Full text of DOI:10.1359/jbmr.2002.17.6.979

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 17, Number 6, 2002
© 2002 American Society for Bone and Mineral Research
Osteoblastic 11␤-Hydroxysteroid Dehydrogenase Type 1
Activity Increases With Age and Glucocorticoid Exposure
MARK S. COOPER,1 ELIZABETH H. RABBITT,1 PHILIPPA E. GODDARD,2 WILLIAM A. BARTLETT,2
MARTIN HEWISON,1 and PAUL M. STEWART1
ABSTRACT
The risk of glucocorticoid-induced osteoporosis increases substantially with age but there is considerable individual
variation. In recent studies we have shown that the effects of glucocorticoids on bone are dependent on autocrine
actions of the enzyme 11␤-hydroxysteroid dehydrogenase type 1 (11␤-HSD1); expression of 11␤-HSD1 in osteo-
blasts (OBs) facilitates local synthesis of active glucocorticoids with consequent effects on osteoblastic proliferation
and differentiation. Using primary cultures of human OBs, we have now characterized the age-specific variation
in osteoblastic 11␤-HSD1 and defined enzyme kinetics and regulation using natural and therapeutic glucocorti-
coids. 11␤-HSD1 reductase activity (cortisone to cortisol conversion) was recognized in all OB cultures and
correlated positively with age (r ؍ 0.58 with all cultures, p < 0.01, and n ؍ 18; r ؍ 0.87 with calcaneal-derived
cultures, p < 0.001, and n ؍ 14). Glucocorticoid treatment caused a time- and dose-dependent increase in
11␤-HSD1 activity over control (e.g., dexamethasone [DEX; 1 ␮M], 2.6-fold ؎ 0.5 (mean ؎ SE), p < 0.001, and n ؍
16; cortisol (100 nM), 1.7-fold ؎ 0.1, p < 0.05, and n ؍ 14). Similar increases in 11␤-HSD1 mRNA expression were
indicated using real-time quantitative reverse-transcription polymerase chain reaction (RT-PCR) analyses (3.5-fold
with DEX, p < 0.01; 2.5-fold with cortisol, p < 0.05). The capacity of 11␤-HSD1 to metabolize the synthetic
glucocorticoids prednisone and prednisolone was investigated in human OBs (hOBs) and fetal kidney-293 cells
stably transfected with human 11␤-HSD1 cDNA. Transfected cells and hOBs were able to interconvert prednisone
and prednisolone with reaction kinetics indistinguishable from those for cortisone and cortisol. To assess the in vivo
availability of substrates for osteoblastic 11␤-HSD1, plasma cortisone and prednisone levels were measured in
normal males before and after oral prednisolone (5 mg). The 9:00 a.m. serum cortisone levels were 110 ؎ 5
nmol/liter and prednisone levels peaked at 78 ؎ 23 nmol/liter 120 minutes after administration of prednisolone.
Thus, therapeutic use of steroids increases substrate availability for 11␤-HSD1 in bone. These studies indicate that
activation of glucocorticoids at an autocrine level within bone is likely to play an important role in the age-related
decrease in bone formation and increased risk of glucocorticoid-induced osteoporosis. (J Bone Miner Res 2002;17:
979–986)
Key words: 11␤-hydroxysteroid dehydrogenase, bone, osteoblasts, glucocorticoids, cortisol, osteoporosis, aging
considerable individual variation in susceptibility to the
effects of glucocorticoids, making it difficult to predict
individual risk of developing glucocorticoid-induced osteo-
porosis. However, now it is clear from epidemiological
studies and placebo-controlled trials of treatments for
glucocorticoid-induced osteoporosis that age is a key deter-
minant of fracture risk.(5–8) Age also is an important factor
INTRODUCTION
LUCOCORTICOIDS HAVE a profound impact on bone,(1,2)
predominantly through modulation of bone formation
G
but also through effects on bone resorption.(3,4) There is
The authors have no conflict of interest.
1Division of Medical Sciences, University of Birmingham, Queen Elizabeth Hospital, Edgbaston, Birmingham, United Kingdom.
2Department of Clinical Biochemistry and Immunology, Birmingham Heartlands and Solihull NHS Trust,
Birmingham, United Kingdom.
979

980
COOPER ET AL.
in fracture risk generally in both men and women, with risk nisone and prednisolone in cells stably transfected with
increasing progressively with age.(9) Although estrogen de- human 11␤-HSD1 cDNA and in primary cultures of hOBs.
ficiency at menopause accounts for part of this risk, an Furthermore, because there is little information on the cir-
age-related decrease in bone formation appears to be partly culating levels of 11␤-HSD1 substrates in normal individ-
responsible.(10,11) Although the changes in bone histomor- uals, we have determined the generation and availability of
phometry seen with age are similar to those seen with cortisone and prednisone in vivo in healthy men before and
glucocorticoid excess,(4) circulating glucocorticoid levels do after a single oral dose of prednisolone.
not appear to increase with age(12) and are normal in the
majority of patients with osteoporosis.
