PPARgamma inhibits osteogenesis via the down-regulation of the expression of COX-2 and iNOS in rats.

Abstract

Peroxisome proliferator-activated receptor gamma (PPARgamma), a ligand-activated transcription factor, is considered as an anti-osteoblastic factor associated with adiposity and the elderly osteoporosis due to a defect in osteoblastogenesis. We have found that oral administration of PPARgamma activator rosiglitazone decreased tibia BMD and serum ALP but left serum calcium and osteoclast marker C-terminal telopeptide unaffected. In addition, we examined the inhibitory mechanisms of PPARgamma on the bone formation by using PPARgamma activators ciglitazone and 15-deoxy-Delta(12,14)-prostaglandin-J2 (15d-PGJ2). Our data indicated that PPARgamma ligands decreased both mineralized bone nodules and alkaline phosphatase (ALP) activities in cultured primary osteoblasts. Reverse transcription polymerase chain reaction (RT-PCR) showed that the expression of bone morphogenetic protein-2 (BMP-2) and osteocalcin (OCN) was inhibited by ciglitizone and 15d-PGJ2. Furthermore, PPARgamma ligands inhibited NF-kappaB associated downstream COX-2 and iNOS osteogenic signaling. The ultrasound (US)-induced elevation of COX-2 and iNOS expression and nitric oxide (NO) production were attenuated in the presence of PPARgamma ligands. Furthermore, local administration of PPARgamma ligands into the metaphysis of rat tibia decreased the bone volume in secondary spongiosa. These results suggest that the activation of PPARgamma inhibits osteoblastic differentiation and the expression of several anabolic mediators involved in bone formation. These data may reflect osteoporosis and less bone formation in the aging people and patients treated with thiazolidinediones.

Full text

Page 1

PPARγ inhibits osteogenesis via the down-regulation of the
expression of COX-2 and iNOS in rats
Tzu-Hung Lina,1, Rong-Sen Yangb,1, Chih-Hsin Tangc, Chih-Peng Lind, Wen-Mei Fua,⁎
aDepartment of Pharmacology, College of Medicine, National Taiwan University, Taipei, Taiwan
bDepartment of Orthopaedics, College of Medicine, National Taiwan University, Taipei, Taiwan
cDepartment of Pharmacology, College of Medicine, China Medical University, Taichung, Taiwan
dDepartment of Anesthesiology, National Taiwan University, Taipei, Taiwan
Received 13 January 2007; revised 24 May 2007; accepted 11 June 2007
Available online 4 July 2007
Abstract
Peroxisome proliferator-activated receptor gamma (PPARγ), a ligand-activated transcription factor, is considered as an anti-osteoblastic factor
associated with adiposity and the elderly osteoporosis due to a defect in osteoblastogenesis. We have found that oral administration of PPARγ
activator rosiglitazone decreased tibia BMD and serum ALP but left serum calcium and osteoclast marker C-terminal telopeptide unaffected. In
addition, we examined the inhibitory mechanisms of PPARγ on the bone formation by using PPARγ activators ciglitazone and 15-deoxy-Δ12,14-
prostaglandin-J2 (15d-PGJ2). Our data indicated that PPARγ ligands decreased both mineralized bone nodules and alkaline phosphatase (ALP)
activities in cultured primary osteoblasts. Reverse transcription polymerase chain reaction (RT-PCR) showed that the expression of bone
morphogeneticprotein-2(BMP-2)andosteocalcin(OCN)wasinhibitedbyciglitizoneand15d-PGJ2.Furthermore,PPARγligandsinhibitedNF-κB
associated downstream COX-2 and iNOS osteogenic signaling. The ultrasound (US)-induced elevation of COX-2 and iNOS expression and nitric
oxide(NO)productionwereattenuatedinthepresenceofPPARγligands.Furthermore,localadministrationofPPARγligandsintothemetaphysisof
rat tibia decreased the bone volume in secondary spongiosa. These results suggest that the activation of PPARγ inhibits osteoblastic differentiation
andtheexpressionofseveralanabolicmediatorsinvolvedinboneformation.Thesedatamayreflectosteoporosisandlessboneformationintheaging
people and patients treated with thiazolidinediones.
© 2007 Elsevier Inc. All rights reserved.
Keywords: PPARγ; Osteoblast maturation; COX-2; Osteocalcin; iNOS
Introduction
Peroxisome proliferator-activated receptor gamma (PPARγ),
a member of the nuclear receptor family of transcription factors,
functions as a heterodimer with a retinoid X receptor by binding
the PPAR responsive element (PPRE) within the promoter of
the target genes. PPARγ is activated by a wide variety of
substances including long chain fatty acids, peroxisome
proliferators, 15-deoxy-Δ12,14-Prostaglandin J2 (15d-PGJ2),
and thiazolidinedione (TZD) compounds [1–3]. The actions of
PPARγ are initially assumed as the regulation of lipid
metabolism and homeostasis, whereas recent studies have
emphasized its important role in the regulation of inflammatory
responses, cellular proliferation, and differentiation [4], as well
as the balance between osteogenesis and adipogenesis [5,6].
Adipocytesandosteoblastshavebeenshowntoshare acommon
progenitor, i.e., mesenchymal stem cells in bone marrow [7,8].
Osteoblasts play a requisite role in bone formation [9], and
PPARγ has been shown to be expressed in osteoblasts [10,11].
The imbalance between adipocytes and osteoblasts is associated
with osteoporosis, especially in the age-related osteoporosis that
is accompanied by an increase in marrow adipocytes [12,13].
Furthermore, in mesenchymal cell lines, activation of PPARγ
stimulates adipogenesis and inhibits osteogenesis [14] and over-
Bone 41 (2007) 562–574
www.elsevier.com/locate/bone
⁎Corresponding author. Department of Pharmacology, College of Medicine,
National Taiwan University, No. 1 Sec. 1 Jen-Ai Road Taipei (100), Taiwan.
Fax: +886 2 23417930.
E-mail address: wenmei@ntu.edu.tw (W.-M. Fu).
1Lin T.H. and Yang R.S. have equal contribution.
8756-3282/$ - see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.bone.2007.06.017

