Osteoblasts are derived originally from pluripotent mesenchymal stem cells on migration into the bone matrix. To elucidate the contribution of classical cadherins in this differentiation pathway, we developed a new protocol for their analysis and studied their specific expressions in various cell lines of the mesenchymal lineage, including osteoblasts. N-cadherin was expressed constitutively in all cell lines examined except an osteocyte-like cell line whereas cadherin-11 was expressed selectively in preosteoblast and preadipocyte cell lines. P-cadherin also was expressed in primary cultures of calvarial cells and mature osteoblasts at a relatively low level compared with N-cadherin and cadherin-11. M-cadherin was expressed only in a premyoblast cell line. We observed the transition of cadherin expression from M-cadherin to cadherin-11 in the premyoblast cell line when osteogenic differentiation was induced by treatment with bone morphogenetic protein 2 (BMP-2), while the expression of N-cadherin remained unchanged. In contrast, when a preadipocyte cell line, which shows a similar pattern of cadherin expression to osteoblasts, was induced to undergo adipogenic differentiation, the expression of N-cadherin and cadherin-11 was decreased. These observations characterize the cadherin expression profile of mesenchymal lineage cells, especially osteoblasts, which regularly express cadherin-11. Cadherin-11 may affect cell sorting, alignment, and separation through differentiation.
Osteoblasts are highly specialized cells that produce several extracellular matrix proteins and contribute to the calcification of tissue. Osteoblasts are thought to originate from mesenchymal stem cells with high potential to differentiate into various cells including stroma, adipocyte, myoblast, chondroblast, fibroblast, and osteoblast.(1,2) During the process of differentiation, immature mesenchymal cells migrate and become incorporated into several tissues including bone, cartilage, muscle, lung, and brain, where they terminally differentiate.(3–6) Similarly, some bone marrow cells become osteoblasts in contact with the bone matrix, subsequently enter bone tissues, and are converted to osteocytes.(7) During this process, osteoblastic cells must be sorted from other mesenchymal cells, migrate, and align with neighboring osteoblasts. Various cell adhesion molecules including cadherins are likely to influence this process.(8)
Classical cadherins are calcium-dependent homophilic adhesion receptors with five extracellular domains, termed EC-1-EC-5, one transmembrane region, and a cytoplasmic domain.(9) The EC-1 domain directly binds to the same cadherin in opposing cells, whereas the other EC domains form an elongated structure in the presence of calcium ion.(10,11) The cytoplasmic domain is composed of a juxtamembrane region, and a β-catenin binding site located at around the 40th amino acid from the COOH-terminus is the most conserved segment among cadherins. In the mouse, nine different classical cadherins have been cloned to date. They have been divided into three groups, types I, II, and III, according to sequence homologies and adhesive stability.(12–20) E-, N-, P-, and R-cadherins belong to the type I family, while K-cadherin, cadherin-8, and cadherin-11 belong to the type II family. VE- and M-cadherins have been proposed to be the type III.(21) The strong homophilic adhesion activity between cadherins contributes not only to homotypic cell integrity but also to the sorting and migration of individual cells. In fact, the expression level and profile of cadherins affect the ability to recognize neighboring cells and migrate.(22–25)
In osteoblasts, the expression of N-cadherin, cadherin-11, R-cadherin, and an isoform of K-cadherin has been reported, and their roles in the maturation of osteoblasts and the activation of osteoclasts were shown.(26,27) However, the expression patterns of these classical cadherins in the differentiation toward osteoblasts from progenitor cells and the transdifferentiation from the other mesenchymal lineages to osteoblasts have not yet been identified. Therefore, this study focuses on investigating the expression of cadherins in the differentiation of osteoblasts from cells of immature mesenchymal lineages and may elucidate the key factor in mesenchymal cell differentiation.
We show that each mesenchymal lineage cell line has a characteristic cadherin expression profile and that cadherin-11 is expressed constitutionally in the osteoblast lineage while N-cadherin is expressed widely in mesenchymal lineage cells. The results point to a mechanism whereby individual mesenchymal cells sort through differentiation and cadherin transition.
