Human osteoblasts produce cathepsin K
Jami Mandelina,b,c,⁎, Mika Hukkanena,c, Tian-Fang Lia, Matti Korhonena, Mikko Liljeströma,b,c,
Tarvo Sillata, Roeland Hanemaaijerd, Jari Saloe, Seppo Santavirtae, Yrjö T. Konttinenb,c,f
aInstitute of Biomedicine/Anatomy, Biomedicum Helsinki, PO Box 63, FI-00014 University of Helsinki, Finland
bDepartment of Medicine/Invärtes Medicin, Helsinki University Central Hospital, Helsinki, Finland
cORTON Orthopaedic Hospital of the Invalid Foundation, Helsinki, Finland
dGaubius Laboratory TNO-PG, Leiden, The Netherlands
eDepartment of Orthopaedics and Traumatology, Helsinki University Central Hospital, Helsinki, Finland
fCOXA, The Joint Replacement Hospital, Tampere University Central Hospital, Tampere, Finland
Received 3 May 2005; revised 17 October 2005; accepted 26 October 2005
Available online 5 December 2005
Healthy bone is a rigid yet living tissue that undergoes continuous remodeling. Osteoclasts resorb bone in the remodeling cycle. They secrete
H+-ions and proteinases to dissolve bone mineral and degrade organic bone matrix, respectively. One of the main collagenolytic proteinase in
osteoclasts is cathepsin K, a member of papain family cysteine proteinases. Recently, it has been shown that osteoblasts may contribute to organic
matrix remodeling. We therefore investigated their ability to produce cathepsin K for this action. Trabecular bone samples were collected from
patients operated due to a fracture of the femoral neck. Part of the bone was decalcified and the rest was used for cell isolation. Sections from the
decalcified bone were immunostained with antibodies against cathepsin K. Isolated cells were characterized for their ability to form mineralized
matrix and subsequently analyzed for their cathepsin K production by Western blotting and quantitative RT-PCR. Osteoblasts, bone lining cells
and some osteocytes in situ showed cathepsin K immunoreactivity and osteoblast-like cells in vitro produced cathepsin K mRNA and released
both 42 kDa pro- and 27 kDa processed cathepsin K to culture media. Osteoblastic cathepsin K may thus contribute to collagenous matrix
maintenance and recycling of improperly processed collagen I. Whether osteoblastic cathepsin K synthesis has consequences in diseases
characterized by abnormal bone matrix turnover remains to be investigated.
© 2005 Published by Elsevier Inc.
Keywords: Cytokines; Cathepsin K; Osteoblasts; Osteoporosis; Stromal cells
Bone remodeling occurs in discrete temporary anatomic
structures called basic multicellular units (BMUs) formed by
bone resorbing osteoclasts and bone-forming osteoblasts. An
imbalance between bone resorption and bone formation can
result in gross perturbations in skeletal structure and function.
Osteoporosis is a common, systemic skeletal disorder charac-
terized by reduction in bone mass and microarchitectural
deterioration of bone tissue leading to an increased susceptibil-
ity to fractures. Like most adult skeletal conditions, osteopo-
rosis is thought to occur due to the excess osteoclastic activity,
which is not fully compensatedby the coupledosteoblastic bone
formation [1–3]. Bone resorption is osteoclast-driven but
osteoblasts control their formation and activity via the receptor
activator of NF-κB ligand (RANKL) system . At resorbing
state, osteoclasts attach to bone surfaces and form different
plasma membrane domains such as ruffled border containing
vacuolar H+-ATPase that enables production of an acidic
subcellular space [5,6]. After demineralization, proteinases, in
particular cathepsin K , initiate the degradation of the organic
collagen type I-rich osteoid matrix in subosteoclastic spaces
It has been thought that the osteoclasts are the sole source of
cathepsin K. However, there is evidence that, e.g. fibroblast-
like synoviocytes from rheumatoid patients produce active
Bone 38 (2006) 769–777
⁎Corresponding author. Institute of Biomedicine/Anatomy, Biomedicum
Helsinki, PO Box 63, FI-00014 University of Helsinki, Finland. Fax: +358 9
E-mail address: email@example.com (J. Mandelin).
8756-3282/$ - see front matter © 2005 Published by Elsevier Inc.
cathepsin K . A recent study demonstrates that the bone
marrow-derived osteoblast-like cells function in vitro as
matrix-degrading cells and can finalize the resorption phase
before synthesizing new mineralized matrix . We hypoth-
esized that osteoblasts may synthesize cathepsin K for this
action. Therefore, we analyzed cathepsin K production in situ
in metaphyseal trabecular bone from patients with osteoporo-
sis, and in osteoblast-like cultures isolated either from
metaphyseal bone or from bone marrow aspirates.
Materials and methods
Immunohistochemical staining of bone samples
Femoral heads from patients (n = 6) undergoing a hemiarthroplasty due
to a fracture of the femoral neck were used to collect bone samples, which
were fixed in formalin, decalcified and embedded in paraffin. The samples
were cut to 4 μm sections, which were de-paraffinized and treated with 5%
pepsin (Merck, Darmstadt, Germany) in 0.2 M HCl solution. The samples
were washed in 10 mM phosphate buffered 150 mM saline pH 7.4 (PBS),
after which the endogenous peroxidase activity was blocked with 0.3%
H2O2in methanol for 25 min, followed by washing in PBS and blocking
with normal horse serum diluted 1:50 in PBS containing 5% bovine serum
albumin (BSA; Sigma, St. Louis, MO) for 40 min in room temperature.