MATERIALS AND METHODS
Cell culture
At a tissue level, glucocorticoid hormone action is in part
determined at a prereceptor level through expression of the
enzyme 11␤-hydroxysteroid dehydrogenase (11␤-HSD),
which interconverts hormonally active cortisol with inactive
cortisone.(13) In recent studies we have characterized the ex-
pression and functional significance of the type 1 and type 2
isozymes of 11␤-HSD in normal human bone and in human
and rodent osteosarcoma cells.(14–17) In adult human bone,
11␤-HSD1 expression is evident in osteoblasts (OBs) where it
appears to function primarily as a reductase (converting corti-
sone to cortisol). In human primary OB cultures, 11␤-HSD1–
mediated oxoreductase activity (cortisone to cortisol conver-
sion) was the predominant activity present and this activity was
sensitively stimulated by proinflammatory cytokines.(17) Ex-
pression of human 11␤-HSD1 cDNA in stably transfected
ROS 17/2.8 osteosarcoma cells sensitized these cells to corti-
sone, stimulating glucocorticoid-sensitive markers of differen-
tiation and inhibiting cell proliferation.(18) The antiproliferative
effect of 11␤-HSD1 expression in OBs has parallels with the
reduction in proliferation rate seen in OBs with age. Therefore,
we have examined changes in 11␤-HSD1 in primary cultures
of human OBs (hOBs) with age. In several systems so far
studied, 11␤-HSD1 expression is increased by glucocorticoids
and thus the enzyme provides a “fast-forward” system for the
local production of glucocorticoids.(19–23) The existence of a
similar mechanism in OBs would have important implications
for bone physiology. In addition, therefore, we have examined
the effects of glucocorticoids on local corticosteroid metabo-
lism in primary cultures of hOBs.
Prereceptor metabolism of cortisone and cortisol is likely
to be of relevance in normal bone physiology but the pre-
dominant oral glucocorticoids used therapeutically are pred-
nisolone (in the United Kingdom) and its metabolite pred-
nisone (in the United States). In a recent study of oral
glucocorticoid usage in the United Kingdom, prednisolone
was used on 99% of occasions.(24) It is likely that in the
United States, data would be similar for prednisone. Both
cortisone and prednisone when given orally require hepatic
11␤-HSD1–mediated conversion to cortisol and pred-
nisolone, respectively, to confer bioactivity. This first-pass
metabolism is highly efficient with almost complete con-
version. However, substantial amounts of cortisone and
Primary cultures of hOBs were generated from normal
trabecular bone specimens obtained at joint replacement
surgery or lower limb amputation by an adaptation of an
established method.(25) The specific diagnoses were total
knee replacement for osteoarthritis (2 subjects), osteosar-
coma (5 subjects), sarcoma (4 subjects), chondrosarcoma (2
subjects), Ewing’s sarcoma (2 subjects), chronic prosthesis
infection (2 subjects), and spindle cell carcinoma (1 sub-
ject). None of the study subjects were systemically unwell,
taking systemic glucocorticoids or other drugs known to
affect bone metabolism. The study was conducted in accor-
dance with the Declaration of Helsinki and was approved by
the local research ethics committee and informed consent
was obtained. Unless stated, all reagents were from the
Sigma Chemical Co. (Poole, UK). Briefly, trabecular bone
was minced, washed extensively in PBS, and treated for 2 h
with 1 mg/ml of collagenase. Chips were then plated in
DMEM/F12 (Gibco, Paisley, UK) supplemented with 10%
FCS and 1ϫ insulin/transferrin/selenium mix. hOB cells
generated by this method had high basal and 1,25-
dihydroxyvitamin D3 [1,25(OH)2D3]–inducible alkaline
phosphatase activity and negligible amounts of adipocyte
phenotypic markers such as glycerol-3-phosphate dehydro-
genase activity and oil red O staining.(17) Cells were treated
with 10Ϫ6–10Ϫ10 M of cortisol, dexamethasone (DEX), and
10 nM of 1,25(OH)2D3 for 24–96 h. In total, cultures from
between 18 individuals were used (10 men and 8 women of
which 4 women were postmenopausal; mean age, 34.9 Ϯ
21.9 years [mean Ϯ SD]). All assays were carried out in
duplicate. A previously reported HEK293 cell line stably
transfected with a full-length human 11␤-HSD1 cDNA(26)
was used to analyze the interconversion of prednisone and
prednisolone. Cells were grown in DMEM supplemented
with 10% FCS and 0.5 mg/ml of G418 and maintained at
37°C in a humidified atmosphere of 5% CO2.
Measurement of enzyme activities
The 11␤-HSD activity assays were undertaken as previ-
prednisone subsequently are generated in the kidney by ously reported(17,26) using appropriate tritiated tracer steroid
dehydrogenation via a related enzyme 11␤-HSD2. Al- followed by separation of steroids from cell culture medium
though the interconversion of endogenous glucocorticoids using thin-layer chromatography (TLC). An in-house
(cortisone and cortisol in man and 11-dehydrocorticosterone method was used to generate [3H]cortisone from [3H]corti-
and corticosterone in rodents) by 11␤-HSD1 has been stud- sol (specific activity, 78.4 Ci/mmol; NEN Life Science
ied in detail,(13) the metabolism of prednisone and pred- Products, Hounslow, UK).(26) 3H[(3,4,7,8]prednisolone was
nisolone by 11␤-HSD1 has not been studied in detail in obtained from Amersham (specific activity, 40 Ci/mmol;
vitro. Therefore, we have assessed the metabolism of pred- Buckinghamshire, UK). [3H]prednisone was generated from

AGE, GLUCOCORTICOIDS, AND OSTEOBLASTIC 11␤-HSD1
981
[3H]prednisolone using an adaptation of the foregoing CAACCACATCACCAACA. Amplification parameters
method.(26) Briefly, 20 ␮l of [3H]prednisolone was con- were 50°C for 2 minutes and 95°C for 10 minutes, and 44
verted to [3H]prednisone using 250 ␮g of protein from cycles of 95°C for 15 s and 60°C for 1 minute. All reactions
homogenized human term placenta in 500 ␮l of total vol-
ume of 0.1 M of phosphate buffer, pH 7.4, with 500 ␮M of
nicotinamide adenine dinucleotide (NAD)ϩ as a cofactor.
