FGF18 accelerates osteoblast differentiation by upregulating
ORIGINAL ARTICLE
FGF18accelerates osteoblast differentiation by upregulating Bmp2expression
Tomoko Nagayama 1,2,Shigeru Okuhara 1,Masato S.Ota 1,Noriko Tachikawa 2,Shohei Kasugai 2,and Sachiko Iseki 11
Section of Molecular Craniofacial Embryology and 2Section of Oral Implantology and Regenerative Dental Medicine,Tokyo Medical and Dental University Graduate School of Medical and Dental Sciences,Tokyo,Japan
ABSTRACT
Fibroblast growth factor (FGF)signaling is
involved in skeletal development.Among total 22FGFs,it is suggested that FGF18functions in promotion of osteoblast dif-ferentiation.In order to elucidate the mechanism of FGF18-dependent acceleration of osteogenesis,we implanted rhFGF18soaked beads over mouse fetal coronal sutures using ex-utero surgery.The coronal suture area comprises the peripheries of the developing frontal and parietal bones,separated by the sutural mesenchyme.rhFGF18accelerated osteogenesis by pro-moting connection of the frontal and parietal bone domains,resulting in elimination of the sutural mesenchyme.Expression of Fgf receptors ,Fgfr1,-2and -3involved in skeletal develop-ment,was maintained or upregulated in the developing bone domains,consistent with enhanced osteogenesis.Bone morpho-genetic protein (Bmp )2was speci?cally upregulated in the skeletogenic layer and the application of Bmp antagonist,rmNoggin,inhibited rhFGF18-dependent upregulation of oste-oblast markers.These results suggest that FGF18accelerates osteogenesis by upregulation of Bmp2as well as maintenance or upregulation of Fgfr1,-2and -3expression in osteoblasts.Key Words:bone morphogenetic protein 2,ex-utero surgery,?broblast growth factor 18,osteoblast differentiation,skull bones
INTRODUCTION
Fibroblast growth factor/FGF receptor (FGF/FGFR)signaling in skeletogenesis has been reported and studied for more than a decade after various mutations in FGFRs were found to be responsible for human congenital anomalies involving skeletal tissues (Morriss-Kay and Wilkie 2005).Of the four FGFRs,activating mutations in FGFR1and -2are responsible for craniosynostosis syndromes,which are characterized by the early bony fusion of the cranial sutures,whereas those in FGFR3are associated with a certain type of craniosynostosis syndrome and dwar?sm.These activating mutations result in increased ligand-binding af?nity,a loss of ligand-receptor speci?city and/or ligand-independent recep-tor dimerization (Miraoui and Marie 2010).These three receptors are expressed in osteoblasts in both human and mouse developing skull bones (Delezoide et al.1998;Iseki et al.1999);Fgfr1is preferentially expressed in differentiating osteoblasts,Fgfr2is transcribed by proliferating preosteoblasts outlining the bone
primordium,expression pattern of Fgfr3is the intermediate between Fgfr1and -2and overlaps with Fgfr1and Fgfr2expression domains in mouse fetal skull (Iseki et al.1999).In endochondral ossi?cation,expression of Fgfr2is found in the condensed mesen-chyme as well as periosteum,Fgfr3is transcribed in chondrocytes,Fgfr1is expressed in hypertrophic chondrocytes (Ornitz and Marie 2002).These observations suggest that each Fgfr signaling func-tions in regulation of osteoblast differentiation.
To date,22ligands of the receptors have been identi?ed (Itoh and Ornitz 2011),and gene-targeted deletion experiments or analysis of spontaneous mouse mutants revealed that some of the ligands were closely associated with skeletal formation (Montero et al.2000;Liu et al.2002;Ohbayashi et al.2002;Harada et al.2009).Fgf18is expressed in the osteogenic mesenchyme and in osteoblasts,and deletion of Fgf18results in delayed osteogenesis suggesting that Fgf18promotes osteoblast differentiation.