After that, the samples were incubated in a humid atmosphere for overnight
in +4°C with 0.17 μg/ml rabbit anti-cathepsin K C-terminal peptide IgG
diluted in 5% BSA-PBS. Purified, irrelevant rabbit IgG (Jackson
Immunoresearch Laboratories, West Grove, PA) was used as a negative
control. After washing, the samples were incubated with biotinylated
secondary antibody for 1 h in room temperature. The antibody was removed
by washing and the samples were incubated with avidin–biotin–peroxidase-
complex solution. The color reaction was developed with 3,3-diaminoben-
zidine (Sigma) peroxidase substrate solution for 1.5 min. The samples were
washed in tap water and counterstained with hematoxylin before
dehydration, clearing and mounting.
Femoral heads from patients (n = 6) undergoing a hemiarthroplasty due to a
fracture of the femoral neck were used for osteoblast isolation. Metaphyseal
trabecular bone was cut to pieces and washed several times with PBS. Samples
were subjected to collagenase (Collagenase XI; Sigma) treatment for 4 times for
15 min each. After washes, the explants were cultured overnight in Dulbecco’s
Modified Eagle Medium (D-MEM) without D-glucose and sodium pyruvate
(Invitrogen, Carlsbad, CA) supplemented with 10% fetal calf serum (FCS) and
10×penicillinandstreptomycin. Onthe followingday, the mediumwas changed
week. Explants were removed after approximately 80% monolayer confluence
was reached. Passages 2–4 were used for subsequent experiments. At each time
point, the media were replaced 24 h prior to the sample collection.
Marrow stromal cell (MSC) isolation
Five milliliters of heparinized bone marrow aspirate was obtained after
informed consent from healthy individuals (n = 3), aged 2–18 years,
undergoing bone marrow harvest for sibling transplantation. The study was
approved by the ethical committee of the Hospital for Children and
Adolescents, Helsinki University Central Hospital, Finland. The aspirate was
suspended in RPMI, and digested with 100 U/ml DNAse I (Roche,
Mannheim, Germany) and 0.2 mg/ml collagenase B (Roche) for 45 min at
room temperature. Mononuclear cells were isolated by density gradient
centrifugation (Ficoll-Paque Plus; Pharmacia Biotech, Uppsala, Sweden),
washed and plated in D-MEM low glucose medium (Invitrogen)
supplemented with 10% FCS. The media was changed after 72 h to
remove non-adherent cells, and thereafter twice weekly. After 3 weeks,
colonies of spindle-shaped cells were formed. The cells were passaged twice
weekly. Cells from passages 4–5 were used for the differentiation
Osteogenic differentiation and TNF-? stimulation
Osteoblasts and MSCs were plated at 103cells/cm2in D-MEM low
glucose medium with 10% FCS and antibiotics, and supplemented with 0.1
μM dexamethasone, 50 μM ascorbic acid-2-phosphate and 10 mM β-
glycerophosphate (all from Sigma). After 3 weeks of culture, the
mineralization was revealed by silver nitrate staining according to the
method of von Kossa. For TNF-α stimulation, osteoblasts were first
cultured in the osteogenic differentiation media for 2 weeks and then
stimulated with recombinant TNF-α (5 ng/ml; R&D Systems, Minneapolis,
MN) for 24 h.
Quantitative reverse transcriptase polymerase chain reaction
Total RNA was isolated by TRIzol reagent (Invitrogen). RNA was
DNase treated (Promega, Madison, WI) and 2 μg was used to prepare
primary cDNA using (dT)12–18 primers and SuperScript enzyme, followed
by RNase H treatment (Invitrogen). Quantitative PCR was run using 0.2 μg
first strand cDNA, 0.5 mM primers and 0.2 mM TaqMan probes (Table 1)
in LightCycler™ PCR mix by LightCycler™ PCR machine (Roche).
Conventional PCR was run using 0.1 μg first strand DNA, 0.4 mM Runx2
primers (Table 1) and AmpliTaq Gold enzyme in AmpliTaq Gold buffer
(Applied Biosystems, Foster City, CA). Serial 1:10 dilutions of human
genomic DNA were used to determine the copy number of the amplicon in
relation to housekeeping gene porphobilinogen deaminase (PBGD) mRNA
copies. The copy numbers of mRNA molecules were determined at least
twice for all samples. The cDNA synthesis reaction was also performed
without RT enzyme followed by amplification of PBGD to exclude the
possibility of genomic DNA contamination.
After media change, the cells were incubated in fresh media for 24
h and the presence of cathepsin K was analyzed from 15 μg of medium
proteins. Samples were boiled for 5 min in SDS gel loading buffer before
electrophoresis. Human giant cell tumor of bone and liver lysates (75 μg)
were used as positive and negative sample controls. After electrophoresis
and blotting, the membrane (Bio-Rad, Richmond, CA) was blocked
overnight using 3% bovine serum albumin (BSA; Sigma) in Tris buffered
saline (TBS) (50 mM Tris–HCl, 150 mM NaCl, pH 7.5) and washed in a
washing buffer (0.1% Tween 20, 50 mM Tris–HCl, 0.5 M NaCl, pH 7.5).
After washes, the membranes were incubated with affinity purified goat
anti-human cathepsin K IgG (0.4 μg/ml; Santa Cruz Biotechnology, Santa
Cruz, CA) in washing buffer containing 2% BSA for 1 h. This was
followed with washes and incubation for 1 h at +22°C with alkaline
phosphatase conjugated rabbit anti-goat IgG (1:5000 in washing buffer
containing 2% BSA; Jackson Immunoresearch Laboratories). After washes,
color was developed in a solution (Alkaline Phosphatase Conjugate
Substrate Kit; Bio-Rad) containing a mixture of 5-bromo-4-chloro-3-indolyl
phosphate (BCIP) and nitroblue tetrazolium (NBT).