Page 2

expression of PPARγ signaling can induce trans-differentiation
of osteoblasts to adipocytes in adipogenic medium [11,15].
The TZD compound rosiglitazone is an FDA-approved oral
insulin-sensitizing antidiabetic agent that is mediated by their
interaction with PPARγ. It has been reported that treatment of
rosiglitazone in non-diabetic mice leads to a significant deteri-
oration in trabecular architecture, decreased bone mineral den-
sity (BMD), suppressed osteoblast differentiation [16,17], and
increased osteoblast apoptosis [18]. Furthermore, recent clinical
trial showed an increase of incidence of fractures in female
patients who underwent long-term treatment of rosiglitazone
[19]. In cell cultures, higher concentrations of TZD compounds
troglitazone and ciglitazone could inhibit the differentiation of
osteoblasts [20]. Akune et al. showed that PPARγ insufficiency
led to an increased osteoblastogenesis in vitro and a higher
trabecular bone volume in vivo [21,22] and the results from
PPARγ insufficient mice indicated that PPARγ negatively
regulated the osteoblastic development and bone formation and
positively regulated adipocyte differentiation. Activation of
PPARγ inhibited the expression and DNA-binding activity of
the transcription factor core-binding-protein-A1 (cbfa1) which
is indispensable for osteoblast differentiation [23,24] and re-
duced osteocalcin expression [25].
Khan et al. demonstrated that activation of PPARγ inhibits
NF-κB nuclear binding to DNA in β-glycerophosphate-treated
osteoblasts [24]. The binding sites for NF-κB are promoter
regionsofgenesencodingvariousosteoblasticproteins,including
cyclooxygenase-2 (COX-2) and inducible nitric oxide synthase
(iNOS) [26]. Prostaglandins are produced by bone cells largely
through the action of inducible COX-2. COX-2 is a regulator of
osteoblast differentiation and plays an important role in response
tomechanicalforce through prostaglandins [27–29]. Nitricoxide
(NO) is produced by bone cells and is a cofactor for the anabolic
responses to mechanical responses [29,30]. Furthermore, NO at
moderate concentrations canmaintainorpromoteproliferationand
prevent the apoptosis of osteoblasts [31,32]. Both COX-2 and
iNOS participate in anabolic response by therapeutic ultrasound
(US) stimulation [33,34].
We here used two kinds of PPARγ activators, TZD compound
ciglitazone and natural ligand 15d-PGJ2, to demonstrate that
activationofPPARγinhibitsosteogenesisbothinvitroandinvivo.
The activation of PPARγ inhibited the expression of COX-2 and
iNOS in the basal level or in respond to US stimulation.
Furthermore, oral administration of rosiglitazone decreased
serum osteoblastic marker alkaline phosphatase (ALP) and tibial
bone mineral density, but left bone resorption unaffected.
Materials and methods
Materials
Rabbit polyclonal antibody for COX-2, N-[2-(cyclohexyloxy)-4-nitrophe-
nyl]-methanesulfonamide (NS-398), and 15d-PGJ2 was purchased from
Cayman Chemical (Ann Arbor, MI, USA). Anti-mouse and anti-rabbit IgG
conjugated with horseradish peroxidase were purchased from Santa Cruz
Biotechnology (Santa Cruz, CA, USA). Mouse anti-rat antibody for α-tubulin
was from Oncogene (Boston, MA, USA). β-Glycerophosphate, ciglitazone,
cetylpyridinium chloride, collagenase, N,N dimethylformamide, fast blue BB salt,
indomethacin, p-nitrophenol phosphate disodium, 2-amino-2-methyl-1-propanol,
3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide, L-ascorbic acid,
and Naphthol AS phosphate were purchased from Sigma Chemical Company (St.
Louis,MO,USA).Wepurchasedα-minimalessentialmedium(α-MEM)andfetal
calf serum from Gibco BRL (Gaithersburg, MD, USA), 2-chloro-5-nitro-N-
phenylbenzamide (GW9662) from Tocris Bioscience (Ellisville, MO, USA),
trypsin and superscript III from Invitrogen (Carlsbad, CA, USA), and TRIZol, tag
