Eur J Oral Sci 2010; 118: 451–459  2010 Eur J Oral Sci

DOI: 10.1111/j.1600-0722.2010.00760.x European Journal of

Printed in Singapore. All rights reserved
Oral Sciences

Margareth V. Tamburstuen1,
Ameloblastin promotes bone growth by Sjur Reppe2, Axel Spahr3, Roya
Sabetrasekh1, Gunnar Kvalheim4,
enhancing proliferation of progenitor Ivan Slaby1, Unni Syversen5, Staale
Petter Lyngstadaas1, Janne E.
Reseland1
cells and by stimulating 1
Department of Biomaterials, Faculty of
Dentistry, University of Oslo (UiO), Oslo,
immunoregulators Norway; 2Department of Medical Biochemistry,
University of Oslo (UiO), Oslo, Norway;
3
Department of Periodontology, Ulm University,
Ulm, Germany; 4Department of Cellular
Therapy, Radiumhospitalet, Oslo University
Tamburstuen MV, Reppe S, Spahr A, Sabetrasekh R, Kvalheim G, Slaby I, Syversen U, Hospital, Oslo, Norway; 5Department of Cancer
Lyngstadaas SP, Reseland JE. Ameloblastin promotes bone growth by enhancing Research and Molecular Medicine, Faculty of
proliferation of progenitor cells and by stimulating immunoregulators. Medicine, Norwegian University of Science and
Eur J Oral Sci 2010; 118: 451–459.  2010 Eur J Oral Sci Technology (NTNU), Trondheim, Norway

In this study, we examined the role of the enamel matrix protein, ameloblastin, in bone
growth and remodelling, and attempted to identify some of the molecular mechanisms
involved in these processes. The effects of recombinant ameloblastin (rAmbn) were
tested in vivo in rats, and in vitro in primary human mesenchymal stem cells, osteo-
blasts, chondrocytes, and osteoclasts. We used a microarray technique to identify
genes that were regulated in human osteoblasts and verified our findings using Prof. Janne E. Reseland, Department of
multiplex protein analysis and real-time RT-PCR. Recombinant ameloblastin was Biomaterials, Faculty of Dentistry, University of
found to stimulate bone healing in vivo, and to enhance the proliferation of mesen- Oslo, PO Box 1109, Blindern, NO-0317 Oslo,
chymal stem cells and osteoblasts, as well as the differentiation of osteoclast precursor Norway
cells in vitro. The most profound effect was on the regulation of genes related to
Telefax: +47–22–852351
immune responses as well as on the expression of cytokines and markers of bone cell E-mail: j.e.reseland@odont.uio.no
differentiation, indicating that ameloblastin has an effect on mesenchymal cell differ-
entiation. A receptor has not yet been identified, but we found rAmbn to induce, Key words: ameloblastin; bone growth; bone
repair; cell proliferation; cytokines
directly and indirectly, signal transducer and activator of transcription (STAT) 1 and 2
and downstream factors in the interferon pathway. Accepted for publication June 2010

