Developmental Regulation of Wnt/�-Catenin Signals Is Required forGrowth Plate Assembly, Cartilage Integrity, andEndochondral Ossification*

Received for publication, December 20, 2004Published, JBC Papers in Press, March 10, 2005, DOI 10.1074/jbc.M414275200

Yoshihiro Tamamura‡, Tomohiro Otani‡, Naoko Kanatani§, Eiki Koyama‡, Jirota Kitagaki¶,Toshihisa Komori§, Yoshihiko Yamada�, Frank Costantini**, Satoshi Wakisaka¶,Maurizio Pacifici‡, Masahiro Iwamoto‡, and Motomi Enomoto-Iwamoto‡ ‡‡

From the ‡Department of Orthopaedic Surgery, Jefferson Medical College, Thomas Jefferson University, Philadelphia,Pennsylvania 19107, the §Department of Oral Cytology and Cell Biology, School of Dentistry, Nagasaki University,Nagasaki 852-8501, Japan, the ¶Osaka University Faculty of Dentistry, Suita, Osaka 565-0871, Japan, the �MolecularBiology Section, National Institutes of Health, NIDCR, Bethesda MD, 20892, and the **Department of Genetics andDevelopment, College of Physicians and Surgeons, Columbia University, New York, New York 10032

Studies have suggested that continuous Wnt/�-cateninsignaling in nascent cartilaginous skeletal elementsblocks chondrocyte hypertrophy and endochondral os-sification, whereas signaling starting at later stagesstimulates hypertrophy and ossification, indicating thatWnt/�-catenin roles are developmentally regulated. Totest this conclusion further, we created transgenic miceexpressing a fusion mutant protein of �-catenin and LEF(CA-LEF) in nascent chondrocytes. Transgenic mice hadsevere skeletal defects, particularly in limbs. Growthplates were totally disorganized, lacked maturing chon-drocytes expressing Indian hedgehog and collagen X,and failed to undergo endochondral ossification. Inter-estingly, the transgenic cartilaginous elements were illdefined, intermingled with surrounding connective andvascular tissues, and even displayed abnormal joints.However, when activated �-catenin mutant (�-�-cate-nin) was expressed in chondrocytes already engaged inmaturation such as those present in chick limbs, chondro-cyte maturation and bone formation were greatly en-hanced. Differential responses to Wnt/�-catenin signalingwere confirmed in cultured chondrocytes. Activation inimmature cells blocked maturation and actually de-stabi-lized their phenotype, as revealed by reduced expressionof chondrocyte markers, abnormal cytoarchitecture, andloss of proteoglycan matrix. Activation in mature cellsinstead stimulated hypertrophy, matrix mineralization,and expression of terminal markers such as metallopro-tease (MMP)-13 and vascular endothelial growth factor.Because proteoglycans are crucial for cartilage function,we tested possible mechanisms for matrix loss. �-�-Cate-nin expression markedly increased expression of MMP-2,MMP-3, MMP-7, MMP-9, MT3-MMP, and ADAMTS5. In con-clusion, Wnt/�-catenin signaling regulates chondrocytephenotype, maturation, and function in a developmen-tally regulated manner, and regulated action by this path-way is critical for growth plate organization, cartilageboundary definition, and endochondral ossification.

Skeletogenesis continues to attract much research interestbecause of its fundamental roles in embryonic development andgrowth and its susceptibility to pathologies. The process ini-tiates with formation of mesenchymal or ectomesenchymal cellcondensations at specific sites and times in the early embryo.As exemplified by the limb, the mesenchymal condensationsthen undergo chondrogenesis and give rise to cartilaginouslong bone anlagen (1, 2). The chondrocytes become organized ingrowth plates, proliferate, and mature into pre-hypertrophicand hypertrophic chondrocytes. Pre-hypertrophic chondrocytesexpress genes such as Indian hedgehog, and hypertrophic chon-drocytes express a number of characteristic genes, includingcollagen X, alkaline phosphatase, osteopontin, and metallopro-tease (MMP)1-13. Eventually, the hypertrophic chondrocytesmineralize their matrix by deposition of apatitic crystals andare replaced by marrow, vascular, and bone cells via an endo-chondral ossification process. Whereas these stepwise endo-chondral events proceed, diarthrodial synovial joints form ateach epiphyseal end. Joint formation begins with appearance ofa mesenchymal interzone that comes to separate adjacent car-tilaginous anlagen (3). Although the above principles and gen-eral mechanisms are clear and well established, much remainsto be understood about skeletogenesis, particularly at the mo-lecular signaling level.

Wnt proteins are powerful secreted signaling factors thatregulate a number of developmental processes (4, 5). The ver-tebrate Wnt family currently comprises 20 members. Wnt pro-teins act by binding to Frizzled and low density lipoproteinreceptor-related protein cell surface receptors (4, 5). Upon Wntbinding, Frizzled receptors transduce signals via the �-catenin-LEF/TCF pathway (5, 6), Ca2�-calmodulin-PKC pathway (7),or JNK-dependent pathway (8). The �-catenin-LEF/TCF path-way is the best characterized to date and is conserved fromDrosophila to humans (5, 6). In the absence of Wnt binding toFrizzled receptors, �-catenin is found to form complexes withcytoplasmic proteins such as glycogen synthase kinase 3�,Axin, and adenomatous polyposis coli, and remains phospho-rylated (9, 10). Phosphorylated cytoplasmic �-catenin mole-cules become ubiquitynated and are directed for proteasome-assisted degradation (11). On the other hand, when Wntproteins bind to the receptors, degradation is inactivated and a“free” form of �-catenin accumulates in the cytoplasm (9, 10,

* This work was supported by National Institutes of Health GrantsAR050507, AR46000, and AR47543 and the Yamanouchi USA Founda-tion. The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby marked“advertisement” in accordance with 18 U.S.C. Section 1734 solely toindicate this fact.

‡‡ To whom correspondence should be addressed: Dept. of OrthopedicSurgery, Jefferson Medical College, Thomas Jefferson University, 1015Walnut St., Curtis Bldg. Suite 501, Philadelphia, PA 19107. Tel.: 215-955-7624; Fax: 215-955-9159; E-mail: Motomi.Iwamoto@jefferson.edu.

1 The abbreviations used are: MMP, metalloprotease; IHH, Indianhedgehog; HA, hemagglutinin; RT, reverse transcriptase; CA-LEF, con-stitutive active form of mouse LEF-1.

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 19, Issue of May 13, pp. 19185–19195, 2005Printed in U.S.A.

This paper is available on line at http://www.jbc.org 19185

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

12). Accumulated �-catenin molecules translocate to the nu-cleus and serve as a co-activator of resident LEF/TCF tran-scriptional factors, resulting in complexes that bind to responsesequences and modulate expression of target genes (6, 12).

