Transcriptional induction of prostaglandin G/H
synthase-2 by basic fibroblast growth factor.
H Kawaguchi, … , L G Raisz, M M Hurley
J Clin Invest.
1995;
96(2)
:923-930.
https://doi.org/10.1172/JCI118140
.
In serum-free mouse osteoblastic MC3T3-E1 cells, basic fibroblastic growth factor (bFGF)
induced mRNA and protein for prostaglandin G/H synthase-2 (PGHS-2), the major enzyme
in arachidonic acid (AA) conversion to prostaglandins. mRNA accumulation peaked at 1 h
with bFGF 1 nM. In cells stably transfected with a 371-bp PGHS-2 promoter-luciferase
reporter, bFGF stimulated luciferase activity, which peaked at 2-3 h with bFGF 1-10 nM. In
the presence of exogenous AA, bFGF stimulated PGE2 production, which paralleled
luciferase activity. In serum-free neonatal mouse calvarial cultures, bFGF stimulated PGE2
production in the absence of exogenous AA. bFGF stimulated PGHS-2 mRNA
accumulation, which peaked at 2-4 h and then decreased; there were later mRNA
elevations at 48 and 96 h that were inhibited by indomethacin. In both MC3T3-E1 cells and
neonatal calvariae, bFGF produced smaller and slower increases in PGHS-1 mRNA levels
than for PGHS-2. bFGF stimulated bone resorption in mouse calvariae with a maximal
increase of 80% at 1 nM. Stimulation was partially inhibited by nonsteroidal
anti-inflammatory drugs. We conclude that bFGF rapidly stimulates PGE2 production in
osteoblasts, largely through transcriptional regulation of PGHS-2, and that prostaglandins
mediate some of bFGF's effects on bone resorption.
Research Article
Find the latest version:
Transcriptional Induction of Prostaglandin
G/H Synthase-2
by Basic
Fibroblast
Growth Factor
Hiroshi Kawaguchi,* Carol C. Pilbeam,* Gloria Gronowicz,* Christine Abreu,*BradleyS.Fletcher,* HarveyR. Herschman,*LawrenceG. Raisz,* and Marja M.Hurley*
*Division of Endocrinology and Metabolism, Departmentof Medicine, UniversityofConnecticutHealth Center, Farmington,
Connecticut06030; and tDepartment of Biological Chemistry, UCLA School of Medicine, LosAngeles, California90024
Abstract
Inserum-free mouse osteoblastic MC3T3-E1 cells, basic fi-broblastic growth factor(bFGF) induced mRNA and pro-teinforprostaglandin G/H synthase-2 (PGHS-2),themajor enzyme inarachidonicacid(AA)conversionto prostaglan-dins. mRNA accumulation peakedat1 hwithbFGF 1nM. In cells stablytransfected with a 371-bp PGHS-2 promoter-luciferase reporter, bFGF stimulated promoter-luciferase
activity,
which peaked at2-3 h with bFGF 1-10 AM. In the presence
of exogenous AA, bFGF stimulatedPGE2 production,which
paralleled luciferase
activity.
Inserum-freeneonatalmousecalvarial cultures,bFGFstimulatedPGE2 productionin the absenceofexogenous AA.bFGFstimulatedPGHS-2mRNA
accumulation, which peakedat2-4 h and then decreased; therewerelatermRNAelevationsat48and96h thatwere
inhibited by indomethacin. In both MC3T3-E1 cells and neonatal calvariae, bFGFproduced smaller and slower
in-creases in PGHS-1 mRNA levels than for PGHS-2. bFGF
stimulated boneresorptioninmousecalvariae witha maxi-mal increase of 80% at 1 nM. Stimulation was partially
inhibited by nonsteroidalanti-inflammatorydrugs. We con-clude thatbFGF rapidlystimulatesPGE2 production in
os-teoblasts, largely through transcriptional regulation of
PGHS-2, and thatprostaglandinsmediate some ofbFGF 's
effects on bone resorption. (J. Clin. Invest. 1995. 96:923-930.) Keywords:osteoblasts* boneresorption* prostaglan-dins *indomethacin * luciferase
Introduction
Basicfibroblast growth factor
(bFGF),'
amemberof thefamilyofheparin-binding growth factors, exertspotentmitogenic
ac-tivityon avariety of cells of mesodermal and ectodermal origin
AddresscorrespondencetoMarjaM.Hurley, M.D., Division of
Endocri-nology and Metabolism, Department of Medicine, University of
Con-necticut Health Center,Farmington,CT06030. Phone: 203-679-2129;
FAX:203-679-1258. H.Kawaguchi'spresent addressisDepartmentof Orthopedic Surgery, Faculty of Medicine, Tokyo University, 7-3-1
Hongo,Bunkyo-Ku, Tokyo 113,Japan.
Receivedfor publication6February 1995andacceptedinrevised form19April 1995.
1. Abbreviations used in thispaper: APC, aphidicolin; bFGF, basic
fibroblastgrowth factor; FGF,fibroblast growthfactor;GAPDH,
glycer-aldehyde-3-phosphate dehydrogenase; NSAID, nonsteroidal
anti-in-flammatory drug; PGHS, prostaglandinG/Hsynthase.
(1-4). Bone cells produce bFGF at10-fold higher levels than theyproduce acidicfibroblast growth factor (FGF) another
hep-arin-binding
factor that exhibits 55% sequenceidentity
tobFGF (5,6). Recently,wefound that bFGFmRNAand protein areexpressed inmouseosteoblastic MC3T3-E1 cells (7).bFGF
isstored in the extracellular matrix in association with heparin sulfate and is activated by proteasesorheparin-like molecules (8-10). bFGF could act as an autocrine/paracrine factor for bone cells. bFGF modulates bone formationthrough the regula-tion ofproliferation and differentiation of cells of osteoblastic lineage in vitro(4). We have also reported that bFGF stimulates boneresorption in cultured fetalratlong bones (11).
Among prostaglandins (PGs),
PGE2
is the major product of osteoblasts (12, 13) and is apotent bone resorber in vitro (12, 14, 15). PGE2 also has mitogenic effects on bone cells(16, 17). Both exogenous bFGF (18-20) and PGE2 (21-24) have been shownto stimulatenewboneformation in vivo. In somecelllines, including fibroblasts and synovial cells, bFGF has been shown tostimulate PGproduction (25-27).
There-fore, it is possible that some of the effects of bFGF onbone metabolismmight be mediated by PG synthesis.
