Regulation of β-Catenin Signaling and Maintenance of Chondrocyte Differentiation by Ubiquitin-independent Proteasomal Degradation of α-Catenin*

Accumulation of β-catenin and subsequent stimulation of β-catenin-T cell-factor (Tcf)/lymphoid-enhancerfactor (Lef) transcriptional activity causes dedifferentiation of articular chondrocytes, which is characterized by decreased type II collagen expression and initiation of type I collagen expression. This study examined the mechanisms of α-catenin degradation, the role of α-catenin in β-catenin signaling, and the physiological significance of α-catenin regulation of β-catenin signaling in articular chondrocytes. We found that both α- and β-catenin accumulated during dedifferentiation of chondrocytes by escaping from proteasomal degradation. β-Catenin degradation was ubiquitination-dependent, whereas α-catenin was proteasomally degraded in a ubiquitination-independent fashion. The accumulated α- and β-catenin existed as complexes in the cytosol and nucleus. The complex formation between α- and β-catenin blocked proteasomal degradation of α-catenin and also inhibited β-catenin-Tcf/Lef transcriptional activity and the suppression of type II collagen expression associated with ectopic expression of β-catenin, the inhibition of proteasome, or Wnt signaling. Collectively, our results indicate that ubiquitin-independent degradation of α-catenin regulates β-catenin signaling and maintenance of the differentiated phenotype of articular chondrocytes.

␤-Catenin interacts with cadherin to participate in cell-cell adhesion and regulates gene expression by acting as a transcriptional co-activator (1,2). In the absence of extracellular stimuli, cytosolic ␤-catenin is phosphorylated by glycogen synthase kinase-3␤, leading to its ubiquitination and subsequent degradation by the 26 S proteasome. However, extracellular stimuli, such as Wnt signaling, lead to the inhibition of glycogen synthase kinase-3␤, escape of ␤-catenin from ubiquitin-dependent proteolytic degradation, and subsequent cytosolic accumulation of ␤-catenin (3). The accumulated ␤-catenin translocates into the nucleus in association with members of the T cell-factor (Tcf) 1 /lymphoid-enhancer-factor (Lef) family of transcription factors, leading to stimulation or suppression of target gene transcription (2). As a transcriptional co-activator, ␤-catenin is involved in the regulation of several biological functions. Our group previously has shown that ␤-catenin regulates maintenance of differentiated phenotypes (4,5) and expression of cyclooxygenase-2 (6) in chondrocytes by acting as a transcriptional co-activator. We have also shown that Wnt-7a causes dedifferentiation of chondrocytes characterized by suppression of type II collagen expression and the onset of type I collagen expression. Accumulation and stimulation of ␤-catenin transcriptional activity by Wnt-7a signaling is sufficient to cause chondrocyte dedifferentiation (4,5).
␤-Catenin signaling can be regulated by a variety of proteins. Cadherins regulate the transcriptional activity of ␤-catenin in a cell adhesion-independent manner (7,8) via sequestration of ␤-catenin in the cytoplasm (9), whereas ␥-catenin increases the turnover of ␤-catenin protein (10). The NEMO-like kinase does not regulate ␤-catenin stability but rather inhibits the interaction between DNA and the ␤-catenin-Tcf/Lef complex by phosphorylating Tcf/Lef proteins (11). During cell-cell adhesion, ␣-catenin associates directly with ␤-catenin, acting to link cadherins to the actin cytoskeleton. However, ␣-catenin is localized not only to areas of cell-cell contact; it is also found in other parts of the cell, such as the cytoplasm and nucleus, suggesting that it participates in the regulation of other cellular functions. Indeed, recent reports indicated that ␣-catenin inhibits ␤-catenin signaling in various cell types. For instance, when ␣-catenin is overexpressed in Xenopus embryo, Wnt signaling is inhibited through direct binding of ␣and ␤-catenin (9,12), perhaps due to the sequestration of ␤-catenin in the cytoplasm (9,13) or inhibition of the interaction between ␤-catenin-Tcf/ Lef complex and target DNA (14). In addition to regulating ␤-catenin signaling, the ␣-catenin protein is actively degraded in cadherin-deficient L cells in a proteasome-independent manner, although ␣-catenin contains the PEST sequence, which is a common target for rapid degradation by calpain-or proteasome-mediated proteolysis (13).