Wnt proteins are involved in skeletogenesis. Seminal obser-vations by Dealy and co-workers (13) first showed that Wnt-7ais expressed by early limb bud ectoderm and influences skeletalpatterning along the dorsal-ventral axis. Wnt-3a was found toshare similar expression patterns and possibly similar roles(14). Subsequent studies indicated that Wnt-1, -4, and -7a exertinhibitory influences on differentiation of mesenchymal cellsinto chondrocytes, whereas Wnt-5a and -5b stimulate it (15,16). Wnt-5a and Wnt-5b were also shown to have strong influ-ences on skeletogenesis in chick and mouse embryos (17–19).Of particular relevance to the present study are two recentreports on Wnt/�-catenin roles in skeletogenesis using trans-genic approaches (20, 21). Akiyama et al. (20) found that con-tinuous expression of an activated �-catenin form (�-�-catenin)in nascent chondrocytes using collagen II promoter sequencesblocks further chondrocyte development and also ossification,as indicated by low expression of chondrocyte master regulatorSox9 and bone regulator Runx2. A detailed analysis of thechondrocyte phenotype and maturation program was, however,not performed. Guo et al. (21) used a similar approach andfound severe defects in long bone joint formation also. On theother hand, previous studies from our groups indicated thatwhen Wnt/�-catenin signaling is experimentally activated atlater stages and specifically in chondrocytes already undergo-ing and advancing toward maturation, activation resulted in amarked stimulation of hypertrophy (22, 23). Together, theabove studies lead to the intriguing and important notion thatWnt/�-catenin signals may not have a single role in skeleto-genesis, but their interventions in the regulation of this com-plex multistep process may be multiple and developmentallyprescribed. The present study was carried out to test thesehypotheses more directly using transgenic approaches in miceand gain-of-function studies in chick embryo limbs and chickchondrocytes in culture. The results provide support for ourhypotheses and reveal also that the Wnt/�-catenin signalingpathway has critical roles in maintenance of cartilage tissueintegrity and formation of proper tissue boundaries.

MATERIALS AND METHODS

Generation of Transgenic Mice—DNA fragments encoding a consti-tutive active form of mouse LEF-1 (CA-LEF) that includes amino acids695–781 of �-catenin fused to the COOH terminus of murine LEF-DNA(24), were cloned into the NotI site of a collagen 2 �1 (Col2�1)-basedexpression vector (25). Synthetic NarI site was introduced into a Hin-dIII site of the above vector. The resulting vector contains the CA-LEFexpression unit including the 5�-NarI site-Col2�1 promoter (nucleotide1940–2971 of M65161), �-globin intron cassette, CA-LEF, SV40poly(A), Col2�1 enhancer (4930–5571 of M65161), NarI site-3� inpNASS� backbone (Clontech). CA-LEF was tagged with HA antigen atthe COOH-terminal site to monitor transgene expression. The expres-sion unit of CA-LEF was excised by NarI digestion, purified, and in-jected into pronuclei of fertilized eggs from F1 hybrid mice (C57BL/6 XC3H). Transgenic embryos were identified by PCR using two pairs ofprimers: 5�-TGCAGCTTTATCCAGGCTGGTCAG-3� and 5�-CACCCAT-CTCATGCTCCATCATCATAGG-3�, which amplify the 345-bp fragmentof CA-LEF; 5�-TGGTCATCATCCTGCCTTTCTC-3� and 5�-ATTAGGT-CACTGTCCGTGTGGG-3�, which amplify the 256-bp fragment of partof the �-globin intron and the LEF-1 DNA binding domain. GenomicDNA was prepared from liver tissue. Expression of the transgene at theprotein level was analyzed by immunostaining for HA antigens. Seven-ty-two of 377 embryos carried CA-LEF transgene and 11 embryosexpressed the transgene products.

Skeletal, Histological, and Immunohistochemical Examinations—For anatomical examination, whole skeletons of E18.5 embryos werestained with alizarin red S and Alcian blue as described (25). For lightmicroscopy, tissues from E18.5 embryos were fixed in 4% paraformal-dehyde, 0.1 M phosphate buffer, and the resulting paraffin sections (6

�m in thickness) were stained with hematoxylin and eosin (H&E). Forimmunohistochemistry, sections were de-masked by treatment with0.2% pepsin in 0.02 N HCl for 15 min at 37 °C, and were incubated withrabbit anti-mouse tenascin-C antibody (26) or anti-MMP-2 antibody(Chemicon International Inc., Temecula, CA) for 16 h at 4 °C. Afterrinsing, sections were incubated with biotinylated anti-rabbit IgG (Vec-tor Laboratories, Inc., Burlingame, CA) and then with Cy3-conjugatedstreptavidin (Jackson ImmunoResearch Laboratories, Inc., West Grove,PA); each incubation was for 30 min at room temperature.

In Situ Hybridization—Tissue sections using in situ hybridizationwas carried out with digoxigenin-conjugated or 35S-labled riboprobes(25, 27). cDNA fragments of mouse Type IX (317–805 of NM_007740)and Type X (1302–1816 of NM_009925) collagens, aggrecan (880–1733of NM_007424), Sox9 (569–1520 of AF421878), IHH (897–1954 ofNM_010544), Gas1 (1559–2452 of NM_008086), GDF5 (1585–2172of U08337), ChM-1 (156–1195 of U43509), and ERG (1007–1029 ofAB07380) were amplified by reverse transcriptase (RT)-PCR from E15mouse limb bud cDNA, subcloned into pGEM-T vector (Promega Corp.,Madison, WI), and used as templates to generate RNA probes. ChickCD44 (438–915 of NM_204980) and Gli3 (688–1131 of AF0022818)were amplified from E8 chick wing bud cDNA by RT-PCR and sub-cloned into pGEM-T vector.

Cell Cultures and Viral Infection—Chondrocytes isolated from thecaudal (for immature cells) and cephalic (for mature cells) portions ofday 17 chick embryo sterna (SPAFAS, c/o) were cultured in high-glucoseDulbecco’s modified Eagle’s medium containing 10% fetal bovine serum(28). The cells were infected with concentrated RCAS virus encoding theactive form of �-�-catenin that lacks the amino-terminal domain (�-catenin) (14) or chick Wnt-14 (29) (kindly provided by Dr. C. J. Tabin,Harvard University) and subcultured at the density of about 1.0–1.5 �104/cm2 by trypsinization after 1 week, and maintained in completemedium containing 10 �g/ml of ascorbic acid. By passage one, morethan 85% of the cells were routinely infected as revealed by immuno-cytochemistry of viral antigens (30); expression of introduced genes wasconfirmed by Western blot. For matrix mineralization, cultures wereprovided with 1 mM �-glycerophosphate (31). Proteoglycan accumula-tion and matrix mineralization in the cell layer were visualized byhistochemistry (30).

Freshly isolated immature and mature chondrocytes were cultureduntil subconfluent and were then transfected with Ax2-Luc promoterreporter vector that contains Axin-2 promoter, exon 1, and intron 1 (32)in the absence or presence of the pUSE expression vector encoding�-�-catenin or the dominant negative form of LEF (that lacks aminoacids 17–264 of murine LEF-1) (24). Transfections were carried outwith TransIT-LT1 transfection reagent (Mirus Bio Corp., Madison, WI)according to the manufacturer’s instructions. Luciferase activity wasmeasured 2 days after transfection by standard methods.

