TISSUE-SPECIFIC STEM CELLS
Impaired Osteoblastogenesis in a Murine Model of Dominant
Osteogenesis Imperfecta: A New Target for Osteogenesis
Imperfecta Pharmacological Therapy
ROBERTA GIOIA,aCRISTINA PANARONI,a,bROBERTA BESIO,aGIOVANNI PALLADINI,a,cGIAMPAOLO MERLINI,a,c
VINCENZO GIANSANTI,dIVANA A. SCOVASSI,dSIMONA VILLANI,eISABELLA VILLA,fANNA VILLA,b,g
PAOLO VEZZONI,b,gRUGGERO TENNI,aANTONIO ROSSI,aJOAN C. MARINI,hANTONELLA FORLINOa
aDepartment of Molecular Medicine, Section of Biochemistry andeDepartment of Health Sciences, Section of
Medical Statistic and Epidemiology, University of Pavia, Pavia, Italy;bMilan Unit, Istituto di Ricerca Genetica e
Biomedica, CNR, Milan, Italy;cAmyloidosis Research and Treatment Center, Biotechnology Research
Laboratories, Fondazione IRCCS Policlinico San Matteo, Pavia, Italy;dMolecular Genetics Institute, CNR, Pavia,
Italy;fBone Metabolic Unit, San Raffaele Scientific Institute, Milan, Italy;gIstituto Clinico Humanitas, Rozzano,
Italy;hBone and Extracellular Matrix Branch, NICHD, NIH, Bethesda, Maryland, USA
Key Words. Osteogenesis imperfecta•Adult stem cell•Osteoblastogenesis•Adipogenesis•Adult stem cells differentiation•Autophagy
The molecular basis underlying the clinical phenotype in
bone diseases is customarily associated with abnormal
extracellular matrix structure and/or properties. More
recently, cellular malfunction has been identified as a con-
comitant causative factor and increased attention has
focused on stem cells differentiation. Classic osteogenesis
imperfecta (OI) is a prototype for heritable bone dyspla-
sias: it has dominant genetic transmission and is caused
by mutations in the genes coding for collagen I, the most
abundant protein in bone. Using the Brtl mouse, a well-
characterized knockin model for moderately severe domi-
nant OI, we demonstrated an impairment in the differen-
osteoblasts. In mutant mesenchymal stem cells (MSCs),
the expression of early (Runx2 and Sp7) and late (Col1a1
and Ibsp) osteoblastic markers was significantly reduced
with respect to wild type (WT). Conversely, mutant MSCs
generated more colony-forming unit-adipocytes compared
to WT, with more adipocytes per colony, and increased
number and size of triglyceride drops per cell. Autophagy
upregulation was also demonstrated in mutant adult
MSCs differentiating toward osteogenic lineage as conse-
quence of endoplasmic reticulum stress due to mutant col-
lagen retention. Treatment of the Brtl mice with the
proteasome inhibitor Bortezomib ameliorated both osteo-
blast differentiation in vitro and bone properties in vivo as
demonstrated by colony-forming unit-osteoblasts assay
and peripheral quantitative computed tomography analy-
sis on long bones, respectively. This is the first report of
impaired MSC differentiation to osteoblasts in OI, and it
identifies a new potential target for the pharmacological
treatment of the disorder. STEM CELLS 2012;30:1465–1476
Disclosure of potential conflicts of interest is found at the end of this article.
Osteogenesis imperfecta (OI) is a heritable bone dysplasia
characterized by reduced bone mass, increased bone fragil-
ity, and skeletal deformity, associated with abnormal type I
collagen. Altered type I collagen quantity, sequence, or
post-translational modification can all be causes of the
Mutations in the genes COL1A1 and COL1A2, coding for
the a chains of type I collagen, are among the most frequent
molecular defects identified in OI patients, with glycine sub-
stitutions constituting more than 85% of reported causative
abnormalities. The presence of glycine at every third amino
acid is in fact essential for triple helical folding, since any
other amino acid in the first position of the classic collagenic
triplet Gly-X-Y represents a steric hindrance to folding .
Among several available murine models for OI, the Brtl
Author contributions: R.G.: collection and assembly of data, data analysis and interpretation, and manuscript writing; C.P., S.V., and
I.V.: collection and assembly of data and data analysis and interpretation; R.B.: collection and assembly of data and manuscript
revision; G.P. and G.M.: provision of study material and data analysis and interpretation; V.G., I.A.S., A.V., and R.T.: data analysis and
interpretation and manuscript revision; P.V.: data analysis and interpretation, financial support, and manuscript revision; A.R.:
conception and design, data analysis and interpretation, and manuscript revision; J.C.M.: data analysis and interpretation and manuscript
writing; A.F.: conception and design, financial support, collection and assembly of data, data analysis and interpretation, and manuscript
Correspondence: Antonella Forlino, PhD, Section of Biochemistry, Department of Molecular Medicine, University of Pavia, Via
Taramelli 3/B, 27100 Pavia, Italy. Telephone: 39-0382-987235; Fax: 39-0382-423108; e-mail: email@example.com
5, 2011; accepted for publication March 25, 2012; first published online in STEM CELLS EXPRESS April 17, 2012. V
1066-5099/2012/$30.00/0 doi: 10.1002/stem.1107
STEM CELLS 2012;30:1465–1476 www.StemCells.com
C AlphaMed Press
mouse  uniquely reproduces the classic OI-causative muta-
tion, a glycine to cysteine substitution in position 349 of the
a1 chain of type I collagen, autosomal dominant transmission,
and nonlethal moderately severe clinical outcome of OI type
IV patients, based on Sillence’s classification .
We have previously characterized Brtl as a valid model
for human OI. Brtl has reduced body weight, reduced bone
mineral density and brittle bones. Brtl femora have compro-
mised geometrical properties with reduced cross-sectional
area, cortical thickness, and bending moment of inertia, with
a negative peak at 2 months of age . Osteoblast and osteo-
clast uncoupling has been recently demonstrated in Brtl
bones, with increased osteoclast number and function and
reduced matrix production by osteoblasts . Impaired osteo-
blast metabolism was also hypothesized, based on microarray
and proteomic analysis of calvarial bone tissues from mutant
mice . Based on histomorphometric analyses, bone loss
and decreased bone mass are characteristic of various patho-
logical conditions including aging and osteoporosis and are
accompanied by more adipocytes at the expense of osteoblast
formation in bone marrow environment [8, 9].
Osteoblasts and adipocytes share a common mesenchymal
bone marrow precursor and have plasticity during differentia-
tion, due to cytoskeleton alteration as well as to structural or
secreted factor changes in the mesenchymal stem cell (MSC)
microenvironment niche [10, 11]. More recently, a well-
defined population of committed preosteoblasts and preadipo-
cytes has been described and isolated , complicating the
scenario, but helping to understand why a decrease in osteo-
blast differentiation is not always coupled to an increased adi-
pocyte formation, as reported more recently in animal models
and human patients [9, 13, 14].
Drug targeting of adult stem cells to control their differen-
tiation is an appealing strategy for regenerative medicine.
However, only a few drugs are known to target stem cell pop-
ulation in vivo and side effects are often a major issue .
The proteasome inhibitor Bortezomib (Btz), widely used in
the treatment of multiple myeloma , was demonstrated to
have, at subtherapeutic doses, an osteoblastogenic effect on
adult murine and human MSCs, at least in part by stabilizing
runt-related transcription factor 2 (RUNX2) and acting
directly on the expression of type I collagen . Because
Btz is already Food and Drug Administration (FDA) approved
for antitumorigenic applications, it is appealing to test its
effect on other diseases, especially using doses well-below
those causing side effects in human patients.
In this study, we demonstrated that the Brtl mouse model
of classic OI has an impaired differentiation of adult bone
marrow progenitor cells. We identified reduced osteogenic
potential and a skewing to adipocytic differentiation in mutant
adult MSCs, determined at least in part by autophagy activa-
tion. Treatment of Brtl mice with the proteasomal inhibitor
Btz rescued the MSC osteoblastogenic capacity in vitro and
ameliorated bone properties in vivo, thus identifying a new
target for OI pharmacological treatment.