Themajorenzymeregulatingthe conversion of arachidonic acid (AA)toPGsisprostaglandin G/H synthase (PGHS), also calledcyclooxygenase,which oxidizes AAtoPGG2 and reduces PGE2 to PGH2. Two forms of PGHS, constitutive PGHS
(PGHS-1) andinducible PGHS (PGHS-2), have been identi-fied.PGHS-lmRNAhas beenidentified in many ovine, murine, and human tissues (28), while the PGHS-2 mRNA was first foundas animmediate-earlygene frommurine fibroblasts and chick embryo fibroblasts (29-31). These two enzymes have 60% homology in nucleic acid and amino acid sequence, al-thoughtheyareproductsof separate genes (28). Bothare
glyco-proteins whose molecular weight is 70-74 kD. PGHS-1 is
con-stitutively expressed, while the recently identified PGHS-2 is likely to be moreimportant in regulating prostaglandin produc-tion by extracellular ligands. Both PGHS-1 and PGHS-2 are
expressed in osteoblastic cells. PGHS-2 is the main enzyme
regulating the production of PGE2 in response to hormones andcytokinesin MC3T3-E1 cell cultures and neonatal mouse calvarial cultures (32-34).
The presentstudywasundertakentoexamine the effects of bFGFonPGproduction and the role of PGs in the responseto
bFGF in bone. Mouse osteoblastic MC3T3-E1 cellswereused
toinvestigatethecellularmechanismsbywhich bFGFregulates PGsynthesis, and neonatalmousecalvarialcultures were used to examine the involvement of PGs inbFGF-stimulated bone
resorption.
Methods
Materials. The PGHS-2 promoter-luciferase fusion genes containing
either 963or371bpof PGHS-2 5'-flanking sequence inpXp-2vector
bFGFonPGHS-2 in Bone 923
J. Clin. Invest.
© The AmericanSociety for Clinical Investigation, Inc.
0021-9738/95/08/0923/08 $2.00
(P2-Luc963 or P2-Luc371, respectively) have been described pre-viously (35). Human recombinant bFGF was provided by California
Biotechnology(Mountain View, CA).MurinePGHS-l cDNA wasthe giftofDrs. David DeWitt andWilliam Smith(Michigan State Univer-sity, East Lansing, MI). Murine PGHS-2 cDNA was obtained from
Oxford Biomedical ResearchInc.(Oxford,MI). Polyclonalrabbit
anti-murine PGHS-2antiserum wasobtainedfromCayman Chemical Co. (AnnArbor, MI).Murineglyceraldehyde-3-phosphatedehydrogenase
(GAPDH)cDNA wasamplified byPCRusingamouseGAPDH control
amplimer setfrom Clontech(Palo Alto, CA).PGE2 antibodywas
pur-chased fromDr. Lawrence Levine (Waltham, MA). Other chemicals
wereobtained from Sigma ChemicalCo. (St. Louis, MO).
StableDNA transfection. The PGHS-2 promoter-luciferase fusion constructs contain either 963 or 371 bp of the 5' flanking sequence immediately proximal to the transcription start site and 70 bp of down-stream untranslated DNA (35). Promoterless luciferase vector was made
by cutting P2-Luc371with BamHI andBglfl, followed by ligationwith T4ligase (GIBCOBRL, Gaithersburg,MD).
PGHS-2promoter-reporter constructs were purified by CsCI banding
and cotransfectedwith pSV2-neo into cultured MC3T3-E1 cellsusing
Lipofectamine (GIBCOBRL). Fortransfection,MC3T3-El cells were
plated in 6well dishes (Costar Corp., Cambridge, MA) at 1 x 105 cells/well andgrownto50-80%confluency (18-24 h)in DMEwithout phenol red (Sigma) containing 10% heat-inactivated FCS (GIBCO
BRL),penicillin(100U/ml), and streptomycin(50ag/ml).Cells were
rinsed twicewithserum-free medium and incubated with 1 ,ugof pro-moter-reporter DNA, 0.067
Isg
ofpSV2-neoDNA, and 8jlofLipofec-taminereagent in 1 mlof serum-free mediumwithoutantibiotics. After
5 hofincubation,asecondml of mediumwith 20% FCS wasadded; 19hlater themedium wasreplaced with fresh complete medium. After 48h, cellsweresplit 1:10 into 100-mm dishes and placedunderselection
with400
pg/ml
ofG418for 2 w.Stablecolonies(>200)werepooled torandomize effects of variationinintegration site. Cells weregrown in culture medium containing 200 ug/mlG418.Cell culture. MC3T3-E1 cells wereplatedin6-well dishes(Costar Corp.) at a density of 50,000cells/cm2 unless stated otherwise and grownfor 6-7 d in DME containing 10% heat-inactivated FCS and
antibiotics.Theywereprecultured foranother 24 h in serum-free.DME
withantibiotics, 1mg/ml BSA (RIA grade; Sigma Chemical Co.), and 100
jig/ml
phosphoascorbic acid (L-ascorbic acid phosphate magnesium salt; WakoPureChemicalIndustries, Osaka, Japan) beforetreatment. Cycloheximide was added 45 min before addition of bFGF. Controlculturesweretreated withappropriate vehicles. Fortestmaterials dis-solvedinethanol,thefinalconcentration of ethanolwas0.1%.
PGHS-2promoteractivity. Luciferase activitywasmeasured in
sol-uble cell extracts prepared with a luciferase detection kit (Promega, Madison, WI) using an automatic injection luminometer (Analytical
LuminescenceLab. Inc., San Diego, CA). Activitywasnormalizedto totalproteins measuredwithaBCAproteinassaykit(Pierce,Rockford, IL). For each experiment, six wellswereanalyzedpertreatmentgroup, and datawereexpressed asmeans±SEM.
Neonatalmousecalvarial culture.Parietal boneswereexcised from
7-d-old CD-i mice (Charles River Laboratories, Wilmington, MA).
Boneswereprecultured for24 h inachemically defined medium, BGJb
(1 mMproline, 3mMphosphate,0.4mML-glutamine, 100
iHg/ml
L-ascorbic acidphosphate,and 1mg/ml BSA)(GIBCO BRL). Theywere
cultured intheexperimentalmedium in 24 well dishes (Costar Corp.) on arocking platform.
Boneresorption assay. Bone resorptionwasmeasured as the release ofpreviously incorporated 45Cafrom neonatal mouse calvariae. Timed pregnant mice wereinjected with 0.05 mCi45Ca onthe 16th day of gestation,andparietalboneswereexcised from 7-d-old neonatal mice. Boneswerepreculturedin BGJb medium for 16-24 h and then cultured with experimentalagents for 96 h, with a medium change after 48 h. Indomethacin oraphidicolinwas added 2 hbefore addition ofbFGF. 45Ca in the medium and in TCAextracts ofbone wasdetermined by
liquidscintillationcounting. 45Carelease in the medium wasexpressed
asthe percentage of totalcountsinthe medium and boneextracts.Data
from multiple experiments were pooled after individual values were normalized to the mean control for each experiment.
