www.elsevier.com/locate/bone

Bone 34 (2004) 940–948

Prostaglandin E2 receptor (EP4) selective agonist (ONO-4819.CD)

accelerates bone repair of femoral cortex after drill-hole injury associated

with local upregulation of bone turnover in mature rats

Masahiro Tanaka,a Akinori Sakai,a,* Soshi Uchida,a Shinya Tanaka,a Masato Nagashima,a

Teruaki Katayama,b Kojiro Yamaguchi,b and Toshitaka Nakamuraa

aDepartment of Orthopaedic Surgery, School of Medicine, University of Occupational and Environmental Health, Yahatanishi, Kitakyushu, JapanbFukui Safety Research Institute, Ono Pharmaceutical Company, Osaka, Japan

Abstract

Prostaglandin E2 (PGE2) is essential for fracture healing. Systemic administration of EP4 ligands such as PGE2 and other synthetic EP4

agonists appears to transduce anabolic signals by binding to receptor EP4. Therefore, the present study was designed to test whether

administration of EP4 agonist accelerates the healing of drill-hole injury in the femoral diaphysis. After surgery, a total of 128 Wistar rats, at

the age of 12 weeks, were assigned to basal control (n = 8), and three groups with respective doses of 0 (vehicle control), 10 (low-dose), and

30 (high-dose) Ag/kg body weight of the agent were subcutaneously injected twice a day. Femoral specimens were obtained at 0, 5, 7, 14, 21,

and 28 days. In EP4 agonist-treated groups, the total bone volume of the regenerating bone in the defect did not significantly differ, but the

regenerated cortical bone volume measured by histomorphometry and cortical bone mineral content (Ct. BMC) by pQCT dose-dependently

increased at 14 and 21 days compared to the control. In the high-dose group, the value of osteoclast surface significantly increased compared

with that in the control at 14 days. Expression levels of osteocalcin and TRAP mRNAs in the injured tissue increased at 14 days. Expression

levels of EP4, BMP-2, and RANKL mRNAs increased at 7 days in the high-dose group. The bone mineral values of the lumbar bone at 28

days, measured by DXA, did not differ in the three groups. These data indicated that systemic administration of EP4 agonist ONO-4819.CD

accelerated cortical bone healing after drill-hole injury by upregulating the local turnover of the regenerating bone.

D 2004 Elsevier Inc. All rights reserved.

Keywords: Histomorphometry; pQCT; Osteoclast; Osteocalcin; RANKL; TRAP

Introduction homozygous for a null mutation in the COX-2 gene [27].

Prostaglandin E2 (PGE2) is a potent agent that stimulates

bone turnover. In rats, systemic administration of PGE2

increased both bone formation and resorption, and conse-

quently led to an increase in bone mass [6]. It also

increased the extent of mineralization and remodeling

activities in a site adjacent to fractured rib in dogs

[16,17]. Fracture healing failed in rats that received a

selective cyclooxygenase-2 (COX-2) inhibitor, which re-

duced endogenous PGE2 production [18], and in mice

8756-3282/$ - see front matter D 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.bone.2004.01.002

* Corresponding author. Department of Orthopaedic Surgery, School of

Medicine, University of Occupational and Environmental Health, 1-1

Iseigaoka, Yahatanishi, Kitakyushu 807-8555, Japan. Fax: +81-93-692-

0184.

E-mail address: a-sakai@med.uoeh-u.ac.jp (A. Sakai).

Local administration of PGE2 in bone undergoing osteot-

omy caused a dose-dependent increase in callus formation

in rabbits [7]. These data indicate that the action of PGE2 is

essential for fracture healing by inducing osteogenesis in

injured bone.

Of the four subtypes of PGE2 receptors (EP1, EP2,

EP3, and EP4) belonging to the G-protein-coupled receptor

family, EP4 is considered the most potent receptor that

mediates the anabolic action in bone. Systemic adminis-

tration of an EP4 antagonist suppressed the increase in

bone mass induced by PGE2 [9]. Subcutaneous injection of

EP4 agonist ONO-4819.CD at the dose of 10 or 30 Ag/kgthree times per day also prevented trabecular bone loss