MATERIALS AND METHODS
Bone morphogenetic protein 2
Bioactive recombinant human bone morphogenetic protein 2 (BMP-2) was kindly provided by Hoechst Marion Roussel Japan Limited (Kawagoe, Japan).
A monoclonal rat-anti-mouse N-cadherin antibody (MNCD-1) was kindly provided by Dr. M. Takeichi (Kyoto University, Kyoto, Japan). A monoclonal mouse-anti-mouse/human cadherin-11 was generated. In brief, the fusion protein of maltose binding protein and the EC-1 domain of mouse cadherin-11, in which the amino acid sequence is the same as that of humans, was injected into cadherin-11 null mutant mice(28) to induce recognition of self antigen. A monoclonal antibody from a resultant hybridoma clone, which reacted to mouse cadherin-11 expressed on transfected L cells, was selected by flow cytometry and then termed OLA-9. OLA-9, in which subclass was determined as immunoglobulin G2bκ (IgG2bκ) with a mouse antibody typing kit (Southern Biotechnology, Birmingham, AL, USA), can be used for flow cytometry, Western blot, immunoprecipitation, and immunohistochemical analyses. A monoclonal anti-β-catenin antibody was purchased from Transduction Laboratories (Lexington, KY, USA). A polyclonal anti-P, -E-, -N-cadherin and a monoclonal anti-mouse P-cadherin (MPCD-1) were purchased from Takara Shuzo (Shiga, Japan).
DNA sequence of mouse classical cadherins
DNA sequences of E-, N-, P-, R-, VE-, K-, and M-cadherin as well as cadherin-8 and -11 are listed in Table 1 with accession numbers.
Table Table 1.. PCR Primers of Mouse Classical Cadherins
Primary culture of mouse calvarial cells
Calvariae isolated from 3 neonatal ddY mice were first incubated for 5 minutes in collagenase solution phosphate-buffered saline (PBS; containing 0.1% collagenase and 0.2% dispase) at 37°C with agitation. They were digested repeatedly four times in a fresh solution under the same condition. Calvarial cells in the suspension were collected by centrifugation and then were cultured overnight in α-modified essential medium (α-MEM) containing 10% fetal calf serum (FCS) in 100-mm dishes. Cells were washed with PBS to eliminate nonadherent cells. Confluent cells were used for preparation of total RNA or cell lysate.
Source of cell lines
All cell lines were maintained in the following medium containing antibiotics (100 U/ml of penicillin-G and 100 μg/ml of streptomycin) at 37°C in a humidified atmosphere of 5% CO2 in air: α-MEM containing 10% FCS for ST2 (mouse bone marrow derived-stromal cell line)(29) and MC3T3-E1 (mouse calvaria-derived osteoblastic cell line),(30) DMEM containing 10% FCS for C3H10T1/2 (mouse embryonal cell line, CCL 226; American Type Culture Collection, Rockville, MD, USA), NIH3T3 (mouse fibroblastic cell line CRL1658) and 3T3-L1 (mouse embryo-derived preadipose-like cell line CL 173), DMEM containing 15% FCS for C2C12 (mouse myoblast cell line CRL 1772), 1:1 mixture of DMEM and F12 containing 5% FCS, 10 μg/ml human transferrin and 3 × 10−8 M sodium selenite for ATDC5 (mouse teratocarcinoma-derived chondrogenic cell line),(31) and α-MEM containing 2.5% FCS and 2.5% calf serum for MLO-Y4 (mouse osteocyte-like cell line),(32) which was kindly provided by Dr. L.F. Bonewald (Texas University, San Antonio, TX, USA).
Treatment of C2C12 with BMP-2
BMP-2 treatment was performed basically as described.(33) In brief, 6 × 105 C2C12 cells were seeded on 60-mm dishes and cultured overnight in DMEM containing 15% FCS. The medium was replaced with DMEM containing 5% FCS with or without 1 μg/ml of BMP-2 for osteoblast or myoblast differentiation, respectively, and the culture continued for various periods.
