Expression Profiles of Receptor Activator of Nuclear FactorkB Ligand, Receptor Activator of Nuclear FactorkB, and

Osteoprotegerin Messenger RNA in Aged andOvariectomized Rat Bones

TOHRU IKEDA,1 MASANORI UTSUYAMA,1,2 and KATSUIKU HIROKAWA1

ABSTRACT

The receptor activator of nuclear factor-kB ligand (RANKL; also known as tumor necrosis factor–relatedactivation-induced cytokine [TRANCE], osteoprotegerin ligand [OPGL], and osteoclast differentiation factor[ODF]) is a transmembrane ligand expressed in osteoblasts and bone marrow stromal cells. It binds to RANK,which is expressed in osteoclast progenitor cells, and induces osteoclastogenesis. OPG, a decoy receptor forRANKL, also binds to RANKL, and competitive binding of RANKL with RANK or OPG is thought toregulate bone metabolism. To investigate roles of the RANKL/RANK/OPG system in pathophysiologicalconditions, the expression of RANKL, RANK, and OPG messenger RNA (mRNA) was analyzed in bones ofaged and ovariectomized rats by means of in situ hybridization. In the control 8-week-old male andsham-operated female rat bones, the expression of RANKL mRNA was detected in hypertrophic chondrocytesof the growth plate and some periosteal and endosteal mesenchymal cells. The expression of RANK mRNAwas detected in osteoclast-like cells and mononuclear cells in contact with the cortical and trabecular bones.The expression of OPG mRNA was detected in proliferating chondrocytes and osteocytes. In the 2.5-year-oldrat bones, the expression of RANKL, RANK, and OPG mRNA tended to decrease except for the endostealregion. In the ovariectomized rat bones, the expression of RANKL, RANK, and OPG mRNA increased, andhigh expression of OPG mRNA was induced in resting chondrocytes and osteocytes. These results suggest thatestrogen deficiency stimulates the RANKL/RANK/OPG system and induces OPG in cells that have beenthought to be less important for bone metabolism. (J Bone Miner Res 2001;16:1416–1425)

Key words: receptor activator of nuclear factor kB ligand, receptor activator of nuclear factor kB, osteo-protegerin, aging, ovariectomy

INTRODUCTION

DISCOVERY OFthe receptor activator of nuclear factorkBligand (RANKL, also called osteoprotegerin ligand

[OPGL]/osteoclast differentiation factor [ODF]/tumor ne-crosis factor [TNF]–related activation-induced cytokine

[TRANCE]) clarified the molecular mechanisms behind theinitial steps of osteoclastogenesis.(1–7) RANKL is a type IITNF-like transmembrane protein, which binds to the recep-tor activator of nuclear factorkB (NF-kB; RANK) or OPG.RANK is expressed in osteoclast progenitor cells and in-duces osteoclastogenesis by binding to RANKL.(5,7,8) OPG

1Department of Pathology and Immunology, Aging and Developmental Science, Graduate School, Tokyo Medical and DentalUniversity, Bunkyo-ku, Tokyo, Japan.

2Department of Membrane Biochemistry, Tokyo Metropolitan Institute of Gerontology, Itabashi-ku, Tokyo, Japan.

JOURNAL OF BONE AND MINERAL RESEARCHVolume 16, Number 8, 2001© 2001 American Society for Bone and Mineral Research

1416

is a secreted TNF receptor–related protein, which was iso-lated as a protein-inhibiting bone resorption.(9,10) BecauseOPG does not have a transmembrane domain, it works as adecoy receptor and inhibits bone resorption.(7,8) Osteoclas-togenesis can be reproduced efficiently by coculture in vitrousing bone marrow stromal cells and bone marrow macro-phages or spleen cells.(11–14) The essential role of bonemarrow stromal cells in in vitro osteoclastogenesis is ex-plained by the fact that RANKL is expressed in the bonemarrow stromal cells on stimulation with 1a,25-dihydroxyvitamin D3 and dexamethasone.(3,7) Actually,bone marrow macrophages or spleen cells differentiatedinto osteoclasts without stromal cells when the soluble formof recombinant RANKL protein was added to the culturemedium.(2,3) The essential role of the RANKL/RANK sys-tem in osteoclastogenesis was confirmed in mice with adisruptedopgl gene. The mice showed severe osteopetrosisand a defect of tooth eruption caused by complete lack ofosteoclasts.(15)