induced by ovariectomy and neurectomy in rats [26]. Local

administration of PGE2 into the midfemur periosteum did

not induce callus formation after periosteal injury in mice

lacking EP4, although callus was formed in mice lacking

M. Tanaka et al. / Bone 34 (2004) 940–948 941

EP1, EP2, and EP3, respectively [26]. Local infusion of

the EP4-selective agonist AE1-329 increased callus forma-

tion in mice [26]. Thus, both systemic and local admin-

istrations of EP4 ligands such as PGE2 and other synthetic

EP4 agonists appear to transduce bone forming signals by

binding to the cell-surface receptor EP4. On the other

hand, it was recently reported that an EP2 selective agonist

CP-533,536 has the ability to heal canine long bone

segmental and fracture model defects [14]. PGE2 enhances

its own production in vitro via upregulation of COX-2

expression in osteoblasts, thus amplifying its own effects

[15]. Expression of EP4 and its regulation by PGE2 has

been observed in osteoblastic cell lines, young rat bone

tissue, and bone marrow osteogenic cells [23–25]. While

the expression of EP4 was minimal in tibia of normal 6-

week-old rats, it was upregulated in diaphyseal and meta-

physeal bone marrow cells after a single injection of PGE2

[25]. However, the effect of these agents on the local

turnover of regenerating bone has not been studied.

We hypothesized that new bone formation could be

enhanced by administration of EP4 agonist through local

upregulation of both osteogenic and osteoclastogenic sig-

nals in injured bone tissue. We tested whether systemic

administration of the EP4 agonist ONO-4819.CD acceler-

ates the regeneration of a cortical bone defect by upregulat-

ing bone turnover in the rat femur. We performed serial

histological and histomorphometrical assessments, bone

mineral measurements by peripheral quantitative computed

tomography (pQCT), and evaluated local mRNA expres-

sion. We also measured systemic bone markers and lumbar

bone mineral by dual energy X-ray absorptiometry (DXA)

in the rats.

Materials and methods

Experimental animals

A total of 128 male Wistar rats weighing 377 F 13 g

(mean F SEM), aged 11 weeks, were purchased from

Table 1

RT-PCR primers used in this study

Gene Primer location Seq

EP4 1086–1105 5V-A1255–1236 5V-G

BMP-2 920–939 5V-C1134–1115 5V-C

Osteocalcin 139–158 5V-T350–331 5V-C

RANKL 15–34 5V-C220–201 5V-G

TRAP 1086–1105 5V-A1299–1280 5V-T

h-actin 451–471 5V-T1089–1069 5V-A

Charles River Japan (Shiga, Japan) and acclimatized for a

week before the beginning of the experiment.

Experimental design

At 12 weeks of age, a hole of approximately 2.0 mm

in diameter penetrating the cortical bone and bone marrow

was made by an electric drill at the anterior portion of the

diaphysis of bilateral femurs, 16 mm above the knee joint

[21]. Rats were then randomized by body weight to 16

groups of eight animals each. One group (baseline con-

trol) was sacrificed at day 0. From the next day after

surgery (day 1), the remaining 15 groups were divided

into three dosing groups, treated with a subcutaneous

injection of the EP4-selective agonist ONO-4819.CD,

methyl 7-[(1R,2R,3R)-3-hydroxy-2-[(E)-(3S)-3-hydroxy-4-

(m-methoxymethylphenyl)-1-butenyl]-5-oxocyclopentyl]-5-

thiaheptanoate (Ono Pharmaceutical Co., Osaka, Japan) in

the back skinfold twice a day at the respective dose of 0

(vehicle control group), 10 (low-dose group), and 30

(high-dose group) Ag/kg BW [26]. Rats from one group

each from vehicle control, low-dose, and high-dose

groups were sacrificed at 5, 7, 14, 21, and 28 days after

surgery. After sacrifice, femoral samples were dissected

out bilaterally; the left femur was immediately stored at

�80jC until use for bone mineral measurements, and the

right femur was processed for histology and mRNA

assessments. The fourth lumbar vertebrae (L4) were also

obtained at 28 days. The rats were weighed every week.

The care and use of animals followed The Guiding

Principles for the Care and Use of Animals, approved

by our university in accordance with the principles of the

Declaration of Helsinki.