Modified urokinase assay to detect active cathepsin K
Cells were gradually changed to serum free D-MEM and after 24 h of
incubation in the serum free media, 10 μl of medium was analyzed. After
collecting the media, the cultures were washed with PBS and lysed in 50 μl
50 mM sodium acetate pH 5.0, 2.5 mM EDTA, 0.01% Tween-80 with three
freezing and thawing cycles and centrifuged. 10 μl of lysate was used to
analysis. The assay  was run in Costar Stripwell plates, which were
coated (2 h, 4°C) with 100 μl of cathepsin K-specific monoclonal antibody
(TNO B1446, 1 μg/ml). This antibody recognizes native cathepsin K, does
not cross-react with cathepsin L or cathepsin S; minimally cross-reacts with
770J. Mandelin et al. / Bone 38 (2006) 769–777
cathepsin V. Binding of cathepsin K was performed by incubation of
purified cathepsin K or biological sample for 16 h at 4°C in capture buffer
(50 mM sodium acetate, pH 5.0, 2.5 mM EDTA, 0.03 mM dithiothreitol,
0.01% gelatin and 0.01% Tween-80), plates were subsequently washed four
times with capture buffer and incubated with detection buffer (50 mM
sodium acetate, pH 6.5, 2.5 mM EDTA, 0.03 mM dithiothreitol, 0.01%
gelatin and 0.01% Tween-80). Cathepsin K activity was measured by
incubation of the captured cathepsin K with a modified prourokinase variant
(UKcatK) in which the endogenous plasmin activation site has been adapted
to a cathepsin K-specific activation site. Activated UKcatK was quantified
using a chromogenic peptide substrate (S-2444; Biosource Europe, Nivelles,
Belgium) and color development was recorded at 405 nm.
Alkaline phosphatase (ALP) activity measurement
ALP was measured after 28 days of culture. Five thousand cells from
differentiation and control experiments were plated on a 96-well plate in triplicate,
the plate was centrifuged and the medium replaced with Fast™ p-Nitrophenyl
Phosphate (pNPP) ALP Substrate (Sigma) containing 2% Triton X-100 (Sigma).
The color was allowed to develop for 30 min and measured at 405 nm.
RANKL and osteoprotegerin ELISA
RANKL ELISA was performed from cell lysates with sRANKL kit
(Biomedica, Vienna, Austria). Samples were diluted 1:5 with PBS. 100 μl of
the diluted sample was added to 96-well microtiter plate wells and 100 μl of
detection antibody was added. A seven point standard diluted from 2 ng/ml to
31.25 pg/ml was prepared from kit stock standard in PBS. Duplicates were used
for samples and standards. Plates were incubated for 16–24 h at +4°C and
washed 3 times with 350 μl washing buffer. After washes, 200 μl streptavidin–
under gentle shaking. Plates were washed three times, 200 μl of tetra methyl
the dark for 20 min. The reaction was stopped with 50 μl stop solution and the
absorbance was measured at 450 nm.
were coated with monoclonal mouse anti-human osteoprotegerin IgG2a(2 μg/ml
in PBS) (R&D Systems, clone 69127.11) overnight at room temperature. The
plates were washed three times with PBS containing 0.05% Tween 20, pH 7.4
and blocked with 1% BSA and 5% sucrose in PBS for 1 h. The blocking buffer
was removed and 100 μl of media samples, diluted 1:100 in 1% BSA in PBS to
adjust the sample to the dynamic range of the assay, was added to each well for
from 2 ng/ml to 31.25 pg/ml was used as a standard. After washes, the bound
osteoprotegerin was detected with biotinylated anti-human osteoprotegerin IgG
(100 ng/ml)(R&D Systems)diluted in 1% BSA in PBS.The plates werewashed
and 1:70,000 diluted alkaline phosphatase conjugated ExtrAvidin (Sigma) was
added. Unbound ExtrAvidin was washed away and the color was developed
using Fast™ pNPP ALP Substrate (Sigma), and the absorbance reaction was
measured at 405 nm.
One-way ANOVAwith Bonferroni’s multiple comparison test was used for
statistical comparisons. Tests were performed withGraphPad Prism version 3.02
for Windows (GraphPad Software, San Diego, CA). All results are expressed as
mean ± SEM.
Osteoblasts, lining cells and some osteocytes contain
To analyze osteoblastic cathepsin K expression in situ, we
Sequences of primers and probes used in RT-PCR
SYBR green I dye used instead of probe
Used only in conventional PCR
The genes and their sequence accession numbers are shown. Primers and probes were designed and tested so that they locate inside one exon in all genes, m = 6-carboxy-fluorescein (FAM); x = 6-carboxy-tetramethyl-
rhodamine (TAMRA); p = phosphorylation.
771J. Mandelin et al. / Bone 38 (2006) 769–777
a well-characterized cathepsin K antibody . Here, we
demonstrate that in osteoporotic bone not only osteoclasts but
also bone lining cells contain cathepsin K. Some osteocytes near
the bone lining cells were also cathepsin K immunoreactive but
controls with irrelevant IgG showed no immunoreactivity (Fig.
1B). We analyzed cathepsin K expression also in intramem-
branous bone by immunostaining samples from patients with
diffuse sclerosing osteomyelitis of the mandible. Bone in this
disease undergoes repeated periods of resorption followed by
repair. Cuboidal osteoblast-like cells and osteocytes showed
similar cathepsin K expression as cells in the osteoporotic bone
Cultured osteoblast-like cells express cathepsin K mRNA and
produce active cathepsin K
In order to characterize and confirm osteoblastic cathep-
sin K expression, we established cultures from the trabecular
bone samples. Cells were derived from the same femoral
heads, which were used for the localization study.
