www.elsevier.com/locate/boneBone 35 (2004) 842–849

Adiponectin and its receptors are expressed in bone-forming cells

Heidi S. Bernera,*, Staale P. Lyngstadaasa, Axel Spahrb, Marta Monjoa, Liv Thommesenc,Christian A. Drevond, Unni Syversenc, Janne E. Reselanda

aOral Research Laboratory, Institute for Clinical Dentistry, University of Oslo, NorwaybDepartment of Conservative Dentistry and Periodontology, University of Ulm, Germany

cDepartment of Cancer Research and Molecular Medicine, University of Science and Technology, Trondheim, NorwaydDepartment of Nutrition, Institute of Basic Medical Sciences, University of Oslo, Norway

Received 22 March 2004; revised 11 June 2004; accepted 14 June 2004

Available online 3 August 2004

Abstract

Adiponectin has until now been considered to be synthesized and secreted exclusively by the adipose tissue, and is reported to influence

energy homeostasis and insulin sensitivity. It is also known that body weight is positively correlated with increased bone mineral density and

decreased fracture risk. The mechanisms explaining this relation, however, are not completely understood. We report a link between

adiponectin and bone homeostasis by demonstrating transcription, translation, and secretion of adiponectin, as well as expression of its

receptors, AdipoR1 and AdipoR2, in bone-forming cells. We show that adiponectin and the receptors are expressed in primary human

osteoblasts from femur and tibia. The phenotype of bone cells was confirmed by the high expression levels of alkaline phosphatase, collagen

type 1, osteocalcin, and CD44, and the formation of mineralization nodules. Immunostaining with monoclonal antibodies also demonstrated

the presence of adiponectin in human osteosarcoma cells and normal osteoblasts. Both mRNA expression and secretion of adiponectin to the

medium increased during differentiation of human osteoblasts in culture. The adiponectin mRNA level increases in osteoblasts cultured 3 and

7 days in the presence of dietary fatty acids and supplementation of culture medium with recombinant adiponectin enhances the proliferation

of murine osteoblasts. The regulation and detailed function of adiponectin in bone still remains obscure, but our findings suggest a functional

role in bone homeostasis. If so, adiponectin may provide an important signal linking fat and body weight to bone density.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Adiponectin; Adipokines; Osteoblasts; Bone; Human

Introduction terminal collagenous repeat and a C-terminal globular

Adiponectin (also called Acrp30, apM1, or adipoQ) was

independently discovered by several groups [1–4] and has

been considered to be synthesized and secreted exclusively

by the adipose tissue. It is abundantly present in plasma [5]

and has been proposed to play important roles in the

regulation of energy homeostasis and insulin sensitivity

[6–8]. However, unlike other adipokines such as leptin,

plasma concentrations of adiponectin are reduced in asso-

ciation with obesity-linked diseases including coronary

artery diseases [9,10] and type 2 diabetes [11].

Adiponectin is a 244-amino acid protein structurally

similar to tumor necrosis factor alpha (TNFa) with an N-

8756-3282/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.bone.2004.06.008

* Corresponding author. Oral Research Laboratory, Institute for

Clinical Dentistry, University of Oslo, PO Box 1109 Blindern, Oslo N-

0317, Norway. Fax: +47-22-85-23-51.

E-mail address: hberner@odont.uio.no (H.S. Berner).

domain [1]. Adiponectin exists as a trimer, hexamer, a high

molecular weight form, and as small proteolytic cleavage

products in both human and mouse plasma [8,12]. The

biological activities of these different forms are poorly

understood, and it is possible that they are specific for

different receptors and cell types. Globular adiponectin

binds more avidly than full-length adiponectin in monocytes

and skeletal muscle membranes, but this pattern is the

opposite in hepatocytes and liver membranes [13]. In vitro

studies have demonstrated that fat cell expression of adipo-

nectin is repressed by factors that are increased in relation to

insulin [14]. In contrast, the only stimulus for adiponectin

gene expression in adipocytes known so far is activation of

PPARg by thiazolidinediones (TZD).

The cloning of two adiponectin receptors, AdipoR1 and

AdipoR2, was recently reported [13]. AdipoR1 is abundant-

ly expressed in muscle, whereas AdipoR2 is predominantly

expressed in the liver [13]. Expression of both receptors has

H.S. Berner et al. / Bone 35 (2004) 842–849 843

also been reported at high levels in human and rat pancreatic

beta cells, and their presence is suggested to be one

mechanism for modulating the effects of circulating adipo-

nectin [15]. Several studies indicate that AdipoR1 and

AdipoR2 serve as receptors for globular and full-length

adiponectin, and that they promote increased AMP kinase

and PPARa ligand activities, as well as fatty acid oxidation

and glucose uptake in the presence of adiponectin [13,16].