Bone metabolic markers

Urinary deoxypyridinoline (DPD) cross-link excretion

levels were measured using an ELISA kit (Pyrilinks-D,

Metra Biosystems, Inc., Mountain View, CA), and the

results were expressed as DPD/creatinine (Cr) ratios

uence Product

length (bp)

GCACAGCACTGCTCAGAGA-3V 170

CGTACCTGGAAGCAAATTC-3VGGAAGCGTCTTAAGTCCAG-3V 215

ATGCCTTAGGGATTTTGGA-3VGACAAAGCCTTCATGTCCA-3V 212

CTAAACGGTGGTGCCATAG-3VCGAGACTACGGCAAGTACC-3V 206

CGCTCGAAAGTACAGGAAC-3VGGGTCCTGCTTATCCCCTA-3V 214

ACCCCAAAACCACAGGGTA-3VTGAGACCTTCAACACCCCAG-3V 639

CTTGCGCTCAGGAGGAGCAA-3V

Table 2

Serial changes in levels of bone metabolic markers

Treatment group 0 days 5 days 7 days 14 days 21 days 28 days

Deoxypyridinoline vehicle control 161.7 F 22.6 160.8 F 17.0 180.5 F 12.0 136.7 F 15.5 78.5 F 9.1 104.9 F 22.4

(nmol/mmol Cr) low-dose 194.0 F 27.4 160.8 F 19.8 122.9 F 14.4 108.1 F 22.9 119.0 F 14.8

high-dose 165.0 F 13.5 131.9 F 6.8 130.3 F 13.1 81.5 F 7.0 89.4 F 10.8

Gla-osteocalcin vehicle control 2708 F 136 3091 F 93 3163 F 176 2304 F 87 3073 F 263 2660 F 172

(ng/ml) low-dose 2928 F 108 3530 F 107 2103 F 153 2696 F 170 2803 F 126

high-dose 2992 F 277 3120 F 82 3111 F 188*** 2449 F 165** 2595 F 117

Data are presented as mean F SEM.

**P < 0.01 vs. vehicle control group at the same time point (Fisher’s PLSD test after ANOVA).

***P < 0.001 vs. vehicle control group at the same time point (Fisher’s PLSD test after ANOVA).

M. Tanaka et al. / Bone 34 (2004) 940–948942

(nmol/l/mmol/l Cr). Serum osteocalcin concentrations were

determined using an enzyme-linked immunosorbent assay

(ELISA) kit with anti-rat Gla-osteocalcin antibody (Rat

Gla-OC Competitive EIA Kit; Takara Bio, Inc., Shiga,

Japan).

Fig. 1. Histological sagittal sections of the injured site of the femur, stained with

group at 5 days, (B) at 7 days or (C) at 28 days. (D-1) Vehicle control group, (D-

group, (E-2) low-dose group, or (E-3) high-dose group at 21 days.

Bone mineral and area of L4 by DXA

L4 body samples were isolated removing the posterior

elements and transverse processes. The values of bone

mineral content (BMC), bone mineral density (BMD), and

hematoxylin and eosin (original magnification � 40). (A) Vehicle control

2) low-dose group or (E) high-dose group at 14 days. (E-1) Vehicle control

M. Tanaka et al. / Bone 34 (2004) 940–948 943

bone mineral area (BMA) were measured in each specimen

by DXA (DCS-600; Aloka, Tokyo) using the small animal

scan mode with irradiation applied anteroposteriorly to the

specimen [13]. The values of BMC (mg), BMD (mg/cm2),

and BMA (cm2) were obtained for each specimen.

Cross-sectional bone mineral and area at the drilled site of

the femur by pQCT

We used pQCT analyses to discriminate the regenerat-

ing bone tissue with the amounts of bone mineral equiv-

alent to trabecular bone and those equivalent to cortical

bone in the whole cross-sectional drilled region. Cross-

sectional scans were made at the drilled sites in the left

Fig. 2. Serial changes in histomorphometry parameters at the injured site. (A) Re

bone volume (Med. BV/TV; %). (C) Regenerating cortical bone volume (Ct. BV

surface/bone surface (Ob.S/BS; %). Data are presented as mean F SEM. *P < 0.0

ANOVA).

femur samples using pQCT (XCT Research SA; Norland

Stratec Medizintechnik, Birkenfeld, Germany). For each

specimen, three consecutive cross-sectional scans were

performed in a 0.5-mm-thick slice with a voxel size of

0.12 � 0.12 � 0.5 mm3. The image representing a mid-

cross-section of the drilled site was chosen for analysis. An

attenuation threshold value of 690 mg/cm3 was used to

define cortical bone, and that of 267 mg/cm3 was used to

define total bone. The difference between the former and

the latter was defined as trabecular bone. The measured

parameters were cortical bone mineral content (Ct. BMC),

cortical bone mineral area (Ct. BMA), trabecular bone

mineral content (Tr. BMC), and trabecular bone mineral

area (Tr. BMA) [5].

