Expression and function of CCN2-derived circRNAs in chondrocytes.

Cellular communication network factor 2 (CCN2) molecules promote endochondral ossification and articular cartilage regeneration, and circular RNAs (circRNAs), which arise from various genes and regulate gene expression by adsorbing miRNAs, are known to be synthesized from CCN2 in human vascular endothelial cells and other types of cells. However, in chondrocytes, not only the function but also the presence of CCN2-derived circRNA remains completely unknown. In the present study, we investigated the expression and function of CCN2-derived circRNAs in chondrocytes. Amplicons smaller than those from known CCN2-derived circRNAs were observed using RT-PCR analysis that could specifically amplify CCN2-derived circRNAs in human chondrocytic HCS-2/8 cells. The nucleotide sequences of the PCR products indicated novel circRNAs in the HCS-2/8 cells that were different from known CCN2-derived circRNAs. Moreover, the expression of several Ccn2-derived circRNAs in murine chondroblastic ATDC5 cells was confirmed and observed to change alongside chondrocytic differentiation. Next, one of these circRNAs was knocked down in HCS-2/8 cells to investigate the function of the human CCN2-derived circRNA. As a result, CCN2-derived circRNA knockdown significantly reduced the expression of aggrecan mRNA and proteoglycan synthesis. Our data suggest that CCN2-derived circRNAs are expressed in chondrocytes and play a role in chondrogenic differentiation. Production and role of CCN2-derived RNAs in chondrocytes.

Do not overwork: cellular communication network factor 3 for life in cartilage.

Cellular communication network factor (CCN) 3, which is one of the founding members of the CCN family, displays diverse functions. However, this protein generally represses the proliferation of a variety of cells. Along with skeletal development, CCN3 is produced in cartilaginous anlagen, growth plate cartilage and epiphysial cartilage. Interestingly, CCN3 is drastically induced in the growth plates of mice lacking CCN2, which promotes endochondral ossification. Notably, chondrocytes in these mutant mice with elevated CCN3 production also suffer from impaired glycolysis and energy metabolism, suggesting a critical role of CCN3 in cartilage metabolism. Recently, CCN3 was found to be strongly induced by impaired glycolysis, and in our study, we located an enhancer that mediated CCN3 regulation via starvation. Subsequent investigations specified regulatory factor binding to the X-box 1 (RFX1) as a transcription factor mediating this CCN3 regulation. Impaired glycolysis is a serious problem, resulting in an energy shortage in cartilage without vasculature. CCN3 produced under such starved conditions restricts energy consumption by repressing cell proliferation, leading chondrocytes to quiescence and survival. This CCN3 regulatory system is indicated to play an important role in articular cartilage maintenance, as well as in skeletal development. Furthermore, CCN3 continues to regulate cartilage metabolism even during the aging process, probably utilizing this regulatory system. Altogether, CCN3 seems to prevent "overwork" by chondrocytes to ensure their sustainable life in cartilage by sensing energy metabolism. Similar roles are suspected to exist in relation to systemic metabolism, since CCN3 is found in the bloodstream.

Gene Expression Analysis of CCNs Along with Odontoblastic Differentiation In Vivo.

Dental pulp cells (DPCs) differentiate into odontoblasts. To observe odontoblastic differentiation, the detection of dentinogenesis-specific molecules such as dentin sialophosphoprotein (DSPP) and the measurement of alkaline phosphatase (ALP) activity are reliable approaches. CCN family member 2 (CCN2) has been proposed as a marker for dentinogenesis. Our recent study revealed that the expression levels of Ccn4, Ccn5, and Ccn6 were changed in accordance with odontoblastic differentiation. Therefore, Ccn4, Ccn5, and Ccn6, as well as Ccn2, could serve as a comprehensive set of markers for dentinogenesis. Here, we describe a method of measuring the Ccns expression levels in differentiating rat DPCs.

Molecular and Genetic Interactions between CCN2 and CCN3 behind Their Yin-Yang Collaboration.