Steady state mRNA analysis. Three wells of cells in 6-well dishes
or6-8 half calvariae were pooled for RNA extraction using the method ofChomczynski and Sacchi (36). Briefly, cells or calvariae were ho-mogenized in 4 M guanidinium thiocyanate, extracted with phenol/
chloroform-isoamyl alcohol (24:1), and RNA precipitated with isopro-panol and washed with 80% ethanol. After quantitation at 260nm,20
,tg
of total RNA was run on a1%agarose-2.2 M formaldehyde gel and transferred to a nylon membrane (Genescreen; New England Nuclear Research, Boston, MA) by positive pressure and fixed to the membrane by ultraviolet irradiation. After 3 h of prehybridization in a 50% for-mamide solution at420C,filters werehybridized overnight in a similar solution in rotating cylinders at420Cwith a random primer[32p]dCTP (New England Nuclear Research) -labeled cDNA probe forPGHS-1,PGHS-2, or GAPDH. Filters were washed in a 1 x SSC, 1% SDS solution,0.1 XSSC, at room temperature. Then the filters were washed three times at650Cwith0.1%SDS. The filters were exposed to XAR-5 film (Eastman Kodak, Rochester, NY) at - 70OC.Signals were
quanti-tated by densitometry(Bio-Rad Laboratories, Richmond, CA), and opti-cal densities forPGHS-1 and PGHS-2 were normalized to the corre-sponding values for GAPDH.
Western blot analysis. MC3T3-E1 cells were plated in 100-mm
dishes (Costar Corp.) at a density of 5,000 cells/cm2 and grown to confluence in DME containing 10% serum. Cells were serum deprived for 24 h and treated for 2, 6, and 24 h with or without bFGF (10nM).
Cells were then washed with PBS, harvested by centrifugation, and extracted with 0.5% Tween 20 in a 20 mM potassium phosphate buffer (pH 7.4) containing 1 mM phenylmethyl-sulfonyl fluoride, 1 mM EDTA, and 1 mM N-ethylmaleimide at4°Cfor 30 min. This mixture was centrifuged at 14,000 g for 30min.The supernatant was dialyzed
againstN-ethylmaleimide without Tween 20 for 16 h, and an aliquot was mixed with DEAE cellulose(200-ulbed volume/mg protein)
pre-equilibrated with the potassium phosphate buffer containing 0.05% Tween 20. DEAE cellulose was precipitated by centrifugation, and
pro-tein in the supernatant was measured by the BCA protein assay kit (Pierce). 25
gg
of protein per treatment group was run on anSDS-polyacrylamide gel(10%)and transferred to apolyvinylidene difluoride membrane. Membranes were incubated with1%nonfat drymilkat 4°C
for 16h toblocknonspecific binding andthentreated with a 1:2,000 dilution of polyclonal rabbit anti-PGHS-2 antiserum or nonimmune rabbit serum. Immunoreactive bands were stained using a Western expo-surechemiluminescent detection kit (Clontech) according to the
manu-facturer'sinstructions.
Immunocytochemistry. MC3T3-E1 cells were plated on glass cov-erslips at adensity of5,000 cells/cm2 and grown in DME with 10% FCS. When at least80% confluent, cells were serum deprived for 24 h andthenculturedfor 6or 24 hwith or without bFGF(10 nM). Cells were fixed with 2% paraformaldehyde in PBS for 30 min at room temperature. Cellswere rinsedtwice for10 min withPBS. Cells were
permeabilized with 0.1% TritonX-100for20 min on ice (37). After a 20-minincubation with 0.1% gelatinin PBS, a 1:40dilution of the
rabbitanti-murine PGHS-2 antiserum or rabbit nonimmune serum was
addedtocellsfor 2 h at room temperature. A 1:200 dilution of
rhoda-mine-conjugated anti-rabbit IgGwasaddedtothe cells for1 h at room temperature. Topreventquenching of thefluorescence,2.5% n-propyl
gallate in 1:1 PBS:glycerol was added. The cells were photographed with a fluorescence microscope(Optiphot;Nikon Inc., Melville, NY).
PGE2 assays. Medium was removed fromculturedcells or bones,
and PGE2 accumulation was measured by radioimmunoassay as de-scribedpreviously (38). In some instances indicated in the text, AA (10
pM)
was added to MC3T3-E1 cell culture to provide substrate for PGE2 production.Statisticalanalysis. Means of groups were compared using ANOVA,
andsignificanceof differences was determined bypost-hoc testing using Bonferroni's method.
Results
TIME(H) 0 0.5 1 2 4 24 bFGF (10n) - + - - + - + - +
PGHS-2 . * 9 9
PGHS-1
GAPDH
_M_
wa_ mIi
28S-
18S-PGHS-1 / GAPDH 1.0 1.0 1.1 1.1 1.0 0.9 1.5 0.8 1.9 0.9 1.7
Figure 1. Time courseoftheeffectofbFGFonsteady state mRNA levelsof PGHS-2andPGHS-1 inMC3T3-E1 cells.MC3T3-E1 cells were serum deprived for 24 h and cultured with or without bFGF (10 nM) for the time indicated. Total RNA was extracted and20 jigof RNA wasfractionatedon a 1% agarosegel,transferredtonylon filters,
andhybridizedtoPGHS-2, PGHS-2,andGAPDH.Thenumbersonthe bottomarethetreated/controlratiosoftheintensity of PGHS-1
normal-izedto that of GAPDHmeasuredbydensitometry.
were maximal at 1 h, remained elevated for at least 4 h, and thendecreasedoverthe next24 h (Fig. 1). Fig. 2 showsthe dose response of the effect of bFGFonPGHS-2mRNA at 1 h oftreatmentincellsinitiallyplatedatdifferent densities(5,000
and50,000 cells/cm2).PGHS-2mRNA wasinduced with 10-10
M bFGF, and accumulation was maximal at 10- M. These effectswereindependent of theplatingdensity. PGHS-I mRNA was expressed constitutively and was not altered by 1 h of treatment with bFGF(Figs. 1 and2). Stimulation of PGHS-I by bFGF only appeared after 2 h and was less than twofold when normalized to GAPDH mRNA. Cycloheximide (3
jig/
ml),aproteinsynthesisinhibitor, induced PGHS-2 mRNA andpotentiated the effect of bFGF (Fig. 3), indicating that new
protein synthesis is not essential for the stimulatory effect of bFGF.