We applied rhFGF18with or without rhBMP2to critical size mouse skull bone defects and found that rhFGF18enhanced and stabilized rhBMP2-dependent bone regeneration (Fujioka-Kobayashi et al.2012).Treatment of critical size bone defects with both rhBMP2and rhFGF18induced thick bone formation with trabecular-like structure,which suggested subsequent bone remodeling.Notably,treatment of rhFGF18alone did not show any effects as seen in phosphate-buffered saline (PBS)beads treated specimens.
These results suggested that rhFGF18promoted further osteo-genesis in which osteogenic differentiation was actively happening.We therefore applied rhFGF18-soaked beads to the fetal mouse coronal suture area where active osteogenic differentiation occurs at the peripheries of developing frontal and parietal bone primordia in order to gain insights into the mechanism of FGF18effects on osteoblast differentiation.
MATERIALS AND METHODS
Animals
C57BL/6mice were used in this study.The morning on which the vaginal plug was identi?ed was designated as embryonic day 0(=E0).All experiments were carried out in accordance with proto-cols certi?ed by the Institutional Animal Care and Use Committee of Tokyo Medical and Dental University.
Subcutaneous insertion of PBS,FGF18or Noggin beads in fetal heads by ex utero surgery
Heparin-coated acrylic beads (diameter 250–300m m,Sigma-Aldrich,St.Louis,MO,USA)were soaked in PBS containing 300ng/m L of rhFGF18(Peprotech,Rocky Hill,NJ,USA)reconsti-tuted in PBS.Af?-Gel Blue Gel beads (diameter,150–300m m;
Correspondence:Sachiko Iseki,Dr,Section of Molecular Craniofacial Embryology,Graduate School of Medical and Dental Sciences,Tokyo Medical and Dental University,1-5-45Yushima,Bunkyo-ku,Tokyo 113-8549,Japan.Email:s.iseki.emb@tmd.ac.jp
Received December 14,2012;revised and accepted January 6,2013.
doi:10.1111/cga.12012Congenital Anomalies 2013;53,83–8883
Bio-Rad Laboratories,Hercules,CA,USA)were soaked in 500ng/m L recombinant mouse(rm)Noggin/Fc chimera(Techne Corporation,Minneapolis,MN,USA)in PBS.Ex utero surgery was carried out on E15.5fetuses,as previously described(Iseki et al. 1999).The fetuses were collected from the dams after48,72or 96h.Fetal heads were directly embedded in https://www.360docs.net/doc/6517323454.html,pound (SAKURA Finetek Japan,Tokyo,Japan)for in situ hybridization. Sections that contained beads were collected from every eighth cutting and placed on glass slides.At least three fetuses were collected for each condition for all experiments.
Skeletal preparations
Preparation of alcian blue-and alizarin red-stained fetal skulls was performed as described(Wood et al.1996)
In situ hybridization
cDNA for Fgfr1,Fgfr2,Fgfr3(IIIc splice variant whole cDNA), Runx2,mouse bone sialoprotein(Bsp),mouse osteocalcin(Ocn), Bmp2,Bmp4and Bmp7were cloned into appropriate vectors.Anti-sense and sense digoxygenin(DIG)-labeled transcripts were gener-ated by RNA polymerases(Roche Diagnostics GmbH,Germany). After hybridization and washing,DIG was detected by immunohis-tochemistry visualized by nitro blue tetrazolium(NBT)/5-bromo-4-chloro-3-indolyl-phosphate(BCIP)(Roche Diagnostics GmbH, Germany)as described(Iseki et al.1999).Colour reaction was carried out the same duration between the PBS-and rhFGF18-treated specimens.
RESULTS
rhFGF18accelerated osteogenesis in the developing coronal suture area resulting in fusion of the frontal and parietal bone primordia
Beads pre-soaked with PBS(control)or rhFGF18were applied to the fetal coronal suture area of E15.5mice by ex utero surgery.The fetal heads were collected at96h(Fig.1A)after bead treatment and subjected to skeletal staining(Fig.1B).Phosphate-buffered saline bead treated fetal skull showed that the edges of developing frontal and parietal bones were in close proximity to each other but main-tained a clear boundary between them,observed as a non-staining thin line(Fig.1C,arrowheads).In contrast,rhFGF18beads appeared to induce fusion between the frontal and parietal bones (Fig.1D,arrow),resulting in obscuration of the boundary when in proximity to placed beads(arrow vs arrowheads).
