate analyses of each
sample. Obtained mRNA level for a gene of interest was normalized to β2-microglobulin
mRNA level in the same sample.
Statistical analysis
This article is protected by copyright. All rights reserved.
Page 13
12
A sample size of 6 mice/group was required in order to detect a difference of 45% in trabecular
bone volume (SD=20%) between groups at the significance level of 0.01 (generally necessary
for Holm-Sidak tests) and a power of 80%. In vitro experiments were performed in triplicate
and independently repeated three times. Statistical analyses were performed using unpaired t-
test for the comparison of 2 groups, and one-way ANOVA followed by post hoc tests (Holm-
Sidak method) for the comparison of multiple genotypes. Interaction between genotype and
Ovx, or between genotype and in vitro treatments were analyzed by using two-way ANOVA.
Results
PDGFRA and PDGFRB are expressed by osteoprogenitors in close proximity to PDGFB-
expressing osteoclasts within human trabecular bone remodeling units
First, we localized expressions of genes encoding the components of the PDGF/PDGFR system
in human bone tissues. As previously shown, PDGFB was expressed by megakaryocytes and
endothelial cells in the bone marrow (Fig. 1A,B) (1). In addition, PDGFB was also expressed
by almost all osteoclasts at the surface of cancellous bone (Fig. 1D,E). More interestingly,
PDGFRA and PDGFRB were expressed by most of osteoprogenitor canopy cells that had been
shown to separate bone remodeling compartments from the bone marrow cavity (Fig. 1C,F,H)
(25). PDGFRA and PDGFRB were also expressed by most osteoblast lineage reversal cells on
the eroded surfaces in the vicinity of PDGFB-expressing TRAP-positive osteoclasts within
trabecular bone remodeling units (Fig. 1C,G,I) (26,27). PDGFRA and PDGFRB mRNA
expression was more abundant in reversal cells present on eroded surfaces than in bone lining
cells on quiescent surfaces (Fig. 1C,G,I).
This article is protected by copyright. All rights reserved.
Page 14
13
Postnatal osteoblast lineage-specific suppression of both PDGFRs increases trabecular
bone mass in male mice
Pdgfra and Pdgfrb mRNAs were highly expressed in primary osteoblasts that expressed Col1a1
(encoding type I collagen; early marker) and Alpl (encoding tissue non-specific alkaline
phosphatase; intermediate marker), but progressively declined as those cells differentiated into
mature osteoblasts expressing Bglap (encoding osteocalcin; late marker) (Fig. S1A).
Accordingly, PDGFRα and PDGFRβ proteins were abundant in early osteoblasts, but no longer
expressed in mature osteoblasts (Fig. S1B). To investigate physiological functions of PDGFRs
in osteoblast biology and bone metabolism, we generated murine models in which tetracycline-
Off-controllable Cre expression and consequent inactivation of floxed Pdgfra or/and Pdgfrb
genes in Osterix-positive cells were induced from 6 weeks to 18 weeks of age by stopping
doxycycline treatment. Conditional deletion of Pdgfra or Pdgfrb was confirmed by real-time
PCR in the osteoblast-rich fraction of tibiae and in primary osteoblasts isolated from Pdgfra
cKO and Pdgfrb cKO mice (Fig. S1C). Consistent with this, intense immunostainings of
PDGFRα and PDGFRβ on trabecular bone surfaces were absent in Pdgfra cKO;Pdgfrb cKO
mice (Fig. S1D). µCT analyses revealed that Pdgfra cKO and Pdgfrb cKO male mice did not
show any changes in trabecular or cortical bone mass and microarchitecture (Fig. 2A-D).
Although Pdgfra cKO;Pdgfrb cKO male mice did not display any change in cortical bone at
tibial midshaft (Fig. 2C,D), they exhibited significant increase in trabecular bone volume at the
proximal metaphysis of tibiae (+53%; Fig. 2A,B) and at the fifth lumbar vertebrae (+47%; Fig.
This article is protected by copyright. All rights reserved.
Page 15
14
S2). Increased trabecular bone mass in Pdgfra cKO;Pdgfrb cKO male mice was associated with
elevated number of trabeculae (Fig. 2B and Fig. S2).
Postnatal osteoblast lineage-specific suppression of both PDGFRs decreases bone
resorption in male mice
Quantitative histomorphometric analyses of femoral secondary spongiosa showed that,
although Pdgfra cKO and Pdgfrb cKO mice did not present changes in parameters of bone
resorption, Pdgfra cKO;Pdgfrb cKO mice demonstrated significant decreases in osteoclast
number and surfaces (Fig. 3A,B). Significant reductions in serum CTX levels and expression
of cathepsin K-encoding gene (CtsK) in bones of Pdgfra cKO;Pdgfrb cKO mice confirmed that
suppression of both PDGFRs in osteoblast lineage cells decreased bone resorption (Fig. S3).
