Bone 32 (2003) 136 141 www.elsevier.

com/locate/bone

The effects of estrogen on osteoprotegerin, RANKL, and estrogen

receptor expression in human osteoblasts
S. Bord,* D.C. Ireland, S.R. Beavan, and J.E. Compston
University of Cambridge School of Clinical Medicine, Addenbrookes Hospital, Box 157, Cambridge CB2 2QQ, UK
Received 28 February 2002; revised 28 August 2002; accepted 8 October 2002

Abstract
Estrogen is essential for bone growth and development and for the maintenance of bone health in adulthood. The cellular responses of
osteoblasts and osteoclasts to estrogen are initiated via two high-affinity receptors (ERs). Osteoblasts synthesize RANKL (receptor activator
of NF-B ligand), necessary for osteoclast formation and function, and osteoprotegerin (OPG), its decoy receptor. To investigate the effects
of estrogen on the expression of OPG, RANKL, and ERs in human osteoblasts, cells were cultured with physiological (1010 M) and
high-dose (107 M) 17-estradiol for 24 and 48 h. Proteins and corresponding mRNA levels were quantitatively determined by
immunocytochemistry and RTPCR. OPG expression was significantly increased three- and sevenfold at 24 h with 1010 M (P 0.05)
and 107 M (P 0.01) estradiol, respectively, compared to untreated cells. Similar but smaller increases were seen at 48 h (P 0.05).
Osteoblasts treated with estradiol demonstrated increased RANKL protein expression at 24 h (P 0.05), but this was not maintained at 48 h.
ER expression was significantly increased by high-dose estradiol (P 0.01) at 24 h and dose-dependently increased at 48 h (P 0.01),
while ER was only increased at 24 h (P 0.01). The estrogen-induced protein expression of ER, OPG, and RANKL was abrogated when
cells were cultured in the presence of the estrogen antagonist ICI 182,780. mRNA levels at 24 h demonstrated a significant suppression of
RANKL with the low-dose but not the high dose. ER mRNA but not ER expression was up-regulated by estrogen. Our results suggest
that estrogen may exert its anti-resorptive effects on bone, at least in part, by stimulating ER and OPG expression in osteoblasts.
2003 Elsevier Science (USA). All rights reserved.

Keywords: Estrogen; Osteoblasts; Osteoprotegerin; RANKL; Estrogen receptors

Introduction mone receptor and a member of a family of activated tran-

It is well established that estrogen has multifunctional
roles influencing growth, differentiation, and function in
many tissues. It is an important factor in the maintenance of
bone health, estrogen deficiency at the time of menopause
being associated with bone loss. Administration of conven-
tional doses of hormone replacement therapy prevents
menopausal bone loss by reducing bone turnover and inhib-
iting osteoclast activity [1], whereas high doses of estrogen
have been shown to exert anabolic skeletal effects in rodents
and postmenopausal women [25]. Estrogens diffuse in and
out of cells, but are retained in target cell nuclei by the
estrogen receptor protein (ER). The ER is a nuclear hor-
* Corresponding author. Fax: 44-0-1223-336846.
E-mail address: sb201@cam.ac.uk (S. Bord).
scription factors that can initiate or enhance the transcrip-
tion of genes. Two ER subtypes have been identified, and
. They are distinct proteins encoded by separate genes and
located on different chromosomes.
RANKL (receptor activator of NF-B ligand), a mem-
brane-bound molecule, is a newly identified member of the
tumor necrosis factor (TNF) ligand family and has been
shown to be crucial for osteoclast formation [6]. Two re-
ceptors for RANKL have been identified, RANK, a mem-
brane-bound signaling receptor expressed on the cell sur-
face of osteoclast progenitors, and osteoprotegerin (OPG), a
secreted cytokine receptor. OPG, a member of the tumor
necrosis factor receptor (TNF-R) superfamily [8,9] acts as a
decoy receptor by blocking the interaction of RANKL with
its functional receptor RANK [10], thereby inhibiting oste-
oclastogenesis.