Adipocytes and osteoblasts are both of mesoderm origin.

It has been questioned how destined the mature phenotypes

of these cells are. Human osteoblasts have been demonstrat-

ed to express leptin, previously considered an adipocyte-

specific gene product [17]. Leptin seems to have at least two

different effects on bone metabolism, a direct stimulatory

effect on osteoblast differentiation, growth, and mineraliza-

tion [17,18], or an indirect suppressive effect on bone

development via the hypothalamus [19].

During osteoblast differentiation, there is a sequential

expression of genes associated with the biosynthesis, orga-

nization, and mineralization of the bone extracellular matrix

[20]. Cells of mesenchymal origin share many character-

istics in gene expression during differentiation, and even

though their phenotypes may differ markedly, mature adi-

pocytes and osteoblasts express and secrete several common

factors that underline the close relationship between these

cells [21]. The possibility remains that adiponectin is also an

adipokine expressed in bone cells. This is underlined by the

recent reported association between serum levels of adipo-

nectin and bone mineral density [22]. The present findings

of adiponectin expression and secretion from bone-forming

cells demonstrate that adiponectin is not adipocyte specific,

but may have important functions in normal bone cells.

Furthermore, the expression of adiponectin in bone cells

appears to be regulated, suggesting that external stimuli

known to regulate adipokines also influence expression in

bone-forming cells.

Materials and methods

Cell cultures

Commercially available primary human osteoblasts from

both femur and tibia of different donors (NHOst cell system;

Cambrex, Walkersville, MD, USA) were grown in osteo-

blast growth media (OGM, Cambrex). Osteoblasts cultured

to facilitate mineralization were exposed to hydrocortisone

hemisuccinate (200 nM) and h-glycerophosphate (10 mM)

(Cambrex) in ambient medium. The phenotype of cells was

characterized based on the expression levels of alkaline

phosphatase (ALP), collagen type 1, osteocalcin, and

CD44, and formation of mineralization nodules. The mouse

osteoblastic cell line MC3T3-E1 was obtained from Deut-

sche Sammlung von Microorganismen und Zellkulturen

(DSMZ No ACC 210; Braunschweig, Germany) and main-

tained in a-MEM (PAA, Linz, Austria) containing 20 mM

HEPES, 10% FCS (PAA), and 1% penicillin–streptomycin

solution. MC3T3-E1 cells cultured to facilitate differentia-

tion and enhanced mineralization were exposed to hydro-

cortisone hemisuccinate (200 nM) and h-glycerophosphate(10 mM) (Cambrex) in ambient medium. The osteosarcoma

cell lines SaOS-2 (ATTC HTB-85) and U-2 OS (ATTC

HTB-96) were obtained from American Type Culture Col-

lection (Rockville, MD). SaOS-2 and U-2 OS were grown in

McCoy’s 5A medium (PAA) supplemented with 10% FCS

and 1% penicillin–streptomycin solution. The monoclonal

lines of human osteosarcoma, OHS [23] and KPDXM [24],

were a kind gift from Dr. Bruland, The Norwegian Radium

Hospital, Norway. The cells were cultured in RPMI 1640

(PAA) with 10% FCS (PAA), 50 IU/ml penicillin, and 50 g/

ml streptomycin. Human SGBS cells (a kind gift from Dr.

Wabitsch, University of Ulm, Germany) were cultured and

differentiated into adipocytes essentially as described pre-

viously [25]. Briefly, cells were cultured in DMED (PAA)

supplemented with 4 mg of biotin, 2 g of D-pantothenate, 2

mM L-glutamine, penicillin (100 units/ml), and streptomy-

cin (100 Ag/ml) supplemented with 10% noninactivated

FCS (10270-106; Invitrogen life technologies, Carlsbad,

CA, USA). All cells were maintained in humidified 95%

air 5% CO2 atmosphere at 37jC. The medium was changed

twice a week and they were subcultured using 0.05% trypsin

with 0.01% EDTA. All media and serum used were pur-

chased from PAA and reagents were obtained from Sigma-

Aldrich (St. Louis, MO, USA) unless otherwise noted.

Rat tissue

The head including the jaws of Sprague–Dawley rats

(Charles River, Sulzfeld, Germany) at postnatal day 10 was

removed and fixed in freshly prepared 4% buffered para-

formaldehyde for about 20 h. The tissues were decalcified in

0.5 M EDTA, rinsed in phosphate-buffered saline, and

finally embedded in paraffin. Semiserial sections (6 Am)

were mounted on aminopropyl-triethoxysilane-coated glass

slides (Sigma). The sections were stained according to van

Gieson or used for immunohistochemistry. All animal

procedures were in accordance with the local ethical com-

mittee (Baden Wurttemberg, Germany).