Further ex vivo gene expression analyses indicated that expressions of pro-osteoclastogenic
genes Rankl and Csf1 were diminished in bones of Pdgfra cKO;Pdgfrb cKO mice whereas that
of Opg remained unchanged (Fig. 3C). However, only the protein level of M-CSF in the bone
marrow was significantly reduced in Pdgfra cKO;Pdgfrb cKO mice in comparison to that of
control mice (Fig. 3D).
Postnatal osteoblast lineage-specific suppression of both PDGFRs does not affect bone
formation in male mice
Dynamic histomorphometric analyses showed that neither Pdgfra cKO, Pdgfrb cKO nor Pdgfra
cKO;Pdgfrb cKO mice displayed modification in mineral apposition rate, mineralizing surfaces
or bone formation rate at trabecular, endosteal and periosteal bone envelopes in comparison to
control mice (Fig. 4A,B and Fig. S4). Furthermore, Pdgfra cKO, Pdgfrb cKO and Pdgfra
This article is protected by copyright. All rights reserved.
Page 16
15
cKO;Pdgfrb cKO mice did not demonstrate any changes in osteoblast number on trabecular
bone surfaces (Fig. 4C) or in serum levels of bone formation marker PINP (Fig. 4D). Moreover,
Pdgfra cKO;Pdgfrb cKO mice showed similar distribution of endomucin-positive blood vessels
in the metaphysis in comparison to control mice (Fig. 4E). Finally, Pdgfra cKO;Pdgfrb cKO
mice had comparable number of osteocytes per trabecular bone surfaces compared with that of
control mice (Fig. 4F), which is coherent with the absence of PDGFRs in osteocytes (Fig.
S1A,B).
Osteoblast lineage-specific suppression of both PDGFRs increases trabecular bone mass
in both gonad-intact and ovariectomized female mice
To test the effect of osteoblast lineage-specific abrogation of PDGFR signaling on ovariectomy-
induced bone loss, we took advantage of the inducible Cre expression in Osx-Cre mice to
suppress PDGFRα and PDGFRβ in osteoblast lineage cells at the time of sham-surgeries and
ovariectomies. Both Osx-Cre and Pdgfra cKO;Pdgfrb cKO female mice experienced similar
trabecular bone loss at proximal metaphysis of tibiae following Ovx (Fig. 5A), as a result of
increased osteoclast number (Fig. 5B). Interestingly, suppression of PDGFR signaling in
osteoblast lineage cells increased trabecular bone volume and trabecular number in both gonad-
intact and Ovx female mice (Fig. 5A). This increase in trabecular bone volume in Sham and
Ovx Pdgfra cKO;Pdgfrb cKO female mice was associated with reduced osteoclast number (Fig.
5B) and unchanged osteoblast number (Fig. 5C). Moreover, decreased bone resorption in Sham
and Ovx Pdgfra cKO;Pdgfrb cKO female mice could be explained by significant reduction in
bone marrow level of M-CSF (Fig. 5D).
This article is protected by copyright. All rights reserved.
Page 17
16
Suppression of PDGFR signaling in osteoblasts inhibits M-CSF production and M-CSF-
mediated osteoclastogenesis in vitro
To confirm and expand in vivo observations, we analyzed the molecular mechanisms by which
PDGFR signaling in osteoblasts could control osteoclastogenesis in vitro. PDGF-AA treatment
did not alter expressions of Csf1, Rankl or Opg in osteoblast cultures (Fig. 6A and Fig. S4). On
the other hand, although PDGF-AB and PDGF-BB did not affect expressions of Rankl and Opg
(Fig. S4), they significantly elevated that of Csf1 (Fig. 6A). In agreement with it, PDGF-BB
stimulated secretion of M-CSF in osteoblast cultures (Fig. 6B). Removal of both PDGFRs in
osteoblasts prevented PDGF-AB- and PDGF-BB-induced Csf1 expression and M-CSF
secretion (Fig. 6A,B). PDGF-BB did not have any effect in basal condition but slightly
enhanced osteoclastogenesis in osteoblasts/osteoclasts co-cultures (Fig. 6C). Conversely,
abrogation of PDGFR signaling in osteoblasts blocked osteoclastogenesis in basal condition, in
presence of calcitriol, or in presence of calcitriol and PDGF-BB (Fig. 6C,D). Interestingly,
defective osteoclastogenesis in absence of PDGFR signaling in osteoblasts could be rescued by
supplementation with recombinant M-CSF (Fig. 6C). Suppression of both PDGFRs in
osteoblasts stopped PDGF-BB-stimulated activation of PI3K-AKT, PLCγ-PKC, SRC and
ERK1/2 signaling pathways (Fig. 6E). While selective inhibitions of PI3K, PKC and SRC did
not affect PDGF-BB-induced expressions of Csf1, selective inhibition of MEK1/2 (mitogen-
activated kinase kinases upstream of ERK1/2) blocked it (Fig. 6F). Together, those results
indicate that PDGF-BB promoted Csf1 expression through PDGFR-ERK1/2 signaling pathway.
This article is protected by copyright. All rights reserved.
Page 18
17
Suppression of PDGFR signaling inhibits osteoblast proliferation and migration, but
stimulates osteoblast differentiation in vitro
Finally, we analyzed effects of PDGFR signaling on osteoblast proliferation, migration and
differentiation in vitro. PDGF-AA treatment did not have any effect on those parameters (Fig.