8756-3282/03/$ see front matter 2003 Elsevier Science (USA). All rights reserved.
doi:10.1016/S8756-3282(02)00953-5
 S. Bord et al. / Bone 32 (2003) 136 141
In vitro studies have demonstrated the synthesis of OPG
by stromal cell lines [11] and osteoblasts [12] and its stim-
ulation by estrogen [13]. Udagawa et al. [14] reported that
osteoblasts constitutively express RANKL mRNA, the bone
resorbing capacity of osteoclasts being enhanced when
cocultured with osteoblasts.
The aim of this study was to investigate the relative
effects of estrogen on ER, OPG, and RANKL mRNA and
protein expression in human osteoblasts.
Materials and methods
Cell culture and immunolocalization
Primary human osteoblasts from two young female do-
nors (1 day and 1 year) were supplied by Clontech Labo-
ratories (Basingstoke, UK), with each grown to confluence
and expanded to yield cells for three replica experiments.
Cells were seeded into eight-well chamber slides (NUNC)
at 104 cells/well in McCoys 5A medium supplemented with
10% heat-inactivated FBS (Life Technologies), penicillin/
streptomycin (Life Technologies), and ascorbic acid (100
mM, Wako, Alpha Labs, Eastleigh, UK). Cultures were
incubated at 37 C in a humidified chamber with 5% CO2.
After 24 h the medium was removed and replaced with
medium minus or plus 17-estradiol (E2, Sigma), at phys-
iological E2 (low dose, 1010 M) or a saturating concentra-
tion of E2 (high dose, 107 M) for 24 or 48 h. Similar
cultures were treated with the estrogen antagonist ICI 182,
780 (108 M, Tocris Cookson) for 24 h.
At the end of the incubation period (24 or 48 h) the
medium was removed and cells were rinsed in phosphate-
buffered saline (PBS), pH 7.4, fixed with 4% paraformal-
dehyde for 5 min at room temperature, and immunostained
using an indirect immunoperoxidase method. Following fix-
ation, cells were washed with PBS and endogenous perox-
idase activity was blocked with ImmunoPure Suppressor
(Pierce) for 30 min at room temperature. Further PBS
washes were followed by blocking with 15% normal horse
serum (Vector Laboratories, Peterborough, UK) for 30 min
at room temperature. Primary antibodies to ER (1.0 g/ml,
rabbit polyclonal HC-20 mapping to the carboxy terminus
of ER and shown not to cross-react with ER) and ER
(1.0 g/ml, rabbit polyclonal H-150 raised against a recom-
binant protein corresponding to amino acids 1150 mapping
at the amino terminus of human ER and shown to be
non-cross-reactive with ER), OPG (1.0 g/ml, goat poly-
clonal N-20 raised against a peptide mapping the amino
terminus of human OPG), and RANKL (1.0 g/ml, goat
polyclonal C-20 raised against a peptide mapping at the
carboxy terminus of human RANKL) were obtained from
Santa Cruz Biotechnology (Santa Cruz, CA). These anti-
bodies, diluted in 0.1% bovine serum albumin (Vector Lab-
oratories), were added and incubated overnight at 4 C in a
humidified chamber. Unbound primary antibody was re-
137
moved by washing in PBS prior to addition of a biotinylated
secondary antibody (horse antirabbit or horse antigoat, Vec-
tor Laboratories, at 3.5 g/ml) for 30 min at room temper-
ature. Following washings with PBS to remove unbound
antibody the signal was amplified by the addition of an
avidin biotin complex (ABC) Elite substrate (Vector Lab-
oratories). Signal was detected using DAB (Vector Labora-
tories). Cells were lightly counterstained with Gills Haema-
toxylin (1:50, Sigma) for 1 min, blued in tap water for 10
min, and air-dried prior to mounting. The specificity of the
antibody reaction was ascertained by substituting the pri-
mary antibody with nonimmune serum at the same IgG
concentration and omission of primary and secondary anti-
bodies.
Quantitation of immunolocalization
Cells were examined by light microscopy with a Nikon
E-800 fitted with a Basler digital camera. The intensity and
extent of staining were measured using Lucia G image
analysis. Thresholds were set to detect only positive stain-
ing with five fields of view examined for each antibody for
each experiment. Experiments were repeated three times.
Results in each group are expressed as the mean relative
change in the E-treated cells compared to untreated cells
(SD).
Cell culture for RNA determination and relative
quantification of gene expression
Cells (as above) were plated at 105 cells/ml and cultured
in T 75 flasks in medium (as above) for 24 h. The medium
was removed and replaced with fresh medium containing no
E2, low-dose E2, or high-dose E2 as above for 24 h. Cells
were removed with TRIzol and RNA was extracted accord-
ing to manufacturers instructions (Life Technologies, Inc.,
Paisley, UK).
To prepare RNA standards PCR products for ER, ER,
OPG, RANKL, and GAPDH were made by RTPCR with
primer sequences shown in Table 1. PCR products were
cloned into the T-tailed vector pCR II-TOPO (Invitrogen),
which has opposing SP6 and T7 RNA polymerase promotor
sites. BamHI restriction sites were added to forward primers
so that the orientation of each insert could be determined by
cutting the plasmids with BamHI. Plasmids containing in-
serts suitable for transcription with SP6 RNA polymerase
were selected and cut with NotI. RNA transcripts were
purified using Microspin S-300 columns (Pharmacia) after
treatment with DNase to remove plasmid DNA. mRNA
levels were determined using one-step RTPCR reagents
(Applied Biosystems) in a Gene Amp 5700 SDS with
Primer Express software used to design primers and probes.
Primers pairs were chosen to include introns in the gene
sequence and probes to span intron exon boundaries (Table
2). A standard curve was included in each assay so that the
overall efficiency of the assay could be calculated. mRNA
138 S. Bord et al. / Bone 32 (2003) 136 141