Bone metabolism must be regulated strictly to maintainbone structures, to heal fractures, and to supply calcium ion.We showed dynamic change in the biological activity ofbone cells under physiological and pathological condi-tions.(16,17) Physiologically, bone formation and bone re-sorption are synchronized nicely, with an imbalance leadingto osteopetrosis or osteoporosis. A decrease in bone volumeoccurs in aged men.(18) In the aged male rat bones, mostosteoblasts looked flat-shaped and expressed much less typeI collagen and osteocalcin messenger RNA (mRNA) thanthe cuboidal osteoblasts in the young rat bones.(16) We alsofound that osteopontin mRNA was dominantly expressed inosteoclast progenitor-like mononuclear cells and osteoclastsrather than osteoblastic cells,(16,17,19)and the expression inthese bone resorbing cells greatly decreased in the aged ratbones.(16) These findings indicate that the bone metabolismin aged rat bones is less active than that in young rat bonesand support that osteoporosis in aged men occurs because ofan imbalance between bone resorption and bone formationat low turnover of bone. Postmenopausal osteoporosis isthought to be caused by estrogen deficiency and is repro-duced in an animal model by ovariectomy.(20) Overproduc-tion and overactivation of osteoclasts have been observed inbones of ovariectomized rats.(17,21–25) Previously, weshowed that both the expression of type I collagen andosteocalcin in osteoblasts and the expression of osteopontinin bone-resorbing cells increased in the ovariectomized ratbones.(17) These data confirmed that postmenopausal osteo-porosis was caused by an imbalance of bone resorption andbone formation under conditions of increased bone turn-over. The importance of the RANKL/RANK system forosteoclastogenesis strongly suggests that the system is as-sociated with the pathogenesis of age-dependent reductionin bone volume and postmenopausal osteoporosis.

In this study, we analyzed the expression of mRNA forRANKL, RANK, and OPG in bones of 8-week-old and2.5-year-old male rats and sham-operated and ovariecto-mized female rats. Relatively constant expression levels ofRANKL and RANK mRNA in the endosteal cells in theaged rat bones indicated a different regulatory mechanismfor the expression of RANKL and RANK mRNA from that

of bone matrix proteins, which was analyzed previously.(16)

On the other hand, the expression of RANKL, RANK, andOPG mRNA increased in the ovariectomized rat bones likethe expression of bone matrix proteins, which was reportedpreviously.(17) These results strongly suggested that theRANKL/RANK/OPG system was closely involved in thepathogenesis of bone loss by aging and estrogen deficiency.

MATERIALS AND METHODS

Animals and tissue preparation

Eight-week-old male and 10-week-old female Wistar ratswere purchased from SRL (Shizuoka, Japan). Male Wistarrats (2.5 years old) were supplied from SPF Aging Farm ofTokyo Metropolitan Institute of Gerontology (Tokyo, Ja-pan). The male Wistar rats from the farm, which werepurchased from SRL, were kept from the time they were 4weeks old, and mean survival of the 2.5-year-old maleWistar rats was 8.8% (Department of Laboratory AnimalScience, Tokyo Metropolitan Institute of Gerontology, un-published data, 1998). From the farm, four healthy 2.5-year-old male rats were supplied for experiments. The 8-week-old and the 2.5-year-old male rats were anesthetized withether and killed by cardiac perfusion. The 10-week-oldfemale Wistar rats were anesthetized and bilaterally ovari-ectomized or sham-operated. These animals were killed bycardiac perfusion 3 weeks after the operation. The distalportion of the femur was dissected from each animal andfixed for 24 h at 4°C in 4% formaldehyde in 0.1 M phos-phate buffer (pH 7.2), which was prepared just before use.All fixed samples were decalcified with 0.5 M EDTA solu-tion at 4°C for up to 1 month. The decalcified samples weredehydrated in ethanol, cleared in chloroform, and embeddedin paraffin, and then 4-mm thick sections were made. Fouranimals for each group were used for the experiments.

Preparation of RNA probes

Complementary DNA (cDNA) fragments of mouseRANKL and RANK were cloned from mouse thymiccDNA, and cDNA fragments of mouse OPG were clonedfrom mouse bone marrow macrophage cDNA by reverse-transcription–polymerase chain reaction (RT-PCR) method.The RT-PCR was carried out with 40 cycles of denaturationat 98°C for 20 s and annealing and extension at 68°C for 2minutes using LATaq polymerase (Takara Co., Shiga,Japan) and the following primers:

● RANKL. RANKL5, 59-ATGAAACAAGCCTTTCAG-GGGGCCGTGCA-39; and RANKL3, 59-TCAGTCT-ATGTCCTGAACTTTGAAAGCCCC-39

● RANK. RANK5, 59-TCACCGGGACTGAAAGCAC-GGTGG-39; and RANK3, 59-TGGGCTCCATCA-CCATGCCAGCAG-39

● OPG. OPG5, 59-GGACAGTTTGCCTGGGACCAA-AGTGAATGC-39; and OPG3, 59-TGAAGCTGTGC-AGGAACCTCATGGTCTTCC-39

The cDNA fragments were cloned into pCR2.1-TOPO vec-tor (Invitrogen, Carlsbad, CA, USA). The vectors including

1417EXPRESSION PROFILES OF RANKL, RANK, AND OPG mRNA

the cDNA fragment for RANKL, RANK, or OPG werelinearized and transcribed using T7 RNA polymerase(Roche Diagnostics GmbH, Mannheim, Germany) to gen-erate uridine 59-triphosphate (UTP)-labeled single-strandedantisense and sense RNA probes, respectively. The [33P]-labeled probes were used for hybridization at a concentra-tion of 1 3 107 cpm/ml in 50% formamide, 10% dextransulfate, 0.6 M NaCl, 10 mM Tris-HCl (pH 7.6), 1 mMdithiothreitol (DTT), and 0.2mg/ml of yeast total RNA.