It appears likely that ␣-catenin regulates ␤-catenin signaling during various cellular processes. However, the regulatory mechanisms underlying the modulation of ␣-catenin protein stability and the physiological function of ␣-catenin regulation of ␤-catenin signaling remain to be elucidated. We, therefore, investigated the mechanisms of ␣-catenin degradation, the role of ␣-catenin in ␤-catenin signaling, and the physiological significance of ␣-catenin regulation of ␤-catenin signaling in articular chondrocytes.
Cell Culture-Individual articular chondrocytes were isolated from joint cartilage slices from 2-week-old New Zealand White rabbits as described previously (15,16). The cells were plated on culture dishes at a density of 5 ϫ 10 4 cells/cm 2 and cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 50 g/ml streptomycin, and 50 units/ml penicillin. The medium was replaced every 2 days until the cultures were tested with the indicated pharmacological reagents in each experiment. A ␤-catenin-negative cell line, human mesothelioma cancer cell (H28), was obtained from American Type Culture Collection (Manassas, VA) and maintained in RPMI 1640 supplemented with 10% fetal calf serum and antibiotics.
Construction and Transfection of Expression Vector-A cDNA for human ␣-catenin was prepared by reverse transcription-polymerase chain reaction (RT-PCR) from HeLa cells with specific primers (sense,  5Ј-GGA AAT CCA TGA CTG CTG TCC ATG CAG GCA-3Ј, and antisense, 5Ј-GGG GTA CCC CTT AGA TGC TGT CCA TAG CTT T-3Ј) designed to introduce EcoRI and KpnI restriction sequences at the 5Ј and 3Ј ends, respectively. The resulting cDNA was cloned into pGFP-C1 vector (Clontech). The non-ubiquitinatable S37A ␤-catenin was constructed as described previously (5). Articular chondrocytes were transfected with expression vectors using the Lipofectamine reagent (Invitrogen) as described previously (17,18).
Knockdown of ␣-Catenin by siRNA-We examined three siRNAs to silence ␣-catenin in chondrocytes and found that one of them caused effective knockdown of ␣-catenin. Briefly, a 21-nucleotide sequence (5Ј-GGA CCT GCT TTC GGA GTA CAT-3Ј) within the coding region of rabbit ␣-catenin-(1038 -1058) (accession number: AB193105) was selected according to the manufacturer's notes (InvivoGen, San Diego, CA). Two complementary oligonucleotides (sense, 5Ј-GTA CCT CGG ACC TGC TTT CGG AGT ACA TTC AAG AGA TGT ACT CCG AAA GCA GGT CCT TTT TGG AAA-3Ј, and antisense, 5Ј-AGC TTT TCC AAA AAG GAC CTG CTT TCG GAG TAC ATC TCT TGA ATG TAC TCC GAA AGC AGG TCC GAG-3Ј; bold characters indicate ␣-catenin mRNA targeting sequences, and italic characters indicate the hairpin loops) were synthesized chemically. A scrambled ␣-catenin siRNA (5Ј-GCT CGC CTA GGT CAG GTA TAT-3Ј) was used as a negative control, which shows no significant homology to known gene sequences and did not regulate ␣-catenin expression. Annealing of sense and antisense oligonucleotides (25 M each) was performed in 0.15 M NaCl at 80°C for 2 min and then maintained until the temperature reached 35°C in a thermal unit. Annealed siRNA was ligated into Acc65I/HindIII-digested psiRNA-hH1GFP:zeo vector (InvivoGen). Transfection was carried out with Lipofectamine reagent as described above.
␤-Catenin Reporter Gene Assay-The transcriptional activity of the ␤-catenin-Tcf/Lef complex was determined using the previously described reporter gene assay (4,5). Briefly, cells were transiently trans-fected with 1 g of the Tcf/Lef reporter genes, TOPFlash (optimal Lef-binding site) or FOPFlash (mutated Lef-binding site) (Upstate Biotechnology Inc., Lake Placid, NY), and l g of pCMV-␤-galactosidase. Luciferase activity was measured and normalized for transfection efficiency using ␤-galactosidase activity.
