Prostaglandins & other Lipid Mediators 89 (2009) 20–25

Contents lists available at ScienceDirect

Prostaglandins and Other Lipid Mediators

Effects of prostaglandin E2 on bone in mice in vivo

Qi Gao, Manshan Xu, Cynthia B. Alander, Shilpa Choudhary,Carol C. Pilbeam, Lawrence G. Raisz ∗

New England Musculoskeletal Institute, University of Connecticut Health Center, 263 Farmington Avenue, Farmington, CT 06030, USA

a r t i c l e i n f o

Article history:Received 4 February 2009Received in revised form 11 March 2009Accepted 14 March 2009Available online 25 March 2009

Keywords:Prostaglandin E2

a b s t r a c t

We have examined the effects of varying doses, schedules and routes of administration of prostaglandinE2 (PGE2) on bone in mice. Male C57BL/6 mice treated with a high dose of PGE2 (6 mg/kg/d) showeddecreased trabecular bone volume (BV/TV) by 14 d, indicating increased bone resorption. However, therewas also stimulation of bone formation at 14 d after 3 d treatment with PGE2, since mineral apposition rate(MAR) and bone formation rate (BFR/BS) were increased. In CD-1 male and female mice, PGE2 (3 mg/kg,2/wk for 4 wk) increased MAR by 50% and BFR/BS by 100%, but there was no significant change in BV/TV.Tibial mRNA showed an increase in BMP-2 and RUNX-2 expression with PGE2. Additional experimentsusing a higher dose or longer exposure did not increase bone mass. We conclude that exposure to high

Gene expression

Bone histomorphometryBone formationBA

doses of PGE2 in mice may be anabolic but is balanced by catabolic effects. Studies of PGE2 in combinationwith an inhibitor of resorption could lead to development of a true anabolic model and permit assessment

E2 rec

1

sibntogotsntPboipi(sT

1d

one resorptionnabolic effect

of the roles of specific PG

. Introduction

Prostaglandins, particularly prostaglandin E2 (PGE2), have beenhown to have complex effects on skeletal tissues, but a major effectn vivo is to increase both bone formation and bone resorption orone remodeling [1]. In many species PGE2 treatment results in aet increase in bone mass, with bone formation stimulated morehan resorption [2,3]. In a recent study in rats continuous infusionf PGE2 produced net bone loss while daily injection led to net boneain [4]. Studies of the mechanisms by which this anabolic effectccurs, utilizing mice with deletion of specific prostaglandin recep-ors, have not been carried out, in part because a suitable model thathows an anabolic effect with an increase in bone mass in mice hasot yet been developed. In the present study we have examinedhe effect of varying doses, schedules and routes administration ofGE2 on bone mass and turnover in both inbred (C57BL/6) and out-red (CD-1) mice. Net bone loss occurred with daily administrationf a high dose (6 mg/kg) of PGE2, but not with less frequent admin-stration of lower doses (3 mg/kg). Brief exposure to high dose orrolonged exposure to lower dose of PGE2 resulted in significant

ncreases in mineral apposition rate (MAR), double-labeled surfacedLS/BS) and bone formation rate (BFR/BS). Moreover, the expres-ion of BMP-2 and RUNX-2 was increased in PGE2 treated animals.his indicates that an anabolic response to PGE2 can occur in mice.

∗ Corresponding author. Tel.: +1 860 679 3851; fax: +1 860 679 2109.E-mail address: raisz@nso.uchc.edu (L.G. Raisz).

098-8823/$ – see front matter © 2009 Elsevier Inc. All rights reserved.oi:10.1016/j.prostaglandins.2009.03.003

eptors and signal transduction pathways.© 2009 Elsevier Inc. All rights reserved.

However, a substantial increase in bone mass may require differ-ent doses or duration of therapy or combined treatment with anantiresorptive agent.

2. Materials and methods

2.1. Animals and materials

PGE2 was obtained from Cayman Chemical (Ann Arbor, MI).Other chemicals were from Sigma (St. Louis, MO). Animals wereobtained from Charles River (Wilmington, MA) and from our ownbreeding units. All animal protocols were approved by the AnimalCare and Use Committee of University of Connecticut Health Center.