After incubation for 6 h at 37°, steroids were extracted into
10 vol of dichloromethane and steroids were separated by
TLC with chloroform/ethanol (92:8) as a mobile phase.
[3H]prednisone was identified using a Bioscan 3000 Imager
(LabLogic, Sheffield, UK) and its identity was confirmed by
UV fluorescence with control lanes containing unlabeled
prednisone and prednisolone. The area of the TLC plate
containing [3H]prednisone was excised and [3H]prednisone
was extracted from the silica using ethanol. TLC analysis
indicated Ͼ99% conversion of [3H]prednisolone to
[3H]prednisone, which then was used as a tracer. A similar
method was used for all other steroid assays. Steroids were
extracted from medium into 10 vol of dichloromethane,
were multiplexed with 18S rRNA control probe (Applied
Biosystems). Data were obtained as Ct values as per man-
ufacturer’s guidelines (cycle number at which logarithmic
PCR plots cross a calculated threshold line) and used to
determine ⌬Ct values (⌬Ct ϭ Ct of target gene minus Ct of
18S). Data were analyzed at the ⌬Ct stage and mean Ϯ SE
⌬Ct values transformed through the Eq. 2Ϫ⌬⌬Ct to give
relative mRNA levels. Results reflect mean data from 12
separate cultures analyzed on three separate occasions in
duplicate.
Measurement of prednisolone to prednisone conversion
in vivo
To investigate whether cortisone and prednisone were
separated by TLC, and fractional interconversion of corti- potential substrates for osteoblastic 11␤-HSD1 in vivo, cir-
sone to cortisol (and prednisone to prednisolone)was as- culating levels of cortisone and prednisone were measured
sessed as outlined previously. Protein in each incubate was after a single 5-mg oral dose of prednisolone taken at 9:00
analyzed using the Bradford method (Bio-Rad Laboratories, a.m. in 12 normal healthy men (age, 32 Ϯ 8 years; mean Ϯ
Munich, Germany),(27) and enzyme activities were ex-
SD). The study had Local Research Ethics Committee ap-
pressed as picomoles of product formed per milligram of
protein per hour (pmol/mg per h).
proval and informed consent was obtained. Corticosteroid
levels were measured by HPLC as previously described.(28)
One hundred microliters of fludrocortisone (40 ␮mol/liter)
was added to 1 ml of serum and mixed with dichlorometh-
ane (10 ml) and then the aqueous phase was discarded. One
milliliter of 0.1 M of sodium hydroxide was added and after
mixing, the aqueous phase was discarded, and this step was
repeated. A similar wash step was repeated twice, replacing
sodium hydroxide with filtered degassed water. After the
wash steps, the dichloromethane layer was evaporated to
dryness and resuspended in 200 ␮l of HPLC mobile phase.
One hundred microliters of aqueous steroid standard or
reconstituted extracted steroid was injected onto a 150
mm ϫ 4.6 mm column packed with 5 ␮m of octadecyl
siloxane (ODS)-2 (LUNA; Phenomenex, Macclesfield,
UK). The mobile phase of tetrahydrofuran/water/methanol
in the ratio 23/65/2 and containing 100 mmol/liter of am-
monium acetate was pumped at 1.0 ml/minute. Steroids
were detected at absorption at 254 nm. Using this assay,
CVs for within-day estimates of precision were 4.4% or less
for the four steroids measured. CVs over time ranged from
6.5 to 8.4% for the steroids measured.
Preparation of RNA from OBs
Total RNA was extracted from hOB cell cultures using an
RNAzol extraction method according to the manufacturers
recommendations and resuspended in diethyl pyrocarbonate
(DEPC)–treated water. RT was performed as described pre-
viously.(14) Briefly, 1 ␮g of total RNA and 0.5 ␮g of random
primers in a final volume of 10 ␮l were incubated at 70°C
for 10 minutes. Primer extension then was performed at
37°C for 60 minutes after addition of reaction buffer con-
taining 50 mM of Tris-HCl (pH 8.3), 50 mM of KCl, 10 mM
of MgCl2, 10 mM of dithiothreitol (DTT), 0.5 mM of
spermidine, 1 mM of each deoxynucleoside triphosphate
(dNTP), 32 U of RNasin ribonuclease inhibitor, and 20 U of
avian megaloblastoma virus reverse transcriptase in a final
volume of 25 ␮l. A 2-␮l aliquot of this reaction was used in
subsequent PCR reactions.