Immunofluorescence Microscopy-Cells were fixed with 3.5% paraformaldehyde in phosphate-buffered saline for 10 min at room temperature. The cells were permeabilized and blocked with 0.1% Triton X-100 and 5% fetal calf serum in phosphate-buffered saline for 30 min. The fixed cells were washed and incubated for 1 h with antibodies against ␣-catenin (BD Transduction Laboratories) or ␤-catenin (Santa Cruz Biotechnology Inc., Santa Cruz, CA). The cells were washed, incubated with rhodamine-or fluorescein-conjugated secondary antibodies for 30 min, and observed under a fluorescence microscope. Where indicated, green florescence protein (GFP)-labeled ␣-catenin was used to determine the localization of transfected ␣-catenin.
Proteasome Activity Assay-Proteasome activity in the cell extracts was determined by fluorometric measurement of 4-methylcoumarinyl-7-amide release from a fluorogenic peptide substrate (Suc-Leu-Leu-Val-Tyr-MCA). Homogenized cell extracts were cleared by centrifugation, and the supernatants were used for determination of enzymatic activity. Proteins (30 g) were mixed with 100 l of assay mixture containing 100 M fluorogenic peptide substrate in 20 mM Tris-HCl, pH 7.8, 5 mM MgCl 2 , 10 mM KCl, and 0.5 mM dithiothreitol. The reactions were incubated for 30 min at 37°C and then quenched with 200 l of ethanol. Fluorescence was measured in a spectrofluorometer using an excitation wavelength of 380 nm and an emission wavelength of 440 nm.
Native Gel Electrophoresis and Fluorogenic Substrate Overlay-Native gel electrophoresis was performed as described by Mahaffey et al. (20). Briefly, samples were analyzed on non-denaturing gels consisting of a 4.5% resolving gel and a 2.5% stacking gel; these were cast in 90 mM Tris-HCl, pH 8.3, 1.6 mM borate, and 0.4 mM EDTA and run 4°C for 700 V-h. After electrophoresis, the proteasome bands (20 S and 26 S) were identified by overlaying the gels with 20 mM Tris-HCl, pH 7.8, 5 mM MgCl 2 , 10 mM KC1, 0.5 mM dithiothreitol, 2 mM ATP, and 100 M fluorogenic peptide and then incubating the gels at 37°C for 30 min. The 26 S proteasome bands were excised, and the proteins were separated by SDS-PAGE and analyzed by Western blotting.
Immunoprecipitation-Chondrocytes were lysed in Nonidet P-40 lysis buffer (1% Nonidet P-40, 150 mM NaCl, 50 mM Tris, pH 8.0) containing protease and phosphatase inhibitors as described above. The lysates were centrifuged at 13,000 ϫ g for 5 min at 4°C to remove cell debris, and the supernatant (500 g) was precleared with rabbit IgG and protein A-Sepharose for 1 h and then incubated with 1.5 g of antibodies against ␣-catenin or ␤-catenin. The immunocomplexes were then precipitated by incubation with protein A-Sepharose for 1 h at 4°C and analyzed by SDS-PAGE and Western blotting.
Western Blot Analysis-Chondrocytes were lysed on ice for 30 min in a buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% Nonidet P-40, and 0.1% SDS supplemented with protease and phosphatase inhibitors as described above. Proteins were fractionated by SDS-PAGE, transferred to a nitrocellulose membrane, and detected using the following antibodies: monoclonal mouse anti-type II collagen (Chemicon, Temecula, CA), mouse monoclonal anti-␤-catenin, -␣catenin, -␥-catenin, -pp120, -N-cadherin, and -ERK-1 (all from BD Transduction Laboratories) and goat polyclonal anti-actin and mouse monoclonal anti-ubiquitin (Santa Cruz Biotechnology Inc.). For the detection of ubiquitin, membranes were sandwiched between several sheets of Whatman 3MM paper and submerged in deionized water, and membrane-bound ubiquitin was heat-activated by autoclaving for 30 min.