Adipose tissue

Subcutaneous adipose tissue was obtained from nondia-

betic patients undergoing mammoplastic surgery at the

Volvat Medical Center. Samples were snap-frozen and

stored at �80jC.

Bone marrow cells and fluid

Femurs from 4-week-old (n = 6) and 14-week-old mice

(n = 5) (NMRI, Bomholt gaard Breeding and Research

Center, Denmark) were flushed with 200 Al PBS. From each

animal, bone marrow cells from one femur were separated

Table 1

Description of the human (h) and murine (m) oligonucleotide primers

Product Sequence Size

(bp)

h Adiponectin 5V-atggtcctgtgatgctttga-3V 343

5V-gttgagtgcgtatgttattttt-3Vh AdipoR1 5V-ttcttcctcatggctgtgatgt-3V 150

5V-agtggacaaaggctgctgcca-3Vh AdipoR2 5V-atagtctcccagtgggacatg-3V 214

5V-aggatccgggcagcataca-3Vm Adiponectin 5V-cattcttttcctgatactggtc-3V 531

5V-gacgttactacaactgaagagc-3Vm AdipoR1 5V-ttcttcctcatggctgtgatgt-3V 200

5V-agccaccttctaggccataac-3Vm AdipoR2 5V-tagcctctatatcaccggagct-3V 260

5V-agaaggcctgcaggacctgca-3V

Fig. 1. Expression of adiponectin, AdipoR1, and AdipoR2 mRNA in

human and murine osteoblasts. Human subcutaneous (sc) adipose tissue

was used as a positive control, whereas H2O was used as negative control.

H.S. Berner et al. / Bone 35 (2004) 842–849844

from the bone marrow fluid by centrifugation and lysed in

200 Al NP40 buffer, whereas cells and fluid from the other

femur were air-dried and stained with Oil-Red-O.

Oil-Red-O staining

From each mouse, the bone marrow fluid with cells from

one femur was air-dried on slides and fixed with 10%

formaldehyde in PBS for 15 min. Fixed cells were rinsed

and stained with Oil-red-O filtered solution (3 mg/ml in

isopropyl alcohol) for 1 h. Cells were then washed with

water and visualized with a phase-contrast microscope and

photographed.

mRNA isolation

Cells were lysed in lysis or binding buffer [100 mM

Tris–HCl, pH 8.0, 500 mM LiCl, 10 mM EDTA, pH 8.0,

0.5 mM dithiothreitol (DTT), and 1% sodium dodecyl

sulfate (SDS)]. mRNA was isolated using magnetic beads

[oligo (dT)25] as described by the manufacturer (Dynal AS,

Oslo, Norway). Beads containing mRNA were resuspended

in 10 mM Tris–HCl, pH 8.0, and stored at �70jC until use.

Ten microliters of the mRNA-containing solution was

applied directly to obtain a first-strand complementary

DNA (cDNA) using the iScript cDNA Synthesis Kit that

contains both oligo(dT) and random hexamer primers (Bio-

Rad, Hercules, CA, USA).

Real-time PCR quantification of gene expression

Reactions were performed and monitored using iCycler

iQ (Bio-Rad). The 2� iQ SYBR Green Supermix was based

on iTaq DNA polymerase (Bio-Rad). cDNA samples (1 Al fora total volume of 25 Al per reaction) were analyzed both for

the genes of interest and the reference genes (h-actin and

GAPDH). The cycling profile was as follows: denaturing at

94jC for 5 min followed by 40 cycles of annealing at 57jCfor 30 s, primer extension at 72jC for 30 s, and denaturing at

95jC for 30 s. Finally, one cycle for 3 min of extension

completed the reaction. Reactions were replicated twice with

duplicates on a 96-well plate. Cycle threshold (Ct) values

were obtained graphically. Gene expression was normalized

to h-actin and GAPDH, and presented as DCt values.

Comparison of gene expression between control and treated

samples was derived from subtraction of control DCt values

from treatment DCt values to give a DDCt value, and relative

gene expression was calculated as 2-DDCt and normalized to

controls. The efficiency of each set of primers was always

higher than 90%. The nucleotide sequences of the oligonu-

cleotides are shown in Table 1. Ten microliters of the

reaction mixture was separated on 2% agarose gels and

stained with ethidium bromide to verify bands.

