Stimulation of Resorption in Cultured Mouse Calvarial Bones by Thiazolidinediones

Stimulation of Resorption in Cultured Mouse CalvarialBones by Thiazolidinediones

A. M. Schwab,* S. Granholm,* E. Persson,* B. Wilkes, U. H. Lerner, and H. H. Conaway

Department of Physiology and Biophysics (A.M.S., B.W., H.H.C.), University of Arkansas for Medical Sciences, Little Rock,Arkansas 72205; and Department of Oral Cell Biology (S.G., E.P., U.H.L.), Umeå University, SE-901 87 Umeå, Sweden

Dosage-dependent release of 45Ca was observed from prela-beled mouse calvarial bones after treatment with two thiazo-lidinediones, troglitazone and ciglitazone. Release of 45Caby ciglitazone was decreased by the osteoclast inhibitorsacetazolamide, calcitonin, 3-amino-1-hydroxypropylidene-1,1-bisphosphonate, and IL-4, but not affected by the peroxi-some proliferator-activated receptor � antagonist, GW 9662,the mitotic inhibitor, hydroxyurea, or indomethacin. En-hanced expression of receptor activator of nuclear factor-�Bligand (RANKL) mRNA and protein and decreased osteopro-tegerin (OPG) mRNA and protein were noted after ciglitazonetreatment of calvariae. Ciglitazone and RANKL each causedincreased mRNA expression of osteoclast markers: calcitoninreceptor, tartrate-resistant acid phosphatase, cathepsin K,matrix metalloproteinase-9, integrin �3, and nuclear factor of

activated T cells 2. OPG inhibited mRNA expression of RANKLstimulated by ciglitazone, mRNA expression of osteoclastmarkers stimulated by ciglitazone and RANKL, and 45Ca re-lease stimulated by troglitazone and ciglitazone. Increasedexpression of IL-1� mRNA by ciglitazone was not linked toresorption stimulated by the thiazolidinedione. Ciglitazonedid not increase adipogenic gene expression but enhancedosteocalcin mRNA in calvariae. In addition to exhibiting sensi-tivity to OPG, data indicate that stimulation of osteoclast dif-ferentiation and activity by thiazolidinediones may occur by anonperoxisome proliferator-activated receptor �-dependentpathway that does not require cell proliferation, prostaglan-dins, or IL-1� but is characterized by an increased RANKL toOPG ratio. (Endocrinology 146: 4349–4361, 2005)

THE PEROXISOME PROLIFERATOR-activated recep-tors (PPARs) are a family of ligand-regulated tran-

scription factors that belong to the nuclear hormone receptorsuperfamily (1). There are three PPARs: �, � (also called �),and � (2–5). The predominant PPAR� isoform is PPAR�1. Asecond isoform, PPAR�2, is expressed principally in adiposetissue (6). Natural PPAR� ligands include fatty acids andfatty acid derivatives (7), such as eicosapentaenoic acid (8),9- and 13-hydroxyoctadecadienoic acid (9), and 15-deoxy-�12,14-prostaglandin J2 (15d-PG-J2) (10). Numerous syntheticligands have also been described. These include drugs suchas the insulin-sensitizing thiazolidinediones (e.g. ciglitazone,troglitazone, rosiglitazone) (10, 11) and nonsteroidal antiin-flammatory compounds like indomethacin and ibuprofen(12).

Mesenchymal stem cells (stromal cells) are capable of de-veloping into adipocytes, osteoblasts, myoblasts, fibroblasts,and chondroblasts (13, 14). PPAR� stimulates adipocyte dif-ferentiation, and it has been posited that the increase in

adipose tissue and decrease in osteoblast differentiation as-sociated with bone loss in conditions such as aging may beoccurring reciprocally, because of increased PPAR� activity(15).

Osteoclasts arise from precursor cells of the monocyte/macrophage lineage (16). Stromal cells/osteoblasts mediatethe formation of fully differentiated bone resorbing oste-oclasts from progenitor cells (16). Key factors regulating os-teoclastogenesis include macrophage colony-stimulatingfactor (M-CSF) and receptor activator of nuclear factor(NF)-�B ligand (RANKL) (16–19). M-CSF is a product ofstromal cells/osteoblasts that enhances macrophage and os-teoclast survival. RANKL is a transmembrane protein pro-duced in stromal cells/osteoblasts. RANKL plays an essen-tial role in osteoclast differentiation and function by bindingto receptor activator of NF-�B (RANK) on osteoclast pro-genitor cells. Mice that do not express RANKL or RANK orare deficient in M-CSF do not have functional osteoclasts anddevelop osteopetrosis (20–23). The interaction betweenRANKL and RANK can be blocked by osteoprotegerin(OPG), a soluble protein released from stromal cells/osteo-blasts that acts as a decoy receptor for RANKL. Mice with atargeted deletion of opg develop multiple fractures and havedecreased trabecular bone volume and numerous osteoclasts(24).

After exposure to RANKL and M-CSF, osteoclast forma-tion and bone resorption in bone marrow coculture experi-ments (human mesenchymal stem cells and CD34� hema-topoietic stem cells) (25) and osteoclastogenesis in primarymurine myeloid cells and Raw cells (26) have been reportedto be blocked by both the natural PPAR� ligand 15d-PG-J2and a thiazolidinedione, ciglitazone. Because only CD34�

First Published Online June 30, 2005* A.M.S., S.G., and E.P. contributed equally to this study.Abbreviations: AHPrBP, 3-Amino-1-hydroxypropylidene-1,1-bisphos-

phonate; C/EBP�, CCAAT/enhancer binding protein �; CT, calcitonin;CTR, CT receptor; D3, 1,25-(OH)2 vitamin D3; 15d-PG-J2, 15-deoxy-�12,14-prostaglandin J2; HU, hydroxyurea; IRAP, IL-1 receptor antagonist protein;M-CSF, macrophage colony-stimulating factor; MMP-9, matrix metallo-proteinase-9; NF, nuclear factor; OPG, osteoprotegerin; OSM, oncostatin M;PPAR, peroxisome proliferator-activated receptor; RANK, receptor activa-tor of NF-�B; RANKL, receptor activator of NF-�B ligand; TRAP, tartrate-resistant acid phosphatase.Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.

0013-7227/05/$15.00/0 Endocrinology 146(10):4349–4361Printed in U.S.A. Copyright © 2005 by The Endocrine Society

doi: 10.1210/en.2005-0601

4349

stem cells expressed PPAR� and no supporting cells werepresent with the murine myeloid cells or RAW 264.7 cells,these studies have indicated that an inhibitory action of thePPAR� ligands is specific to the osteoclast lineage. Experi-ments in mouse whole-bone marrow cultures have alsopointed to an inhibitory role for the PPAR� ligands (27). Inmouse whole-bone marrow cultures, it has been found thatthiazolidinedione drugs can decrease the number of tartrate-resistant acid phosphatase (TRAP�) multinucleated oste-oclasts formed in response to 1,25-(OH)2 vitamin D3 (D3) andPTH, and dose dependently decrease basal as well as D3 andPTH-induced resorption on dentine slices (27).

However, in contrast to the studies showing inhibition ofbone resorption, thiazolidinedione treatment of cocultures ofmurine bone marrow-derived BMS2 adipocytes and primarybone marrow cells has been found to have no effect on theformation of TRAP� osteoclasts in the presence of D3 (28).Furthermore, an investigation testing the effectiveness ofBRL49653 (rosiglitazone) in intact and ovariectomized ratshas shown no inhibition of bone resorption in intact animalsand a substantial increase in bone resorption in estrogen-deprived animals (29).

Not all actions of compounds known to be PPAR� ligandsare mediated by PPAR� activation (30–38). In the case of thethiazolidinedione, ciglitazone, inhibition of cholesterol bio-synthesis in cultured Chinese hamster ovary cells (35), stim-ulation of MAP kinase cascades in astrocytes and preadipo-cytes (36), decreased growth and increased apoptosis of renalinterstitial fibroblasts (37), and decreased expression of in-hibitor of DNA binding (Id2) in human aortic smooth musclecells (38) have all been reported previously as PPAR�-inde-pendent mechanisms.

The effects of PPAR� ligands on bone resorption and ex-pression of RANKL, OPG, and RANK have not been eval-uated in calvarial bone explants. In the present investigation,the neonatal mouse calvarial model was used to: 1) test theeffects of thiazolidinediones and 15d-PG-J2 on bone resorp-tion, 2) determine whether resorptive effects are PPAR� de-pendent or independent, and 3) correlate changes in resorp-tion with expression of regulators of osteoclastogenesis, suchas RANKL, OPG, and RANK.

Materials and MethodsExperimental animals

Animal experiments were conducted in accordance with acceptedstandards of humane animal care and use as deemed appropriate by theAnimal Care and Use Committees of the University of Arkansas MedicalSciences Campus (Little Rock) and by Umea University, Sweden.