In situ hybridization

Treatment of the slides and hybridization conditions wereas described previously.(19) Briefly, after removal of theparaffin with xylene, the sections were treated with 2mg/mlof proteinase K (Sigma Chemical Co., St. Louis, MO, USA)at 37°C for 15 minutes, fixed for 15 minutes with the samefixatives as described previously, treated with 0.25% aceticanhydride in 0.1 M triethanol amine-HCl (pH 8.0) for 10minutes, dehydrated in ethanol, and air-dried. The hybrid-ization solution was spread over the sections and the slideswere incubated overnight at 60°C in plastic humidifiedboxes. After hybridization, the sections were washed twicein 50% formamide solution containing 23 SSC and treatedwith 12.5 mg/ml of RNAse A (Sigma Chemical Co.). Thesections were then washed twice in 23 SSC and twice in0.23 SSC at 60°C for 30 minutes each time, dehydrated inethanol, and finally dipped in NTB-3 emulsion (EastmanKodak, Rochester, NY, USA) diluted 1:1 with a 2% glyc-erol solution. The dipped slides were exposed at 4°C for 3weeks. The exposed slides were developed in D-19 devel-oper, fixed in F-5 fixative, and finally counterstained withhematoxylin and eosin.

RESULTS

The expression profiles of RANKL, RANK, and OPGmRNA were summarized in Table 1. The grading of theexpression level represents total amount of the expression ineach region. Profiling of the sites and cells that appear in thetable were shown in Fig. 1.

The expression of RANKL mRNA

The expression levels of RANKL mRNA were comparedbetween the femurs of 8-week-old and 2.5-year-old malerats. In the 8-week-old rat bones, RANKL mRNA wasdetected most prominently in a subset of periosteal cells ofthe metaphyseal cortical bone (Fig. 2A, arrows). Lowerlevels of expression were detected in a subset of endostealcells (Fig. 2A, arrowheads). In addition, RANKL mRNAwas detected in hypertrophic chondrocytes of the growthplate (Fig. 2C, arrows). In the primary spongiosa, RANKLmRNA was not detected clearly (Fig. 2C). In the 2.5-year-old rat bones, RANKL mRNA was detected in a subset ofperiosteal (Fig. 2B, an arrow) and endosteal cells (Fig. 2B,arrowheads) of the metaphyseal cortical bones. The expres-sion profile in the 2.5-year-old rat bones was similar to thatin the 8-week-old rat bones except for the periosteal region.In the periosteal region, cells with a positive signal were

fewer than in the 8-week-old rat bones although the expres-sion level in each cell was similar to the 8-week-old ratbones (Figs. 2A and 2B, arrows). In the 2.5-year-old ratbones, the growth plate was composed of several layers ofresting and/or proliferative chondrocytes, and hypertrophicchondrocytes were not present. In this region, little RANKLmRNA was expressed (Fig. 2D).

Next, the expression of RANKL mRNA was analyzed inthe femurs of sham-operated and ovariectomized femalerats. Ten-week-old female rats were sham-operated orovariectomized and killed 3 weeks after each operation. Theexpression of RANKL mRNA was detected in a subset ofperiosteal (Figs. 2E and 2F, arrows) and endosteal (Fig. 2F,arrowheads) cells of the sham-operated and ovariectomizedrat bones. The expression level was similar in this regionbetween the sham-operated and ovariectomized rat bones(Figs. 2E and 2F). In the growth plate of the sham-operatedrat bones, RANKL mRNA was detected in the hypertrophicchondrocytes (Fig. 2G, arrows). In the ovariectomized ratbones, both the number of cells expressing RANKL mRNAand the expression level of RANKL mRNA in each cellincreased in comparison to the sham-operated rat bones(Fig. 2H, arrows). In addition to being up-regulated in thehypertrophic chondrocytes, the expression of RANKL

TABLE 1. SUMMARY OF PROFILES OF THEEXPRESSION OF

RANKL, RANK, AND OPGMRNA

8 w 2.5 y sham ovx

RANKLCortical periosteuma 11 1 1 1Cortical endosteuma 1 1 1 1Trabecular endosteumb 2 2 2 1Growth plate hypertrophic

chondrocytes 1 2 1 11Osteocytes 2 2 2 2

RANKCortical periosteuma 111 1 11 11Cortical endosteuma 11 1 1 1Trabecular endosteumc 1 1 1 11Growth plate 2 2 2 2Osteocytes 2 2 2 2

OPGCortical periosteuma 1 2 2 2Trabecular endosteumc 2 2 2 1Growth plate proliferative

chondrocytes 1 1 2 2Growth plate resting

chondrocytes 2 2 2 1Articular proliferative

chondrocytes 1 2 1 1Articular resting

chondrocytes 2 2 2 111Osteocytes 1 2 1 11

8w, 8-week-old rat; 2.5 y, 2.5-year-old rat; sham, sham-operatedrat; OVX, ovariectomized rat.

a Metaphyseal region.b Primary spongiosa.c Primary and secondary spongiosas.