2.2. Experiments

A series of 6 experiments were carried out to examine the effectof mouse strain, age and gender, as well as dose, duration and routeof administration of PGE2 on skeletal responses. All experimentscompared PGE2 treated groups with a vehicle treated control group.PGE2 was dissolved in ethanol and diluted 1:20 with PBS. Dose wasbased on average body weight of the group. Injections were subcu-taneous (sc) except in experiment 4, where given intraperitoneally.

2.2.1. Experiment 1: daily injection for 3, 6 or 12 d2-month-old male C57BL/6 mice (6/group) were divided into

groups A–D. Group A, vehicle daily for 12 days. Group B, PGE2(6 mg/kg sc) for 3 days followed by vehicle. Group C, PGE2 for 6

ther L

deh

2

vd

2

o

2

it

2

P

2

mf

2

ow

2

uarAm

2

2

wis

2

tSvaraGadsp

2

fia

Q. Gao et al. / Prostaglandins & o

ays followed by vehicle. Group D, PGE2 for 12 days. Animals wereuthanized at day 14. Bones were analyzed by micro CT and staticistomorphometry.

.2.2. Experiment 2: daily Injection for 3 d2-month-old C57BL/6 male mice (10/group) were injected with

ehicle or PGE2 (6 mg/kg sc) for the first 3 days and euthanized atay 14. Analysis included dynamic histomorphometry.

.2.3. Experiment 3: weekly injection for 8 weeks7–9-month-old female C57BL/6 mice (12/group) were injected

nce a week with vehicle or PGE2 (6 mg/kg sc).

.2.4. Experiment 4: 2 injections/wk for 4 weeks2-month-old male and female, CD-1 mice (6/group) were

njected twice a week with vehicle or PGE2 (3 mg/kg) by intraperi-oneal (IP) injection for 4 weeks.

.2.5. Experiment 5: 2 injections/wk for 4 weeks2-month-old female, CD-1 mice (6–7/group) were injected with

GE2 (6 m/kg sc) twice a week for 4 weeks.

.2.6. Experiment 6: Daily injection for 2 weeks4-month-old CD-1 COX-2 wild type (WT) and knockout (KO)

ice (4–7/group) were injected with vehicle or PGE2 (3 mg/kg sc)or 14 days.

.2.7. ProceduresAll mice (except Experiment 1) received 10 mg/kg body weight

f calcein in 2% NaHCO3 7 days before sacrifice and 90 mg/kg bodyeight of xylenol orange, 2 days before sacrifice, intraperitoneally.

.2.8. Serum studiesBlood was obtained by heart puncture at the time of euthanasia

nder carbon dioxide anesthesia. Serum was collected and storedt −80 ◦C. Serum calcium (Eagle Diagnostics, Desoto, TX), tartrateesistant acid phosphatase 5B (TRAP, IDS Inc. EIA kit, Fountain HillsZ) and N-terminal procollagen peptide (PINP, IDS Inc. EIA kit) wereeasured per manufacturer’s instructions.

.3. Densitometry analysis

.3.1. In vivo DXABone mineral density (BMD) and bone mineral content (BMC)

ere measured by a PIXImus 2 densitometer (LUNAR, GE Med-cal Systems), under isoflurane anesthesia. Individual mice werecanned the day before treatment and 2 days before sacrifice.

.3.2. Micro CTCortical and trabecular morphometry of the femur were quan-

ified using microCT (microCT40; Scanco Medical AG, Bassersdorf,witzerland). 3D images were reconstructed using standard con-olution back projection algorithms with Shepp and Logan filteringnd rendered at a discrete density of 578,704 voxels/m3 (isomet-ic 12-�m voxels). Threshold segmentation of bone from marrownd soft tissue was performed in conjunction with a constrainedaussian filter to reduce noise. Volumetric regions for trabecularnalysis of the distal femur were selected within the endosteal bor-ers to include the central 80% of vertebral height and secondarypongiosa of femoral metaphyses located 960 �m from the growthlate.