Ameloblastin was previously regarded as a tooth-specific
protein (1–4), mainly found to be synthesized and
secreted into developing enamel matrix by the enamel-
forming ameloblasts. However, it has also been detected
in mesenchymal pulp cells, including pre-odontoblasts
and young odontoblasts, during dentin formation (5–7).
Recent reports showing the expression of ameloblastin
during craniofacial bone development in rats (8), and the
expression and secretion of ameloblastin from cultured
human mesenchymal precursor cells, osteoblasts, and
osteoclasts (9), indicate a role for ameloblastin in bone
formation and regeneration.
Multiple isoforms of ameloblastin have been identified
in the developing enamel of the species studied, and each
isoform may potentially serve a unique physiological
role (10–12). A three-dimensional model of ameloblastin
(13) supports earlier experimental observations suggest-
ing that ameloblastin is a bipolar, two-domain protein
that interacts with Ca2+ ions (14, 15). The primary
structure of ameloblastin generates two chemically and
physically distinct regions – a basic N-terminal region
and an acidic C-terminal region – which can be liberated
by proteolysis from the larger precursor molecule during
physiological processing of enamel or bone matrix (13).
A receptor specific for ameloblastin has so far not been
identified.
The physiological role for the ameloblastin protein in
hard tissue development and maintenance remains un-
known, although it has been suggested to control apatite
crystal growth, to determine enamel prismatic structure
and to be involved in cell signaling during dental devel-
opment (16–19). Ameloblastin is also found to have a
role in pulpal healing and to induce secondary dentin
formation in miniature pigs (20). In the present study we
attempted to clarify the role of ameloblastin in bone
formation and repair and to identify some of the
molecular mechanisms involved.
Material and methods
Bone-healing model
All animal procedures were carried out in accordance with,
and approved by, the local ethical committee of Ulm
University, Germany, and were performed in accordance
with German and European Union (EU) regulations.
Adult female Sprague–Dawley rats (n = 4), each weigh-
ing approximately 200 g, were injected subcutaneously with
0.05 mg kg)1 of atropine (Braun, Melsungen, Germany)
15 min before, and then anaesthetized with xylazine
(12 mg kg)1) and ketamine (75 mg kg)1) 5 min before,
starting the surgical procedure. The clean-shaven heads and
452 Tamburstuen et al.
necks, as well as the oral cavities, of the animals were
washed preoperatively with ethanol/chlorhexidine (5.0 g of
chlorhexidine digluconate in 1,000 ml of ethanol; Pharma-
cia, Uppsala, Sweden), and all experiments were performed
under aseptic conditions. An infrared lamp was used to
maintain body temperature.
A horizontal incision of about 10 mm was made along the
buccodistal aspect of the mandible behind and below the
location of the molar teeth. After performing a blunt sub-
cutaneous dissection to mobilize the skin, the distal part of
the jaw bone was uncovered. Damage to the major blood
vessels and nerves, as well as perforation of the oral cavity,
were avoided. After exposing the buccal bone, a cylindrical
defect with a diameter of 2.5 mm and penetrating the ramus
was drilled using a low-speed trepan burr and irrigated with
copious amounts of sterile saline. The periosteum covering
the area of the holes was completely removed to ensure that
the defects healed with fibrous tissue filling and not by direct
bone formation. Bilateral defects were created in each ani-
mal: one served as the control and one as the test. Either
100 ll of PBS or 100 ll of recombinant ameloblastin
(rAmbn) (5, 20) (1 mg ml)1 in PBS) was added to precut
soft lyophilized collagen sheets (Lyoplant; Braun Aescu-
lap,Tutsingen, Germany) immediately before implantation
in the defects. After this procedure the soft tissue, muscles,
and blood vessels were carefully repositioned and the wound
was closed by sutures (Perma hand Seide 4-0; Ethicon,
Norderstedt, Germany). After surgery, the animals were
placed in separate cages under an infrared lamp to regain
consciousness. After 14 d, the animals were killed and the
segments of the lower jaws containing the bone defects were
removed en bloc together with adjacent teeth and alveolar
bone, fixed in freshly prepared 4% buffered paraformalde-
hyde for about 16 h, decalcified in 12.5% EDTA for 14 d,
rinsed in 1% PBS for 1 d, and then embedded in paraffin.
RNase-free water was used for all solutions. Semi-serial
sections (7 lm) were mounted on aminopropyltriethoxysi-
lane (Sigma)-coated glass slides. The segments were stained
with hematoxylin and eosin, and evaluated by light
microscopy. The percentage of new bone formation in the
control defects (n = 4) and in rat rAmbn (rrAmbn)-treated
defects (n = 4) was quantified.
Cell cultures
Human mesenchymal stem cells (MSCs) from two donors
were purchased from Cambrex Bio Science (Walkersville,
MD, USA), and additional MSCs from four donors which
were also used in the experiments were established at
Radiumhospitalet, Oslo, Norway. Following informed
consent, bone marrow aspirates were collected simulta-
neously with adipose tissue (obtained by liposuction) from
two patients. The MSCs from bone marrow were initially
cultured in minimum essential medium, alpha modification
(MEM-a) (Invitrogen, Grand Island, NY, USA) containing
20% fetal calf serum (FCS) (PAA Laboratories, Pasching,
Austria), 100 U ml)1 of penicillin, and 0.1 mg ml)1 of
streptomycin (Invitrogen Life Technologies, Madison, WI,
USA), while adipose tissue-derived MSCs were cultured in
DulbeccoÕs modified EagleÕs medium (DMEM)/nutrient
mixture F-12 Ham (1:1, vol/vol) (Sigma-Aldrich, St Louis,
MO, USA) supplemented with 20% FCS, 100 U ml)1 of
penicillin, and 0.1 mg ml)1 of streptomycin. The bone
marrow-derived MSCs were thereafter subcultured in
MEM-a containing 10% FCS, 100 U ml)1 of penicillin, and
0.1 mg ml)1 of streptomycin, and the adipose-derived
MSCs were subcultured in DMEM (PAA Laboratories)
containing 10% FCS, 100 U ml)1 of penicillin, and
100 lg ml)1 of streptomycin. The immunophenotypes of
the undifferentiated MSCs from bone marrow and adipose
tissue were characterized, by flow cytometry, to be CD45
low or CD45 negative, and CD10, CD13 and CD90 posi-
tive; moreover, their ability to differentiate in osteogenic
and adipose directions, as well as to undergo clonogenic
growth in a low-density colony-forming unit fibroblast
(CFU-f) assay, were tested (21, 22). To maintain the via-
bility of the stem cells, only half of the cell culture medium
was replaced twice weekly with fresh cell culture medium.
Commercially available primary osteoblasts from both
femur and tibia of different donors (NHOst cell system;
Cambrex Bio Science) were grown in osteoblast growth
medium (Cambrex Bio Science). Osteoblasts (NHO) cul-
tured to facilitate differentiation were exposed to hydro-
cortisone hemisuccinate (200 nM) and b-glycerophosphate
(10 mM) (Cambrex Bio Science) in the ambient medium.
Commercially available human primary chondrocytes
(NHACs) from knee joint were purchased from two
different companies (hCa were from Cell Applications,
San Diego, CA, USA; and NHAC-kn were from Camb-
rex Bio Science). The cells were grown in chondrocyte
growth medium (obtained from Cell Applications and
Cambrex Bio Science). Chondrocyte differentiation med-
ium containing 5% fetal bovine serum (FBS), 0.1% gen-
tamicin sulfate/amphotericin B, 0.5% transforming
growth factor-b1 (TGF-b1), 0.2% of the long R3 analog
of insulin-like growth factor 1 (R3 IGF-1), 0.2% insulin,
0.2% transferrin, and 2.5% ascorbic acid, was added to
chondrocyte cell cultures to facilitate differentiation.
Osteoclasts were differentiated from human peripheral
mononuclear cells (PBMCs) in medium to which was added
macrophage-colony stimulating factor (M-CSF) and recep-
tor activator of nuclear factor-kappaB ligand (RANKL)
(50 ng ml)1 of each). The cells were stained for tartrate-
resistant acid phosphatase (TRAP) activity using Naphthol
AS-BI phosphate and Fast Garnet in the presence of sodium
tartrate, as described by the manufacturer (Sigma-Aldrich).
The number of TRAP-positive and multinuclear (three or
more nuclei) cells in each well were counted, and these were
regarded as genuine osteoclasts.
Recombinant human Ambn (rhAmbn) was used in the
experiments with NHOs. Recombinant human Ambn
(1-447 amino acids) (a gift from Dan Deutsch, Institute for
Dental Sciences, Hebrew University, Jerusalem, Israel) was
expressed in the cytoplasm of Spodoptera frugiperda (SF9)
insect cells. This recombinant protein was purified and
characterized by SDS-PAGE, western blotting, and elec-
trospray ionization-mass spectrometry (ESI-MS/MS), to
confirm that the product was the full-length human
ameloblastin protein (Rosenfeld E. et al., unpublished
article). Cultures of MSCs, NHACs and PBMCs were
stimulated using rrAmbn (5, 20). Both recombinant proteins
(rhAmbn and rrAmbn) were added at final concentrations of
2–50 lg ml)1, and the cell culture medium and cell lysates
were harvested after 6 and 12 h, and 1, 2, 3, and 7 d of
incubation. Detailed information related to cell type, con-
centration of rAmbn, and incubation time is stated in the
description of the assays used and in the presentation of the
results.
To evaluate whether rrAmbn contained microbial factors
from the host cells that induced a cytokine response,
rrAmbn (2 and 10 lg ml)1) was added to U937 cells
(2 · 105 cells ml)1), a human leukemic monocyte lymphoma
cell line (ATCC, Manassas, VA, USA), and the effect on
cytokines in the cell culture medium was compared with the

