Full text of DOI:10.1359/jbmr.2001.16.1.24

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 16, Number 1, 2001
© 2001 American Society for Bone and Mineral Research
Tensile Stress Induces Bone Morphogenetic Protein 4 in
Preosteoblastic and Fibroblastic Cells, Which Later
Differentiate into Osteoblasts Leading to Osteogenesis in
the Mouse Calvariae in Organ Culture
MIKA IKEGAME,1,2 OSAMU ISHIBASHI,1 TATSUYA YOSHIZAWA,1 JUNKO SHIMOMURA,1
TOSHIHISA KOMORI,3 HIDEHIRO OZAWA,2 and HIROYUKI KAWASHIMA1
ABSTRACT
Mechanical stress is an important factor controlling bone remodeling, which maintains proper bone mor-
phology and functions. However, the mechanism by which mechanical stress is transduced into biological
stimuli remains unclear. Therefore, the purpose of this study is to examine how gene expression changes with
osteoblast differentiation and which cells differentiate into osteoblasts. Tensile stress was applied to the cranial
suture of neonatal mouse calvaria in a culture by means of helical springs. The suture was extended gradually,
displaying a marked increase in cell number including osteoblasts. A histochemical study showed that this
osteoblast differentiation began in the neighborhood of the existing osteoblasts, which can be seen by 3 h. The
site of osteoblast differentiation moved with time toward the center of the suture, which resulted in an
extension of osteoid. Scattered areas of the extended osteoid were calcified by 48 h. Reverse-transcription
polymerase chain reaction (RT-PCR) revealed that tensile stress increased bone morphogenetic protein 4
(BMP-4) gene expression by 6 h and it remained elevated thereafter. This was caused by the induction of the
gene in preosteoblastic cells in the neighborhood of osteoblasts and adjacent spindle-shaped fibroblastic cells.
These changes were evident as early as 3 h and continued moving toward the center of the suture. The
expression of Cbfa1/Osf-2, an osteoblast-specific transcription factor, followed that of BMP-4 and those cells
positive with these genes appeared to differentiate into osteoblasts. These results suggest that BMP-4 may play
a pivotal role by acting as an autocrine and a paracrine factor for recruiting osteoblasts in tensile stress–
induced osteogenesis. (J Bone Miner Res 2001;16:24–32)
Key words: mechanical stress, osteoblast differentiation, osteogenesis, bone morphogenetic protein 4, osteo-
blast precursors
in immobilization, results in disuse osteopenia or osteopo-
rosis. Similarly, orthodontic tooth movement necessitates
bone resorption at the pressure side with concomitant bone
formation at the tension side. Osteoblasts are the bone-
INTRODUCTION
ECHANICAL STRESS is known to influence both bone
formation and bone resorption. Tensional stress pro-
M
motes bone formation, while the lack of physical stress, as forming cells responsible for bone remodeling, a complex
1Department of Pharmacology, Niigata University, Niigata-City, Niigata, Japan.
2Department of Oral Anatomy I, Faculty of Dentistry, Niigata University, Niigata-City, Niigata, Japan.
3Department of Molecular Medicine, Osaka University Medical School, Suita-City, Osaka, Japan.
24

TENSILE STRESS INDUCES BMP-4–MEDIATED OSTEOGENESIS
25
series of events whereby bone is first resorbed and then new
bone is deposited. Because osteoblasts and osteocytes are
believed to be cells responsive to mechanical stress, under-
standing the mechanism(s) by which externally applied
forces are transduced to those cells to yield a series of
biochemical events such as bone remodeling is, therefore,
important.
Recent studies, using in vitro mechanically loaded bone
tissues and/or osteoblastic cells, as well as in vivo animal
models with repetitive compressive mechanical stress ap-
plication, have shown responses similar to that observed
when cells have been exposed to extracellular ligands such
as growth factors, neuropeptides, or hormones, all modula-
tors of cellular phenotype. These changes include produc-
tion of cyclic adenosine monophosphate (cAMP),(1,2) inosi-
tol phosphates,(1,3) 1,2-diacyglycerol,(4) prostaglandins,(5–9)
and nitric oxide.(10–12) Activation of protein kinase C(4) and
ion channels(9,13,14) and an increase of intracellular calci-
um(15,16) have been reported as well. It also has been shown
that expressions of many genes or gene products are regu-
lated by mechanical stress. These genes include early re-
sponsive genes such as c-fos(17–19) and egr-1,(20,21) genes
involved in the eicosanoid metabolism(5) and nitrogen oxide
synthesis,(22) and genes encoding growth factors such as
insulin-like growth factor I (IGF-I)(17,22,23) and transforming
growth factor ␤ (TGF-␤).(7,18) However, these responses
also have been found in many other tissues and so far no
osteoblast-specific events that eventually lead to osteogen-
esis induced by mechanical stress have been shown. To
elucidate the mechanism(s) whereby mechanical stress is
transduced into osteogenesis, we selected an organ culture
system in which mouse calvarial sutures were incubated
with or without tensile stress. This experimental system was
reported by Hickory and Nanda,(24) in which tensile stress
stimulated bone formation in rat calvarial sutures. We chose
this experimental system for the following reasons. First,
calvarial sutures contain not only bone tissues but also
surrounding soft tissues including fibroblastic cells, which
could differentiate into osteoblasts. This system is more
complex and a much better model of the in vivo situation
compared with the osteoblastic cell culture system. Second,
it is less complex and better suited for microscopic obser-
vations of cell differentiation compared with in vivo exper-
iments. Third, because osteogenesis of calvariae is devoid
of endochondral ossification, the system is simpler than
those including other parts of bone such as long bone
FIG. 1. Application of tensile stress to calvarial sutures in
culture. (A) Schematic illustration of the helical springs
used. See text for details. (B) Calvarial suture was removed
from mice and tensile stress was applied as illustrated. The
suture was suspended in the medium of each well.