MATERIALS AND METHODS
Brtl mice and their wild-type (WT) littermates were used for this
study . Mice were genotyped by PCR as previously reported
. Unless otherwise noted all mice used for this study were 2-
month-old females. All animal procedures were approved by the
local Pavia City Hall authorities (reference no. 7287/00) and by
the Italian Ministry of Health according to Italian law (D.L. 116/
92, Art. 12).
Fluorescence-Activated Cell Sorting
Suspensions of bone marrow cells were obtained from hind limbs
of Brtl (n ¼ 8) and WT mice (n ¼ 7) as described . Mature
hematopoietic cells were depleted using the Mouse Hematopoietic
Progenitor Enrichment kit on Robosep Instrument (StemCell
Technologies, Vancouver, Canada, www.stemcell.com); biotinyl-
ated CD31 and CD117 antibodies were added to the kit to
remove endothelial cells and hematopoietic precursors, respec-
tively. The Lin?, CD117?, and CD31?bone marrow mesenchy-
mal/progenitor-enriched stem cells were further characterized by
a seven color staining approach, using Streptavidin-Peridinin
Chlorophyll-a Protein (SAv-PerCP; BD Pharmingen, BD Bio-
sciences, Franklin Lakes, NJ, www.bdbiosciences.com) against all
biotinylated antibodies to stain unwanted cells still present after
depletion, monoclonal phycoerythrin (PE)-Cy7-labeled CD45,
CD31, and Ter119 antibodies, monoclonal (allophycocyanin
[APC]-780)-conjugated anti-CD117, monoclonal FITC-conjugated
anti-Sca1, monoclonal PE-conjugated anti-CD105, monoclonal
APC-conjugated anti-CD49e, and monoclonal Pacific Blue-conju-
gated anti-CD29 (antibodies were purchased by eBiosciences, San
Diego, CA, www.ebioscience.com and BioLegend, San Diego,
CA, www.biolegend.com). Nonspecific staining was excluded
using rat-specific isotype immunoglobulin controls. The analyses
were performed by FACScanto II flow cytometer (Becton Dickin-
son France SAS, Le Pont-De-Claix, France, www.bd.com), using
the FACSDiva software supplied by the manufacturer. At least
5 ? 104events per sample were acquired.
Primary MSC Culture
A fixed midspine segment (?4 cm) from WT (n ¼ 11) and Brtl
(n ¼ 11) mice was dissected, cleaned from soft tissues and
mechanically crushed in Phosphate Buffer Saline (PBS) (Sigma-
Aldrich, St. Louis, MO, http://www.sigmaaldrich.com) containing
0.5% heat-inactivated fetal calf serum (hiFCS; Euroclone, Milan,
Italy, www.euroclonegroup.it), 100 U/ml penicillin, and 100 lg/
invitrogen.com). Cells were passed through a 40 lm nylon mesh
filter (Millipore, Billerica, MA, http://www.millipore.com) and
centrifuged at 450g for 5 minutes at 4?C. Red blood cell lysis
was performed with Red Blood Cell Lysing Buffer (Sigma-
Aldrich) for 5 minutes at RT. Finally, cells were resuspended in
MesenCult medium (StemCell Technologies) supplemented with
2 mM L-glutamine (Sigma-Aldrich), MSC Stimulatory Supple-
ments (StemCell Technologies), and antibiotics as above and
seeded for the different experiments detailed below. The medium
was replaced after 24 hours to eliminate nonadherent cells.
Proliferation was evaluated using the CellTiter 96 AQueous One
Solution Cell Proliferation Assay (Promega, Madison, WI, http://
www.promega.com) according to the manufacturer’s protocol.
Bone marrow MSCs were isolated as described above (WT n ¼
10; Brtl n ¼ 11) and plated at 5 ? 105cells per square centime-
ter in 10 cm Petri dishes. Complete MesenCult medium was
changed twice a week. At confluence, cells were trypsinized and
seeded in triplicate in 96-well plates at 1.5 ? 104cells per square
centimeter in 100 ll of complete MesenCult medium. After 1
and 5 days, 20 ll of MTS solution were added to each well. Af-
ter 4 hours of incubation at 37?C, absorbance at 490 nm was
measured using an ELISA plate reader.
Bone marrow MSCs were isolated as described above (WT n ¼
7; Brtl n ¼ 7) and plated in triplicate at 5 ? 105cells per square
centimeter in 24-well plates. From the third day after plating,
cells were cultured with a-Minimum Essential Medium (MEM)
(Gibco, Invitrogen) with 20% hiFCS and antibiotics. Medium was
Impaired Osteoblastogenesis in Osteogenesis Imperfecta
changed three times a week and cells cultured at 37?C in a humidi-
fied 5% CO2incubator. On day 9 after seeding, cells were fixed
with methanol for 5 minutes and stained with 5% Giemsa (Sigma-
Aldrich). Colonies containing more than 50 cells were counted and
their number was normalized to the DNA content per well.
In Vitro Adipogenic Differentiation
Bone marrow MSCs were isolated and plated as described above
for colony-forming unit-fibroblasts (CFU-F) (WT n ¼ 11; Brtl n
¼ 10). At day 6, the medium was replaced with adipocyte induc-
tive medium: a-MEM with 20% hiFCS, antibiotics, 5 ? 10?8M
dexamethasone (Sigma-Aldrich), 0.5 mM 3-isobutyl-1-methylxan-
thine (IBMX; Sigma-Aldrich), 10 lg/ml insulin (Sigma-Aldrich),
and 125 lM indomethacin (Sigma-Aldrich). At day 12, inductive
medium was replaced with adipocyte maintenance medium: a-
MEM with 20% hiFCS, antibiotics, 5 ? 10?8M dexamethasone,
and 10 lg/ml insulin. The medium was changed three times a
week. After 16 days, cells were fixed with formalin solution 10%
neutral buffered (Sigma Aldrich) for 30 minutes and adipocytes
stained with 3% (wt/vol) oil red O (Sigma-Aldrich) in 60% iso-
propanol. Counterstaining was performed with Gill’s no. 3 hema-
toxylin (Sigma-Aldrich). Only colonies containing more than 20
adipocytes were counted and their number was normalized to the
DNA content per well. Wells containing less than 1 lg of total
DNA were not considered, since associated with impaired cell
Adipocyte Morphometric Measurements
High-resolution light microscopy images of in vitro bone marrow
MSCs differentiated into adipocytes were taken with the DFC480
digital camera and analyzed with LEICA Application Suite v3.00
software (Leica Microsystems GmbH, Wetzlar, Germany, www.le-
ica-microsystems.com). Colony size and number of cells per col-
ony (WT n ¼ 30 colonies; Brtl n ¼ 33 colonies, from four mice
each) were measured on ?4 magnification images. Adipocyte size
(WT n ¼ 1,153 adipocytes from eight mice; Brtl n ¼ 1,007 adipo-
cytes from 11 mice) was measured on ?10 magnification images.
Drop number per cell (n ¼ 55 adipocytes from four mice, for both
WT and Brtl) was evaluated on ?20 magnification images.
In Vitro Osteoblastic Differentiation
Bone marrow MSCs were isolated as described above for CFU-F
(WT n ¼ 7; Brtl n ¼ 7). From day 3, cells were cultured in
osteoblast inductive medium: a-MEM with 20% hiFCS, antibiot-
ics, 5 ? 10?8M dexamethasone, 0.2 mM ascorbic acid (Sigma-
Aldrich), and 10 mM b-glycerophosphate (Sigma-Aldrich). The
medium was changed three times a week. On day 21, cells were
fixed in 10% formalin solution neutral buffered (Sigma-Aldrich)
for 1 hour and Von Kossa staining with 5% silver nitrate (Merck,
Whitehouse Station, NJ, www.merck.com) was performed. Coun-
terstaining was performed with Nuclear Fast Red (Sigma-
Aldrich). The calcified extracellular matrix appeared as black
nodules. Images were acquired with a digital scanner with a
1,200 dpi resolution and analyzed with LEICA Application Suite
v3.00 software. Mineralized spots were manually delimited and
the percentage of mineralized area was calculated on the total
well area. Wells containing less than 1 lg of total DNA were not
considered, since associated with impaired cell growth.