7A-C). PDGF-AB and PDGF-BB treatments enhanced osteoblast proliferation and migration
(Fig. 7A,B), and inhibited osteoblast differentiation, as shown by decreased ALP activity (Fig.
7C). In contrast, removal of PDGFRs in osteoblasts inhibited osteoblast proliferation and
migration (Fig. 7A,B), but stimulated osteoblast differentiation and BMP2-induced
mineralization (Fig. 7C,D).
Discussion
Osteoprogenitors and osteoblasts express PDGFRα and β, but overlapping and distinct
functions of those two receptor tyrosine kinases in osteoblast biology and bone remodeling
remained to be elucidated. Here, we report that PDGFRA and PDGFRB were expressed by
osteoprogenitors in close proximity to PDGFB-expressing osteoclasts within human trabecular
bone remodeling units. Moreover, we show that postnatal genetic ablation of single PDGFRα
or β under the control of the Osx promoter did not affect bone metabolism, but that postnatal
removal of both PDGFRs in osteoblast lineage cells increased trabecular bone mass by lowering
Csf1 expression, M-CSF secretion, osteoclast number and therefore bone resorption, without
having any effect on bone formation. Abrogation of PDGFR signaling in osteoblasts prevented
calcitriol-mediated osteoclastogenesis in co-cultures, while its activation promoted ERK1/2-
mediated Csf1 expression.
This article is protected by copyright. All rights reserved.
Page 19
18
The identification of osteoclasts as a cellular source of PDGF-BB, and osteoblast lineage
canopy and reversal cells as PDGF-responsive cells at the surface of human cancellous bone
indicates that the PDGF/PDGFR signaling system is operative within trabecular bone
remodeling compartment. To investigate functions of PDGFRs in osteoblast lineage cells, we
generated mice in which PDGFRs-encoding genes were selectively deleted in Osx-positive
cells. Surprisingly, while Pdgfra cKO or Pdgfrb cKO mice did not exhibit any bone phenotype,
mice lacking both receptors in Osx-positive cells displayed increased trabecular bone mass due
to reduced bone resorption and unchanged bone formation. The absence of bone phenotype in
Pdgfrb cKO mice is consistent with previous findings (10,11). Given that PDGF-AA only binds
to PDGFRαα, the absence of bone phenotype in Pdgfra cKO mice indicates that PDGF-AA
plays minor role in osteoblast development and function, which is coherent with the absence of
PDGF-AA effect on osteoblast cultures. Altogether, our results also show that either PDGFRα
or PDGFRβ can compensate for the loss of the other receptor in osteoblast lineage cells and
that both receptors redundantly regulate osteoclastogenesis and bone resorption. The fact that
absence of both PDGFRs in osteoblast lineage cells only affected cancellous bone in axial and
appendicular skeletal sites may be explained by its predominant effect on bone resorption and
by cellular sources of PDGFs, i.e. osteoclasts, which are more abundant in the trabecular
compartment than in cortical one (4-6).
Postnatal removal of PDGFRs in Osx-positive cells decreased osteoclast number and bone
resorption, an effect that was associated with reduced M-CSF production, indicating that
PDGFRα and PDGFRβ in osteoblast lineage cells positively regulate osteoclastogenesis and
bone resorption by elevating Csf1 expression. This notion is further supported by our in vitro
This article is protected by copyright. All rights reserved.
Page 20
19
investigations. First, suppression of PDGFR signaling in osteoblasts prevented calcitriol-
mediated osteoclastogenesis in co-cultures. Second, supplementation with M-CSF in this
situation was sufficient to restore osteoclast formation. Third, PDGF-AB and PDGF-BB could
stimulate Csf1 expression in osteoblasts. Our findings are in agreement with increased Csf1
expression and osteoclast number in mice transplanted with Pdgfb-overexpressing
hematopoietic stem cells that preferentially colonize bone marrow niches (14). Expression of
Rankl was also reduced in bones of mice lacking both PDGFRs in Osx-positive cells. This
observation could be consistent with stimulated Rankl expression induced by PDGF-BB in
osteoblast-like cell lines, or by Pdgfb overexpression or aberrant osteoblastic PDGFR signaling
in mice (14-16), but this has not been confirmed in vitro or at the protein level in vivo. Moreover,
ERK1/2 signaling pathway was required for PDGF-BB-induced Csf1 expression in osteoblast
cultures. That information provides a molecular mechanism by which ERK1/2 signaling in
osteoprogenitors regulates bone resorption (28,29). Our data are also in accordance with the
requirement of osteoblast lineage-derived M-CSF for osteoclast development in the bone
microenvironment (30). In addition, our results suggest that osteoclasts can control their own
optimal development and maintenance as well as those of neighboring osteoclasts within the
same bone remodeling unit by releasing PDGF-BB. In this context, however, a role of
osteoblast lineage-secreted M-CSF in the regulation of osteoclast migration, activity and
survival cannot be ruled out (31).