BMP-4 may play a role as an autocrine and paracrine factor
in the differentiation of preosteoblasts and fibroblasts into
osteoblasts, thereby stimulating osteogenesis induced by
tensile stress in the organ culture of the calvarial suture.
MATERIALS AND METHODS
Organ culture and mechanical force device
Calvariae were aseptically removed from neonatal ddY
tissues. Fourth, this organ culture system is free from sys- mice (3–4 days old) and the cranial suture of each calvaria
temic regulation such as hormonal control. was incubated in a 24-well plate (Corning Costar, Corning,
Using this model, the purpose of the present study was to NY, USA) with or without mechanical stress at 37°C and
examine the time course of osteogenesis induced by me- 5% CO2 in air in BGJb medium (Gibco BRL, Life Tech-
chanical stress with particular interest as to which cell or nologies, Inc., Rockville, MD, USA) supplemented with
cells can differentiate into osteoblasts, and which genes, 10% fetal bovine serum (FBS; Gibco BRL), penicillin G
which are specific to osteoblast function or osteogenesis, are (100 U/ml; Gibco BRL), and streptomycin (100 ␮g/ml;
regulated. Here we present data showing that tensile stress Gibco BRL) for the period indicated. Tensile stress was
application to mouse calvarial sutures induced differentia- applied to each cranial suture using a helical spring made of
tion and growth of osteoblasts, which eventually lead to Elgiloy orthodontic wire (Elgiloy 0.01 in, semiresilient;
osteogenesis. Before osteoblasts were recruited, bone mor- Rocky Mountain, Denver, CO, USA) according to the
phogenetic protein 4 (BMP-4) gene expression was up- method of Hickory and Nanda(24) with a slight modification
regulated or induced in the preosteoblast and fibroblastic (Fig. 1). The magnitude of each helical spring was adjusted
cells surrounding bone tissues, which was followed by to give 0.2 g tensile stress when the distance at the top of
induction of Cbfa1/Osf-2 expression in these cells. Thus, each arm (length, 4 cm) was set at 5 mm (Fig. 1). The

26
IKEGAME ET AL.
magnitude of the spring was decreased at a rate of 0.01
g/mm (data not shown). For the control, springs maintained
at 5 mm by means of adhesive tape, thus giving 0 g of
tensile strength, were used (Fig. 1). After incubation, cranial
sutures were subjected to histochemical analyses, in situ
hybridization, or reverse-transcription polymerase chain re-
action (RT-PCR).
Histochemical studies
Cranial sutures were fixed for 1 h at 4°C with a fixative
containing 4% paraformaldehyde, embedded in optimum
cutting temperature (O.C.T.) compound (Tissue-Tek; Miles,
Inc., Elkhart, IN, USA) and 8-␮m cryosections were cut by
Cryotome (Shandon, Elgany, France). The sections were
then subjected to histochemical analyses. Alkaline phospha-
tase activity was detected according to the azo-dye method
of Burstone.(25) The same sections were then subjected to
von Kossa staining to detect calcification. Methylene blue
was used for the counterstaining. To detect cartilage as well
as calcified tissues in the samples, some of sections were
subjected to alizarin red and Alcian blue staining.