effect of lipopolysaccharides (LPS; 1 and 10 lg ml)1)
(Sigma-Aldrich), as a positive control, and with untreated
cells, as a negative control, after 1, 2, and 3 d of incubation.
Cell viability was confirmed by monitoring the release of
lactate dehydrogenase (LDH) into the medium using the
microplate-based Cytotoxicity Detection Kit (LDH)
(Boehringer, Mannheim, Germany). According to the
manufacturerÕs protocol, 50-ll aliquots of cell culture
medium were used and the absorbance was read on a
microplate reader at 450 nm. The number of cells at the
start, and after 3 d of incubation, was calculated using the
Contess automated cell counter (Invitrogen).
Proliferation assay
The proliferation rate of NHOs and MSCs was measured
using [3H]thymidine incorporation. Subconfluent cells were
incubated for 24, 48, and 72 h with growth medium con-
taining rAmbn (10 lg ml)1), and with a control containing
growth medium only, and then pulsed with 1 lCi/well of
[3H]thymidine, 12 h before harvesting. The medium was
removed and the cells were washed twice with ice-cold PBS
and twice with ice-cold 5% trichloracetic acid to remove
unincorporated [3H]thymidine. The cells were solubilized in
125 ll of 1 M sodium hydroxide (NaOH), and 100 ll of the
solubilized cell solution was transferred to 2 ml of scintillation
fluid (Lumagel LSC; GE, Groningen, the Netherlands) and
counted for 3 min in a liquid scintillation counter (Packard
1900 TR; Packard Instruments, Meriden, CT, USA).
Wound-healing assay
NHOs were seeded in four, two-well glass slides (Chamber
Slides system; Lab-Tek, Thermo Fisher Scientific, Roskilde,
Denmark). The cells were rinsed in PBS, and a pipette tip
was used to make straight scratches in the confluent cell
cultures. The cells were rinsed again in PBS, and on each
slide the cells in one chamber were incubated with cell cul-
ture medium (control) and the cells in the other chamber
were incubated with medium containing 10 lg ml)1 of
rrAmbn. Images of the scratches were taken at time 0, and
after 6, 12, and 24 h of incubation, using the Olympus
Cell8 software system with 40· magnification. The wound
healing or closing of the scratches in defined surface areas
after 6, 12, and 24 h were compared with the individual
areas at time-point 0.
Differentiation assays
Human PBMCs were cultured for 12 d in medium
containing M-CSF and RANKL (50 ng ml)1 of each), with
and without 10 or 50 lg ml)1 of rAmbn. The number of
osteoclasts was evaluated visually and counted manually.
Mesenchymal stem cells were cultured to 90% confluence
before administration of 10 lg ml)1 of rrAmbn. Cell culture
medium and cell lysates were harvested after 1, 2, 3, and 7 d.
The phenotype of the cells was investigated based on the
changes in the expression levels of collagen type 1, runt-
related transcription factor 2 (Runx-2), bone morphogenetic
protein 2 (BMP2), osteocalcin, and CD44.
mRNA isolation
MSCs, NHOs and NHACs were lysed in lysis/binding buffer
[100 mM Tris–HCl (pH 8.0), 500 mM LiCl, 10 mM EDTA
Ameloblastin promotes bone growth 453
(pH 8.0), 0.5 mM dithiothreitol (DTT), and 1% SDS].
mRNA was isolated using magnetic beads [oligo (dT)25], as
described by the manufacturer (Dynal, Oslo, Norway). Beads
containing mRNA were resuspended in 10 mM Tris–HCl
(pH 8.0) and stored at )70C until use. One microlitre of the
mRNA-containing solution was applied directly to obtain a
first-strand cDNA using the iScript cDNA Synthesis Kit with
both oligo (dT) and random hexamer primers (Bio-Rad,
Hercules, CA, USA).
Real-time PCR
RT-PCR reactions were performed and monitored using
iCycler iQ (Bio-Rad). The iQ SYBR Green Supermix was
based on iTaq DNA polymerase (Bio-Rad). cDNA samples
were analyzed for the genes of interest and for the reference
genes glyceraldehyde-3-phosphate dehydrogenase (gapdh)
and b-actin. The amplification program consisted of a pre-
incubation step for denaturation of the template cDNA
(3 min, 95C), followed by 40 cycles [50 cycles for human
ameloblastin mRNA (AMBN)], consisting of a denaturation
step (15 s, 95C), an annealing step (30 s 60C), and an
extension step (30 s 72C). After each cycle, fluorescence
was measured at 72C. A negative control without a cDNA
template was run in each assay. Samples were run in
duplicate. To allow relative quantification after PCR,
standard curves were constructed from the standard reac-
tions for each target and housekeeping genes by plotting
cycle threshold (Ct) values, that is, the cycle number at
which the fluorescence signal exceeds the background vs. log
cDNA dilution. The Ct readings for each of the unknown
samples were then used to calculate the amount of either the
target gene or the housekeeping gene relative to the
standard. mRNA levels were calculated as the ratio con-
centration for the target genes relative to the housekeeping
genes. Oligonucleotide primer sequences used for the real-
time RT-PCR and the specific parameters are shown in
Table 1. Real-time efficiencies were calculated from the
given slopes in the iCycler software using serial dilutions,
showing all the high real-time PCR efficiency rates of the
investigated transcripts, and high linearity (r = ±0.99)
when different concentrations were used. PCR products
were subjected to a melting curve analysis on the iCycler
and subsequently to 2% agarose/Tris–acetate–EDTA
(TAE) gel electrophoresis to confirm amplification speci-
ficity, and amplicon size, respectively (see Table 1).
Expression of AMBN was confirmed by cDNA sequencing
(Lark Technologies, Hope End Takeley, Essex, UK).
Microarray analysis
The effect of rhAmbn on gene expression in primary human
osteoblasts was studied upon stimulation of osteoblasts for
24 h and compared with the expression in untreated cells.
RNA was isolated using TRIzol reagent (Gibco-BRL Life
Technologies, Gaithersburg, MD, USA), according to the
manufacturerÕs instructions, and further purified using the
RNeasy kit (Qiagen, Valencia, CA, USA) in order to
remove organic components.
Double-stranded cDNA and biotin-labelled cRNA
probes were created from 5 lg of total RNA using the
Superscript Choice system (Invitrogen) and the Bioarray
(Enzo Biochem, New York, NY, USA), respectively. Pro-
cedures were carried out according to recommendations
from Affymetrix (Santa Clara, CA, USA). This cRNA was
454 Tamburstuen et al.
Table 1
Primers used in real-time PCR quantification
Gene Primer sequence
AMBN S 5¢-AGCCATGTTTCCAGGATTTG-3¢
A 5¢-TGCACCTCCTTCTTCGTTCT-3¢
Runx2 S 5́-GCCTTCAAGGTGGTAGCCC-3́
A 5́-CGTTACCCGCCATGACAGTA-3́
CD44 S 5́-CAAGTTTTGGTGGCACACAGC-3́
A 5́- GAAGCAATATGTGTCATACTGGGA G-3́
Osteocalcin S 5́-GAAGCCCAGCGGTGCA-3́
A 5́-CACTACCTCGCTGCCCTCC-3́
GAPDH S 5́-TGCACCACCAACTGCTTAGC-3́
A 5́-GGCATGGACTGTGGTCATGAG-3
b-actin S 5́-CTGGAACGGTGAAGGTGACA-3́
A 5́-AAGGGACTTCCTGTAACAATGCA-3́
hybridized to Hu-133 A chips (Affymetrix), containing
cDNA oligonucleotides representing more than 22,000
transcripts, followed by washing and staining on the
GeneChips Fluidics Station 450 (Affymetrix), according to
the manufacturerÕs instructions. The chips were scanned on
the Affymetrix GeneArray 2500 scanner. The P-value for
global analysis takes into account the number of genes
associated with a given function or pathway, as well as the
amount of information in the Ingenuity annotations for the
associated genes in the given function or pathway. The data
sets were processed using the Affymetrix mas 5.0 software,
and signal values representing the expression level of each
transcript were generated.
Data evaluation via Ingenuity pathways analysis
The significance of the canonical pathways was calculated
using the right-tailed FisherÕs exact test. The P-value was
calculated by comparing the number of user-specified genes
of interest (i.e. the regulated genes) that participate in a
given function or pathway, relative to the total number of
occurrences of these genes in all functional/pathway anno-
tations stored in the Ingenuity Pathways Knowledge Base.
Only annotations that had more Functions/Canonical
Pathways Analysis Genes than expected by chance (Ôright-
tailedÕ annotations) were used. The P-value for Global
analysis takes into account the number of genes associated
with a given function or pathway as well as the amount of
information in the Ingenuity annotations for the associated
genes in the given function or pathway.
Determination of cytokine levels in the cell culture
medium
Multi-analyte profiling of the level of cytokines in the cell
culture medium of MSCs, NHO and U937 cells was per-
formed on the Luminex-100 system (Luminex, Austin, TX,
USA). Aliquots of rAmbn (10 lg ml)1) were tested in the
same assay to determine the presence of contamination that
affected the analysis. Calibration microspheres for classifi-
cation and reporter readings, as well as sheath fluid, were
also purchased from Luminex. Acquired fluorescence data
were analyzed using the Starstation software (Version 2.0;
Applied Cytometry Systems, Sheffield, UK). The amount of
various cytokines in the culture media was measured using
the Human Adipocyte Lincoplex Kit (Linco, St Charles,
Amplicon
Species size (bp)
Human 142
Human 67
Human 352
Human 70
Human 87
Human 140
MO, USA) and Invitrogen Human Cytokine Twenty-
Five-Plex (Invitrogen), according to the manufacturerÕs
instructions.
Data presentation
Statistical comparisons between treatments were performed
using the non-parametric Mann–Whitney Rank Sum test
and anova on ranks; however, when both normality and
equal variance tests passed, the StudentÕs t-test was used.
Differences were considered statistically significant at
P < 0.05.
Results
rAmbn induces new bone formation in vivo
Treatment with rrAmbn (Fig. 1A) induced new bone
formation within the collagen carrier. The new bone
has the appearance of woven or trabecular bone with
bone marrow spaces (Fig. 1A). In the PBS controls,
only small amounts of bone growing from the margins
out, as appositional growth from the old bone margins,
could be observed. These collagen sponges were mostly
surrounded by dense, cell-poor, fibrous tissues, some-
times with signs of inflammatory cell infiltrate
(Fig. 1B).
New bone formation was significantly (P = 0.006)
higher in the treated defects (60.5 ± 9.8% of the area)
(n = 4) compared with the control (25.3 ± 13.6% of the
area) (n = 4).
Effect of rAmbn on proliferation, migration, and
differentiation of bone cells
Recombinant human ameloblastin (10 lg ml)1) signifi-
cantly enhanced the number of osteoblasts after 2 d
(P = 0.005) and 3 d (P = 0.04) (Fig. 2A), and rrAmbn
(10 lg ml)1) enhanced the number of MSCs after 3 d
(P < 0.001) (Fig. 2B), compared with untreated
control cells at the same time-points. Recombinant
rat ameloblastin (10 lg ml)1) also enhanced the
migration of NHOs and reduced the surface areas in a