generating total bone volume (T. BV/TV; %). (B) Regenerating medullary

/TV; %). (D) Osteoclast surface/bone surface (Oc.S/BS; %). (E) Osteoblast

5 vs. vehicle control group at the same time point (Fisher’s PLSD test after

one 34 (2004) 940–948

Histology and histomorphometry

Five of the right femur samples from each group

obtained at each time point were decalcified and embed-

ded in paraffin. The specimens were cut along the long

axis of the femur including the midregion of the drilled

site at the anterior cortex. Five-micrometer-thick sections,

thus obtained, were then stained for tartrate-resistant acid

phosphatase (TRAP), hematoxylin-eosin, respectively [11].

The sections were examined under a light microscope.

For histomorphometry of newly formed bone tissue, we

measured the regenerating bone area at the drilled site at

100-fold magnification. Histomorphometry was performed

with a semiautomatic image-analysis system linked to the

light microscope (Cosmozone 1S, Nikon, Tokyo) [11]. We

measured the total bone area (T. BV/TV; %) in the defect

made by drilling. Then, we measured the medullary bone

(Med. BV/TV; %) in the bone marrow and cortical bone

volume (Ct. BV/TV; %) in the drilled cortical tissue area

separately. We also measured osteoclast surface/bone

M. Tanaka et al. / B944

Fig. 3. Serial changes in cross-sectional bone mineral and area at the injured site

BMC). (B) Cortical bone mineral area (Ct. BMA). (C) Trabecular bone mineral

presented as mean F SEM. *P < 0.05, **P < 0.01, ***P < 0.001 vs. vehicle co

surface (Oc.S/BS; %), and osteoblast surface/bone surface

(Ob.S/BS; %) in the regenerating bone in the total bone

area.

Quantitative real-time reverse-transcriptase-polymerase

chain reactions (RT-PCR)

The wound regions of three right femur specimens from

each group obtained at each time point were excised, and then

frozen and powdered in liquid nitrogen using a frozen cell

crusher. Total RNAwas extracted using an acid guanidinium

thiocyanate–phenol–chloroform method after homogeniz-

ing [2]. The isolated RNA was then cleaned up using the

RNeasy kit (Qiagen, Hilden, Germany). First-strand cDNA

was reverse-transcribed from total RNA (1 Ag) using Molo-

ney murine leukemia reverse transcriptase (SuperScript; Life

Technologies, Inc., Rockville, MD) and oligo(dT) 12–18

primer (Life Technologies). Quantitative PCR analysis was

performed using an iCycler apparatus (Bio-Rad laboratories,

Hercules, CA) associated with the iCycler Optical System

of the femur measured by pQCT. (A) Cortical bone mineral content (Ct.

content (Tr. BMC). (D) Trabecular bone mineral area (Tr. BMA). Data are

ntrol group at the same time point (Fisher’s PLSD test after ANOVA).

M. Tanaka et al. / Bone 34 (2004) 940–948 945

Interface software (version 3.0; Bio-Rad). The quantitative

PCRs for EP4, BMP-2, osteocalcin, receptor activator of

nuclear factor-nB ligand (RANKL), TRAP, and h-actin wereperformed in 20 Al with approximately 7.5 ng cDNA, 0.5 pM

primers, and 10Al iQ SYBR Green Supermix (Bio-Rad)

containing 0.4 mM of each dNTP (dATP, dCTP, dGTP, and

dTTP), iTaq DNA polymerase, and SYBR Green l dye. The

sequences of primers used in this study are shown in Table 1.

These primers were designed by Primer 3 software (White-

head Institute/MIT Center for Genome Research, Cambridge,

MA) and synthesized at the Sigma Genosys Japan Company

Fig. 4. Serial changes in relative expression levels of mRNA measured by quantitat

Data are presented as mean F SEM. *P < 0.05, **P < 0.01, ***P < 0.001 vs. vehi

(Hokkaido, Japan). h-actin was used as an internal control.

The amplification conditions were an initial 3 min at 95jC,and then 40–50 cycles of denaturation at 95jC for 30 s,

annealing at 60jC for 30 s, and extension at 72jC for 30 s.