.3.3. Static histomorphometryLeft femurs were dissected free of surrounding tissue and

xed in freshly prepared 4% paraformaldehyde in PBS, pH 7.4,t 4 ◦C for 5 days. The femurs were decalcified in EDTA/NH3OH,

ipid Mediators 89 (2009) 20–25 21

dehydrated and embedded in paraffin. The femurs were sectionedlongitudinally in the area where the central vein is located, stainedfor TRAP and counterstained with hematoxylin (Fisher Scientific,Fair Lawn, NY).

2.3.4. Dynamic histomorphometryRight femurs were dissected free of surrounding tissue and fixed

in 70% ethanol, dehydrated and embedded in methyl methacrylate.Five micrometers longitudinal sections of undecalcified bones wereprepared using a Reichert-Jung Polycut S microtome (Reichert-Jung, Germany) with a D profile knife (Delaware Diamond KnivesCorp., Wilmington, DE). Histomorphometric measurements weremade in a blinded, nonbiased manner using the OsteoMeasurecomputerized image analysis system (OsteoMetrics, Atlanta, GA)interfaced with a Nikon E400 microscope (Nikon, Melville, NY). Theterminology and units used are those recommended by the Histo-morphometry Nomenclature Committee of the American Societyfor Bone and Mineral Research.

2.3.5. mRNA isolation and quantificationIn selected experiments tibiae were homogenized in 2 ml TRIzol

Reagent (Invitrogen) by a PowerGen polytron (Fisher Scientific) toextract RNA. The RNA quantity was determined. The total mRNA(100 ng per final reaction) was reversed transcribed to CDNA usingHigh Capacity cDNA Archive Kit (Applied Biosystems). PCR was per-formed on an Applied Biosystems 7300 QT-PCR System. Relativequantification method was used with GAPDH as the endogenouscontrol.

2.3.6. StatisticsAll values are expressed as the mean ± standard error of the

mean (SEM). One way and two way ANOVA with Bonferroni’s cor-rection or unpaired two-tailed student’s t-test using Sigma Stat(version 2.03, San Rafael, CA) were used to determine statisticalsignificance as appropriate.

3. Results

3.1. Experiment 1

This experiment was based on prior studies showing maximalanabolic effects of PGE2 in rats at 6 mg/kg [5], as well as studiesshowing increased bone formation rates after brief (5 day) treat-ment with PTH in mice [6] and after brief mechanical loadingattributed to endogenous prostaglandin production [7]. Increasingduration of treatment led to a progressive decrease in trabecu-lar bone volume (BV/TV), associated with a decrease in trabecularnumber (TbN) and an increase in trabecular spacing (TbSP) with nochange in trabecular thickness (TbTh) (Fig. 1), consistent with boneloss due to increased resorption. Micro-CT showed a similar trendfor decreasing trabecular number and increasing trabecular spac-ing and no change in cortical thickness (data not shown). FemurBMC increased by 21 ± 7% in vehicle treated mice, 34 ± 7% in micetreated with PGE2 for 3 days, but the increase was significantly lessin animals treated for 6 and 12 days, 9 ± 2 and 7 ± 1% respectively(p < 0.05 by t-test).

3.2. Experiment 2

Since 3 day treatment with PGE2 did not significantly decrease

bone mass in Experiment 1, the effects of this period of treatmenton dynamic histomorphometry were assessed. In this experiment,there was a significant decrease in BV/TV as well as in TbN withan increase in trabecular spacing (Tb/Sp) at 14 d (Fig. 2). However,there was also a 2 fold increase in mineral apposition rate (MAR) and

22 Q. Gao et al. / Prostaglandins & other Lipid Mediators 89 (2009) 20–25

F of PGf olumea ANO

bp(

3

m

Ffia

ig. 1. Effects of 3, 6 and 12 days of PGE2 treatments on the distal femurs. 6 mg/kgemurs were fixed at day 14. Data obtained by static histomorphometry. (A) Bone vre means ± SEM for (n = 4–6) mice. aSignificantly different from vehicle by one way

one formation rate (BFR/BS) (Fig. 2). BMD by DXA and micro-CTarameters as well as serum calcium were not changed significantlydata not shown).