A A
B B
Fig. 1. Bone defect in rat jaw 2 wk after surgery. Bone growth
in the defect with the collagen sponge wetted in a recombinant
rat ameloblastin (rrAbbn)/PBS solution (1 mg of rrAmbn ml)1
of PBS) immediately before implantation (A) and in the control
defect with collagen sponges wetted in PBS only (B). The
images are from a representative specimen (only low-resolution
images exist because the material was lost during relocation).
wound-healing assay, significantly (P < 0.001), after
24 h compared with the control (Fig. 2C).
The number of differentiated osteoclasts was higher
(P = 0.002) in the presence of 10 lg ml)1 of rrAmbn
compared with the control, but not significantly different
in the presence of 50 lg ml)1 of rrAmbn (Fig. 3).
In MSCs the expression of collagen type I was reduced
(P = 0.05) 7 d after incubation with rrAmbn, whereas
the expression of osteocalcin (P = 0.05) and of CD44
(P < 0.009) was enhanced after 1 d compared with
untreated cells (Fig. 4A). Recombinant human ameob-
lastin also stimulated the expression of CD44 mRNA in
cultured primary human osteoblasts, and rrAmbn stim-
ulated the expression of CD44 mRNA in chondrocytes
(data not shown), whereas no significant effect was
observed on the expression of Runx2 mRNA in either of
the cell types studied. Recombinant human ameloblastin
significantly enhanced the expression of BMP2 mRNA
in osteoblasts at all time-points tested (P = 0. 024 for all
time-points) (Fig. 4B).
Microarray analysis demonstrated that rhAmbn
activated the expression of a large number of genes in
primary human osteoblasts. Genes related to major bio-
logical networks that were significantly regulated more
than twofold, as determined using the mas 5.0 software,
are listed in Table 2. Most pronounced was the regulation
of genes involved in immune responses, and in particular,
the expression levels of interleukin (IL)-6, IL-8, and IL-1
were significantly enhanced (P < 0.001) in the cells. A
subset of regulated genes was also confirmed using real-
time RT-PCR, and rhAmbn was found to increase the
expression of IL6 mRNA (P < 0.001) after 24 h of
Ameloblastin promotes bone growth 455
C
Fig. 2. Effect of recombinant ameloblastin (rAmbn) on prolif-
eration and migration. [3H]Thymidine incorporation in osteo-
blasts (NHOs) (A) and mesenchymal stem cells (MSCs) (B)
incubated with recombinant ameloblastin (rAmbn) [human
rAmbn (hrAmbn) and rat rAmbn (rrAmbn), respectively]
(10 lg ml)1) and without rAmbn (control) for 1, 2, and 3 d.
The results are expressed as the percentage of [3H]thymidine
incorporation in cells stimulated with rAmbn relative to un-
treated control cells at each time point. Migrated cell-surface
area, expressed as a percentage of the total surface area (C),
after 6, 12, and 24 h, with and without rrAmbn (10 lg ml)1).
*P < 0.05, **P < 0.01, ***P < 0.001.
incubation (data not shown). Microarray analysis was
also confirmed by immunoassays, and the IL-6 level in the
cell culture medium was enhanced in both NHOs (not
significant, day 3; P = 0.001, for days 1 and 7) and
MSCs (P = 0.004 for days 1, 2, and 3; P = 0.01 for day
7) upon administration of rhAmbn and rrAmbn com-
pared with untreated cells (Fig. 5A). In the cell culture
medium of undifferentiated chondrocytes, an increase of
more than eightfold (P = 0.02) in the concentration of
IL-6 was found after only 6 h of incubation compared
with untreated cells (data not shown). Whereas the
concentration of IL-8 in the cell culture medium of NHOs
was not significantly different from the untreated controls,
stimulation with rrAmbn caused an increase in the con-
centration of IL-8 in the culture medium of MSCs
(P = 0.029 for days 2 and 3; not significant for days 1 and
7) compared with the control (Fig. 5B). An enhanced level
of monocyte chemotactic protein-1 (MCP-1) was also
456 Tamburstuen et al.
Fig. 3. Effect of recombinant rat ameloblastin (rrAmbn) on
osteoclast differentiation. Human peripheral blood mono-
nuclear cells (PBMCs) were cultured in medium containing
macrophage-colony stimulating factor (M-CSF) and receptor
activator of nuclear factor-kappaB ligand (RANKL)
(50 ng ml)1 of each) for 12 d with or without rrAmbn (10 and
50 lg ml)1). Multinuclear (three or more nuclei) tartrate-resis-
tant acid phosphatase (TRAP)-positive cells were counted and
regarded as osteoclasts. The data represent the mean ± SD of
three individual experiments. **P < 0.01.
found in MSCs after 3 d of incubation (P = 0.029)
compared with untreated cells (Fig. 5C). Recombinant rat
ameloblastin did not contain any contaminants that
affected the immunoassay of cytokines, and had no effects
on cytokine production compared with untreated U937
cells (data not shown). By contrast, LPS (both 1 and
10 lg ml)1) induced a 150-fold increase in IL-6, an
increase of more than 10-fold in IL-8, and a 30-fold increase
in the secretion of MCP-1 into the cell culture medium
(P < 0.001, respectively) compared with untreated U937
cells (P < 0.001, respectively) (data not shown).
The rhAmbn-induced signal changes in genes involved
in signalling and metabolic pathways are given in Table 3.
Importantly, the expression levels of genes implicated in
interferon signalling were altered. An ameloblastin
receptor has so far not been identified; however, rhAmbn
stimulated significantly the expression of signal transducer
and activator of transcription (STAT) 1 (P < 0.001),
STAT2 (P = 0.001), and downstream factors in the
interferon pathway. However, no effect on the expression
of the receptors for interferon-c (IFNRa and IFNRb) or
interferon-a/b (IFNR1 and 2) could be detected.
Discussion
In the present study we demonstrate that rAmbn stim-
ulates bone growth and repair in experimental defects in
mandibular bone in rats. We also found that rAmbn
enhances the number of bone precursor cells and osteo-
blasts, as well as osteoclast-like cells in vitro. This stim-
ulation of cell growth was followed by the enhanced
expression and secretion of factors involved in immune
responses.
It has previously also been demonstrated that amel-
oblastin is expressed during bone repair (8) and that mice
with a mutant ameloblastin gene appear to have reduced
bone growth (23). We found that rAmbn induced new
bone formation within the defect, and not as apposi-
tional growth from old bone margins. Taken together
with the observed enhancement in proliferation of pro-
genitor cells and osteoblasts, and the migration of cells,
B
Fig. 4. The effect of recombinant rat Ambn (rrAmbn;
10 lg ml)1) on the mRNA expression of Runx2, osteocalcin
(OC), collagen type 1, and CD44 in mesenchymal stem cells
(MSC) (A), and of recombinant human Ambn (rhAmbn) on
bone morphogenetic protein 2 (BMP2) in osteoblasts (NHOs)
(B) after 1, 3, and 7 d of incubation. The expression is
calculated relative to the expression of two housekeeping genes
in each sample, and presented as a percentage of the relative
expression in untreated cells at each time point. The data
represent the mean ± SD of four individual experiments with
different donors. *P < 0.05, **P < 0.01.
our results indicate that ameloblastin has the potential to
induce cell recruitment and growth both in vivo and
in vitro.
We used both rhAmbn and rrAmbn; rhAmbn was
produced in S. frugiperda (Sf9) insect cells and rrAmbn
was expressed as a fusion protein with thioredoxine in
Escherichia coli. Thioredoxine has previously been found
to be inactive in eukaryotic cells (5, 20), and we found no
indications of any contamination from the expression
host affecting the cytokine detection in our system
because the cytokine secretion from monocytes was
unaltered by rrAmbn administration. Ameloblastin has
potential phosphorylation sites and a putative O-linked
glycosylation site (24), which might result in structural
differences between rhAmbn and rrAmbn proteins.
Although the primary structure of ameloblastin is highly
conserved between species, rodent ameloblastin contains
an integrin-binding domain and a thrombospondin cell-
adhesion domain, sequences that are not conserved in
the human ameloblastin molecule. The human amelo-
blastin homolog has a unique 26-amino acid insert in the
middle part of the molecule that appears to be a dupli-
cation of short exon 7 (11, 25). Despite the differences in
host expression systems, and the differences between
human and rat sequences, we obtained the same results
for both recombinant proteins used.
Recombinant ameloblastin enhanced the growth of
both MSCs and mature osteoblasts. We found that the
expression of CD44 mRNA was considerably increased
upon stimulation with rrAmbn and rhAmbn. CD44
 Ameloblastin promotes bone growth 457