Materials

Recombinant mouse RANKL, mouse OPG fused to human IgG1 Fc(OPG/Fc chimera), recombinant human IL-1�, recombinant human IL-1receptor antagonist protein (IRAP), and immunoassay kits for mouseOPG and mouse RANKL were purchased from R&D Systems (Abing-don, UK); fetal bovine serum from ICN Pharmaceuticals, Inc. (CostaMesa, CA); ciglitazone, pioglitazone, troglitazone, acetazolamide, hy-droxyurea (HU), GW 9662 (Sigma no. M-6191), and acid-free fetal BSAfrom Sigma Chemical Co. (St. Louis, MO); �-MEM, TRIzol LS Reagent,and oligonucleotide primers from Invitrogen/Life Technologies (Pais-ley, UK); fluorescent-labeled probes (reporter fluorescent dye VIC at the5�end and quencher fluorescent dye TAMRA at the 3�end) and TaqMan

Universal PCR Master Mix from Applied Biosystems (Warrington, UK);the first strand cDNA synthesis Kit and PCR Core kit from Roche(Mannheim, Germany); Triton X-100, HotStar Taq polymerase, and QIA-quick purification kit from Qiagen/VWR International (Stockholm,Sweden); Thermo Sequenase-TM II DYEnamic ET terminator cycle se-quencing kit from Amersham Biosciences (Uppsala, Sweden); [45Ca]CaCl2 from Amersham Life Science (Little Chalfont, Buckinghamshire,UK); 15d-PG-J2 from Calbiochem (La Jolla, CA); synthetic bovine PTH1-34 from Bachem (Bubendorf, Switzerland); and culture dishes andmultiwell plates from Costar (Cambridge, MA). Salmon calcitonin (CT)was kindly supplied by Novartis (Basel, Switzerland); indomethacin byMerck, Sharp & Dohme (Haarlem, The Netherlands); D3 by Hoff-mann-La Roche (Basel, Switzerland); and the bisphosphonate 3-amino-1-hydroxypropylidene-1,1-bisphosphonate (AHPrBP) from HenkelKGaA (Dusseldorf, Germany).

Bone organ culture

Bone resorption was assessed by analyzing mineral mobilization incultured mouse calvarial bones. Parietal bones from 6- to 7-d-old CD-1or CsA mice were dissected and cut into four pieces. The bones werepreincubated for 18–24 h in �-MEM containing 0.1% albumin and 1�mol/liter indomethacin. After preincubation, the bones were exten-sively washed and subsequently cultured for up to 120 h in multiwellculture dishes containing 2.0 ml indomethacin-free medium, with orwithout test substances (39). The bones were incubated in the presenceof 5% CO2 in humidified air at 37 C.

Measurements of mineral release

Mineral mobilization was assessed by analyzing the release of 45Cafrom bones prelabeled in vivo. In most experiments, 2- to 3-d-old micewere injected with 1.5 �Ci 45Ca and the amount of radioactivity in boneand culture media was analyzed by liquid scintillation at the end of theculture period. For the time course experiments, the mice were injectedwith 12.5 �Ci 45Ca, and radioactivity was analyzed at different timepoints by withdrawal of small amounts of culture media. Release ofisotope was expressed as the percentage release of the initial amount ofisotope (calculated as the sum of radioactivity in medium and bone afterculture) (39). In some experiments, the data were recalculated, and theresults were expressed as percentage of control, which was set at 100%.This allowed for accumulation of data from several experiments.

RNA isolation and first strand cDNA synthesis

Before RNA isolation, five calvarial halfs per group were preincu-bated for 18–24 h in �-MEM containing 0.1% albumin and 1 �mol/literindomethacin before subsequent incubation in 24-well plates in theabsence or presence of ciglitazone and D3. Total RNA was extracted fromindividual bones with TRIzol LS reagent by following the manufactur-er’s protocol. The RNA was quantified spectrophotometrically, and theintegrity of the RNA preparations was examined by agarose gel elec-trophoresis. Only RNA preparations showing intact species were usedfor subsequent analysis. One microgram of total RNA was reverse tran-scribed into single-stranded cDNA with a 1st Strand cDNA Synthesis Kitusing oligo(dT)15 primers. After incubation at 25 C for 10 min and at 42C for 60 min, the avian myeloblastosis virus reverse transcriptase wasdenatured at 99 C for 5 min. The cDNA was kept at �20 C until usedfor PCR.

Semiquantitative RT-PCR

For semiquantitative RT-PCR analysis, RNA from four to five bonesper each group were pooled. The PCR for CCAAT/enhancer bindingprotein � (C/EBP�), cathepsin K, GAPDH, IL-6, IL-1�, IL-11, LIF, os-teocalcin, oncostatin M (OSM), PPAR�1, PPAR�2, TNF-�, TRAP,RANKL, and Wnt10b were performed using PCR standard protocol. Inthe PCR for IL-1�, the final MgCl2 concentration was changed from 1.5–1mm. The conditions for PCR were denaturing at 94 C for 2 min, annealingat various temperatures for 40 sed, followed by elongation at 72 C for60 sec; in subsequent cycles, denaturing was performed at 94 C for 40 sec.The PCRs for CT receptor (CTR), NF of activated T cells 2 (NFAT2),integrin �3, and matrix metalloproteinase-9 (MMP-9) were initiated with

4350 Endocrinology, October 2005, 146(10):4349–4361 Schwab et al. • Thiazolidinediones Stimulate Resorption

hot start at 94 C for 15 min, using HotStar Taq polymerase. Annealingtemperatures were 54 C (OSM), 57 C (C/EBP�, cathepsin K, GAPDH,IL-11, integrin �3, LIF, MMP-9, and Wnt10b), 58 C (TRAP), 60 C (IL-1�),61 C (NFAT2 and TNF-�), 63 C (IL-6, PPAR�1, and PPAR�2), 64 C(osteocalcin), 65 C (RANKL), and 67 C (CTR). The PCR for RANKL wasperformed with a step-down technology in which the primer annealingtemperature was decreased by 5 C every five cycles down to 45 C. Thesequences of primers, positions of the 5� and 3� ends of the predicted PCRproducts, GenBank accession numbers, and estimated fragment lengthsare listed in Table 1. The expressions of these factors were compared atthe logarithmic phase of the PCR. No amplification was detected insamples where the RT reaction had been omitted (data not shown). ThePCR products were separated by electrophoresis in 1.5% agarose gels

and visualized using ethidium bromide. The identity of the PCR prod-ucts was confirmed using a QIAquick purification kit and a ThermoSequenase-TM II DYEnamic ET terminator cycle sequencing kit withsequences analyzed on an ABI377 XL DNA sequencer.

Quantitative real-time PCRs

Quantitative PCR analyses on cDNA from individual bones ofRANKL, RANK, OPG, TRAP, cathepsin K, NFAT2, and �-actin mRNAwere performed using the TaqMan Universal PCR Master Mix kit andABI PRISM 7900 HT Sequence Detections System and software (AppliedBiosystems, Foster City, CA) as described previously (40). The sequencesof the primers and probes, the GenBank accession numbers, and the

TABLE 1. Sequences of primers used in semiquantitative RT-PCR, positions of the 5� and 3� ends of predicted nucleotides, GenBankaccession numbers, and estimated fragment length

mRNA species Sequence (5�–3�) 5� and 3� ends GenBank accession no. Size (bp)

C/EBP� 820–1008 NM007678 189Sense AAAGCCAAGAAGTCGGTGGAAntisense CAGTTCACGGCTCAGCTGTT

Cathepsin K 552–889 NM007802 338Sense CTTGTGGACTGTGTGACTAntisense AACACTGCATGGTTCACA

CTR 1483–1650 U185421 168Sense TGCTGGCTGAGTGCAGAAACCAntisense GGCCTTCACAGCCTTCAGGTAC

IL-1� 8161–12240 AF010237 451Sense TCCAACCCAGATCAGCACCTTACAntisense AGTCCCCGTGCCAGGTGC

IL-6 1328–7622 M20572 638Sense ATGAAGTTCCTCTCTGCAAGAGACTAntisense CACTAGGTTTGCCGAGTAGATCTC

IL-11 221–483 NM008350 263Sense CACAGATGAGAGACAAATTCCAntisense AGACATCAAGAGCTGTAAACG

Integrin �3 268–470 AF026509 203Sense ATTGAGTTCCCAGTCAGTGAGAntisense GACAGGTCCATCAAGTAGTAG

LIF 368–600 A01690 233Sense TCCCTGACCAATATCACCAntisense TTGTATGTCCCCAGAAGC

MMP9 1564–2142 NM013599 579Sense TCTGAGGCCTCTACAGAGTCTAntisense CTCATGGTCCACCTTGTTCAC

NFAT2 1685–1776 AF239169 92Sense TGAGGCTGGTCTTCCGAGTTAntisense CGCTGGGAACACTCGATAGG