Table 1
Sequences of cloning primers

Gene Forward primer Reverse primer

GAPDH TGAAGGTCGGAGTCAACGGATTTG GTTGGTGGTGCAGGAGGCATTGCT

ER* AATTCAGATAATCGACGCCAG GTGTTTCAACATTCTCCCTCCTC
ER* TAGTGGTCCATCGCCAGTTAT GGGAGCCACACTTCACCAT
OPG GTCAAGCAGGAGTGCAATCG GAGTTGATTCACTGTTTCCGG
RANKL ATCCCACTCGGTTCCCATA CCCTGACCAATACTTGGTGC

* Arts et al. (1997) [20].

levels were quantified using the comparative threshold-cy-
cle (CT) method [7]. First the amount of target mRNA in
each sample was normalized to the amount of housekeeper
mRNA (GAPDH), designated as calibrator, to give CT (CT
target CTGAPDH). Second, the amounts of target mRNA in
the samples were compared using the formula Amount of
target mRNA 2CT, where CT CTsample1
CTsample2 (calibrator), assuming that the efficiencies of the
PCR reaction were close to 1. The efficiency of each assay
was calculated using the formula E 101/S 1, where S
is the slope of the standard curve.
Four replicates were performed for each experimental
point and experiments repeated three times. The results
shown (Fig. 3) are of a representative experiment.
Statistical analysis
Statistical analysis for protein and mRNA data was per-
formed using the approximate test for unequal variance
based on the t distribution [15].
P 0.01). At 48 h there was a more modest but still
significant increase in OPG expression in the estrogen-
treated cells (low dose, 1.2 0.09, high dose, 1.3 0.16,
P 0.05) (Fig. 1). RANKL expression was elevated in the
presence of E2 at 24 h but not at 48 h. At 24 h there was a
2.35-fold (0.74, P 0.05) increase with low-dose E2 and
a 2.1-fold (0.67, P 0.05) increase with high-dose E2
compared to untreated cells (Fig. 1). ER expression varied
with time and dose of E2 treatment. At 24 h high-dose E2
elicited a 2.05-fold (0.11, P 0.01) increase in ER ex-
pression but there was no change in the low-dose E2-treated
cells. At 48 h a dose-dependent increase was observed (low
dose, 2.14 0.44, high-dose, 2.57 0.68, P 0.01) (Fig. 1).
ER exhibited a significant dose-dependent increase at 24 h
(low dose, 1.72 0.20, high dose, 2.05 0.44, P 0.01) but
had returned to basal levels at 48 h (Fig. 1).
Cells cocultured with E2 and the estrogen antagonist ICI
182,780 showed no estrogen-induced increase in OPG,
RANKL, or ER expression. In all cultures the expression
levels remained at basal levels (Fig. 2).