PGHS-2protein levels in
MC3T3-El
cells were measured by Westernblotting (Fig.4). Immunoreactiveproteinwas in-ducedby bFGF(10 nM)at2 h andwas still presentat24 h.Multiplebandswerepresentat24h, perhapsdueto
deglycosyl-ationduring culture (39). The localization of PGHS-2protein
wasexaminedby immunocytochemistryat6 and 24 h of culture
(Fig. 5). Immunofluorescence microscopy demonstratedan
in-crease in cytoplasmic PGHS-2 levels at 6 h with bFGF (10 nM). Staining was still present at24 h. The nuclear staining
bFGF[Log(M)] C -12 -11 -10 -9 -8 C -12 -11 -10 -9 -8 PGHS-2
PGHS-1
GAPDH
PlatingDensit(cells/c')
Figure2. Doseresponse of the effect of bFGFonsteady statemRNA levels of PGHS-2 andPGHS-l inMC3T3-E1 cells at 1 hof culture. Cellswereplatedat5,000or50,000 cells/cm2and cultured for 6 d. Cellswere serumdeprivedfor 24 h and cultured for 1 hwith various concentrationsofbFGF. 20pgof RNAwashybridizedtoPGHS-2,
PGHS-1,and GAPDH.
bFGF - + - +
CHX - - + + PGHS-2
PGHS-1
GAPDH
Figure 3. Effect of cycloheximide (CHX) on
steadystate mRNA levels of PGHS-2 and
PGHS-1inthe presence and absence of bFGF inMC3T3-El cells. Cells were precultured for 24 h and cultured for1 hwith bFGF (10 nM) in the presence and absence of cycloheximide
(3jig/ml). Cycloheximidewaspulsed into the culture45minbefore bFGFtreatment.20jig ofRNA washybridizedtoPGHS-2, PGHS-I,
andGAPDH.
seenin control cultures was also seenwith preimmune serum
(datanotshown), indicating that this wasnonspecific.
Medium PGE2 was not detectable (< 0.1 nM) in
serum-free cell cultures, despite the constitutive expression of
PGHS-1 mRNAand the induction of PGHS-2 by bFGF. However, if cellsweregiven exogenous AA (10jzM),thesubstrate for both
PGHS-l
and PGHS-2, during the last 10 min of culture,PGE2
production was increased by bFGF (10 nM) treatmentasearly
as2 h(Table I).
To examine the transcriptional regulation of PGHS-2 by
bFGF, PGHS-2 promoter-luciferase fusion genes containing
963or371 bp of5' flanking sequence (P2-Luc963 orP2-Luc 371,respectively) were stably transfected into MC3T3-E1 cells. Basic FGF (10 nM) stimulated luciferase activity in P2-Luc 963 and P2-Luc 371to asimilarextent at3 h(treated/control
ratios for luciferaseactivitywere4.7±0.3 and 4.9±0.1, respec-tively), suggesting that abFGF response element is likely to
reside within the proximal 371 bp. Fig. 6 shows the timecourse
of the effect of bFGF on luciferase activity in cells transfected with P2-Luc 371. Significant stimulation was observed at 1 h, reached a maximum at 2-3 h, and decreased to the control level at 24 h.
Luciferase activity of P2-Luc 371 transfected cells and promoterless-Luc transfected cells were measured with or
with-outbFGFat3 h of cultureatdifferentplatingdensities of5,000
and50,000
cells/cm2
(Fig. 7). Cellswereculturedcontinuously in the presence of arachidonic acid ( 10jM),
and mediumPGE2
levels were also measured at 3 h. There were similar dose-dependent responses for both luciferase activity and PGE2
lev-TIME(h) 2 6 24 bFGF - +
200-
97-
68-
43-
29-2 6 24
a L
PGHS-2
AbNon-immune
Figure 4. Timecourseof the effect of bFGFonimmunoreactive PGHS-2proteininMC3T3-E1 cells. Cellswere serumdeprivedfor 24 h and thentreated for2, 6,and 24 h withorwithoutbFGF(10nM).Proteins wereextractedasshown inMethods, and 25
jIg
ofproteinwasloadedon a10%SDS-polyacrylamide geland transferredto apolyvinylidene
difluoride membrane. Membraneswereincubated with 1% nonfatdry
milktoblocknonspecific bindingand thenwithpolyclonalrabbit anti-murine PGHS-2 antiserumornonimmune rabbitserum(both 1:2,000 dilutions). Immunoreactivebandswerestainedusingalkaline
phospha-taseimmunostainingand detectedbychemiluminescence.
Control
6h
24 h
els. There was anincrease in PGE2production with bFGF (10 nM) of 10-fold or more andanincrease in PGHS-2 promoter activity of sevenfold or more. The similarities in these
re-sponses, along with the twofold or less increase in PGHS-1 mRNA levels at 3 h (Fig. 1), suggest that the stimulation of
PGE2 productionwaslargely dependentontranscriptional
regu-lation of PGHS-2. Plating density did not influence bFGF's effectoneitherPGE2 productionorPGHS-2 promoteractivity.
Cells transfected with the promoterless luciferase construct
showed minimal luciferaseactivity, butproducedPGE2 levels similartothe P2-Luc 371 transfected cells.
The neonatalmousecalvarial culture systemcanbe usedto
examine and compare physiological responses of bone cells
(such ascellularproliferation), bone formation(measuredby collagen synthesis), and bone resorption (measured by 45Ca release from prelabeled calvariae). Because bFGF and PGE2
have similar effects on someresponses, weused the calvarial culture systemto determine ifPGFa mediated these responses
to bFGF. 7-d-old neonatal mouse calvariae were cultured in serum-free media. PGHS-2mRNA was notdetectable infreshly
dissected calvariae butwas induced in control cultures during thefirst24hinculture,aspreviously reported (33).The
sponta-neouslyinducedPGHS-2 mRNA levels in control cultures
de-Table I. PGE2 ProductionbySerum-freeMC3T3-EI Cells Treatedfor2h withbFGF (10nM)andforthe Last 10minof
Culture with Arachidonic Acid(AA; 10
pM)
orVehicle(Ethanol)Treatment PGE2 (nM)
Control <0.1
bFGF <0.1
Control +AA 6.2±3.6
bFGF+AA 16.0±3.6*
Valuesaremeans±SEM;n = 3. *Significantly different from control
+AA;P<0.05.
bFGF
Figure 5.
Immunofluores-cencemicroscopyof PGHS-2 staining in
MC3T3-E1 cellstreated with orwithoutbFGF for
6or24 h. MC3T3-El cells were serum-deprivedfor 24h,cultured for6or24 h with orwithout bFGF (10nM), and thenfixed, permeabilized, and incu-bated withpolyclonal
rab-bitanti-murinePGHS-2 antiserum ornonimmune
rabbit serum and then treatedwith
rhodamine-conjugated anti-rabbit
IgG. Anincrease in PGHS-2staining is apparent in the cytoplasmofthecells.The nuclearstainingwhich is seenincontrolcultures wasalso observed in non-immune cells x 200 (data notshown).
creased after 24 h ofculture, eventuallybecoming undetectable. Hence, the calvariae were given a 24 h preculture in control mediato allow the spontaneous induction of PGHS-2 mRNA
to decrease so as not to obscure induction by bFGF. After preculture, calvariae were placed in culture with or without bFGF and cultured for another 96 h, with a medium change after 48 h(Figs. 8 and 9). At thebeginningof treatment, control cultures still had elevated levels of PGHS-2mRNA(Figs. 8A
and9). bFGF stimulated PGHS-2 mRNA levels with a biphasic time course. Levels peakedat2-4h, decreased to undetectable
at24h,and then increased againat48 h. Afterachangetofresh
300
-a Control C
g
20
bFGF(10
nM)
000
'e 0 100 1 00.