OSM 267–608 D31942 342Sense GTGGCCTTCCCCAGTGAGGAAntisense TGAGCCCATGAAGCGATGGTATCC

Osteocalcin 280–826 L24431 233Sense TCTGACAAAGCCTTCATGTCCAntisense AAATAGTGATACCGTAGATGCG

PPAR�1 391–745 U01841 354Sense TTCTGACAGGACTGTGTGACAGAntisense ATAAGGTGGAGATGCAGGTTC

PPAR�2 34–384 NM011146 350Sense GCTGTTATGGGTGAAACTCTGAntisense ATAAGGTGGAGATGCAGGTTC

RANKL 6186–6485 Y00467 307Sense GGCAGGTCTACTTTGGAGTCATTGCAntisense ACATTCCAGGCTCCAGTGAATTCGG

TRAP 1072–1384 NM007388 313Sense AAATCACTCTTCAAGACCAGAntisense TTATTGAACAGCAGTGACAG

Wnt10b 1420–2080 U30464 661Sense CTGCCACTGTCGTTTCCACTGAntisense AGACCCTTTCAACAACTGAACG

GAPDH 957–1223 M32599 267Sense ACTTTGTCAAGCTCATTTCCAntisense TGCAGCGAACTTTATTGATG

Schwab et al. • Thiazolidinediones Stimulate Resorption Endocrinology, October 2005, 146(10):4349–4361 4351

positions of the 5�and 3�ends of the predicted nucleotides are given inTable 2. To control for variability in amplification due to differences instarting mRNA concentrations, �-actin was used as an internal standard.The relative expression of target mRNA was computed from the targetthreshold cycle values and �-actin threshold cycle values using thestandard curve method (User Bulletin #2, Applied Biosystems).

Protein analyses

The protein synthesis of RANKL and OPG was assessed by measur-ing the levels of RANKL and OPG in calvarial bones using commerciallyavailable ELISA kits (41). After preincubation, eight calvarial halves pergroup were individually incubated in 24-well plates in the absence orpresence of 10�5 m ciglitazone and 10�8 m D3 for 48 h. After treatmentwith 0.2% Triton X-100, extracted bone samples were analyzed using themanufacturer’s protocols for the ELISAs. The sensitivities of the immu-noassays are 5 pg/ml. D3 was used as a positive control and, as expected,resulted in an increased protein level of RANKL and a decreased levelof OPG in calvarial bones.

Statistical analysis

Statistical analysis of multiple treatment groups was performed withTukey’s pair-wise comparison after one-way ANOVA of logarithmictransformed data.

ResultsStimulation of neonatal mouse calvarial bone resorption bythiazolidinediones

PPAR� ligands (ciglitazone, troglitazone, pioglitazone,and 15d-PG-J2) were tested in neonatal mouse calvarial bonecultures to determine whether they could stimulate 45Carelease. The results of this study are shown in Fig. 1A. Allagents were tested at concentrations ranging from 10�7 to 5 �10�5 m. Calvarial bones were cultured for 120 h with the testsubstances. Pioglitazone and 15d-PG-J2 had no significant

affect on 45Ca release at any concentration tested. In contrast,significant stimulations of 45Ca release were noted with tro-glitazone (at 10�6 and 5 � 10�6 m) and ciglitazone (at 5 �10�6, 10�5, and 5 � 10�5 m). Release of 45Ca caused bytroglitazone and ciglitazone was biphasic, with peak releaseby each agent decreased at higher concentrations. The de-crease in release may have been due to toxicity of the thia-zolidinediones because at 10�4 m ciglitazone, only a verysmall amount of RNA could be extracted from calvarialbones.

A time-course study showed that 10�5 m ciglitazone sig-nificantly enhanced (P � 0.01) 45Ca release at 72 and 120 h(Fig. 1B). Additionally, two different concentrations of PTH(10�10 and 10�8 m) were found to cause time-dependentstimulations of 45Ca release (Fig. 1B). The response caused by10�10 m PTH was significantly enhanced (P � 0.01) at 72 and120 h (Fig. 1B), whereas that caused by 10�8 m PTH wassignificantly increased (P � 0.01) at 24 (the first time mea-sured), 72, and 120 h.

The stimulation of 45Ca release in calvarial bones byciglitazone in the presence of different inhibitors, CT,AHPrBP, acetazolamide, indomethacin, HU, and IL-4

Effects on 45Ca release caused by addition of CT (10�9 m)and AHPrBP (10�4 m) to calvariae treated for 144 h with 10�8

m PTH and 10�5 m ciglitazone are shown in Fig. 2A. AHPrBPresulted in inhibition of resorption caused by ciglitazone andthe positive control, PTH, throughout the experimental pe-riod. Inhibition of both ciglitazone and PTH was also ob-served with CT, but unlike the inhibition found with the

TABLE 2. Sequences of primers and probes used in quantitative real-time PCR, positions of the 5� and 3� ends of predicted nucleotidesand GenBank accession numbers

mRNA species Sequence (5�–3�) 5� and 3� ends GenBank accession no.

Cathepsin K 606–695 NM007802Sense primer ATATGTGGGCCAGGATGAAAGTTAntisense primer TCGTTCCCCACAGGAATCTCTProbe CCACGGCAAAGGCAGCTAAATGCA

NFAT2 1685–1776 AF239169Sense primer TGAGGCTGGTCTTCCGAGTTAntisense primer CGCTGGGAACACTCGATAGGProbe CCGGACGCTGTCTCTCCAGGTGG

OPG 845–958 U94331Sense primer AGCTGCTGAAGCTGTGGAAAntisense primer TGTTCGAGTGGCCGAGATProbe CCAAGACATTGACCTCTGTGAAAGCA

RANK 1422–1593 AF019046Sense primer TGCCTACAGCATGGGCTTTAntisense primer AGAGATGAACGTGGAGTTACTGTTTProbe CCAGTGAAGCAGCAGCCAGCAT

RANKL 606–680 AF053713Sense primer TGGAAGGCTCATGGTTGGATAntisense primer CATTGATGGTGAGGTGTGCAAProbe AGGCTTGCCTCGCTGGGCCAC

TRAP 771–847 BC019160Sense primer CGACCATTGTTAGCCACATACGAntisense primer TCGTCCTGAAGATACTGCAGGTTProbe CACTGCCTACCTGTGTGGACATGA

�-Actin 471–563 M12481Sense primer GGACCTGACGGACTACCTCATGAntisense primer TCTTTGATGTCACGCACGATTTProbe CCTGACCGAGCGTGGCTACAGCTTC


bisphosphonate, escape from the inhibitory action of CToccurred in calvarial bones treated with ciglitazone and PTH.

Fig. 2 (B–E) shows the effects on 45Ca release caused by

additions of 10�4 m acetazolamide, 10�6 m indomethacin,10�3 m HU and 10 ng/ml IL-4, respectively, to calvarial bonestreated with 10�5 m ciglitazone, 10�8 m PTH or 10�8 m D3 for120 h. Significant decreases (P � 0.01) in 45Ca release wereobserved when the osteoclast inhibitors acetazolamide andIL-4 were added to calvarial bones treated with either PTHor D3, but no changes in mineral release stimulated by PTHand D3 were noted in the presence of the inhibitor of pros-taglandin biosynthesis, indomethacin, or the inhibitor ofDNA synthesis, HU. Similar results were obtained in cal-varial bones treated with 10�5 m ciglitazone. Acetazolamideand IL-4 proved to be potent (P � 0.01) inhibitors of resorp-tion stimulated by ciglitazone, whereas indomethacin andHU had no effect on 45Ca release caused by thethiazolidinedione.

The mRNA expression of RANKL, OPG, and RANK andprotein formation of RANKL and OPG in neonatal mousecalvariae treated with ciglitazone

In agreement with earlier studies (41), real-time quantita-tive PCR analysis revealed that treatment for 48 h with 10�8

m D3 increased expression (P � 0.01) of RANKL mRNA inneonatal mouse calvariae (Fig 3A). In addition, there was asignificantly increased (P � 0.01) mRNA expression ofRANKL after treatment of calvarial bones for 48 h with 10�5

m ciglitazone (Fig. 3A). However, although 10�8 m D3 alsocaused an increased mRNA expression (P � 0.01) of RANKin calvarial bones at 48 h, no significant change in RANKmRNA expression was observed in 10�5 m ciglitazone-treated calvariae (Fig. 3B). In contrast, when the mRNA ex-pression of OPG was evaluated, significant decreases (P �0.01) were noted after treatment with both 10�5 m ciglitazoneand the positive control, 10�8 m D3 (Fig. 3C).

Earlier studies (41) have shown that treatment with 10�8

m D3 increases RANKL protein formation, whereas decreas-ing OPG protein formation in neonatal mouse calvariae. Agood correlation between mRNA expression and proteinsynthesis was also noted with ciglitazone. Treatment of cal-varial bones for 48 h with 10�5 m ciglitazone resulted in asignificantly increased (P � 0.01) protein level of RANKLand a significantly decreased (P � 0.01) protein level for OPG(Fig. 3, D and E).