RNA Isolation and RT-PCR—Total RNA was isolated by the guani-dine isothiocyanate method (31). First strand cDNA was synthesizedfrom 1 �g of total RNA with 1 �M of random 9-mer primer (PerkinElmerLife Sciences) using SuperScript IITM reverse transcriptase (Invitrogen)at 42 °C for 45 min. Subsequent amplification was performed withPremix Taq (TaKaRa EX Taq version, Takara Mirus Bio Inc., Madison,WI) for 25–35 cycles under the following conditions: 94 °C for 30 s,58–60 °C for 1 min, and 72 °C for 1 min. Primer sequences for RT-PCRamplification are follows: 5�-TTCTGCTCACTTGGATCCACTG-3� and5�-CGATCTCCATCGTGACGTTGTA-3� for 13–324 of vascular endothe-lial growth factor (AB011078), 5�-GCAACAGAAACGAGGAGCAAAG-CAT-3� and 5�-CAGGGGAAGAAGTAGGTTCACCG TC-3� for 438–915of CD44 (NM_204980), 5�-AGCGGATGCTCAAGGGGTTCTATG-3� and5�-GGTGCAGATGGCACATTTGATGAG-3� for 1814–2409 of chordin(NM_204980), 5�-TGTCTATCTCCCGCTGTCTGATGC-3� and 5�-CGA-CTTCGGTCTTGACTTTGGTGCG-3� for 688 –1131 of Gli3 (AF-0022818), 5�-TCCTTTGCTTCTGGATTTCACGGTG-3� and 5�-CTCTG-GTCTTTCTGTGGGTGCCATT-3� for 523–960 of MMP3 (XM_417175),5�-TGCAACAGCCCAGATGTGGAATCC-3� and 5�-CCTCGCCTATCA-TCCCAATGAAGTG-3� for 295–805 of MMP7 (Ensemble data base ID:ENSGALT00000027770), 5�-GCCCTATGAAGACATCCGCCAGAA-3�and 5�-TGGAAGTTCTCCGTGTCCATCCACT-3� for 84–407 of MT2-MMP (ENSGALT 00000001489), 5�-TCACTCCCCTCACCTTCACCCA-GAT-3� and 5�-ACCCATCGCTGTTGCCACCATTG-3� for 452–856 ofMMP9 (AF222690), 5�-GCCTTGTGCCTCCTTGTCTTGGTT-3� and 5�-AAAGTTCCACCGTGAGTTGTTGCGT-3� for 1837–2365 of cMT3-MMP(NM_205197), 5�-GGAGGCGTTCAGTTTGCTTATCGTC-3� and 5�-TC-CACCACACACCCCACATTTATCATA-3� for 832–1257 of ADAMTS5(ENSGALT00000025470). The resulting products were sequenced toconfirm specific amplification of the target transcripts. All other PCRprimers were described previously (22, 33).

Wnt/�-Catenin Signals and Limb Cartilage Development19186

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

RESULTS

Skeletal Dis-morphogenesis in Transgenic Mice—First, wegenerated transgenic mice in which a constitutive active formof �-catenin (CA-LEF) was expressed in nascent chondrocytesin developing mouse embryo skeletal elements, using cartilagecharacteristic collagen II promoter/enhancer sequences to drivethe transgene (25). CA-LEF is a fusion protein consisting of theDNA binding site of mouse LEF-1 and the transactivationdomain of �-catenin, and has activity in chondrocytes similar to�-�-catenin used in previous studies (24).

Transgenic mice expressing CA-LEF died at birth of respi-ratory failure and displayed severe dwarfism with shortenedlimbs and a flattened skull (Fig. 1B) compared with wild typelittermates (Fig. 1A). Skeletal staining with alizarin red andAlcian blue revealed that mineralization was markedly re-duced in trunk and limbs and affected ribs, vertebrae, and longbones (Fig. 1, A and B). Under-mineralization was also evidentin the cranial base (Fig. 1D, arrow) compared with wild type(Fig. 1C), but calvaria mineralization appeared largely unaf-fected (Fig. 1, A and B). In addition to mineralization defects,transgenic skeletons displayed abnormalities in overall pat-terning, development, and growth, particularly in the limbs.Transgenic scapula, humerus, radius, and ulna were consider-ably short (Fig. 1H) and hind limb elements were sharplyhypoplastic and totally un-mineralized (Fig. 1J) compared withcorresponding wild type elements (Fig. 1, G and I). Ulna andradius had undergone fusion in some embryos (Fig. 1H, arrow)and digit joints were poorly formed (Fig. 1, H and J, arrow-heads). Interestingly, the extent and overall patterns of Alcianblue staining suggested that transgenic cartilages and theirextracellular matrices were actually not uniform and homoge-nous (Fig. 1L, arrows), although wild type cartilage displayedtypical seamless matrix homogeneity (Fig. 1K).

Growth Plate Defects—Histologically, long bone elements inwild type E18.5 limbs had a typical organization consisting of awell formed articular cap and joints, well defined zones ofchondrocyte maturation in the growth plates, and active endo-chondral ossification in the diaphysis (Figs. 2, A, C, E, and G).In contrast, transgenic elements were completely disorganized(Fig. 2, B, D, F, and H). There was no clear histological demar-cation of articular cap and joints, and typical growth plateswere absent. Most if not all the chondrocytes, regardless oftheir localization with a given element, displayed uniform,round and relatively small-sized morphology and shape (Fig. 2,D and H, arrows), thus differing sharply from wild type chon-drocytes displaying their characteristic and marked changes incytoarchitecture and phenotype in different growth plate zones(Fig. 2, C and G). Interestingly, transgenic cartilage was notclearly distinguishable from adjacent tissues, and there wascommon occurrence of an apparent inter-mixing of cartilagetissues with adjacent connective and vascular tissues (Fig. 2, Dand H, arrowheads). The latter probably accounts for the lackof uniform Alcian blue staining seen anatomically in similartransgenic specimens above (Fig. 1L).

To verify that the above histological defects were associatedwith expression of CA-LEF in cartilage, sections were pro-cessed for immunohistochemistry using antibodies to HA tagpresent in the CA-LEF-encoding transgene. Transgenic carti-lages were clearly stained by the HA antibodies, althoughsurrounding tissues were not (Fig. 2J). In addition, every tissuein the wild type specimens was negative (Fig. 2I), reaffirmingspecificity of expression and staining.

Gene Expression Patterns—The severe histological defectsand chondrocyte cellular uniformity seen in transgenic growthplate cartilage suggested that the maturation program of chon-drocytes had been markedly affected by CA-LEF expression. To

FIG. 1. Skeletal structure analyses. Skeletons of E18.5 CA-LEFtransgenic mouse embryos (TG; B, D, F, H, J, and L) and wild typelittermates (WT; A, C, E, G, I, and K) were stained with Alcian blue andalizarin red. A and B, whole body view; C and D, back side view ofcalvaria; E and F, vertebrae and ribs; G and H, forelimb; I and J, hindlimb; K and I, magnified view of hind limb.

Wnt/�-Catenin Signals and Limb Cartilage Development 19187

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

characterize such possible changes, longitudinal sections ofE18.5 limb skeletal elements were processed for histochemicalstaining with Safranin O to depict matrix proteoglycan depo-sition and for in situ hybridization to define the molecular cellphenotypes. Wild type specimens stained strongly with Safra-nin O both in their articular cap and growth plate, and lowstaining was seen in subadjacent endochondral bone and sur-rounding connective tissues as expected (Fig. 3A). In contrast,transgenic cartilages stained in a dull and substandard man-ner, and staining of cartilaginous masses was often not uniform

(Fig. 3B, arrows). In situ hybridization showed that wild typearticular cap and growth plate chondrocytes strongly expressedgenes encoding cartilage-characteristic matrix macromole-cules, such as Collagen IX and Aggrecan, and showed a pre-scribed reduction in expression of these genes in the hypertrophiczone (Fig. 3, C and E). The cartilage master regulator Sox9displayed an identical and typical pattern (Fig. 3G), and theimportant signaling factor Indian hedgehog (IHH) was expressedin the pre-hypertrophic zone (Fig. 3I). Transgenic cartilage dis-played very low and scattered expression of collagen IX and

FIG. 2. Histological analysis of limbskeletons. Longitudinal sections of fore(A–D) and hind (E–J) limb skeletons fromE18.5 CA-LEF transgenic mouse embryos(TG; B, D, F, H, and J) and the wild typelittermates (WT; A, C, E, G, and I) weresubjected to H&E staining (A–H) or im-munostaining with anti-HA antibody (Iand J). Bars represent 1 mm for A, B, E,and F, and 0.2 mm for C, D, and G–J.