RT-PCR
Cranial sutures were freed from calcified tissues under a
dissection microscope and total RNA was extracted using
Trizol (Gibco BRL) according to the manufacturer‘s proto-
col. Then, first-strand complementary DNA (cDNA) was
synthesized using oligo (dT) (Gibco BRL) and Superscript
II (Gibco BRL) as a primer and a reverse transcriptase,
respectively. For PCR reactions, aliquots of synthesized
cDNA were added to a PCR reaction mixture (0.04 ␮l/␮l)
containing 3Ј- and 5Ј-primer (0.2 ␮M each), deoxynucleo-
side triphosphate (dNTP) mixture (0.2 mM each; Gibco
BRL), and Taq polymerase (0.05 U/␮l) in Taq polymerase
buffer (1ϫ; Gibco BRL). Cycling conditions were 94°C/30
s, 55°C/30 s, and 72°C/60 s for alkaline phosphatase (AL-
Pase), type I collagen, osteocalcin, glyceraldehyde-3-
phosphate dehydrogenase (GAPDH), and ␤-actin for 25
cycles, and 94°C/30 s, 57°C/30 s, and 72°C/60 s for BMP-4
for 32 cycles, respectively (Fig. 2 lower panel). Products
were separated on a 1% agarose gel. Primers for PCR are as
follows: oligonucleotides 5Ј-GACTGGTACTCGGATAA-
CGAGATGC-3Ј and 5Ј-TGCGGTTCCAGACATAGTGG-3Ј
corresponding to the 735–759 nucleotides (nt) and 980–999 nt
of mouse ALPase gene, respectively; 5Ј-ACCATCTGG-
CATCTCATGGC-3Ј and 5Ј-GCAACACAATTGCACCTG-
AGG-3Ј corresponding to the 666–685 nt and 847–867 nt of
mouse type I collagen gene, respectively; 5Ј-TGCGCTCTG-
TCTCTCTGACC-3Ј and 5Ј-CTGTGACATCCATACTTGC-
AGG-3Ј corresponding to the 87–167 nt and 422–443 nt of
mouse osteocalcin gene, respectively; 5Ј-CAGCGGTCAGG-
AAGAAGAATAAG-3Ј and 5Ј-TCTGCACAATGGCATG-
GTTG-3Ј corresponding to the 1615–1638 nt and 1776–1795
nt of mouse BMP-4 gene, respectively; 5Ј-CAAGAAGGT-
GGTGAAGCAGG-3Ј and 5Ј-ATTGAGAGCAATGCCA-
GCC-3Ј corresponding to the 814–833 nt and 934–952 nt of
mouse GAPDH gene; 5Ј-CAGGAGATGGCCACTGCC-
FIG. 2. Primers and conditions used for PCRs. See text
for details.
sponding to the 753–773 nt and 1007–1027 nt of mouse
␤-actin, respectively (Fig. 2).
Probe preparation
Digoxigenin (DIG)-labeled sense and antisense single-
stranded RNA probes were prepared with a DIG RNA
Labeling Kit (Boehringer Mannheim, Mannheim, Germany)
according to the manufacturer’s instructions. For BMP-4, a
456-base pair (bp) cDNA fragment (366–821 nt of acces-
sion number X56848) was obtained by RT-PCR and sub-
cloned into pGEM-T Vector System (Promega, Madison,
WI, USA). After transcription, 40 U of RNAse-free DNAse
(Boehringer Mannheim) were added to the reaction and an
additional 10-minute incubation was done at 37°C. Tran-
scription products were recovered by ethanol precipitation,
and precipitates were washed once in 70% ethanol, air-
dried, and resuspended in 100 ␮l diethylpyrocarbonate
(DEPC)-treated water. For Cbfa1/Osf-2 a 0.3-kb fragment
of exon I of mouse cDNA(26) was subcloned into pBlue-
script KS(Ϫ) (Stratagene, La Jolla, CA, USA). The sub-
cloned fragments were confirmed to be identical to the
respective cDNA sequence by an ABI PRISM 377 DNA
Sequencing System (Perkin Elmer Co., Foster City, CA,
USA).
In situ hybridization
After incubation, cranial sutures were fixed with 4%
GCA-3Ј and 5Ј-TCCTTCTGCATCCTGTCAGCA-3Ј corre- paraformaldehyde in DEPC-treated phosphate-buffered sa-

TENSILE STRESS INDUCES BMP-4–MEDIATED OSTEOGENESIS
27
line (PBS) at 4°C overnight, embedded in O.C.T. com-
pound, and 6-␮m cryosections were cut by Cryotome. Cryo-
sections were washed in PBS, treated with 0.2N HCl for 10
minutes, and incubated with 0.5 ␮g/ml proteinase K (Gibco
BRL) in 10 mM Tris-HCl (pH 8.0) at 37°C for 30 minutes.
Cryosections were then refixed with 4% paraformaldehyde
in PBS at room temperature for 20 minutes and further
treated with 0.25% acetic anhydride in 0.1 M triethanol-
amine for acetylation. Acetylated cryosections were hybrid-
ized with probes at a final concentration of 1.0 ␮g/ml in a
hybridization mixture containing 50% deionized form-
amide, 2% blocking reagent (Boehringer Mannheim), 5ϫ
SSC (1ϫ SSC ϭ 0.15 M NaCl, pH 7.5, 15 mM sodium
citrate), 0.02% sodium dodecyl sulfate (SDS), 0.1%
N-lauroylsarcosine, 200 ␮g/ml yeast transfer RNA (tRNA),
and 100 ␮g/ml of salmon sperm DNA at 55°C for 18–20 h,
and then washed twice with 0.1ϫ SSC at 60°C for 15
minutes followed by RNAse (Wako Chemicals, Osaka,
Japan) treatment at 37°C for 30 minutes. The sections were
washed further twice with 2ϫ SSC at 60°C for 15 minutes
and then twice with 0.1ϫ SSC at 60°C for 15 minutes.
Specific transcripts were detected with anti-DIG–
conjugated ALPase antibody by the manufacturer’s protocol
(DIG Detection Kit; Boehringer Mannheim). The sections
were counterstained with nuclear fast red.