The mRNA expression of CTR, TRAP, cathepsin K, MMP-9,integrin �3, and NFAT2 in neonatal mouse calvarial bonestreated with ciglitazone or D3

Data showing that inhibitors of osteoclast formation suchas IL-4 can block calvarial bone resorption stimulated byciglitazone suggest that the resorption is dependent on os-teoclast differentiation. This was supported by semiquanti-tative RT-PCR analysis, which revealed that treatment ofmouse calvarial bones with either 10�5 m ciglitazone or 10�8

m D3 for 48 h would increase mRNA expression of theosteoclast markers integrin �3, TRAP, cathepsin K, MMP-9,NFAT2, and CTR (Fig. 4A). In Fig. 4 (B–D), real-time, quanti-tative PCR analysis showed that 10�5 m ciglitazone and 10�8 mD3 treatment of calvarial bones increased (P � 0.01) mRNAexpression of TRAP (B), cathepsin K (C), and NFAT2 (D).

FIG. 1. Concentration-dependent stimulation of 45Ca release fromneonatal mouse calvarial bones (A). Different concentrations of 15d-PG-J2, pioglitazone, troglitazone, and ciglitazone were added at timezero, and bones were cultured for 120 h. Values are based on dataobtained in three different experiments, with each point representingthe mean of 14–15 bones. To accumulate data from different exper-iments, the release of 45Ca in controls was set at 100%. Vertical bars,SEM. The release of 45Ca was significantly different from control re-lease in bones treated with 10�6 and 5 � 10�6 M troglitazone and 5 �10�6, 10�5, and 5 � 10�5 M ciglitazone. Time course (B) of 45Carelease from neonatal mouse calvarial bones treated with ciglitazone(10�5 M) and two different concentrations of PTH (10�10 and 10�8 M).The values are based on data obtained in two separate experiments,with each point representing the mean of 14–15 bones. Vertical bars,SEM. The releases of 45Ca caused by 10�10 M PTH and 10�5 M cigli-tazone were significantly (**, P � 0.01) enhanced at 72 and 120 h. Theresponse caused by 10�8 M PTH was significantly increased at 24, 72,and 120 h (**, P � 0.01).


Inhibitory effect of OPG on 45Ca release stimulated byciglitazone, troglitazone, and RANKL and mRNAexpression of osteoclast genes increased by ciglitazoneand D3

The decoy receptor, OPG, is thought to decrease osteoclas-togenesis by binding RANKL and decreasing activation ofRANK. OPG has been shown previously (40, 41) to be a goodinhibitor of 45Ca release stimulated by agents such as D3,

IL-11, and RANKL in neonatal mouse calvariae. In thepresent study, addition of OPG to cultured calvarial bonesfor 96 h caused inhibition of resorption stimulated by 10�5

m ciglitazone. The inhibition was concentration dependent,with statistical significance (P � 0.05) observed at 100 and 300ng/ml OPG (Fig. 5A). OPG (300 ng/ml) also caused signif-icant (P � 0.01) inhibition of 45Ca release stimulated by 10�6

m troglitazone (Fig. 5B) and RANKL (200 ng/ml; Fig. 5C)

FIG. 2. Stimulation of percent 45Ca re-lease at 24–144 h by 10�5 M ciglitazoneand 10�8 M PTH in the absence and pres-ence of 10�9 M CT and 10�4 M AHPrBP(A). Data are derived from one experi-ment and based on means of six to eightbones. Vertical bars, SEM. After precul-ture of bones in indomethacin as de-scribed in Materials and Methods,bones were preincubated for an addi-tional 24 h in control medium or me-dium containing either PTH (10�8 M) orciglitazone (10�5 M) to initiate bone re-sorption. After this prestimulation pe-riod, CT and AHPrBP were added, andsmall aliquots of media were removedat 24–144 h for evaluation of 45Ca re-lease. Control release of 45Ca was sig-nificantly (P � 0.01) different from re-lease in ciglitazone- and PTH-treatedcalvariae at 48–144 h. Significant inhi-bition (P � 0.01) of ciglitazone and PTHtreatment was noted in the presence ofAHPrBP at 48–144 h. Significant inhi-bition (P � 0.01) of ciglitazone and PTHtreatment was also noted in the pres-ence of CT at 48–144 h. However, al-though the magnitude of inhibitioncaused by AHPrBP increased from 48–144 h, inhibition was not as pro-nounced, with escape occurring in cal-varial bones treated with CT for 48–144h. Effects on 45Ca release caused by ad-dition of 10�4 M acetazolamide (B), 10�6

M indomethacin (C), 10�3 M HU (D), and10 ng/ml IL-4 (E) to neonatal mouse cal-varial bones treated with 10�5 M cigli-tazone, 10�8 M PTH, or 10�8 M D3 for120 h. Values are based on data ob-tained in two experiments with each in-hibitor, with each point representingthe mean of 10–14 bones. To accumu-late data from different experiments,the release of 45Ca in controls was set at100%. Vertical bars, SEM. Acetazol-amide (B) and IL-4 (E) significantly(P � 0.01) inhibited the increases (P �0.01) in 45Ca release stimulated by cigli-tazone, PTH, and D3, but inhibition wasnot noted when either indomethacin (C)or HU (D) were added to calvariaetreated with ciglitazone, PTH, and D3.


When the ability of OPG (300 ng/ml) to block gene expres-sion stimulated by 10�5 m ciglitazone for 48 h was evaluated,decreased mRNA expression of integrin �3, TRAP, cathepsin K,CTR, NFAT2, MMP-9, and RANKL was noted (Fig. 5D). Similar

results were observed after stimulation of gene expression byRANKL (200 ng/ml) for 48 h. OPG (300 ng/ml) decreasedmRNA expression of integrin �3, TRAP, cathepsin K, CTR,NFAT2, and MMP-9 (Fig. 5E) stimulated by RANKL.

FIG. 3. The mRNA expression of RANKL (A),RANK (B), and OPG (C) and ELISA analysis of pro-tein formation of RANKL (D) and OPG (E) in normal(control) calvarial bones and mouse calvariaetreated with 10�5 M ciglitazone and 10�8 M D3. Forthe mRNA expression of RANKL (A), RANK (B), andOPG (C), data are derived from means of four to fivebones treated for 48 h. Total RNA was extractedfrom each bone and cDNA synthesized. Real-time,quantitative PCR analysis was performed usingcDNA from individual bones. RANKL, RANK, andOPG levels were normalized to �-actin. The effectsof ciglitazone and D3 were calculated by setting theratios obtained in control bones to 100%. Verticalbars, SEM. Ciglitazone had statistically significant(P � 0.01) effects on the mRNA expression ofRANKL and OPG. D3 had statistically significant(P � 0.01) effects on the mRNA expression ofRANKL, RANK, and OPG. For the ELISA analysisof RANKL (A) and OPG (B) protein, calvarial boneswere cultured for 48 h in the absence and presenceof 10�5 M ciglitazone and 10�8 M D3. After culture,calvarial bones were extracted with 0.2% TritonX-100 for 24 h at room temperature. Values repre-sent the means of eight calvarial halves per group.Vertical bars, SEM. Ciglitazone and D3 increased(P � 0.01) RANKL protein formation and decreased(P � 0.01) OPG protein formation.


Effects of ciglitazone on cytokine expression

IL-1, TNF, and members of the IL-6 family of cytokines(IL-6, IL-11, LIF, and OSM) are good stimulators of boneresorption (40–42) with actions that can suppress PPAR�activity (43, 44). Therefore, experiments were performed todetermine whether bone resorption stimulated by ciglita-zone might be indirect, due to increased expression ofcytokines.

No change in the mRNA expression of either IL-6 or IL-11was observed in three independent experiments testing cigli-tazone (Fig. 6A). Similarly, expression of mRNA for LIF andOSM was unaffected by ciglitazone treatment for 48 h (Fig.6A). In contrast, the mRNA expression of TNF-� was de-creased and that of IL-1� increased in three experiments (Fig.6A).

To determine whether IL-1� might be responsible for re-sorption stimulated by ciglitazone, calvarial bones were pre-incubated with IRAP for 3 h before addition of either 10�5 mciglitazone or IL-1� for 96 h (45). IRAP blocked 45Ca releasestimulated by IL-1� (data not shown) but not release stim-ulated by ciglitazone (Fig. 6B).

Effects of ciglitazone on adipogenic and osteoblasticgene expression

PPAR� increases adipogenesis and may suppress osteo-blastogenesis (15). In calvarial bones treated for 48 h with10�5 m ciglitazone, decreased mRNA expression for bothPPAR�1 and PPAR�2 was found (Fig. 7). This was in contrastto treatment with 10�8 m D3, which had no effect on themRNA expression of PPAR�1, but increased that of PPAR�2(Fig. 7). When looking at other genes involved in adipogen-esis, it was found that C/EBP� (46) was not affected by eitherciglitazone or D3 treatment but that mRNA expression ofWnt10b (47) was decreased by both agents (Fig. 7). Evalua-tion of osteocalcin, a gene involved in osteoblast differenti-ation (48), revealed that mRNA expression was increased byboth ciglitazone and D3 (Fig. 7).