1418 IKEDA ET AL.

mRNA was induced in a subset of cells in the primaryspongiosa of the ovariectomized rat bones (Fig. 2H, arrow-heads). In the primary spongiosa of the sham-operated ratbones, RANKL mRNA was not clearly detected (Fig. 2G).

The expression of RANK mRNA

In the metaphyseal cortical bone of the 8-week-old rats,high levels of expression of RANK mRNA were detected inrelatively small cells (Fig. 3A, arrows) and in large cells(Fig. 3A, arrowheads) in the periosteum and endosteum. Inthe periosteum of the 2.5-year-old rat bones, both the num-ber of cells expressing RANK mRNA and the expressionlevel in each cell were less than those in the 8-week-old ratbones, but in the endosteum of the 2.5-year-old rat bones,the number of cells expressing RANK mRNA changed littleand only the level of the expression of RANK mRNAdecreased in comparison to the 8-week-old rat bones (Figs.3B and 3C). In the 2.5-year-old rat bones, RANK mRNAwas detected in mononuclear cells (Fig. 3C, arrows) andmultinuclear cells (Figs. 3B and 3C, arrowheads). The ex-pression profile in the trabecular bones of each animal was

similar to that in the metaphyseal cortical bones (data notshown).

In the sham-operated and ovariectomized rat bones,RANK mRNA was detected in small (Figs. 3D and 3E,arrows) and large (Figs. 3D and 3E, arrowheads) cells in theperiosteum and endosteum of the metaphyseal trabecularbones. The level of expression in these regions was similarbetween sham-operated and ovariectomized rat bones.However, in the trabecular bones, the expression in eachcell was stronger in the ovariectomized rats than in thesham-operated rats (Figs. 3F and 3G).

The expression of OPG mRNA

OPG mRNA was detected in the cartilage rather than inthe bone. In the 8-week-old rat bones, relatively strongexpression of OPG mRNA was detected in the proliferativechondrocytes of the articular cartilage (Fig. 4A, arrows). Inaddition, the mRNA also was detected in some proliferativechondrocytes of the growth plate (Fig. 4C, arrows). In the2.5-year-old rat bones, OPG mRNA also was detected in theproliferative chondrocytes of the atrophic growth plate (Fig.4D, arrows), but the number of cells expressing OPGmRNA and the expression level in each cell were similar tothose in the 8-week-old rat bones. In the articular cartilage,the expression was undetectable (Fig. 4B). Outside of thecartilage region in the 8-week-old rats, OPG mRNA wasdetected in a subset of cells in the periosteum of the me-taphyseal cortical bones (Fig. 4E, arrows). In this region, themRNA also was detected in a subset of osteocytes (Fig. 4E,an arrowhead). In the 2.5-year-old rat bones, the expressionwas not detected clearly in the metaphyseal cortical boneregion (Fig. 4F). In the trabecular bones of the 8-week-oldand 2.5-year-old rats, little OPG mRNA was expressed(data not shown).

In the articular region of the sham-operated rat bones,OPG mRNA was detected in the proliferative chondrocyticcells (Fig. 5A, arrowheads). In addition, the mRNA wasdetected in some osteocytes in the metaphyseal and diaphy-seal cortical bones (Fig. 5C, arrowheads). In the region ofthe growth plate, little OPG mRNA was seen in the sham-operated rats (Fig. 5E) in contrast to 8-week-old and 2.5-year-old male rats (Figs. 4C and 4D). In the sham-operatedrats, the expression also was weak in the trabecular boneregion (Fig. 5G). Three-weeks after the ovariectomy, theexpression of OPG mRNA greatly increased in some re-gions of the bone. Prominent expression was detected inresting chondrocytes of the articular cartilage (Fig. 5B,arrows). In this region, the mRNA also was detected insome proliferative chondrocytes (Fig. 5B, arrowheads). Inthe cortical and trabecular bones, both the number of osteo-cytes expressing OPG mRNA and the expression level ofOPG mRNA in each cell increased (Fig. 5D, arrowheads).In the growth plate region, weak expression was detected insome resting chondrocytes (Fig. 5F, arrows). In addition,many cells expressing OPG mRNA were seen in the pri-mary (Fig. 5H) and secondary spongiosa (data not shown)of the ovariectomized rat bones.

FIG. 1. Profiling of the sites and cells in the femur analyzed in thisstudy. 1, cortical periosteum; 2, cortical endosteum; 3, trabecularendosteum; 4, growth plate resting chondrocyte; 5, growth plate pro-liferative chondrocyte; 6, growth plate hypertrophic chondrocyte; 7,articular resting chondrocyte; 8, articular proliferative chondrocyte; 9,osteocyte.

1419EXPRESSION PROFILES OF RANKL, RANK, AND OPG mRNA

For all these experiments, sense control hybridization wasperformed and no specific signal was detected (data notshown).