Unexpectedly, abrogation of PDGFR signaling in osteoblast lineage cells did not affect bone
formation. First, distribution of endomucin-positive blood vessels remained unchanged in bone
metaphyses of mice lacking PDGFRs in Osx-positive cells, indicating that PDGFR signaling in
This article is protected by copyright. All rights reserved.
Page 21
20
osteoblast lineage cells is not required for the formation of local blood vessels (6,9). Second,
osteoblast number was not affected in mice lacking both PDGFRs in Osx-positive cells,
showing that PDGFR signaling in osteoblast lineage cells is not involved in osteoprogenitor
proliferation and recruitment to bone surfaces during physiological bone modeling and
remodeling. Those data confirm that PDGFR signaling governs osteoprogenitor expansion and
recruitment by acting in earlier progenitors than Osx-positive cells in physiological condition
(9,32). Third, bone formation remained unchanged on trabecular surfaces despite diminished
bone resorption in mice lacking PDGFRs in osteoblast lineage cells, suggesting an uncoupling
between bone resorption and formation. That may be explained by a cell-autonomous
stimulatory effect of PDGFR suppression on osteoblast differentiation, as observed in our in
vitro analyses and in previous investigations (4,12). This explanation is supported by a recent
report showing that deletion of PDGFRβ in Osx-positive cells leads to accelerated osteoblast
differentiation and callus mineralization during fracture healing (10).
Conditional suppression of PDGFRα and β in Osx-positive cells decreased bone marrow level
of M-CSF, osteoclast number and bone resorption, and increased trabecular bone mass in both
gonad-intact and Ovx female mice, thus confirming results obtained with male mice. However,
it did not prevent Ovx-induced bone loss, which is somehow coherent with the literature.
Indeed, estrogen deficiency reduces bone marrow level of PDGF-BB, and supplementation with
PDGF-BB in this situation averts cancellous bone loss by restoring pericyte-mediated pro-
osteogenic vascularization (6,33). Perhaps more conceivable, local administration of PDGF-
BB could be used to enhance fracture repair in diabetic and osteoporotic patients (10,34,35). In
that context, our results imply that such pharmacological intervention must respect adequate
This article is protected by copyright. All rights reserved.
Page 22
21
timing to stimulate callus vascularization, expansion and recruitment of osteoprogenitors,
without affecting later callus mineralization and remodeling (10). Finally, excessive PDGFR
activity is involved in several cancers, including those that give rise to osteolytic lesions such
as multiple myeloma and bone metastases from breast and prostate cancers (2,36-38). Our
findings reveal that PDGFR inhibition could have additional beneficial effect by lowering the
occurrence of osteolytic lesions associated with those types of cancers.
This article is protected by copyright. All rights reserved.
Page 23
22
Acknowledgments
We would like to thank Pierre Apostolides for his expert technical assistance. This work was
supported by the Swiss National Science Foundation (310030-166341), by the Fondation pour
la Recherche sur l’Ostéoporose et les Maladies Osseuses, and by the Novartis Foundation for
medical-biological research (Project number 16C212 - Basel, Switzerland), by the VELUX
FOUNDATION (VELUX25723), and by the Region of Southern Denmark Research Fund
(18/17871).
Authors’ roles: Study design: CT. Study conduct: JB, CT, CE and CMA. Data collection: JB,
CT, CMA and TLA. Data analysis: JB, CT, CMA and TLA. Data interpretation: JB, CT, CMA
and TLA. Drafting manuscript: JB, CT, CMA and TLA. Revising manuscript content: JC, CT,
CE, CMA and TLA. Approving final version of manuscript: JB, JC, CT, CE, CMA and TLA.
CT takes responsibility for the integrity of the data analysis.
This article is protected by copyright. All rights reserved.
Page 24
23
References
1. Caplan AI, Correa D. PDGF in bone formation and regeneration: new insights into a
novel mechanism involving MSCs. J Orthop Res. 2011;29(12):1795-803.
2. Andrae J, Gallini R, Betsholtz C. Role of platelet-derived growth factors in physiology
and medicine. Genes Dev. 2008;22(10):1276-312.
3. Alvarez RH, Kantarjian HM, Cortes JE. Biology of platelet-derived growth factor and
its involvement in disease. Mayo Clin Proc. 2006;81(9):1241-57.
4. Kubota K, Sakikawa C, Katsumata M, Nakamura T, Wakabayashi K. Platelet-derived
growth factor BB secreted from osteoclasts acts as an osteoblastogenesis inhibitory
factor. J Bone Miner Res. 2002;17(2):257-65.
5. Sanchez-Fernandez MA, Gallois A, Riedl T, Jurdic P, Hoflack B. Osteoclasts control
osteoblast chemotaxis via PDGF-BB/PDGF receptor beta signaling. PLoS One.
2008;3(10):e3537.
6. Xie H, Cui Z, Wang L, Xia Z, Hu Y, Xian L, et al. PDGF-BB secreted by preosteoclasts
induces angiogenesis during coupling with osteogenesis. Nat Med. 2014;20(11):1270-
8.