Only h-actin was annealed at 65jC for 30 s. All PCR

reactions were performed in triplicate.

Statistical analyses

Results are expressed as the mean F SEM. Statistical

analysis was performed using Fisher’s PLSD test after

ive RT-PCR. (A) EP4, (B) BMP-2, (C) RANKL, (D) osteocalcin, (E) TRAP.

cle control group at the same time point (Fisher’s PLSD test after ANOVA).

M. Tanaka et al. / Bone 34 (2004) 940–948946

ANOVA at each time point. P values less than 0.05 were

considered significant. Calculations were performed with a

StatView 4.5 J program (Abacus Concepts, Inc., Berkeley,

CA) on a Macintosh computer.

Results

Body weight and bone metabolic markers

Body weight was not significantly different among the

groups. The body weight at the end of the study were 481

F 11, 466 F 7, and 473 F 11 g for vehicle-control, low-

dose, and high-dose groups, respectively. The mean uri-

nary DPD concentration was not significantly different

among the three groups. The serum Gla-osteocalcin levels

fluctuated (Table 2).

Bone mineral values of L4 by DXA

The values of L4 BMC, BMD, and BMA at 28 days were

not significantly different among the three groups (data not

shown).

Histology and histomorphometry at the injured site

In all groups, regenerating medullary bone appeared at

5 days (Fig. 1A), filled up the drill-hole defect in the bone

marrow by 7 days (Fig. 1B). The enchondral bone forma-

tion did not occur in the medullary region. Then, the

medullary bone mass began to reduce, and newly formed

bone in the drilled cortical area began to increase (Figs.

1C–E). In 14 days, the regenerating bone in the drilled

cortical tissue area increased (Fig. 1D), and newly formed

cortical bone appeared in 21 days (Fig. 1E).

The mean values of T. BV/TV and Med. BV/TV were

not significantly different among the three groups (Figs.

2A, B). In the high-dose group, the value of Ct. BV/TV

was significantly increased compared with the control at

14 days (Figs. 1D and 2C). At 21 days, newly formed

cortical bone tissue filled up the defect in both the high-

and low-dose groups, and their values of Ct. BV/TV were

significantly increased compared to the control (Figs. 1E

and 2C). At 28 days, the values did not significantly differ

among the three groups (Figs. 1C and 2C). In the high-

dose group, the value of Oc.S/BS was significantly in-

creased compared with that in the control at 14 days (Fig.

2D). In the high-dose group, the value of Ob.S/BS was

significantly decreased compared with that in the control at

5 days (Fig. 2E).

Cross-sectional bone mineral values at injured site of the

femur by pQCT

The mean values of Ct. BMC at the whole drill-hole

defect in the high- and low-dose groups were significantly

higher than that in the control group at 21 days (Fig. 3A).

The mean value of Ct. BMA in the high-dose group was

also larger than the control at 21 days (Fig. 3B). The mean

values of Tr. BMC at the whole drill-hole defect in the high-

dose group were significantly larger than the values of the

control group at 14, 21, and 28 days (Fig. 3C). The mean

values of Tr. BMA in the high-dose groups were signifi-

cantly larger than the values in the controls at 21 and 28

days (Fig. 3D). In the low-dose group, the value was smaller

than the control at 5 days.

Expression of mRNA measured by quantitative RT-PCR

Compared to the control group, the expression levels of

EP4, BMP-2, and RANKL mRNAs in the high-dose group

were significantly higher at 7 days (Figs. 4A–C). Further-

more, the expression levels of osteocalcin and TRAP

mRNAs in the high- and low-dose groups at 14 days were

significantly higher than the control (Figs. 4D, E).

Discussion

Our study demonstrated that systemic administration of

the EP4 agonist ONO-4819.CD by twice daily subcutaneous

injections accelerated cortical bone healing by upregulating

the local bone turnover of the regenerating bone in the

femoral diaphysis after drill-hole injury in mature rats. In

histology and histomorphometry, the amounts of the regen-

erating bone tissue in the medullary area did not differ, but

those in the drilled cortical area increased at 14 days in the

high-dose group, filling up the cortical defect at 21 days in

the high- and low-dose groups. In the control group, defect

filling occurred at 28 days. In the high-dose group, the value

of Oc.S/BS was significantly increased compared with that

in the control at 14 days. In the pQCT analyses, the area and

total mineral values of bone tissue equivalent to cortical

bone increased, reaching a plateau level at 21 days in EP4-

treated groups. Expression of EP4, BMP-2, and RANKL

mRNAs in the injured bone tissue increased at 7 days in the

high-dose group. TRAP and osteocalcin mRNA expression

increased at 14 days in the EP4 agonist-treated groups.