DNA polymerase, and PCR primers from MDBIO (Taipei, Taiwan).
Primary osteoblast cultures
Primary osteoblastic cells were obtained from the calvaria of Day-21 fetal rats
[35]. In brief, the calvaria of fetal rats were dissected with aseptic technique. The
soft tissues were removed under dissecting microscope. The calvaria were divided
intosmallpiecesandweretreatedwith1mg/mlcollagenasesolutionfor20–30min
at 37 °C. The next two 20 minute sequential collagenase digestions were then
pooledandfilteredthrough70μmnylonfilters(Falcon,BDBiosciences,SanJose,
CA, USA). The cells were grown on the plastic cell culture dishes in 95% air–5%
CO2with α-MEM, which was supplemented with 10% heat-inactivated fetal calf
serum, 100 U/ml penicillin, and 100 μg/ml streptomycin (pH adjusted to 7.6). The
cell medium was changed every 3 days. The characteristics of osteoblasts were
confirmed by morphology and the expression of ALP.
Measurement of mineralized nodule formation
Osteoblasts were grown for 48 h in 10% FCS/α-MEM in cell culture dishes.
Cells were then trypsinized and plated onto 24-well culture plates at a density of
2×104cells/well in 10% FCS/α-MEM supplemented with 50 μg/ml L-ascorbic
acid and 10 mM β-glycerophosphate [35]. The culture medium was changed
every 3 days and test substances were newly replaced. After 2 weeks, the cells
and matrix were washed twice with 20 mM Tris-buffered saline (TBS)
containing 0.15 M NaCl (pH 7.4) and fixed in ice-cold 75% (vol/vol) ethanol for
30 min and then air-dried. Calcium deposition was determined using
quantitative alizarin red-S staining. In brief, the ethanol-fixed cells and matrix
were stained for 1 h with 40 mM alizarin red-S (pH 4.2) and extensively rinsed
with distilled water. After photography, the bound staining was eluted with 10%
(wt/vol) cetylpyridinium chloride. The alizarin red-S in samples was quantified
by measuring absorbance at 550 nm and calculated according to a standard
curve. One mole alizarin red-S selectively binds about 2 mol of calcium.
Measurement of cytosolic ALP activity and ALP staining
On Day-14, cultured osteoblasts were washed with PBS and then lysed with
buffer containing 0.9% NaCl, 0.6% Tris, 1 mM EGTA, 1% NP-40, and 0.25%
deoxycholate dissolved in RIPA at 4 °C for 30 min. The cell lysates were then
sonicated on ice bath, centrifuged at 1500×g for 5 min, and measured the ALP
activityinsupernatantbyALPassaymixturescontaining0.1M2-amino-2-methyl-
1-propanol, 1 mM MgCl2, and 8 mM p-nitrophenyl phosphate disodium. After
incubationat37°Cfor10–30min,thereactionwasstoppedwith0.1NNaOHand
the absorbance was read at 405 nm. A standard curve was prepared with p-
nitrophenol. Each value was normalized to the protein concentration. For ALP
staining, cultured osteoblasts were fixed on Day-14 by using buffered acetone
(0.03 M sodium citrate, 0.03 M citric acid, and 60% acetone) for 30 s. Cells were
thenwashedtwicebydistilledwater.Cellswerestainedwithstainingreagent(6mg
Napthol AS phosphate, 0.1 ml N,N dimethylformamide, and 20 mg fast blue BB
salt in 20 ml 0.5 M Tris buffer, pH was adjusted to 10.2) for 30 min. Stained
osteoblastswerephotographedusingaIX70microscope(Olympus,Tokyo,Japan).
Ultrasound (US) treatment
Cells (105cells/35 mm dish) were cultured for 2 days and exposed to US
stimulation. AUV-sterilizedtransducer (Exogene2000;SmithandNephewInc.,
Memphis, TN, USA) that generated 1.5 MHz US in a pulsed-wave mode (200-
μs burst width with repetitive frequency of 1 kHz at the intensity of 30 mW/cm2)
was immersed vertically into each culture dish and placed to just contact the
surface of the medium. The distance between the transducer and the cells was
approximately 5 to 6 mm. Exposure time was 20 min for all experiments [33].
563T.-H. Lin et al. / Bone 41 (2007) 562–574