Ubiquitin-independent Proteasomal Degradation of ␣-Catenin in Articular
Chondrocytes-Based on previous reports that ␤-catenin levels were regulated by proteasomal degradation (3), we first examined whether the proteasomal degradation pathway regulates ␣-catenin protein level in primary cultured articular chondrocytes. As expected, ␤-catenin protein levels were significantly increased by 26 S proteasome inhibitors (ALLnL and MG132) but not by caspase (Z-VAD-fmk and DEVD-fmk) or calpain (ALLM) inhibitors (Fig. 1A). Protein levels of ␣-catenin were also significantly increased by the inhibition of the 26 S proteasome (Fig. 1A), suggesting that both ␣and ␤-catenins are degraded via proteasomal pathways. Treatment with a translation inhibitor (cycloheximide) reduced levels of ␣and ␤-catenin; this reduction was blocked by the inhibition of the 26 S proteasome with MG132 ( Fig.  1B), confirming that both ␣and ␤-catenins are posttranslationally regulated via the proteasomal degradation pathway. Post-translational regulation of ␣and ␤-catenin was additionally supported by our observations that their transcript levels were not affected by proteasome inhibition (Fig. 1C).
To characterize the proteasomal degradation of ␣-catenin, we next examined its physical association with the 26 S proteasome. For this purpose, chondrocytes were treated with ALLnL or MG132 to inhibit proteasome activity ( Fig. 2A), and total cell lysates were separated by native gel electrophoresis. The 26 S proteasome band was identified by fluorogenic substrate overlay (Fig. 2B, upper panel). Proteins localized in the 26 S proteasome band were separated by SDS-PAGE. Western blotting indicated that both ␣and ␤-catenins were located in the 26 S proteasome fraction (Fig. 2B, lower panel), suggesting that ␣and ␤-catenins may associate with the 26 S proteasome. Proteasome inhibition with ALLnL and MG132 caused accumulation of ␣and ␤-catenins in a dose- (Fig. 2C) and time- (Fig. 2D) dependent manner. Accumulated ␤-catenin was detected in the gel as a ladder form consistent with ubiquitination (3). Interestingly, we did not detect any electrophoretic mobility shift of ␣-catenin following proteasome inhibition (Fig. 2, C and D). Because the electrophoretic pattern of ␣-catenin suggested non-ubiquitination, we further tested for ubiquitination of ␣and ␤-catenin by immunoprecipitation and Western blotting. As shown in Fig. 2E, Western blot analysis indicated that ␤-catenin, but not ␣-catenin, was ubiquitinated in chondrocytes, suggesting that ␣-catenin is degraded by the 26 S proteasome in a ubiquitin-independent fashion.
␤-Catenin Inhibits ␣-Catenin Degradation-We next examined whether accumulation of ␣and ␤-catenin proteins is associated with each other in chondrocytes. For this purpose, we first used immunofluorescence microscopy and Western blot analysis to determine the expression levels of ␣and ␤-catenins in chondrocytes transfected with wild-type ␣-catenin or S37A ␤-catenin (which cannot be ubiquitinated due to the loss of its phosphorylation site). Double immunostaining of chondrocytes (Fig. 3A) and Western blot analysis (Fig. 3B) revealed that cells overexpressing S37A ␤-catenin showed high levels of ␣-catenin protein, whereas overexpression of ␣-catenin did not affect ␤-catenin protein levels. Because the above results suggest that accumulation of ␣-catenin is associated with the observed increases in ␤-catenin protein levels, the association of ␣and ␤-catenin accumulation was further characterized by uncoupling the degradation of ␣-catenin from that of ␤-catenin using a ␤-catenin-negative cell line (H28). As shown in Fig.  3C, ␣-catenin was accumulated in the ␤-catenin-negative cell line by the inhibition of proteasome, indicating that ␣-catenin degradation via the proteasomal pathway is independent of the existence of ␤-catenin protein. The role of ␤-catenin accumulation in ␣-catenin degradation was further examined by ectopic expression of S37A ␤-catenin in H28 cells. Ectopically expressed S37A ␤-catenin binds to endogenous ␣-catenin and caused the increase in ␣-catenin protein level (Fig.  3D), suggesting that the existence of ␤-catenin protein and its binding to ␣-catenin inhibit proteasomal degradation of ␣-catenin. This also suggests that the observed increase in ␣-catenin protein level in ␤-catenin-overexpressing chondrocytes is due to the inhibition of ␣-catenin degradation by ␤-catenin.