7. Matthews BG, Grcevic D, Wang L, Hagiwara Y, Roguljic H, Joshi P, et al. Analysis of
alphaSMA-labeled progenitor cell commitment identifies notch signaling as an
important pathway in fracture healing. J Bone Miner Res. 2014;29(5):1283-94.
8. Kusumbe AP, Ramasamy SK, Itkin T, Mae MA, Langen UH, Betsholtz C, et al. Age-
dependent modulation of vascular niches for haematopoietic stem cells. Nature.
2016;532(7599):380-4.
This article is protected by copyright. All rights reserved.
Page 25
24
9. Gao B, Deng R, Chai Y, Chen H, Hu B, Wang X, et al. Macrophage-lineage TRAP+
cells recruit periosteum-derived cells for periosteal osteogenesis and regeneration. J
Clin Invest. 2019;129(6):2578-94.
10. Bohm AM, Dirckx N, Tower RJ, Peredo N, Vanuytven S, Theunis K, et al. Activation
of Skeletal Stem and Progenitor Cells for Bone Regeneration Is Driven by PDGFRbeta
Signaling. Dev Cell. 2019;51(2):236-54 e12.
11. Tokunaga A, Oya T, Ishii Y, Motomura H, Nakamura C, Ishizawa S, et al. PDGF
receptor beta is a potent regulator of mesenchymal stromal cell function. J Bone Miner
Res. 2008;23(9):1519-28.
12. Wang X, Matthews BG, Yu J, Novak S, Grcevic D, Sanjay A, et al. PDGF Modulates
BMP2-Induced Osteogenesis in Periosteal Progenitor Cells. JBMR Plus.
2019;3(5):e10127.
13. Soriano P. The PDGF alpha receptor is required for neural crest cell development and
for normal patterning of the somites. Development. 1997;124(14):2691-700.
14. Chen W, Baylink DJ, Brier-Jones J, Neises A, Kiroyan JB, Rundle CH, et al. PDGFB-
based stem cell gene therapy increases bone strength in the mouse. Proc Natl Acad Sci
U S A. 2015;112(29):E3893-900.
15. O'Sullivan S, Tay ML, Lin JM, Bava U, Callon K, Cornish J, et al. Tyrosine Kinase
Inhibitors Regulate OPG through Inhibition of PDGFRbeta. PLoS One.
2016;11(10):e0164727.
This article is protected by copyright. All rights reserved.
Page 26
25
16. Bartelt A, Behler-Janbeck F, Beil FT, Koehne T, Muller B, Schmidt T, et al. Lrp1 in
osteoblasts controls osteoclast activity and protects against osteoporosis by limiting
PDGF-RANKL signaling. Bone Res. 2018;6:4.
17. Lassen NE, Andersen TL, Pløen GG, Søe K, Hauge EM, Harving S, et al. Coupling of
Bone Resorption and Formation in Real Time: New Knowledge Gained From Human
Haversian BMUs. J Bone Miner Res. 2017;32(7):1395-405.
18. Rodda SJ, McMahon AP. Distinct roles for Hedgehog and canonical Wnt signaling in
specification, differentiation and maintenance of osteoblast progenitors. Development.
2006;133(16):3231-44.
19. Tallquist MD, Soriano P. Cell autonomous requirement for PDGFRalpha in populations
of cranial and cardiac neural crest cells. Development. 2003;130(3):507-18.
20. Schmahl J, Rizzolo K, Soriano P. The PDGF signaling pathway controls multiple
steroid-producing lineages. Genes Dev. 2008;22(23):3255-67.
21. Davey RA, Clarke MV, Sastra S, Skinner JP, Chiang C, Anderson PH, et al. Decreased
body weight in young Osterix-Cre transgenic mice results in delayed cortical bone
expansion and accrual. Transgenic Res. 2012;21(4):885-93.
22. Bouxsein ML, Boyd SK, Christiansen BA, Guldberg RE, Jepsen KJ, Muller R.
Guidelines for assessment of bone microstructure in rodents using micro-computed
tomography. J Bone Miner Res. 2010;25(7):1468-86.
23. Dempster DW, Compston JE, Drezner MK, Glorieux FH, Kanis JA, Malluche H, et al.
Standardized nomenclature, symbols, and units for bone histomorphometry: a 2012
This article is protected by copyright. All rights reserved.
Page 27
26
update of the report of the ASBMR Histomorphometry Nomenclature Committee. J
Bone Miner Res. 2013;28(1):2-17.
24. Thouverey C, Caverzasio J. Suppression of p38alpha MAPK Signaling in Osteoblast
Lineage Cells Impairs Bone Anabolic Action of Parathyroid Hormone. J Bone Miner
Res. 2016;31(5):985-93.
25. Kristensen HB, Andersen TL, Marcussen N, Rolighed L, Delaisse JM. Osteoblast
recruitment routes in human cancellous bone remodeling. Am J Pathol.
2014;184(3):778-89.
26. Andersen TL, Abdelgawad ME, Kristensen HB, Hauge EM, Rolighed L, Bollerslev J,
et al. Understanding coupling between bone resorption and formation: are reversal cells
the missing link? Am J Pathol. 2013;183(1):235-46.