Osteoblast markers are upregulated in rhFGF18treated coronal suture area
We studied the expression pattern of osteoblast markers,Runx2, Bsp and Ocn(early,differentiated and mature stage osteoblast markers,respectively)on horizontal sections at72h and96h using in situ hybridization.The schematic in Figure2A shows a typical developing coronal suture area where frontal and parietal bones are in close proximity to each other,overlapping at their edges and sandwiching the coronal suture mesenchyme.There are?ve tissue layers in the cross section depicted,the skin(s),the connective tissue(c),the skeletogenic layer(sk)containing the developing frontal(f)and parietal(p)bones,the meninges(m)and the brain (br).We found that after PBS bead treatment,the expression pattern of the markers remained unchanged compared to the unoperated side at both72h and96h(Fig.2B,D,F,H,J,L).In contrast,at 72h rhFGF18bead treatment clearly showed thickening of the expression domains of Runx2,Bsp and Ocn in the skeletogenic layer of the operated side(Fig.2C,G,K double arrows).Further-more,Runx2expression was also induced within the suture(Fig.2C arrowhead),connecting expression domains of the frontal and pari-etal bones.Bsp and Ocn expression domains were still clearly separated at the coronal suture(Fig.2G,K,arrowheads).
At96h,expression domains of all three markers were further thickened(Fig.2E,I,and M)with fusion occurring between the
Fig.1rhFGF18accelerated osteogenesis across the coronal suture.(A)Lateral
view of the mouse fetal head96h after
the bead implant operation.Double
arrowheads indicate the implanted beads
in the coronal suture area under the skin.
The yellow line shows the plane of
sectioning in other?gures.(B)Whole-
mount alcian blue and alizarin red
skeletal staining96h after operation.
Square indicates the coronal suture area
as shown in(C and D).(C and D)Mag-
ni?ed images of the coronal suture area
squared in(B)96h after phosphate-
buffered saline(PBS)(C)and rhFGF18
(D)bead treatments.The coronal
suture(arrowheads)is clearly observed
between the frontal(f)and parietal(p)
bones in the PBS-treated fetus(C).
rhFGF18treatment resulted in a more
ambiguous delineation of the coronal
suture(D,arrow).e,eye.Scale bars:
2mm(A and B),300m m(C and
D).
84T.Nagayama et al.
frontal and parietal expression domains of Bsp and Ocn (arrow-heads)as expected from whole skeletal staining (Fig.1).rhFGF18upregulated Fgfr1,-2,and -3expression in the skeletogenic layer
We previously reported that Fgfr1,-2and -3are expressed in osteoblasts (Iseki et al.1999),therefore the expression of these genes was investigated in the coronal suture area 72h after bead treatment.As expected,PBS bead treatment did not alter expression of these genes (Fig.3A,C,E).Treatment with rhFGF18beads caused an expansion in frontal and parietal expression domains of the Fgfrs examined (Fig.3B,D,F),consistent with the changes observed in osteoblast markers (see Fig.2).Fgfr1expression domains did not show complete fusion;however,the suture mes-enchyme seemed to upregulate Fgfr1expression (Fig.3B).Frontal and parietal expression domains of Fgfr2and -3were connected
as
Fig.2Gene expression of osteogenic markers in the coronal suture area at 72and 96h after phosphate-buffered saline (PBS)or rhFGF18bead implantation.
(A)A schematic image of the coronal suture after horizontal sectioning (as indicated by yellow line in Fig 1A).In situ hybridization showing the expression of Runx2(B–E),Bsp (F–I)and Ocn (J–M)on horizontal coronal sections obtained from PBS treatment at 72h (B,F and J)and 96h (D,H and L)and rhFGF18treatment at 72h (C,G and K)and 96h (E,I and M).Arrowheads indicate the coronal sutures,double arrows indicate thickened expression domain in the skeletogenic layer.Operated or unoperated refers to bead-implanted coronal or its opposite coronal (control),respectively.b;beads,br;brain,c;connective tissue layer,f;frontal bone,m;meningeal layer,p;parietal bone,s;skin,sk;skeletogenic layer.Scale bar:300m m.