Results
Protein expression
The result for untreated cells in each group was given a
value of 1, and data for treated cells in each group were
expressed as a relative ratio. OPG protein expression mea-
sured at 24 h demonstrated a significant 3-fold increase with
low-dose E2 compared to untreated cells (3.00 1.05, P
0.05) and a 7-fold increase with high-dose E2 (6.9 1.80,
mRNA expression
mRNA data were collected at 24 h to determine that
changes in protein levels were as a result of changes in
transcription rather than proteasome-mediated receptor deg-
radation or translation. The amount of target mRNA for
each sample was normalized to the amount of the house-
keeper mRNA (GAPDH). GAPDH expression did not
change significantly with estradiol treatment. Approximate
relative levels of mRNA expression with GAPDH as 1 were
OPG, 0.25; ER, 0.00025; ER, 0.00004; and RANKL,

Table 2
GeneAmp 5700 SDS primers and probes

Gene Forward primer Reverse primer Probe

GAPDH TTTTAACTCTGGTAAAGTGGATATTGTTG TGACGGTGCCATGGAATTT ATTGACCTCAACTACATGGTTTACA

TGTTCCAATATG
ER TGCTTCAGGCTACCATTATGGA GTTTTTATCAATGGTGCACTCGTA ACACATATAGTCGTTATGTCCTTG
AATACTTCTCTTGAAGAA
ER TCAAAAGAGTCCCTGGTGTGAAG CTCTTTGAACCTGGACCAGTAACAG CGGTTCCCACTAACCTTCCTTTTC
AGTGTCTCT
OPG GAGATAGAGTTCTGCTTGAAACA CCATCTGGACATCTTTTGCAAA TGCAAGCTGGAACCCCAGAGCG
RANKL CAGCCTTTTGCTCATCTCACTATTAA GGTACCAAGAGGACAGACTCACTTTA ACCGACATCCCATCTGGTTCCC
 S. Bord et al. / Bone 32 (2003) 136 141 139

Fig. 1. Immunohistochemistry of osteoprotegerin (OPG), RANKL, and estrogen receptor (ER) and protein expression by osteoblasts at 24 and 48 h was
quantitatively measured by image analysis. Results from untreated cells were given a value of 1 and estrogen-treated cells a relative ratio. Results SD, *
P 0.05, ** P 0.01, compared to untreated cells.

0.00002. To compare changes in expression levels with
treatment, the result for untreated cells in each group was
given a value of 1, and data for treated cells in each group
were expressed as a relative ratio. At 24 h RANKL mRNA
expression was suppressed by low-dose E2 (P 0.05) but
not by by high-dose E2. OPG, ER, and ER mRNA
expressions were unchanged at 24 h with high-dose E2
treatment. With low-dose E2 treatment only ER was in-
creased (P 0.05) (Fig. 3).
Discussion
This in vitro study demonstrated that estrogen elicits
changes in cellular protein and mRNA expression in human
osteoblastic cells with OPG and ER protein expression
significantly increased. The estrogen-induced changes in
ER expression and up-regulation of the decoy receptor OPG
suggest that osteoblasts may be involved in the mediation of
the estrogen-induced anabolic skeletal effects in bone. We
and other groups have demonstrated the presence of ERs in
osteoblasts [16 18] and Udagawa et al. [14] reported the
constitutive expression of RANKL by osteoblasts. In addi-
tion, various reports show the induction of OPG by estrogen
[13,19]. However, to our knowledge, this is the first report
of estrogen-induced changes in OPG, RANKL, and ER
expression measured in the same human osteoblast cultures.
The use of image analysis in our study provided quantitative
data for protein synthesis that adds to the observational
reports in other studies.
We have previously reported the presence of both ER
and ER in human osteoblastic cells in vitro and in devel-
oping human bone in vivo [16]. The current study demon-
strates that in vitro osteoblastic cells respond initially to
estrogen by increasing ER and protein expression. By
48 h ER expression had returned to basal levels, whereas
ER expression continued to rise. From this study it was not
possible to establish the time point of maximal mRNA
expression or at which time point this was reflected by
maximal protein expression. However, the mRNA data at
24 h were generally in concordance with the 48-h protein
expression. ER showed a significant increase at 24 h,
which agreed with the protein data at 48 h, while ER