RESULTS
Tensile stress induces osteoblast differentiation and
osteogenesis in calvarial suture
Figure 3 depicts cranial sutures of calvariae with or
without mechanical loading. Short trabeculae(27) stained
with alizarin red extend from the edge of the bone (Figs. 3A
and 3B, arrows). These trabeculae were visible under a
dissecting microscope without staining and they were ex-
tended further by mechanical loading (Figs. 3C–3E, arrows;
tensile stress was applied outward). Sutures were extended
in a time-dependent manner as shown in Fig. 3F.
Figure 4 illustrates cross-sections of calvarial sutures with
ALPase activity and von Kossa staining. The calcified bone
and partly calcified osteoid were covered with oval-shaped
ALPase positive osteoblasts, which were particularly abun-
dant at the growing tip of the osteoid (Fig. 4A). Between the
osteoblasts layer and periosteum or dura mater and at the
center of the suture, many fibroblastic cells were seen
among the fibrous matrices, and none of these cells were
ALPase positive (Fig. 4A). Tensile stress increased the
width of the suture in a time-dependent manner (Figs. 3 and
4). This increase of suture width was accompanied by cell
proliferation evaluated by immunohistochemistry using the
proliferating cell nuclear antigen (PCNA) monoclonal anti-
body (Ab-1; Oncogene Research Products, Cambridge,
MA, USA; data not shown). As described previously, the
tensile stress application resulted in a continuous time-
FIG. 3. Tensile stress elongated the calvarial sutures in a
time-dependent manner. Calvariae from 3- to 4-day-old
mice were cultured as described in the text. (A) Alizarin red
and Alcian blue staining of the suture at time 0. Blue
staining indicates the presence of cartilaginous tissue (orig-
inal magnification ϫ12). (B) Higher magnification of the
parietal bone (B) and suture (S) framed in panel A. Note
alizarin red–stained short trabeculae extended from the pa-
rietal bone (arrowheads; original magnification ϫ40). (C–
E). Calvariae under tensile stress at 0, 24, and 48 h, respec-
tively. The short trabeculae can be seen without staining
(arrowheads; original magnification ϫ40). (F) Suture width
was determined as the distance between the two calcified
bone edges and plotted against the duration under tensile
stress. Each point and bar represents a mean of five exper-
iments Ϯ SD, respectively.

28
IKEGAME ET AL.
FIG. 4. Tensile stress increased in osteoblast number and tissues as the calvarial sutures increased in size. Cross-sections
of calvariae were subjected to ALPase activity and von Kossa staining. (A) Calvaria at time 0. Note the oval-shaped
ALPase positive osteoblasts surrounding the bone (B) and osteoid (*) and preosteoblasts (slant arrow; S, suture). (B–C)
Calvaria at 1, 6, 24, and 48 h under tensile stress, respectively. (B) At 1 h, osteoblasts, as well as fibroblastic cells, were
flattened and elongated. The osteoid and ALPase positive preosteoblast (slant arrow) layer also were extended toward the
center of the suture. (C) At 6 h, the flattened ALPase positive cell (slant arrows) layer further extended toward the center
and some osteoid is formed. (D) At 24 h, the osteoid (*) and preosteoblast (slant arrows) layer further increased in length.
(E) At 48 h, the osteoid thickened, part of which became calcified (arrowheads). The center of the suture is indicated by
a perpendicular arrow (original magnification ϫ280).
dependent elongation of the suture in which most of the sion of osteocalcin and osteopontin was confirmed in these
cells became flattened and elongated (Figs. 4B–4E). The newly recruited osteoblasts (M. Ikegame, O. Ishibashi, T.
osteoid covered with ALPase positive cells extended toward Yoshizawa, H. Ozawa, and H. Kawashima, unpublished
the median line of the suture at 1 h after the start of data, 1999). In contrast to these calvarial sutures, those
mechanical loading (Fig. 4B). At 6 h, the flattened ALPase without tensile stress showed practically no significant
positive cell layer extended further toward the center (Fig. changes in histological appearance, osteoblast number, and
4C), although most parts of the extended cell layer were osteoid area from time 0–24 h incubation. However, at 48 h,
without osteoid and were thinner compared with the previ- a minimal increase was observed in the thickness of the
ously existing layer with osteoid (Figs. 4C and 4B), indi- bone edge in which part of the osteoid had been calcified
cating that these ALPase positive cells are preosteoblasts. without mechanical loading (data not shown). ALPase ac-
This also was supported by the fact that they were negative tivity was detected in fibroblastic cells surrounding cartilage
with osteocalcin and osteopontin (data not shown), both of and along the dura mater, but these activities did not change
which are typical markers of osteoblasts. By contrast, ex- with mechanical loading (Figs. 4C and 4D). From these
pression of osteocalcin and osteopontin was detected in the observations, it is possible that a portion of the newly
mature osteoblasts on the mineralized bone (M. Ikegame, O. recruited osteoblasts due to tensile stress were differentiated
Ishibashi, T. Yoshizawa, H. Ozawa, and H. Kawashima, from fibroblastic cells.