Wnt/�-Catenin Signals and Limb Cartilage Development19188

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

aggrecan (Fig. 3, D and F), poor expression of Sox9 (Fig. 3H), andundetectable expression of IHH (Fig. 3J). Not surprisingly then,expression of genes associated with chondrocyte hypertrophy and

endochondral ossification, which is Collagen X, was completelyabsent in transgenic cartilage (Fig. 3L), whereas the gene exhib-ited a strong and restricted pattern in wild type (Fig. 3K).

FIG. 3. Histochemical and in situhybridization characterizations ofgrowth plate phenotype. Longitudinalsections of hind limbs from E18.5 CA-LEFtransgenic mouse embryos (TG; B, D, F,H, J, and L) or tibia from wild type litter-mates (WT; A, C, E, G, I, and K) weresubjected to Safranin O staining (A andB) or in situ hybridization for Type IXcollagen (C and D), aggrecan (E and F),Sox9 (G and H), Indian hedgehog (I andJ), and Type X collagen (K and L) genes.Bar represents 0.2 mm.

Wnt/�-Catenin Signals and Limb Cartilage Development 19189

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

In addition to the above defects in growth plate gene expres-sion, we observed changes in expression of developing synovialjoint genes. For example, GDF5 transcripts were much moreabundant in transgenic tissues and were present throughoutthe entire cartilaginous elements, including tarsal, metatarsal,and phalangeal elements (Fig. 4B, arrows), whereas GDF5transcripts in wild type specimens displayed were characteris-tically limited to nascent joints and their interzones (Fig. 4A,arrows). Similar abnormal expression patterns were observedfor two other joint markers: growth arrest specific gene 1(GAS1) (34) and ERG (35) (Fig. 4, C–F).

Blood Vessel Analysis—To further examine the mechanismsleading to tissue boundary alterations at epiphyseal ends andalong the shaft of long bone elements, we processed longitudi-nal sections for immunostaining with antibodies to tenascin-C,a protein characteristic of developing articular cartilage andpresent at tissue-tissue interfaces also (25, 36). Wild type E18.5limb sections displayed strong and characteristic tenascin-Cstaining, with the protein decorating the epiphyseal articularlayer and metaphyseal perichondrium (Fig. 5A). In transgenicsections, however, the protein was much more abundant andwidely distributed in cartilage and surrounding tissues (Fig. 5,B and F), a clear sign of chaotic tissue-tissue interfaces andrelationships. Thus, we next examined the distribution of bloodvessels, using antibodies to the cell surface marker CD34(which is expressed by mesenchymal and vascular endothelialcells) and CD31 (which is restricted to endothelial cells). Inwild type sections, CD34- or CD-31-positive cells were presentin periarticular tissue, perichondrium, and periosteum (Fig. 5,G and J, arrows) and were absent from articular cap cartilageitself and most of the growth plate cartilage (Fig. 5, G and J).There was, however, clear and well defined staining at the

chondro-osseus boundary reflecting ongoing vascular invasionand endochondral bone and marrow development (Fig. 5, G andJ, arrowheads). In transgenic specimens, CD34- or CD-31-positive cells were distributed in a broader, ill defined andscattered manner (Fig. 5, H, I, K, and L). Some of the positivecells were in very close proximity and essentially within thecartilaginous tissue, appearing to be in the process of activelyentering it (Fig. 5, H and K, arrows). Given the nature of CD-31and CD-34 markers, the positive cells intermingling with car-tilage likely represented a mixture of vascular and mesenchy-mal cells.

We asked then whether abnormal vascular distribution oftransgenic cartilage is associated with, and may be because of,changes in expression of genes regulating angiogenesis. ChM-Iis known as an angiogenic inhibitor and was originally identi-fied in bovine cartilage (37). In wild type sections, ChM-1transcripts were abundant and present from epiphyseal artic-ular cartilage through much of the underlying growth plate,with an obvious reduction in the hypertrophic zone (Fig. 6A).ChM-1 expression was significantly reduced in transgenic spec-imens (Fig. 6B). We also examined the distribution of MMP-2,a protease responsible for extracellular matrix degradationduring vascular formation (38). MMP-2 staining was clear andprominent in pre-hypertrophic and hypertrophic zones of wildtype growth plates (Fig. 6, C and E), but was much moreprominent and actually widespread in transgenic tissue andwas not restricted to cartilage only (Fig. 6, D and F).

Chick Limb Studies—The transgenic data above indicatethat continuous expression of CA-LEF in nascent chondrocytesdriven by collagen II gene promoter sequences blocks furtherchondrocyte maturation and hypertrophy in transgenic mice,whereas our previous studies suggested that CA-LEF expres-sion in cultured growth plate chondrocytes (already engaged inand pursuing maturation) accelerated their hypertrophy. Thedata raised the possibility that chondrocyte responses to Wnt/�-catenin signaling activation depend on the cell developmen-tal stage/status at the moment of activation. To test this pos-sibility, we misexpressed a constitutive-active form of�-catenin (�-�-catenin) in day 5.5 chick embryo wing cartilag-inous skeletal anlagen; at this stage of embryogenesis, hu-merus, radius, and ulna contain: diaphyseal growth plates withchondrocytes already engaged in maturation; and epiphysealcaps containing resting/pre-articular chondrocytes. We rea-soned that simultaneous �-�-catenin misexpression in diaphy-seal and epiphyseal chondrocytes may elicit different responsesand have different developmental consequences in diaphysisversus epiphysis.

Retroviral RCAS particles encoding �-�-catenin were micro-injected in the vicinity of cartilaginous anlagen in day 5.5 chickembryo limbs in ovo, and embryos were monitored and exam-ined over time. Longitudinal sections of control day 11 embryosrevealed that ulna had a typical elongated shape with an epiph-yseal articular cap (Fig. 7C), a prominent growth plate extend-ing from metaphysis to diaphysis (Fig. 7A), and a small amountof invading bone/marrow cell population in the diaphyseal cen-ter (Fig. 7, A and D, arrow). In �-�-catenin-expressing ulna,however, there was extensive ossification and marrow deposi-tion in the diaphysis (Fig. 7, B and F). However, the epiphysealregion had not undergone ossification and marrow deposition,and was instead composed of seemingly immature cartilagewith ill-defined boundaries and weakly staining with SafraninO (Fig. 7E, arrowheads). Joint markers genes CD44 (39) andGli3 (40) were widely expressed (Fig. 7, I and J), whereas theywere confined to developing articular layers in control (Fig. 7, Gand H). Immunostaining with �-catenin antibodies (that rec-ognize �-�-catenin as well) showed that the protein was widely

FIG. 4. Expression of joint marker genes. Longitudinal sections ofhind limbs from E18.5 CA-LEF transgenic mouse embryos (TG; B, D,and F) or metatarsals in wild type littermates (A, C, and E) weresubjected to in situ hybridization for GDF5 (A and B), GAS1 (C and D),and ERG (E and F). Bar represents 0.5 mm.

Wnt/�-Catenin Signals and Limb Cartilage Development19190

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

distributed in epiphyseal and diaphyseal cartilage and adja-cent tissues in virally infected specimens (Fig. 7, M and N),whereas it was much less abundant and restricted in distribu-tion in control tissues as expected (Fig. 7, K and L).