Analysis of 11␤-HSD1 mRNA expression
Quantitative PCR analysis of 11␤-HSD1 mRNA expres-
sion was carried out using an ABI7700 Sequence Detection
system (Applied Biosystems Warrington, UK). Briefly, re-
actions were performed in 25-␮l volumes on 960-well
plates, in reaction buffer containing TaqMan Universal PCR
Master Mix, 3 mM of Mn(Oac)2, 200 ␮M of dNTP, 1.25 U
of AmpliTaq Gold polymerase, 1.25 U of AmpErase UNG,
100 nmol of TaqMan probe, 900 nmol of primers, and 100
ng of cDNA. 11␤-HSD1 primer and probe sequences are as
Statistical analysis
Results from in vitro and in vivo experiments are ex-
pressed as mean Ϯ SE. The effect of age on 11␤-HSD1
activity was assessed using the Pearson product moment
correlation. The effects of glucocorticoids on 11␤-HSD1
activity were assessed using Kruskal-Wallis ANOVA on
ranks using Dunn’s test for comparison with control cells.
follows: forward primer, AGGAAAGCTCATGGGAG- Comparisons at the RNA level were performed using paired
GACTAG; reverse primer, ATGGTGAATATCATCAT- t-tests. All statistical tests were carried out using SigmaStat
GAAAAAGATTC; TaqMan probe, CATGCTCATTCT- 2.03 software (SPSS Science, Chicago, IL, USA).

982
COOPER ET AL.
FIG. 2. (A) Dose-response curves for 11␤-HSD1 activity induction
with cortisol and DEX. Values are mean Ϯ SE, n ϭ 5, and mean age
of 23 years. Values of p refer to comparison to control untreated level
by ANOVA. (B) Comparison of 11␤-HSD1 mRNA expression in
primary OB cultures treated with glucocorticoids assessed by real-time
quantitative RT-PCR. Expression has been adjusted relative to control
untreated cells. Values reflect mean Ϯ SE (n ϭ 12) values at the ⌬Ct
stage (see text).
FIG. 1. Relationship between 11␤-HSD1 reductase activity and age
in primary cultures generated from (A) all trabecular bone specimens
and (B) samples taken exclusively from the calcaneus.
RESULTS
Expression of 11␤-HSD1 in OBs and relationship
to age
All primary cultures of hOBs showed 11␤-HSD1 reduc-
tase activity as measured by conversion of cortisone to
cortisol (3.4 Ϯ 0.4 pmol/mg per h at 250 nM cortisone).
This activity correlated with the age of the subject from
which the bone tissue was derived (Fig. 1A). This relation-
ship was even stronger for samples derived from the calca-
neus (Fig. 1B). Although there was a trend toward increased
11␤-HSD1 mRNA expression with age (r ϭ 0.35; n ϭ 10),
this relationship was not statistically significant. There was
no difference in 11␤-HSD1 activity between OBs derived
from men and women (p ϭ 0.84) and treatment of cells with
0.01–100 nM of 17␤-estradiol for 48 h had no effect on
11␤-HSD1 activity (data not shown). Dehydrogenase activ-
ity also was evident in primary cultures of OBs (cortisol to
FIG. 3. Time course of 11␤-HSD1 activity induction with 1 ␮M of
cortisone conversion at 500 nM of cortisol, 5.4 Ϯ 1.4 DEX. Values are mean Ϯ SE and n ϭ 5. The p values refer to
comparison to control untreated level by ANOVA.
pmol/mg per h) but this correlated poorly with reductase
activity (r ϭ Ϫ0.04). This dehydrogenase activity appeared
to be caused by 11␤-HSD1 rather than 11␤-HSD2 because,
as previously reported,(15) no 11␤-HSD2 mRNA was de-
tectable in these cultures.
were paralleled by changes in mRNA levels assessed by
“real-time” quantitative PCR. 11␤-HSD1 mRNA levels
were 3.5-fold those of untreated levels (Ϫ1 SE 2.5-fold; ϩ1
SE 5.0-fold) with 1 ␮M of DEX (p Ͻ 0.01) and 2.5-fold
those of control (Ϫ1 SE 1.6-fold; ϩ1 SE 3.9-fold) with 100
nM of cortisol (p Ͻ 0.05; Fig. 2B). The time course of
induction of 11␤-HSD1 reductase activity showed increased
activity at 24 h and maximal activity at 48–72 h (Fig. 3).
Effect of glucocorticoids on 11␤-HSD1 expression
in OBs
Treatment with either 100 nM of cortisol or 1 ␮M of
DEX for 48 h significantly induced 11␤-HSD1 activity
(1.7-fold Ϯ 0.1 with cortisol, p Ͻ 0.01, and n ϭ 13;
2.6-fold Ϯ 0.5 with DEX, p Ͻ 0.001, and n ϭ 15). In
contrast, treatment with 10 nM of 1,25(OH)2D3 had no
effect on 11␤-HSD1 activity (1.2-fold Ϯ 0.2, NS, and n ϭ
12). The induction after incubation with both DEX and
Interconversion of prednisone and prednisolone
The interconversion of prednisone and prednisolone ini-
cortisol was dose dependent (Fig. 2A), with half-maximal tially was assessed in stable transfectant variants of
induction evident at between 1 and 10 nM for DEX and 10 HEK293 cells expressing human 11␤-HSD1. Using these
and 100 nM for cortisol. There was no effect of age on cells, both 11␤-HSD1 reductase and dehydrogenase activi-
either the absolute or the relative increase in 11␤-HSD1 ties could be established with both cortisone/cortisol and
activity with glucocorticoids. Changes in reductase activity prednisone/prednisolone. The kinetics of these reactions for

AGE, GLUCOCORTICOIDS, AND OSTEOBLASTIC 11␤-HSD1
983
FIG. 5. (A) Prednisolone and cortisol and (B) prednisone and corti-
sone levels in 12 normal men after ingestion of 5 mg of prednisolone
orally at 9:00 a.m. All measurements were made using HPLC as
described in the text and values are mean Ϯ SE.