Cellular communication network factor (CCN) 2 and 3 are the members of the CCN family that conduct the harmonized development of a variety of tissues and organs under interaction with multiple biomolecules in the microenvironment. Despite their striking structural similarities, these two members show contrastive molecular functions as well as temporospatial emergence in living tissues. Typically, CCN2 promotes cell growth, whereas CCN3 restrains it. Where CCN2 is produced, CCN3 disappears. Nevertheless, these two proteins collaborate together to execute their mission in a yin-yang fashion. The apparent functional counteractions of CCN2 and CCN3 can be ascribed to their direct molecular interaction and interference over the cofactors that are shared by the two. Recent studies have revealed the mutual negative regulation systems between CCN2 and CCN3. Moreover, the simultaneous and bidirectional regulatory system of CCN2 and CCN3 is also being clarified. It is of particular note that these regulations were found to be closely associated with glycolysis, a fundamental procedure of energy metabolism. Here, the molecular interplay and metabolic gene regulation that enable the yin-yang collaboration of CCN2 and CCN3 typically found in cartilage development/regeneration and fibrosis are described.

Cellular communication network factor 3 in cartilage development and maintenance.

Cellular communication network factor (CCN) 3 is one of the classical members of the CCN family, which are characterized by common molecular structures and multiple functionalities. Although this protein was discovered as a gene product overexpressed in a truncated form in nephroblastoma, recent studies have revealed its physiological roles in the development and homeostasis of mammalian species, in addition to its pathological association with a number of diseases. Cartilage is a tissue that creates most of the bony parts and cartilaginous tissues that constitute the human skeleton, in which CCN3 is also differentially produced to exert its molecular missions therein. In this review article, after the summary of the molecular structure and function of CCN3, recent findings on the regulation of ccn3 expression and the roles of CCN3 in endochondral ossification, cartilage development, maintenance and disorders are introduced with an emphasis on the metabolic regulation and function of this matricellular multifunctional molecule.

Odontoblast differentiation is regulated by an interplay between primary cilia and the canonical Wnt pathway.

Primary cilium is a protruding cellular organelle that has various physiological functions, especially in sensory reception. While an avalanche of reports on primary cilia have been published, the function of primary cilia in dental cells remains to be investigated. In this study, we focused on the function of primary cilia in dentin-producing odontoblasts. Odontoblasts, like most other cell types, possess primary cilia, which disappear upon the knockdown of intraflagellar transport protein 88. In cilia-depleted cells, the expression of dentin sialoprotein, an odontoblastic marker, was elevated, while the deposition of minerals was slowed. This was recapitulated by the activation of canonical Wnt pathway, also decreased the ratio of ciliated cells. In dental pulp cells, as they differentiated into odontoblasts, the ratio of ciliated cells was increased, whereas the canonical Wnt signaling activity was repressed. Our results collectively underscore the roles of primary cilia in regulating odontoblastic differentiation through canonical Wnt signaling. This study implies the existence of a feedback loop between primary cilia and the canonical Wnt pathway.

RFX1-mediated CCN3 induction that may support chondrocyte survival under starved conditions.

Cellular communication network factor (CCN) family members are multifunctional matricellular proteins that manipulate and integrate extracellular signals. In our previous studies investigating the role of CCN family members in cellular metabolism, we found three members that might be under the regulation of energy metabolism. In this study, we confirmed that CCN2 and CCN3 are the only members that are tightly regulated by glycolysis in human chondrocytic cells. Interestingly, CCN3 was induced under a variety of impaired glycolytic conditions. This CCN3 induction was also observed in two breast cancer cell lines with a distinct phenotype, suggesting a basic role of CCN3 in cellular metabolism. Reporter gene assays indicated a transcriptional regulation mediated by an enhancer in the proximal promoter region. As a result of analyses in silico, we specified regulatory factor binding to the X-box 1 (RFX1) as a candidate that mediated the transcriptional activation by impaired glycolysis. Indeed, the inhibition of glycolysis induced the expression of RFX1, and RFX1 silencing nullified the CCN3 induction by impaired glycolysis. Subsequent experiments with an anti-CCN3 antibody indicated that CCN3 supported the survival of chondrocytes under impaired glycolysis. Consistent with these findings in vitro, abundant CCN3 production by chondrocytes in the deep zones of developing epiphysial cartilage, which are located far away from the synovial fluid, was confirmed in vivo. Our present study uncovered that RFX1 is the mediator that enables CCN3 induction upon cellular starvation, which may eventually assist chondrocytes in retaining their viability, even when there is an energy supply shortage.

Bipartite regulation of cellular communication network factor 2 and fibroblast growth factor 1 genes by fibroblast growth factor 1 through histone deacetylase 1 and fork head box protein A1.