DISCUSSION

In the 8-week-old rat bones, strong expression ofRANKL and RANK mRNA was detected in the periosteumand endosteum of the metaphyseal cortical bones. In the2.5-year-old rat bones, the expression of RANKL andRANK mRNA in the periosteum greatly decreased, but theexpression in the endosteum was relatively stable. Meta-

physeal cortical bone is the region where bone reconstruc-tion actively occurs during bone development.(26) There-fore, the high level of expression of RANKL and RANKmRNA in the epiphyseal cortical bone of the 8-week-oldrats was thought to be induced by active reconstruction inthe developing bone. In addition, the decreased expressionof RANKL mRNA in the growth plate of the 2.5-year-oldrat bones might be caused by a less active bone metabolismthan that in the 8-week-old rat bones. Except for the regionswhere bone metabolism was influenced greatly by thegrowth of bone, the difference in the expression level ofRANKL and RANK mRNA between young and old rat

FIG. 2. The expression ofRANKL mRNA in the femurs of(A and C) 8-week-old male, (Band D) 2.5-year-old male, (E andG) sham-operated female, (F andH) and ovariectomized rats. (Aand B) The expression in the peri-osteal (arrows) and endosteal (ar-rowheads) cells of the metaphy-seal cortical bones. (C) Theexpression in the hypertrophicchondrocytes (arrows) of the de-veloping growth plate. (D) Noexpression of RANKL mRNA inthe atrophic growth plate regionof the 2.5-year-old rat bone. (Eand F) The expression in the me-taphyseal cortical bone region ofthe sham-operated and ovariecto-mized rats. Arrows indicate theexpression in the periosteum andarrowheads indicate the expres-sion in the endosteum. (G and H)The expression in hypertrophicchondrocytes of the growth plate(arrows). In the ovariectomizedrat bone, strong expression inthe hypertrophic chondrocytes(panel H, arrows) and inductionof RANKL mRNA in the cells onthe trabecular bones (panel H, ar-rowheads) are seen (3300).

1420 IKEDA ET AL.

bones was not great. Considering the marked decrease in theexpression of mRNA for bone matrix proteins in aged ratbones,(16) the decrease in bone resorption may be milderthan that of bone formation in the aged rat bones. This mayexplain the pathogenesis of osteoporosis in aged men.

We previously reported that the expression of bone ma-trix proteins increased in ovariectomized rat bones in addi-tion to the number and activity of osteoclasts.(17) The dataindicated that osteopenia in ovariectomized rat bones wasinduced by an imbalance of bone resorption and bone for-mation under conditions of high turnover of bone. TheRANKL/RANK system, a key regulator of osteoclastogen-esis, was expected to be stimulated in the ovariectomized rat

bones. Actually, in the ovariectomized rat bones, an in-crease in the expression of RANKL mRNA was seen in thehypertrophic chondrocytes of the growth plate, and induc-tion of the expression was seen in the cells of the primaryspongiosa (Figs. 2G and 2H). An increase in the expressionof RANK mRNA was seen in the cells of the trabecularbones (Figs. 3F sand 3G). However, in the cortical boneregion, the expression profiles of RANKL and RANKmRNA changed little between sham-operated and ovariec-tomized rat bones (Figs. 2E, 2F, 3D, and 3E). In the ovari-ectomized rats, bone volume decreased more prominently inthe trabecular bones than in the cortical bones, at least up to6 weeks after the operation.(17) Therefore, an increase in the

FIG. 3. The expression ofRANK mRNA in the femurs of(A) 8-week-old male, (B and C)2.5-year-old male, (D and F)sham-operated female, and (Eand G) ovariectomized rats. (A)Strong expression of RANKmRNA in the small (arrows) andlarge (arrowheads) cells in themetaphyseal cortical bone ofthe 8-week-old rat. (B and C) Theexpression in the osteoclast-likecells (arrowheads) and mononu-clear cells (arrows) in the perios-teum (B) or endosteum (C) of themetaphyseal cortical bone of the2.5-year-old rat. (D and E) Theexpression in the metaphysealcortical bones of the sham-operated (D) and ovariectomized(E) rats. Signals in the small (ar-rows) and large (arrowheads)cells are seen in both specimens.(F and G) The expression ofRANK mRNA in cells on the sur-face of the trabecular bones (ar-rows). (G) Strong expression isseen in the ovariectomized ratbone (3300).

1421EXPRESSION PROFILES OF RANKL, RANK, AND OPG mRNA

expression of RANKL and RANK mRNA in the trabecularbones rather than in the cortical bones was thought toexplain why the increase in bone resorption caused byestrogen deficiency was more severe in the trabecular bonesthan in the cortical bones.(17) The mechanism of the differ-ential regulation of the expression of RANKL and RANKmRNA in different portions of the bone was not clear, butbecause the ovariectomy induced systemic hormonalchange, the differential response of trabecular bone cellsand cortical bone cells to estrogen deficiency might becaused by the difference of cell populations between thetrabecular bone region and cortical bone region. In addition,increased expression of RANKL mRNA in the hypertrophicchondrocytes of the ovariectomized rat bones suggests thatabnormally excessive bone resorption occurs just after en-dochondral bone formation in the region.

Also, it was interesting that the expression level of RANKmRNA was high in osteoclastic cells and osteoclastprogenitor-like mononuclear cells on the surface of thecortical and trabecular bones but not in the bone marrowcells. Our PCR analyses indicated that bone marrow mac-rophages express RANK mRNA (Ikeda and Utsuyama,unpublished data, 2000), and the expression was thought tobe enhanced on the surface of the cortical and trabecular

bones. These results suggest that the expression of RANKmRNA increases to a level high enough to induce osteoclas-togenesis on the surface of the bone matrix, and this mightexplain why osteoclasts are formed only on the surface ofthe bone matrix.