Table 2 A
The 10 most significant physiological functions affected upon
stimulation of primary human osteoblasts with recombinant
human ameloblastin (rhAmbn) (10 lg ml )1)

Number Significance
of genes range
Function affected (P-values)

Immunological disease 96 9.7 · 10)17–1.6 · 10)3

Inflammatory disease 103 4.3 · 10)16–1.7 · 10)3
Connective tissue disorder 78 1.2 · 10)15–1.6 · 10)3
Organismal injury and 52 2.2 · 10)15–1.1 · 10)3 B
abnormalities
Skeletal and muscle disorders 92 2.5 · 10)14–1.6 · 10)3
Infection mechanism 43 6.5 · 10)14–7.8 · 10)4
Inflammatory response 73 1.0 · 10)12–1.7 · 10)3
Cellular movement 67 7.3 · 10)12–1.6 · 10)3
Hematological system 58 7.3 · 10)12–1.8 · 10)3
development and function
Hematopoiesis 44 7.3 · 10)12–1.3 · 10)3

Genes that had a log2 value of at least ± 1.0 upon stimulation

with rhAmbn compared with no rhAmbn added were selected for C
analysis and association with functions within the Ingenuity
Pathways Knowledge Base. FisherÕs exact test was used to
calculate a P-value determining the probability that each
biological function assigned to the data set occurs by chance alone.