Total RNA was isolated using TriReagent (Sigma-Aldrich)
according to the manufacturer’s protocol from bone marrow
MSCs undifferentiated (WT n ¼ 5; Brtl n ¼ 5), differentiated to-
ward osteoblasts (WT n ¼ 5; Brtl n ¼ 5) and toward adipocytes
(WT n ¼ 9; Brtl n ¼ 11), from bone marrow MSCs differentiated
toward osteoblasts obtained from Btz and placebo-treated mice
(WT n ¼ 7; Brtl n ¼ 4; BrtlþBtz n ¼ 8), from primary calvarial
osteoblasts (at least three independent experiments) isolated as
described in Supporting Information Methods, from fresh bone
marrow from long bones (WT n ¼ 6; Brtl n ¼ 6), and from long
bone diaphysis (WT n ¼ 11; Brtl n ¼ 12). DNase digestion was
performed using the Turbo DNA Free Kit (Ambion, Applied Bio-
systems, Austin, TX, http://www.ambion.com). RNA integrity
was verified using the Bioanalyzer 2100 (Agilent Technologies,
Santa Clara, CA, http://www.agilent.com). Real Time PCR meas-
urements were done on the Mx3000P Stratagene thermocycler
using commercially available TaqMan primers and probes
(Applied Biosystems) and TaqMan Universal PCR Master
Mix (Applied Biosystems). All reactions were performed in
triplicate. Expression levels for Runx2 (Mm01269515_mH),
(Mm00492555_m1), Cathepsin K (CtsK) (Mm00484039_m1),
Pparg2 (Mm00440940_m1), fatty acid binding protein 4 (Fabp4)
(Mm01295675_g1), and Cfd (Mm00442664_m1) were evaluated.
Gapdh (Mm99999915_g1) was used as normalizer. Relative
expression levels were calculated using the DDCt method.
Total proteins were extracted from bone marrow MSCs (WT n ¼
6; Brtl n ¼ 7) differentiated toward osteoblasts using 10 mM Tris/
HCl pH 7.6, 5 mM ethylenediaminetetraacetic acid (EDTA) pH
8.0, 140 mM NaCl, 0.5% NP40 added with protease inhibitors
(130 mM benzamidine, 2 mM N-ethylmalemide (NEM), 5 mM
EDTA, 1 mM Phenylmethylsulfonyl fluoride (PMSF)) and 1.6 mM
Na3VO4. Proteins were quantitated by the RC DC Protein Assay
(Bio-Rad, Hercules, CA, http://www.bio-rad.com); 20 lg (for heat
shock protein 47 [HSP47], autophagy-related gene 7 [ATG7],
Beclin1 [BECN1], and LC3) or 40 lg (for PARP1 and CASP3)
were separated by SDS-PAGE in denaturing conditions using 7.5%
(for HSP47, ATG7, BECN1, and PARP1) or 12% (for LC3 and
CASP3) gels and electro-transferred to a PVDF membrane (Amer-
sham, GE Healthcare, Chalfont St. Giles, U.K., www.gehealthcare.-
com) for 2 hours at 100 V. After washing with 50 mM Tris pH
7.5, 150 mM NaCl, 0.1% Tween 20 (TBS-T), membranes were
incubated 1 hour at RT with 5% dried milk in the same buffer. Pri-
mary antibodies against PARP1, CASP3, ATG7, LC3, BECN1
(Cell Signaling Technology, Danvers, MA, http://www.cellsignal.-
com), and a-tubulin (Santa Cruz Biotechnology, Santa Cruz, CA,
www.scbt.com) were diluted 1:1,000 in TBS-T containing 5%
dried milk, with the exception of a-tubulin (2.5% BSA); anti-
HSP47 antibody (Santa Cruz Biotechnology) was diluted 1:10,000
in TBS-T, 1% dried milk. The primary antibody incubation was
performed overnight at 4?C. The secondary antibody conjugated
with horseradish peroxidase (ECL anti-mouse Peroxidase Labeled,
Amersham, GE Healthcare; donkey anti-rabbit IgG-HRP, Santa
Cruz) was incubated at RT for 1 hour. The signal was detected
with ECL Western Blotting Detection Reagents (Amersham, GE
Healthcare). X-rays were acquired by VersaDoc 3000 (BioRad)
and band intensity was evaluated by QuantityOne software (Bio-
Rad). a-Tubulin was used to normalize protein loading.
Brtl and WT MSCs differentiated toward osteoblasts as described
above were plated 5 ? 105cells per square centimeter in glass
Petri dishes (WillCo-dish; Willcowells, Amsterdam, The Nether-
lands, www.willcowells.com). For HSP47 (Santa Cruz Biotech-
nology) and COL1 (Abcam, Cambridge, U.K., http://www.
abcam.com) costaining, cells were washed with PBS and fixed
for 20 minutes in 4% paraformaldehyde (PFA; Sigma-Aldrich).
Fixed cells were washed with PBS, permeabilized with 0.2% Tri-
ton X-100 for 10 minutes on ice, washed again with PBS and
0.05% Tween, and incubated in blocking buffer containing 5%
BSA, 0.3 M glycine in PBS, and 0.05% Tween for 1 hour at RT.
Cells were then incubated in 1% Bovine Serum Albumin (BSA)
in PBS and 0.05% Tween 20 buffer containing primary antibodies
(1:100) overnight in a humidified chamber at 4?C. After washing,
cells were incubated with 1:100 FITC anti-mouse (Sigma-
Aldrich) (for HSP47) and 1:500 Alexa Fluor-647 anti-rabbit
(Invitrogen) (for COL1) conjugated secondary antibodies in 1%
BSA in PBS and 0.05% Tween 20 buffer for 1 hour in a
Gioia, Panaroni, Besio et al.
humidified chamber at RT. The nuclei were counterstained by
40,6-diamidino-2-phenylindole 0.5 lg/ml (DAPI; Sigma-Aldrich).
Slides were mounted using Mowiol. The sections were examined
by TCS SP2-Leica confocal microscope (Leica). Colocalization
was evaluated on single planes.
In Vivo Treatment with Btz
Five-week-old female mice were treated with 0.3 mg/kg Btz (WT
n ¼ 16; Brtl n ¼ 15) or saline control (placebo) (WT n ¼ 17;
Brtl n ¼ 16) three times a week by i.p. injection. The dose for
antitumor effect in mice is 0.6–1.3 mg/kg. After 3 weeks of
treatment, animals were sacrificed and bone marrow MSCs iso-
lated as described above. CFU-F, CFU-adipocytes (CFU-A), and
were analyzedas described
Bone Histology and Bone Length
For histochemical studies, left tibiae from Btz treated (Brtl ¼ 2)
and untreated (WT n ¼ 4; Brtl n ¼ 6) mice were dissected and
fixed in 4% PFA in PBS, decalcified in 14% EDTA pH 7.1 for
21 days and processed for light microscopy, according to standard
procedure. Ten micrometer sections were stained with H&E and
used for morphometric measurements. Pictures were taken using
a DFC480 digital camera (Leica) connected to a light microscope
(Dialux 20, Leica). The resting, proliferative, and hypertrophic
zones were delineated on the basis of cell morphology, as
reported elsewhere  and measured as described in . The
measurements were performed using the Leica Application Suite
v3.00 image analysis software (Leica). At least 20 vertical meas-
urements per zone per section were performed. An average of six
sections per animal was considered and measurements were per-
formed in the medial zone of the growth plate. Femur length was
measured by using a calipser accurate to 1/20 mm (WT n ¼ 15;
Brtl n ¼ 12; BrtlþBtz n ¼ 9).