0
0 6 12 18 24
Hours in
Culture
Figure 6. Timecourseof theeffect of bFGFonPGHS-2 promoter
activityofMC3T3-E1 cellsstablytransfected with the371-bp PGHS-2promoter-luciferasefusionconstruct(P2-Luc371).Cellswerestably
transfected withaPGHS-2promoter-luciferasefusionconstruct con-taining371bpof the5'flankingsequence and 70bpof downstream untranslatedDNA,and selected with G418asdescribed in Methods.
Transfectedcellswere serumdeprived for24 hand treated withor withoutbFGF(10 nM)for the time indicated. Assay for luciferase
P2-Luc 371 Promoteress_ -Luc
Il ]
5K 50K 5K 50K
i'c
=e
400-*,S 20
a00. 0
60
Ec
'E
_-. 20
maa
ol
T
C -12 11 10 4 8 C -12-11-104 8 C 4 c 4
Log
[bFGF
(M)]
Figure7. Doseresponseoftheeffectof bFGFat3 h on PGHS-2 promoteractivity and mediumPGE2 levelsin MC3T3-E1 cell culture
stably transfectedwith P2-Luc371or apromoterless luciferase.Cells wereplated at5,000or50,000 cells/cm2andculturedfor6 d. Cells were serumdeprivedfor 24 h and culturedfor3 h withvarious concen-trations ofbFGF.The assay for luciferaseactivitywasperformedon soluble cell extracts and normalized to totalproteins. Medium PGE2
wasmeasuredbyradioimmunoassay inthe presence of AA(10AM).
media with bFGF, PGHS-2mRNAlevelswereagain elevatedat
96 h. PGHS-1 mRNA levelswerealsostimulated by bFGF but
only after 4 h, and this increase was maintained. In contrast
with MC3T3-E1 cell cultures, measurable PGE2 accumulated in the medium without the addition of arachidonic acid.A dose-response study showed that at 4 h, the induction of PGHS-2 and PGHS-2 mRNA and
PGE2
production weremaximum atbFGF 1 nM (Fig. 8 B).
In additiontothe 4.2-kb PGHS-2band, asmaller(2.5 kb)
A
T1E (H) 0 0.5 1 2 4 8 12 24 48 96
bFGF + + +- + -+ . + - +
__H.-_
PGHS-1
GAPDH
28S-_
18s-PE2(A)
2bFQ + + + + + . + . + +
TI~zEKo 085 1 2 4 8 12 24 4 9
TIME(H) 4 48 96
NSAIDs - - I I F F I I F - I I F F
bFGF - +- + + +
PGHS-2
GAPDH
P0O421GAPDH 1.053 0 330974.40.0 105 0.00.30.02.40.00.50.004
Figure9. Effectsoftwo structurally unrelated NSAIDs,indomethacin andflurbiprofen,onbFGF-stimulatedPGHS-2mRNAlevelsincultured
neonatal mousecalvariae at 4, 48, and 96 h.Calvariaewereprecultured
for 24 h and cultured with or without bFGF(1 nM)for the time indi-cated. 1 MM indomethacin (I)and 1 ,uMflurbiprofen(F) were added 2hbefore adding bFGF. Total RNA was extracted, and 20
jsg
of RNA wasfractionated on a 1% agarose gel, transferred tonylonfilters, andhybridizedtoPGHS-2 andGAPDH.
mRNA was present in RNA extracted from calvariaetreated for48hwith bFGF. This bandincreased at 96h.Thesmaller
band probably reflects a different site of polyadenylation of PGHS-2 mRNA as reported previously for murine PGHS-2 (40).
Becausewehavepreviously found that endogenous prosta-glandinscanstimulate PGHS-2 mRNA in both MC3T3-E1 cells (32,41) and mouse calvarial cultures (33), we examined the effects oftwostructurally unrelated nonsteroid anti-inflamma-torydrugs (NSAIDs) indomethacin and flubiprofen (bothat 1
MLM)
on theinduction of PGHS-2 inmouse calvariae cultures by bFGF (Fig. 9). At4h theinductionwasonlyslightlyinhib-ited. At 48 and 96 h, both NSAIDs inhibited the bFGF induction ofPGHS-2 mRNA accumulation by about 80%, suggesting that the recurrence of PGHS-2 induction at later time points was
probably dueto autoamplification by the production of prosta-glandins.
Both bFGF and PGE2canhave potent effectsonosteoblastic proliferation and synthesis of collagen. We have previously reported that bFGFinhibited collagen synthesis in MC3T3-E1 cells but that this inhibitionwas notPG dependent (42). In the present study, we examined the effect of bFGF (10 nM) on
Figure8.(A)Time courseofthe
ef-fectofbFGFonsteadystate mRNA levels ofPGHS-2andPGHS-lin cul-tured neonatal mousecalvariae. 7-d-old neonatal mousecalvariaewere precultured for24 hand cultured with
orwithoutbFG1 (1 nM)for the time
B
indicated. Time 0 is time of bFGF ad-bFGFLog(M)l
C -11-10-9 -8 dition. Total RNAwasextracted and20jigof RNAwas fractionatedon a PGHS-2 1%agarose gel, transferred to nylon
M
filters,andhybridizedtotheprobesPGHS1 _ for PGHS-2,PGHS-l,and GAPDH.
(B)Dose response of the effect of
GAPDH bFGF on steady state mRNA levels of PGHS-2 and PGHS-l in cultured
20* I I neonatalmousecalvariaeat4hof Medium culture. Calvariaewereprecultured
PGE2 10- for 24 h and then treated for 4 h with 0
-11-0 --
various concentrations of bFGF. 20C -11 -10- -8 /g of RNAu washybridizedto
PGHS-bFGF[LO9(M)] 2,PGHS-l, andGAPDH.
bFGFonPGHS-2 in Bone 927
30
F
b
*
bFGF
o-Indo
+ bFGF
co a
co
20-Control±SEM
...
__.. ...10
.
.
-12
-11
-10
-9
-8
bFGF [Log
(M)]
Figure 10. Dose response of the effect of bFGF on45Careleasefrom prelabeled neonatal mouse calvariae in the presenceandabsenceof indomethacinat96hof culture.45Caprelabeledcalvariae were
precul-turedfor 16-24h andthenculturedwithexperimentalagentsfor 96h with a medium change after 48 h. Treatment with indomethacin(11sM) wasbegun2 h before the addition ofbFGF(1 nM). 45Cainmedium andTCA extracts ofbonewasdeterminedbyliquid scintillation
count-ingand thecumulativepercentage of45Carelease was calculated. Data
from multiple experimentswerepooled aftercalculating individual
val-ues as apercentage ofthemeancontrol for eachexperiment. Similar
resultswere seen at48hofculture, althoughbFGF stimulationwas less than thatat96h. a,significant effect of bFGF: P< 0.05.b,P
<0.01.c,significant effect of indomethacin: P<0.01.