Differential Responses in Immature and Mature Chondro-cytes in Vitro—To test even more directly whether chondrocyteresponses to Wnt/�-catenin signal activation are developmen-tally regulated, populations rich in immature or mature chon-drocytes isolated from caudal and cephalic sternal regions (31)were infected with �-�-catenin-encoding RCAS viruses chon-drocytes and were monitored over time in culture. Companioncontrol cultures infected with insert-less viruses behaved asexpected over time; the caudal cells maintained an immatureand differentiated phenotype (Fig. 8A), whereas the cephaliccells displayed a hypertrophic phenotype (Fig. 8D). The re-sponses of the two cell populations to �-�-catenin overexpres-sion differed markedly. The immature cells became fibroblastic

and flat-shaped (Fig. 8B), but the mature cells remained dif-ferentiated (Fig. 8E). Similar responses were seen when joint-associated Wnt-14 (29) was overexpressed in place of �-�-cate-nin (Fig. 8, C and F). Histochemical staining showed that�-�-catenin overexpression reduced proteoglycan content andaccumulation in both immature and mature cultures (Fig. 8G,AB), but stimulated matrix calcification in mature culturesonly (Fig. 8G, AR). Again, Wnt-14 overexpression elicited sim-ilar responses (Fig. 8G).

To further characterize the cell phenotype, we determinedexpression of various chondrocyte markers by RT-PCR. Over-expression of �-�-catenin or Wnt-14 inhibited aggrecan andtype IX collagen gene expression in both immature and maturechondrocytes (Fig. 9, A and B, respectively). However, it stim-ulated markers of terminal chondrocyte maturation (MMP-13and vascular endothelial growth factor) in mature cultures(Fig. 9D) consistent with data on matrix calcification (Fig. 8G,

FIG. 5. Distribution of tenascin-C, CD34, and CD31. Longitudinal sections of hind limbs from E18.5 CA-LEF transgenic mouse embryos (TG;B, C, E, F, H, I, K, and L) or tibias from wild type littermate (A, D, G, and J) were subjected to H&E staining (A–C) or immunostaining fortenascin-C (TN; D and E), CD34; G and H) and CD31 (J and K). C is a magnified image of the boxed area in B. F, I, and L are phase-contrast imagescorresponding to E, H, and K, respectively. Bar represents 0.2 mm for A, B, and D–L.

Wnt/�-Catenin Signals and Limb Cartilage Development 19191

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

AR), but had no major effect in immature cultures (Fig. 9C).Type X collagen expression, which normally decreases in themineralizing zone of growth plate, was decreased by �-�-cate-nin or Wnt-14 (Fig. 9D).

To make sure that �-�-catenin and Wnt-14 are able to acti-vate the Wnt/�-catenin pathway in immature and maturechondrocytes equally well, control cultures and cultures over-expressing �-�-catenin or Wnt-14 were transfected with a pro-moter reporter construct for Axin-2, a direct target of thatpathway (32). Basic activity of the Axin-2 reporter was slightlyhigher in control mature than immature chondrocytes (Fig.8H), a likely reflection of endogenous differences in ongoingsignaling. Reporter activity was enhanced about 5-fold by �-�-catenin and about 3-fold by Wnt-14 in immature and maturecultures, indicating that the two populations did possess sim-ilar Wnt/�-catenin signal activation capacities. This was con-firmed by similar responses to overexpression of Wnt-3a orWnt-8 (Fig. 8H), both of which are well known activators of theWnt/�-catenin pathway. Increases in reporter activity were allcounteracted by co-expression of dominant-negative LEF (Fig.8H), re-affirming specificity of responses.

MMP Gene Expression—The extracellular matrix is crucialfor cartilage structure and function, and thus our data indicat-ing a major loss of proteoglycan content and accumulation incultures overexpressing �-�-catenin or Wnt-14 was of particu-lar interest. Thus, we asked whether this response is accom-panied by increases in gene expression for metalloproteases,enzymes intimately linked to matrix turnover, and remodeling(41–43). Indeed, �-�-catenin or Wnt-14 overexpression led tosharp increases in expression of MMP-2, -7, and -9, MT3-MMP,and ADAMTS5 in both immature and mature chondrocytecultures (Fig. 9, E and F). MMP-3 expression was stimulated inmature chondrocytes only (Fig. 9F), whereas expression of MT-2-MMP was increased in mature cultures but decreased in im-mature cultures (Fig. 9, E and F).

DISCUSSION

We show here that misexpression of CA-LEF driven bytype II collagen promoter/enhancer sequences causes seriousdefects in organization, structure, histology, and molecularphenotype in skeletal elements and inhibits endochondralossification in transgenic mice. The chondrocytes remain im-mature, fail to become organized in growth plates, do notundergo an orderly process of maturation, and even fail toassemble typical epiphyseal articular caps. In addition, thecartilage tissue does not establish clear and defined tissueboundaries, resulting in infiltration and invasion by sur-rounding tissues and abnormal distribution of blood vesselsand perichondrial tissues. In contrast, when �-�-catenin isoverexpressed in the developing chick embryo limb long boneanlagen, endochondral ossification and bone deposition areenhanced in diaphyseal regions, whereas immature but dis-organized cartilage persists at the epiphyseal regions. Thedata strongly indicate that the responses of chondrocytes toWnt/�-catenin signaling and the phenotypic and functionalconsequences are strictly dependent on stage/status of thecells at the moment of signal activation. This is reaffirmed bythe very distinct responses to signal activation seen in cul-tures of chick immature and mature chondrocytes. The datacorrelate quite well with our previous data showing thatwhereas much of endogenous �-catenin is cytoplasmic inproliferative and pre-hypertrophic zones of the growth plate,it shifts to a nuclear localization in hypertrophic chondro-cytes (22). It is clear then that Wnt/�-catenin signals regulatechondrocyte behavior and function in a developmentally reg-ulated and maturation-dependent manner and that tight con-trol of this signaling pathway is required for normal chon-drocyte behavior and maturation and normal skeletogenesis.

Chondrocyte De-stabilization and Tissue Boundary Loss—Aprominent and obvious feature of transgenic CA-LEF-express-ing cartilaginous elements is that their chondrocytes are rela-tively uniform in size and shape, do not display the changes incytoarchitecture and volume characteristic of normal growthplates, and are tightly packed together, probably a result ofsuboptimal deposition of extracellular matrix. The tissue itselfis not homogenous and not well separated from adjacent tis-sues. These histological and cellular alterations resemble thoseseen in thanatophoric dysplasia or achondroplasia (44). Interms of gene expression, the transgenic chondrocytes displaypoor expression of such typical genes as Aggrecan, Collagen IX,and Sox9, and virtually undetectable expression of IHH andCollagen X. These data indicate that the transgenic chondrocyteswere not only unable to mature and become hypertrophic, butwere also likely to be undergoing phenotypic de-stabilization. Thelatter is suggested by the poor expression of Sox9, a transcriptionfactor that acts as a master regulator of chondrocyte phenotype(45), a change also documented by Akiyama et al. (20).