unpublished data, 1999). At 24 h the extended layer of
ALPase positive cells thickened and some new osteoid was
added (Fig. 4D). At 48 h the ALPase positive cell layer
increased further in length and thickness, which was accom-
Tensile stress up-regulates BMP-4 expression in
calvariae
panied by an increase of osteoid (Fig. 4E). Scattered newly
To identify key genes responsible for the differentiation
calcified areas also were evident in the increased osteoid as and the growth of osteoblasts induced by tensile stress,
evidenced with von Kossa staining (Fig. 4E). Gene expres- calvarial sutures were dissected under a dissecting micro-

TENSILE STRESS INDUCES BMP-4–MEDIATED OSTEOGENESIS
29
tensile stress. Figure 6I illustrates exactly from where Figs.
6A–6F were taken. A schematic illustration of the profiles
of Figs. 6A–6F also is given in Fig. 7. BMP-4 expression
was detected in osteoblasts surrounding the bone tissues and
osteoid, in preosteoblasts and in some of fibroblastic cells
near the periosteal surface and the central region of the
suture in the control (time 0; Figs. 6B and 7). On the other
hand, Cbfa1/Osf-2 expression was localized specifically to
osteoblasts and preosteoblasts in the control (Figs. 6A and
7). At 3 h under tensile stress BMP-4 expression was
increased in preosteoblastic cells located in front of osteo-
blasts and was induced in the adjacent spindle-shaped fi-
broblastic cells near the central region of the suture (Figs.
6D and 7). In contrast, the expression of Cbfa1/Osf-2 did
not change significantly (Figs. 6C and 7). At 6 h both
preosteoblastic and fibroblastic cells with positive BMP-4
expression further increased in number as the ALPase pos-
itive cell layer and osteoid extended (Figs. 4C, 6F, and 7).
Cbfa1/Osf-2 expression also was induced in the newly
recruited preosteoblasts (Figs. 6E and 7), which were likely
to be the fibroblastic cells with positive BMP-4 expression
at 3 h (Figs. 6D and 7). To ensure that the expression of
BMP-4 precedes the Cbfa1 expression, we counted the
number of cells that express BMP-4 or Cbfa1 in four dif-
ferent sets of plates including those shown in Figs. 6A–6F.
The actual counting was carried out as follows. The back-
ground level was set to the strongest staining on the plate
incubated with sense probe, and subtraction was done on a
computer screen using the National Institutes of Health
(NIH) image 1.60. Then the remaining positive cells were
counted. As shown in Fig. 6J, BMP-4 expression slightly
increased at 3 h and significantly elevated at 6 h, whereas
Cbfa1 expression showed a tendency to increase at 6 h but
did not reach a significant level.
FIG. 5. Tensile stress up-regulated the BMP-4 gene ex-
pression in calvariae. Total RNA was extracted from cul-
tured calvariae at each time point indicated and was sub-
jected to RT-PCR. Each column and bar represents a mean
of three different experiments and SD. *Significantly dif-
ferent from the control at p Ͻ 0.05. See text for details.
scope (Fig. 3). Cryosections of the dissected sutures con-
firmed that the tissues contain only fibrous matrix and
osteoid and are devoid of calcified matrix (data not shown).
The sutures were then subjected to RNA extraction fol-
lowed by RT-PCR.
DISCUSSION
The present study shows that mechanical stress induces
osteoblast differentiation, which then leads to osteogenesis
in mouse calvarial sutures in culture. This osteoblast differ-
entiation seemed to be accompanied by an increase in
BMP-4 gene expression when evaluated by RT-PCR of the
cultured suture. Our study also showed that calvarial sutures
express the BMP-4 gene even without mechanical loading.
Ever since molecular cloning and characterization of the
BMP gene family and its products, one of the major con-
cerns in the field of bone mineral research has been whether
or not one or more members of the BMP gene family play
a role in modeling or remodeling bone in neonatal and adult
life. It is well established that BMP family members play
critical roles in patterning and developing of skeletal tissues
during embryogenesis.(28,29) Site-specific expression of
gene family members also has been shown in the develop-
ing bone tissues in the embryo.(30) However, many studies
failed to show BMP gene expression in adult bone tissues.
The only exception was that the BMP-4 gene was expressed
in tissues surrounding healing bone after artificial bone
fracture and it disappeared as healing progressed.(31) Our
Up to 6 h after the start of tensile stress, no change in the
expression of either ALPase, type I collagen, or osteocalcin
gene was observed (data not shown). In contrast, PCR
products corresponding to BMP-4 slightly increased at 3 h,
although this increase was not significant. The expression
significantly increased at 6 h reaching approximately 3.5-
fold of the control (Fig. 5) and this level of BMP-4 expres-
sion was maintained until 24 h. In another set of experi-
ments a similarly elevated expression of BMP-4 gene was
observed at 48 h (data not shown). Because the tissue
preparation contains heterogenous types of cell (Fig. 4), it is
impossible to identify cells in which the BMP-4 expression
is up-regulated. To address this question we next examined
gene expression in individual cells using an in situ hybrid-
ization technique.