Fibroblast growth factor 1 (FGF-1) is the first FGF family member, and it induces proliferation of fibroblasts and other types of the cells. However, recent studies are uncovering unexpected functions of this molecule. Our previous study redefined this growth factor as a catabolic molecule produced in cartilage upon metabolic insult. Indeed, FGF-1 was found to repress the gene expression of cellular communication network factor 2 (CCN2), which protects and regenerates cartilage, amplifying its own production through positive feedback regulation. In the present study, we investigated the molecular mechanism of this bipartite CCN2 repression and FGF1 activation by FGF-1 in chondrocytes. Repression of CCN2 and induction of FGF1 in human chondrocytic cells were both partly abolished by valproic acid, an inhibitor of histone deacetylase 1 (HDAC1), indicating the involvement of chromatin remodeling by histone acetylation in this system. In contrast, RNA degradation analysis suggested no contribution of post-transcriptional regulation of the mRNA stability to the effects conferred by FGF-1. Suspecting a regulation by a specific transcription factor, we next sought a candidate in silico from a large dataset. As a result, we found fork head box protein A1 (FOXA1) as the transcription factor that bound to both CCN2 and FGF1 loci. Functional analysis demonstrated that FOXA1 silencing significantly attenuated the CCN2 repression and FGF1 induction caused by FGF1. These findings collectively indicate that the bipartite regulation by FGF-1 is enabled by the combination of chromatin remodeling by HDACs and transcriptional modulation by FOXA1 with unknown transcriptional coactivators of opposite functionalities.

Regulation of cellular communication network factor 2 (CCN2) in breast cancer cells via the cell-type dependent interplay between CCN2 and glycolysis.

OBJECTIVES: Anti-osteoclastic treatments for breast cancer occasionally cause medication-related osteonecrosis of the jaw. Moreover, elevated glycolytic activity, which is known as the Warburg effect, is usually observed in these breast cancer cells. Previously, we found that cellular communication network factor 2 (CCN2) production and glycolysis enhanced each other in chondrocytes. Here, we evaluated the interplay between CCN2 and glycolysis in breast cancer cells, as we suspected a possible involvement of CCN2 in the Warburg effect in highly invasive breast cancer cells. METHODS: Two human breast cancer cell lines with a distinct phenotype were used. Glycolysis was inhibited by using 2 distinct compounds, and gene silencing was performed using siRNA. Glycolysis and the expression of relevant genes were monitored via colorimetric assays and quantitative RT-PCR, respectively. RESULTS: Although CCN2 expression was almost completely silenced when treating invasive breast cancer cells with a siRNA cocktail against CCN2, glycolytic activity was not affected. Notably, the expression of glycolytic enzyme genes, which was repressed by CCN2 deficiency in chondrocytes, tended to increase upon CCN2 silencing in breast cancer cells. Inhibition of glycolysis, which resulted in the repression of CCN2 expression in chondrocytic cells, did not alter or strongly enhanced CCN2 expression in the invasive and non-invasive breast cancer cells, respectively. CONCLUSIONS: High CCN2 expression levels play a critical role in the invasion and metastasis of breast cancer. Thus, a collapse in the intrinsic repressive machinery of CCN2 due to glycolysis may induce the acquisition of an invasive phenotype in breast cancer cells.

Retrotransposons Manipulating Mammalian Skeletal Development in Chondrocytes.

Retrotransposons are genetic elements that copy and paste themselves in the host genome through transcription, reverse-transcription, and integration processes. Along with their proliferation in the genome, retrotransposons inevitably modify host genes around the integration sites, and occasionally create novel genes. Even now, a number of retrotransposons are still actively editing our genomes. As such, their profound role in the evolution of mammalian genomes is obvious; thus, their contribution to mammalian skeletal evolution and development is also unquestionable. In mammals, most of the skeletal parts are formed and grown through a process entitled endochondral ossification, in which chondrocytes play central roles. In this review, current knowledge on the evolutional, physiological, and pathological roles of retrotransposons in mammalian chondrocyte differentiation and cartilage development is summarized. The possible biological impact of these mobile genetic elements in the future is also discussed.

A Tumor Suppressor Gene Product, Platelet-Derived Growth Factor Receptor-Like Protein Controls Chondrocyte Proliferation and Differentiation.