OPG mRNA was shown to be expressed in cartilages ina mouse embryo,(9) but the expression profile in maturebone tissue had not been reported. In this study, we showedthat OPG mRNA was expressed in proliferative chondro-cytes, osteocytes, and a subset of periosteal cells in the8-week-old rat bones. Relatively strong expression was seenin the proliferative chondrocytes and osteocytes. OPGworks as a decoy receptor for RANKL and inhibits oste-oclastogenesis by competing with RANK when binding toRANKL.(8) The expression profile of OPG mRNA wasdifferent from that of RANK mRNA. The expression ofOPG mRNA in the proliferative chondrocytes may notaffect osteoclastogenesis considering the fact that there is novascularization in the cartilage, but there is a possibility thatOPG is accumulated in the cartilage and protects it fromresorption. Surprisingly, in the ovariectomized rat bones,strong expression of OPG mRNA was induced in the restingchondrocytes (Fig. 5B). In human osteoblastic cells, theexpression of OPG mRNA was indicated to be stimulated

FIG. 4. The expression of OPGmRNA in the femurs of (A, C,and E) 8-week-old male and (B,D, and F) 2.5-year-old male rats.(A) The expression in prolifera-tive chondrocytes in the articularregion of the 8-week-old rat (ar-rows). (B) No expression of OPGmRNA in the articular region ofthe 2.5-year-old rat. (C and D)The expression in the prolifera-tive chondrocytes in the growthplate of the (C, arrows) 8-week-old and (D, arrows) 2.5-year-oldrat bones. (E) The expression ofOPG mRNA in a subset of peri-osteal cells (arrows) and an osteo-cyte (an arrowhead) of the me-taphyseal cortical bone of the8-week-old rat. (F) Equivalent re-gion to panel E in the 2.5-year-old rat. No expression of OPGmRNA in this region (3300).

1422 IKEDA ET AL.

by estrogen.(27,28) In addition, serum OPG was shown toincrease in women with postmenopausal osteoporosis.(29)

Our data seemed to conflict with these data. In this study,we could not detect the expression of OPG mRNA inosteoblastic cells, and in vivo change in expression of OPGmRNA in osteoblasts remained unknown. The meaning ofthis strong induction of OPG mRNA in the resting chon-drocytes and osteocytes was not clear, but our resultsstrongly suggested that regulation of OPG expression inchondrocytes and osteocytes was different from that ofosteoblasts.

The expression of OPG mRNA in the osteocytes is inter-esting. The increase in the number of osteocytes expressing

OPG mRNA in the ovariectomized rat bones suggests aprotective reaction against excess bone resorption caused byovariectomy. The excess bone resorption exposes the osteo-cytes expressing OPG mRNA at the surface of the bone,which may inhibit osteoclastogenesis. In the ovariectomizedrat bones, the expression of OPG mRNA also was inducedin cells on the surface of the trabecular bones (Fig. 5H). Inthis study, we could not characterize these cells, but therewas a possibility that they were identical to the osteocytesexpressing OPG mRNA, which were exposed at the surfaceof the bone.

There were some inconsistencies in the expression be-tween 8-week-old male rat bones and sham-operated female

FIG. 5. The expression of OPGmRNA in the femurs of (A, C, E,and G) sham-operated female and(B, D, F, and H) ovariectomizedrats. (A and B) The expression ofOPG mRNA in the articular re-gion. Proliferative chondrocyteswith positive signal (arrowheads)are seen in both (A) sham-operated and (B) ovariectomizedrat bones. Note the high level ofexpression in the resting chondro-cytes in the ovariectomized ratbone shown in panel B (arrows).(C and D) The expression of OPGmRNA in the osteocytes (arrow-heads) of the cortical bones in the(C) sham-operated and (D) ovari-ectomized rats. An increase in thenumber of osteocytes with a pos-itive signal is observed in theovariectomized rat bone shown inpanel D. (E and F) Upper portionof the growth plate and a part ofthe epiphysis. (E) No expressionof OPG mRNA in the sham-operated rat bone and (F) weakexpression in the resting chondro-cytes (arrows) in the ovariecto-mized rat bone. (G and H) Lowerportion of the growth plate and apart of the primary spongiosa ofthe metaphysis. (G) No expres-sion of OPG mRNA in the sham-operated rat bone (H) and mRNAexpression in a subset of cells onthe surface of the primary spon-giosa (arrows) in the ovariecto-mized rat bone (3300).

1423EXPRESSION PROFILES OF RANKL, RANK, AND OPG mRNA

rat bones. There are three possibilities to explain theseinconsistencies: difference between 8-week-old rat bonesand 13-week-old rat bones, difference between nontreatedrat bones and sham-operated rat bones, or difference be-tween male rat bones and female rat bones. In the prolifer-ative chondrocytes, OPG mRNA was detected only in themale rat bones and was not detected in the female rat bonesirrespective of age and operation. Therefore, it was sug-gested that the difference in the expression of OPG mRNAin the proliferative chondrocytes was caused by sex differ-ence.