.3. Experiment 3

To determine whether fewer injections for a longer periodight increase bone mass, 7–9-month-old female C57BL/6 mice

ig. 2. Effects of 3 days of PGE2 treatments on the distal femurs. 6 mg/kg of PGE2 or vehicxed at day 14. Data were obtained by dynamic histomorphometry. (A) Bone volume; (B)re means ± SEM for (n = 9–10) mice. aSignificantly different from vehicle by t-test, p < 0.0

E2 or vehicle were given subcutaneously two 2-month-old C57BL/6 male mice. All; (B) trabecular number; (C) trabecular spacing and (D) trabecular thickness. Data

VA, p < 0.05, bSignificantly different from vehicle by t-test, p < 0.05.

(12/group) were injected once a week with vehicle or PGE2 (6 mg/kgsc). Baseline BV/TV was substantially lower in these animals thanin young males and was further decreased by PGE2 as was TbN,

while TbSp was increased (Table 1). There were small, insignificantincreases in MAR and BFR/BS. Micro-CT of the distal femur and thethird lumbar vertebrae showed no significant changes, except thatcortical bone thickness (CtTH) in the mid-shaft of the femur wasdecreased in treated mice (201 ± .5 vs. 185 ± .2, p < .05 by t-test).

le were given subcutaneously to 2-month-old C57BL/6 male mice. All femurs weretrabecular number; (C) mineral apposition rate and (D) bone formation rate. Data

5.

Q. Gao et al. / Prostaglandins & other Lipid Mediators 89 (2009) 20–25 23

Table 1Experiment 3.

Group BV/TV (%) TbTh (�m) TbN (mm−2) TbSp (�m) MAR (�/day) BFR/BS (�3 �2/day) Ca (mg/dL) Phos (mg/dL) TRAP (U/L)

Veh 2.5 ± 0.5 28.1 ± 1.6 0.87 ± 0.14 1583 ± 480 0.80 ± 0.05 0.24 ± 0.03 9.6 ± 0.1 4.2 ± 0.1 4.8 ± 0.4P .03 ±D

Ss

3

tot(e

Fam

Fwa

GE2 3 d 1.1 ± 0.3a 29.3 ± 0.6 0.36 ± 0.07a 3429 ± 533a 1

ata are mean ± SEM for (n = 8–10) mice.a Significantly different from vehicle treated group by t-test, p < 0.05.

erum calcium and phosphorus were unaffected, while serum TRAPhowed a significant increase.

.4. Experiment 4

To assess whether changing mouse strain or dose might alter

he responses, we conducted an experiment in which 2-month-ld male and female outbred CD-1 mice (6/group) were injectedwice a week with vehicle or PGE2 (3 mg/kg) by intraperitonealIP) route for 4 weeks. Analysis included dynamic histomorphom-try and tibial mRNA levels by real time QT-PCR. There were no

ig. 3. Effects of 4 weeks of PGE2 treatments on the distal femurs. 3 mg/kg of PGE2 or vehit week 4. Data were obtained by dynamic histomorphometry. (A) Bone volume; (B) trabeans ± SEM for (n = 6) mice. aSignificantly different from vehicle by t-test, p < 0.05.

ig. 4. Effects of 4 weeks of PGE2 treatments on tibial RNA. 3 mg/kg of PGE2 or vehicle weere pooled; samples from whole tibia were reverse transcript and performed by real tim

Significantly different from vehicle by t-test, p < 0.05. p = 0.09.

0.10 0.35 ± 0.04 10.1 ± 0.1a 4.6 ± 0.2 6.0 ± 0.3a

significant differences in BMC or BMD by DXA, and no changes inserum calcium, PINP or TRAP. However by dynamic histomorphom-etry there were significant increases in MAR and BFR/BS as well asin mineralizing surface (MS/BS) (Fig. 3). There was no significantchange in BV/TV in the males, while there was a small increase inthe females which did not reach statistical significance (p = 0.06).