Peripheral Quantitative Computed Tomography
Analysis of Tibia and Femur
Long bones (femur and tibia) obtained from Btz (WT n ¼ 13;
Brtl n ¼ 14) and placebo (WT n ¼ 18; Brtl n ¼ 16) treated mice
were analyzed by peripheral quantitative computed tomography
(pQCT). pQCT measurements were performed using a Stratec
Research SAþ pQCT scanner (Stratec Medizintechnik GmbH,
Pforzheim, Germany, www.stratec-med.com) with 70 lm voxel
size and 3 mm/second scan speed. The scans were performed at
the proximal metaphysis of tibiae, distal metaphysis of femura,
and diaphysis of both bones. The scans were analyzed with
pQCT software 6.00B using contour mode two and peel mode
two with a threshold of 350 mg per cubic centimeter for trabecu-
lar and total bone and with a threshold of 600 mg per cubic centi-
meter for cortical bone. For metaphyseal analysis, a distance
from growth plate of 2.0 mm for tibia and 3.0 mm for femur was
Statistical comparisons of stem cell differentiation between two
groups were based on parametric unpaired t test (with Sat-
terthwaite’s correction for degrees of freedom if it needed) and
values were reported as mean 6 SEM. Relationship between adi-
pocytes and osteoblasts was evaluated with nonparametric Spear-
man’s correlation coefficient. Comparisons among three groups
were evaluated using the parametric or the equivalent nonpara-
metric analysis of variance followed by multiple comparison tests
(Bonferroni’s test or Wilcoxon’s test, respectively). A p-value
less than .05 was considered statistically significant (two-side),
except for multiple comparison test where the significant level
was p < .025. The analyses were made using STATA10 release
and SigmaPlot Statistics 11.0.
Characterization of Unstimulated Bone Marrow
Murine MSCs are traditionally isolated from the low-density
mononuclear cell population of bone marrow based on their
selective plastic adherence, a characteristic not shared by the
marrow hematopoietic cells. MSC number is then extrapolated
from the number of colonies, named CFU-F, that they are
able to form in vitro [22–25]. Using this approach to evaluate
the unstimulated MSC population in Brtl and WT mice, the
number of Giemsa-positive CFU-F obtained from plastic ad-
herent bone marrow cells was counted. A normalization
against the total amount of DNA per well was introduced to
correct for plating and/or proliferation differences. The CFU-
F number generated by mutant cells (21.6 6 1.7) and WT
samples (17.7 6 1.9; t ¼ ?1.47; p ¼ .15) was comparable
(Fig. 1A, 1B).
Because the CFU-F assay has documented limitations for
evaluation of heterogeneous cell populations, which contain a
mix of progenitors at different stages of commitment and are
often contaminated by hematopoietic cells [8, 12], a second
strategy using flow cytometry was adopted. An enriched pro-
genitor population was obtained by hematopoietic and endo-
thelial cell depletion and by staining with specific mesenchy-
mal progenitor cell surface markers .
Flow cytometry analysis was performed on freshly iso-
lated bone marrow from Brtl and WT animals to determine
the percentage of cells negative for hematopoietic (Lin,
CD45, CD117, and Ter119) and endothelial (CD31) surface
markers and expressing the stem cell antigens Sca1, CD105,
CD49e, and CD29. Using this criterion, similar values for the
Ter119?, CD31?, CD105þ, CD49eþ, and CD29þ) were again
detected in WT and Brtl animals (WT 0.58 6 0.09; Brtl 0.57
6 0.09; t ¼ 0.04; p ¼ .96) (Fig. 1C). Only 30% of this popu-
lation was positive for Sca1 marker with no difference
approaches supported the similarity of MSC/progenitor cell
population in Brtl and WT marrow in vivo.
(Lin?, CD45?, CD117?,
Because differences in cell proliferation could affect the in
vitro evaluation of cellular differentiation, the growth rate of
cultured MSCs was evaluated. No significant difference was
detected between the growth rates of adherent cells from Brtl
and WT marrow, although slightly higher cell numbers were
observed in mutant MSCs on day 5 (t ¼ ?1.76; p ¼ .09)
(Fig. 1D). To account for this difference, we normalized the
CFU-F and CFU-A values using DNA content per well.
Brtl MSCs have Decreased Osteoblastogenesis
To evaluate the differentiation of Brtl and WT MSCs toward
the osteoblast lineage, cells were cultured in osteogenic induc-
tive medium for 21 days. The distinct mineralized nodules
that stained positive with Von Kossa (Fig. 2A) were digitally
analyzed. The percent mineralized area in Brtl was about half
that of WT cultures (WT 3.92% 6 0.98%; Brtl 1.41% 6
0.51%; t ¼ ?2.27; p ¼ .03) (Fig. 2B).
To determine whether the reduced Brtl mineralized area
was due to a defect in osteoblastogenesis, the expression of
early (Runx2 and Sp7) and late (Col1a1 and Ibsp) osteoblastic
genes was analyzed by real time PCR. Mutant MSCs revealed
lower expression levels after osteogenic differentiation for all
Impaired Osteoblastogenesis in Osteogenesis Imperfecta
the tested markers, compared to WT (Fig. 2C). Interestingly,
lower expression of osteoblast-specific marker genes was as
well detected in nondifferentiated Brtl cells, compared to WT.
Only Runx2 expression was equivalent in nondifferentiated
Brtl and WT cells (Fig. 2D). In cultured calvarial osteoblasts,
no difference was detected in Col1a1 and Ibsp expression
between mutant and WT mice (Supporting Information Fig.
S1 and Supporting Information Methods). These data demon-
strated that the compromised osteoblastogenesis in differenti-
ated Brtl cells reflected a reduction in the number of commit-
ted osteoprogenitors present in the bone marrow population.
The more sensitive and specific transcript assay showed dif-
ferences not detectable by fluorescence-activated cell sorting
Brtl MSCs have Increased Adipogenesis In Vitro
Mutant and WT MSCs were evaluated for their adipogenic
differentiation potential by culturing the bone marrow plastic
adherent cells in adipogenic medium for 16 days. Oil red O
staining demonstrated the intracellular lipid droplets charac-
teristic of adipocyte formation (Fig. 2E). Brtl MSCs generated
nearly twice the number of adipocytic colonies, normalized to
the DNA content per well, generated by the WT samples
(2.3 6 0.3 and 1.3 6 0.1, respectively; t ¼ ?2.70; p ¼ .009)
qPCR was used to compare the expression level of early
(Pparc2) and late (Adipsin/Csd and Fabp4) markers of adipo-
genic differentiation.Surprisingly,giventhe substantial
increase in Brtl adipogenic colonies, these transcripts were all
higher in WT than in mutant cells (Table 1). However, to
account for differences in the contribution of mature adipo-
cytes, we normalized gene expression to the expression of
Fabp4, which is specific for differentiated adipocytes .
The normalized expression level of the early adipocyte
marker Pparc2 in Brtl samples was more than double the WT
level (Table 1), suggesting that the mutant population con-
tained more immature adipocytes. Further morphometric anal-
ysis supported this observation. A marker of adipocyte matu-
ration is the presence of a single or small number of large
triglyceride drops occupying most of the cytosol. The number
of triglyceride drops per adipocyte was indeed significantly
increased in mutant cells (Table 1, Supporting Information
Fig. S2), as were adipocyte size and adipocyte number per
colony. No difference was detected in adipocyte colony size
and no relationship was identified between increased CFU-A
number and decreased mineralized area in mutant samples
(Spearman’s Rho ¼ 0.08; p ¼ .75). Interestingly oil red O
staining of bone cryosections at most revealed a slight
decrease of adiposity in bone marrow of Brtl mice (Support-
Autophagy Activation in Mutant MSCs Differenti-
ated Toward Osteoblasts
Brtl osteoblasts are characterized by partial intracellular reten-
tion of mutant collagen causing endoplasmic reticulum (ER)
(B): CFU-F quantitation. (C): Examples of fluorescence-activated cell sorting analysis of mesenchymal/progenitor stem cells. Top: Lin?, CD31?,
CD117?, CD45?, and Ter119?population. Middle: among the previous population CD29þand CD49eþcells. Bottom: among the previous
population, CD105þcells in red, control unstained cells in white. (D): Proliferation analysis. ^ Brtl MSCs, h WT MSCs. Abbreviation: WT,
Characterization of bone marrow mesenchymal stem cells (MSCs). (A): Giemsa stained colony-forming unit-fibroblasts (CFU-F).