Differentiation of 3T3-L1
3T3-L1 cells (6 × 105) were seeded on 60-mm dishes and cultured overnight in DMEM containing 10% FCS. For differentiation, cells were treated with DMEM containing 5% FCS, 100 ng/ml of insulin, 0.5 mM of 3-isobuthyl-1-methyl-xanthine (IBMX), and 1 μM of dexamethasone (Dex) for various periods.
Total RNA was isolated from cultured cells by the guanidine isothiocyanate extraction method. For preparation of RNA from a mouse embryo, an E17 embryo frozen in liquid N2 was pulverized, and total RNA was extracted by the guanidine isothiocyanate extraction method.
Reverse-transcription polymerase chain reaction Southern blot analysis
First-strand complementary DNAs (cDNAs) were prepared from 1 μg of total RNA using Super Script II RNaseH reverse transcriptase (Life Technologies, Gaithersburg, MD, USA). The cDNA was amplified with the following primers: for actin, 5′-catcgtgggccgctctaggcacca-3′and 5′-cggttggccttagggttcaggggg-3′; for osteocalcin, 5′-tctgacaaagccttcatgtcc-3′and 5′-aaatagtgataccgtagatgcg-3′; and for alkaline phosphatase (ALP), 5′-ccctgaaactccaaaagctc-3′and 5′-tctggtggcatctcgttatc-3′. The cDNA for each cadherin was amplified using primers as indicated in Table 1. Polymerase chain reaction (PCR) for cadherin fragments was carried out for 22 cycles of denaturation at 95°C for 1 minute, annealing at 60°C for 1 minute, and extension at 72°C for 1 minute in a tube or separate tubes, and PCR for other genes was performed separately in the same condition. The resultant PCR products were separated on a 1.2% agarose gel and transferred onto a nylon membrane. For Southern blotting, a digoxigenin (DIG) Luminescent Detection Kit for Nucleic Acids (Boehringer Mannheim, Mannheim, Germany) and DIG-labeled probe generated using PCR DIG PROBE Synthesis kit (Boehringer Mannheim) were applied to detect desired PCR products by probing each cloned PCR fragment, which was already confirmed as an intact gene fragment by DNA sequencing.
Northern blot analysis
Total RNA (10 μg) was run on 1.2% agarose-formaldehyde gels and transferred onto Hybond-N+ membranes (Amersham Pharmacia, Buckinghamshire, UK). The membranes were prehybridized for 1 h at 42°C in DIG Easy Hyb (Boehringer Mannheim) and hybridized with [3P]-labeled probes overnight at 42°C. The labeled probes were prepared using the DNA random priming kit (Takara Shuzo) and [α-32P]deoxycytosine triphosphate (dCTP; Amersham Pharmacia). The membranes were washed twice at room temperature in 2× SSC/0.1% sodium dodecyl sulfate (SDS) and twice at 55°C in 0.2× SSC/0.1% SDS. Signals were detected by autoradiography at −80°C or analyzed using the combination of an Imaging Plate and BAS2000 (Fuji, Tokyo, Japan). DNA fragments for probes of β-actin, osteocalcin, ALP, M-cadherin, and N-cadherin were prepared from PCR products using primers as indicated in the previous section and in Table 1. For preparation of myogenin and aP2 probes, PCRs by the primers 5′-acactgagggagaagcgcag-3′,5′-ctgggtgttagccttatgtg-3′ and 5′-acaaggaaagtggcaggcat-3′,5′-accaccagcttgtcaccatc-3′, respectively, were performed, and then cDNA products were sequenced. All PCR products were inserted into the pPICT-2 vector. A full-length mouse cadherin-11 cDNA was used as a probe.(19)
Sequence analysis of PCR products
The DNA fragments cloned into the XcmI site of the pPICT-2 vector were sequenced using the ThermoSequenase kit (Amersham Pharmacia) and a DNA sequencer LIC-4200L(S)-1 (Li-Cor, Lincoln, NE, USA).