FIG. 4. Kinetics of 11␤-HSD1 in HEK293 cells stably transfected
with human 11␤-HSD1 cDNA. (A) 11␤-HSD1 reductase activity for
cortisone and prednisone. (B) 11␤-HSD1 dehydrogenase activity for
cortisol and prednisolone. Sone-sol indicates conversion of prednisone
to prednisolone; sol-sone indicates conversion of prednisolone to pred-
nisone.
concentration decreased steadily with time post-prednisolone
whereas cortisone concentrations decreased by a small amount
initially but then remained steady such that 300 minutes post-
prednisolone cortisol levels were only slightly higher than
cortisone levels.
each pair of substrates were indistinguishable (Fig. 4). In
primary OBs, kinetics for cortisone/cortisol and prednisone/
prednisolone conversion also were similar (Km for cortisone
to cortisol, 640 Ϯ 60 nM; Km for prednisone to pred-
nisolone, 730 Ϯ 110 nM) and did not differ from enzyme
kinetics in transfected cells.
DISCUSSION
In this study, we have established a correlation between
11␤-HSD1 activity in primary cultures of hOBs and age.
Additionally, we have indicated that glucocorticoids in-
crease the activity and expression of 11␤-HSD1 in OBs and
that relatively high levels of substrate for this enzyme are
Measurement of 11␤-HSD1 substrate levels in vivo
To assess the in vivo availability of substrates and products available both under normal conditions and after the admin-
of 11␤-HSD1, serum glucocorticoids were determined in 12 istration of low doses of prednisolone. In view of the effect
normal men before and after a single 5-mg oral dose of of 11␤-HSD1 expression on OB proliferation and differen-
prednisolone. The 9:00 a.m. circulating cortisol and cortisone tiation,(18) these data suggest that increasing 11␤-HSD1
concentrations were 431 Ϯ 30 nmol/liter and 110 Ϯ 5 nmol/ activity could be an important component of the increase in
liter, respectively. After ingestion of 5 mg of prednisolone, fracture risk for both idiopathic and glucocorticoid-induced
serum prednisolone peaked at 60 minutes postingestion (Fig. osteoporosis that occurs with aging. Furthermore, the reg-
5A) and prednisone levels peaked at 120 minutes (Fig. 5B). ulation of local glucocorticoid metabolism by glucocorti-
Greater variation between individuals was apparent for pred- coids themselves suggests that a “feed-forward” effect may
nisone levels compared with those for prednisolone. Cortisol magnify the exposure of OBs to glucocorticoids during

984
COOPER ET AL.
treatment with these drugs or in states of endogenous glu- ple, in recent studies using primary cultures of human
cocorticoid excess. preadipocytes, we have shown increased cortisone to corti-
Fracture risk increases substantially with age in both men sol metabolism after treatment with cortisol.(19) This sug-
and women. A component of this increased risk may be gests that a fast-forward intracrine mechanism plays a piv-
attributable to sex-hormone deficiency at menopause but a otal role in adipocyte differentiation; unlike control cells,
large component of this risk appears to be independent of preadipocytes pretreated with cortisol were sensitive to the
this effect. A characteristic feature of aging in the skeleton supposedly inactive cortisone.(23) These effects may explain
is a reduction in bone formation, an effect attributed to the central obesity seen in states of glucocorticoid excess;
decreased OB function.(11,29) Thus, aging shares many of differential expression of 11␤-HSD1 in various fat depots
the features of excess glucocorticoids but circulating levels may provide a mechanism for differences in regional fat
of glucocorticoids do not change significantly with ag- distribution. This hypothesis has been supported by the
ing.(30) Our data suggest that some of the features of aging development of a viscerally obese phenotype in mice with
on bone may in fact be caused by increased intracellular targeted overexpression of 11␤-HSD1 in adipose tissue.(31)
glucocorticoid levels. This raises the possibility that inhib- Furthermore, these mice had increased levels of corticoste-
itors of 11␤-HSD1 could ameliorate some of the effects of rone in fat tissue, proving that 11␤-HSD1 expression in vivo
aging on the skeleton. In previous studies we have shown directly impacts on local glucocorticoid levels. Reductase
that inhibition of 11␤-HSD1 in vivo significantly alters and dehydrogenase 11␤-HSD1 activities have been reported
biochemical bone markers in young subjects(14) but this has in primary cultures of skin fibroblasts, and both increased
not been examined in older individuals. Estrogen had no threefold after treatment with 100 nM of DEX with a similar
effect on 11␤-HSD1 activity in cultured cells, suggesting time course to that shown in this study.(20) Regulation of
that there is unlikely to be a direct effect of postmenopausal 11␤-HSD1 by glucocorticoids also has been reported in
estrogen deficiency on 11␤-HSD1 activity. 11␤-HSD1 ac- some tissues unrelated to the OB lineage, including primary
tivity in primary cultures of OBs from premenopausal cultures of rat hepatocytes(21,22) and rat hepatoma-derived
women was 2.1 Ϯ 1.0 pmol/mg per h and 4.4 Ϯ 0.9 cell lines.(32) However, in vivo, DEX modulation of hepatic
pmol/mg per h in postmenopausal women (difference NS). 11␤-HSD1 expression in the rat appears to be more com-
It remains possible that conditions predisposing to surgery plex with the suggestion of decreased activity.(33)
could have introduced bias, for example, osteoarthritis as an
The metabolism of prednisone and prednisolone by 11␤-
indication was more likely in older subjects. However, none HSD1 previously has not been explored in detail in vitro
of the subjects had conditions known to affect bone metab- although reductase activity for prednisone in the liver is
olism and the same general pattern of increasing activity evident from the rapid generation of prednisolone from oral
with age was seen in conditions that were represented across prednisone, an effect comparable with the conversion of
the age range.