Regeneration of the trabecular bone in the medullary

region after injury was not affected by administration of the

EP4 agonist. It started within 5 days after injury, reaching a

peak value in 7 days. The appearance of newly formed bone

in the cortical defect at 7 days did not differ among the three

groups. These findings were compatible with the reported

data in this model [21,22]. Thus, it is obvious that EP4

agonist did not accelerate the bone formation in medullary

callus after bone and marrow injury. However, the subse-

quent increase in the amounts of bone tissue in the drilled

cortical region was accelerated by EP4 agonist. In the

control group, the bone tissue gradually increased from 14

to 28 days in accordance with a gradual decrease in the bone

mass in the medullary region. EP4 agonist administration

M. Tanaka et al. / Bone 34 (2004) 940–948 947

increased the amounts of bone tissue in the drilled cortical

defect from 14 days in the high-dose and from 21 days in

the low-dose group. Thus, EP4 agonist dose-dependently

accelerated the bone tissue repair, filling the cortical defect

with newly formed bone earlier than in the control. PGE2

administered in the first half of the healing period after

fracture reportedly did not show any effect, but when given

in the second half of the period, PGE2 stimulated callus

formation in rabbits [7]. The synthetic EP4-selective agonist

AE1-329 increased callus volume at the periosteal injured

site in mice femur [26]. Thus, the increase in the amounts of

newly formed bone tissue at the cortical defect in the present

study appeared to be consistent with the anabolic effects of

other EP4 agonists in the fracture model.

The tissue volume of newly formed bone (Ct. BV/TV) in

the drilled cortical area increased from 14 days, but the

amounts of bone mineral at the cortical region (Ct. BMC)

increased at 21 days in the high-dose group. This apparent

discrepancy may be related to the degree of mineralization

of newly formed bone. The degree of mineralization tended

to be reduced in the bone tissue rapidly formed by injecting

human parathyroid hormone(1–34) in humans [8,10]. It was

also observed that the mineral content per volume of

fracture callus formed during PGE2 treatment was reduced

[7]. Thus, the apparent discrepancy in increases in bone

mineral and tissue volume in the newly formed cortical bone

in this study may reflect the reduced mineralization of bone

tissue by the high-dose treatment with ONO-4819.CD.

Expression of osteocalcin, TRAP, and RANKL mRNAs

markedly increased within 7 days at the injured site in all

three groups. These increases in mRNAs of bone markers

were associated with the medullary bone formation. In EP4

agonist-treated groups, however, increases in osteocalcin

and TRAP mRNA expression were sustained at 14 days,

when expression of these markers reduced in the control

group. The upregulation of osteocalcin and TRAP mRNAs

was apparently consistent with the acceleration of the

turnover in the medullary bone and newly formed cortical

bone. Systemic markers of bone resorption such as urinary

DPD were shown to increase after forearm and ankle

fractures in humans [3,4,19]. They also changed during

the healing process of fractures in the extremities [12].

Although serum Gla-osteocalcin levels fluctuated, urinary

DPD did not change after drill-hole injury in the control

group of rats. Thus, it seems that drill-hole injury in the

bilateral femurs does not substantially affect the systemic

bone turnover in mature rats.

An increase in EP4 mRNA expression observed in the

high-dose group at 7 days may be related to the concomitant

increases in BMP-2 and RANKL mRNAs and subsequent

upregulation in osteocalcin and TRAP mRNA expression.

Reportedly, PGE2 administration enhanced BMP-2 mediat-

ed osteoblast differentiation, suggesting that BMP signaling

events may be downstream of EP4 signaling [20,27]. Since

upregulation of EP4 mRNA by PGE2 administration has

been observed in immature osteoblastic cell lines and bone

marrow osteoblast precursor cells in young mice [25], it is

possible that administration of synthetic EP4 ligand ONO-

4819.CD locally increased the expression of EP4 mRNA in

osteoblastic cells at the site of bone and bone marrow injury.