Inability of GW 9662 to inhibit bone resorption stimulatedby ciglitazone

Extensive use has been made of GW 9662, an irreversibleinhibitor of PPAR�, to determine whether responses of var-ious ligands are PPAR� independent or dependent (25, 38,49, 50). Figure 8 shows that treatment of calvarial bones with10�5 and 10�6 m GW 9662 had no effect on basal resorption,resorption stimulated by 10�5 and 3 � 10�6 m ciglitazone, orresorption stimulated by 10�8 m PTH for 96 h. These resultssuggest that bone resorption stimulated by ciglitazone occursby a PPAR�-independent mechanism.

Discussion

Three thiazolidinediones (ciglitazone, troglitazone, andpioglitazone) and a natural PPAR� ligand, 15d-PG-J2, wereevaluated for their ability to stimulate 45Ca release fromprelabeled neonatal mouse calvarial bones. There were nochanges in 45Ca release noted when either 15d-PG-J2 or pio-glitazone was tested, but increases in 45Ca release were foundafter treatment with both ciglitazone and troglitazone. Re-lease of 45Ca after treatment of calvarial bones with ciglita-zone and troglitazone was biphasic, with peak release of eachdrug (5 � 10�6 m for troglitazone and 10�5 m for ciglitazone)depressed at increased concentrations. This decrease in re-lease may have been due to toxicity of the compounds be-cause it was found that only a very small amount of RNAcould be extracted from calvarial bones treated with 10�4 mciglitazone.

The greatest increase of 45Ca release in calvarial bones wasobserved with ciglitazone, the prototype thiazolidinedione(51). Resorption stimulated by ciglitazone was blocked byseveral well-documented inhibitors of osteoclastic bone re-sorption, the carbonic anhydrase inhibitor, acetazolamide(52), the bisphosphonate, AHPrBP (53), the polypeptide hor-mone, CT (54), and IL-4 (26, 55). Inhibition by these agents

FIG. 4. Semiquantitative RT-PCR analysis showing mRNA expres-sion of integrin �3, TRAP, cathepsin K, MMP-9, NFAT2, and CTRafter treatment with 10�5 M ciglitazone and 10�8 M D3 for 48 h (A).Real-time, quantitative PCR analysis showing the effects of 10�5 Mciglitazone and 10�8 M D3 on mRNA expression of TRAP (B), cathep-sin K (C), and NFAT2 (D) in calvarial bones treated for 48 h. Levelsof the markers were normalized to �-actin. The effects of ciglitazoneand D3 were calculated by setting the ratios obtained in control bonesto 100%. Vertical bars, SEM. Ciglitazone and D3 significantly en-hanced (**, P � 0.01) the mRNA expression of TRAP, cathepsin K, andNFAT2.


suggests that the resorptive action of ciglitazone in neonatalmouse calvarial bones was mediated by stimulation of os-teoclast differentiation and activity.

Unlike the sustained inhibition of bone resorption foundwith acetazolamide, AHPrBP, and IL-4, the inhibition by CTwas transient. This is characteristic of the escape phenome-non (56, 57) known to occur with CT.

Indomethacin had no effect on 45Ca release stimulated byciglitazone, suggesting that the resorptive effect of ciglita-zone was not dependent on prostaglandin biosynthesis, a

view supported by the observation that ciglitazone did notenhance mRNA expression of cyclooxygenase-2 (data notshown). In addition to inhibiting prostaglandin production,indomethacin can serve as a PPAR� ligand (12). However,the concentration of indomethacin needed for PPAR� acti-vation is significantly higher (58) than the concentration(10�6 m) employed to inhibit cyclooxygenase activity (40, 59).

A mitotic inhibitor, HU, was also found to have no effecton the release of 45Ca in bone explants treated with ciglita-zone. The resorptive actions of agents such as the thyroid

FIG. 5. Dose response of the inhibitory effect ofOPG on 45Ca release stimulated by ciglitazone inmouse calvarial bones (A). Different concentrationsof OPG were incubated with 10�5 M ciglitazone for120 h. Significant inhibition (**, P � 0.05) of 45Carelease was observed at 100 and 300 ng/ml OPG.Each point represents the mean of six bones. Ver-tical bars, SEM. OPG (300 ng/ml) inhibited (**, P �0.01) 45Ca release stimulated by 10�6 M troglitazone(B) and 200 ng/ml RANKL (C) in neonatal mousecalvarial bones. OPG was incubated with troglita-zone and RANKL for 96 h. Values are based on sixbones per group. Vertical bars, SEM. Semiquantita-tive RT-PCR analysis of calvarial bones showingthat 300 ng/ml OPG decreased mRNA expression ofRANKL stimulated by 10�5 M ciglitazone (D) for48 h and the osteoclast markers (integrin �3, TRAP,cathepsin K, MMP-9, NFAT2, and CTR) stimulatedby 10�5 M ciglitazone (D) and 200 ng/ml RANKL (E)for 48 h.


hormones (60) and transforming growth factor-� (61) areinhibited by mitotic inhibitors in mouse calvariae. In con-trast, osteolytic compounds like PTH and D3 can stimulatecalvarial bone resorption by mechanisms that are not de-pendent on cell replication (62). Thus, ciglitazone may bestimulating resorption by enhancing differentiation and/orfusion of mononuclear preosteoclasts to terminally differen-tiated multinucleated osteoclasts at a postmitotic step inmouse calvarial bones.

To evaluate the possible involvement of cytokines in theresorptive action of ciglitazone, mRNA expression of IL-1�,TNF-�, IL-6, IL-11, LIF, and OSM was determined after cigli-tazone treatment. TNF-� expression decreased, and nochanges in mRNA expression of IL-6 cytokines (IL-6, IL-11,LIF, and OSM) were noted, but ciglitazone increased mRNAof IL-1�. However, 45Ca release stimulated by ciglitazone

was not affected by an inhibitor (IRAP) of IL-1�, suggestingthat IL-1� was not involved in the resorption stimulated byciglitazone.

Constitutive expression of RANKL, OPG, and RANKmRNA was observed in the calvarial bones. No change inRANK mRNA expression was noted after ciglitazone treat-ment of calvariae, but the thiazolidinedione was found toincrease RANKL and decrease OPG mRNA expression in

FIG. 6. Semiquantitative RT-PCR analysis of mRNA expression ofIL-6, IL-11, LIF, OSM, TNF-�, and IL-1� in calvarial bones treatedfor 48 h with 10�5 M ciglitazone (A). Ciglitazone exposure had no effecton mRNA of IL-6 family cytokines (IL-6, IL-11, LIF, and OSM) butdecreased mRNA of TNF-� and increased mRNA expression of IL-1�(A). The lack of an effect by IRAP, the IL-1 receptor antagonist, on45Ca release (**, P � 0.01) stimulated by 10�5 M ciglitazone (B). Aftertreatment with 100 ng/ml IRAP for 3 h, calvarial bones were culturedfor an additional 96 h with 100ng/ml IRAP and 10�5 M ciglitazone.Values are based on six bones per group. Vertical bars, SEM.

FIG. 7. Semiquantitative RT-PCR analysis of the mRNA expressionof adipogenic markers (PPAR�1, PPAR�2, Wnt10b, and C/EBP�) andthe osteogenic marker, osteocalcin, after treatment for 48 h with 10�5

M ciglitazone and 10�8 M D3. Ciglitazone decreased mRNA expressionof PPAR�1, PPAR�2, and Wnt10b, had no effect on C/EBP�, andincreased expression of osteocalcin. D3 decreased mRNA expressionof Wnt10b, had no effect on the mRNA expression of C/EBP� orPPAR�1, and increased expression of PPAR�2 and osteocalcin.

FIG. 8. The inability of GW 9662, an irreversible inhibitor of PPAR�,to alter resorption in normal (control) calvarial bones and calvarialbones treated with either ciglitazone or PTH for 96 h. Data are derivedfrom means of five to six bones The effects of ciglitazone and PTH werecalculated by setting the ratios obtained in control bones to 100%.Vertical bars, SEM. GW 9662 (10�5 and 10�6 M) had no statistical effecton control resorption, resorption stimulated by 3 � 10�6 and 10�5 Mciglitazone, or resorption stimulated by 10�8 M PTH.


calvarial explants. Furthermore, analyses revealed an in-crease in RANKL protein and a decrease in OPG protein afterexposure to ciglitazone. The increased RANKL to OPG ratioobserved after ciglitazone treatment provides an explanationfor why the thiazolidinedione was found to be a good stim-ulator of calvarial bone resorption.