Acceleration of osteogenesis by FGF1885
observed for Runx2expression domains in Figure 2(Fig.3D,F).Besides the thickening of the expression domain,Fgfr1and -3expression appeared to be upregulated and expression levels of Fgfr2was maintained.
Upregulation of Bmp2,but not Bmp4or -7was associated with rhFGF18-dependent acceleration of osteogenesis
In postnatal mice,acceleration of osteogenesis by rhFGF18depends on rhBMP2-induced osteogenesis (Fujioka-Kobayashi et al.2012).We investigated the expression of Bmp2,-4and -7that
are known to be involved in osteogenesis 72h after PBS or rhFGF18bead treatment (Fig.4A–F).Bmp2expression was upregulated in the frontal and parietal bone domains in the skel-etogenic layer by rhFGF18beads (Fig.4D),whereas PBS beads showed no effect (Fig.4A).Bmp4and -7were mainly expressed in the meningeal layer and their expression was not particularly altered by rhFGF18or PBS treatment (Fig.4B,C,E,F).
In order to con?rm that rhFGF18induces acceleration of osteo-genesis via upregulation of the Bmp2,we co-implanted beads car-rying rmNoggin,an inhibitor of Bmp signaling,in addition to rhFGF18beads.At 48h after rhFGF18bead implantation,Bmp2was upregulated in the skeletogenic layer including the coronal sutural mesenchyme (Fig.4G)and Runx2expression domains of frontal and parietal bones were thickened and fused (Fig.4H,arrows).The tip of the Bsp expression domain of frontal and parietal bones was thickened (Fig.4I,arrows).Ocn expression was weaker compared to the unoperated side at this stage (Fig.4J).Notably,rmNoggin beads treatment alone did not appear to affect expression of Bmp2,Runx2,Bsp and Ocn compared to the unoperated side (Fig.4K–N).The combination of rhFGF18and rmNoggin inhibited the effects of rhFGF18on the upregulation of Runx2,Bsp and Ocn expression (Fig.4P,Q and R,respectively).Addition of rmNoggin beads did not have any effect on Bmp2upregulation by rhFGF18treatment (Fig.4O).
DISCUSSION
In this study,we showed that rhFGF18accelerated osteogenic dif-ferentiation in the mouse fetal coronal suture area where active osteogenesis was occurring,resulting in fusion of the frontal and parietal bone domains.This result coincides with the hypothesis deducted by bone regeneration study on postnatal mice (Fujioka-Kobayashi et al.2012).
The treatment upregulated or maintained expression of Fgfr1,-2and -3,as well as Bmp2in the skeletogenic layer.Since active osteogenesis is already occurring in the coronal suture area and Bmp2expression is present in osteoblasts,it is not clear if Bmp2upregulated in osteoblasts by rhFGF18resulted in secretion of more Bmp2protein,thereby initiating osteoblast differentiation in surrounding undifferentiated mesenchymal cells,or rhFGF18increases the number of osteoblasts that express Bmp2.Interest-ingly,rmNoggin treatment did not seem to affect osteogenic differ-entiation in the skull bone by itself;however,it demolished not only the rhFGF18effect on upregulation of osteogenic markers but also their endogenous expression on co-administration.Noggin is known to negatively regulate Bmp signaling by binding to Bmps and preventing them from binding to their receptors (Groppe et al.2002).We speculate that under normal conditions,binding of endogenous Bmp protein to its receptor is mediated or supported by extracellular factors,and in the presence of exogenous rhFGF18these interactions may be disrupted or modi?ed,allowing rmNog-gin to bind to Bmps.
Upregulation of Bmp2seemed to be the consequence of enhanced rhFGF18/Fgfr signaling.We observed that
rhFGF18
Fig.3Expressions of Fgfrs in the coronal suture at 72h after phosphate-buffered saline (PBS)or rhFGF18bead implantation.In situ hybridization of Fgfr1(A and B),Fgfr2(C and D)and Fgfr3(E and F).PBS (A,C and E)and rhFGF18(B,D and F)treatment.Oper-ated or unoperated refers to bead-implanted coronal or its opposite coronal (control),respectively.Arrowheads indicate the coronal suture.b;beads,br;brain,c;connective tissue layer,f;frontal bone,m;meningeal layer,p;parietal bone,s;skin,sk;skeletogenic layer.Scale bar:300m m.