Gioia, Panaroni, Besio et al.
stress activation . Recently, it has been reported that mu-
tant collagen with triple helical defects can activate autophagy
. Impairment of autophagy has also been related to com-
promised adipocyte differentiation in knockout murine models
[29, 30]. Thus, we investigated the presence of ER stress in
mutant MSCs cultured in osteogenic medium. The expression
of HSP47, a well-known collagen-specific ER chaperone ,
was significantly increased in Brtl cells (Fig. 3A, 3B). Confo-
cal analysis of the intracellular distribution of HSP47 by im-
munofluorescence showed that HSP47 and type I procollagen
colocalized in the ER in both WT and mutant differentiated
MSCs, but in mutant cells, type I procollagen appeared more
prone to accumulate in aggregates within the cells (Fig. 3C).
Finally, we investigated several characteristic autophagic
players: BECN1, ATG7, and LC3-II, by Western blot. Their
expression was increased in Brtl MSCs following in vitro
osteoblastic differentiation (Fig. 4A, 4B). Further support for
autophagy activation in differentiated Brtl osteoblastic cells is
provided by the increase of transcript for the lysosomal
enzyme CtsK, specifically involved in collagen degradation
Table 1. Quantitative polymerase chain reaction on RNA
extracted from mesenchymal stem cells differentiated into
adipocytes and adipocytic morphological parameters
1.51 6 0.44
2.61 6 0.73
1.97 6 0.36
0.89 6 0.23
1.58 6 0.41
5.41 6 0.54
0.42 6 0.08
0.80 6 0.42
0.44 6 0.16
2.15 6 0.65
1.70 6 0.60
6.44 6 0.74
218.23 6 44.16 486.90 6 81.26.006
3867.49 6 88.41 4181.57 6 114.84.03
43.09 6 2.66 60.32 6 4.28
Abbreviation: WT, wild type.
(C, D): Expression level of early (Runx2 and Sp7) and late (Col1a1 and Ibsp) osteoblast-specific genes in (C) MSCs induced toward osteoblastic
differentiation and in (D) undifferentiated MSCs. (E): CFU-A oil red O staining. (F): CFU-A quantitation. *, p < .01;#, p < .05. Abbreviations:
CFU-A, colony-forming unit-adipogenic; CFU-adipocytes CFU-O, CFU-osteoblasts; WT, wild type.
Mesenchymal stem cell (MSC) osteoblastic and adipocytic differentiation. (A): CFU-O Von Kossa staining. (B): CFU-O quantitation.
Impaired Osteoblastogenesis in Osteogenesis Imperfecta
(Fig. 4C, Supporting Information Methods and Supporting In-
formation Fig. S4).
Autophagy is often linked to apoptosis so we also checked
the protein expression level of the classic apoptotic markers
CASP3 and PARP1 by Western blot. No difference was
detected between mutant MSCs and WT following osteoblas-
tic differentiation (Supporting Information Fig. S5).
Btz Stimulates Osteoblastogenesis in Brtl MSCs
The proteasomal inhibitor Btz, widely used for multiple my-
eloma treatment, has recently been demonstrated to stimulate
osteogenic differentiation in murine and human MSCs, at
least in part by stabilizing RUNX2 protein . To determine
whether mutant MSCs respond to Btz in vitro, MSC cultures
sis of the expression of the ER stress-related marker HSP47 in mesenchymal stem cells (MSCs) differentiated toward the osteoblastic lineage. *,
p < .05. (C): Double fluorescence using antibodies against HSP47 (green) and type I collagen (red) in MSCs differentiated toward osteoblast lin-
eage. Green bars represent some of the selected points of colocalization shown below in the graphs. ?40 magnification. Abbreviations: HSP47,
heat shock protein 47; WT, wild type.
Mutant collagen retention in the endoplasmic reticulum (ER). (A): Representative Western blot of HSP47. (B): Densitometric analy-
Gioia, Panaroni, Besio et al.
were established from Brtl and WT mice treated for three
weeks by i.p. injection of Btz. No difference was detected in
the number of CFU-F between MSCs from treated and
untreated animals (WT 19.46 6 1.69; Brtl 20.20 6 2.06; Brtl
þ Btz 19.40 6 2.06; p > .90). Also, when MSCs were cul-
tured under adipogenic conditions, there was no difference in
the number of adipocyte colonies (WT 1.57 6 0.18; Brtl 2.21
6 0.22; Brtl þ Btz 2.22 6 0.24; F ¼ 2.84; p ¼ .06). Interest-
ingly, under osteogenic conditions,
increased the mineralized area in vitro, normalizing the mu-
tant samples to WT level (WT 3.53% 6 0.44%; Brtl 2.02%
6 0.30%; Brtl þ Btz 4.40% 6 0.66%; p < .001) (Fig. 5A).
Similarly, it increased the expression level of the early (Sp7,
Brtl 0.56 6 0.15; Brtl þ Btz 0.69 6 0.13) and late (Col1a1,
Brtl 0.36 6 0.23; Brtl þ Btz 0.64 6 0.23) osteoblastic genes.
Considering that this effect was noted on cultured cells from
animals injected for 3 weeks with drug or placebo, a long-
term effect of the drug on cellular plasticity is suggested.
Btz Increases Total, Trabecular, and Cortical Den-
sity at Femoral and Tibial Metaphyses
pQCT was performed on femur and tibia of treated and
untreated WT and Brtl mice. At both distal femoral and prox-
imal tibial metaphyses, total, trabecular, and cortical density
were increased in mutant drug versus placebo-treated mice,
reaching values similar to WT mice, although no significant
difference was detected between mutant treated and untreated
animals for all the reported bone parameters. Thus, the signifi-
cant difference present between Brtl and WT untreated ani-
mals was eliminated (Fig. 5B, 5C). No differences were
detected at the diaphyses for all the tested parameters (Sup-
porting Information Table S1).
Examination of histological sections of bone metaphyses
from untreated WT and Brtl and treated Brtl mice did not
reveal any evident effect of the treatment in the trabecular
compartment. A higher growth plate was present in untreated
Brtl (176.4 6 4.2 lm) with respect to WT mice (153.0 6 2.0
lm; t ¼ ?4.34; p < .001), due to an increase of the hyper-
trophic zone in mutant animals (Brtl 80.7 6 1.9 lm; WT
61.3 6 2.3 lm; t ¼ ?6.59; p < .001). Interestingly, this latter
difference was rescued by the drug treatment (Brtl þ Btz 68.2
6 2.4 lm; t ¼ 1.90; p ¼ .07) (Fig. 5D). No effect of Btz
administration on femur length was anyway detected (Brtl
13.06 6 0.15 mm, Brtl þ Btz 13.30 6 0.24 mm; t ¼ 0.91;
p ¼ .4; WT 14.30 6 0.17 mm).
The relationship between osteoblast and adipocyte differentia-
tion in adult bone marrow is still subject of debate. A recipro-
cal relationship between osteogenesis and adipogenesis was
attributed to the selective differentiation of multipotent MSCs
into either osteoblasts or adipocytes at the expense of the al-
ternative lineage . More recently, the existence of inde-
pendent preosteoblastic and preadipocytic cell populations
within the bone marrow has been demonstrated, revealing a
more complex differentiation picture, in which the bone mar-
row environment contains, together with multipotent stem
cells, also clonal subpopulations of cells already committed to
specific lineages that may undergo independent changes dur-
ing aging or bone diseases .