At the time of dissection, low numbers of osteoclasts werepresent in the 6- to 7-d-old calvariae used for study. Thesewere lost during the preculture period when the calvarialbones are exposed to basic medium and indomethacin. Dur-ing subsequent treatment with osteoclastic stimuli, enhanced45Ca release is due to increased osteoclast formation frommononucleated osteoclast progenitor cells (Lerner, U. H., P.Lundberg, and M. Ransjo, unpublished data). In addition tostimulating expression of RANKL, ciglitazone increased ex-pression of the osteoclast markers integrin �3 (63, 64), TRAP(65, 66), cathepsin K (67), CTR (54), NFAT2 (68), and MMP-9(69). RANKL also increased expression of integrin �3, TRAP,cathepsin K, CTR, NFAT2, and MMP-9 in calvariae. In-creased expression of the osteoclast markers by ciglitazoneand RANKL is evidence that both agents stimulate osteoclastdifferentiation and function in calvariae and supports therole of RANKL in the stimulation of osteoclast differentiationand function by ciglitazone. Moreover, the importance of theRANKL-OPG-RANK system was further emphasized by ob-servations showing that the stimulatory effects of troglita-zone and ciglitazone on bone resorption, the stimulation ofRANKL by ciglitazone, and the stimulation of osteoclastmarkers (integrin �3, TRAP, cathepsin K, CTR, NFAT2 andMMP-9) by ciglitazone and RANKL were all inhibited byexogenous OPG.

In the murine marrow-derived mesenchymal progenitorcell line U-33/�2, treatment with either a thiazolidinedione,rosiglitazone, or natural ligands, 9-hydroxyoctadecadienoicacid and 15d-PG-J2, stimulates differentiation to adipocytes,while blocking differentiation to osteoblasts (15). To bettercharacterize how ciglitazone stimulates resorption, mRNAexpression of adipogenic and osteogenic markers was eval-uated in calvariae after ciglitazone treatment. For the adi-pogenic markers, mRNA of Wnt10b (47), PPAR�1 andPPAR�2 were decreased, but there was no change in expres-sion of C/EBP�, a stimulator of terminal adipocyte differ-entiation (46). Decreased expression of PPAR� may be suf-ficient to promote osteoblastogenesis (70) and the decrease inPPAR� mRNA and lack of an effect on C/EBP� noted withciglitazone, coupled with stimulation of mRNA for osteo-calcin, an important marker of osteoblast differentiation (48),support the contention that the thiazolidinedione increasesosteoblastogenesis in calvarial bones.

There are numerous examples of non-PPAR�-dependentactions of natural and synthetic PPAR� ligands (30–38). Inthe present study, several lines of evidence suggest that theincreases in resorption stimulated by troglitazone and cigli-tazone in calvarial bones are not PPAR�-dependent effects.First, in dose-response experiments where troglitazone andciglitazone were found to be stimulators of 45Ca release, thenatural PPAR� ligand, 15d-PG-J2, and another thiazo-lidinedione, pioglitazone, had no effect on calvarial boneresorption. Second, stimulation of PPAR� can increase adi-pogenesis, but the effects of ciglitazone on mRNA of adipo-

genic and osteogenic markers in calvarial bones suggest thatciglitazone stimulates osteoblast differentiation. Third, se-lective activation of PPAR� is thought to occur when thia-zolidinedione drugs stimulate PPAR receptors, but the irre-versible PPAR� antagonist GW 9662 (40) had no effect onbone resorption stimulated by ciglitazone.

The reason why stimulation rather than inhibition of os-teoclast-mediated resorption was observed with ciglitazoneand troglitazone in the present study appears to be due to theability of calvarial osteoblasts to produce a RANKL to OPGratio after treatment that favors resorption, together withpossible differences in periosteal osteoclast precursor cellsand the less differentiated osteoclast precursor cells found inbone marrow and spleen cell cultures. It is possible that thelack of an inhibitory effect by either ciglitazone or troglita-zone is due to differences in expression of PPAR�. Prelim-inary analysis has indicated that expression of PPAR�1mRNA is similar in calvariae and mouse spleen cells afterciglitazone treatment but that expression of PPAR�2 mRNAis less in calvariae in comparison with the hematopoietic cells(Lerner, U. H., and H. H. Conaway, unpublished data).

Animal studies have suggested that bone loss can resultfrom thiazolidinedione use. In agreement with our observa-tion of ciglitazone and troglitazone stimulating bone resorp-tion, an in vivo study in rats (29) has found no evidence forrosiglitazone to inhibit bone resorption in intact animals,whereas in ovariectomized rats, treatment with the thiazo-lidinedione caused significant stimulation of bone resorp-tion, accompanied by decreased bone mass in the tibia, fe-mur, and lumbar spine of the animals. Furthermore,although no effect on trabecular bone volume has been notedafter troglitazone administration in mice (71), several studieshave shown significant bone loss after treatment of mice withrosiglitazone (72–75). Decreased bone formation has usuallybeen suggested as the reason for the skeletal loss that occursin mice treated with rosiglitazone, but a recent report hasindicated that increased bone resorption can also play aprominent role (75). In humans, results have been mixed,with thiazolidinedione treatment being suggested to bothprotect against bone loss (76) and increase bone loss (77).Watanabe et al. (76) have reported that troglitazone treatmentof type 2 diabetics for 1 yr decreases serum leptin and pre-vents bone loss. In contrast, results from the Health, Aging,and Body Composition cohort study (77) have suggested thatthiazolidinedione (troglitazone, rosiglitazone, and pioglita-zone) treatment of type 2 diabetic patients for longer than 24months decreases bone mineral density of the femoral neckand total hip. It is possible that the differences noted afterthiazolidinedione treatment in humans are related to thedifferent durations of treatment, but additional studies willbe necessary to clarify this point.

Acknowledgments

The authors gratefully acknowledge the technical help of Mrs. IngridBostrom, Inger Lundgren, and Birgit Andertun.

Received May 18, 2005. Accepted June 21, 2005.Address all correspondence and requests for reprints to: Howard H.

Conaway, Ph.D., Department of Physiology and Biophysics, Universityof Arkansas for Medical Sciences, 4301 West Markham Street, Slot 505,Little Rock, Arkansas 72205. E-mail: [email protected].


This project was supported by grants from the Swedish Science Coun-cil (Project 07525), the Swedish Rheumatism Association, the Royal 80Year Fund of King Gustav V, and the County Council of Vasterbottenand Salus Ansvar.

References

1. Aranda A, Pascual A 2001 Nuclear hormone receptors and gene expression.Phys Rev 81:1269–1304

2. Isseman I, Green S 1990 Activation of a member of the steroid hormonereceptor superfamily by peroxisome proliferators. Nature 347:645–650

3. Graves RA, Tontonoz P, Spiegelman BM 1992 Analysis of a tissue specificenhancer: ARF6 regulates adipogenic gene expression. Mol Cell Biol 12:1202–1208

4. Dreyer C, Krey G, Keller H, Givel F, Helftenbein G, Wahli W 1992 Controlof the peroxisomal �-oxidation pathway by a novel family of nuclear hormonereceptors. Cell 68:879–887

5. Schmidt A, Endo N, Rutledge SJ, Vogel R, Shinar D, Rodan GA 1992 Iden-tification of a new member of the steroid hormone receptor superfamily thatis activated by a peroxisome proliferator and fatty acids. Mol Endocrinol6:1634–1641

6. Auboeuf D, Rieusset J, Fajas L, Vallier P, Frering V, Riou JP, Staels B,Auwerx J, Laville M, Vidal H 1997 Tissue distribution and quantification ofthe expression of PPARs and LXRs in humans: no alterations in adipose tissueof obese and NIDDM patients. Diabetes 46:1319–1327

7. Kliewer SA, Sundseth SS, Jones SA, Brown PJ, Wisely GB, Koble CS,Devchand P, Wahli W, Willson TM, Lenhard JM, Lehmann JM 1997 Fattyacids and eicosanoids regulate gene expression through direct interactionswith peroxisome proliferator-activated receptors � and �. Proc Natl Acad SciUSA 94:4318–4323

8. Yu K, Bayona W, Kallen CB, Harding HP, Ravera CP 1995 Differential ac-tivation of peroxisome proliferator-activated receptors by eicosanoids. J BiolChem 270:23975–23983

9. Nagy L, Tontonez P, Alvarez JG, Chen H, Evans RM 1998 Oxidized LDLregulates macrophage gene expression through ligand activation of PPAR�.Cell 93:229–240

10. Forman BM, Tontonez P, Chen J, Brun RP, Spiegelman BM, Evans RM 199515-Deoxy-�12,14-prostaglandin J2 is a ligand for the adipocyte determinationfactor PPAR�. Cell 83:803–812

11. Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson TM,Kliewer SA 1995 An antibiotic thiazolidinedione is a high affinity ligand forperoxisome proliferator-activated receptor � (PPAR�). J Biol Chem 270:12953–12956