?
Fig.4Involvement of Bmp signaling in rhFGF18induced effects on osteoblast marker expression.Expression of Bmp2(A and D),Bmp4(B and E)and Bmp7(C and F)in the coronal suture area 72h after phosphate-buffered saline (PBS)(A–C)or rhFGF18(D–F)bead implantation determined by in situ hybridization.Expression of Bmp2(G,K and O),Runx2(H,L and P),Bsp (I,M and Q)and Ocn (J,N and R),after application of rhFGF18(G–J),rmNoggin (K–N)or both rhFGF18and rmNoggin (O–R)beads to the coronal suture area.Blue beads are rmNoggin-soaked af?gel beads and white beads are either PBS-or rhFGF18-soaked heparin-coated acrylic beads.Operated or unoperated refers to bead-implanted coronal or its opposite coronal (control),respectively.Arrowheads indicate the coronal suture.b;beads,br;brain,c;connective tissue layer,f;frontal bone,m;meningeal layer,p;parietal bone,s;skin,sk;skeletogenic layer.Scale bar:300m m.
86
T.Nagayama et al.
Acceleration of osteogenesis by FGF1887
treatment maintained and/or upregulated the expression of Fgfrs1-3 up to96h after the bead treatment until the fusion of the bone domains.Our previous study showed that rhFGF2treatment on mouse fetal coronal sutures upregulated Fgfr1and downregulated Fgfr2in the skeletogenic layer,subsequently inhibiting completion of osteoblast differentiation,mineralization(Iseki et al.1999).Our preliminary result shows that rhFGF2treatment on the fetal coronal suture area downregulates Fgfr3expression(Nagayama and Iseki, unpubl.data,2012).These observations suggest that maintenance of Fgfr2expression and/or upregulation of Fgfr3expression directly or indirectly led to Bmp2upregulation.It has been sug-gested that the interaction between Bmp signalling and Fgf signal-ling is often found in development(Bénazet and Zeller2009).Kim et al.(1998)reported that rhFGF4accelerates the approximation of the parietal bone margin and rhBMP4increases the tissue volume by inducing Msx transcription expression in a fetal skull organ culture system.In endochondral ossi?cation,FGF and BMP signals work in an antagonistic relationship on chondrocyte proliferation (Minina et al.2002).Both signals apparently work in different context in a stage-or tissue-dependent manner.It has to be clari?ed which Fgfr mediates the signal to upregulate Bmp2expression in the?rst place and how sustained expression of Bmp2as well as Fgfr2and/or Fgfr3is controlled up to72h even after cease of FGF diffusion from the beads at48h(Iseki et al.1999).Further studies are also required to elucidate that the rhFGF18effects on accelera-tion of osteoblast differentiation are due to only upregulation of Bmp2expression or induction of other changes in osteoblast differ-entiation processes.
Previous studies and our results suggest that different combina-tion of Fgf/Fgfr induces distinct effects in a cell.However,com-pensation among Fgfrs is suggested by the observation that bone marrow cells from Fgfr3-null mice express Fgfr1and-2,whereas those from wild-type mice express only Fgfr3(Valverde-Franco et al.2004).Wang et al.(2001)tested the property of FGFR3sig-naling in chondrocytes by expressing a chimeric receptor that com-prises the mutated extracellular and transmembrane domains of FGFR3(G380R)and the intracellular domain of Fgfr1in prolifer-ating chondrocytes in the long bone in vivo.They did not?nd the difference in cell proliferation pattern of the chondrocytes that received the chimeric receptor mediated signals.Since Fgfr1-3are expressed in osteoblasts,further studies will be required to deter-mine how distinct combinational Fgf/Fgfr signaling based on ligand-receptor af?nities reveals the different effects on osteogenic differentiation.