Western blot analysis
Cells were washed twice with cold PBS and scraped off the dish. Collected cells were lysed in 100 μl of PBS and an equal volume of lysis buffer (2% NP-40, 20 mM Tris-Cl, pH 8.0, 150 mM NaCl, 5 mM MgCl2, 5 mM EDTA, 2 mM NaN3, 1 mM phenylmethylsulfonyl fluoride [PMSF], and 2 μg/ml leupeptin). Samples were stored on ice for 30 minutes, and centrifuged at 15,000 rpm for 15 minutes. The supernatant was transferred to a new tube and the amount of protein was determined by the Bradford method using bovine serum albumin (BSA) as a standard. An equal volume of 2× sample buffer was added and the mixture was boiled for 5 minutes. The samples were separated on 7.5% polyacrylamide SDS gels and transferred to nitrocellulose filters. After incubation with the respective antibodies, signals were detected using the enhanced chemiluminescence (ECL) system (Amersham Pharmacia).
Establishment of a reverse-transcription-PCR system for the detection of classical cadherins
To date, nine classical cadherins, E-, N-, P-, R-, VE-, K-, M-cadherin, and cadherin-8 and -11 have been cloned and sequenced. From the published sequences, eight kinds of primers were designed and synthesized to examine the expression of classical cadherins in mesenchymal cells (Table 1). Figure 1 shows the cDNA fragments amplified from total RNA of a mouse whole embryo. The bands were amplified as a single band of desired size and subsequently cloned and sequenced. The sequence of each band was identical to the targeted cadherin (data not shown). Thus, this reverse-transcription (RT)-PCR system can detect all published classical cadherins on the same gel.
Expression of classical cadherins in mesenchymal cell lines
We investigated the expression of classical cadherins in several mesenchymal cell lines. These mouse cell lines are C3H10T1/2 (embryonal cell line), ST2 (stromal cell line), ATDC5 (chondrogenic cell line), NIH3T3 (fibroblastic cell line), C2C12 (myoblast cell line), 3T3-L1 (preadipose-like cell line), and MC3T3-E1 (osteoblastic cell line). Figure 2A shows the results of RT-PCR Southern blot analyses, using cDNAs from a mouse whole embryo E17, as positive controls in each panel. N-cadherin was present in all cell lines whereas E-cadherin was absent. Cadherin-11 was detected in C3H10T1/2, ST2, 3T3-L1, and MC3T3-E1 but not in ATDC5, NIH3T3, and C2C12. M-cadherin was detected only in C2C12. The expression level of P-, R-, VE-, K-cadherin, and cadherin-8 may be too low to obtain signals at this exposure time. To detect the expression of cadherin-11 protein, we first established a monoclonal antibody against mouse cadherin-11 as described in the Materials and Methods section and confirmed that mouse cadherin-11 was detected at 120 kDa (data not shown). This antibody was applied to detect cadherin-11 in mesenchymal cell lines, and positive signals of cadherin-11 were obtained from C3H10T1/2, ST2, 3T3-L1, and MC3T3-E1 (Fig. 3A, top). On the other hand, the expression of N-cadherin was found in all cell lines as a 140-kDa band (Fig. 3A, middle). Expression of β-catenin also was examined because the expression level of β-catenin corresponds to that of all cadherins. Results showed that there was good correlation in the expression level between β-catenin and classical cadherins (Fig. 3A, bottom). All results from Western blot analyses were consistent with the data obtained by RT-PCR Southern blot analyses shown in Fig. 1.
Cadherin expression in osteoblasts
The expression of cadherins was examined in osteoblasts by RT-PCR Southern and Western blot analyses. As shown in results from PT-PCR Southern blot analysis (Fig. 2), primary calvarial osteoblast (COB) cells expressed N-, P-, VE-, K-cadherin, and cadherin-8 and -11, whereas MC3T3-E1 expressed only N-cadherin and cadherin-11. Thus, to examine the cadherin expression during the osteoblast differentiation, we used long-term cultured MC3T3-E1 cells and the osteocyte-like cell line MLO-Y4. As reported previously,(30) MC3T3-E1 cells displayed mature osteoblast characteristics along with up-regulation of ALP and osteocalcin after the long-term culture. In the long-term culture with repeated passages, MC3T3-E1 cells uniformly expressed N-cadherin and cadherin-11 (Fig. 2B). Interestingly, P-cadherin expression was observed from day 8 and was slightly up-regulated in the longer culture (Fig. 2B). However, no expression of classical cadherins was observed in MLO-Y4 (Fig. 2C). The low expression of β-catenin at the protein level in MLO-Y4 also indicated the absence of classical cadherins (Fig. 3B). To examine the expression level of cadherin-11 messenger RNA (mRNA) quantitatively at the stage during the osteoblast differentiation of MC3T3-E1 cells in the long-term culture from 3 to 53 days, the Northern blot analysis was performed (Fig. 2D). The cadherin-11 expression was slightly up-regulated at 5–11days, and then more increased at the later stage of differentiation (40 days and 46 days). The expression pattern in the Northern blot analysis showed a good correlation to that from RT-PCR Southern blot analysis, suggesting that the result from RT-PCR Southern blot analysis was semiquantitative.