Tensile stress sequentially up-regulates BMP-4 and
Cbfa1/Osf-2 expression in preosteoblastic and
fibroblastic cells in the calvariae
Figure 6 depicts the expression profiles of BMP-4 and RT-PCR analysis thus provides the first evidence that nor-
Cbfa1/Osf-2 messenger RNAs (mRNAs) in calvariae under mal bone tissues from neonatal animals express the BMP-4

30
IKEGAME ET AL.
FIG. 6. Sequential up-regulation/induction of BMP-4 and Cbfa1/Osf-2 gene expression preceded osteoblast differenti-
ation in calvariae under tensile stress. (A and B) Time 0; (C and D) 3 h; (E–H) 6 h. (A, C, and E) Antisense Cbfa1/Osf-2
probe; (G) sense Cbfa1/Osf-2 probe; (B, D, and F) antisense BMP-4 probe; (H) sense BMP-4 probe (B, parietal bone; S,
suture; POB, preosteoblast; FB, fibroblast; arrows, center of the sutures). (I) Schematic overview showing where panels
A–H were taken from. (J) Changes in cell numbers that are positive for BMP-4 or Cbfa1 gene expression. At time 0, BMP-4
expression was detected in osteoblasts surrounding the osteoid and in preosteoblasts and in some of fibroblastic cells (B).
On the other hand, Cbfa1/Osf-2 expression was limited to osteoblasts and preosteoblasts (A). At 3 h, BMP-4 expression
was induced in preosteoblastic cells and their adjacent spindle-shaped cells (D, arrowheads). Expression of Cbfa1/Osf-2
gene did not significantly change from that at time 0 (C). At 6 h although BMP-4 expression was diminished in some
osteoblasts, it was elevated in preosteoblasts and spindle-shaped fibroblastic cells (F, arrowheads). Expression of Cbfa1/
Osf-2 also was induced in preosteoblasts and spindle-shaped fibroblastic cells where BMP-4 expression became evident at
3 h (E, arrowheads). The signals of Cbfa1/Osf-2 and BMP-4 disappeared when corresponding sense probes were used (G
and H; original magnification ϫ350). The details as to where pictures of A–H were taken are schematically illustrated by
frames in I. Each arrow in the frame indicates the center of the suture. Cell numbers positive for BMP-4 gene expression
or for Cbfa1 gene expression were counted in four sets of plates including those shown in A–F and statistically analyzed
in J. The counting was carried out for each sample in the area between the right end of the POB layer and the center of
the suture. **Significantly different from control (0 g) at p Ͻ 0.01. POB, preosteoblast.

TENSILE STRESS INDUCES BMP-4–MEDIATED OSTEOGENESIS
31
expression were induced during distraction osteogenesis in
rats.(32) Our data are consistent with this report, although we
could not detect up-regulation of BMP-2 expression by
RT-PCR. The time frame of the BMP-4 expression in the
present study and that in the distraction study cited previ-
ously are different; in our study BMP-4 gene expression
began to rise at 6 h under tensile stress and remained
elevated until 48 h, whereas in the distraction study the gene
expression was slightly elevated at 4 days after osteotomy
and it declined at 7 days. By this time the soft external callus
and the periosteal as well as endosteal bony callus were
formed. Distraction caused separation of those calluses,
between which a fibrous interzone was formed. The carti-
lage callus at the mouth of osteotomy site was resorbed
gradually and replaced by new bone through endochondral
ossification. The BMP-4 expression was elevated at this
stage (10 days after distraction started) and the expression
remained elevated throughout the distraction phase. In con-
trast to the osteotomy followed by distraction, application of
tensile stress in the present study does not involve any
FIG. 7. Schematic illustration of events observed in Fig. trauma and it does not cause any serious damage to sutures.
6. OB, osteoblast; POB, preosteoblast; FB, fibroblast.
In addition, osteogenesis in the calvaria is not endochondral
but membranous ossification. Therefore, times required for
the gene expression and osteogenesis in these two systems
gene. However, because the suture contained a heteroge- may not be the same. Nonetheless, BMP-4 expression ap-
nous cell population, we next tried to identify cell types that pears to play an important role in both membranous and
express this gene using an in situ hybridization technique. endochondral ossification induced by tensile stress. Further
As shown in Fig. 6B, BMP-4 gene expression was de- study is needed for a clear understanding of the exact role of
tected in osteoblasts surrounding osteoid and in their adja- BMP-4 in these systems.
cent preosteoblastic and spindle-shaped fibroblastic cells.