Quantitative RT-PCR showed that in these cultures
(n = 6), cathepsin K mRNA copy number was 382 ± 124
per one PBGD mRNA copy (Fig. 2A and Table 2). Since
mRNA expression does not always lead to protein
translation, we then analyzed the cultures for the presence
of cathepsin K. All samples contained 27 kDa processed
cathepsin K protein and those having high mRNA
expression contained also 42 kDa pro-cathepsin K (Fig.
2B). Cultured cells also contained cathepsin K as revealed
by immunostaining (Fig. 2C). To further confirm these
findings, we analyzed cathepsin K activity in the samples
with a modified urokinase assay, which specifically detects
cathepsin K enzyme activity . Cell lysates contained
0.437 ± 0.034 ng/ml and media samples 0.006 ± 0.002 ng/
ml active cathepsin K.
Cultured osteoblast-like cells synthesize mineralized matrix
To confirm the osteoblastic nature of the cultured cells,
we analyzed some typical osteoblast markers in the cultures.
We stimulated the cells in osteogenic differentiation media
. After 21 days, the cells generated trabecular structures
and formed a mineralized matrix as seen by staining with
silver method of von Kossa (Figs. 3A–C). ALP activity, a
marker for active mineralization and osteoblast maturation,
in osteogenic media was 0.616 ± 0.055 absorbance units in
a colorimetric assay compared to 0.106 ± 0.059 (P b 0.001)
absorbance units in control cultures (Fig. 4A). These cells
also expressed low amounts of osteocalcin mRNA (not
shown) and clear Runx2 expression was detected when the
cells were cultured in osteogenic differentiation media as
revealed by RT-PCR (Fig. 4B). Cells grown in the
osteogenic media expressed considerable amounts of colla-
gen type I after 14 days of stimulation (115 ± 20 mRNA
copies per one PBGD mRNA copy) whereas after 28 days
the expression was reduced to 4 ± 1 (P b 0.001) mRNA
copies per one PBGD mRNA copy.
Bone-forming cells decrease their cathepsin K expression
To further investigate the balance of bone formation and
cathepsin K production, we then investigated how cathepsin
K mRNA expression is affected when cultured cells are
induced to synthesize mineral matrix. Interestingly, cathepsin
K mRNA copy number was 8 ± 2 and 17 ± 7 per one PBGD
mRNA copy after 14 and 28 days of culture in the osteogenic
Fig. 1. Cathepsin K immunoreactivity in bone samples. (A) Lining cells on the
osteoporotic metaphyseal trabeculae display cathepsin K immunoreactivity
(arrows). Some osteocytes in the underlying matrix express also with the
cathepsin K immunoreactivity (arrowheads). (B) A serial section from the same
sample shows no immunoreactivity with an irrelevant rabbit IgG. (C)
Intramembranous bone samples from a patient with diffuse sclerosing
osteomyelitis of the mandible show similar cathepsin K immunoreactivity.
Cuboidal osteoblast-like cells (arrows) and some superficial osteocytes (arrow-
heads) show strong immunoreactivity, while osteocytes in deeper bone display
none (open arrowheads). Original magnification 25×.
772J. Mandelin et al. / Bone 38 (2006) 769–777
differentiation media, respectively (Fig. 2A). This is signif-
icantly lower than the expression levels observed in the basal
(382 ± 124 per one PBGD mRNA copy) medium (P b 0.01
for both time points).
Cathepsin K expression is not related to collagenase-3 mRNA
To reveal how cathepsin K mRNA expression is related
to the expression of other collagenolytic enzymes, we
analyzed collagenase-3 mRNA expression in the cultures.
Collagenase-3 belongs to the family of matrix metallopro-
teinases (MMPs), is regulated in osteoblast-like cells and has
been associated to extracellular matrix degradation during
fetal ossification [14,15]. Collagenase-3 mRNA expression
in the basal culture was 5 ± 2 copies per one PBGD mRNA
copy. In osteogenic differentiation media, the expression was
28 ± 9 copies after 14 days and 4 ± 1 copies after 28 days
(P b 0.01).
MSC-derived osteoblast produces cathepsin K
To analyze if the expression of cathepsin K is unique to
osteoblast-like cells derived from the osteoporotic bone, we
also generated osteoblast-like cells from MSCs (n = 3) .
After 21 days of differentiation in osteogenic medium, these
cells generated mineralized trabecular structures as revealed
by von Kossa staining (Figs. 3D–F) and had increased ALP
activity (Fig. 4A). Cathepsin K expression after 28 days
culture was also evident in these cells (28 ± 4 mRNA
copies per one PBGD mRNA copy); however, this was
significantly lower than the expression levels in the
metaphyseal trabecular bone-derived cells cultured under
basal conditions (P b 0.05, Fig. 2A). MSC-derived
osteoblasts also released 27 kDa processed cathepsin K to
the culture medium as revealed by Western blotting (not
shown). ALP activity and type I collagen expression profile
were similar to osteogenic media stimulated cells derived
from osteoporotic bone.