In this study, we analyzed profiles of the expression ofRANKL, RANK, and OPG mRNA in the 8-week-old and2.5-year-old male rat bones and in the sham-operated andovariectomized female rat bones. The expression ofRANKL, RANK, and OPG mRNA was relatively stablymaintained in the endosteal region. In addition, high expres-sion of OPG mRNA was induced in resting chondrocytesand osteocytes in the ovariectomized rat bones.

ACKNOWLEDGMENTS

We thank Sachiko Seki and Isao Inada for their technicalassistance. This work was supported by the Research Grantfor Longevity Sciences (10–3) from the Ministry of Healthand Welfare.

REFERENCES

1. Anderson DM, Maraskovsky E, Billingsley WL, Dougall WC,Tometsko ME, Roux ER, Teepe MC, DuBose RF, Cosman D,Galibert L 1997 A homologue of the TNF receptor and itsligand enhance T cell growth and dendritic-cell function. Na-ture 390:175–179.

2. Lacey DL, Timms E, Tan H-L, Kelley MJ, Dunstan CR,Burgess T, Elliott R, Colombero A, Elliott G, Scully S, Hsu H,Sullilvan J, Hawkins N, Davy E, Capparelli C, Eli A, QianY-X, Kaufman S, Sarosi I, Shalhoub V, Senaldi G, Guo J,Delaney J, Boyle WJ 1998 Osteoprotegerin ligand is a cyto-kine that regulates osteoclast differentiation and activation.Cell 93:165–176.

3. Yasuda H, Shima N, Nakagawa N, Yamaguchi K, Kinosaki M,Mochizuki S, Tomoyasu A, Yano K, Goto M, Murakami A,Tsuda E, Morinaga T, Higashio K, Udagawa N, Takahashi N,Suda T 1998 Osteoclast differentiation factor is a ligand forosteoprotegerin/osteoclastogenesis-inhibitory factor and isidentical to TRANCE/RANKL. Proc Natl Acad Sci USA95:3597–3602.

4. Wong BR, Josien R, Lee SY, Sauter B, Li HL, Steinman RM,Choi Y 1997 TRANCE (tumor necrosis factor [TNF]-relatedactivation-induced cytokine), a new TNF family member pre-dominantly expressed in T cells, is a dendritic cell-specificsurvival factor. J Exp Med186:2075–2080.

5. Hsu H, Lacey DL, Dunstan CR, Solovyev I, Colombero A,Timms E, Tan HL, Elliott G, Kelley MJ, Sarosi I, Wang L, XiaX-Z, Elliott R, Chiu L, Black T, Scully S, Capparelli C,Morony S, Shimamoto G, Bass MB, Boyle W 1999 Tumornecrosis factor receptor family member RANK mediates os-teoclast differentiation and activation induced by osteoprote-gerin ligand. Proc Natl Acad Sci USA96:3540–3545.

6. Lum L, Wong BR, Joisen R, Becherer D, Erdjument-BromageH, Schlondorff J, Tempst P, Choi Y, Blobel CP 1999 Evidencefor a role of a tumor necrosis factor-a (TNF-a)-converting

enzyme-like protease in shedding of TRANCE, a TNF familymember involved in osteoclastogenesis and dendritic cell sur-vival. J Biol Chem274:13613–13618.

7. Suda T, Takahashi N, Udagawa N, Jimi E, Matthew T,Gillespie T, Martin TJ 1999 Modulation of osteoclast differ-entiation and function by the new members of the tumornecrosis factor receptor and ligand families. Endocr Rev20:345–357.

8. Nakagawa N, Kinosaki M, Yamaguchi K, Shima N, Yasuda H,Yano K, Morinaga T, Higashio K 1998 RANK is the essentialsignaling receptor for osteoclast differentiation factor in oste-oclastogenesis. Biochem Biophys Res Commun253:395–400.

9. Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang M-S,Luthy R, Nguyen HQ, Wooden S, Bennett L, Boone T, Shi-mamoto G, DeRose M, Elliott R, Colombero A, Tan H-L, TrailG, Sullivan J, Davy E, Bucay N, Renshaw-Gegg L, HughesTM, Hill D, Pattison W, Campbell P, Sander S, Van G,Tarpley J, Derby P, Lee R, Amgen EST Program, Boyle WJ1997 Osteoprotegerin: A novel secreted protein involved in theregulation of bone density. Cell89:309–319.

10. Yasuda H, Shima N, Nakagawa N, Mochizuki S, Yano K,Fujise N, Sato Y, Goto M, Yamaguchi K, Kuriyama M, KannoT, Murakami A, Tsuda E, Morinaga T, Higashio K 1998Identity of osteoclastogenesis inhibitory factor (OCIF) andosteoprotegerin (OPG): A mechanism by which OPG/OCIFinhibits osteoclastogenesis in vitro. Endocrinology139:1329–1337.

11. Udagawa N, Takahashi N, Akatsu T, Sasaki T, Yamaguchi A,Kodama H, Martin TJ, Suda T 1989 The bone marrow-derivedstromal cell lines MC3T3–G2/PA6 and ST2 supportosteoclast-like cell differentiation in cocultures with mousespleen cells. Endocrinology125:1805–1813.