Because the expression of the P-cadherin in osteoblasts has not yet been reported, we examined P-cadherin expression at the protein level. A monoclonal antibody against P-cadherin was used to detect P-cadherin (Fig. 3C, top). P-cadherin was not detected in the 5-day and 11-day culture of MC3T3-E1, although distinct and faint signals were observed in the placenta and in COBs, respectively. Polyclonal antibodies against P-, E-, and N-cadherin, which do not cross-react with cadherin-11 (data not shown), detected the 120-kDa band in COB and MC3T3-E1 at 11 days culture but the signals were quite weak compared with the 140-kDa band assigned to N-cadherin (Fig. 3C, bottom). Because E-cadherin is not expressed in osteoblasts, the signal detected as a 120-kDa band would be P-cadherin. These results show that P-cadherin is expressed in osteoblasts at the very low level compared with N-cadherin.
Changes in cadherin expression during osteoblast differentiation
It is well known that the implantation of BMP-2 into muscle tissues induces osteogenesis in vivo,(34) and myogenic C2C12 cells are converted to the osteoblastic lineage in vitro by BMP-2 treatment.(33) We assessed whether the switch of cadherin expression occurs during the transdifferentiation of C2C12 in Northern blot analysis (Fig. 4A). As shown in Fig. 4B, the cadherin-11 expression increased 2.5- and 4.5-fold after treatment with 1 μg/ml of BMP-2 for 3 days and 6 days, respectively. In further treatment or nontreatment of BMP-2, the cadherin-11 expression at 9 days was the same as that at 6 days in either case (data not shown). Concomitantly, the M-cadherin expression was significantly reduced at the RNA level. Furthermore, the expression of N-cadherin was reduced 0.8- and 0.5-fold at 3 days and 6 days, respectively (Figs. 4A and 4B). We subsequently reconfirmed these results at the protein level. Western blotting shows that N-cadherin was detected from all the samples (Fig. 4C, top) and that the strong cadherin-11 signal was obtained from the BMP-2-treated cells (Fig. 4C, middle). Although N-cadherin mRNA was decreased in BMP-2 treatment for 6 days, N-cadherin protein expression appeared somewhat increased, which may be explained by formation of the stabilized cell structure constructed by both N-cadherin and cadherin-11. Constitutive expression of β-catenin (Fig. 4C, bottom) indicates the switch of cadherin expression from M-cadherin to cadherin-11. Although the cadherin-11 RNA expression increased slightly in controls after 3-day and 6-day culture, this was not significant because the cadherin-11 expression at the protein level was not observed at 6 days (Fig. 4C).
Classical cadherins are not expressed in mature adipocytes
The expression profiles revealed that cadherins in preosteoblasts were almost the same as those in preadipocytes, reflecting a common lineage.(35,36) Therefore, we examined whether cadherin expression changes during adipogenic differentiation. 3T3-L1 was used as a model cell line to study adipogenesis because it easily differentiates into mature adipocytes on treatment with Dex, IBMX, and insulin.(37) As indicated by Northern blot analysis, the expression level of N-cadherin RNA was reduced to 70% compared with that of control cells during the adipogenic differentiation. Cadherin-11 RNA was reduced to 30% and 5% of control levels at 3 days and 6 days, respectively (Figs. 5A and 5B). At the protein level, both N-cadherin and cadherin-11 were reduced more markedly than at the RNA level, indicating that the reduced expression of cadherins is caused not only by transcriptional regulation but also by posttranscriptional modification (Fig. 5C). The morphological change of 3T3-L1 cells after 6 days of treatment was shown in Fig. 5D, indicating round cells of adipocytes (Fig. 5D, panel c). In this process, classical cadherins may not be replaced, because β-catenin also is reduced at the protein level as the cells lose apparent adhesiveness during the adipogenic differentiation.