27. Abdelgawad ME, Delaisse JM, Hinge M, Jensen PR, Alnaimi RW, Rolighed L, et al.
Early reversal cells in adult human bone remodeling: osteoblastic nature, catabolic
functions and interactions with osteoclasts. Histochem Cell Biol. 2016;145(6):603-15.
28. Matsushita T, Chan YY, Kawanami A, Balmes G, Landreth GE, Murakami S.
Extracellular signal-regulated kinase 1 (ERK1) and ERK2 play essential roles in
osteoblast differentiation and in supporting osteoclastogenesis. Mol Cell Biol.
2009;29(21):5843-57.
29. Kim JM, Yang YS, Park KH, Oh H, Greenblatt MB, Shim JH. The ERK MAPK
Pathway Is Essential for Skeletal Development and Homeostasis. Int J Mol Sci.
2019;20(8). pii: E1803.
This article is protected by copyright. All rights reserved.
Page 28
27
30. Abboud SL, Woodruff K, Liu C, Shen V, Ghosh-Choudhury N. Rescue of the
Osteopetrotic Defect in op/op Mice by Osteoblast-Specific Targeting of Soluble
Colony-Stimulating factor-1. Endocrinology. 2002;143(5):1942-9.
31. Fuller K, Owens JM, Jagger CJ, Wilson A, Moss R, Chambers TJ. Macrophage
Colony-Stimulating Factor Stimulates Survival and Chemotactic Behavior in Isolated
Osteoclasts. J Exp Med. 1993;178(5):1733-44.
32. Park D, Spencer JA, Koh BI, Kobayashi T, Fujisaki J, Clemens TL, et al. Endogenous
bone marrow MSCs are dynamic, fate-restricted participants in bone maintenance and
regeneration. Cell Stem Cell. 2012;10(3):259-72.
33. Mitlak BH, Finkelman RD, Hill EL, Li J, Martin B, Smith T, et al. The effect of
systemically administered PDGF-BB on the rodent skeleton. J Bone Miner Res.
1996;11(2):238-47.
34. Al-Zube L, Breitbart EA, O'Connor JP, Parsons JR, Bradica G, Hart CE, et al.
Recombinant human platelet-derived growth factor BB (rhPDGF-BB) and beta-
tricalcium phosphate/collagen matrix enhance fracture healing in a diabetic rat model.
J Orthop Res. 2009;27(8):1074-81.
35. Hollinger JO, Onikepe AO, MacKrell J, Einhorn T, Bradica G, Lynch S, et al.
Accelerated fracture healing in the geriatric, osteoporotic rat with recombinant human
platelet-derived growth factor-BB and an injectable beta-tricalcium phosphate/collagen
matrix. J Orthop Res. 2008;26(1):83-90.
This article is protected by copyright. All rights reserved.
Page 29
28
36. Lev DC, Kim SJ, Onn A, Stone V, Nam DH, Yazici S, et al. Inhibition of platelet-
derived growth factor receptor signaling restricts the growth of human breast cancer in
the bone of nude mice. Clin Cancer Res. 2005;11(1):306-14.
37. Dolloff NG, Shulby SS, Nelson AV, Stearns ME, Johannes GJ, Thomas JD, et al. Bone-
metastatic potential of human prostate cancer cells correlates with Akt/PKB activation
by alpha platelet-derived growth factor receptor. Oncogene. 2005;24(45):6848-54.
38. Coluccia AM, Cirulli T, Neri P, Mangieri D, Colanardi MC, Gnoni A, et al. Validation
of PDGFR beta and c-Src tyrosine kinases as tumor/vessel targets in patients with
multiple myeloma: preclinical efficacy of the novel, orally available inhibitor dasatinib.
Blood. 2008;112(4):1346–56.
This article is protected by copyright. All rights reserved.
Page 30
29
Figure legends
Fig. 1. PDGFB mRNA expression in TRAP-positive osteoclasts and PDGFRA and PDGFRB
mRNA expression in proximate canopy and reversal cells in iliac crest bone biopsy specimens
from four human volunteers without known musculoskeletal disease. The mRNA expression
was visualized by in situ hybridization (red dots) combined with TRAP immunostaining (black
stain). Images are representative of those obtained from the four patients. Scale bars are 20 µm.
(A, B) PDGFB mRNA expression in megakaryocytes (MKC) and vascular structures (VS)
within bone marrow. (C) Overview image showing the location of the osteoclastic zone
illustrated in adjacent sections in D, F and H. (D, E) PDGFB mRNA expression in TRAP-
positive osteoclasts (OC). Note the red dots within the black osteoclasts. (F, G) PDGFRA
mRNA expression in elongated canopy cells (red arrows) above osteoclasts (OC) and reversal
cells (black arrows) on eroded surfaces close to osteoclasts. (H, I) PDGFRB mRNA expression
in elongated canopy cells (red arrows) above osteoclasts and reversal cells (black arrows) on
eroded surfaces close to osteoclasts. (J, K) Images from negative controls.