The platelet-derived growth factor receptor-like (PDGFRL) gene is regarded as a tumor suppressor gene. However, nothing is known about the molecular function of PDGFRL. In this study, we initially clarified its function in chondrocytes. Among all cell lines examined, the PDGFRL mRNA level was the highest in chondrocytic HCS-2/8 cells. Interestingly, the proliferation of chondrocytic HCS-2/8 cells was promoted by PDGFRL overexpression, whereas that of the breast cancer-derived MDA-MB-231 cells was inhibited. Of note, in PDGFRL-overexpressing HCS-2/8 cells, the expression of chondrocyte differentiation marker genes, SOX9, ACAN, COL2A1, COL10A1, and ALP, was decreased. Moreover, we confirmed the expression of PDGFRL mRNA in normal cartilage tissue and chondrocytes. Eventually, the expression of PDGFRL mRNA in condrocytes except in the case of hypertrophic chondrocytes was demonstrated in vivo and in vitro. These findings suggest that PDGFRL plays the different roles, depending upon cell types. Particularly, in chondrocytes, PDGFRL may play a new and important role which is distinct from the function previously reported. J. Cell. Biochem. 118: 4033-4044, 2017. © 2017 Wiley Periodicals, Inc.

Analysis of Transcytosis of CCN2 by Chondrocytes.

Transcytosis is a mechanism for the transcellular transport of biomolecules. Analysis of transcytosis is frequently performed in cells with distinct polarity, such as brain endothelial cells. However, in cells without evident polarity, analysis of transcytosis has not been performed. Here, we describe a method for analyzing transcytosis of a CCN family protein through chondrocytic cells having no apparent polarity.

CCN2 as a novel molecule supporting energy metabolism of chondrocytes.

CCN2/connective tissue growth factor (CTGF) is a unique molecule that promotes both chondrocytic differentiation and proliferation through its matricellular interaction with a number of extracellular biomolecules. This apparently contradictory functional property of CCN2 suggests its certain role in basic cellular activities such as energy metabolism, which is required for both proliferation and differentiation. Comparative metabolomic analysis of costal chondrocytes isolated from wild-type and Ccn2-null mice revealed overall impaired metabolism in the latter. Among the numerous metabolites analyzed, stable reduction in the intracellular level of ATP, GTP, CTP, or UTP was observed, indicating a profound role of CCN2 in energy metabolism. Particularly, the cellular level of ATP was decreased by more than 50% in the Ccn2-null chondrocytes. The addition of recombinant CCN2 (rCCN2) to cultured Ccn2-null chondrocytes partly redeemed the cellular ATP level attenuated by Ccn2 deletion. Next, in order to investigate the mechanistic background that mediates the reduction in ATP level in these Ccn2-null chondrocytes, we performed transcriptome analysis. As a result, several metabolism-associated genes were found to have been up-regulated or down-regulated in the mutant mice. Up-regulation of a number of ribosomal protein genes was observed upon Ccn2 deletion, whereas a few genes required for aerobic and anaerobic ATP production were down-regulated in the Ccn2-null chondrocytes. Among such genes, reduction in the expression of the enolase 1 gene was of particular note. These findings uncover a novel functional role of CCN2 as a metabolic supporter in the growth-plate chondrocytes, which is required for skeletogenesis in mammals.

Novel role of miR-181a in cartilage metabolism.

Micro RNA (miRNA) is a small non-coding post-transcriptional RNA regulator that is involved in a variety of biological events. In order to specify the role of miRNAs in cartilage metabolism, we comparatively analyzed the expression profile of known miRNAs in chicken sternum chondrocytes representing early and late differentiation stages. Interestingly, none of the miRNAs displaying strong expression levels showed remarkable changes along with differentiation, suggesting their roles in maintaining the homeostasis rather than cytodifferentiation of chondrocytes. Among these miRNAs, miR-181a, which is known to play critical roles in a number of tissues, was selected and was further characterized. Human microarray analysis revealed remarkably stronger expression of miR-181a in human HCS-2/8 cells, which strongly maintained a chondrocytic phenotype, than in HeLa cells, indicating its significant role in chondrocytes. Indeed, subsequent investigation indicated that miR-181a repressed the expression of two genes involved in cartilage development. One was CCN family member 1 (CCN1), which promotes chondrogenesis; and the other, the gene encoding the core protein of aggrecan, a major cartilaginous proteoglycan, aggrecan. Based on these findings, negative feedback system via miR-181a to conserve the integrity of the cartilaginous phenotype may be proposed.