has previously been demonstrated to be required for

proliferation and homing of mesenchymal precursor
cells (26). The stimulated expression of osteocalcin and
BMP2 mRNA species suggests that ameloblastin plays
a role in the osteogenic commitment of mesenchymal
cells. Bone morphogenetic proteins have a central role
in bone development and osteoblast differentiation (27),
but require interactions with other growth factor-
activated signals. Bone morphogenetic protein 2 has
been reported to stimulate the transcription of Runx2
(28, 29), which, in turn, regulates the bone-specific
transcription of osteocalcin (30, 31). The enhanced
expression of BMP2 was only verified in ameloblastin-
treated osteoblasts. The transcriptional level was too
low to be quantified in primary MSC cultures in the
present experiments.
Recombinant rat ameloblastin induced an invert dose-
dependent effect on the differentiation of osteoclast
precursors, with the highest effect observed at the lowest
dose tested. Similar dose-dependent effects have been
reported using other stimulatory factors, such as IL-32
(32). Stimulation of osteoclast differentiation and activ-
ity is a crucial step in bone remodeling and repair (33).
We found rAmbn to stimulate pro-inflammatory
proteins, such as IL-6, IL-8, and MCP-1. Interleukin-6 is
highly expressed in the bone marrow stroma (34) and is
known for its role in bone homeostasis (35), and in
particular bone turnover, by stimulating osteoclast
differentiation (36). However, IL-6 has been reported to
be both necessary and sufficient for enhanced MSC
proliferation and differentiation into osteoblasts (37, 38),
protection of MSC from apoptosis, inhibition of adipo-
genic and chondrogenic differentiation of MSCs, and
increasing the rate of in vitro wound healing of MSCs
(39). Interleukin-6 is also found to stimulate the secretion
Fig. 5. The effect of recombinant Ambn (rAmbn) on the
concentrations of interleukin (IL)-6 (A), IL-8 (B), and mono-
cyte chemotactic protein-1 (MCP-1) (C) in the cell culture
medium of human primary osteoblasts (NHOs) (human
rAmbn) and mesenchymal stem cells (MSCs) (rat rAmbn). The
amount of protein is presented as a percentage of the level in the
cell culture medium of untreated cells (control) at the same time
points of more than three individual experiments. *P < 0.05,
**P < 0.01, ***P < 0.001.
of IL-8 (40) and MCP-1 (41) in bone marrow cells.
Interleukin-8 has been reported to both directly stimulate
(42) and suppress osteoclastogenesis, as well as to affect
the migration of osteoclasts (43) and other cells in bone
marrow (44). Both IL-8 and MCP-1 have been found to
enhance bone marrow cell migration (44, 45), and MCP-
1, in particular, is assumed to have a role in progenitor
cell recruitment (46–48).
We observed a rapid filling of the bony defect in rats,
indicating that ameloblastin has a strong effect on
osteogenic cells. Results from studies in the rat may,
however, be difficult to relate directly to studies
performed on human primary cells in vitro. Our data
indicate that ameloblastin has an effect on the expression
of genes related to growth and repair in human primary
cells. Besides the direct effect on proliferation and
differentiation, ameloblastin may also have an indirect
effect on the recruitment and stimulation of osteoclasts
as well as osteoblast precursors.
Receptors for ameloblastin or its splice products have
never been identified. We found the expression of STAT1
458 Tamburstuen et al.