Fig. 2. Postnatal osteoblast lineage-specific suppression of PDGFR signaling increases
trabecular bone mass in male mice. (A) Representative reconstructed µCT images of proximal
tibiae from 18-week-old male Osx-Cre, Pdgfra cKO, Pdgfrb cKO and Pdgfra cKO;Pdgfrb cKO
mice, in which Cre expression and consequent conditional gene deletion were induced (by
stopping doxycycline treatment) from 6 weeks to 18 weeks of age. (B) Trabecular bone
microarchitecture measured at proximal tibiae. µCT parameters include BV/TV: bone
volume/total volume, Tb.Th: trabecular thickness, Tb.N: trabecular number. (n=6 per group).
This article is protected by copyright. All rights reserved.
Page 31
30
(C) Representative reconstructed µCT images of tibial midshafts from 18-week-old male Osx-
Cre, Pdgfra cKO, Pdgfrb cKO and Pdgfra cKO;Pdgfrb cKO mice, in which Cre expression and
consequent conditional gene deletion was induced from 6 weeks to 18 weeks of age. (D)
Cortical bone microarchitecture measured at tibial midshaft. µCT parameters include Ct.TV:
cortical total volume, Ct.BV: cortical bone volume and Ct.Th: cortical thickness (n=6 per
group). Comparisons between the multiple genotypes were analyzed by one-way ANOVA
followed by Holm-Sidak post hoc tests.
Fig. 3. Postnatal osteoblast lineage-specific suppression of PDGFR signaling decreases bone
resorption. (A) Histomorphometric parameters of trabecular bone resorption measured at the
secondary spongiosa of distal femurs from 18-week-old male Osx-Cre, Pdgfra cKO, Pdgfrb
cKO and Pdgfra cKO;Pdgfrb cKO mice (n=6 per group). Cre expression and consequent
conditional gene deletion were induced (by stopping doxycycline treatment) from 6 weeks to
18 weeks of age. N.Oc/Pm: osteoclast number/bone perimeter, Oc.S/BS: osteoclast
surface/bone surface. (D) Representative images of TRAP-stained histological sections of distal
femurs (n=6 per group). (E) Quantitative RT-PCR analyses of Rankl, Opg and Csf1 expressions
in proximal tibial metaphyses (n=6 per group). (F) Levels of RANKL, OPG and M-CSF
detected by ELISA in bone marrow supernatants (n=6 per group). Comparisons between the
multiple genotypes were analyzed by one-way ANOVA followed by Holm-Sidak post hoc tests.
Fig. 4. Postnatal osteoblast lineage-specific suppression of PDGFR signaling does not affect
bone formation. (A-C) Histomorphometric parameters of trabecular bone formation measured
This article is protected by copyright. All rights reserved.
Page 32
31
at the secondary spongiosa of distal femurs from 18-week-old male Osx-Cre, Pdgfra cKO,
Pdgfrb cKO and Pdgfra cKO;Pdgfrb cKO mice (n=6 per group). Cre expression and consequent
conditional gene deletion were induced (by stopping doxycycline treatment) from 6 weeks to
18 weeks of age. (A) Tb.MAR: trabecular mineral apposition rate, Tb.MS/BS: trabecular
mineralizing surfaces/bone surfaces, Tb.BFR: trabecular bone formation rate. (B)
Representative images of calcein-labeled histological sections of distal femurs. (C) Ob.N/Pm:
osteoblast number/bone perimeter. (D) Serum levels of PINP detected by ELISA (n=6 per
group). (E) Immunohistochemical staining of endomucin (red) and nuclei (blue) on histological
sections of distal femurs and quantification of endomucin-positive area using ImageJ software
(n=3 per group). (F) N.Ocy/BS: number of osteocytes/bone surface (n=5 per group).
Comparisons between the multiple genotypes were analyzed by one-way ANOVA followed by
Holm-Sidak post hoc tests.
Fig. 5. Postnatal osteoblast lineage-specific suppression of PDGFR signaling increases
trabecular bone mass in both gonad-intact and ovariectomized female mice. Twelve-week-old
female Osx-Cre and Pdgfra cKO;Pdgfrb cKO mice were Sham-operated or Ovx and sacrificed
8 weeks later. Cre expression and consequent conditional gene deletion were induced (by
stopping doxycycline treatment) one week prior Sham and Ovx operations. (A) Trabecular bone
microarchitecture measured at proximal tibiae (n=6 per group). µCT parameters include
BV/TV: bone volume/total volume, Tb.Th: trabecular thickness, Tb.N: trabecular number,
Tb.Sp: trabecular separation. (B, C) Histomorphometric parameters of trabecular bone
resorption (B) and formation (C) measured at the secondary spongiosa of distal femurs (n=6
This article is protected by copyright. All rights reserved.
Page 33
32
per group). (F) Levels of RANKL, OPG and M-CSF detected by ELISA in bone marrow
supernatants (n=6 per group). The two-way ANOVA did not detect significant interaction
between effects of genotype and those of Ovx. The comparison between the two genotypes (p
values in bold) or between Sham and Ovx operations (p values in italic) was performed using
unpaired t-test.