An imbalance between osteoblast and adipocyte differen-
tiation from a common adult bone marrow precursor was
associated with the bone loss typical of aging, steroid use,
hypercortisolism, and antidiabetic glitazone treatment .
Changes in both soluble factors and/or extracellular microen-
vironment appear to influence skewing toward adipocytic line-
age versus osteoblast maturation in osteoporotic patients .
OI is a heritable ‘‘brittle bone disease,’’ whose main fea-
tures are bone fragility and skeletal deformity accompanied
by a severe osteoporotic phenotype. Thus, it is tempting to
speculate that an abnormal relationship between marrow
osteoblastogenesis and adipogenesis could contribute to OI
bone phenotype. To investigate this possibility, we used a mu-
rine model for OI, the Brtl mouse . Adult bone marrow
stem cells from mutant mice showed a reduced capacity to
differentiate toward osteoblasts. After osteoblastogenic stimu-
lation in culture, Brtl MSCs produced a significantly reduced
mineralization area, associated with decreased expression of
early (Runx2 and Sp7) and late (Col1a1 and Ibsp) osteoblast-
specific genes. Conversely, Brtl cells have increased ability to
differentiate toward adipogenic lineage (AdL), since the num-
ber of CFU-A obtained in culture from Brtl MSCs was signif-
icantly higher than in WT littermates. However, Spearman’s
analysis did not identify a correlation between decreased
osteoblast and increased adipocyte maturation in Brtl mice.
This is consistent with the hypothesis that not all preosteo-
blastic cells retain plasticity to transdifferentiate to adipocytic
cells. Thus, it is feasible that, depending on the stage of dif-
ferentiation impairment, mutant preosteoblasts can either
become fat cell precursors or undergo to a different fate.
Interestingly, the early (Pparc2) and late (Adipsin/Cfd)
adipogenic genes were more highly expressed in WT MSCs
differentiated in vitro toward adipogenic lineage compared to
Brtl, contrasting with the histological findings. However,
when these data were normalized for the expression of
Fabp4, a gene expressed only in mature adipocytes, to correct
for the proportion of mature adipocytes in the culture, Pparc2
was found to be increased in Brtl samples compared to WT.
This observation suggests that a greater number of adipocyte
precursor cells were present in Brtl MSCs in culture. Two dif-
ferent interpretations of this result are possible: there may be
an impairment in adipogenic differentiation of mutant cells,
sentative Western blots. (B): Densitometric analysis of the expression
levels of autophagic markers in mesenchymal stem cells (MSCs) dif-
ferentiated toward osteoblasts. (C): Expression analysis of CtsK by
qPCR in MSCs differentiated toward osteoblasts. *, p < .05. Abbrevi-
ations: ATG7, autophagy-related gene 7; BECN1, Beclin1; CtsK, Ca-
thepsin K; LC3-II, microtubule-associated protein1 light chain 3;
qPCR, Quantitative Polymerase Chain Reaction.
Expression analysis of autophagy markers. (A): Repre-
Impaired Osteoblastogenesis in Osteogenesis Imperfecta
or, alternatively, a higher number of mutant preosteoblastic
cells may be directed toward AdL, due to impaired osteoblast
maturation. This latter hypothesis would explain both the
increased preadipocytic cells in mutant culture as well as the
higher number of CFU-A obtained in the in vitro MSC differ-
entiation assay. This hypothesis is also supported by the mor-
phometric analysis: adipocytes derived from mutant MSCs
showed a higher number of cytoplasmic lipid drops than WT
adipocytes, suggesting an immature status. On the other hand,
bone marrow adiposity in vivo was not increased in Brtl mice
as demonstrated by oil red O staining of bone cryosections,
thus suggesting that the higher number of preadipocytes
detected in vitro in Brtl mice could migrate from the marrow
environment to the peripheral adipose tissue [35, 36]. This
supports the observation of a thicker subcutaneous fat pad in
Brtl mice (Marini JC and Forlino A personal communication).
Brtl and WT bone marrow MSCs generated similar num-
bers of CFU-F in culture and have similar cell surface
markers. However, a detailed molecular analysis of specific
osteogenic genes revealed that mutant MSC population con-
tained a reduced number of cells expressing the early and late
osteogenic genes, suggesting in vivo distinctions in the mar-
row microenvironment. What is the cause of the impaired
MSC osteoblastic differentiation? We know that osteoblasts
from Brtl mice are subjected to ER stress, caused by selective
retention and intracellular degradation of mutant collagen,
resulting in ER enlargement and cellular malfunction .
Since the major protein synthesized and secreted by osteo-
blasts is type I collagen, it is possible that increased collagen
synthesis during differentiation may compromise their matura-
tion from marrow precursors by triggering increased collagen
degradation. Ishida et al. reported that mutant type I collagen
that assembled in triple helices inside the ER formed insolu-
ble aggregates that were eliminated by autophagy . Our
examination of autophagic-related markers in adult MSCs fol-
lowing in vitro osteo-induction revealed an upregulation of
the autophagic proteins BECN1, ATG7, and LC3-II in mutant
stem cells by Western blot as well as an increase of the lyso-
somal collagenolytic enzyme CtsK expression by real time
PCR. This is consistent with recent proposals of autophagy as
a mechanism through which cells regulate proliferation, death,
and differentiation during both embryogenesis and postnatal
Autophagy has also been linked to adipogenic differentia-
tion, since its impairment in Atg5 and Atg7 knockout mice
compromises fat cell maturation [29, 30]. Furthermore,
autophagy also favors adipogenic differentiation by affecting
cellular shape [38–40]. Thus, our hypothesis is that the ER
stress caused by intracellular retention of mutant collagen
triggers activation of autophagy in order to eliminate the
abnormal collagen, causing preosteoblasts transdifferentiation
into preadipocytic cells (Fig. 6).
It is plausible that abnormal extracellular matrix can act
in concert with intracellular events to influence and modulate
stem cell differentiation, as already described in decorin and
biglycan knockout mice. Mice lacking biglycan showed age-
related osteoporosis due to a reduced capacity to produce
bone marrow stromal cells . Analysis of decorin/biglycan
double knockout mice revealed that these two proteoglycans
are relevant to guarantee the correct number of osteoprogeni-
tors , thus underscoring the importance of the microenvir-
onment in influencing the fate of adult stem cells. In Brtl
mice, mutant collagen is secreted and incorporated into the
extracellular matrix, whose structure and properties are thus
compromised, justifying a further detailed investigation of the
extracellular matrix (ECM) effect on MSC differentiation.
The discovery that adult stem cell fate can be influenced
by intracellular and microenvironment factors represents a
new and exciting tool for regenerative medicine. The identifi-
cation of novel substances that influence stem cell fate toward
specific lineages is an active research field. The possibility of
‘‘rediscovering’’ pharmacological molecules already FDA
approved for ‘‘novel health application’’ is particularly appeal-
ing. In this arena, Btz caught our attention. The clinically
available proteasomal inhibitor Btz is widely administrated to
treat multiple myeloma and relapse/refractory mantle cell
lymphoma . By reversibly binding to the chymotrypsin-
like activity of the 20S proteasome, it inhibits 20S proteolytic
function. Btz induces apoptosis in tumor cells via the intrinsic
stem cells (MSCs) and on in vivo bone properties. (A): Quantitation
of CFU-O from MSCs of untreated WT (?) and treated (þ) and
untreated (?) Brtl mice. (B): Representative peripheral quantitative
computed tomography metaphyseal scans. The right panel shows the
scale of volumetric mineral density. (C): Quantitation of bone den-
sity.*, p < .05. (D): Representative hematoxylin and eosin stained
sections of tibia metaphysis (top, ?2.5 magnification) and growth
plate (bottom, ?25 magnification). Abbreviations: CFU-O, colony-
forming unit-osteoblasts; WT, wild type.