Role of LRP1 in transport of CCN2 protein in chondrocytes.

Low-density lipoprotein receptor-related protein 1 (LRP1) is known to be a receptor for signal transmission and endocytosis. We have previously reported that LRP1 regulates WNT-ß-catenin and protein kinase C signaling in chondrocytes, represses the hypertrophy of chondrocytes during endochondral ossification and that LRP1 is colocalized with a ligand, CCN family member 2 (CCN2; also known as connective tissue growth factor, CTGF), which conducts endochondral ossification, in chondrocytes. However, the role of LRP1 in the endocytic transport of CCN2 in chondrocytes is not yet understood. In the present study, we investigated the interaction between LRP1 and CCN2 during endocytic trafficking. Small interfering RNA (siRNA)-mediated knockdown of LRP1 in chondrocytic HCS-2/8 cells showed that the amount of exogenous CCN2 binding and/or incorporation was decreased in the LRP1 downregulated cells. Importantly, we observed that CCN2 internalization in chondrocytes was dependent on clathrin, and internalizated CCN2 was colocalized with an early or recycling endosome marker. Transcytosis of CCN2 through HCS-2/8 cells was confirmed by performing experiments with a trans-well apparatus, and the amount of transcytosed CCN2 was decreased by an LRP1 antagonist. These findings rule out possible leakage and confirm the crucial involvement of LRP1 during experimental transcytosis. Moreover, under hypoxic conditions that mimic the cartilaginous microenvironment, the level of LRP1 and the amount of transcytosed CCN2 increased, and these increases were neutralized by treatment with the LRP1 antagonist. The distribution of LRP1 and its antagonist in the growth plate in vivo was consistent with that of CCN2 in this tissue, which is produced by and transported by LRP1 from the chondrocytes in the prehypertrophic layer. These findings suggest that LRP1 mediates the transcytosis of CCN2, which might be a crucial event that determines the distribution of CCN2 in cartilage.

Mechanical stretch increases CCN2/CTGF expression in anterior cruciate ligament-derived cells.

Anterior cruciate ligament (ACL)-to-bone interface serves to minimize the stress concentrations that would arise between two different tissues. Mechanical stretch plays an important role in maintaining cell-specific features by inducing CCN family 2/connective tissue growth factor (CCN2/CTGF). We previously reported that cyclic tensile strain (CTS) stimulates α1(I) collagen (COL1A1) expression in human ACL-derived cells. However, the biological function and stress-related response of CCN2/CTGF were still unclear in ACL fibroblasts. In the present study, CCN2/CTGF was observed in ACL-to-bone interface, but was not in the midsubstance region by immunohistochemical analyses. CTS treatments induced higher increase of CCN2/CTGF expression and secretion in interface cells compared with midsubstance cells. COL1A1 expression was not influenced by CCN2/CTGF treatment in interface cells despite CCN2/CTGF stimulated COL1A1 expression in midsubstance cells. However, CCN2/CTGF stimulated the proliferation of interface cells. Our results suggest that distinct biological function of stretch-induced CCN2/CTGF might regulate region-specific phenotypes of ACL-derived cells.

Identification of miR-1 as a micro RNA that supports late-stage differentiation of growth cartilage cells.

The process of endochondral ossification is strictly regulated by a variety of extracellular and intracellular factors. Recently, it has become recognized that specific miRNAs are involved in this process by regulating the expression of the relevant genes at the post-transcriptional level. In this present study we obtained the first evidence of the involvement of a specific micro RNA (miRNA) in the regulation of the chondrocyte phenotype during late stages of differentiation. By use of the microarray technique, miR-1 was identified as this miRNA, the expression of which was most repressed upon hypertrophic differentiation. Transfection of human chondrocytic HCS-2/8 cells and chicken normal chondrocytes with miR-1 led to repressed expression of aggrecan, the major cartilaginous proteoglycan gene. Therefore, miR-1 was found to be involved in the regulation of the chondrocytic phenotype and thus to play an important role in chondrocytes during the late stage of the differentiation process, maintaining the integrity of the cartilage tissue.