However, a significant increase in EP4 and BMP-2 mRNA

expression was not observed in the low-dose group, which

also showed subsequent increases in the volume of cortical

bone as well as in osteocalcin and TRAP mRNA expression.

The production of some growth factors, cytokines, and their

cognate receptors, including transforming growth factors

and BMPs, is elevated in and around the site of bone injury

[1]. Thus, the increases in EP4 and BMP-2 mRNAs found in

the high-dose group may not be the direct effect of the EP4

agonist, but rather be due to other factors secondarily

stimulated by EP4 signals. The observation in this model

may be inconsistent with Zhang et al.’s [27] report that

BMP-2 is a target gene for PGE2-induced bone formation.

Since bone mineral values of L4 measured by DXA were

not affected by systemic administration of the EP4 agonist,

the effect of the agent could be due to local upregulation of

these osteogenic and osteoclastogenic signals at the injured

site.

In conclusion, the present study demonstrated that

subcutaneous injection of the EP4 agonist ONO-4819.CD

dose-dependently accelerated the healing of the cortical

bone defect after drill-hole injury in rat femur by stimu-

lating local bone resorption and formation. Local increases

of TRAP and osteocalcin mRNAs in the injured bone

tissue were compatible with these changes. Increases in

EP4, BMP-2, and RANKL expression were observed

before the local increase in bone turnover. Systemic

administration of the EP4 agonist appeared to stimulate

healing of bone injury by locally upregulating the turnover

of the regenerating bone.

Acknowledgments

The authors are grateful for the invaluable assistance of

Drs. Satoshi Ikeda, Shigeki Nishida, Shinobu Arita, Makoto

Watanuki, Hajime Ohtomo, Shojiro Akahoshi, and Toshi-

haru Mori. The authors also greatly appreciate the technical

assistance of Erika Kobayashi and Keiko Shigemoto. This

study was supported in part by Grants-in-Aid from the Japan

Ministry of Education and Science to Toshitaka Nakamura

(Grant No. 12470313) and Akinori Sakai (Grant No.

14370475).

References

[1] Barnes GL, Kostenuik PJ, Gerstenfeld LC, Einhorn TA. Growth

factor regulation of fracture repair. J Bone Miner Res 1999;

14:1805–15.

[2] Chomczynski P, Sacchi N. Single-step method of RNA isolation by

acid guanidinium thiocyanate–phenol–chloroform extraction. Anal

Biochem 1987;162:156–9.

M. Tanaka et al. / Bone 34 (2004) 940–948948

[3] Ingle BM, Hay SM, Bottjer HM, Eastell R. Changes in bone mass and

bone turnover following forearm fracture. Osteoporos Int 1999;

10:399–407.

[4] Ingle BM, Hay SM, Bottjer HM, Eastell R. Changes in bone mass and

bone turnover following ankle fracture. Osteoporos Int 1999;

10:408–15.

[5] Jamsa T, Koivukangas A, Kippo K, Hannuniemi R, Jalovaara P,

Tuukanen J. Comparison of radiographic and pQCT analyses of heal-

ing rat tibial fractures. Calcif Tissue Int 2000;66:288–91.

[6] Jee WS, Ma YF. The in vivo anabolic actions of prostaglandins in

bone. Bone 1997;4:297–304.

[7] Keller J, Klamer A, Bak B, Suder P. Effect of local prostaglandin E2

on fracture callus in rabbits. Acta Orthop Scand 1993; 64:59–63.

[8] Kneissel M, Boyde A, Gasser JA. Bone tissue and its mineralization

in aged estrogen depleted rats after long-term intermittent treatment

with parathyroid hormone (PTH) analog SDZ PTS 893 or human

PTH(1–34). Bone 2001;28:237–50.

[9] Machwate M, Harada S, Leu CT, Seedor G, Labelle M, Gallant M,

et al. Prostaglandin receptor EP4 mediates the bone anabolic effects

of PGE2. Mol Pharmacol 2001;60:36–41.

[10] Misof BM, Roschger P, Cosman F, Kurland ES, Tesch W, Messmer P,

et al. Effects of intermittent parathyroid hormone administration on

bone mineralization density in iliac crest biopsies from patients with

osteoporosis: a paired study before and after treatment. J Clin Endo-

crinol Metab 2003;88:1150–6.