Page 3

Control samples were prepared in the same manner without US exposure. Cells
were harvested at 1 and 3 h after US stimulation.
Measurement of cell viability by 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl
tetrazolium bromide (MTT) assay
MTT was used as an indicator of cell viability as determined by its
mitochondrial-dependentreductiontoformazone.Osteoblastswereplatedonto
96-well culture plates at a density of 104cells/well in 10% FCS/α-MEM. Cells
were grown for 24 h and then treated with ciglitazone or 15d-PGJ2 for another
24h.Afterwashingthecells,100μlα-MEMcontaining0.5mg/mlofMTTwas
added to each well. Cells were incubated for 60 min at 37 °C, the supernatant
was then removed, and the formed blue crystals in viable cells were solubilized
with100μldimethylsulfoxide(DMSO).Thesupernatantwastransferredto96-
well plate and the absorbance of each well was measured at 550 nm.
Cytotoxicity measurement by lactate dehydrogenase (LDH) assay
Cellulardamage was evaluated by LDH leakage. For LDH assay, osteoblasts
wereplatedonto 96-wellculture platesat adensityof 104cells/wellin10%FCS/
α-MEM. Cells were grown for 24 h and then treated with ciglitazone or 15d-
PGJ2 for another 24 h. Culture medium (40 μl) was carefully transferred to an
optically clear 96-well plate. The activity of LDH in the supernatant was
determined using LDH detection kit (Roche Applied Sciences, Indianapolis,
USA). The activity of released LDH was read at 490 nm.
Cell proliferation assay using BrdU
For the 5-bromo-2′-deoxyuridine (BrdU) assay (Roche Applied Sciences,
Indianapolis, IN, USA), cells were seeded onto 96-well plates at a density of 5000
cells/well.Cellswereculturedonedaywith10%FCS/α-MEM(90μl).Themedium
waschangedandcellswerethentreatedwithtestsubstancesfor24h.BrdU(100μM
stock solution,10μl) wasaddedtoeach wellandthe cells were incubatedforfurther
4h.Afterremovalofthemedium,cellswerefixedandtheDNAwasdenaturedusing
the FixDenat reagent. The FixDenat reagent was removed after 30 min incubation,
anti-BrdU-POD solution was added, and the plate was incubated for 90 min at room
temperature.Thecellswerethenwashedthreetimeswithwashingsolution,100μlof
the substrate solution was added to each well, and relative light units were measured
by using a microplate luminescence reader.
Western blot analysis
Osteoblasts were seeded onto 6-well plates at a density of 8×104cells/well.
Cells were incubated with 50 μg/ml L-ascorbic acid and 10 mM β-glyceropho-
sphateandtestsubstanceswereaddedfor3days.CellswerethenwashedwithPBS
and lysed for 30 min at 4° with lysis buffer as described previously. Equal protein
was applied per lane, and electrophoresis was performed under denaturing
conditions ona 10% SDS gel and transferred toan immobilon polyvinyldifluoride
(PVDF) membranes at 4 °C overnight. The blots were blocked with 4% BSA for
1hatroomtemperatureandthenprobedwithrabbitanti-ratantibodyagainstCOX-
2 (1:1000) for 1 h at room temperature. After three washes, the blots were
subsequentlyincubated witha donkey anti-rabbit or sheep anti-mouse peroxidase-
conjugated secondary antibody (1:1000) for 1 h at room temperature. The blots
were visualized by enhanced chemiluminescence using Kodak X-OMAT LS film
(Kodak,Rochester,NY,USA).Fornormalizationpurposes,thesameblotwasalso
probed with mouse anti-rat α-tubulin antibody (1:1000).
Reverse transcriptase polymerase chain reaction (RT-PCR) for mRNA
analysis
Osteoblasts were seeded on 6-well plate at a density of 8×104cells/well in
10% FCS/α-MEM supplemented with 50 μg/ml L-ascorbic acid and 10 mM β-
glycerophosphate, and test substances were added. Total RNAwas extracted by