Immunohistochemistry

Cells were grown on Lab-Tek chamber slides (Myriad

Industries, CA, USA). After cell culturing, the medium was

removed and chambers detached from the slides. Slides

were air dried and placed in ice-cold acetone for 10 min.

Sections were then stained manually using the InnoGenex

IHC Kit (InnoGenex, San Ramon, CA, USA) following the

instructions in the commercial manual. Briefly, sections

were incubated with adiponectin antibody (1:1000) or

casein antibody as irrelevant control (1:1000) in room

temperature for 1 h followed by incubation with biotinylated

secondary antibody at room temperature for 30 min. The

immunoreaction was finally visualized with diaminobenzi-

dine (0.7 mg/ml; Sigma). The human adipocyte cell line

SGBS [25] was used as positive control.

Tyramide signal amplification (TSA) Biotin system (Per-

kin-Elmer, Boston, MA, USA) was used according to

manufacturer instructions to staining rat sections. Briefly,

sections were deparaffinized in xylene and rehydrated in

decreasing concentrations of ethanol. Endogenous peroxi-

dase activity was quenched with 1% H2O2. Sections were

then treated with Proteinase K (0.1 mg/ml; Sigma) followed

H.S. Berner et al. / Bone 35 (2004) 842–849 845

by incubation for 30 min with the blocking reagent provided

in the kit. Sections were incubated overnight at 4jC with

adiponection antibody (1:500) diluted in the blocking

reagent. Control sections were incubated in the absence of

primary antibody. The immunoreaction was detected by

subsequent incubation with horseradish peroxidase (HRP)-

labeled goat anti-rabbit antibody (DakoCytomation Den-

mark A/S, Glostrup, Denmark), Biotinyl Tyramide, and

HRP-labeled streptavidine. The immunoreaction was finally

visualized with diaminobenzidine (0.7 mg/ml; Sigma).

All antibodies used were purchased from Chemicon

International (Temevula, CA, USA) unless otherwise noted.

Quantitation of adiponectin in culture media

Concentration of secreted adiponectin from human osteo-

blast was measured by a competitive radioimmunoassay

(human adiponectin RIA kit, Linco, St. Charles, MI, USA)

with recombinant 125I-adiponectin as tracer, following the

instruction manual. Murine adiponectin released into the

Fig. 2. Immunostaining of adiponectin in vitro in adipocytes and human osteoblas

for the negative controls (right panel).

medium was quantified by an Enzyme Amplified Sensitivity

Immunoassay kit (mouse adiponectin ELISA kit, Linco).

The assay was performed following the manufacturer’s

directions. Relative abundance of adiponectin was calculat-

ed as total protein ratios.

Proliferation assay

MC3T3-E1 cells (2 � 103 cells/well) were seeded out in

96-well plates and cultured for 24 h. Then the cells were

washed once with serum-free medium before addition of

new medium with or without recombinant mouse adiponec-

tin in the range of 0.01–2.0 Ag/ml (BioVendor Laboratory,

Heidelberg, Germany). After 2 h, BrdU (5-bromo-2V-deoxy-uridine)-labeling solution (Roche Molecular Biochemicals,

Mannheim, Germany) was added and the cells were cultured

for an additional 18 h. At the end of the pulse, the cells were

washed, fixed, and denatured, and an ELISA utilizing an

anti-BrdU peroxidase-conjugated antibody was used to

measure the amount of incorporated BrdU.

ts and in vivo in mandibular bone of rat (left panel). No staining was shown

Bone 35 (2004) 842–849

Statistical computations

Data obtained in real-time PCR (DCt values), ELISA,

and proliferation passed normality and equal variance test.

Statistical comparison between groups and treatments was

performed using Student’s t test. A probability of less than

or equal to 0.05 was considered significant.

H.S. Berner et al. /846

Fig. 4. Adiponectin levels in serum, bone marrow fluid, and lysed bone

marrow cells from 4-week-old (left panel, n = 6) and 14-week-old (right

panel, n = 5) mice. aIndicates significant difference ( P < 0.05) in

adiponectin concentration between BM cells and BM fluid, whereasbindicates significant difference ( P < 0.05) between young and old

animals.

Results

Expression of adiponectin and adiponectin receptors in

osteoblasts

Adiponectin and its receptors, AdipoR1 and AdipoR2,

were expressed in primary human osteoblasts from both

femur and tibia of different donors (Fig. 1). The expression

of adiponectin and AdipoR1 was confirmed in the murine

osteoblast cell line MC3T3-E1 (Fig. 1). Traditional endpoint

techniques where the results are evaluated after the PCR

reaction has reached a plateau indicate similar levels of

adiponectin expression in both bone and adipose tissue (Fig.