12. Lehmann JM, Lenhard JM, Oliver BB, Ringold GM, Kliewer SA 1997 Per-oxisome proliferator-activated receptors � and � are activated by indomethacinand other non-steroidal anti-inflammatory drugs. J Biol Chem 272:3406–3410

13. Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD,Moorman MA, Simonetti DW, Craig S, Marshak DR 1999 Multilineage po-tential of adult human mesenchymal stem cells. Science 284:143–147

14. Prockop DJ 1997 Marrow stromal cells as stem cells for nonhematopoietictissues. Science 276:71–74

15. Lecka-Czernik B, Gubrij I, Moerman EJ, Kajkenova O, Lipschitz DA, Mano-lagas SC, Jilka RL 1999 Inhibition of Osf2/Cbfa1 expression and terminalosteoblast differentiation by PPAR�2. J Cell Biochem 74:357–371

16. Teitelbaum SL, Ross FP 2003 Genetic regulation of osteoclast developmentand function. Nat Rev Genet 4:638–649

17. Takahashi N, Udagawa N, Takami M, Suda T 2002 Cells of bone: osteoclastgeneration. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bonebiology. 2nd ed. San Diego: Academic Press; 109–126

18. Boyle WJ, Simonet WS, Lacey DL 2003 Osteoclast differentiation and acti-vation. Nature 423:337–342

19. Lerner UH 2004 New molecules in the tumor necrosis factor ligand andreceptor superfamilies with importance for physiological and pathologicalbone resorption. Crit Rev Oral Biol Med 15:64–81

20. Kong Y-Y, Yoshida H, Sarosi I, Tan H-L, Timms E, Capparelli C, Morony S,Oliveira-dos-Santos AJ, Van G, Itie A, Khoo W, Wakeham A, Dunstan CR,Lacey DL, Mak TW, Boyle WJ, Penninger JM 1999 OPGL is a key regulatorof osteoclastogenesis, lymphocyte development and lymph-node organogen-esis. Nature 397:315–323

21. Dougall WC, Glaccum M, Charrier K, Rohrbach K, Brasel K, De Smedt T,Daro E, Smith J, Tometsko ME, Maliszewski CR, Armstrong A, Shen V, BainS, Cosman D, Anderson D, Morrissey PJ, Peschon JJ, Schuh J 1999 RANK isessential for osteoclast and lymph node development. Genes Dev 13:2412–2424

22. Li J, Sarosi I, Yan X-Q, Morony S, Capparelli C, Tan H-L, McCabe S, ElliottR, Scully S, Van G, Kaufman S, Juan S-C, Sun Y, Tarpley J, Martin L,Christensen K, McCabe J, Kostenuik P, Hsu H, Fletcher F, Dunstan CR,Lacey DL, Boyle WJ 2000 RANK is the intrinsic hematopoietic cell surfacefactor that controls osteoclastogenesis and regulation of bone mass and cal-cium metabolism. Proc Natl Acad Sci USA 97:1566–1571

23. Yoshida H, Hayashi S, Kunisada T, Ogawa M, Nishikawa S, Okamura H,Sudo T, Shultz D, Nishikawa S 1990 The murine mutation osteopetrosis is in

the coding region of the macrophage colony stimulating factor gene. Nature345:442–444

24. Bucay N, Sarosi I, Dunstan CR, Morony S, Tarpley J, Capparelli C, ScullyS, Tan HL, Xu W, Lacey DL, Boyle WJ, Simonet WS 1998 Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification.Genes Dev 12:1260–1268

25. Mbalaviele G, Abu-Amer Y, Meng A, Jaiswal R, Beck S, Pittenger MF,Thiede MA, Marshak DR 2000 Activation of peroxisome proliferator-acti-vated receptor-� pathway inhibits osteoclast differentiation. J Biol Chem 275:14388–14393

26. Bendixen AC, Shevde NK, Dienger KM, Willson TM, Funk, CD, Pike JW2001 IL-4 inhibits osteoclast formation through a direct action on osteoclastprecursors via peroxisome proliferator-activated receptor �1. Proc Natl AcadSci USA 98:2443–2448

27. Okazaki R, Toriumi M, Fukumoto S, Miyamoto M, Fujita T, Tanaka K,Takeuchi Y 1999 Thiazolidinediones inhibit osteoclast-like cell formation andbone resorption in vitro. Endocrinology 140:5060–5065

28. Kelly KA, Tanaka S, Baron R, Gimble JM 1998 Murine bone marrow stro-mally derived BMS2 adipocytes support differentiation and function of oste-oclast-like cells in vitro. Endocrinology 139:2092–2101

29. Sottile V, Seuwen K, Kneissel M 2004 Enhanced marrow adipogenesis andbone resorption in estrogen-deprived rats treated with the PPAR� agonistBRL49653 (rosiglitazone). Calcif Tissue Int 75:329–337

30. Hunter JG, van Delft MF, Rachubinski RA, Capone JP 2001 Peroxisomeproliferator-activated receptor � ligands differentially modulate cell differen-tiation and MyoD gene expression via peroxisome proliferator-activated re-ceptor �-dependent and -independent pathways. J Biol Chem 276:38297–38306

31. Jozkowicz A, Dulak J, Prager M, Nanobashvili J, Nigisch A, Winter B, WeigelG, Huk I 2001 Prostaglandin-J2 induces synthesis of interleukin-8 by endo-thelial cells in a PPAR� independent manner. Prostaglandins Other LipidMediat 66:167–177

32. Wilmer WA, Dixon C, Lu L, Hilbelink T, Rovin BH 2001 A cyclopentenoneprostaglandin activates mesangial MAP kinase independently of PPAR�. Bio-chem Biophys Res Commun 281:57–62

33. Ward JE, Haslinda G, Harris T, Bonacci JV, Stewart AG 2004 PPAR� ligands,15-deoxy-�12,14 prostaglandin J2 and rosiglitazone regulate human culturedairway smooth muscle proliferation through different mechanisms. Br J Phar-macol 141:517–525

34. Davies GF, Khandelwal RL, Wu L, Juurlink BH, Roesler WJ 2001 Inhibitionof phosphoenolpyruvate carboxykinase (PEPCK) gene expression by trogli-tazone: a peroxisome proliferator-activated receptor � (PPAR�)-independent,antioxidant-related mechanism. Biochem Pharmacol 62:1071–1079

35. Wang M, Wise SC, Leff T, Su TZ 1999 Troglitazone, an antidiabetic agent,inhibits cholesterol biosynthesis through a mechanism independent of per-oxisome proliferator-activated receptor �. Diabetes 48:254–260

36. Lennon AM, Ramauge M, Dessouroux A, Pierre M 2002 MAP kinase cascadesare activated in astrocytes and preadipocytes by 15-deoxy-�12,14-prostaglan-din J2 and the thiazolidinedione ciglitazone through peroxisome proliferatoractivator receptor �-independent mechanisms involving reactive oxygenatedspecies. J Biol Chem 277:29681–29685

37. Parameswaran N, Hall CS, Bomberger JM, Sparks HV, Jump DB, SpielmanWS 2003 Negative growth effects of ciglitazone on kidney interstitial fibro-blasts: role of PPAR�. Kidney Blood Press Res 26:2–9

38. Zhu X, Lin Y, Zhang J, Fu M, Mao Z, Chen YE 2003 Thiazolidinediones, a classof anti-diabetic drugs, inhibit Id2 expression through a PPAR�-independentpathway in human aortic smooth muscle cells. Cell Mol Life Sci 60:212–218

39. Lerner UH 1987 Modifications of the mouse calvarial technique improve theresponsiveness to stimulators of bone resorption. J Bone Miner Res 2:375–383

40. Ahlen J, Anderson S, Mukohyama H, Roth C, Backman A, Conaway HH,Lerner UH 2002 Characterization of the bone-resorptive effect of interleukin-11in cultured mouse calvarial bones. Bone 31:242–251

41. Palmqvist P, Persson E, Conaway HH, Lerner UH 2002 IL-6, leukemia in-hibitory factor, and oncostatin M stimulate bone resorption and regulate theexpression of receptor activator of NF-�B ligand, osteoprotegerin, and receptoractivator of NF-�B in mouse calvariae. J Immunol 169:3353–3362

42. Horowitz J, Lorenzo JA 2002 Local regulators of bone: IL-1, TNF, lymphotoxin,interferon-g, IL-8, IL-10, IL-4, the LIF/IL-6 family, and additional cytokines. In:Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. 2nd ed. SanDiego: Academic Press; 961–978

43. Tanaka T, Itoh H, Doi K, Fukunaga Y, Hosoda K, Shintani M, Yamashita J,Chun T-H, Inoue M, Masatsugu K, Sawada N, Saito T, Inoue G, NishimuraH, Yoshimasa Y, Nakao K 1999 Down regulation of peroxisome proliferator-activated receptor� expression by inflammatory cytokines and its reversal bythiazolidinediones. Diabetologia 42:702–710