TRIZol Kit on Day-3 and was quantified by adding 1 μl sample to 79 μl RNase-
free water and the absorbance was measured in a RNA/DNA calculator
(GeneQuant Pro, GE Healthcare, Piscataway, NJ, USA) at 260 and 280 nm. Two
micrograms of total RNAwas used for RT-PCRby usingSuperScriptIII RT-PCR
kit. Amplification by Tag DNA polymerase was accomplished at 26–35 cycles,
whichwas within the linear changeof PCR product. In brief, sense and antisense
primers were added and the following profile was used: 1 cycle for 94 °C for
3 min, followed by set cycles of 94 °C for 30 s, 55–58 °C for 1 min, and 72 °C
for 1 min. All primers used are listed as follows:
GAPDH (609 bp) sense, GCCATCAACGCCCCTTCATTGAC; antisense,
ACGGAAGGCCATGCCAGTGAGCTT, 26 cycles (55 °C);
(265 bp) sense, ACTCCCATTCCTTTGACATC; antisense,
TCCCCACAGACTCGGCACTC, 32 cycles (56 °C);
Osteocalcin (416 bp) sense, TGAGGACCCTCTCTCTGCTC; antisense,
ATTTTGGAGCAGCTGTGCCG, 34 cycles (55 °C);
COX-2 (702 bp) sense, TGATGACTGCCCAACTCCCATG; antisense,
AATGTTGAAGGTGTCCGGCAGC, 31 cycles (58 °C);
iNOS(651 bp) sense ACAACAGGAACCTACCAGCTCA; antisense,
GATGTTGTAGCGCTGTGTGTCA, 31 cycles (58 °C);
BMP2(440 bp) sense, GAGTTTGAGTTGAGGCTGCTC; antisense,
TGAGTCACTAACCTGGTGTCC, 34 cycles (58 °C).
PPARγ
PCR products were then separated electrophoretically in a 2% agarose DNA
gel and stained with ethidium bromide. mRNA levels were normalized to the
levels of GAPDH.
Quantitative real-time PCR
The RNAwas obtained as described above and 2 μg total RNAwas used for
RT-PCR by using SuperScriptIII followed by PCR in real time using an SYBR
GreenPCR mastermix (AppliedBiosystems,CA, USA)andan ABI Prism 7500
sequence detection system (Applied Biosystems, CA, USA). Amplification was
performed in the following cycling conditions: 50 °C for 2 min and 95 °C for
10 min and then 40 cycles of 95 °C for 15 s followed by 60 °C for 1 min. The
optimal concentrations of primers and templates that were used in each reaction
were established based on the standard curve created before the reaction and
corresponded to the nearly 100% efficiency of the reaction. The reference
household gene used to normalize the amount of mRNA in the cultures was β-
actin. The fold change in gene expression relative to the control was calculated
by 2
−ΔΔCT. Primers used are listed as follows:
β-actin, sense GCTCCTCCTGAGCGCAAGTA;
antisense GGCCAGGATAGAGCCACCA;
Osteocalcin, sense GGCTTCCAGGACGCCTACA;
antisense CATGCCCTAAACGGTGGTG;
COX-2, sense TACTGTGTAGCTCCCCTTCG;
antisense TGCCCAGAACTACCCACTAA;
iNOS, sense GGAGCATCACCCCTGTGT
antisense GGTCTTCCAGGGCTCGAT;
BMP-2, sense AAGGCACCCTTTGTATGTGG;
antisense CATGCCTTAGGGATTTTGGA;
Determination of nitrite concentration
The oxidation of L-arginine by NOS results in the production of citrulline and
unstablemoleculenitricoxide.Nitricoxideisrapidlyconvertedtostablemetabolite:
nitriteandnitrate.NitriteinculturesupernatantswasmeasuredbytheGriessreaction.
Briefly, 100 μl culture supernatants were mixed with an equal volume of Griess
reagent (0.5% sulfanilamide, 2.5% H3PO4, and 0.05% naphthylethylene diamine in
H2O) and incubated for 5 min at room temperature. Absorbance was determined at
550 nm and compared with a standard curve obtained using sodium nitrite.
Local injection of drugs into the tibia of young rats and bone
histomorphometry
Three-week-old male Sprague–Dawleyrats weighing 55 g to 92 g were used
for local injection. A cannula (21G) was done from the posteriolateral side into
the proximal tibial metaphysis in both limbs of young rats [35]. The cannula had
its outer end in the subcutaneous tissue. Ciglitazone (10 μM or 100 μM, 10 μl),
564T.-H. Lin et al. / Bone 41 (2007) 562–574