1). However, using real-time PCR we quantified the adipo-

nectin mRNA levels in mature osteoblasts from both the

tibia and femur to be about 3% of the level found in human

subcutaneous adipose tissue. Adiponectin was visualized in

vitro in normal osteoblasts and in vivo in mandibular bone

of 10-day-old rat by immunohistochemistry (Fig. 2). The

protein was also detected in culture media from human

osteosarcoma cells SaOs-2 (4 ng/ml), KPDXM (138 ng/ml),

and OHS (138 ng/ml). During a differentiation period of 5

weeks, the amount of protein secreted to the culture media

from primary human osteoblasts was in the range of 105–

130 ng/ml (Fig. 3). The expression of adiponectin mRNA

increased 3.7-fold during the same period of differentiation

Fig. 3. The level of adiponectin mRNA (5) and secreted protein (x) fromhuman osteoblasts differentiated over a 5-week period. The expression was

calculated relative to the level of 24 h. mRNA expression was normalized

to h-actin and GAPDH, whereas relative abundance of adiponectin protein

was calculated as total protein ratios. The data represent means F SD of

three similar experiments with duplicate samples.

when normalized to the expression of h-actin and GAPDH

(Fig. 3). A similar mRNA expression profile was observed

for differentiated osteoblasts from both femur as well as

tibia.

In the bone marrow, adiponectin may be secreted from

the surrounding bone cells, like osteoblasts or mesenchymal

stem cell-derived adipocytes. To evaluate whether adipo-

nectin was secreted from bone or bone marrow cells in vivo,

bone marrow from 4- and 14-week-old mice was harvested.

In 4-week-old mice, the amount of adiponectin in diluted

bone marrow fluid was significantly higher (150 F 28 ng/

ml) (P < 0.001) than the amount in lysed bone marrow cells

(20 F 5 ng/ml) (Fig. 4). A similar finding was also observed

for 14-week-old mice. Furthermore, the concentration of

adiponectin in bone marrow fluid in 4-week-old mice was

higher (150 F 28 ng/ml) (P = 0.002) than in the fluid of 14-

week-old mice (97 F 3 ng/ml) (Fig. 4). The level of

Fig. 5. Dietary fatty acid stimulation of adiponectin mRNA expression.

Human femur osteoblasts were incubated with 0.5 mM palmitic acid (in

ratio 2.5:1 with BSA) for 1, 3, and 7 days, and mRNA levels were

measured using real-time PCR. Relative expression levels were calculated

as percentage of cells incubated with BSA (control). The data represent

means F SD of three independent cell experiments with triplicate.

*Indicates significant difference at P < 0.05.

Fig. 6. Effect of adiponectin on cell growth. DNA synthesis analyzed by

BrdU incorporation during DNA replication after 24-h incubation with

recombinant adiponectin in the range of 0.01–2.0 Ag/ml in MC3T3-E1

cells. Values represent the average F SD of one experiment with four

parallels. *Indicates significant difference at P < 0.05.

H.S. Berner et al. / Bone 35 (2004) 842–849 847

adiponectin was lower in serum than in bone marrow fluid

in mice irrelevant of age.

Regulation of adiponectin expression by fatty acids

The expression of adiponectin in bone-forming cells was

found to be regulated by dietary fatty acids. Treatment of

human osteoblasts with 0.5 mM palmitic acid significantly

stimulated adiponectin mRNA production to 190% (P <

0.02) after 1 day and 280% (P < 0.03) after 3 days, relative

to cells treated with BSA as control (Fig. 5).

Adiponectin stimulates cell growth

We found that after 24-h incubation with low concen-

trations of adiponectin, DNA synthesis was significantly

(P < 0.001) enhanced 1.8-fold compared to cells incubated

without adiponectin, reflecting an increased growth rate of

MC3T3-E1 cells (Fig. 6).

Discussion

We report here the first demonstration of the transcrip-

tion, translation, and secretion of adiponectin in vitro from

primary human osteoblasts. Additionally, we show expres-

sion of adiponectin in vivo in rat mandibular bone. Adipo-

nectin expression in osteoblasts is much lower than in

subcutaneous adipose tissue, indicating that adipocytes

may be the major contributor of circulating adiponectin,

although our data clearly demonstrate that adiponectin is not

adipocyte specific. As in adipocytes [26], adiponectin syn-

thesis in osteoblasts corresponds to the stage of cell differ-

entiation in culture. The number of adipocytes in the bone

marrow increases with age, and adiponectin has previously

been found in bone marrow fat cells [27]. Surprisingly, the

level of adiponectin was higher in diluted bone marrow fluid

than in serum in both young and older animals. The

detection of adiponectin in red bone marrow of young

animals [28] and the fact that the level was higher in 4-

week-old compared to 14-week-old animals confirm that

bone cells contribute to the production of adiponectin found

in bone marrow.