[3H]thymidine (TdR) incorporation into MC3T3-E1 cells and
culturedmousecalvariae in the presence and absence of indo-methacin(1 MM). bFGF stimulated [3H]TdR incorporationin both cell and organ culture (4.6+0.5-fold and 3.8+0.2-fold,
respectively)at24h, and this stimulationwas notinhibitedby
indomethacin in either system(4.7+0.5-foldand3.8+0.2-fold,
respectively).
Both bFGF andPGE2arepotent stimulators of bone resorp-tion. To examine thepossiblerole ofPGE2inbFGF-stimulated resorption, wemeasured 45Carelease fromprelabeledneonatal
mousecalvariaeat48and 96 h of culture. MediumPGE2
accu-mulationwasalsomeasuredatthesetimepoints,in the absence of exogenous arachidonic acid. bFGF increased the release of 45Ca dose dependently, with amaximal increase of 80% at 1 nMafter96 h of culture (Fig. 10 and Table II).Indomethacin (1,iM)blockedPGE2production, abrogatedtheeffect of bFGF 0.1 nM, and decreased the effects ofhigher concentrations of bFGF, suggesting that bFGF effects on bone resorption are
partially dependentonPGproduction.Similar resultswere seen at48 hofculture (datanotshown).
Toexamine the roleof the mitogenic effect of bFGF in bone resorption,weadded 30
MiM
aphidicolin (APC), an inhibitor of DNAsynthesis, to the neonatal mouse calvarial cultures. At 48and96 h, APC decreased[3H]TdRincorporation into calvariae
by 92-96% (datanotshown).bFGF still increased bone resorp-tion in the presence of APC (TableH). However, APC alone increased PGE2 production and bone resorption; both effects
wereblockedby indomethacin (Table II). Therefore,weadded both APC and indomethacin and found that the combination abrogated the effects of bFGFon boneresorption.
TableII. Effects of bFGF(I0-9M)on
4"Ca
Releasefrom Prelabeled Neonatal Mouse Calvariae and Medium PGE2 Level in the Presence and Absenceof Indomethacin (Indo: 10-6 M) and/orAphidicolin (APC; 3 XJ0- M)45Ca release(%ofcontrol) Medium PGE2 Treatment n 48 h 96 h (nM)
Control 32 100±4 100±4 0.7±0.1
bFGF 32 155±5 (55%)*§ 180±5(80%)*§ 36.6±6.7*§
Indo 21 85±40 92±4 N.D.
Indo +bFGF 20 110±6(29%)§ 124±6(36%)*§ N.D. APC 24 131±5* 142±7* 7.3±1.6*
APC + bFGF 24 146±4 (12%)* 187±5(32%)*§ 143.0±23.3*§ Indo +APC 19 95±7 102±6 N.D. Indo + APC
+ bFGF 19 105±4(11%) 105±4 (3%) N.D.
Values are the mean±SEMfornbone cultures.Data for45Carelease were normalized to 100 for those of the control. The mean control valuesofthe percentage of45Carelease were 11.9±0.4% for 48 h and 17.8±0.7%for 96hofculture. The data in parentheses indicate the percentage of stimulation of45Careleaseproduced by bFGF.Medium
PGE2wasmeasured at 2 d of culture. *Significantdifference from control; P<0.01. *P <0.05. § Significant effectof bFGF; P<0.01. N.D., not detectable (<0.1 nM).
Discussion
Thisstudy has shown that bFGF is a potent stimulator of PGHS-2 mRNA andproteinaccumulation in osteoblastic MC3T3-E1 cells. In thesecells, PGHS-2 mRNAwas notdetectable under control conditions. In contrast, PGHS-1 mRNA was
constitu-tively expressed. Basic FGF transiently induced PGHS-2
mRNA in MC3T3-E1 cells, with levels peaking at 1 h and undetectableat24 h. This inductionwas notblocked by cyclo-heximide. The induction of PGHS-2mRNA wasaccompanied
byinduction of PGHS-2protein,whichwasstill detectableat24 h. In MC3T3-E1 cellsstablytransfected witha371-bp PGHS-2
promoter-luciferase reporter construct, bFGF also rapidly and
transiently stimulated luciferase activity. We conclude that bFGFregulatesPGHS-2expressioninlargepartthrougha
tran-scriptional mechanism. Despite the induced expression of PGHS-2, PGE2productionwas notmeasurable (<0.1 nM) up
to 24h inthese cultures unless arachidonic acidwasgiven as
substrate, suggesting that substrate release may be a limiting
factor for PGproductioninthese cells under these conditions.
Inthe presence of exogenous arachidonicacid,bFGF-stimulated PGE2 production paralleled the stimulation of PGHS-2 pro-moter-luciferase activity, indicating a major dependence on
transcriptionalregulation of PGHS-2.
The5'flanking regionof the PGHS-2 promotercontaining
the 371-bp proximal to the transcription start site appears to
containone or moremajorbFGF response elements.Computer
search of this region has identified an AP-l-like site, which may bind the Fos-Jun dimer and two NF-IL6 consensus
se-quences. Basic FGF has been shownto induce c-fos andc-jun
andNF-IL6transcripts in MC3T3-E1 cells (43, 44). However,
include an SP-1 site, an ATF/CRE region, and an E-box or helix-loop-helix protein-binding element. The ATF/CRE site has been recently shown to mediate v-src induction of the mu-rine PGHS-2 promoter(45).
BasicFGFalsorapidly stimulated PGHS-2 mRNA levels in neonatal mousecalvarial cultures. Basic FGF increased medium PGE2 levels in these cultures, which were measurable in the absence of exogenous substrate, and stimulation of PGE2 pro-duction paralleled the stimulation of PGHS-2 mRNA expres-sion. The mitogenic effects of bFGF inMC3T3-El cells and theinhibitory effects of bFGFoncollagen synthesis in MC3T3-El cells (42) have been showntobeindependent of PG tion. However, in the present study, the inhibition of PG produc-tionby NSAIDs reduced bFGF-stimulated resorption in neona-talmousecalvariae, indicatingthat theresorptiveeffect of bFGF was mediated in part by PGs. Although an inhibitor of DNA
synthesisalonedidnotblockPG-mediatedresorption,inhibition of both DNAsynthesis and PGproductioncompletely blocked
bFGF-stimulated resorption. Hence,wepropose that bFGF has
adirecteffectonresorption through stimulation of proliferation ofosteoclast precursors andanindirect effect mediatedby PGs. We have shown by histological examination that bFGF
in-creases thenumber of osteoclast-like cells in cultures of 21-d fetalratcalvariae(46). Inaprevious study in cultured fetalrat
long bones,wefound thatbFGF-stimulated boneresorptionwas
independent of PGs ( 11 ). Many stimulators of resorption show differentdegrees of dependence onendogenous prostaglandin
production in different model systems, and even in the same
model systems. In general, endogenous prostaglandin produc-tion is more likely to play a role in resorptive responses in neonatal mousecalvariae than in fetal ratlong bones. Similar
observations have been made for thyroid hormone (47) and interleukin 1 (48, 49). These differences may reflect the fact that thetwo culture systems contain adifferentpopulation of osteoclast precursor cells ordifferences in other cell
popula-tions, particularly since the neonatal mouse calvariae contain
somehematopoetic marrow, while fetalratlong bones donot.