Thrombopoietic-mesenchymal interaction that may facilitate both endochondral ossification and platelet maturation via CCN2.

CCN2 plays a central role in the development and growth of mesenchymal tissue and promotes the regeneration of bone and cartilage in vivo. Of note, abundant CCN2 is contained in platelets, which is thought to play an important role in the tissue regeneration process. In this study, we initially pursued the possible origin of the CCN2 in platelets. First, we examined if the CCN2 in platelets was produced by megakaryocyte progenitors during differentiation. Unexpectedly, neither megakaryocytic CMK cells nor megakaryocytes that had differentiated from human haemopoietic stem cells in culture showed any detectable CCN2 gene expression or protein production. Together with the fact that no appreciable CCN2 was detected in megakaryocytes in vivo, these results suggest that megakaryocytes themselves do not produce CCN2. Next, we suspected that mesenchymal cells situated around megakaryocytes in the bone marrow were stimulated by the latter to produce CCN2, which was then taken up by platelets. To evaluate this hypothesis, we cultured human chondrocytic HCS-2/8 cells with medium conditioned by differentiating megakaryocyte cultures, and then monitored the production of CCN2 by the cells. As suspected, CCN2 production by HCS-2/8 was significantly enhanced by the conditioned medium. We further confirmed that human platelets were able to absorb/uptake exogenous CCN2 in vitro. These findings indicate that megakaryocytes secrete some unknown soluble factor(s) during differentiation, which factor stimulates the mesenchymal cells to produce CCN2 for uptake by the platelets. We also consider that, during bone growth, such thrombopoietic-mesenchymal interaction may contribute to the hypertrophic chondrocyte-specific accumulation of CCN2 that conducts endochondral ossification.

Role of the low-density lipoprotein receptor-related protein-1 in regulation of chondrocyte differentiation.

The low-density lipoprotein receptor-related protein 1 (LRP1) is known as an endocytic and signal transmission receptor. We formerly reported the gene expression and the localization of LRP1 in cartilage tissue and chondrocytes, but its roles in the differentiation of chondrocytes remained to be investigated. Here, in order to address this issue, we employed RNAi strategy to knockdown lrp1 in chondrocytic cells and obtained findings indicating a critical role therein. As a result of lrp1 knockdown, aggrecan and col2a1 mRNA levels were decreased. However, that of col10a1 or mmp13 mRNA was rather increased. Under this condition, we performed a promoter assay for Axin2, which is known to be induced by activation of the WNT/beta-catenin (betacat) signaling pathway. Thereby, we found that Axin2 promoter activity was enhanced in the lrp1 knockdown cells. Furthermore, when the WNT/beta-catenin pathway was activated in chondrocytic cells by WNT3a or SB216763, which inhibits the phosphorylation of GSK3beta, the mRNA levels of aggrecan and col2a1 were decreased, whereas that of mmp13 was increased. Additionally, the level of phosphorylated protein kinase C (PKC) zeta was also decreased in the lrp1 knockdown cells. When the phosphorylation of PKCzeta was selectively inhibited, aggrecan and col2a1 mRNA levels decreased, whereas the mmp13 mRNA level increased. These data demonstrate that LRP1 exerts remarkable effects to retain the mature phenotype of chondrocytes as a critical mediator of cell signaling. Our findings also indicate that the onset of hypertrophy during endochondral ossification appears to be particularly dependent on the WNT and PKC signaling initiated by LRP1.

Distribution, gene expression, and functional role of EphA4 during ossification.

EphA4 receptor tyrosine kinase has been shown to be critically involved in neural tissue development. Here, we found EphA4 was also distributed among hypertrophic chondrocytes and osteoblasts in the growth plate of developing mouse long bones. In vitro evaluation revealed that ephA4 expression was elevated upon hypertrophic differentiation of chondrocytes and that markedly stronger expression was observed in osteoblastic SaOS-2 than chondrocytic HCS-2/8 cells. Of note, RNAi-mediated silencing of ephA4 in SaOS-2 cells resulted in the repression of osteocalcin gene expression and alkaline phosphatase activity. Interestingly, confocal laser-scanning microscopic analysis revealed the presence of EphA4 molecules in the nucleus as well as on the surface of SaOS-2 cells. These findings are the first indication of a critical role of EphA4 in ossification, especially at the final stage in which osteoblasts and hypertrophic chondrocytes play major roles.