Generally, it is believed that mesenchymal cells including osteoblasts are differentiated from common progenitor cells, and the demonstration of bone marrow cells as the source of various mesenchymal tissues has brought the hypothesis that bone marrow contains mesenchymal stem cells.(3,5,38) However, little is known about the nature of mesenchymal stem cells and the mechanism of their recruitment through differentiation toward several cell types including osteoblasts. Therefore, for a more complete understanding of osteoblastogenesis, the mechanism of their differentiation from mesenchymal stem cells is of great importance. In this study we focused on the expression of classical cadherins in various mesenchymal cell lines, because the expression level and profile of classical cadherins is essential for cell scattering,(39) migration, and sorting. Therefore, it is reasonable to link the cadherin activity with mesenchymal cell migration and terminal differentiation.
Although several classical cadherins have been cloned and their expression patterns were reported, there is no single method to study the expression profiles of these molecules at once. This prompted us to establish a detection system using RT-PCR Southern blot analysis. We prepared seven cell lines: ATDC5, NIH3T3, C2C12, 3T3-L1, and MC3T3-E1, as representative precursor cells of chondroblasts, fibroblasts, myoblasts, adipocytes, and osteoblasts, respectively, and ST2 as stromal cells and C3H10T1/2 as mesenchymal progenitors. Our cadherin detection system found that all cell lines expressed N-cadherin but not E-cadherin and the specific expression of M-cadherin in C2C12 myoblasts. These observations corresponded to three previous findings. First, the expression of N-cadherin was widely distributed in mesenchymal tissues.(40) Second, E-cadherin suppressed the expression of T-brachyury and thus inhibited mesenchymal differentiation of embryonic stem cells.(41) Third, M-cadherin is expressed in satellite cells and is involved in myogenesis including myotube formation.(42–45) However, one interesting and new feature of classical cadherins was the selective expression of cadherin-11 in the precursor cell line of adipocyte or osteoblast lineage, the stromal cell line, and the mesenchymal progenitor cell line. All these cell lines are closely related because the mesenchymal progenitor cell line C3H10T1/2 has the potential to differentiate into several mesenchymal lineages including osteoblasts.(46–48) Stromal cells also have the potential to show osteoblast phenotypes.(49) The recent cloning of cells that either undergo adipogenesis or osteogenesis indicates a reciprocal relationship between the osteoblast and adipocyte lineages.(50) Although preadipocytes and preosteoblasts have similar profiles of cadherin expression as described previously, the reduced expression of both N-cadherin and cadherin-11 during the adipogenic differentiation of 3T3-L1 indicates that the cadherins of mature adipocytes differ from those of osteoblasts. As a consequence, each mesenchymal lineage has a characteristic profile with various expression levels. This was illustrated by the transition of cadherin expression from M-cadherin to cadherin-11 under the transdifferentiation from the myogenic to the osteogenic lineage. Importantly, expression of N-cadherin was relatively stable during transdifferentiation. Prior studies have shown that cadherin-11 is also a mesenchymal-specific cadherin; however, it should be noted that the expression pattern of cadherin-11 in tissues and embryos is not identical to that of N-cadherin.(51) For example, the sclerotome only continues to express cadherin-11 because of somatogenesis until chondrogenesis begins.(51,52) These data suggested that cadherin-11 is expressed consistently throughout osteoblast differentiation whereas N-cadherin is expressed more universally throughout mesenchymal lineages.