Fig. 6. Suppression of PDGFR signaling in osteoblasts inhibits M-CSF production and M-CSF-
mediated osteoclastogenesis in vitro. Primary osteoblasts were isolated from long bones of
Pdgfraf/f;Pdgfrbf/f mice and infected with empty or Cre-expressing adenoviruses to obtain
Pdgfraf/f;Pdgfrbf/f (control) and ΔPdgfra;ΔPdgfrb (without PDGFRs) osteoblasts respectively.
(A) Quantitative RT-PCR analysis of Csf1 expression in confluent cultures of Pdgfraf/f;Pdgfrbf/f
and ΔPdgfra;ΔPdgfrb osteoblasts treated with 25 ng/mL PDGF-AA, PDGF-AB or PDGF-BB
or their vehicle (Veh) solution for 8 hours. (B) Levels of M-CSF detected by ELISA in the
culture medium of Pdgfraf/f;Pdgfrbf/f and ΔPdgfra;ΔPdgfrb osteoblasts treated with Veh or 25
ng/mL PDGF-BB for 48 hours. (C) Number of TRAP-positive multinucleated cells in
osteoblasts/osteoclasts co-cultures treated with Veh, 10 ng/mL PDGF-BB or/and 10-8 M 1,25-
dihydroxyvitamin D3 (Vit.D3), and supplemented or not with 20 ng/mL M-CSF for 8 days. (D)
Representative images of TRAP-positive multinucleated cells in co-cultures with Veh or 10-8
M Vit.D3 for 8 days. (E) Western blot analyses of PDGFR signaling in Pdgfraf/f;Pdgfrbf/f and
ΔPdgfra;ΔPdgfrb osteoblasts treated with Veh or 25 ng/mL PDGF-BB at indicated times. (F)
Quantitative RT-PCR analysis of Csf1 expression in confluent cultures of Pdgfraf/f;Pdgfrbf/f
osteoblasts pretreated with 5 µM LY294002 (PI3K inhibitor), 5 µM U0126 (MEK1/2 inhibitor),
This article is protected by copyright. All rights reserved.
Page 34
33
5 µM GO6983 (PKC inhibitor) and 5 µM SU6656 (Src family kinase inhibitor) for 30 minutes
and then, treated with Veh or 25 ng/mL PDGF-BB for 8 hours. Data are represented as mean ±
SD. # p≤0.002 versus Veh, * p≤0.002 versus Pdgfraf/f;Pdgfrbf/f.
Fig. 7. Suppression of PDGFR signaling in osteoblasts stimulates osteoblast differentiation in
vitro. Primary osteoblasts were isolated from long bones of Pdgfraf/f;Pdgfrbf/f mice and infected
with empty or Cre-expressing adenoviruses to obtain Pdgfraf/f;Pdgfrbf/f (control) and
ΔPdgfra;ΔPdgfrb (without PDGFRs) osteoblasts respectively. (A) Number of primary
Pdgfraf/f;Pdgfrbf/f and ΔPdgfra;ΔPdgfrb osteoblasts treated with Veh, 25 ng/mL PDGF-AA,
PDGF-AB or PDGF-BB for 3 days. (B) Quantification of cell migration rates in scratched sub-
confluent cultures of Pdgfraf/f;Pdgfrbf/f and ΔPdgfra;ΔPdgfrb osteoblasts treated with Veh, 25
ng/mL PDGF-AA, PDGF-AB or PDGF-BB for 3 days. (C) Alkaline phosphatase (ALP)
activity of Pdgfraf/f;Pdgfrbf/f and ΔPdgfra;ΔPdgfrb osteoblasts incubated in osteogenic
medium and treated with Veh, 25 ng/mL PDGF-AA, PDGF-AB or PDGF-BB for 6 days. (D)
Representative images of matrix mineralization by Pdgfraf/f;Pdgfrbf/f and ΔPdgfra;ΔPdgfrb
osteoblasts treated with Veh or 200 ng/mL BMP2 for 14 days. Data are represented as mean ±
SD. # p≤0.001 versus Veh, * p≤0.001 versus Pdgfraf/f;Pdgfrbf/f.
This article is protected by copyright. All rights reserved.
Page 35
JBMR_4150_Fig. 1 (revised).tif
This article is protected by copyright. All rights reserved.
Page 36
JBMR_4150_Fig. 2 (revised).tif
This article is protected by copyright. All rights reserved.
Page 37
JBMR_4150_Fig. 3 (revised).tif
This article is protected by copyright. All rights reserved.
Page 38
JBMR_4150_Fig. 4 (revised).tif
This article is protected by copyright. All rights reserved.
Page 39
JBMR_4150_Fig. 5 (revised).tif
This article is protected by copyright. All rights reserved.
Page 40
JBMR_4150_Fig. 6_revised_2.tif
This article is protected by copyright. All rights reserved.
Page 41
JBMR_4150_Fig. 7_revised_2.tif
This article is protected by copyright. All rights reserved.