The molecular regulation of plasma adiponectin levels is

largely unknown. Fatty acids enhanced adiponectin expres-

sion in bone cells, suggesting that also in these cells

adiponectin may be regulated by external stimuli known

to modulate adipokines. To our knowledge, this is also the

first report demonstrating a regulatory effect of fatty acids

on adiponectin mRNA expression. Fatty acids provide

increased energy supply to cells and enhance adiponectin

expression in osteoblasts. Our findings are in accordance

with recent observations among healthy individuals where

plasma-free fatty acid concentration correlated positively

with plasma adiponectin concentration [29]. It has previ-

ously been shown that adiponectin gene expression is

reduced by TNFa [26], IL-6 [30], h-adrenergic agonists

[31], and glucocorticoids [32], whereas TZD agonists of

proliferator-activated receptor (PPARg) increase adiponec-

tin expression and plasma levels [33]. Dietary fatty acids

may induce mRNA expression of PPARg and potentially

activate PPARa [34].

The observed transcription of both adiponectin receptors

in human osteoblasts indicates paracrine or endocrine

effects of adiponectin on bone-forming cells. In murine

osteoblasts, only AdipoR1 mRNA transcripts were detected,

which is in agreement with previous findings of AdipoR1 as

the most ubiquitously expressed receptor in both human and

murine bone tissue [13].

Administration of recombinant adiponectin enhances the

proliferation of MC-3T3-E1, indicating a functional role for

adiponectin in bone remodeling. Previously, there have been

some conflicting reports concerning the effect of adiponec-

tin on growth of endothelial cells [35,36]. The discrepancy

may be due to variations in cell systems used or it may

suggest cell-specific effects of adiponectin on growth.

The molecular control of human mesenchymal stem cell

differentiation into osteoblasts and adipocytes is not known,

but adipokines might play a role in the regulation of the

mature endpoint of these cells. We observe an increase in

osteoblast proliferation, whereas Yokota et al. [37] found

that adiponectin inhibited adipocyte differentiation in long-

term bone marrow cultures. It has previously been specu-

lated that adiponectin affects the bone marrow environment

by favoring myeloid cell production at the expense of B

lymphopoiesis [38], but adiponectin may also contribute in

the regulation of the mature stage of mesenchymal stem

cells. Adiponectin is structurally similar to TNFa, RANKL,

and osteoprotegerin, all of which are involved in osteoclas-

togenesis. Adiponectin may regulate the TNFa-induced

nuclear transcription factor-nB activation [39,40], the sig-

H.S. Berner et al. / Bone 35 (2004) 842–849848

naling pathway of RANKL. Adiponectin may have impor-

tant functions related to bone growth and maintenance;

however, the role of adiponectin in osteoclastogenesis

remains to be characterized.

Adiponectin may, by binding to its receptors, directly

influence the metabolism of cells in bone, as described

earlier in muscle and liver cells [7,8]. The regulation of bone

metabolism by adipokines is largely unknown, and the

observed expression and secretion of adiponectin by bone-

forming cells add more complexity, and redundancy, to this

intriguing issue.

Acknowledgments

We are grateful to Gita Pezeshki at the Department of

Conservative Dentistry and Periodontology, University of

Ulm, Germany, for excellent help with the immunohisto-

chemistry of rat sections. This work was supported by EU

Grant, QLK3-CT-2001-00090, the Freia Chocolade Fabriks

Medicinske Fond, and Throne Holst Foundation for

Nutrition Research.

References

[1] Hu E, Liang P, Spiegelman BM. AdipoQ is a novel adipose-specific

gene dysregulated in obesity. J Biol Chem 1996;271:10697–703.

[2] Nakano Y, Tobe T, Choi-Miura NH, Mazda T, Tomita M. Iso-

lation and characterization of GBP28, a novel gelatin-binding

protein purified from human plasma. J Biochem (Tokyo) 1996;

120:803–12.

[3] Maeda K, Okubo K, Shimomura I, Funahashi T, Matsuzawa Y,

Matsubara K. cDNA cloning and expression of a novel adipose

specific collagen-like factor, apM1 (AdiPose Most abundant Gene

transcript 1). Biochem Biophys Res Commun 1996;221:286–9.

[4] Scherer PE, Williams S, Fogliano M, Baldini G, Lodish HF. A

novel serum protein similar to C1q, produced exclusively in adipo-

cytes. J Biol Chem 1995;270:26746–9.