In conclusion, we showed that tensile stress induces os-
The expression was increased by tensile stress in preosteo- teoblast differentiation and osteogenesis in mouse calvarial
blastic and fibroblastic cells, whereas it remained un- sutures in culture and also induces BMP-4 expression in
changed or diminished in osteoblasts covering the existing preosteoblasts and fibroblasts, which appeared to differen-
bone tissue. Cell number with elevated BMP-4 expression tiate eventually into osteoblasts. BMP-4 may act as an
in both cell types further increased as time of mechanical autocrine and paracrine factor in tensile stress–induced os-
loading extended. Cbfa1/Osf-2 is an osteoblast-specific teogenesis.
transcription factor and under the control of the BMP gene
family. The expression of this gene followed the BMP
expression under mechanical loading (Figs. 6 and 7). This
ACKNOWLEDGMENTS
sequence of events is consistent with the fact that osteoblast
differentiation and osteogenesis were observed in Fig. 6.
Based on these observations, we propose that tensile
stress increases BMP-4 expression in preosteoblasts and
fibroblasts, and BMP-4, in turn, activates Cbfa1/Osf-2 ex-
pression, thereby inducing osteoblast differentiation. Thus,
BMP-4 may act as an autocrine and paracrine factor in the
osteogenesis induced by tensile stress. Fibroblastic cells that
became BMP-4 positive do not necessarily differentiate into
osteoblasts. Cells that can differentiate into osteoblasts ap-
peared to be limited to the spindle-shaped cells that lined
along the extension of existing bone tissue in the direction
of the center of the suture, while BMP-4 positive fibroblas-
tic cells at the periosteal surface did not appear to change
their fate. Therefore, there must be determinants other than
BMP-4 that control the fate of cells to differentiate into
osteoblasts. Thus, the exact mechanism by which tensile
stress induces BMP-4 expression and how BMP-4 stimu-
lates osteogenesis are subjects of future study.
This work was supported by a Grant-in Aid for Scientific
Research from the Ministry of Education, Science and Cul-
ture of Japan, a grant from the “Ground Research for Space
Utilization” promoted by National Space Development
Agency of Japan and Japan Space Forum, and grants from
The Foundation for Total Health Promotion; The Nakatomi
Foundation; Glaxo Wellcome Co., Ltd., Japan; Hoechst
Marion Roussel Ltd., Japan; and Yamanouchi Co. Ltd.,
Japan.
REFERENCES
1. Sandy JR, Meghji S, Farndale RW, Meikle MC 1989 Dual
elevation of cyclic AMP and inositol phosphates in response to
mechanical deformation of murine osteoblasts. Biochim Bio-
phys Acta 1010:265–269.
2. Binderman I, Zor U, Kaye AM, Shimshoni Z, Harrel A,
Somjen D 1988 The transduction of mechanical force into
biochemical events in bone cells may involve activation of
phospholipase A2. Calcif Tissue Int 42:261–266.
While we were preparing this article, a paper was pub-
lished showing that BMP-2 and BMP-4 but not BMP-6,
BMP-7, and growth and differentiation factor-5 (GDF-5)
3. Jones DB, Nolte H, Scholubbers JG, Turner E, Veltel D 1991
Biochemical signal transduction of mechanical strain in
osteoblast-like cells. Biomaterials 12:101–110.

32
IKEGAME ET AL.
4. Sandy JR, Meikle MC, Farndale RW 1993 Pertussis toxin- 19. Raab-Cullen DM, Thiede MA, Peterson DN, Kimmel DB,
sensitive activation of phospholipase A2 can be resolved from
phosphoinositidase C in primary cultures of mouse osteoblasts
using indomethacin. J Bone Miner Res 8:543–552.
Recker RR 1994 Mechanical loading stimulates rapid changes
in periosteal gene expression. Calcif Tissue Int 55:473–478.
20. Dolce C, Kinniburgh AJ, Dziak R 1996 Immediate early-gene
identification in rat osteoblastic cells after mechanical defor-
mation. Arch Oral Biol 41:1101–1108.
5. Klein-Nulend J, Burger EH, Semeins CM, Raisz LZ, Pilbeam
CC 1997 Pulsating fluid flow stimulates prostaglandin release
and inducible prostaglandinG/H synthase mRNA expression in 21. Ogata T 1997 Fluid flow induces enhancement of the Egr-1
primary mouse bone cells. J Bone Miner Res 12:45–51.
mRNA level in osteoblast-like cells: Involvement of tyrosine
6. Smalt R, Mitchell FT, Howard RL, Chambers TJ 1997 Induc-
kinase and serum J Cell Physiol 170:27–34.
tion of NO and prostaglandin E2 in osteoblasts by wall-shear 22. Ochoa JA, Sanders AP, Heck DA, Hillberry BM 1991 Stiff-
stress but not mechanical strain. Am J Physiol 273:E751–
ening of the femoral head due to inter-trabecular fluid and
E758.
intraosseous pressure. J Biomech Eng 113:259–262.
7. Neidlinger-Wilke C, Stalla I, Claes L, Brand R, Hoellen I, 23. Mikuni-Takagaki Y, Suzuki Y, Kawase T, Saito S Distinct
Rubenacker S, Adam M, Kinzl L 1995 Human osteoblasts
from younger normal and osteoporotic donors show differ-
responses of different populations of bone cells to mechanical
stress. Endocrinology 137:2028–2035.
ences in proliferation and TGF beta-release in response to 24. Hickory WB, Nanda R 1987 Effect of tensile force magnitude
cyclic strain. J Biomech 28:1411–1418.
on release of cranial suture cells into S phase. Am J Orthod
8. Klein-Nulend J, van der Plas A, Semiens CM, Ajubi NE,
Dentofacial Orthop 91:328–334.