Fig. 2. Immunohistochemistry of OPG, RANKL, ER and ER protein

expression by osteoblasts at 24 h with and without the estrogen antagonist
ICI 182,780 was quantitatively measured by image analysis. Results from Fig. 3. OPG, RANKL, ER, and ER mRNA expression by osteoblasts
untreated cells were given a value of 1 and estrogen-treated cells a relative was measured at 24 h. Results were normalized to GAPDH. Results from
ratio. Results SD, * P 0.05, ** P 0.01, compared to untreated untreated cells were given a value of 1 and estrogen-treated cells a relative
cells. ratio. Results SD, * P 0.05, compared to untreated cells.
140 S. Bord et al. / Bone 32 (2003) 136 141
mRNA demonstrated no change, which was reflected by the
protein expression at 48 h. Although the results for OPG and
ER demonstrated the same trends as the protein data these
results were not significant due to the wide variability be-
tween samples.
Studies of human breast cancer tissue have shown that
ER-negative cells are more proliferative and less differen-
tiated than ER-positive cells. In the ER-positive tumors
the amount of ER mRNA was correlated to age and post-
menopausal status, higher levels being demonstrated in el-
derly women [11]. In studies conducted in our laboratory we
have found that proliferating osteoblastic cells from young
donors do not show increasing ER/ER with time in
culture, unlike cells from older donors, which mineralize
more rapidly and express increasing ER mRNA [17].
Consistent with this, Arts et al. [20], using a mineralizing
immortalized osteoblast-like cell line, showed differential
mRNA expression of ER and ER during cell differenti-
ation. ER expression was found to increase during culture,
whereas ER levels, after an initial increase, stayed con-
stant. In our current study, using young donor osteoblasts,
ER continued to rise, while ER had returned to basal
values at 48 h, indicating that the cells had not reached the
mineralizing stage and collagen synthesis was not down-
regulated. Together these reports suggest a potential mech-
anism for the observed anabolic effect produced by high-
dose estrogen, with the ER/ER ratio dependent on the
age of the person, the dose, and the length of time of
estrogen treatment. However, our complete understanding
of the mechanisms of regulation of ERs has yet to be
established. A recent study by Waters and colleagues [21]
using osteoblast cell lines stably expressing either ER or
ER identified responses to estrogen that were both ER
isoform-specific and differentiation stage-dependent, which
is consistent with reports from Ireland et al. [17]. Bonnelye
and Aubin [22] have shown that the estrogen receptor-
related receptor is codistributed with ER and ER in
different cohorts of osteoblastsic cells, raising the possibil-
ity that the receptors function in different pairs in different
groups of osteoblasts. These observations give rise to some
intriguing questions.
RANKL, an essential factor for osteoclast formation and
activity, and OPG are regulated by various hormones, cy-
tokines, and mesenchymal transcription factors. However,
OPG, which acts as a soluble neutralizing receptor to
RANKL, is produced by cells of the osteoblastic lineage and
therefore likely to be a major regulator of bone metabolism.
Severe osteoporosis develops in mice with OPG gene dele-
tion [23], whereas mice overexpressing OPG develop os-
teopetrosis [22]. Recently Croucher et al. [24] reported
expression of RANKL in myeloma cells in mice. The in-
crease in osteoclast number and the associated development
of bone disease, osteolytic lesions, and decrease in bone
volume were prevented by treatment with recombinant
OPG, indicating the importance of OPG in vivo. Hofbauer
et al. [13], using an osteoblast-like cell line stably trans-
fected with ER, showed that 17-estradiol induced a dose-
dependent increase in OPG protein secretion after 24 h of
culture. This increase was completely abrogated by cocul-
ture with the estrogen antagonist ICI 182,780. They also
demonstrated that the estrogen-induced effect was specific
to 17-estradiol, as the estrogen isomer 17-estradiol had