cortisone to cortisol.(34) One in vitro study has examined
The adverse effects of elevated therapeutic and endoge- metabolism of these steroids in yeast cells overexpressing
nous glucocorticoids on bone are well established. Recently, human 11␤-HSD1 cDNA and microsomal preparations of
the relationship between fracture risk and glucocorticoid human liver.(35) It was found that the Km for prednisone was
dose during treatment has been studied in a large cohort of an order of magnitude higher than that for cortisone. How-
patients.(5) This study showed that fracture risk is signifi- ever, kinetic constants in these studies were much higher
cantly elevated with even low doses of glucocorticoid and than those reported in human liver or other intact cells
that this risk increases within a short time of commencing expressing 11␤-HSD1 and, in contrast to intact systems,
oral glucocorticoids. Data from the placebo-controlled trials dehydrogenase activity was prominent, suggesting that nor-
of treatments for glucocorticoid-induced osteoporosis have mal enzymatic activity had been lost in these studies. Our
emphasized that fracture risk is very low in younger men data, in intact cells, suggest that the metabolism of
and women, suggesting that age plays an important role in cortisone/cortisol and prednisone/prednisolone by 11␤-
susceptibility. However, for an individual, the relationship HSD1 is very similar. In transfected kidney cells and pri-
between glucocorticoid dose and effects on bone have been mary OBs, kinetics are virtually identical for these steroids.
more difficult to estimate and there appear to be large This suggests that for equivalent levels of these steroids
differences in susceptibility between individuals. Our data their metabolism in vivo will be similar.
suggest that one reason for these differences is that in the
Because prednisone is almost completely converted to
face of similar extracellular levels of glucocorticoids, intra- prednisolone on first-pass hepatic metabolism, we examined
cellular levels of steroid may vary between individuals as a the generation of prednisone after an oral dose of pred-
consequence of differences in 11␤-HSD1 expression. Fur- nisolone. Furthermore, in contrast to the USA, prednisolone
thermore, the intracellular metabolism of glucocorticoids is the most commonly prescribed therapeutic corticosteroid
appears to be a dynamic process, varying with age and in the UK. Circulating levels of prednisone rose during
prevailing glucocorticoid level and thus further complicat- treatment with low doses of prednisolone. Although this
ing the relationship between glucocorticoid dose and glu- highlights the potential for production of a substrate for
cocorticoid action on OBs.
osteoblastic 11␤-HSD1, it is important to point out that
Glucocorticoid-induced increases in 11␤-HSD1 enzyme these levels do not accurately reflect glucocorticoid bio-
expression have been reported previously in other cell types, availability because there is extensive binding to serum
including cells developmentally related to OBs. For exam- proteins, especially cortisol binding globulin (CBG). This

AGE, GLUCOCORTICOIDS, AND OSTEOBLASTIC 11␤-HSD1
985
paracrine mechanisms of glucocorticoid-induced osteoporosis.
Endocrinology 140:4382–4389.
binding is more extensive for cortisol (95%) and pred-
nisolone (50%) than for cortisone and prednisone (both
Ͻ5%).(36–38) Therefore, the actual bioavailability of corti-
sone (and prednisone) under these conditions will be closer
to that of cortisol (and prednisolone) than predicted from the
total level alone.(39)
4. Carbonare LD, Arlot ME, Chavassieux PM, Roux JP, Portero
NR, Meunier PJ 2001 Comparison of trabecular bone micro-
architecture and remodeling in glucocorticoid-induced and
postmenopausal osteoporosis. J Bone Miner Res 16:97–103.
5. van Staa TP, Leufkens HG, Abenhaim L, Zhang B, Cooper C
2000 Use of oral corticosteroids and risk of fractures. J Bone
Miner Res 15:993–1000.
6. Adachi JD, Bensen WG, Brown J, Hanley D, Hodsman A,
Josse R, Kendler DL, Lentle B, Olszynski W, Ste-Marie LG,
Tenenhouse A, Chines AA 1997 Intermittent etidronate ther-
apy to prevent corticosteroid-induced osteoporosis. N Engl
J Med 337:382–387.
7. Saag KG, Emkey R, Schnitzer TJ, Brown JP, Hawkins F,
Goemaere S, Thamsborg G, Liberman UA, Delmas PD, Mal-
ice MP, Czachur M, Daifotis AG 1998 Alendronate for the
prevention and treatment of glucocorticoid-induced osteopo-
rosis. Glucocorticoid-Induced Osteoporosis Intervention Study
Group. N Engl J Med 339:292–299.
8. Wallach S, Cohen S, Reid DM, Hughes RA, Hosking DJ, Laan
RF, Doherty SM, Maricic M, Rosen C, Brown J, Barton I,
Chines AA 2000 Effects of risedronate treatment on bone
density and vertebral fracture in patients on corticosteroid
therapy. Calcif Tissue Int 67:277–285.
9. Cooper C, Melton JL III 1996 Magnitude and impact of
osteoporosis and fractures. In: Marcus R, Feldman D, Kelsey
J (eds.) Osteoporosis. Academic Press, San Diego, CA, USA,
pp. 421–425.