Frangos JA, Nijweide PJ, Burger EH 1995 Sensitivity of 25. Burstone MS 1962 Alkaline phosphatase, naphthol AS-BI
osteocytes to biomechanical stress in vitro. FASEB J 9:441–
445.
9. Rawlinson SC, Pitsillides AA, Lanyon LE 1996 Involvement
phosphate method. In: Enzyme Histochemistry and Its Appli-
cation to the Study of Neoplasms. Burstone MS (ed.) Aca-
demic Press, New York, NY, USA, pp. 275–276.
of different ion channels in osteoblasts’ and osteocytes’ early 26. Xiao ZS, Thomas R, Hinson TK, Quarles LD 1998 Genomic
responses to mechanical strain. Bone 19:609–614.
structure and isoform expression of the mouse, rat and human
10. Johnson DL, McAllister TN, Frangos JA 1996 Fluid flow
Cbfa1/Osf2 transcription factor. Gene 214:187–197.
stimulates rapid and continuous release of nitric oxide in 27. Young RW 1959 The influence of cranial contexts on postnatal
osteoblasts. Am J Physiol 271:E205–E208. growth of the skull in the rat. Am J Anat 105:383–416.
11. Pitsillides AA, Rawlinson SC, Suswillo RF, Bourrin S, Zaman 28. Kingsley DM 1994 What do BMPs do in mammals? Clues
G, Lanyon LE 1995 Mechanical strain-induced NO production from the mouse short-ear mutation. Trends Genet 10:16–21.
by bone cells: A possible role in adaptive bone (re)modeling. 29. Matzuk MM 1995 Functional analysis of mammalian mem-
FASEB J 9:1614–1622.
12. Klein-Nulend J, Semeins CM, Ajubi NE, Nijweide PJ, Burger
bers of the transforming growth factor-␤ superfamily. Trends
Endocrinol Metab 6:120–127.
EH 1995 Pulsating fluid flow increases nitric oxide (NO) 30. Wozney JM 1993 Bone morphogenetic proteins and their gene
synthesis by osteocytes, but not periosteal fibroblasts–
correlation with prostaglandin upregulation. Biochem Biophys
Res Commun 217:640–648.
expression. In: Cellular and Molecular Biology of Bone. Noda
M (ed.) Academic Press, New York, NY, USA, pp. 131–167.
31. Nakase T, Nomura S, Yoshikawa H, Hashimoto J, Hirota S,
Kitamura Y, Oikawa S, Ono K, Takaoka K 1994 Transient and
localized expression of bone morphogenetic protein 4 messen-
ger RNA during fracture healing. J Bone Miner Res 9:651–
659.
13. Duncan RL, Kizer N, Barry EL, Friedman PA, Hruska KA
1996 Antisense oligo-deoxynucleotide inhibition of
a
swelling-activated cation channel in osteoblast-like osteosar-
coma cells. Proc Natl Acad Sci USA 93:1864–1869.
14. Wiltink A, Nijweide PJ, Scheenen WJ, Ypey DL, Van Duijn B 32. Sato M, Ochi T, Nakase T, Hirota S, Kitamura Y, Nomura S,
1995 Cell membrane stretch in osteoclasts triggers a self-
reinforcing Ca2ϩ entry pathway. Pflugers Arch 429:663–671.
15. Xia SL, Ferrier J 1992 Propagation of a calcium pulse between
osteoblastic cells. Biochem Biophys Res Commun 186:1212–
1219.
Yasui N 1999 Mechanical tension-stress induces expression of
bone morphogenetic protein (BMP)-2 and BMP-4, but not
BMP-6, BMP-7, and GDF-5 mRNA, during distraction osteo-
genesis. J Bone Miner Res 14:1084–1095.
16. McDonald F, Somasundaram B, McCann TJ, Mason WT,
Meikle MC 1996 Calcium waves in fluid flow stimulated
osteoblasts are G protein mediated. Arch Biochem Biophys
326:31–38.
17. Glanstchnig H, Varga F, Rumper M, Klaushofer K 1996
Prostacyclin (PGI2): A potential mediator of c-fos expression
induced by hydrostatic pressure in osteoblastic cells. Eur J Clin
Invest 26:544–548.
Address reprint requests to:
Hiroyuki Kawashima, Ph.D.
Department of Pharmacology
Faculty of Dentistry
5274 2-Bancho, Gakkocho-dori
Niigata, Niigata 951-8514, Japan
18. Jagger CJ, Chow JW, Chambers TJ 1996 Estrogen suppresses
activation but enhances formation phase of osteogenic re-
sponse to mechanical stimulation in rat bone. J Clin Invest Received in original form December 3, 1999; in revised form July
98:2351–2357. 3, 2000; accepted August 24, 2000.