Bortezomib effect on cultured bone marrow mesenchymal
Gioia, Panaroni, Besio et al.
mitochondrial, extrinsic death receptor pathway and ER stress
response pathways . It was demonstrated that Btz has ana-
bolic effect on bone by enhancing specifically osteoblasts ac-
tivity and differentiation . Improvement of bone properties
[46, 47], decreased expression of Wnt inhibitors such as
DKK-1  and sclerostin  as well as increased level of
osteoblast activity-specific markers  were reported in
humans after Btz treatment. More interesting for us, Mukher-
jee et al. have recently demonstrated that, at subtherapeutic
doses, Btz is effective in stimulating in vitro and in vivo adult
stem cells osteoblastic differentiation in both mice and
humans . Furthermore, the inhibition of proteasomal ac-
tivity stabilized p57KIP2, an important factor in osteoblasts
differentiation , and an increase in osteoblast number was
also reported following Btz administration .
These observations prompted our attempt to rescue the
impaired differentiation of Brtl MSCs toward osteoblasts
using Btz. Indeed, following Btz treatment, we observed a
significant increase in mineralization nodule formation by
Von Kossa staining in mutant MSCs, normalizing to WT val-
ues. We did not detect post-treatment changes in CFU-F and
CFU-A; in contrast to the report of Mukherjee et al. CFU-A
in treated and placebo Brtl mice were comparable to each
other, and statistically higher than WT. This could be due to
the different murine strains used in the two studies, since we
also found no difference between WT mice treated with Btz
or placebo (data not shown). Interestingly, pQCT analysis
revealed an amelioration of bone properties in Brtl metaphy-
ses, where cellular proliferation/differentiation is more active.
Btz-treated Brtl mice had total, trabecular, and cortical density
values comparable to WT. Furthermore, histological examina-
tion of the growth plate revealed that in untreated Brtl mice,
the hypertrophic zone was significantly higher than in WT,
similar data were reported for both oim/oim mice and OI
patient [21, 53]. Interestingly, in Brtl mice, Btz treatment
brought the hypertrophic zone height close to WT value.
The described defect in mutant osteoblast maturation as
potential contributing factor to the OI phenotype in Brtl mice
and the possibility to enhance osteoblastogenesis with specific
drugs are in favor of the use of cellular therapy in the treat-
ment of this disease. MSC transplantation, by providing nor-
mal precursor cells, has the potentiality to replace defective
osteoblasts with healthy ones able to properly produce normal
collagen. Indeed, few attempts of cell therapy had been per-
formed both in OI patients and in OI murine models both pre-
natal and after birth with promising results associated to bone
properties improvement [19, 54–56]. Unfortunately, all trials
were characterized by low donor cell engraftment independ-
ently from sources (total bone marrow, bone marrow stromal
cells, MSCs, hematopoietic stem cells, and fetal blood stem
cells), thus the possibility to enhance with specific molecules
MSC differentiation toward osteoblasts in vivo after stem cell
transplantation represents a novel approach and an appealing
tool for OI cellular treatment.
We identified abnormalities of adult stem cells differentiation
in a murine model for classic OI, and we associated ER stress
leading to autophagy with this differentiation defect. Abnor-
mal stem cell differentiation was also revealed as a novel
pathway for treatment of OI, which could be used in combi-
nation with cellular therapeutic approaches. The effectiveness
of novel drugs with reduced side effects, such as the second
generation of proteasomal inhibitors, opens an attractive new
research area for the field.
We thank Dr. Lupi A. for initial set up of MSCs culture and Dr.
Vaghi P. (Centro Grandi Strumenti, University of Pavia, Italy)
for technical assistance for Confocal Microscopy. This workwas
ferentiation, probably due to collagen retention and endoplasmic reticulum stress responsible for autophagy upregulation, either undergo adipo-
genic lineage or undergo a different unknown fate. Abbreviations: MSC, mesenchymal stem cell; WT, wild type.
Model proposed for MSC impairment in osteogenic differentiation. We propose that preosteoblasts that are not able to complete dif-
Impaired Osteoblastogenesis in Osteogenesis Imperfecta
supported by PRIN 2008 (2008XA48SC) to A.F., by PRIN
200999KRFW-002 to P.V., by Fondazione Cariplo 2007 to A.F.
and P.V., by Progetto Premiale CNR Invecchiamento to P.V.,
and by Progetto Regione Lombardia (cod. SAL/45) ‘‘Dalla sci-
enza dei materiali alla medicina molecolare’’ to A.F. G.V. was
supportedby CollegioGhislieri,Pavia, Italy, fellowship.
DISCLOSURE OF POTENTIAL
CONFLICTS OF INTEREST
The authors indicate no conflict of interest.
1 Forlino A, Cabral WA, Barnes AM et al. New perspectives on osteo-
genesis imperfecta. Nat Rev Endocrinol 2011;7:540–557.
Marini JC, Forlino A, Cabral WA et al. Consortium for osteogenesis
imperfecta mutations in the helical domain of type I collagen: Regions
rich in lethal mutations align with collagen binding sites for integrins
and proteoglycans. Hum Mutat 2007;28:209–221.
Forlino A, Porter FD, Lee EJ et al. Use of the Cre/lox recombination
system to develop a non-lethal knock-in murine model for osteogene-
sis imperfecta with an alpha1(I) G349C substitution. Variability in
phenotype in BrtlIV Mice. J Biol Chem 1999;274:37923–37931.
Sillence DO, Senn A, Danks DM. Genetic heterogeneity in osteogene-
sis imperfecta. J Med Genet 1979;16:101–116.
Kozloff KM, Carden A, Bergwitz C et al. Brittle IV mouse model for
osteogenesis imperfecta IV demonstrates postpubertal adaptations to
improve whole bone strength. J Bone Miner Res 2004;19:614–622.
Uveges TE, Collin-Osdoby P, Cabral WA et al. Cellular mechanism
of decreased bone in Brtl mouse model of OI: Imbalance of decreased
osteoblast function and increased osteoclasts and their precursors. J
Bone Miner Res 2008;23:1983–1994.
Forlino A, Tani C, Rossi A et al. Differential expression of both
extracellular and intracellular proteins is involved in the lethal or non-
lethal phenotypic variation of BrtlIV, a murine model for osteogenesis
imperfecta. Proteomics 2007;7:1877–1891.
Bellantuono I, Aldahmash A, Kassem M. Aging of marrow stromal
(skeletal) stem cells and their contribution to age-related bone loss.
Biochim Biophys Acta 2009;1792:364–370.
Gimble JM, Zvonic S, Floyd ZE et al. Playing with bone and fat. J
Cell Biochem 2006;98:251–266.
10 Taipaleenmaki H, Abdallah BM, AlDahmash A et al. Wnt signalling
mediates the cross-talk between bone marrow derived pre-adipocytic
and pre-osteoblastic cell populations. Exp Cell Res 2011;317:
11 Mendez-Ferrer S, Michurina TV, Ferraro F et al. Mesenchymal and
haematopoietic stem cells form a unique bone marrow niche. Nature
12 Post S, Abdallah BM, Bentzon JF et al. Demonstration of the presence
of independent pre-osteoblastic and pre-adipocytic cell populations in
bone marrow-derived mesenchymal stem cells. Bone 2008;43:32–39.
13 Tornvig L, Mosekilde LI, Justesen J et al. Troglitazone treatment
increases bone marrow adipose tissue volume but does not affect tra-
becular bone volume in mice. Calcif Tissue Int 2001;69:46–50.
14 Harslof T, Wamberg L, Moller L et al. Rosiglitazone decreases bone
mass and bone marrow fat. J Clin Endocrinol Metab 2011;96:
15 Xu Y, Shi Y, Ding S. A chemical approach to stem-cell biology and
regenerative medicine. Nature 2008;453:338–344.
16 Chen D, Frezza M, Schmitt S et al. Bortezomib as the first proteasome
inhibitor anticancer drug: Current status and future perspectives. Curr
Cancer Drug Targets 2011;11:239–253.
17 Mukherjee S, Raje N, Schoonmaker JA et al. Pharmacologic targeting
of a stem/progenitor population in vivo is associated with enhanced
bone regeneration in mice. J Clin Invest 2008;118:491–504.