12. Udagawa N, Takahashi N, Akatsu T, Tanaka H, Sasaki T,Nishihara T, Koga T, Martin TJ, Suda T 1990 Origin ofosteoclasts: Mature monocytes and macrophages are capableof differentiating into osteoclasts under a suitable microenvi-ronment prepared by bone marrow-derived stromal cells. ProcNatl Acad Sci USA87:7260–7264.

13. Shioi A, Ross FP, Teitelbaum SL 1994 Enrichment of gener-ated murine osteoclasts. Calcif Tissue Int55:387–394.

14. Inoue M, Namba N, Chappel J, Teitelbaum SL, Ross FP 1998Granulocyte macrophage-colony stimulating factor recipro-cally regulates associated integrins on murine osteoclast pre-cursors. Mol Endocrinol12:1955–1962.

15. Kong Y-Y, Yoshida H, Sarosi I, Tan H-L, Timms E, Cappar-elli C, Morony S, Oliveria-dos-Santos AJ, Van G, Itie A, KhooW, Wakeham A, Dunstan CR, Lacey DL, Mak TW, Boyle WJ,Penninger JM 1999 OPGL as a key regulator of osteoclasto-genesis, lymphocyte development and lymph-node organogen-esis. Nature397:315–323.

16. Ikeda T, Nagai Y, Yamaguchi A, Yokose S, Yoshiki S 1995Age-related reduction in bone matrix protein mRNA expres-sion in rat bone tissues: Application of histomorphometry to insitu hybridization. Bone16:17–23.

17. Ikeda T, Yamaguchi A, Yokose S, Nagai Y, Yamato H,Nakamura T, Tsurukami H, Tanizawa T, Yoshiki S 1996Change in biological activity of bone cells in ovariectomizedrats revealed by in situ hybridization. J Bone Miner Res11:780–788.

18. Melton LJ, Atkinson EJ, O’Conner MK, O’Fallon WM, RiggsBL 1998 Bone density and fracture risk in men. J Bone MinerRes13:1915–1923.

19. Ikeda T, Nomura S, Yamaguchi A, Suda T, Yoshiki S 1992 Insitu hybridization of bone matrix protein in undecalcified adultrat bone sections. J Histochem Cytochem40:1079–1088.

20. Kalu DN 1991 The ovariectomized rat model of postmeno-pausal bone loss. Bone Miner15:175–192.

1424 IKEDA ET AL.

21. Wronski TJ, Cintro´n M, Dann LM 1988 Temporal relationshipbetween bone loss and increased bone turnover in ovariecto-mized rats. Calcif Tissue Int43:179–183.

22. Wronski TJ, Dann LM, Horner SL 1989 Time course ofvertebral osteopenia in ovariectomized rats. Bone10:295–301.

23. Turner RT, Vandersteenhoven JJ, Bell NH 1987 The effects ofovariectomy and 17b-estradiol on cortical bone histomor-phometry in growing rats. J Bone Miner Res2:115–122.

24. Wronski TJ, Cintro´n M, Dann LM 1988 Estrogen treatmentprevents osteopenia and depress bone turnover in ovariecto-mized rats. Endocrinology123:681–686.

25. Takano-Yamamoto T, Rodan G 1990 Direct effects of 17b-estradiol on trabecular bone in ovariectomized rats. Proc NatlAcad Sci USA87:2172–2176.

26. Jee WSS The skeletal tissues. In: Weiss L (ed.) Histology, Celland Tissue Biology, 5th ed. The MacMillan Press, London,UK, p. 200.

27. Hofbauer LC, Khosla S, Dunstan CR, Lacey DL, SpelsbergTC, Riggs BL 1999 Estrogen stimulates gene expression andprotein production of osteoprotegerin in human osteoblasticcells. Endocrinology140:4367–4370.

28. Hofbauer LC, Gori F, Riggs BL, Lacey DL, Dunstan CR,Spelsberg TC, Khosla S 1999 Stimulation of osteoprotegerin

ligand and inhibition of osteoprotegerin production by glu-cocorticoids in human osteoblastic lineage cells: Potentialparacrine mechanism of glucocorticoid-induced osteoporosis.Endocrinology140:4382–4389.

29. Yano K, Tsuda E, Washida N, Kobayashi F, Goto M, Harada A,Ikeda K, Higashio K, Yamada Y 1999 Immunological character-ization of circulating osteoprotegerin/osteoclastogenesis inhibi-tory factor: Increased serum concentrations in postmenopausalwomen with osteoporosis. J Bone Miner Res14:518–527.

Address reprint requests to:Tohru Ikeda, Ph.D.

Department of Pathology and ImmunologyAging and Developmental Science

Graduate School, Tokyo Medical and Dental University1–5-45 Yushima, Bunkyo-ku

Tokyo 113-8519, Japan

Received in original form October 4, 2000; in revised form Feb-ruary 13, 2001; accepted March 22, 2001.

1425EXPRESSION PROFILES OF RANKL, RANK, AND OPG mRNA