10. Brockstedt H, Kassem M, Eriksen EF, Mosekilde L, Melsen F
1993 Age- and sex-related changes in iliac cortical bone mass
and remodeling. Bone 14:681–691.
11. Clarke BL, Ebeling PR, Jones JD, Wahner HW, O’Fallon
WM, Riggs BL, Fitzpatrick LA 1996 Changes in quantitative
bone histomorphometry in aging healthy men. J Clin Endocri-
nol Metab 81:2264–2270.
12. Ferrari E, Cravello L, Muzzoni B, Casarotti D, Paltro M,
Solerte SB, Fioravanti M, Cuzzoni G, Pontiggia B, Magri F
2001 Age-related changes of the hypothalamic-pituitary-
adrenal axis: Pathophysiological correlates. Eur J Endocrinol
144:319–329.
Therefore, our data suggest that the impact of glucocor-
ticoids on an OB will depend on both the endocrine level of
active glucocorticoid and autocrine generation of glucocor-
ticoid. Autocrine production of active glucocorticoid is
likely to be higher for cortisone and prednisone than for
DEX or deflazacort (which are both poor substrates for
11␤-HSD1). Few studies have explored the differences be-
tween these glucocorticoids with respect to their relative
capacity to cause deleterious effects on bone. A notable
exception is deflazacort, a derivative of prednisolone. Initial
studies with deflazacort suggested a steroid-sparing effect
on bone(40–42) although the robustness of this data has been
questioned.(43) Deflazacort does not appear to be a substrate
for 11␤-HSD1(35); thus, use of this agent would not be
complicated by local generation of active glucocorticoid by
OBs. The existence of steroid-modifying enzymes in bone
should lead to a better evaluation of differential effects of
glucocorticoids on bone and potentially could be used in a
more rational design of synthetic agents that have reduced
effects on bone.
These data indicate that biologically inactive forms of
endogenous and commonly used synthetic glucocorticoids
are substrates for osteoblastic 11␤-HSD1; activity of this
enzyme increases with age and after treatment with thera-
peutic glucocorticoids. Glucocorticoid generation within
OBs via 11␤-HSD1 expression has important consequences
on OB function and may be an important determinant of the
effects of age and glucocorticoids on bone.
13. Stewart PM, Krozowski ZS 1999 11␤-Hydroxysteroid dehy-
drogenase. Vitam Horm 57:249–324.
ACKNOWLEDGMENTS
14. Cooper MS, Walker EA, Bland R, Fraser WD, Hewison M,
Stewart PM 2000 Expression and functional consequences of
11␤-hydroxysteroid dehydrogenase activity in human bone.
Bone 27:375–381.
15. Bland R, Worker CA, Noble BS, Eyre LJ, Bujalska IJ, Shep-
pard MC, Stewart PM, Hewison M 1999 Characterization of
11␤-hydroxysteroid dehydrogenase activity and corticosteroid
receptor expression in human osteosarcoma cell lines. J En-
docrinol 161:455–464.
16. Eyre LJ, Rabbitt EH, Bland R, Hughes SV, Cooper MS,
Sheppard MC, Stewart PM, Hewison M 2001 Expression of
11␤-hydroxysteroid dehydrogenase in rat osteoblastic cells:
Pre-receptor regulation of glucocorticoid responses in bone.
J Cell Biochem 81:453–462.
We thank the staff of the Department of Musculoskeletal
Pathology, Royal Orthopedic Hospital, Birmingham, for
help with the preparation of clinical materials and the staff
of the Wellcome Trust Clinical Research Facility for help
with the clinical studies. This study is funded by the Med-
ical Research Council, United Kingdom. Mark S. Cooper is
a Medical Research Council Clinical Training Fellow and
Paul M. Stewart is an MRC Senior Clinical Fellow.
REFERENCES
17. Cooper MS, Bujalska I, Rabbitt E, Walker EA, Bland R,
Sheppard MC, Hewison M, Stewart PM 2001 Modulation of
11␤-hydroxysteroid dehydrogenase isozymes by proinflam-
matory cytokines in osteoblasts: An autocrine switch from
glucocorticoid inactivation to activation. J Bone Miner Res
16:1037–1044.
18. Rabbitt E, Lavery GG, Walker EA, Cooper MS, Stewart PM,
Hewison M 2002 Pre-receptor regulation of glucocorticoid
action by 11␤-hydroxysteroid dehydrogenase: A novel deter-
minant of cell proliferation. FASEB J 16:36–44.
1. Canalis E 1996 Clinical review 83: Mechanisms of glucocor-
ticoid action in bone: Implications to glucocorticoid-induced
osteoporosis. J Clin Endocrinol Metab 81:3441–3447.
2. Manolagas SC 2000 Birth and death of bone cells: Basic
regulatory mechanisms and implications for the pathogenesis
and treatment of osteoporosis. Endocr Rev 21:115–137.
3. 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: Potential

986
COOPER ET AL.
19. Bujalska IJ, Kumar S, Stewart PM 1997 Does central obesity 33. Jamieson PM, Chapman KE, Seckl JR 1999 Tissue- and
temporal-specific regulation of 11␤-hydroxysteroid dehydro-
genase type 1 by glucocorticoids in vivo. J Steroid Biochem
Mol Biol 68:245–250.
reflect “Cushing’s disease of the omentum”? Lancet 349:
1210–1213.
20. Hammami MM, Siiteri PK 1991 Regulation of 11
␤-hydroxysteroid dehydrogenase activity in human skin fibro-
blasts: Enzymatic modulation of glucocorticoid action. J Clin
Endocrinol Metab 73:326–334.