TNF-? does not significantly change cathepsin K production in
Synoviocytes isolated from patients suffering from
rheumatoid arthritis increase their cathepsin K production
when stimulated with TNF-α . We therefore tested if
TNF-α had an effect on osteoblast-like cells. The cells were
first induced in osteogenic media for 21 days and then
stimulated with TNF-α for 24 h. Without TNF-α stimula-
tion, the expression level of cathepsin K was 24 ± 12
mRNA copies per one PBGD mRNA copy and after TNF-α
stimulation the corresponding copy number was 97 ± 75
mRNA copies per one PBGD mRNA copy (not significant,
NS). When the cells were cultured on basal conditions and
stimulated with TNF-α for 24 h, cathepsin K expression
level was 162 ± 36 mRNA copies per one PBGD mRNA
copy vs. 256 ± 116 mRNA copies per one PBGD mRNA
TNF-? does not stimulate RANKL but stimulates
osteoprotegerin production in cultured osteoblast-like cells
The balance of RANKL and osteoprotegerin controls
osteoblast formation and it has been shown that TNF-α
increases RANKL production in human osteoblast-like cells
. We therefore investigated the production of RANKL
and osteoprotegerin protein and mRNA in osteoblast-like
cells derived from trabecular bone. In basal culture
conditions, the expression of RANKL was 0.002 ± 0.0009
mRNA copies per one PBGD mRNA copy and after TNF-α
stimulation the expression was 0.002 ± 0.0003 mRNA
copies per one PBGD mRNA copy (NS). The corresponding
Fig. 2. Cultured osteoblast-like cells express and produce cathepsin K. (A) Cathepsin K mRNA expression in osteoblast-like cells (n = 6) derived from osteoporotic
bone (OB basal). The expression is significantly higher than that of 28 days cultures (n = 6) in osteogenic media (OB in osteogenic; P b 0.01) or that of MSCs (MSC in
osteogenic; P b 0.05) cultured in osteogenic media (n = 3). (B) Western blot showing released 42 kDa pro- and 27 kDa processed cathepsin K protein in osteoblast-like
cells culturedin basal (lane 1; 15 μg of total protein loaded) medium.Cells cultured in osteogenicmedia express lowamount of cathepsinK mRNA and secrete 27 kDa
processed cathepsin K in the culture media (lane 2). Negative sample control, liver lysate (lane 3; 75 μg of total protein load), does not contain cathepsin K. Positive
samplecontrol,giantcell tumorof bonelysate(lane4;75 μgof total proteinloaded),contains42 kDapro-and27kDa activecathepsinK.Molecularweightmarker(S)
with 37 kDa and 25 kDa bands. (C) Immunolocalization of cathepsin K in osteoblast-like cells.
773J. Mandelin et al. / Bone 38 (2006) 769–777
osteoprotegerin expression levels were 66 ± 37 and
111 ± 51 mRNA copies per one PBGD mRNA copy
To analyze the corresponding protein production in these
cells, we performed ELISA assays. RANKL was analyzed from
cell lysates and osteoprotegerin from culture media. The amount
of RANKL in basal cultures was 0.4 ± 0.1 mg/ml. After TNF-α
stimulation, the amount of RANKL was 0.5 ± 0.1 (NS). The
amount of osteoprotegerin protein in basal cultures was
11.4 ± 1.8 mg/ml. After TNF-α stimulation, the amount of
osteoprotegerin was 24.4 ± 5.1 mg/ml (P b 0.01).
The organic matrix of bone is mainly constituted of type I
collagen. Cathepsin K is unique among mammalian proteinases
since it can degrade type I collagen at several sites inside the
helical region . It is distinguished from classical human
cysteine proteases like cathepsins L and B by its increased
stability at neutral pH as well as its ability to efficiently
hydrolyze type I collagen and elastin at pH values above 6.0
. Lack of cathepsin K is associated to osteopetrosis and
pycnodysostosis  and cathepsin K overexpression results in
enhanced trabecular bone turnover in mice . Early studies
have suggested that osteoclasts are the only cells to produce
cathepsin K in human bone . However, recent studies
indicate that also other cells than osteoclasts can produce
cathepsin K [9,13,21] and participate in extracellular matrix
degradation during bone remodeling [10,22]. Here, we show
that osteoblasts may also be a source of cathepsin K in human
bone. Although there are strong evidence that matrix metallo-
proteinases are responsible for extracellular matrix degradation
in bone [22,23], a recent study of collagen degradation in
pycnodysostosis suggests that cathepsin K is a crucial protease
for collagen degradation in human fibroblasts but not in the
corresponding mouse model . The use of animal models in
several previous studies may explain why the expression and
function of cathepsin K have been exclusively linked to
osteoclasts. It can be stated that human osteoblasts contain
cathepsin K in vivo, although the biological relevance for the
presence cathepsin K in human osteoblasts requires further
Cathepsin K copy numbers in primary osteoblast-like
cells, cultured in basal conditions, were over 70-fold higher
when compared to that of collagenase-3. It therefore appears
that high cathepsin K expression in basal culture conditions
is not related to high expression level of collagenase-3.
Rather the results are in accordance with the observation
that collagenase-3 mRNA expression is linked to matrix
formation in vivo . Actually, in basal conditions, the
primary osteoblast-like cultures produced similar amounts of
cathepsin K mRNA as in vitro generated osteoclasts, which
we have analyzed in our previous work . Although the
copy numbers are not directly comparable, it can be stated
that cultured osteoblast-like cells produce considerable
amounts of cathepsin K mRNA. More importantly, primary
cultures both produced and released cathepsin K enzyme
and with the in vitro activity measurements we were able to
confirm that the enzyme is active. The amount of active
enzyme was quite low in the culture media compared to that
of cell extracts. However, culture time was relatively short,
24 h, and the initial volume was 40-fold higher in the
lysates when compared to media samples. If the volume
difference is taken into account, the activity in culture media
is half of that in cell lysates. At this stage, it is not clear
why cells, which are induced to form mineralized matrix,
reduce their cathepsin K expression. However, it may be
biologically unreasonable to form, and at the same phase to
degrade the newly synthesized organic matrix. Nevertheless,
these results indicate that osteoblasts are able to control the
expression of cathepsin K depending on the stage of their
The osteoblast-like cells were derived from trabecular
bone explants. It can be speculated that a small number of
osteoclasts and/or their precursors of the monocyte/macro-
phage lineage were also able to attach to the wells.