18 Uveges TE, Kozloff KM, Ty JM et al. Alendronate treatment of the
brtl osteogenesis imperfecta mouse improves femoral geometry and
load response before fracture but decreases predicted material proper-
ties and has detrimental effects on osteoblasts and bone formation. J
Bone Miner Res 2009;24:849–859.
19 Panaroni C, Gioia R, Lupi A et al. In utero transplantation of adult
bone marrow decreases perinatal lethality and rescues the bone pheno-
type in the knockin murine model for classical, dominant osteogenesis
imperfecta. Blood 2009;114:459–468.
20 Vanky P, Brockstedt U, Hjerpe A et al. Kinetic studies on epiphyseal
growth cartilage in the normal mouse. Bone 1998;22:331–339.
21 Evans KD, Lau ST, Oberbauer AM et al. Alendronate affects long
bone length and growth plate morphology in the oim mouse model for
Osteogenesis Imperfecta. Bone 2003;32:268–274.
22 Sun S, Guo Z, Xiao X et al. Isolation of mouse marrow mesenchymal
progenitors by a novel and reliable method. Stem Cells 2003;21:
23 Anjos-Afonso F, Siapati EK, Bonnet D. In vivo contribution of murine
mesenchymal stem cells into multiple cell-types under minimal dam-
age conditions. J Cell Sci 2004;117:5655–5664.
24 Quarto N, Longaker MT. FGF-2 inhibits osteogenesis in mouse adi-
pose tissue-derived stromal cells and sustains their proliferative and
osteogenic potential state. Tissue Eng 2006;12:1405–1418.
25 Meirelles Lda S, Nardi NB. Murine marrow-derived mesenchymal
stem cell: Isolation, in vitro expansion, and characterization. Br J Hae-
26 Krings A, Rahman S, Huang S et al. Bone marrow fat has brown adi-
pose tissue characteristics, which are attenuated with aging and diabe-
tes. Bone 2012;50:546–552.
27 Forlino A, Kuznetsova NV, Marini JC et al. Selective retention and
degradation of molecules with a single mutant alpha1(I) chain in the
Brtl IV mouse model of OI. Matrix Biol 2007;26:604–614.
28 Ishida Y, Yamamoto A, Kitamura A et al. Autophagic elimination of
misfolded procollagen aggregates in the endoplasmic reticulum as a
means of cell protection. Mol Biol Cell 2009;20:2744–2754.
29 Baerga R, Zhang Y, Chen PH et al. Targeted deletion of autophagy-
related 5 (atg5) impairs adipogenesis in a cellular model and in mice.
30 Zhang Y, Goldman S, Baerga R et al. Adipose-specific deletion of
autophagy-related gene 7 (atg7) in mice reveals a role in adipogenesis.
Proc Natl Acad Sci USA 2009;106:19860–19865.
31 Ishida Y, Nagata K. Hsp47 as a collagen-specific molecular chaper-
one. Methods Enzymol 2011;499:167–182.
32 Kassem M, Abdallah BM, Saeed H. Osteoblastic cells: Differentiation
and trans-differentiation. Arch Biochem Biophys 2008;473:183–187.
33 Abdallah BM, Kassem M. New factors controlling the balance
between osteoblastogenesis and adipogenesis. Bone 2012;50:540–545.
34 Savopoulos C, Dokos C, Kaiafa G et al. Adipogenesis and osteoblas-
togenesis: Trans-differentiation in the pathophysiology of bone disor-
ders. Hippokratia 2011;15:18–21.
35 Crossno JT, Jr., Majka SM, Grazia T et al. Rosiglitazone promotes de-
velopment of a novel adipocyte population from bone marrow-derived
circulating progenitor cells. J Clin Invest 2006;116:3220–3228.
36 Hausman GJ, Hausman DB. Search for the preadipocyte progenitor
cell. J Clin Invest 2006;116:3103–3106.
37 Cecconi F, Levine B. The role of autophagy in mammalian develop-
ment: Cell makeover rather than cell death. Dev Cell 2008;15:
38 Kilian KA, Bugarija B, Lahn BT et al. Geometric cues for directing
the differentiation of mesenchymal stem cells. Proc Natl Acad Sci
39 Bamber BA, Rowland AM. Shaping cellular form and function by
autophagy. Autophagy 2006;2:247–249.
40 Singh R, Xiang Y, Wang Y et al. Autophagy regulates adipose mass
and differentiation in mice. J Clin Invest 2009;119:3329–3339.
41 Chen XD, Shi S, Xu T et al. Age-related osteoporosis in biglycan-de-
ficient mice is related to defects in bone marrow stromal cells. J Bone
Miner Res 2002;17:331–340.
42 Bi Y, Stuelten CH, Kilts T et al. Extracellular matrix proteoglycans
control the fate of bone marrow stromal cells. J Biol Chem 2005;280:
43 Genin E, Reboud-Ravaux M, Vidal J. Proteasome inhibitors: Recent
advances and new perspectives in medicinal chemistry. Curr Top Med
44 Adams J. The development of proteasome inhibitors as anticancer
drugs. Cancer Cell 2004;5:417–421.
45 Zangari M, Terpos E, Zhan F et al. Impact of bortezomib on bone
health in myeloma: A review of current evidence. Cancer Treat Rev
2012. DOI: 10.1016/j.ctvr.2011.12.007.
46 Zangari M, Yaccoby S, Pappas L et al. A prospective evaluation of
the biochemical, metabolic, hormonal and structural bone changes
associated with bortezomib response in multiple myeloma patients.
47 Terpos E, Christoulas D, Kokkoris P et al. Increased bone mineral
density in a subset of patients with relapsed multiple myeloma who
received the combination of bortezomib, dexamethasone and zole-
dronic acid. Ann Oncol 2010;21:1561–1562.
48 Heider U, Kaiser M, Mieth M et al. Serum concentrations of DKK-1
decrease in patients with multiple myeloma responding to anti-my-
eloma treatment. Eur J Haematol 2009;82:31–38.
Gioia, Panaroni, Besio et al.
49 Terpos E, Christoulas D, Katodritou E et al. Elevated circulating scle- Download full-text
rostin correlates with advanced disease features and abnormal bone
remodeling in symptomatic myeloma: Reduction post-bortezomib
monotherapy. Int J Cancer 2011;2011:27342. DOI: 10.1002/ijc27342.
50 Ozaki S, Tanaka O, Fujii S et al. Therapy with bortezomib plus dexa-
methasone induces osteoblast activation in responsive patients with
multiple myeloma. Int J Hematol 2007;86:180–185.
51 Kim M, Nakamoto T, Nishimori S et al. A new ubiquitin ligase
involved in p57KIP2 proteolysis regulates osteoblast cell differentia-
tion. EMBO Rep 2008;9:878–884.
52 Giuliani N, Morandi F, Tagliaferri S et al. The proteasome inhibitor
bortezomib affects osteoblast differentiation in vitro and in vivo in
multiple myeloma patients. Blood 2007;110:334–338.
53 Sanguinetti C, Greco F, De Palma L et al. Morphological changes in
growth-plate cartilage in osteogenesis imperfecta. J Bone Joint Surg
54 Horwitz EM, Prockop DJ, Gordon PL et al. Clinical responses to bone
marrow transplantation in children with severe osteogenesis imper-
fecta. Blood 2001;97:1227–1231.
55 Horwitz EM, Gordon PL, Koo WK et al. Isolated allogeneic bone
marrow-derived mesenchymal cells engraft and stimulate growth in
children with osteogenesis imperfecta: Implications for cell therapy of
bone. Proc Natl Acad Sci USA 2002;99:8932–8937.
56 Vanleene M, Saldanha Z, Cloyd KL et al. Transplantation of human fetal
blood stem cells in the osteogenesis imperfecta mouse leads to improve-
ment in multiscale tissue properties. Blood 2011;117:1053–1060.
Impaired Osteoblastogenesis in Osteogenesis Imperfecta