In addition to poor maturation, chondrocytes in CA-LEFtransgenic mice express joint marker genes widely and ectopi-cally, including GDF5, GAS1, and ERG. These genes arethought to coordinate multiple events needed for joint forma-tion, including formation of interzone, organization of jointstructures, and maintenance of joint function (27, 34, 35, 46).Our findings are consistent and agree well with recent reportsin which Wnt-14 and �-catenin signaling were found to induceectopic joint marker expression (21, 29). These authors inter-preted the findings to indicate that Wnt-14 and �-catenin arenot only involved in joint formation, but may actually be suffi-cient to induce it. This conclusion, however, may be premature.In addition to inducing ectopic joint markers, overexpressedWnt-14 and active �-catenin actually disturbed joint, particu-larly in the digits where the skeletal elements were fused and

FIG. 6. Expression and distribution of ChM-1 and MMP2. Lon-gitudinal sections of hind limbs from E18.5 CA-LEF transgenic mouseembryos (TG; B, D, and F) or tibias from wild type littermates (A, C, andE) were subjected to in situ hybridization for ChM-1 (A and B) orimmunostaining for MMP2 (C and D). E and F are phase-contrastimages corresponding to C and D, respectively. Bar represents 0.2 mm.

Wnt/�-Catenin Signals and Limb Cartilage Development19192

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

the chondrocytes were uniform and packed together. We ob-served a similar disturbance of joint formation in CA-LEF-expressing transgenic cartilaginous limb elements, but did notdetect de novo joint formation at ectopic sites. Thus, whereasWnt/�-catenin signaling pathway is certainly important forjoint formation, it may be insufficient to induce it and maydepend on upstream mechanisms dictating site and timing ofjoint formation initiation.

For the most part cartilage is an avascular tissue, and inva-sion of cartilage by blood vessels is normally restricted to hy-pertrophic cartilage during endochondral bone deposition.Thus, lack of clear tissue boundaries and apparent intermin-gling of cartilage with surrounding perichondrial/connectivetissues in CA-LEF transgenic mice represent a major depar-ture from normal cartilage physiology. Given the immuno-staining patterns we observe, it is likely that transgenic carti-lage had been invaded by CD34-positive mesenchymal cellsand CD-34/CD-31 vascular endothelial cells. This abnormalprocess may reflect several factors, including: substandard ma-trix deposition that may not constitute an effective barrier;abnormally high and widespread expression of MMP-2 likely to

further degrade and weaken the matrix; and deficient expressionof anti-angiogenic factors such as ChM-1 calcified cartilage. In-terestingly, however, the transgenic tissue does not express In-dian hedgehog in appreciable amounts (Fig. 3J) and did notexpress vascular endothelial growth factor either (not shown).Indian hedgehog has several roles in the growth plate and one islikely to be angiogenic (47). Vascular endothelial growth factor isnormally produced by hypertrophic chondrocytes and is thoughtto be a primary inducer of vascular invasion (48, 49). Thus, itseems likely that vascularization and invasion of transgenic car-tilage are not directly related to the orderly invasion normallyoccurring at the chondro-osseus border during endochondral os-sification, because they appear to be independent of factors suchas IHH, vascular endothelial growth factor, and MMP-13.

Transgenic cartilage invasion might mimic pathological sit-uations such as those seen in joint degenerative disorders,which exhibit sever loss of cartilage matrix and tissue by itselfand invasion by the surrounding tissues. In the present studywe demonstrated that �-�-catenin overexpression stronglystimulates expression of matrix degradation enzymes such asMMPs and MT-MMPs in chondrocytes. We found also that

FIG. 7. Histological, histochemical, and in situ hybridization analyses of chick developing limbs. Longitudinal sections of ulnas fromE11 chick embryos infected with �-�-catenin RCAS virus (�-�-catenin; B, E, F, I, J, M, and N) or inset-less virus (Control; A, C, D, G, H, K, andL) were subjected to H&E staining (A, B, D, and F), Safranin O staining (C and E), in situ hybridization for CD44 (G and I) or gli3 (H and J), orimmunostaining for �-catenin (K–N). Bars represent 0.4 mm for A and B, and 0.2 mm for C–J.

Wnt/�-Catenin Signals and Limb Cartilage Development 19193

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Wnt/�-catenin signal stimulates expression of MMPs and ag-grecanases in mouse chondrogenic cells.2 In osteoarthritis andrheumatoid arthritis, up-regulation of a variety of matrix deg-radation enzymes including MMPs and aggrecanases are cru-cial and considered responsible for cartilage matrix loss (42,50). Interestingly, we found that osteoarthritic cartilage dis-plays signs of Wnt/�-catenin activation.3 Other investigatorshave reported that Wnt-1, an activator of the �-catenin signal,and other Wnts are up-regulated in synovial cells in rheuma-toid arthritis (51), and that functional variants within sFRP3,which is an endogenous antagonist of the Wnt/�-catenin signal,are associated with hip osteoarthritis in females (52). ThusWnt/�-catenin signals may activate cartilage matrix catabo-lism and may have roles in cartilage destruction under patho-logical conditions.

Control of Wnt/�-Catenin Signaling and Action—The data

described here and previous reports by others make it quiteclear that restriction, activation, and modulation of Wnt/�-catenin signaling are crucial for chondrogenesis and skeleto-genesis (20–23). The pathway must be kept relatively inactivefor condensed mesenchymal cells to differentiate into chondro-cytes and for chondrocytes to be able to assemble into welldefined skeletal anlagen with characteristic growth plates. Thepathway, however, must be activated to allow maturing chon-drocytes to complete their development into hypertrophic min-eralizing cells, which in turn orchestrate their replacementwith endochondral bone and marrow. What could then be pos-sible mechanisms activating and modulating Wnt/�-cateninsignaling in developing chondrocytes? It should be pointed outthat a number of Wnt proteins, including Wnt-2b, Wnt-3a,Wnt-4, Wnt-7b, and Wnt-8, are expressed by developing chon-drocytes or neighboring cells in vivo (14, 18, 22, 53, 54) and thatthese proteins all have the ability to activate Wnt/�-cateninsignaling (53–57). In addition, there are well known extracel-lular components that can dampen the degree of signaling byaffecting Wnt binding to their receptors (58). These compo-nents include soluble Frps, Dickkopf (DKK), and WIF. Inter-estingly, onset of Frp3/Frzb-1 expression coincides with onsetof chondrogenesis (59, 60), but expression is down-regulated inthe hypertrophic zone of the growth plate, and Frzb-1 neutral-izes the inhibition of chondrogenesis by misexpression of Wnt-8(22). Thus, it is possible that continuous Frzb-1 expression indeveloping chondrocytes may contribute to prevent Wnt actionand may maintain the cells in a healthy and functional state.

2 Y. Tamamura, T. Otani, M. Iwamoto, and M. Enomoto-Iwamoto,manuscript in preparation.

3 T. Otani, M. Iwamoto, and M. Enomoto-Iwamoto, submitted forpublication.

FIG. 8. Differential effects of �-�-catenin and Wnt-14 on chon-drocyte phenotype and maturation. A–G, freshly isolated imma-ture and mature chondrocytes were infected with RCAS virus encoding�-�-catenin or Wnt-14, or insert-less virus (Control). After 3 days cellswere re-plated at a density of 1.0–1.5 � 105 cells/cm2 and cultured inDulbecco’s modified Eagle’s medium containing 10% fetal bovine serumand 10 �g/ml ascorbic acid an additional 4 days. On the last day,cultures received 1 mM �-glycerophosphate to induce matrix calcifica-tion. Cultures were examined by phase microscopy (A–F) and were thenstained with Alcian blue (AB) or alizarin red (AR) (G). H, immature andmature chondrocytes were transfected with the indicated DNA plas-mids along with Axin-2 reporter DNA 2 days after plating, and lucifer-ase activity was measured 2 days later.