mRNA expression in different culture conditions (copies/porphobilinogen deaminase (PBGD) mRNA copy)
Culture conditionCathepsin KCol α1(I) Collagenase-3Osteoprotegerin RANKL
OB in basal medium
OB in basal medium + TNF-α
OB in osteogenic medium for 14 days
OB in osteogenic induction for osteogenic 28 days
MSC in osteogenic induction for 14 days
MSC in osteogenic induction for 28 days
OB in osteogenic medium for 21 days
OB in osteogenic induction for 21 days + TNF-α
In vitro differentiated osteoclastsd
382 ± 124a
162 ± 36
8 ± 2
17 ± 7
10 ± 6
28 ± 4
24 ± 12
97 ± 75
324 ± 183
10 ± 4
115 ± 20b
4 ± 1
405 ± 57
9 ± 2
5 ± 2
28 ± 9c
4 ± 1
92 ± 75
0.5 ± 0.1
66 ± 37
111 ± 51
0.002 ± 0.0009
0.002 ± 0.0003
ND = not done.
aCathepsin K expression in osteoblast cultured in basal media is significantly higher (P b 0.01) than that of osteoblasts cultured in osteogenic media for 14, 21 or
28 days or that of MSCs cultured in osteogenic induction for 28 days (P b 0.05).
bType I collagen expression in osteoblast or MSCs cultured in osteogenic media for 14 days is significantly higher (P b 0.01) than that of those cells cultured in
osteogenic media for 28 days.
cCollagenase-3 expression in osteoblast or MSCs cultured in osteogenic media for 14 days is significantly higher (P b 0.01) than that of those cells cultured in
osteogenic media for 28 days.
dThe level of cathepsin K expression is similar to that of in vitro differentiated osteoclast, which we have published in our previous study .
774 J. Mandelin et al. / Bone 38 (2006) 769–777
However, the cells were used in passages 2–4 and it is
usually impossible to detach macrophages or osteoclasts
from the wells by conventional trypsinization. Thus, there is
no reason to believe that these cells could be the source of
cathepsin K in the present cultures. Although the purity of
the cultures was not assured with specific markers for
osteoblasts or macrophages, we showed that the cultures
were able to form mineralized matrix under appropriate
conditions. This and the significantly increased ALP activity
during mineralization further confirm the osteoblastic nature
of the cells. Also, the expression profile of type I collagen
is in line with previous studies which have shown that type
I collagen expression is induced during active extracellular
matrix deposition and decreased during mineralization and
osteoblast differentiation to quiescent lining cell or osteocyte
. The production of cathepsin K in differentiated marrow
stromal cells further confirms that osteoblast-like cells can
produce cathepsin K.
In osteoporosis, bone turnover has changed in favor of
bone resorption so that bone-forming osteoblasts are not
able to compensate the resorption . This leads to
trabecular thinning with structural damage. At present, it is
generally accepted that the osteoclast is the only cell that is
able to resorb mineralized bone. Several cytokines and
hormones are known to influence bone metabolism and to
modulate osteoclastic resorption in osteoporosis. Pro-inflam-
matory cytokines, like TNF-α, are known to be able to
affect the balance of RANKL and osteoprotegerin .
They have also an effect on osteoclastic bone resorption
[28–30] and apparently on cathepsin K production in
rheumatoid synoviocytes . We were not, and have not
been , able to demonstrate that TNF-α stimulation
increases RANKL or cathepsin K production in cultured
human osteoblast-like cells. In this study, the concentration
of TNF-α and duration of the stimulation were the same as
in the study of Hou et al. .
Fig. 3. Cultured osteoblast-like cells from the metaphyseal trabecular bone samples (A–C) and from MSCs (D–F) form mineralized matrix resembling trabecular
structures. (A and D) Cells grown for 21 days in osteogenic media form dense aggregates with mineral deposits seen as black precipitate as revealed by von Kossa
staining. (B and E) Same field in showing nuclear DAPI staining reveals dense packed cells and empty spaces without cells. Note that in this field the black precipitate
is also visible. (C and F) Cell cultures in control media grow uniformly and reach confluence at 21 days as shown after von Kossa and DAPI staining. Black precipitate
indicating mineral deposition is not seen in these cultures. Original magnification 12.5×.
775 J. Mandelin et al. / Bone 38 (2006) 769–777
It is well recognized that RANKL is the key factor to
induce osteoclastogenesis locally. Decoy receptor for RANKL
is osteoprotegerin [2,31]. In our experiments, cultured
osteoblasts produced low amounts of RANKL. Both
RANKL mRNA and protein levels are near the detection
limit whereas osteoprotegerin mRNA and protein levels were
high. Naturally, we cannot exclude the possibility that
RANKL expression is significantly decreased in culture
conditions when compared to in vivo situation like Kiviranta
and his co-workers have shown in cultured mouse osteoblasts
. However, it has been shown that in osteoporosis serum
osteoprotegerin levels are increased  and the number of
BMUs is reduced . Osteoprotegerin administration
decreases the number of osteoclasts in mice  and
osteoclastic bone resorption markers in postmenopausal
women . There is even evidence that osteoprotegerin
administration protects bone in TNF-α induced bone
resorption . Thus, the increased serum osteoprotegerin
levels in humans may reflect a compensatory response against
enhanced osteoclastic bone resorption and the resultant loss of
bone. In this respect, our data suggest that osteoblast-like
cells may also participate in organic matrix degradation
during remodeling cycle by producing collagenolytic cathep-
sin K. Osteoblasts, although their number may be reduced in
osteoporotic patients, cover trabeculae and are able to locally
decrease extracellular pH , a requirement for optimal
cathepsin K activity for collagen degradation. It must also be
noted that the slight but detectable progressive metabolic
acidosis during ageing [38,39] may contribute to bone surface
demineralization seen in patients with osteoporosis .