[5] Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, et al.

Paradoxical decrease of an adipose-specific protein, adiponectin, in

obesity. Biochem Biophys Res Commun 1999;257:79–83.

[6] Combs TP, Berg AH, Obici S, Scherer PE, Rossetti L. Endogenous

glucose production is inhibited by the adipose-derived protein Acrp30.

J Clin Invest 2001;108:1875–81.

[7] Berg AH, Combs TP, Du X, Brownlee M, Scherer PE. The adipocyte-

secreted protein Acrp30 enhances hepatic insulin action. Nat Med

2001;7:947–53.

[8] Fruebis J, Tsao TS, Javorschi S, Ebbets-Reed D, Erickson MR, Yen

FT, et al. Proteolytic cleavage product of 30-kDa adipocyte comple-

ment-related protein increases fatty acid oxidation in muscle and

causes weight loss in mice. Proc Natl Acad Sci U S A 2001;98:

2005–10.

[9] Ouchi N, Kihara S, Arita Y, Maeda K, Kuriyama H, Okamoto Y,

et al. Novel modulator for endothelial adhesion molecules: adipo-

cyte-derived plasma protein adiponectin. Circulation 1999;100:

2473–6.

[10] Kumada M, Kihara S, Sumitsuji S, Kawamoto T, Matsumoto S, Ouchi

N, et al. Association of hypoadiponectinemia with coronary artery

disease in men. Arterioscler Thromb Vasc Biol 2003;23:85–9.

[11] Hotta K, Funahashi T, Arita Y, Takahashi M, Matsuda M, Okamoto Y,

et al. Plasma concentrations of a novel, adipose-specific protein, adi-

ponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol

2000;20:1595–9.

[12] Kishida K, Nagaretani H, Kondo H, Kobayashi H, Tanaka S, Maeda

N, et al. Disturbed secretion of mutant adiponectin associated with the

metabolic syndrome. Biochem Biophys Res Commun 2003;306:

286–92.

[13] Yamauchi T, Kamon J, Ito Y, Tsuchida A, Yokomizo T, Kita S, et al.

Cloning of adiponectin receptors that mediate antidiabetic metabolic

effects. Nature 2003;423:762–9.

[14] Fasshauer M, Klein J, Neumann S, Eszlinger M, Paschke R. Hormon-

al regulation of adiponectin gene expression in 3T3-L1 adipocytes.

Biochem Biophys Res Commun 2002;290:1084–9.

[15] Kharroubi I, Rasschaert J, Eizirik DL, Cnop M. Expression of adipo-

nectin receptors in pancreatic beta cells. Biochem Biophys Res Com-

mun 2003;312:1118–22.

[16] Tomas E, Tsao TS, Saha AK, Murrey HE, Zhang Cc, Itani SI, et al.

Enhanced muscle fat oxidation and glucose transport by ACRP30

globular domain: acetyl-CoA carboxylase inhibition and AMP-acti-

vated protein kinase activation. Proc Natl Acad Sci U S A 2002;99:

16309–13.

[17] Reseland JE, Syversen U, Bakke I, Qvigstad G, Eide LG, Hjertner O,

et al. Leptin is expressed in and secreted from primary cultures of

human osteoblasts and promotes bone mineralization. J Bone Miner

Res 2001;16:1426–33.

[18] Reseland JE, Gordeladze JO. Role of leptin in bone growth: central

player or peripheral supporter? FEBS Lett 2002;528:40–2.

[19] Ducy P, Amling M, Takeda S, Priemel M, Schilling AF, Beil FT, et al.

Leptin inhibits bone formation through a hypothalamic relay: a central

control of bone mass. Cell 2000;100:197–207.

[20] Stein GS, Lian JB, Stein JL, Van Wijnen AJ, Montecino M. Tran-

scriptional control of osteoblast growth and differentiation. Physiol

Rev 1996;76:593–629.

[21] Dennis JE, Charbord P. Origin and differentiation of human and mu-

rine stroma. Stem Cells 2002;20:205–14.

[22] Lenchik L, Register TC, Hsu FC, Lohman K, Nicklas BJ, Freedman

BI, et al. Adiponectin as a novel determinant of bone mineral density

and visceral fat. Bone 2003;33:646–51.

[23] Fodstad O, Brogger A, Bruland O, Solheim OP, Nesland JM, Pihl A.

Characteristics of a cell line established from a patient with multiple

osteosarcoma, appearing 13 years after treatment for bilateral retino-

blastoma. Int J Cancer 1986;38:33–40.

[24] Larsen RH, Bruland OS, Hoff P, Alstad J, Lindmo T, Rofstad EK.