FIG. 9. Gene expression in chondrocyte cultures. Freshly iso-lated immature (A, C, and E) and mature chondrocytes (B, D, and F)were infected with RCAS viruses encoding �-�-catenin or Wnt-14 orwith insert-less virus (Control). Three days later, cells were re-plated ata density of 1.0–1.5 � 105 cells/cm2 and cultured in Dulbecco’s modifiedEagle’s medium containing 10% fetal bovine serum and 10 �g/ml ascor-bic acid for an additional 4 days. Total RNAs were subjected to RT-PCRanalysis for chondrocyte phenotypic markers (A and B), markers ofmaturation (C and D), and matrix degradation enzymes (E and F).

Wnt/�-Catenin Signals and Limb Cartilage Development19194

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

The cells would be able to form well defined cartilaginousanlagen, organize growth plates, and establish well definedtissue boundaries. Frzb-1 down-regulation in the hypertrophiczone, however, would allow strong Wnt action, permitting thecells to terminally differentiate and favoring the normal tran-sition from hypertrophic cartilage to bone.

An additional possible regulator of Wnt/�-catenin signalingpathway actions and roles is Wnt-5a. This Wnt is first ex-pressed in cartilage anlagen and expression shifts to perichon-drium with further development (16, 19). Deletion of theWnt-5a gene in mice causes a phenotype quite similar to thatseen in CA-LEF transgenic mice here (17). Wnt-5a-null mouseembryos are dwarfs with shortened and deformed limbs, anddisplay cartilaginous elements with low expression of Sox9,IHH, PTH/PTHrp receptor, and Type X collagen and no endo-chondral ossification (17). Wnt-5a has been shown to inhibitthe Wnt/�-catenin pathway through the calmodulin-dependentprotein kinase II pathway or an unknown pathway that inhib-its �-catenin degradation (61). These findings suggest thatWnt-5a could serve as a strong inhibitor of Wnt/�-catenin sig-naling during cartilage development. Indeed the limb buds ofWnt-5a-null mouse embryos display ectopic up-regulation ofWnt/�-catenin as revealed by promoter reporter activity (61).Thus, Wnt-5a may be part of the normal mechanisms involvingalso Frzb-1 by which Wnt/�-catenin signaling is maintainedrelatively low at the initial stages of skeletogenesis and carti-lage development, and is then markedly up-regulated in hyper-trophic mineralizing cartilage.

Last, it is important to discuss Sox9. The factor is continu-ously expressed starting with pre-cartilaginous mesenchymalcells up to pre-hypertrophic chondrocytes, and is then down-regulated in hypertrophic cells (62). Sox9 inhibits Wnt/�-cate-nin signals by competing with �-catenin binding to LEF/TCFproteins, and experimental ablation of Sox9 in cartilage causesa phenotype similar to that following chronic activation of theWnt/�-catenin pathway (20). Because Sox9 expression precedesdifferentiation of mesenchymal cells into chondrocytes, thismolecule also could antagonize the Wnt/�-catenin signal fromonset of cartilage formation up until the pre-hypertrophicstage. Terminal hypertrophic chondrocyte maturation wouldthus entail a down-regulation of Sox9, in coordination withdown-regulations in Wnt-5a and Frzb1 and concurrent boost inWnt/�-catenin signaling.

Acknowledgments—We thank Dr. C. L. Cepko (Harvard MedicalSchool) for RCASBP(A), Dr. A. Hecht (Max Plank Institute of Immu-nology) for LEF mutant constructs, Dr. C. Tabin (Harvard MedicalSchool) for chick Wnt-14 RCAS, and Dr. T. Yochida (Mie University) foranti-tenascin antibody.

REFERENCES

1. Horton, W. A. (1993) in Connective Tissue and Its Heritable Disorders (Royce,P. M., and Steinmann, B., eds) pp. 73–84, Wiley-Liss, Inc., New York

2. Karsenty, G. (2003) Nature 423, 316–3183. Pacifici, M., Koyama, E., Iwamoto, M., and Gentili, C. (2000) Connect. Tissue

Res. 41, 17–7844. Nusse, R., and Varmus, H. E. (1992) Cell 69, 1073–10875. Huelsken, J., and Birchmeier, W. (2001) Curr. Opin. Genet. Dev. 11, 547–5536. Eastman, Q., and Grosschedl, R. (1999) Curr. Opin. Cell. Biol. 11, 233–2407. Miller, J. R., Hocking, A. M., Brown, J. D., and Moon, R. T. (1999) Oncogene 18,

7860–78728. Veeman, M. T., Axelrod, J. D., and Moon, R. T. (2003) Dev. Cell 5, 367–3779. Nakamura, T., Hamada, F., Ishidate, T., Anai, K., Kawahara, K., Toyoshima,

K., and Akiyama, T. (1998) Genes Cells 3, 395–40310. Kishida, S., Yamamoto, H., Ikeda, S., Kishida, M., Sakamoto, I., Koyama, S.,

and Kikuchi, A. (1998) J. Biol. Chem. 273, 10823–1082611. Aberle, H., Bauer, A., Stappert, J., Kispert, A., and Kemler, R. (1997) EMBO

J. 16, 3797–380412. Willert, K., and Nusse, R. (1998) Curr. Opin. Genet. Dev. 8, 95–10213. Dealy, C. N., Roth, A., Ferrari, D., Brown, A. M., and Kosher, R. A. (1993)

Mech. Dev. 43, 175–18614. Kengaku, M., Capdevila, J., Rodriguez-Esteban, C., De La Pena, J., Johnson,

R. L., Belmonte, J. C., and Tabin, C. J. (1998) Science 280, 1274–127715. Rudnicki, J. A., and Brown, A. M. (1997) Dev. Biol. 185, 104–11816. Church, V., Nohno, T., Linker, C., Marcelle, C., and Francis-West, P. (2002)

J. Cell Sci. 115, 4809–481817. Yang, Y., Topol, L., Lee, H., and Wu, J. (2003) Development 130, 1003–101518. Hartmann, C., and Tabin, C. J. (2000) Development 127, 3141–315919. Kawakami, Y., Wada, N., Nishimatsu, S. I., Ishikawa, T., Noji, S., and Nohno,

T. (1999) Dev. Growth Differ. 41, 29–4020. Akiyama, H., Lyons, J. P., Mori-Akiyama, Y., Yang, X., Zhang, R., Zhang, Z.,

Deng, J. M., Taketo, M. M., Nakamura, T., Behringer, R. R., McCrea, P. D.,and de Crombrugghe, B. (2004) Genes Dev. 18, 1072–1087

21. Guo, X., Day, T. F., Jiang, X., Garrett-Beal, L., Topol, L., and Yang, Y. (2004)Genes Dev. 18, 2404–2417

22. Enomoto-Iwamoto, M., Kitagaki, J., Koyama, E., Tamamura, Y., Wu, C.,Kanatani, N., Koike, T., Okada, H., Komori, T., Yoneda, T., Church, V.,Francis-West, P. H., Kurisu, K., Nohno, T., Pacifici, M., and Iwamoto, M.(2002) Dev. Biol. 251, 142–156

23. Kitagaki, J., Iwamoto, M., Liu, J. G., Tamamura, Y., Pacifici, M., and Enomoto-Iwamoto, M. (2003) Osteoarthritis Cartilage 11, 36–43

24. Vleminckx, K., Kemler, R., and Hecht, A. (1999) Mech. Dev. 81, 65–7425. Ueta, C., Iwamoto, M., Kanatani, N., Yoshida, C., Liu, Y., Enomoto-Iwamoto,