Although evidence is still lacking on the physiological role of
osteoblast-produced cathepsin K, it is feasible to speculate
that this may reflect matrix maintenance and degradation of
damaged and exposed collagen I fibrils, or recycling of
improperly processed/folded (pro-)collagen I. Whether these
mechanisms have consequences in diseases characterized by
abnormal matrix turnover remains as a subject of further
Mr. Erkki Hänninen and Mr. Mikko Rautia are greatly
appreciated for their expert technical help. This study is
supported by the Academy of Finland, TEKES, Sigrid Jusélius
Foundation, Invalid Foundation, Helsinki University Central
Hospital and Biomedicum Helsinki research funds. During this
study, Jami Mandelin has been funded by young investigator
research grant from Emil Aaltonen Foundation.
 Harada S, Rodan GA. Control of osteoblast function and regulation of
bone mass. Nature 2003;423:349–55.
 Boyle WJ, Simonet WS, Lacey DL. Osteoclast differentiation and
activation. Nature 2003;423:337–42.
 Rodan GA, Raisz LG, Bilezikian JP. Pathophysiology of osteoporosis. In:
Bilezikian JP, Raisz LG, Rodan GA, editors. Principles of bone biology.
San Diego, CA: Academic Press; 2002. p. 1275–90.
 Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess T, et al.
Osteoprotegerin ligand is a cytokine that regulates osteoclast differenti-
ation and activation. Cell 1998;93:165–76.
 Vaananen HK, Zhao H, Mulari M, Halleen JM. The cell biology of
osteoclast function. J Cell Sci 2000;113:377–81.
by a polarized vacuolar proton pump. Science 1989;245:855–7.
 Gelb BD, Shi GP, Chapman HA, Desnick RJ. Pycnodysostosis, a
lysosomal disease caused by cathepsin K deficiency. Science 1996;273:
 Yamaza T, Goto T, Kamiya T, Kobayashi Y, Sakai H, Tanaka T. Study
of immunoelectron microscopic localization of cathepsin K in
osteoclasts and other bone cells in the mouse femur. Bone 1998;23:
 Hou WS, Li W, Keyszer G, Weber E, Levy R, Klein MJ, et al. Comparison
of cathepsins K and S expression within the rheumatoid and osteoarthritic
synovium. Arthritis Rheum 2002;46:663–74.
 Mulari MT, Qu Q, Harkonen PL, Vaananen HK. Osteoblast-like cells
complete osteoclastic bone resorption and form new mineralized bone
matrix in vitro. Calcif Tissue Int 2004;75:253–61.
 Pittenger MF, Mackay AM, Beck SC, Jaiswal RK, Douglas R, Mosca JD,
et al. Multilineage potential of adult human mesenchymal stem cells.
 Lindeman JHN, Hanemaaijer R, Mulder A, Dijkstra PDS, Szuhai K,
Bromme D, et al. Cathepsin K is the principal protease in giant cell tumor.
Am J Pathol 2004;165:593–600.
 Konttinen YT, Mandelin J, Li TF, Salo J, Lassus J, Liljestrom M, et al.
Acidic cysteine endoproteinase cathepsin K in the degeneration of the
superficial articular hyaline cartilage in osteoarthritis. Arthritis Rheum
 Stahle-Backdahl M, Sandstedt B, Bruce K, Lindahl A, Jimenez MG, Vega
JA, et al. Collagenase-3 (MMP-13) is expressed during human fetal
ossification and re-expressed in postnatal bone remodeling and in
rheumatoid arthritis. Lab Invest 1997;76:717–28.
 Johansson N, Saarialho-Kere U, Airola K, Herva R, Nissinen L,
Westermarck J, et al. Collagenase-3 (MMP-13) is expressed by
Fig. 4. Cultured osteoblast-like cells and MSCs increase their alkaline
phosphatase activity and Runx2 expression when induced to form mineralized
matrix. (A) Alkaline phosphatase activity in osteoblast-like cells (n = 6) derived
from osteoporotic bone (OB control) under control cultures is significantly
lower than that of osteogenic media (OB in osteogenic; P b 0.001; n = 6) or that
of MSCs (MSC in osteogenic; P b 0.001; n = 3) cultured in osteogenic media.
(B) Representative samples from osteoblast-like cells and MSCs cultured in
osteogenic induction media (lanes 1 and 2, respectively) and in control medium
(lanes 3 and 4, respectively) expressing Runx2. Negative PCR control without
template (lane N) and 100 bp DNA standard (lanes S) displaying 100, 200, 300
and 400 bp bands.
776 J. Mandelin et al. / Bone 38 (2006) 769–777
hypertrophic chondrocytes, periosteal cells, and osteoblasts during Download full-text
human fetal bone development. Dev Dyn 1997;208:387–97.
 Hofbauer LC, Lacey DL, Dunstan CR, Spelsberg TC, Riggs BL, Khosla S.
Interleukin-1beta and tumor necrosis factor-alpha, but not interleukin-6,
stimulate osteoprotegerin ligand gene expression in human osteoblastic
cells. Bone 1999;25:255–9.
 Garnero P, Borel O, Byrjalsen I, Ferreras M, Drake FH, McQueney MS, et
al. The collagenolytic activity of cathepsin K is unique among mammalian
proteinases. J Biol Chem 1998;273:32347–52.