There is a possibility that the expression patterns of cadherins we observed are not characteristic of each cell line. To confirm the entire results, cadherin expression in a primary culture of calvarial cells was investigated. We found more variable cadherin expression profiles than those observed in MC3T3-E1. The most likely explanation for this is that the young calvaria contain other contaminating tissues. For example, the bands of cadherin-8 and VE-cadherin could be explained by contamination of brain and vascular endothelial cells.(16,18,53) Alternatively, the primary culture may contain osteoblasts at various developmental stages with different profiles of cadherin expression. To test the later possibility, cadherin expression in undifferentiated or differentiated MC3T3-E1 cells and the osteocyte-like cell line MLO-Y4 was examined; up-regulation of P-cadherin was observed in differentiated MC3T3-E1 cells. Although this is the first report of P-cadherin expression in osteoblasts, the low expression level suggests that the contribution of P-cadherin to cell adhesion is minor compared with that of other cadherins. The fact that there is no skeletal abnormality in P-cadherin null mutant mice supports our speculation about little function of P-cadherin in osteogenesis.(54) In addition, the intact K-cadherin appears to be expressed at a very low level in primary osteoblasts because the band intensity was comparable with that of P-cadherin in RT-PCR Southern blot analysis, and no detectable band was observed in MC3T3-E1 samples. An isoform of K-cadherin was reported to be expressed in osteoblasts, which acted not only in homophilic cell adhesion but also in osteoclast maturation.(26) However, the involvement of the intact K-cadherin in osteogenesis is still open to question. Recently, it was reported that osteoblasts expressed N-cadherin, cadherin-11, and R-cadherin,(27) but in our study, R-cadherin was not observed in differentiated osteoblasts or even in primary cultures of calvarial cells. Considering that R-cadherin was down-regulated on treatment with BMP-2, this cadherin may work at an earlier stage of osteoblast development or its expression may be specific for osteoblasts in the long bone. Consequently, the main homophilic adhesion receptors in osteoblasts may be N-cadherin and cadherin-11. The MC3T3-E1 cell line already is committed to the osteoblastic lineage; therefore, both cadherin-11 and N-cadherin are expressed dominantly, and no switch differentiation of MC3T3-E1 cells to the other lineages has been observed. In concert with these classical cadherin activities, the preosteoblasts align on the bone matrix and then further differentiate into osteocytes. In this study, the osteocyte-like cell line MLO-Y4 did not express classical cadherins at all. In addition, we found that classical cadherins were not expressed in mature human osteocytes. However, both cadherin-11 and N-cadherin were expressed in human osteoblasts.(55) It is likely that the formation of adherence junction is reduced markedly in this process, which would be an important mechanism for osteocytes to enter the osteoid. These results are consistent with data obtained from immunostaining in mice: the strong expression was observed at the outer periosteal layer of calvaria in the sagittal suture and in the periosteal and endosteal surfaces of the femoral diaphysis, whereas the fainter expression of cadherin-11 was observed at the primary spongiosum in femoral trabecular bone, and, moreover, cadherin-11 was not clearly detectable in osteocytes.(57) Although the role of type I cadherins, including N-cadherin, in osteogenesis has been studied by HAV peptide perturbation,(27,58) the contribution of cadherin-11 in osteogenesis has yet to be elucidated because HAV is replaced by QAV in cadherin-11. Thus, we further test the effect of QAV in cell-to-cell adhesion. To show the cadherin function in the osteoblast differentiation from results obtained in this work, comparing the expression of cadherins in various unrelated cells of different origin, is limited. One of the best approaches for investigating the involvement of cadherin-11 in osteogenesis is to establish mutant null mice. Our unpublished results show a reduction in bone density in cadherin-11 null mutant mice, suggesting that cadherin-11 plays an important role in osteogenesis.(57) Thus, we may expect that cadherins directly induce differentiation and are able to determine the cell lineage of mesenchymal cells, which will be shown in further experiments.
The authors acknowledge Dr. M. Takeichi and Dr. O. Chisaka for providing monoclonal N-cadherin antibody and cadherin-11 null mutant mice, and Dr. L.F. Bonewald for providing MLO-Y4. We thank Dr. Steve Byers for critical review of this article. This work was supported by grants from the Japan Ministry of Education, Science, Sports and Culture and Mitsubishi Foundation.