no stimulatory effect. Using a primary cell line from young
human donors we have demonstrated similar effects to the
transfected cells used by Hofbauer et al. [12,13]. We quan-
tified cellular synthesis of OPG in response to estrogen and
show that protein expression is significantly increased in
estrogen-treated cells at both 24 and 48 h. However, OPG
synthesis was measured only in osteoblasts and it is possible
that levels of secreted OPG in the culture medium may
differ between treated and untreated cultures. Further work
is required to determine this fact. RANKL is a functional
protein when expressed at the cell membrane. In our study
intense staining was seen on the cell surface of many os-
teoblasts and image analysis detection thresholds were set to
identify these areas of dark staining. However, some intra-
cellular intense staining was observed. As this might be
representative of RANKL trafficking to the cell surface
these measurements were retained.
Both doses of estrogen produced an increase in RANKL
protein synthesis at 24 h. As estrogen is known to exert an
anti-resorptive effect on bone this finding was unexpected.
However, protein levels had returned to basal values by
48 h, leaving only elevated OPG. Thus, in this system,
estrogen could alter the OPG/RANKL ratio against oste-
oclastogenesis.
Interestingly in a study by Collin-Osdoby and colleagues
[25] increases in RANKL and OPG mRNA expression were
seen in endothelial cells following an inflammatory stimu-
lus. The RANKL/OPG ratio increased because OPG levels
declined. In the same study experiments performed on
HBMSC cultured in the presence of PTH and vitamin D
also showed a reciprocal expression pattern, with an early
rise in OPG followed by a declined together with a delayed
and sustained rise in RANKL. In contrast HMVEC treated
with these calciotrophic agents showed no change in their
expression patterns, indicting that different stimuli may
affect the OPG/RANKL ratio in different ways and in dif-
ferent cell types.
To investigate the mechanism by which estrogen may
induce protein expression, osteoblasts were cultured with
estradiol in the presence or the absence of the estrogen
antagonist ICI 182,780. The low- and high-dose estrogen-
induced expression of OPG, RANKL, and ER was abro-
gated by the estrogen antagonist, suggesting that the actions
of estrogen on these cytokines are mediated by ERs.
In summary, the data presented here suggest that es-
trogen may exert its anti-resorptive effects on bone, at
least in part, by stimulating ER and OPG expression in
osteoblasts.
 S. Bord et al. / Bone 32 (2003) 136 141
Acknowledgments
This work was funded by the Wellcome Trust.
References
[1] Vedi S, Compston JE. The effects of long-term hormone replacement
therapy on bone remodelling in postmenopausal women. Bone 1996;
19:5359.
[2] Chow J, Tobias JH, Colston KW, Chambers TJ. Estrogen maintains
trabecular volume in rats not only by suppression of resorption but
also by stimulation of bone formation. J Clin Invest 1992;89:74 8.
[3] Edwards MW, Bain SD, Bailey MC, Lantry MM, Howard GA. 17
Estradiol stimulation of endosteal bone formation in the ovariecto-
mised mouse: an animal model for the evaluation of bone-targeted
estrogens. Bone 1992;13:29 34.
[4] Vedi S, Purdie DW, Ballard P, Bord S, Cooper AC, Compston JE.
Bone remodelling and structure in postmenopausal women treated
with long-term, high-dose estrogen therapy. Osteoporos Int 1999;10:
52 8.
[5] Wahab M, Ballard P, Purdie DW, Cooper A, Wilson JC. The effect
of long-term oestradiol implantation on bone mineral density in
postmenopausal women who have undergone hysterectomy and bi-
lateral oophorectomy. Br J Obstet Gynaecol 1997;104:728 31.
[6] Fuller K, Wong B, Choi Y, Chambers TJ. TRANCE is necessary and
sufficient for osteoblast-mediated activation of bone resorption. J Exp
Med 1998;188:9971001.
[7] Fink L, Seeger W, Ermert L, Hanze J, Stahl U, Grimminger F,
Kummer W, Bohle RM. Real-time quantitative RT-PCR after laser-
assisted cell picking. Nature Med 1998;4:1329 33.
[8] Simonet WS, Lacey DL, Dunstan CR, Kelly M, Chang MS, Luthy R,