Inactivation of human osteosarcoma cells in vitro by 211At-TP-3

monoclonal antibody: comparison with astatine-211-labeled bovine

serum albumin, free astatine-211 and external-beam X rays. Radiat

Res 1994;139:178–84.

[25] Wabitsch M, Brenner RE, Melzner I, Braun M, Moller P, Heinze E,

et al. Characterization of a human preadipocyte cell strain with high

capacity for adipose differentiation. Int J Obes Relat Metab Disord

2001;25:8–15.

[26] Kappes A, Loffler G. Influences of ionomycin, dibutyryl-cycloAMP

and tumour necrosis factor-alpha on intracellular amount and secre-

tion of apM1 in differentiating primary human preadipocytes. Horm

Metab Res 2000;32:548–54.

[27] Yokota T, Meka CS, Kouro T, Medina KL, Igarashi H, Takahashi M,

et al. Adiponectin, a fat cell product, influences the earliest lympho-

cyte precursors in bone marrow cultures by activation of the cyclo-

oxygenase-prostaglandin pathway in stromal cells. J Immunol 2003;

171:5091–9.

[28] Compston JE. Bone marrow and bone: a functional unit. J Endocrinol

2002;173:387–94.

[29] Bernstein EL, Koutkia P, Ljungquist K, Breu J, Canavan B, Grins-

poon S. Acute regulation of adiponectin by free fatty acids. Metabo-

lism 2004;53:790–3.

[30] Fasshauer M, Kralisch S, Klier M, Lossner U, Bluher M, Klein J, et al.

Adiponectin gene expression and secretion is inhibited by interleukin-

H.S. Berner et al. / Bone 35 (2004) 842–849 849

6 in 3T3-L1 adipocytes. Biochem Biophys Res Commun 2003;301:

1045–50.

[31] Fasshauer M, Klein J, Neumann S, Eszlinger M, Paschke R. Adipo-

nectin gene expression is inhibited by beta-adrenergic stimulation via

protein kinase A in 3T3-L1 adipocytes. FEBS Lett 2001;507:142–6.

[32] Halleux CM, Takahashi M, Delporte ML, Detry R, Funahashi T,

Matsuzawa Y, et al. Secretion of adiponectin and regulation of

apM1 gene expression in human visceral adipose tissue. Biochem

Biophys Res Commun 2001;288:1102–7.

[33] Combs TP, Wagner JA, Berger J, Doebber T, Wang WJ, Zhang BB,

et al. Induction of adipocyte complement-related protein of 30 kilo-

daltons by PPARgamma agonists: a potential mechanism of insulin

sensitization. Endocrinology 2002;143:998–1007.

[34] Hihi AK, Michalik L, Wahli W. PPARs: transcriptional effectors of

fatty acids and their derivatives. Cell Mol Life Sci 2002;59:790–8.

[35] Ouchi N, Kobayashi H, Kihara S, Kumada M, Sato K, Inoue T, et al.

Adiponectin stimulates angiogenesis by promoting cross-talk be-

tween AMP-activated protein kinase and Akt signaling in endothelial

cells. J Biol Chem 2004;279:1304–9.

[36] Brakenhielm E, Veitonmaki N, Cao R, Kihara S, Matsuzawa Y, Zhi-

votovsky B, et al. Adiponectin-induced antiangiogenesis and antitu-

mor activity involve caspase-mediated endothelial cell apoptosis. Proc

Natl Acad Sci U S A 2004;101:2476–81.

[37] Yokota T, Meka CS, Medina KL, Igarashi H, Comp PC, Takahashi M,

et al. Paracrine regulation of fat cell formation in bone marrow cul-

tures via adiponectin and prostaglandins. J Clin Invest 2002;109:

1303–10.

[38] Yokota T, Oritani K, Takahashi I, Ishikawa J, Matsuyama A,

Ouchi N, et al. Adiponectin, a new member of the family of

soluble defense collagens, negatively regulates the growth of mye-

lomonocytic progenitors and the functions of macrophages. Blood

2000;96:1723–32.

[39] Ouchi N, Kihara S, Arita Y, Okamoto Y, Maeda K, Kuriyama H, et al.

Adiponectin, an adipocyte-derived plasma protein, inhibits endothe-

lial NF-kappaB signaling through a cAMP-dependent pathway. Cir-

culation 2000;102:1296–301.

[40] Tsao TS, Murrey HE, Hug C, Lee DH, Lodish HF. Oligomerization

state-dependent activation of NF-kappa B signaling pathway by adi-

pocyte complement-related protein of 30 kDa (Acrp30). J Biol Chem

2002;277:29359–62.