␣-Catenin Inhibits ␤-Catenin Signaling and ␤-Catenin-mediated Dedifferentiation of Articular Chondrocytes-We next examined the functional significance of the proteasomal degradation of ␣-catenin in articular chondrocytes. First, we examined the intracellular localization of ␣and ␤-catenins. Immunofluorescence microscopy revealed that ␣-catenin was localized throughout cells, whereas ␤-catenin was detected mainly at areas of cell-cell contact. However, when proteasome activity was blocked by MG132 treatment, levels of ␣and ␤-catenins were significantly increased in the nucleus (Fig. 5A). Cell fractionation experiments indicated that following proteasome inhibition, accumulated ␣and ␤-catenins were localized in both the cytosolic fraction and the nuclear fraction (Fig. 5B). Co-immunoprecipitation of ␣and ␤-catenins suggested that these proteins exist as complexes in both the cytosolic and the nuclear fractions (Fig. 5C).
The functional significance of ␣-catenin degradation by the proteasomal pathway was next examined in terms of its role in Wnt-and ␤-catenin signaling. As expected, accumulation of ␤-catenin following proteasome inhibition by MG132 led to stimulation of ␤-catenin-Tcf/Lef transcriptional activity (Fig.  6A, left panel). ␤-Catenin accumulation also inhibited type II collagen expression (Fig. 6A, right panel), which is consistent with our previous observation (4,5). The significance of ␣-catenin accumulation and association with ␤-catenin was determined by overexpressing ␣-catenin followed by proteasome inhibition. As shown in Fig. 6B, ectopic expression of ␣-catenin blocked the MG132-induced increase of ␤-catenin-Tcf/Lef transcriptional activity and also significantly blocked the inhibition of type II collagen expression. The partial recovery of type II collagen expression (Fig. 6B) may be due to low ␣-catenin transfection efficiency, whereas the complete inhibition of ␤-catenin-Tcf/Lef transcriptional activity could be due to our co-transfection of the reporter gene and ␣-catenin.
Because the above results suggest that ␣-catenin may inhibit ␤-catenin signaling, the role of ␣-catenin in ␤-catenin signaling was directly examined by overexpressing ␣and/or ␤-catenin.
As shown in Fig. 7A, overexpression of ␣-catenin alone led to decreased basal levels of ␤-catenin-Tcf/Lef transcriptional activity (Fig. 7A, upper panel) and increased type II collagen expression (Fig. 7A, lower panel). Ectopic expression of S37A ␤-catenin led to increased ␤-catenin transcriptional activity and suppression of type II collagen expression (Fig. 7B). Coexpression of ␣-catenin and S37A ␤-catenin blocked the S37A ␤-catenin-induced increase in ␤-catenin-Tcf/Lef complex transcription activity and the inhibition of type II collagen expression (Fig. 7C), indicating that ␣-catenin inhibits ␤-catenin signaling and subsequent dedifferentiation of chondrocytes. The inhibitory action of ␣-catenin in ␤-catenin signaling was further confirmed by examining the role of ␣-catenin in Wnt signaling. Infection of cells with an adenovirus carrying Lef-1 (Fig. 8A) or treatment of cells with Wnt-7a-conditioned medium (Fig. 8B) led to increased ␤-catenin-Tcf/Lef transcriptional activity (left panels) and suppression of type II collagen expression (right panels). Ectopic expression of ␣-catenin blocked the increase of transcriptional activity and rescued type II collagen expression (Fig. 8), indicating that ␣-catenin inhibits Wnt signaling by disturbing the transcriptional activity of ␤-catenin.