21. Jamieson PM, Chapman KE, Edwards CR, Seckl JR 1995
11␤-hydroxysteroid dehydrogenase is an exclusive 11␤-
reductase in primary cultures of rat hepatocytes: Effect of
physicochemical and hormonal manipulations. Endocrinology
136:4754–4761.
22. Liu YJ, Nakagawa Y, Nasuda K, Saegusa H, Igarashi Y 1996
Effect of growth hormone, insulin and dexamethasone on
11␤-hydroxysteroid dehydrogenase activity on a primary cul-
ture of rat hepatocytes. Life Sci 59:227–234.
23. Bujalska IJ, Kumar S, Hewison M, Stewart PM 1999 Differ-
entiation of adipose stromal cells: The roles of glucocorticoids
and 11␤-hydroxysteroid dehydrogenase. Endocrinology 140:
3188–3196.
24. van Staa TP, Leufkens HG, Abenhaim L, Begaud B, Zhang B,
Cooper C 2000 Use of oral corticosteroids in the United
Kingdom. QJM 93:105–111.
25. Robey PG, Termine JD 1985 Human bone cells in vitro. Calcif
Tissue Int 37:453–460.
26. Bujalska I, Shimojo M, Howie A, Stewart PM 1997 Human
11␤-hydroxysteroid dehydrogenase: Studies on the stably
transfected isoforms and localization of the type 2 isozyme
within renal tissue. Steroids 62:77–82.
27. Bradford MM 1976 A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal Biochem 72:248–254.
28. Bayly GR, Bartlett WA, Jones AF 1997 A non-isotopic
method for estimating 11␤-hydroxysteroid dehydrogenase ac-
tivity in vivo. Ann Clin Biochem 34:521–526.
29. Martinez ME, Del Campo MT, Medina S, Sanchez M,
Sanchez-Cabezudo MJ, Esbrit P, Martinez P, Moreno I, Ro-
drigo A, Garces MV, Munuera L 1999 Influence of skeletal
site of origin and donor age on osteoblastic cell growth and
differentiation. Calcif Tissue Int 64:280–286.
30. Van Cauter E, Leproult R, Kupfer DJ 1996 Effects of gender
and age on the levels and circadian rhythmicity of plasma
cortisol. J Clin Endocrinol Metab 81:2468–2473.
31. Masuzaki H, Paterson J, Shinyama H, Morton NM, Mullins JJ,
Seckl JR, Flier JS 2001 A transgenic model of visceral obesity
and the metabolic syndrome. Science 294:2166–2170.
32. Napolitano A, Voice MW, Edwards CR, Seckl JR, Chapman
KE 1998 11␤-hydroxysteroid dehydrogenase 1 in adipocytes:
Expression is differentiation-dependent and hormonally regu-
lated. J Steroid Biochem Mol Biol 64:251–260.
34. Jenkins JS, Sampson PA 1967 Conversion of cortisone to
cortisol and prednisone to prednisolone. BMJ 2:205–207.
35. Hult M, Jornvall H, Oppermann UC 1998 Selective inhibition
of human type 1 11␤-hydroxysteroid dehydrogenase by syn-
thetic steroids and xenobiotics. FEBS Lett 441:25–28.
36. Walker BR, Campbell JC, Fraser R, Stewart PM, Edwards CR
1992 Mineralocorticoid excess and inhibition of 11␤-
hydroxysteroid dehydrogenase in patients with ectopic ACTH
syndrome. Clin Endocrinol (Oxf) 37:483–492.
37. Gambertoglio JG, Amend WJ Jr, Benet LZ 1980 Pharmaco-
kinetics and bioavailability of prednisone and prednisolone in
healthy volunteers and patients: A review. J Pharmacokinet
Biopharm 8:1–52.
38. Ballard PL 1979 Delivery and transport of glucocorticoids to
target cells. In: Baxter JD, Rousseau GG (eds.) Glucocorticoid
Hormone Action. Springer-Verlag, Berlin, pp. 25–48.
39. Andrews RC, Walker BR 1999 Glucocorticoids and insulin
resistance: Old hormones, new targets. Clin Sci (Colch) 96:
513–523.
40. Gennari C, Imbimbo B, Montagnani M, Bernini M, Nardi P,
Avioli LV 1984 Effects of prednisone and deflazacort on
mineral metabolism and parathyroid hormone activity in hu-
mans. Calcif Tissue Int 36:245–252.
41. Lo CV, Bonucci E, Imbimbo B, Ballanti P, Tartarotti D,
Galvanini G, Fuccella L, Adami S 1984 Bone loss after glu-
cocorticoid therapy. Calcif Tissue Int 36:435–438.
42. Lippuner K, Casez JP, Horber FF, Jaeger P 1998 Effects of
deflazacort versus prednisone on bone mass, body composi-
tion, and lipid profile: A randomized, double blind study in
kidney transplant patients. J Clin Endocrinol Metab 83:3795–
3802.
43. Anonymous 1999 Deflazacort—an alternative to pred-
nisolone? Drug Ther Bull 37:57–58.
Address reprint requests to:
Paul M. Stewart, M.D., F.R.C.P., F.Med.Sci.
Endocrinology
Division of Medical Sciences
University of Birmingham
Queen Elizabeth Hospital
Edgbaston
Birmingham B15 2TH, UK
Received in original form November 28, 2001; in revised form
February 8, 2002; accepted March 13, 2002.