Control females had significantly lower BV/TV and TbN than malesby 2 way ANOVA (p < 0.05). Measurement of gene expression in thetibiae of these animals showed a significant increase in the expres-sion of BMP-2 and Runx-2 (Fig. 4). RANKL expression also increasedbut did not attain statistical significance (p = 0.09).

cle were given ip to 2-month-old CD-1 male and female mice. All femurs were fixedecular number; (C) mineral apposition rate and (D) bone formation rate. Data are

re given ip on 2-month-old CD-1 male and female mice. Male and female samplese PCR. (A) BMP2; (B) RUNX2; (C) RANKL. Data are means ± SEM for (n = 12) mice.

24 Q. Gao et al. / Prostaglandins & other Lipid Mediators 89 (2009) 20–25

Table 2Experiment 5.

Group BV/TV (%) TbTh (�m) TbN (mm−2) TbSp (�m) MAR (�/day) BFR/BS (�3 �2/day) Ca (mg/dL) PINP (U/L) TRAP (U/L)

Veh 4.7 ± 0.5 32 ± 3 1.5 ± 0.1 657 ± 51 1.09 ± 0.05 0.39 ± 0.02 10.2 ± 0.3 76 ± 7 25 ± 4PGE2 4 wk 4.3 ± 0.5 27 ± 1 1.6 ± 0.2 656 ± 80 1.45 ± 0.10a 0.54 ± 0.04a 10.4 ± 0.1 84 ± 4 24 ± 5

Data are mean ± SEM for (n = 6) mice.a Significantly different from vehicle treated group determined by unpaired two tailed Student t-test, p < 0.05.

Table 3Experiment 6.

Group BV/TV (%) TbTh (�m) TbN (mm−2) TbSp (�m) MAR (�/day) BFR/BS (�3 �2/day) Ca (mg/dL) PINP (U/L) TRAP (U/L)

WT Veh 7.8 ± 1.9 37 ± 5 2.0 ± 0.4 529 ± 149 1.01 ± 0.08 0.25 ± 0.03 9.9 ± 0.1 34 ± 2 6 ± 1WT PGE2 14 d 7.8 ± 1.1 36 ± 3 2.2 ± 0.2 438 ± 40 1.64 ± 0.05a 0.59 ± 0.06a 10.0 ± 0.1 36 ± 2 18 ± 1a

KO Veh 9.8 ± 1.0 37 ± 2 2.6 ± 0.3 357 ± 40 0.84 ± 0.03 0.24 ± 0.04 10.3 ± 0.1 43 ± 2 10 ± 2K 1.80

D

3

e1wB3t

3

mMmKSw

4

dsdmtdsiwbacotCtbciftm

m

O PGE2 14 d 9.9 ± 1.4 37 ± 3 2.7 ± 0.2 350 ± 33

ata are mean ± SEM for (n = 4–7) mice.a Significantly different from vehicle treated group by t-test, p < 0.05.

.5. Experiment 5

To determine whether a higher dose of PGE2 would eithernhance or diminish the anabolic effect, 2-month-old female CD-mice (6–7/group) were injected with PGE2 (6 mg/kg sc) twice aeek for 4 weeks. There were significant increases in MAR andFR/BS, but the changes were smaller than were observed withmg/kg (Table 2). There were no significant changes in static his-

omorphometric parameters.

.6. Experiment 6

To determine whether decreasing endogenous prostaglandinsight affect the responses we compared WT and COX-2 KO mice.AR and BFR/BS were significantly increased in both WT and KOice, with a slightly but not significantly greater response in the

O animals (Table 3). BV/TV was not increased in either group.erum calcium and P1NP were not significantly altered but TRAPas increased more than 3 fold in both WT and KO animals.