44. Suzawa M, Takada I, Yanagisawa J, Ohtake F, Ogawa S, Yamauchi T, Kad-owaki T, Takeuchi Y, Shibuya H, Gotoh Y, Matsumoto K, Kato S 2003Cytokines suppress adipogenesis and PPAR� function through the TAK1/TAB1/NIK cascade. Nat Cell Biol 5:224–230

45. Ohlin A, Sjogren U, Lerner UH 1999 Bone resorbing activity released fromzymosan-activated mouse peritoneal macrophages: the role of prostanoids andinterleukin-1. J Infl Res 48:181–192

46. Wu Z, Rosen ED, Brun R, Hauser S, Adelmont G, Troy A E, McKeon C,


Darlington GJ, Spiegelman BM 1999 Cross-regulation of C/EBP� and PPAR�controls the transcriptional pathway of adipogenesis and insulin sensitivityMol Cell 3:151–158

47. Bennett CN, Longo KA, Wright WS, Suva, LJ, Lane TF, Hankenson, KD,MacDougald OA 2005 Regulation of osteoblastogenesis and bone mass byWnt10b. Proc Natl Acad Sci USA 102:3324–3329

48. Garcia T, Roman-Roman S, Jackson A, Theilhaber J, Connolly T, Spinella-Jaegle S, Kawai S, Courtois B, Bushnell S, Aberval M, Call K Baron R 2002Behavior of osteoblast, adipocyte, and myoblast markers in genome-wideexpression analysis of mouse calvaria primary osteoblasts in vitro. Bone 31:205–211

49. Willson TM, Brown PJ, Sternbach DD, Henke BR 2000 The PPARs: fromorphan receptors to drug recovery. J Med Chem 43:527–550

50. Huang JT, Welch JS, Ricote M, Binder CJ, Willson TM, Kelly C, Witztum JL, Funk CD, Conrad D, Glass CK 1999 Interleukin-4-dependent productionof PPAR-� ligands in macrophages by 12/15-lipoxygenase. Nature 400:378–382

51. Willson TM, Cobb JE, Cowan DJ, Wiethe RW, Correa ID, Prakash SR, BeckKD, Moore LB, Kliewer SA, Lehmann JM 1996 The structure-activity rela-tionship between peroxisome proliferator-activated receptor � agonism andthe antihyperglycemic activity of thiazolidinediones. J Med Chem 99:665–668

52. Vaananen K, Zhao H 2002 Osteoclast function. Biology and mechanisms. In:Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. 2nd ed. SanDiego: Academic Press; 127–140

53. Beck Ev, Pieterman E, Cohen L, Lowik C, Papapoulos S 1999 Nitrogen-containing bisphosphonates inhibit isopentenyl pyrophosphate isomerase/farnesyl pyrophosphate synthase activity with relative potencies correspond-ing to their antiresorptive potencies in vitro and in vivo. Biochem Biophys ResCommun 255:491–494

54. Nicholson GC, Livesey SA, Moseley JM, Martin TJ 1986 Actions of calcitonin,parathyroid hormone, and prostaglandin E2 on cyclic AMP formation inchicken and rat osteoclasts. J Cell Biochem 31:229–241

55. Lerner UH, Conaway HH Interleukin-4 and -13 inhibit bone resorption andexpression of RANK in mouse calvarial bones. American Society of Bone andMineral Research Annual Meeting, 2000, p 387 (Abstract SU230)

56. Wener JA, Gorton SJ, Raisz LG 1972 Escape from inhibition of resorption incultures of fetal rat bone treated with calcitonin and parathyroid hormone.Endocrinology 90:752–759

57. Tashjian AH, Wright DR, Ivey JL, Pont A 1978 Calcitonin binding sites inbone: relationship to biological response and escape. Recent Prog Horm Res34:285–334

58. Willson TM, Wahli W 1997 Peroxisome proliferator-activated receptor ago-nists. Curr Opin Chem Biol 1:235–241

59. Pilbeam C, Harrison JR, Raisz LG 2002 Prostaglandins and bone metabolism.In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. 2nd ed.San Diego: Academic Press; 979–994

60. Conaway HH, Ransjo M, Lerner UH 1998 Prostaglandin-independent stim-ulation of bone resorption in mouse calvariae and in isolated rat osteoclasts bythyroid hormones (T4 and T3). Proc Soc Exp Biol Med 217:153–161

61. Lerner UH 1996 Transforming growth factor-� stimulates bone resorption inneonatal mouse calvariae by a prostaglandin unrelated but proliferation de-pendent pathway. J Bone Miner Res 11:1628–1639

62. Lerner UH, Hanstrom L 1989 Stimulation and inhibition of mitotic activity in

cultured calvarial bones by cyclic AMP analogues and phosphodiesteraseinhibitors is unrelated to the delayed resorptive effect of cyclic AMP. Exp ClinEndocrinol 8:89–95

63. Hynes RO 2002 Integrins: bidirectional, allosteric signaling machines. Cell110:673–687

64. Inoue M, Namba N, Chappel J, Teitelbaum SL, Ross FP 1998 Granulocytemacrophage-colony stimulating factor reciprocally regulates �v-associated in-tegrins on murine osteoclast precursors. Mol Endocrinol 12:1955–1962

65. Hayman AR, Jones SJ, Boyde A, Foster D, Colledge WH, Carlton MB, EvansMJ, Cox TM 1996 Mice lacking tartrate-resistant acid phosphatase (Acp 5) havedisrupted endochondral ossification and mild osteopetrosis. Development122:3151–3162

66. Angel NZ, Walsh N, Forwood MR, Ostrowski MC, Cassady AI, Hume DA2000 Transgenic mice overexpressing tartrate-resistant acid phosphatase ex-hibit an increased rate of bone turnover. J Bone Miner Res 15:103–106

67. Saftig P, Hunziker E, Wehmeyer O, Jones S, Boyde A, Rommerskirch W,Moritz JD, Schu P, von Figura K 1998 Impaired osteoclastic bone resorptionleads to osteopetrosis in cathepsin K-deficient mice. Proc Natl Acad Sci USA95:13453–13458

68. Takayanagi H, Kim S, Koga T, Nishina H, Isshiki M, Yoshida H, Saiura A,Isobe M, Yokochi T, Inoue JI, Wagner EF, Mak TW, Kodama T, TaniguchiT 2002 Induction and activation of the transcription factor NFATc1 (NFAT2)integrate RANKL signaling in terminal differentiation of osteoclasts. Dev Cell3:889–901

69. Lerner UH 2000 Osteoclast formation and resorption. Matrix Biol 19:107–12070. Akune T, Ohba S, Kamekura S, Yamaguchi M, Chunk U-i, Kubota N,

Terauchi Y, Harada Y, Azuma Y, Nakamura K, Kadowaki T, Kawaguchi H2004 PPAR� insufficiency enhances osteogenesis through osteoblast formationfrom bone marrow progenitors. J Clin Invest 113:846–855

71. Tornvig L, Mosekilde Li, Justesen J, Falk E, Kassem M 2001 Troglitazonetreatment increases bone marrow adipose tissue volume but does not affecttrabecular bone volume in mice. Calcif Tissue Int 69:46–50

72. Rzonka SO, Suva LJ, Gaddy DC, Montague B, Lecka-Czernik B 2004 Boneis a target for the anti-diabetic compound rosiglitazone. Endocrinology 145:401–406

73. Ali AA, Weinstein RS, Stewart SA, Parfitt AM, Manolagas SC, Jilka RL 2005Rosiglitazone causes bone loss in mice by suppressing osteoblast differenti-ation and bone formation. Endocrinology 146:1226–1235

74. Soroceanu MA, Mioa D, Bai X-Y, Su H, Goldzman D, Karplis AC 2004Rosiglitazone impacts negatively on bone by promoting osteoblast/osteocyteapoptosis. J Endocrinol 183:203–216

75. Lecka-Czernik B, Lazarenko OP, Rzonka SO, Teng K, Suva LJ 2004 The effectof rosiglitazone, an anti-diabetic PPAR-� agonist, on bone mass and architec-ture in growing, adult, and aged mice. Am Soc Bone Miner Res 19:M054(Abstract)

76. Watanabe S, Yasuhiro T, Seiji F, Fujita H, Nakano T, Fujita T 2003 Decreasein serum leptin by troglitazone is associated with preventing bone loss in type2 diabetic patients. J Bone Miner Metab 21:166–177

77. Schwartz AV, Sellmeyer DE, Feingold KR, Strotmeyer E, Resnick HE, Car-bone L, Beamer BA, Lane NE Thiazolidinedione (TZD) use and bone mineraldensity in older adults with diabetes. American Diabetes Association 62ndScientific Sessions, 2002, p A237 (Abstract 961-P)

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving theendocrine community.


Stimulation of Resorption in Cultured Mouse Calvarial Bones by Thiazolidinediones

Documents