M., Ohmori, T., Enomoto, H., Nakata, K., Takada, K., Kurisu, K., andKomori, T. (2001) J. Cell Biol. 153, 87–100

26. Kalembey, I., Yoshida, T., Iriyama, K., and Sakakura, T. (1997) Int. J. Dev.Biol. 41, 569–573

27. Iwamoto, M., Higuchi, Y., Koyama, E., Enomoto-Iwamoto, M., Kurisu, K., Yeh,H., Abrams, W. R., Rosenbloom, J., and Pacifici, M. (2000) J. Cell Biol. 150,27–40

28. Pacifici, M., Golden, E. B., Adams, S. L., and Shapiro, I. M. (1991) Exp. CellRes. 192, 266–270

29. Hartmann, C., and Tabin, C. J. (2001) Cell 104, 341–35230. Enomoto-Iwamoto, M., Iwamoto, M., Mukudai, Y., Kawakami, Y., Nohno, T.,

Higuchi, Y., Takemoto, S., Ohuchi, H., Noji, S., and Kurisu, K. (1998) J. CellBiol. 140, 409–418

31. Iwamoto, M., Shapiro, I. M., Yagami, K., Boskey, A. L., Leboy, P. S., Adams,S. L., and Pacifici, M. (1993) Exp. Cell Res. 207, 413–420

32. Jho, E. H., Zhang, T., Domon, C., Joo, C. K., Freund, J. N., and Costantini, F.(2002) Mol. Cell. Biol. 22, 1172–1183

33. Iwamoto, M., Kitagaki, J., Tamamura, Y., Gentili, C., Koyama, E., Enomoto,H., Komori, T., Pacifici, M., and Enomoto-Iwamoto, M. (2003) OsteoarthritisCartilage 11, 6–15

34. Lee, K. K., Leung, A. K., Tang, M. K., Cai, D. Q., Schneider, C., Brancolini, C.,and Chow, P. H. (2001) Dev. Biol. 234, 188–203

35. Dhordain, P., Dewitte, F., Desbiens, X., Stehelin, D., and Duterque-Coquil-laud, M. (1995) Mech. Dev. 50, 17–28

36. Pacifici, M., Iwamoto, M., Golden, E. B., Leatherman, J. L., Lee, Y. S., andChuong, C. M. (1993) Dev. Dyn. 198, 123–134

37. Hiraki, Y., Tanaka, H., Inoue, H., Kondo, J., Kamizono, A., and Suzuki, F.(1991) Biochem. Biophys. Res. Commun. 175, 971–977

38. Jain, R. K. (2003) Nat. Med. 9, 685–69339. Edwards, J. C., Wilkinson, L. S., Jones, H. M., Soothill, P., Henderson, K. J.,

Worrall, J. G., and Pitsillides, A. A. (1994) J. Anat. 185, 355–36740. Iwasaki, M., Le, A. X., and Helms, J. A. (1997) Mech. Dev. 69, 197–20241. Visse, R., and Nagase, H. (2003) Circ. Res. 92, 827–83942. Caterson, B., Flannery, C. R., Hughes, C. E., and Little, C. B. (2000) Matrix

Biol. 19, 333–34443. Cawston, T., Billington, C., Cleaver, C., Elliott, S., Hui, W., Koshy, P., Shingle-

ton, B., and Rowan, A. (1999) Ann. N. Y. Acad. Sci. 878, 120–12944. Horton, W. A., and Hecht, J. T. (1993) in Connective Tissue Heritable and Its

Disorders (Royce, P. M., and Steinmann, B., eds) pp. 641–676, Wiley-Liss,Inc., New York

45. de Crombrugghe, B., Lefebvre, V., and Nakashima, K. (2001) Curr. Opin. CellBiol. 13, 721–727

46. Storm, E. E., and Kingsley, D. M. (1999) Dev. Biol. 209, 11–2747. Kanda, S., Mochizuki, Y., Suematsu, T., Miyata, Y., Nomata, K., and Kan-

etake, H. (2003) J. Biol. Chem. 278, 8244–824948. Gerber, H. P., Vu, T. H., Ryan, A. M., Kowalski, J., Werb, Z., and Ferrara, N.

(1999) Nat. Med. 5, 623–62849. Maes, C., Stockmans, I., Moermans, K., Van Looveren, R., Smets, N., Carme-

liet, P., Bouillon, R., and Carmeliet, G. (2004) J. Clin. Investig. 113, 188–19950. Sandell, L. J., and Aigner, T. (2001) Arthritis Res. 3, 107–11351. Sen, M., Lauterbach, K., El-Gabalawy, H., Firestein, G. S., Corr, M., and

Carson, D. A. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 2791–279652. Loughlin, J., Dowling, B., Chapman, K., Marcelline, L., Mustafa, Z., Southam,

L., Ferreira, A., Ciesielski, C., Carson, D. A., and Corr, M. (2004) Proc. Natl.Acad. Sci. U. S. A. 101, 9757–9762

53. Kawakami, Y., Capdevillar, J., Bilscher, D., Itoh, T., Esteban, J. C., andBelmonte, J. C. (2001) Cell 104, 891–900

54. Hu, H., Hilton, M. J., Tu, X., Yu, K., Ornitz, D. M., and Long, F. (2004)Development 132, 49–60

55. Kishida, M., Koyama, S., Kishida, S., Matsubara, K., Nakashima, S., Higano,K., Takada, R., Takada, S., and Kikuchi, A. (1999) Oncogene 18, 979–985

56. Surendran, K., McCaul, S. P., and Simon, T. C. (2002) Am. J. Physiol. 282,F431–F441

57. Hsieh, J. C., Rattner, A., Smallwood, P. M., and Nathans, J. (1999) Proc. Natl.Acad. Sci. U. S. A. 96, 3546–3551

58. Kawano, Y., and Kypta, R. (2003) J. Cell Sci. 116, 2627–263459. Wada, N., Kawakami, Y., Ladher, R., Francis-West, P. H., and Nohno, T.

(1999) Int. J. Dev. Biol. 43, 495–50060. Hoang, B. H., Thomas, J. T., Abdul-Karim, F. W., Correia, K. M., Conlon, R. A.,

Luyten, F. P., and Ballock, R. T. (1998) Dev. Dyn. 212, 364–37261. Topol, L., Jiang, X., Choi, H., Garrett-Beal, L., Carolan, P. J., and Yang, Y.

(2003) J. Cell Biol. 162, 899–90862. Bi, W., Deng, J. M., Zhang, Z., Behringer, R. R., and de Crombrugghe, B. (1999)

Nat. Genet. 22, 85–89

Wnt/�-Catenin Signals and Limb Cartilage Development 19195

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Pacifici, Masahiro Iwamoto and Motomi Enomoto-IwamotoToshihisa Komori, Yoshihiko Yamada, Frank Costantini, Satoshi Wakisaka, Maurizio

Yoshihiro Tamamura, Tomohiro Otani, Naoko Kanatani, Eiki Koyama, Jirota Kitagaki,Assembly, Cartilage Integrity, and Endochondral Ossification

-Catenin Signals Is Required for Growth PlateβDevelopmental Regulation of Wnt/

doi: 10.1074/jbc.M414275200 originally published online March 10, 20052005, 280:19185-19195.J. Biol. Chem. 

  10.1074/jbc.M414275200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/280/19/19185.full.html#ref-list-1

This article cites 60 references, 18 of which can be accessed free at

by guest on April 9, 2020

http://ww

w.jbc.org/

Dow

nloaded from