. Discussion

The experiments presented here were carried out in an effort toevelop an anabolic model of PGE2 treatment in mice. Substantialtimulation of bone formation and increased bone mass have beenemonstrated in several other species, but in our experiments inice bone loss occurred using protocols that were similar to those

hat produced bone gain in rats. Our major conclusion is that atoses and schedules similar to those which were anabolic in otherpecies the catabolic effect is sufficient to blunt a gain in bone massn mice. Nevertheless, substantial stimulation of bone formationas demonstrated by dynamic histomorphometry. This was seen inoth C57BL/6 and CD-1 mice, and with different dosage schedulesnd routes of administration. Since our initial experiments werearried out using high doses in C57/BL6 mouse, which have ratesf bone turnover [8] and might be more susceptible to stimula-ion of bone resorption, we carried out additional experiments inD-1 mice using lower doses, both intraperitoneally and subcu-aneously. While there was a suggestion of increased trabecularone volume in females in experiment 4, this was not signifi-ant and repeated studies in these mice failed to show a definitencrease in bone mass. We also noted that while both male and

emale mice showed increases in bone formation rates with PGE2reatment that the females had higher turnover and lower bone

ass.Based on the enhanced anabolic effect of PTH in COX-2 KO

ice [9], we also tested the possibility that the failure to demon-

± 0.17a 0.65 ± 0.07a 10.0 ± 0.2 41 ± 2 36 ± 4a

strate an anabolic effect was due to the presence of endogenousprostaglandins. We did observe a slightly greater increase in MARand BFR/BS with PGE2 treatment of COX-2 knockout mice, but therewas no increase in bone mass. Moreover, the resorptive responseas measured by TRAP was substantial in both WT and KO mice.Agonists which engage one of the two receptors that are believedto be responsible for stimulating osteogenesis [10], the EP2 andEP4 receptors, have been shown to increase bone mass when giveneither systemically or locally [11,12]. We have been able to find onlyone report of anabolic effect of PGE2 in mice in vivo [13]. In this studywhen a low dose of PGE2 was injected over the calvarium there wasan increase in trabecular bone mass of distal femur. Recently, con-tinuous PGE2 infusion was shown to produce bone loss in rats whileintermittent administration resulted in bone gain [4,14]. It is pos-sible that PGE2 levels are more prolonged in mice than in rats aftersubcutaneous injection, although in our studies administration asinfrequently as once a week subcutaneously or twice a week by theintraperitoneal route still did not produce an anabolic effect.

Although there are studies suggesting a role for the other PGE2receptors such as, EP1 and EP3, experiments using selective agonistsor knockouts of specific receptors in mice have indicated that thisrole is probably minor [15–17]. Separate analysis of the catabolicand anabolic effects of PGE2 in vivo in mice will be important in fur-ther study of the role of specific prostaglandin receptors. Both globaland targeted knockouts of these receptors have been developed.Overlapping roles for the EP2 and EP4 receptors, which are linkedto adenylyl cyclase, have been suggested [10,18,19], but differencesin the relative anabolic and catabolic effects of EP2 or EP4 receptoractivation have also been reported [10,18,20]. Further experimentsusing different doses and duration of treatment in mice might yielda model in which bone mass is substantially increased. Alterna-tively it should also be possible to examine the anabolic effect inthe absence of excessive bone resorption by combined treatmentwith a resorption inhibitor such as a bisphosphonate. For example,combined treatment with PGE2 and risedronate has been shown toproduce a greater increase in bone mass in rats than either agentalone [21]. While systemic therapy with prostaglandins or selectiveagonists is unlikely to be a clinical approach to stimulating boneformation, because of the many other cell targets that would beaffected, understanding the pathways of the anabolic effect couldlead to new approaches to therapy.

Acknowledgements

We thank Ms. Barbara Capella for her careful preparation of themanuscript. We thank Dr. Douglas Adams and Ms. Vilmaris Diaz-Doran for the microCT analyses.

ther L

(

R

[

[

[

[

[

[

[

[

[

[

[20] Raisz LG, Woodiel FN. Effects of selective prostaglandin EP2 and EP4 recep-

Q. Gao et al. / Prostaglandins & o

Supported by NIH grants AR-18063 (L.G.R.) and DK-48361C.C.P.).

eferences

[1] Norrdin RW, Jee WS, High WB. The role of prostaglandins in bone in vivo.Prostaglandins Leukot Essent Fatty Acids 1990;41:139–49.

[2] Jee WS, Ma YF. The in vivo anabolic actions of prostaglandins in bone. Bone1997;21:297–304.