Page 4

15d-PGJ2 (100 μM, 10 μl), or PPARγ antagonist GW9662 (30 μM, 10 μl) was
percutaneously injected into the proximal tibia through the cannula for 7
consecutive days (once/day). For histological analysis, the hindlimbs were fixed
using PBS containing 4% paraformaldehyde at 4 °C for 48 h, decalcified using
10% Na2EDTA at 4 °C for 14 days, dehydrated in increasing concentration of
ethanol, and embedded in paraffin. The serial histological sections were cut
longitudinally (4 μm) and stained with Meyer's hematoxylin–eosin solution.
Images of the growth plate and proximal tibia were photographed using an IX70
microscope. Measurement of bone volume was performed on the secondary
spongiosa, which is located 1.0 to 3.0 mm distal to the epiphyseal growth plate
andischaracterizedbyanetworkoflargertrabeculae.Bonevolumewascalculated
usingimageanalysissoftware(ImageProPlus3.5,MediaCybernetics,MD,USA)
andexpressedaspercentageofbonearea.Allmeasurementsweredoneinasingle-
blind fashion.
Oral treatment of rosiglitazone in adult rats and serum biochemistry
and bone mineral density measurement
Three-month-old male Sprague–Dawley rats weighing 237 g to 262 g
were used for oral treatment of rosiglitazone. Rosiglitazone maleate (Avandia,
GlaxoSmithKline, King of Prussia, PA, USA) at the doses of 150 mg/kg
(containing 4 mg rosiglitazone) was suspended in 0.9% saline solution and
administered by gavage daily for 14 consecutive days (once/day). Control
groups received vehicle in a comparable volume (6 ml/kg body weight). The
rats were anesthetized and sacrificed on Day-15. The blood sample was
quickly obtained from left ventricle. Serum samples were prepared by
centrifuging blood (1500×g for 5 min) with vacutainer tubes (Becton
Dickinson, Franklin Lakes, NJ, USA). In addition, the tibia and femur were
also removed and cleaned of soft tissue for bone mineral density (BMD)
measurement.
Serum calcium, GPT (glutamic pyruvic transaminase), and ALP levels were
determined using an auto dry chemistry analyzer (SPOTCHEM SP-4410,
Arkray Inc., Kyoto, Japan). Serum C-terminal telopeptides of type-I collagen
levels were measured by Serum CrossLaps ELISA assay for the evaluation of
bone resorption (Nordic Bioscience Diagnostics, Herlev, Denmark). Bone
mineral density of the tibia and femur was measured with a dual-energy x-ray
absorptiometer (DEXA, XR-26; Norland, Fort Atkinson, WI, USA). The mode
adapted to the measurements of small subjects was adopted. A coefficient of
variation of 0.7% was calculated from daily measurements of BMD on a lumbar
phantom for more than 1 year. Proximal tibia, proximal femur, whole tibia, and
whole femur were scanned, and BMD was then measured by absorptiometer
[35].
Statistics
The values given weremeans±SEM.The significanceof difference between
the experimental group and control was assessed by Student's t test. The
difference is significant if the P value is less than 0.05.
Results
Effect of ciglitazone and 15d-PGJ2 on osteoblastic
differentiation and proliferation
We first examined whether PPARγ is expressed in primary
cultured rat osteoblasts. By using RT-PCR, PPARγ mRNAwas