Table 3 Sciences, Hebrew University-Hadassah Faculty of Dental

Medicine, Israel) for the kind gift of human recombinant
The 10 canonical pathways affected most significantly upon ameloblastin, and Britt Mari Kvam and Aina Mari Lian
stimulation of primary human osteoblasts with recombinant (Faculty of Dentistry, Oral Research Laboratory, University of
human ameloblastin (rhAmbn) Oslo (UiO), Oslo, Norway) for skilful technical assistance. We
also thank Bente Marie Berntsen Jacobsen and Lise Aagaard
Ratio of Significance
Willadsen (Norwegian Radium Hospital, Oslo, Norway), and
Pathway affected/total (P values)
Kari Slørdahl (Norwegian University of Science and Technol-
Interferon signalling 12/30 5.3 · 10)14 ogy (NTNU), Trondheim, Norway) for isolation and charac-
Role of pattern recognition 14/86 1.0 · 10)10 terization of human mesenchymal stem cells and osteoclasts,
receptors in recognition of respectively.
bacteria and viruses
Antigen-presentation pathway 9/39 2.5 · 10)9
Activation of IRF by cytosolic 11/74 2.2 · 10)8
pattern recognition receptors References
Hepatic fibrosis/hepatic stellate 12/135 5.7 · 10)6 1. Krebsbach PH, Lee SK, Matsuki Y, Kozak CA, Yamada
cell activation KM, Yamada Y. Full-length sequence, localization, and
)6
TREM1 signaling 8/69 6.6 · 10 chromosomal mapping of ameloblastin. A novel tooth-specific
Dendritic cell maturation 12/172 1.4 · 10)5 gene. J Biol Chem 1996; 271: 4431–4435.
Atherosclerosis signaling 9/110 4.0 · 10)5 2. Fong CD, Hammarstrom L, Lundmark C, Wurtz T, Slaby
Graft-vs.-host disease signaling 6/45 4.4 · 10)5 I. Expression patterns of RNAs for amelin and amelogenin in
Protein ubiquitination pathway 12/201 5.4 · 10)5 developing rat molars and incisors. Adv Dent Res 1996; 10: 195–
200.
Genes that had a log2 value of at least ± 1.0 upon stimulation 3. Lee SK, Krebsbach PH, Matsuki Y, Nanci A, Yamada KM,
with rhAmbn compared with no rhAmbn added were selected Yamada Y. Ameloblastin expression in rat incisors and human
for analysis and association with canonical pathways within the tooth germs. Int J Dev Biol 1996; 40: 1141–1150.
Ingenuity Pathways Knowledge Base. The significance of the 4. Cerny R, Slaby I, Hammarstrom L, Wurtz T. A novel gene
association between the data set and the canonical pathway was expressed in rat ameloblasts codes for proteins with cell binding
measured in two ways. (1) A ratio of the number of genes from domains. J Bone Miner Res 1996; 11: 883–891.
the data set that map to the pathway divided by the total 5. Fong CD, Cerny R, Hammarstrom L, Slaby I. Sequential
number of genes that map to the canonical pathway is expression of an amelin gene in mesenchymal and epithelial
displayed. (2) FisherÕs exact test was used to calculate a P-value cells during odontogenesis in rats. Eur J Oral Sci 1998; 106:
determining the probability that the association between the 324–330.
6. Hao J, He G, Narayanan K, Zou B, Lin L, Muni T,
genes in the data set and the canonical pathway is explained by
Ramachandran A, George A. Identification of differentially
chance alone. IRF, interferon regulatory factor; TREM1, the expressed cDNA transcripts from a rat odontoblast cell line.
triggering receptor expressed on myeloid cells. Bone 2005; 37: 578–588.
7. Begue-Kirn C, Krebsbach PH, Bartlett JD, Butler WT.
Dentin sialoprotein, dentin phosphoprotein, enamelysin and
and STAT2, as well as the expression of their down-
ameloblastin: tooth-specific molecules that are distinctively
stream factors in the interferon pathway, to be highly expressed during murine dental differentiation. Eur J Oral Sci
enhanced in osteoblasts. This effect can either be direct 1998; 106: 963–970.
or indirect; however, we found no effect on the expres- 8. Spahr A, Lyngstadaas SP, Slaby I, Pezeshki G. Ameloblastin
sion of the interferon receptors. The immune and skeletal expression during craniofacial bone formation in rats. Eur J
Oral Sci 2006; 114: 504–511.
systems have a variety of regulatory molecules, such as 9. Tamburstuen MV, Reseland JE, Spahr A, Brookes SJ,
cytokines, in common. Most known cytokines and Slaby I, Kvalheim G, Snead ML, Lyngstadaas SP.
growth factors use the Janus kinase (Jak)–STAT Ameloblastin expression and putative autoregulation in
signaling pathway to transmit extracellular signals to the mesenchymal cells suggest a role in early bone formation and
healing. Bone 2010; in press.
nucleus (49) Interleukin-6 mainly induces STAT3
10. Simmons D, Gu TT, Krebsbach PH, Yamada Y, Macdou-
activation (50, 51), but has been found to cross-regulate gall M. Identification and characterization of a cDNA for
STAT1 activation (51, 52). Although interferons are mouse ameloblastin. Connect Tissue Res 1998; 39: 3–12.
known to play crucial roles in the regulation of immune 11. Macdougall M, Simmons D, Gu TT, Forsman-Semb K,
responses, interferon signaling also interferes directly Mardh CK, Mesbah M, Forest N, Krebsbach PH, Yamada
Y, Berdal A. Cloning, characterization and immunolocaliza-
with bone resorption and osteoclast differentiation tion of human ameloblastin. Eur J Oral Sci 2000; 108: 303–310.
through RANKL (53), and interferes with bone growth 12. Lee SK, Kim SM, Lee YJ, Yamada KM, Yamada Y, Chi JG.
and osteoblast differentiation through Runx2 (54). The The structure of the rat ameloblastin gene and its expression in
growth-promoting effects of ameloblastin may be related amelogenesis. Mol Cells 2003; 15: 216–225.
13. Vymetal J, Slaby I, Spahr A, Vondrasek J, Lyngstadaas SP.
to the early steps of bone healing and growth by
Bioinformatic analysis and molecular modelling of human
enhancing proliferation and activating bone progenitor ameloblastin suggest a two-domain intrinsically unstructured
cells; however, the strong effect on immune modulators calcium-binding protein. Eur J Oral Sci 2008; 116: 124–134.
appears to differ from other growth-promoting factors 14. Murakami C, Dohi N, Fukae M, Tanabe T, Yamakoshi Y,
such as BMPs, fibroblast growth factors (FGFs), and Wakida K, Satoda T, Takahashi O, Shimizu M, Ryu OH,
Simmer JP, Uchida T. Immunochemical and immunohisto-
insulin growth factor (IGF). chemical study of the 27- and 29-kDa calcium-binding proteins
and related proteins in the porcine tooth germ. Histochem Cell
Acknowledgements – The project received financial support Biol 1997; 107: 485–494.
from the Norwegian Cancer Society, Norwegian Research 15. Iwata T, Yamakoshi Y, Hu JC, Ishikawa I, Bartlett JD,
Council and EU Grant QLK3-CT-2001-00090. We thank Prof. Krebsbach PH, Simmer JP. Processing of ameloblastin by
Dan Deutsch (Dental Research Laboratory, Institute of Dental MMP-20. J Dent Res 2007; 86: 153–157.