[3] Sibonga JD, Zhang M, Ritman EL, Turner RT. Restoration of bone massin the severely osteopenic senescent rat. J Gerontol A Biol Sci Med Sci2000;55:B71–78.

[4] Tian XY, Zhang Q, Zhao R, et al. Continuous PGE2 leads to net bone loss whileintermittent PGE2 leads to net bone gain in lumbar vertebral bodies of adultfemale rats. Bone 2008;42:914–20.

[5] Jee WS, Ke HZ, Li XJ. Long-term anabolic effects of prostaglandin-E2 on tibialdiaphyseal bone in male rats. Bone Miner 1991;15:33–55.

[6] Keller H, Kneissel M. SOST is a target gene for PTH in bone. Bone2005;37:148–58.

[7] Forwood MR. Inducible cyclo-oxygenase (COX-2) mediates the inductionof bone formation by mechanical loading in vivo. J Bone Miner Res1996;11:1688–93.

[8] Glatt V, Canalis E, Stadmeyer L, Bouxsein ML. Age-related changes in trabec-ular architecture differ in female and male C57BL/6J mice. J Bone Miner Res2007;22:1197–207.

[9] Choudhary S, Huang H, Raisz L, Pilbeam C. Anabolic effects of PTH incyclooxygenase-2 knockout osteoblasts in vitro. Biochem Biophys Res Commun2008;372:536–41.

10] Alander CB, Raisz LG. Effects of selective prostaglandins E2 receptor agonists oncultured calvarial murine osteoblastic cells. Prostaglandins Other Lipid Mediat2006;81:178–83.

[

ipid Mediators 89 (2009) 20–25 25

11] Paralkar VM, Borovecki F, Ke HZ, et al. An EP2 receptor-selective prostaglandinE2 agonist induces bone healing. Proc Natl Acad Sci U S A 2003;100:6736–40.

12] Yoshida K, Oida H, Kobayashi T, et al. Stimulation of bone formation and pre-vention of bone loss by prostaglandin E EP4 receptor activation. Proc Natl AcadSci U S A 2002;99:4580–5.

13] Kato N, Hasegawa U, Morimoto N, et al. Nanogel-based delivery systemenhances PGE2 effects on bone formation. J Cell Biochem 2007;101:1063–70.

14] Tian XY, Zhang Q, Zhao R, et al. Continuous infusion of PGE2 is catabolic witha negative bone balance on both cancellous and cortical bone in rats. J Muscu-loskelet Neuronal Interact 2007;7:372–81.

15] Sarrazin P, Hackett JA, Fortier I, Gallant MA, de Brum-Fernandes A. Role of EP2and EP4 prostaglandin receptors in reorganization of the cytoskeleton in maturehuman osteoclasts. J Rheumatol 2004;31:1598–606.

16] Tsujisawa T, Inoue H, Nishihara T. SC-19220 antagonist of prostaglandinE2 receptor EP1, inhibits osteoclastogenesis by RANKL. J Bone Miner Res2005;20:15–22.

17] Take I, Kobayashi Y, Yamamoto Y, et al. Prostaglandin E2 strongly inhibits humanosteoclast formation. Endocrinology 2005;146:5204–14.

18] Choudhary S, Alander C, Zhan P, Gao Q, Pilbeam C, Raisz L. Effect of deletionof the prostaglandin EP2 receptor on the anabolic response to prostaglandinE2 and a selective EP2 receptor agonist. Prostaglandins Other Lipid Mediat2008;86:35–40.

19] Li M, Healy DR, Li Y, et al. Osteopenia and impaired fracture healing in aged EP4receptor knockout mice. Bone 2005;37:46–54.

tor agonists on bone resorption and formation in fetal rat organ cultures.Prostaglandins Other Lipid Mediat 2003;71:287–92.

21] Ma YF, Lin BY, Jee WS, et al. Prostaglandin E2 (PGE2) and risedronate was supe-rior to PGE2 alone in maintaining newly added bone in the cortical bone siteafter withdrawal in older intact rats. J Bone Miner Res 1997;12:267–75.