Nguyen HQ, Wooden S, Bennett L, Boone T, Shimamoto G, DeRose
M, Elliot R, Colombero A, Tan HL, Trail G, Sullivan J, Davey E,
Bucay N, Renshaw-Gegg L, Hughes TM, Hill D, Pattison W, Camp-
bell P, Sander S, Van G, Tarpley J, Derby P, Lee R, Boyle WJ, &
Amgen EST Programme. Osteoprotegerin a novel secreted protein
involved in the regulation of bone density. Cell 1997;89:309 19.
[9] Tsuda E, Goto M, Mochizuki S-I, Yano K, Kobayashi F, Morinaga T,
Higashio K. Isolation of a novel cytokine from human fibroblasts that
specifically inhibits osteoclastogenesis. Biochem Biophys Res Com-
mun 1997;234:137 42.
[10] Anderson DM, Maraskovsky E, Billingsley WL, Dougall WC, Tom-
etsko ME, Roux ER, Teepe MC, DuBose RF, Cosman D, Galibert L.
A homologue of the TNF receptor and its ligand enhance T-cell
growth and dendritic-cell function. Nature 1999;390:1759.
[11] Bieche I, Parfait B, Laurendeau I, Girault I, Vivaud M, Lidereau.
Quantification of estrogen receptor and expression in sporadic
breast cancer. Oncogene 2001;20:8109 55.
[12] Hofbauer LC, Dunstan CR, Spelsberg TC, Riggs BL, Khosla S.
Osteoprotegerin production by human osteoblast lineage cells is stim-
141
ulated by vitamin D, bone morphogenetic protein and cytokines.
Biochem Biophys Res Commun 1998;250:776 81.
[13] Hofbauer LC, Khosla S, Dunstan CR, Lacey DL, Spelsberg TC,
Riggs BL. Estrogen stimulates gene expression and protein produc-
tion of osteoprotegerin in human osteoblastic cells. Endocrinology
1999;140:436770.
[14] Udagawa N, Takahashi N, Jimi E, Matsuzaki K, Tsurukai T, Itoh K,
Nakagawa N, Yasuda H, Goto M, Tsuda E, Higashio K, Gillespie
MT, Martin TJ, Suda T. Osteoblasts/stromal cells stimulate osteoclast
activation through expression of osteoclast differentiation factor/
RANKL but not macrophage colony-stimulating factor. Bone 1999;
25:51723.
[15] Armitage P, Berry G. Statistical inference. In: Statistical methods in
medical research, 3rd ed. Blackwell; 1994, p. 1157.
[16] Bord S, Horner A, Beavan S, Compston J. Estrogen receptors and
are differentially expressed in developing human bone. J Clin
Endocrinol Metab 2001;86:2309 14.
[17] Ireland DC, Bord S, Beavan SR, Compston JE. Effects of estrogen on
collagen synthesis by cultured human osteoblasts depend on the rate
of cellular differentiation. J Cell Biochem 2002;9999:17.
[18] Onoe Y, Miyaura C, Ohta H, Nozawa S, Suda T. Expression of
estrogen receptor in rat bone. Endocrinology 1997;138:4509 12.
[19] Saika M, Inoue D, Kido S, Matsumoto T. 17-Estradiol stimulates
expression of osteoprotegerin by a mouse stromal cell line, ST-2, via
estrogen receptor-. Endocrinology 2001;142:220512.
[20] Arts J, Kuiper GJM, Janssen JMMF, Gustafsson J-A, Lowik CWCM,
Pols HAP, Van Leeuwen JPTM. Differential expression of estrogen
receptors and mRNA during differentiation on human osteoblast
SV-HFO cells. Endocrinology 1997;138:506770.
[21] Waters KM, Rickard DJ, Riggs BL, Khosla S, Katzenellenbogen JA,
Katzenellenbogen BS, Moore J, Spelsberg TC. Estrogen regulation of
human osteoblast function is determined by the stage of differentia-
tion and the estrogen receptor isoform. J Cell Biochem 2001;83:448
62.
[22] Bonnleye E, Aubin JE. Differential expression of estrogen receptor-
related receptor and estrogen receptor and in osteoblasts in vivo
and in vitro. J Bone Miner Res 2002;17:13921400.
[23] Bucay N, Sarosi I, Dunstan CR, Morony S, Tarpley J, Capparelli C,
Scully S, Tan HL, Xu W, Lacey DL, Boyle WJ, Simonet WS.
Osteoprotegerin-deficient mice develop early onset osteoporosis and
arterial calcification. Genes Dev 1998;12:1260 8.
[24] Coucher PI, Shipman CM, Lippitt J, Perry M, Asosingh K, Hijzen A,
Brabbs AC, Van Beek EJR, Holen I, Skerry TM, Dunstan CR, Russell
GR, Van Camp B, Vanderkerken K. Osteoprotegerin inhibits the
development of osteolytic bone disease in multiple myeloma. Blood
2001;98:3534 40.
[25] Collin-Osdoby P, Rothe L, Anderson F, Nelson M, Maloney W,
Osdoby P. Receptor activator of NF-B and osteoprotegerin expres-
sion by human microvascular endothelial cells, regulation by inflam-
matory cytokines, and role in human osteoclastogenesis. J Biol Chem
2001;276:20659 72.