Although the induction of PGHS-2 mRNAbybFGF in cal-varial cultures haddecayed by24h,therewas arebound induc-tionat48 and96 h. This rebound could be inhibited by treatment with NSAIDs and hencewasprobablyduelargelyto
autoampli-ficationby endogenousPGs.Autoamplification, perhapsin
con-junction with the smaller,slower,butmoresustained increases in
PGHS-l
expression, may play an important role inmain-taining physiologicresponses mediatedby PGs,suchas bFGF-stimulatedresorption,whichareinitiatedby transient inductions of PGHS-2expression.
Acknowledgments
WethankOlga Voznesensky, KristinaMarcello,andCynthiaAlander
fortheir expert technicalassistance.Wealso thank Dr. Yoshinori Itoh forhelpful discussionofexperimental protocols.
References
1. Gospodarowicz, D., N. Ferrara, L. Schweigerer, and G. Neufeld. 1987. Structuralcharacterization andbiological functions of fibroblastgrowth factor. Endocr. Rev. 8:95-114.
2. Thomas,K. A. 1987. Fibroblastgrowth factors.FASEB(Fed. AnL Soc.
Exp.Biol.)J.1:434-440.
3.Rifkin, D.B., and D.Moscatelli. 1989. Recentdevelopmentin the cell biology of basic fibroblast growth factor. J. Cell Biol. 109:1-6.
4. Canalis, E.,M. Centrella, and T.L. McCarthy. 1988. Effects of basic fibroblastgrowthfactor in vitro. J. Clin. Invest. 81:1572-1577.
5. Seyedin, S. M., T. C. Thomas, A. Y. Thompson, D. M. Rosen, and K. A. Diez. 1985. Purification and characterization of two cartilage-inducing factors from bovine demineralized bone. Proc.NatL.Acad.Sci. USA. 82:2267-2271.
6. Globus,R. K., J. Plouet,and D. Gospodarowicz. 1989. Cultured bovine bone cells synthesizefibroblastgrowth factor and store it in their extracellular
matrix. Endocrinology. 124:1539-1547.
7. Hurley, M. M., C. Abreu, G. A.Gronowicz, H. Kawaguchi, and J. A. Lorenzo. 1994.Expression and regulation of basic fibroblast growth factor mRNA levels in mouse osteoblasticMC3T3-Elcells. J.Biol. Chem. 269:9392-9396.
8.Bashkin, P.,S. Doctrow, M.Klagsbrun,C. M.Svahn,J.Folkman,and I. Vlodavsky.1989.Basicfibroblastgrowth factor binds to subendothelial matrix and is released by heparitinase and heparin-like molecules. Biochemistry. 28:1737-1743.
9.Saksela,O., and D. Rifkin. 1990. Release of basic fibroblastgrowth factor-heparin sulfate complexes from endothelial cells by plasminogen activator-medi-atedproteolytic activity.J.Cell Biol. 110:767-775.
10.Hauschka, P. V., A. E. Mavrakos, M. D.lafrati,S. E. Doleman, and M.
Klagsburn.1986. Growth factors in bone matrix: isolation ofmultipletypes of
affinity chromatographyonheparin sepharose. J. Biol. Chem.261:12665-12674.
11.Simmons, H. A., and L. G. Raisz. 1991. Effects of acid and basic fibroblast growth factorand heparinonresorptionofculturedfetalratlongbones. J. Bone. Miner. Res.6:1301-1305.
12.Kawaguchi,H., C. C. Pilbeam, J. R. Harrison, and L. G. Raisz. The role
ofprostaglandinsin theregulationof bone metabolism. Clin.Orthop.Relat. Res. In press.
13.Raisz,L.G.,J. Y.Vanderhoek,H. A.Simmons,B. E.Kream,and K. C. Nicolaou. 1979.Prostaglandin synthesis byfetal rat bone invitro:evidence fora role ofprostacyclin. Prostaglandins. 17:905-914.
14.Klein,D.C.,and L. G. Raisz.1970.Prostaglandins:stimulation of bone
resorptionin tissue culture.Endocrinology.86:1436-1440.
15.Raisz,L.G.,and F. N. Woodiel. 1989. Effects of alteration in the
cyclopen-taneringonboneresorptiveactivity of prostaglandin. Prostaglandins.
37:229-236.
16. Hakeda,Y., T. Yoshino, Y.Nakatani, N.Kurihara,N. Maeda,and M.
Kumegawa. 1986. Prostaglandin E2stimulates DNAsynthesis byacyclic
AMP-independent pathwayin osteoblastic cloneMC3T3-E1 cells. J. Cell. Physiol.
128:155-161.
17.Feyen,J. H. M., A. Di Bon, A. Van derPlas, C.W.G. M.Lowick,and P. J.Nijweide. 1985. Effectsof exogenousprostanoidsontheproliferationof
osteoblast-likecells in vitro.Prostaglandins. 30:827-840.
18. Kawaguchi, H.,T.Kurokawa, K.Hanada, Y.Hiyama,M.Tamura,E.
Ogata,and T.Matsumoto. 1994. Stimulationof fracture repair byrecombinant human basicfibroblastgrowth factor in normal andstreptozotocin-diabeticrats.
Endocrinology. 135:774-781.
19.Mayahara, H.,T. Ito, H.Nagai,H.Miyajima,R.Tsukuda,S.Taketomi,
J.Mizoguchi,and K. Kata. 1993. In vivo stimulation of endosteal bone formation
bybasic fibroblastgrowthfactor in rats. Growth Factors.9:73-80.
20. Aspenberg, P., and L. S. Lohmander. 1989. Fibroblast growth factor
stimulatesbone formation. ActaOrthop. Scand.60:473-476.
21. Akamine, T.,W. S. S.Jee, H. Z. Ke,X. J.Li,and B. Y. Lin. 1992.
Prostaglandin E2prevents bone loss and adds extra bone to immortalized diatal femoralmethaphysicsin female rats. Bone. 13:11-22.
22.Jee,W.S. S.,K.Ueno, Y. P. Deng, and D. M. Woodbury.1985. The
effectsofprostaglandineE2 ingrowingrats:increasedmetaphysealhardtissue and corticoendostealboneformation. Calcif.Tissue Int.37:148-156.
23.Keller, J.,A.Klamer,B.Bak,and P. Suder.1993. Effects of local
prosta-glandin E2onfracture callus in rabbits. ActaOrthop.Scand. 64:59-63. 24.Miller, S.C., andS. C. Marks. 1993. Local stimulation of new bone
formation by prostaglandin El: quantitative histomorphometryandcomparisonof
-delivery by minipumpsand controlled-releasepellets.Bone. 14:143-152. 25.Candela, M., S. C.Barker,and L. R. Ballou. 1993. Fibroblastgrowth factor increasesTNF alphareceptorexpressionin humanfibroblasts. Mol. Cell. Biochem. 120:43-50.