Finally,it is important to clarify the fate of the sutural mesen-chyme in rhFGF18induced frontal and parietal bone fusion as it could provide some insight into the pathological condition of craniosynostosis patients caused by FGFR mutations(Morriss-Kay and Wilkie2005).Although there is an argument if the sutural mesenchymal cells are the source of osteoblasts for?anking skull bones,it has not yet been revealed.Therefore,the loss of the sutural mesenchyme could be a result of differentiation of the cells into osteoblasts,cell migration out of the area or cell death.Histological observation and Caspase-3detection did not indicate that rhFGF18induced cell death(data not shown).Further studies will reveal the character of the sutural mesenchyme.
REFERENCES
Bénazet JD,Zeller R.2009.Vertebrate limb development:moving from classical morphogen gradients to an integrated4-dimentional patterning system.Cold Spring Harb Perspect Biol1:a001339.
Delezoide AL,Benoist-Lasselin C,Legeai-Mallet L et al.1998.Spatio-temporal expression of FGFR1,2and3genes during human embryo-fetal ossi?cation.Mech Dev77:19–30.
Fujioka-Kobayashi M,Ota MS,Shimoda A et al.2012.Cholesteryl group-and acryloyl group-bearing pullulan nanogel to deliver BMP2and FGF18 for bone tissue engineering.Biomaterials33:7613–7620.
Groppe J,Greenwald J,Wiater E et al.2002.Structural basis of BMP signalling inhibition by the cystine knot protein Noggin.Nature420:636–642.
Harada M,Murakami H,Okawa A et al.2009.FGF9monomer-dimer equilibrium regulates extracellular matrix af?nity and tissue diffusion. Nat Genet41:289–298.
Iseki S,Wilkie AO,Morriss-Kay GM.1999.Fgfr1and Fgfr2have distinct differentiation-and proliferation-related roles in the developing mouse skull vault.Development126:5611–5620.
Itoh N,Ornitz DM.2011.Fibroblast growth factors:from molecular evolution to roles in development,metabolism and disease.J Biochem 149:121–130.
Kim HJ,Rice DP,Kettunen PJ,Thesleff I.1998.FGF-,BMP-and Shh-mediated signalling pathways in the regulation of cranial suture morpho-genesis and calvarial bone development.Development125:1241–1251. Liu Z,Xu J,Colvin JS,Ornitz DM.2002.Coordination of chondrogenesis and osteogenesis by?broblast growth factor18.Genes Dev16:859–869.
Minina E,Kreschel C,Naski MC,Ornitz DM,V ortkamp A.2002.Interac-tion of FGF,Ihh/Pthlh,and BMP signaling integrates chondrocyte prolif-eration and hypertrophic differentiation.Dev Cell3:1–20.
Miraoui H,Marie PJ.2010.Fibroblast growth factor receptor signaling crosstalk in skeletogenesis.Sci Signal3:re9.
Montero A,Okada Y,Tomita M et al.2000.Disruption of the?broblast growth factor-2gene results in decreased bone mass and bone formation. J Clin Invest105:1085–1093.
Morriss-Kay GM,Wilkie AO.2005.Growth of the normal skull vault and its alteration in craniosynostosis:insights from human genetics and experimental studies.J Anat207:637–653.
Ohbayashi N,Shibayama M,Kurotaki Y et al.2002.FGF18is required for normal cell proliferation and differentiation during osteogenesis and chondrogenesis.Genes Dev16:870–879.
Ornitz DM,Marie PJ.2002.FGF signaling pathways in endochondral and intramembranous bone development and human genetic disease.Genes Dev16:1446–1365.
Valverde-Franco G,Liu H,Davidson D et al.2004.Defective bone miner-alization and osteopenia in young adult FGFR3-/-mice.Hum Mol Genet 13:271–284.
Wang Q,Green RP,Zhao G,Ornitz DM.2001.Differential regulation of endochondral bone growth and joint development by FGFR1and FGFR3 tyrosine kinase domains.Development128:3867–3876.
Wood HB,Ward SJ,Morriss-Kay GM.1996.Effects of all-trans-retinoic acid on skeletal pattern,5′HoxD gene expression,and RAR beta2/beta4 promoter activity in embryonic mouse limbs.Dev Genet19:74–84.
88T.Nagayama et al.