[11] Murakami H, Nakamura T, Tsurukami H, Abe M, Barbier A, Suzuki

K. Effects of tildronate on bone mass, structure, and turnover at the

proximal tibia of growing rats after sciatic neurectomy. J Bone Miner

Res 1994;9:1355–64.

[12] Ohishi T, Takahashi M, Kushida K, Hoshino H, Tsuchikawa T, Naitoh

K, et al. Changes of biochemical markers during fracture healing.

Arch Orthop Trauma Surg 1998;118:126–30.

[13] Okimoto N, Tsurukami H, Okazaki Y, Nishida S, Sakai A, Ohnishi H,

et al. Effects of a weekly injection of human parathyroid hormone(1–

34) and withdrawal on bone mass, strength, and turnover in mature

ovariectomized rats. Bone 1998;22:523–31.

[14] Paralkar VM, Borovecki F, Ke HZ, Cameron KO, Lefker B, Grasser

WA, et al. An EP2 receptor-selective prostaglandin E2 agonist indu-

ces bone healing. Proc Natl Acad Sci U S A 2003;100:6736–40.

[15] Pilbeam CC, Raisz LG, Voznesensky O, Alander CB, Delman BN,

Kawaguchi H. Autoregulation of inducible prostaglandin G/H syn-

thase in osteoblastic cells by prostaglandins. J Bone Miner Res

1995;10:406–14.

[16] Shih MS, Norrdin RW. Effect of prostaglandin E2 on rib fracture

healing in beagles: histomorphometric study on periosteum adjacent

to the fracture site. Am J Vet Res 1986;47:1561–4.

[17] Shih MS, Norrdin RW. PGE2 induces regional remodeling changes in

haversian envelope: a histomorphometric study of fractured ribs in

beagles. Bone Miner 1986;1:227–34.

[18] Simon AM, Manigrasso MB, O’Conor JP. Cyclo-oxygenase 2 func-

tion is essential for bone fracture healing. J Bone Miner Res

2002;17:963–76.

[19] Takahashi M, Kushida K, Hoshino H, Ohishi T, Inoue T. Evaluation

of bone turnover in postmenopause, vertebral fracture, and hip frac-

ture using biochemical markers for bone formation and resorption. J

Endocrinol Invest 1997;20:112–7.

[20] Takiguchi T, Kobayashi M, Nagashima C, Yamaguchi A, Nishihara T,

Hasegawa K. Effect of prostaglandin E2 on recombinant human bone

morphogenetic protein-2-stimulated osteoblastic differentiation in hu-

man periodontal ligament cells. J Periodontal Res 1999;34:431–6.

[21] Uchida S, Doi Y, Kudo H, Furukawa H, Nakamura T, Fujimoto S.

Transient expression of activin beta A mRNA on osteoprogenitor

cells in rat bone regeneration after drill-hole injury. Bone 2000;27:

81–90.

[22] Uchida S, Sakai A, Kudo H, Otomo H, Watanuki M, Tanaka M, et al.

Vascular endothelial growth factor is expressed along with its recep-

tors during the healing process of bone and bone marrow after drill-

hole injury in rats. Bone 2003;32:491–501.

[23] Weinreb M, Suponitzki I, Keila S. Systemic administration of PGE2

increases the osteogenic capacity of bone marrow. Bone 1997;20:

521–6.

[24] Weinreb M, Grosskopf A, Shir N. The anabolic effect of PGE2 in rat

bone marrow cultures is mediated via the EP4 receptor subtype. Am J

Physiol 1999;276:E376–8.

[25] Weinreb M, Machwate M, Shir N, Abramovitz M, Rodan GA, Harada

S. Expression of the prostaglandin E2 (PGE2) receptor subtype EP4

and its regulation by PGE2 in osteoblastic cell lines and adult rat bone

tissue. Bone 2001;28:275–81.

[26] Yoshida K, Oida H, Kobayashi T, Maruyama T, Tanaka M, Katayama

T, et al. Stimulation of bone formation and prevention of bone loss by

prostaglandin E EP4 receptor activation. Proc Natl Acad Sci U S A

2002;99:4580–5.

[27] Zhang X, Schwartz EM, Young DA, Puzas JE, Rosier RN, O’Keefe

RJ. Cyclooxygenase-2 regulates mesenchymal cell differentiation into

the osteoblast lineage and is critically involved in bone repair. J Clin

Invest 2002;109:1405–15.