16. Lyngstadaas SP. Synthetic hammerhead ribozymes as tools in
gene expression. Crit Rev Oral Biol Med 2001; 12: 469–478.
17. Nanci A, Zalzal S, Lavoie P, Kunikata M, Chen W,
Krebsbach PH, Yamada Y, Hammarstrom L, Simmer JP,
Fincham AG, Snead ML, Smith CE. Comparative immuno-
chemical analyses of the developmental expression and distri-
bution of ameloblastin and amelogenin in rat incisors.
J Histochem Cytochem 1998; 46: 911–934.
18. Snead ML. Enamel biology logodaedaly: getting to the root of
the problem, or ‘‘whoÕs on first...’’. J Bone Miner Res 1996; 11:
899–904.
19. Dhamija S, Liu Y, Yamada Y, Snead ML, Krebsbach PH.
Cloning and characterization of the murine ameloblastin
promoter. J Biol Chem 1999; 274: 20738–20743.
20. Nakamura Y, Slaby I, Spahr A, Pezeshki G, Matsumoto K,
Lyngstadaas SP. Ameloblastin fusion protein enhances pulpal
healing and dentin formation in porcine teeth. Calcif Tissue Int
2006; 78: 278–284.
21. Gimble J, Guilak F. Adipose-derived adult stem cells: isola-
tion, characterization, and differentiation potential. Cytother-
apy 2003; 5: 362–369.
22. Castro-Malaspina H, Gay RE, Resnick G, Kapoor N,
Meyers P, Chiarieri D, Mckenzie S, Broxmeyer HE, Moore
MA. Characterization of human bone marrow fibroblast
colony-forming cells (CFU-F) and their progeny. Blood 1980;
56: 289–301.
23. Wazen RM, Moffatt P, Zalzal SF, Yamada Y, Nanci A. A
mouse model expressing a truncated form of ameloblastin
exhibits dental and junctional epithelium defects. Matrix Biol
2009; 28: 292–303.
24. Kobayashi K, Yamakoshi Y, Hu JC, Gomi K, Arai T, Fukae
M, Krebsbach PH, Simmer JP. Splicing determines the
glycosylation state of ameloblastin. J Dent Res 2007; 86: 962–967.
25. Toyosawa S, Fujiwara T, Ooshima T, Shintani S, Sato A,
Ogawa Y, Sobue S, Ijuhin N. Cloning and characterization of
the human ameloblastin gene. Gene 2000; 256: 1–11.
26. Khaldoyanidi S, Denzel A, Zoller M. Requirement for
CD44 in proliferation and homing of hematopoietic precursor
cells. J Leukoc Biol 1996; 60: 579–592.
27. Li X, Cao X. BMP signaling and skeletogenesis. Ann N Y Acad
Sci 2006; 1068: 26–40.
28. Lian JB, Stein GS, Javed A, Van Wijnen AJ, Stein JL,
Montecino M, Hassan MQ, Gaur T, Lengner CJ, Young
DW. Networks and hubs for the transcriptional control of
osteoblastogenesis. Rev Endocr Metab Disord 2006; 7: 1–16.
29. Hassan MQ, Tare RS, Lee SH, Mandeville M, Morasso MI,
Javed A, Van Wijnen AJ, Stein JL, Stein GS, Lian JB. BMP2
commitment to the osteogenic lineage involves activation of
Runx2 by DLX3 and a homeodomain transcriptional network.
J Biol Chem 2006; 281: 40515–40526.
30. Yu S, Jiang Y, Galson DL, Luo M, Lai Y, Lu Y, Ouyang
HJ, Zhang J, Xiao G. General transcription factor IIA-gamma
increases osteoblast-specific osteocalcin gene expression via
activating transcription factor 4 and runt-related transcription
factor 2. J Biol Chem 2008; 283: 5542–5553.
31. Paredes R, Arriagada G, Cruzat F, Olate J, Van WA, Lian
J, Stein G, Stein J, Montecino M. The Runx2 transcription
factor plays a key role in the 1alpha,25-dihydroxy Vitamin
D3-dependent upregulation of the rat osteocalcin (OC) gene
expression in osteoblastic cells. J Steroid Biochem Mol Biol
2004; 90: 269–271.
32. Mabilleau G, Sabokbar A. Interleukin-32 promotes osteo-
clast differentiation but not osteoclast activation. PLoS ONE
2009; 4: e4173.
33. Khosla S. Minireview: the OPG/RANKL/RANK system.
Endocrinology 2001; 142: 5050–5055.
34. Ogasawara H, Tsuji T, Hirano D, Aoki Y, Nakamura M,
Kodama H. Induction of IL-6 production by bone marrow
stromal cells on the adhesion of IL-6-dependent hematopoietic
cells. J Cell Physiol 1996; 169: 209–216.
35. Manolagas SC. Role of cytokines in bone resorption. Bone
1995; 17: 63S–67S.
36. Rozen N, Ish-Shalom S, Rachmiel A, Stein H, Lewinson D.
Interleukin-6 modulates trabecular and endochondral bone
Ameloblastin promotes bone growth 459
turnover in the nude mouse by stimulating osteoclast differen-
tiation. Bone 2000; 26: 469–474.
37. Erices A, Conget P, Rojas C, Minguell JJ. Gp130 activation
by soluble interleukin-6 receptor/interleukin-6 enhances osteo-
blastic differentiation of human bone marrow-derived mesen-
chymal stem cells. Exp Cell Res 2002; 280: 24–32.
38. Laflamme C, Rouabhia M. Effect of BMP-2 and BMP-7
homodimers and a mixture of BMP-2/BMP-7 homodimers on
osteoblast adhesion and growth following culture on a collagen
scaffold. Biomed Mater 2008; 3: 15008.
39. Pricola KL, Kuhn NZ, Haleem-Smith H, Song Y, Tuan RS.
Interleukin-6 maintains bone marrow-derived mesenchymal
stem cell stemness by an ERK1/2-dependent mechanism. J Cell
Biochem 2009; 108: 577–588.
40. Rougier F, Cornu E, Praloran V, Denizot Y. IL-6 and IL-8
production by human bone marrow stromal cells. Cytokine
1998; 10: 93–97.
41. Biswas P, Delfanti F, Bernasconi S, Mengozzi M, Cota M,
Polentarutti N, Mantovani A, Lazzarin A, Sozzani S, Poli
G. Interleukin-6 induces monocyte chemotactic protein-1 in
peripheral blood mononuclear cells and in the U937 cell line.
Blood 1998; 91: 258–265.
42. Bendre MS, Montague DC, Peery T, Akel NS, Gaddy D,
Suva LJ. Interleukin-8 stimulation of osteoclastogenesis and
bone resorption is a mechanism for the increased osteolysis of
metastatic bone disease. Bone 2003; 33: 28–37.
43. Lassus J, Waris V, Xu JW, Li TF, Hao J, Nietosvaara Y,
Santavirta S, Konttinen YT. Increased interleukin-8 (IL-8)
expression is related to aseptic loosening of total hip replace-
ment. Arch Orthop Trauma Surg 2000; 120: 328–332.
44. Van Eeden SF, Terashima T. Interleukin 8 (IL-8) and the
release of leukocytes from the bone marrow. Leuk Lymphoma
2000; 37: 259–271.
45. Wang L, Li Y, Chen X, Chen J, Gautam SC, Xu Y, Chopp M.
MCP-1, MIP-1, IL-8 and ischemic cerebral tissue enhance
human bone marrow stromal cell migration in interface culture.
Hematology 2002; 7: 113–117.
46. Gordon RJ, Mcgregor AL, Connor B. Chemokines direct
neural progenitor cell migration following striatal cell loss.
Mol Cell Neurosci 2009; 41: 219–232.
47. Fuller K, Owens JM, Chambers TJ. Macrophage inflamma-
tory protein-1 alpha and IL-8 stimulate the motility but sup-
press the resorption of isolated rat osteoclasts. J Immunol 1995;
154: 6065–6072.
48. Zhang F, Tsai S, Kato K, Yamanouchi D, Wang C, Rafii S,
Liu B, Kent KC. Transforming growth factor-beta promotes
recruitment of bone marrow cells and bone marrow-derived
mesenchymal stem cells through stimulation of MCP-1
production in vascular smooth muscle cells. J Biol Chem 2009;
284: 17564–17574.
49. Ivashkiv LB, Hu X. Signaling by STATs. Arthritis Res Ther
2004; 6: 159–168.
50. Li Y, Backesjo CM, Haldosen LA, Lindgren U. IL-6
receptor expression and IL-6 effects change during osteoblast
differentiation. Cytokine 2008; 43: 165–173.
51. Costa-Pereira AP, Tininini S, Strobl B, Alonzi T, Schlaak
JF, IsÕharc H, Gesualdo I, Newman SJ, Kerr IM, Poli V.
Mutational switch of an IL-6 response to an interferon-gamma-
like response. Proc Natl Acad Sci USA 2002; 99: 8043–8047.
52. Nishimura R, Moriyama K, Yasukawa K, Mundy GR,
Yoneda T. Combination of interleukin-6 and soluble interleu-
kin-6 receptors induces differentiation and activation of
JAK-STAT and MAP kinase pathways in MG-63 human
osteoblastic cells. J Bone Miner Res 1998; 13: 777–785.
53. Takayanagi H, Kim S, Matsuo K, Suzuki H, Suzuki T, Sato
K, Yokochi T, Oda H, Nakamura K, Ida N, Wagner EF,
Taniguchi T. RANKL maintains bone homeostasis through
c-Fos-dependent induction of interferon-beta. Nature 2002;
416: 744–749.
54. Kim S, Koga T, Isobe M, Kern BE, Yokochi T, Chin YE,
Karsenty G, Taniguchi T, Takayanagi H. Stat1 functions as
a cytoplasmic attenuator of Runx2 in the transcriptional
program of osteoblast differentiation. Genes Dev 2003; 17:
1979–1991.