26.Goddard, D.H., S. L.Grossman, R. Newton, M. A. Clark, and J. S. Bomalaski. 1992.Regulation ofsynovial cell growth: basicfibroblast growth factorsynergizeswithinterleukin-lB stimulating phospholipaseA2 enzyme
activ-ity, phospholipase A2 activatingproduction and release ofprostaglandinE2 by rheumatoid arthritissynovialcells in culture.Cytokine.4:377-384.
27.Tamura,K.,R.Asakai,and R. Okamoto. 1991. Basic fibroblastgrowth
factor in rat corpus luteum stimulates prostaglandin F2a.production. Biochem.
Biophys.Res. Commun. 178:393-399.
28.Smith,W. L. 1992.Prostanoidbiosynthesisand mechanisms of action. Am.J.Physiol. 263:F181-F191.
29.Kujubu, D. A., B. S.Fletcher, B. C. Varnum,R. W. Lim, and H. R. Herschman. 1991.TIS1O,aphorbolester tumorpromoter-induciblemRNAfrom Swiss 3T3cells,encodesanovelprostaglandin synthase/cyclooxygenase
homo-logue.J.Biol. Chem.266:12866-12872.
30.O'Banion,M.K.,H. B.Sadowski,V.Winn,and D. A.Young.1991.A serumandglucocorticoid-regulated4-kilobasemRNA encodesa cyclooxygenase-relatedprotein.J.BioL Chem. 266:23261-23267.
31.Xie, W., J. G. Chipman, D. L.Robertson, R. L. Erikson, and D. L.
Simmons. 1991.Expressionof amitogen-responsivegene encoding
prostaglan-ding synthase is regulated by mRNA splicing. Proc. Natl. Acad. Sci. USA. 88:2692-2696.
32. Pilbeam, C. C., H. Kawaguchi, Y. Hakeda, 0.S. Voznesensky, C. B. Alander, and L. G. Raisz. 1993. Differentialregulationof inducible and constitu-tiveprostaglandin endoperoxidesynthase in osteoblasticMC3T3-Elcells. J.Biol. Chem. 268:25643-25649.
33. Kawaguchi, H., L. G. Raisz,0.S.Voznesensky, C. B. Alander, Y. Hakeda, andC. C. Pilbeam. 1994. Regulations of the twoprostaglandinG/Hsynthases
by parathyroid hormone, interleukin-1, cortisol andprostaglandinE2 in cultured neonatal mouse calvariae. Endocrinology135:1157-1164.
34. Harrison, J. R., J. A. Lorenzo, H. Kawaguchi,L. G.Raisz, and C. C. Pilbeam. 1994.Stimulation of prostaglandin E2productionbyinterleukin-la and
transforming growthfactor ain osteoblasticMC3T3-El cells. J. Bone Miner. Res.9:817-823.
35. Fletcher, B. S., D. A. Kujubu, D. M. Perrin, and H. R. Herschman. 1992. Structureofthemitogen-inducibleTIS10 gene and demonstration that the TIS 10-encoded protein is a functional prostaglandin G/H synthase. J. Biol. Chem. 267:4338-4344.
36.Chomczynski,P., and N. Sacchi. 1987.Single-stepmethodof RNA isola-tion byacidguanidinium thiocyanate-phenol-chloroformextraction. Anal. Bio-chem. 162:156-159.
37. Hill, D. L., M. Logan, D. Ong, D. DeSousa, and A. M. Gonzales. 1992. Basicfibroblastgrowthfactor issynthesizedandreleasedbyisolated ovinefetal growth plate chondrocytes: potential role as an autocrine mitogen. Growth Fac-tors. 6:277-294.
38. L. G. Raisz, and H. A. Simmons. 1985. Effects ofparathyroidhormone andcortisol onprostaglandin production byneonatal rat calvaria in vitro. Endocr. Res. 11:59-74.
39.Otto, J.C., D.L. DeWitt, and W. L. Smith. 1993. N-glycosylationof
prostaglandin endoperoxide synthases-l and -2 and their orientations in the
endo-plasmicreticulum. J. Biol.Chem.268:18234-18242.
40.O'Banion,M.K., V. D.Winn, andD.A. Young. 1992. cDNAcloning
andfunctional activity of aglucocorticoid-regulated inflammatory cyclooxygen-ase.Proc. Natl. Acad. Sci. USA. 89:4888-4892.
41. Pilbeam, C. C., L.G. Raisz,0.Voznesensky, C. B. Alander, B. N. Delman, and H. Kawaguchi. 1995. Autoregulation of inducible prostaglandin G/H synthase byprostaglandinsinosteoblastic cells. J. Bone Miner. Res. 10:406-414.
42.Hurley, M. M., C. Abreu, J. R. Harrison, A. C. Lichtler, L. G. Raisz, and B. E. Kream. 1993. Basic fibroblast growth factor inhibits type I collagen gene
expressioninosteoblasticMC3T3-E1cells.J.Biol. Chem. 268:5588-5593. 43.Hurley, M. M., C. Abreu, J. Lorenzo, A. Ray, and G. Gronowicz. 1994.
RegulationofNF-IL6andIL-6 expression by basic fibroblast growth factor in osteoblastic MC3T3-E1 cells. J. Bone Miner. Res. 9 (Suppl. 1):S164.(Abstr.)
44.Okazaki,R., K.Ikeda,A.Sakamoto, T. Nakano, K. Morimoto, T. Kikuchi, K.Urakawa, E. Ogata, and T. Matsumoto. 1992. Transcriptional activation of c-fos and c-jun protooncogenes by serum growth factors in osteoblast-like
MC3T3-El cells. J. Bone Miner. Res. 7:1149-1154.
45. Xie, W., B. S. Fletcher, R. D. Anderson, and H. R. Herschman. 1994. v-srcinduction of theTISlOPGS2 prostaglandinsynthase gene is mediated by an ATF/CREtranscriptionresponse element. Mol. Cell. Biol. 14:6531-6539.
46. Hurley, M. M., M. Kessler, G. Gronowicz, and L. G. Raisz. 1992. The interaction of heparin and basic fibroblast growth factor on collagen synthesis in 21-day fetal rat calvariae.Endocrinology.130:2675-2682.
47.Kawaguchi,H., C. C. Pilbeam,F. N.Woodiel, and L. G. Raisz. 1992.
Comparisonof the effects of3,5,3-triiodothyroaceticacid and triiodothyronine on boneresorptionincultured fetal ratlongbonesand neonatal mouse calvariae. J. BoneMiner. Res. 9:247-253.
48.Raisz, L. G., R. A. Luben, G. R. Mundy, J. W. Dietrich, J. E. Horton, and C. L. Trummel. 1975. Effect of osteoclast activating factor from human leukocytes on bone metabolism. J. Clin. Invest. 56:408-413.