detected in rat calvaria osteoblasts in different generations
(Fig. 1). In order to examine the effects of PPARγ activators
on osteoblastic function, two specific PPARγ activators, TZD
compound ciglitazone or natural ligand 15d-PGJ2, were used
in this study. The differentiation and maturation of osteoblasts
were evaluated by the assay of alkaline phosphatase activity
and mineralized bone nodule formation using alizarin red-S
staining. To examine the effect of PPARγ ligands on the
differentiation in primary cultured rat osteoblasts, cells were
incubated for 14 days with osteogenic medium containing β-
glycerophosphate and L-ascorbic acid and PPARγ ligands.
Both ciglitazone and 15d-PGJ2 concentration-dependently
reduced mineralization (Figs. 2A and E) and ALP activity
(Figs. 2B–D and F), suggesting that the activation of PPARγ
inhibits the maturation of primary osteoblasts.
Several studies indicate that activation of PPARγ is
associated with cell death [36,37]. The LDH release and cell
viability were then measured in cells exposed to ciglitazone and
15d-PGJ2 for 24 h. Both ciglitazone and 15d-PGJ2 did not
cause cell death and induce a loss of osteoblastic viability when
the concentration was lower than 10 μM (Figs. 2G and H).
However, 10 μM 15d-PGJ2 can cause cell death when serum
was depleted from culture medium (cell viability: 55.3±12.8%,
compared with control, n=3).
To examine the effects of PPARγ ligands on the cell
proliferation, we used BrdU technique that is based on the
incorporation of the pyrimidine analogue BrdU instead of
thymidine into the DNA of proliferating cells. Ciglitazone
and 15d-PGJ2 inhibited BrdU incorporation in primary
osteoblasts cultured in medium without differentiation agents
(Fig. 2I). However, ciglitazone did not affect proliferation
rate when osteoblasts were cultured in the differentiating
medium (in the presence of β-glycerophosphate plus
ascorbic acid) and 15d-PGJ2 at 10 μM still inhibited the
proliferation rate of osteoblasts in the differentiation stage
(Fig. 2J). These results indicate that PPARγ ligands inhibited
mineralization mainly via interrupting the differentiation of
osteoblasts.
L-
Activation of PPARγ inhibited the expression of COX-2 and
BMP-2
Inducible COX-2 is involved in the anabolic effects of
osteoblasts in response to mechanical forces [38]. To examine
the effects of PPARγ activators on the expression of COX-2,
osteoblastswereincubatedwithosteogenic mediumandPPARγ
activators for 3 days. 15d-PGJ2 significantly decreased both the
mRNA and protein levels of COX-2 and ciglitazone at higher
concentration of 10 μM also exerted similar effects (Fig. 3). In
addition,quantitativePCRanalysisalsodemonstratedthedown-
Fig. 1. Expression of PPARγ mRNA in rat calvaria osteoblasts. Rat calvaria
osteoblasts were incubated in α-MEM containing 10% FBS for 3 days and RNA
was extracted by TRIZol. (A) Rat calvaria osteoblasts expressed PPARγ mRNA
in different cycle number of PCR reaction (n=3). (B) PPARγ mRNA was
observed in different generations of cultured rat calvaria osteoblasts (n=3).
565T.-H. Lin et al. / Bone 41 (2007) 562–574

Page 5

566T.-H. Lin et al. / Bone 41 (2007) 562–574