 Bromme D, Okamoto K, Wang BB, Biroc S. Human cathepsin O2,
a matrix protein-degrading cysteine protease expressed in osteoclasts.
frugiperda and characterization of the enzyme. J Biol Chem 1996;271:
 Kiviranta R, Morko J, Uusitalo H, Aro HT, Vuorio E, Rantakokko J.
Accelerated turnover of metaphyseal trabecular bone in mice over-
expressing cathepsin K. J Bone Miner Res 2001;16:1444–52.
 Littlewood-Evans A, Kokubo T, Ishibashi O, Inaoka T, Wlodarski B,
Gallagher JA, et al. Localization of cathepsin K in human osteoclasts
by in situ hybridization and immunohistochemistry. Bone 1997;20:
 Mandelin J, Li T-F, Hukkanen M, Liljestrom M, Salo J, Santavirta S, et al.
Interface tissue fibroblasts from loose total hip replacement prosthesis
produce receptor activator of nuclear factor-kB ligand, osteoprotegerin,
and cathepsin K. J Rheumatol 2005;32:713–20.
 Everts V, Delaisse JM, Korper W, Jansen DC, Tigchelaar-Gutter W, Saftig
P, et al. The bone lining cell: its role in cleaning Howship’s lacunae and
initiating bone formation. J Bone Miner Res 2002;17:77–90.
 Parikka V, Vaananen A, Risteli J, Salo T, Sorsa T, Vaananen HK, et al.
Human mesenchymal stem cell derived osteoblasts degrade organic bone
matrix in vitro by matrix metalloproteinases. Matrix Biol 2005;24:438–47.
 Everts V, Hou WS, Rialland X, Tigchelaar W, Saftig P, Bromme D, et al.
Cathepsin K deficiency in pycnodysostosis results in accumulation of non-
digested phagocytosed collagen in fibroblasts. Calcif Tissue Int 2003;73:
 Jimenez MJ, Balbin M, Lopez JM, Alvarez J, Komori T, Lopez-Otin
C. Collagenase 3 is a target of Cbfa1, a transcription factor of the runt
gene family involved in bone formation. Mol Cell Biol 1999;19:
 Paloneva J, Mandelin J, Kiialainen A, Bohling T, Prudlo J, Hakola P, et al.
DAP12/TREM2 deficiency results in impaired osteoclast differentiation
and osteoporotic features. J Exp Med 2003;198:669–75.
 Aubin JE. Regulation of osteoblast formation and function. Rev Endocr
Metab Disord 2001;2:81–94.
 Jimi E, Akiyama S, Tsurukai T, Okahashi N, Kobayashi K, Udagawa
N, et al. Osteoclast differentiation factor acts as a multifunctional
regulator in murine osteoclast differentiation and function. J Immunol
 Fuller K, Murphy C, Kirstein B, Fox SW, Chambers TJ. TNFalpha
potently activates osteoclasts, through a direct action independent of
and strongly synergistic with RANKL. Endocrinology 2002;143:
 Lam J, Takeshita S, Barker JE, Kanagawa O, Ross FP, Teitelbaum SL.
TNF-alpha induces osteoclastogenesis by direct stimulation of macro-
phages exposed to permissive levels of RANK ligand. J Clin Invest Dec
 Simonet WS, Lacey DL, Dunstan CR, Kelley M, Chang MS, Luthy R, et
al. Osteoprotegerin: a novel secreted protein involved in the regulation of
bone density. Cell 1997;89:309–19.
 Kiviranta R, Morko J, Alatalo SL, NicAmhlaoibh R, Risteli J, Laitala-
Leinonen T, et al. Impaired bone resorption in cathepsin K-deficient mice
is partially compensated for by enhanced osteoclastogenesis and increased
expression of other proteases via an increased RANKL/OPG ratio. Bone
 Yano K, Tsuda E, Washida N, Kobayashi F, Goto M, Harada A, et al.
Immunological characterization of circulating osteoprotegerin/osteoclas-
togenesis inhibitory factor: increased serum concentrations in postmeno-
pausal women with osteoporosis. J Bone Miner Res 1999;14:518–27.
 Ott SM. Histomorphometric analysis of bone remodeling. In: Bilezikian
JP, Raisz LG, Rodan GA, editors. Principles of bone biology. San Diego,
CA: Academic Press; 2002. p. 303–19.
 Bekker PJ, Holloway D, Nakanishi A, Arrighi M, Leese PT, Dunstan
CR. The effect of a single dose of osteoprotegerin in postmenopausal
women. J Bone Miner Res 2001;16:348–60.
 Morony S, Capparelli C, Lee R, Shimamoto G, Boone T, Lacey DL, et al.
A chimeric form of osteoprotegerin inhibits hypercalcemia and bone
resorption induced by IL-1beta, TNF-alpha, PTH, PTHrP, and 1, 25(OH)
2D3. J Bone Miner Res 1999;14:1478–85.
 George M, Stein B, Muller O, Weis-Klemm M, Pap T, Parak WJ, et al.
Metabolic activation stimulates acid secretion and expression of matrix
degrading proteases in human osteoblasts. Ann Rheum Dis 2004;63:
 Frassetto LA, Morris Jr RC, Sebastian A. Effect of age on blood acid-base
composition in adult humans: role of age-related renal functional decline.
Am J Physiol 1996;271:F1114–22.
 Sebastian A, Harris ST, Ottaway JH, Todd KM, Morris Jr RC. Improved
mineral balance and skeletal metabolism in postmenopausal women
treated with potassium bicarbonate. N Engl J Med 1994;330:1776–81.
777J. Mandelin et al. / Bone 38 (2006) 769–777