In an attempt to elucidate the physiological significance of ␣-catenin degradation and its regulation of ␤-catenin signaling, we finally examined whether endogenous ␣-catenin can , cytosolic and nuclear fractions were prepared, and the levels of ␣-catenin and ␤-catenin were determined by Western blotting. Purity of nuclear and cytosolic fractions was determined by detecting lamin and tubulin, respectively. C, chondrocytes were treated with 10 M MG132 for 24 h. ␣-Catenin or ␤-catenin was immunoprecipitated (IP) from cytosolic or nuclear fractions, and co-precipitation of ␤-catenin or ␣-catenin, respectively, was determined by Western blotting (WB).
regulate ␤-catenin signaling and chondrocyte differentiation by using siRNA to knock down the endogenous ␣-catenin in chondrocytes. As shown in Fig. 9A, down-regulation of endogenous ␣-catenin increased the basal level of ␤-catenin-Tcf/Lef transcriptional activity (left panel) and caused reduction in type II collagen expression level (right panel). Knockdown of ␣-catenin also increased Wnt-7a-induced activation of ␤-catenin-Tcf/Lef transcriptional activity and promoted the inhibition of type II collagen expression caused by Wnt-7a treatment (Fig. 9B). Following incubation for 24 h, the transcriptional activity of ␤-catenin was determined by reporter gene assay (upper panels), and expression levels of GFP-␣catenin, ␤-catenin, and type II collagen were determined by Western blotting (lower panels). C, chondrocytes were transfected with S37A ␤-catenin with or without ␣-catenin expression vectors (2 g). Following incubation for 24 h, the transcriptional activity of ␤-catenin was determined by reporter gene assay (upper panel), and expressions of ␣-catenin, ␤-catenin, and type II collagen were determined by Western blotting (lower panel).

FIG. 8. ␣-Catenin inhibits Wnt signaling and Wnt-induced inhibition of type II collagen expression.
A, chondrocytes were transfected with empty vector (Control and Lef-1) or with a vector encoding ␣-catenin. Following incubation for 24 h, the cells were infected with empty adenovirus or with an adenovirus carrying Lef-1. The transcriptional activity of ␤-catenin was determined by reporter gene assay (left panel), and expression levels of ␣-catenin, ␤-catenin, type II collagen, and Lef-1 (determined with anti-hemagglutinin (HA) antibody) were determined by Western blotting (right panel). B, chondrocytes were transfected with empty vector or ␣-catenin expression vector. Following incubation for 24 h, the cells were treated with 200 l of Wnt-7a-conditioned medium for 24 h. The transcriptional activity of ␤-catenin was determined by reporter gene assay (left panel), and expressions of ␣-catenin (determined with anti-GFP antibody), ␤-catenin, and type II collagen were determined by Western blotting (right panel). ERK was employed as a loading control.

DISCUSSION
We previously showed that stimulation of ␤-catenin-Tcf/Lef transcriptional activity by ectopic expression of ␤-catenin or Wnt-7a treatment caused dedifferentiation of articular chondrocytes (4,5). Here, we investigated the regulatory mechanisms of ␣-catenin degradation in articular chondrocytes and examined the role of ␣-catenin in ␤-catenin signaling and its physiological significance in chondrocyte differentiation. We found that ␣-catenin was actively degraded by a ubiquitin-independent proteasomal pathway and that accumulated ␣-catenin inhibited the transcriptional activity of the ␤-catenin-Tcf/Lef complex, thereby blocking ␤-cateninor Wnt-induced dedifferentiation of primary cultured articular chondrocytes.
Ubiquitin-independent Proteasomal Degradation of ␣-Catenin-This is the first report that ␣-catenin is degraded via a proteasomal pathway, as evidenced by our observation that the inhibition of proteasome caused accumulation of ␣-catenin. This conclusion is further supported by our finding that ␣-catenin was associated with the proteasome fraction. Proteasomes play an essential role in the rapid elimination of short-lived key regulatory proteins. A major determinant for protein half-life is the presence of degradation signals, such as the ubiquitin fusion degradation signal, the PEST sequence, and the destruction box (2). It was previously shown that ␣-catenin, which contains a high quality PEST sequence (residues 633-651), was actively degraded in L cells (13). However, the previous report found that various proteasome and calpain inhibitors did not affect the level of endogenous ␣-catenin proteins in L cells (13). In contrast, under our experimental conditions, proteasome inhibition increased ␣-catenin protein levels not only in primary chondrocytes but also in HTB-94 chondrosarcoma, Balb/c, and NIH3T3 cells (data not shown), suggesting that proteasomal degradation of ␣-catenin is not restricted to primary cultured articular chondrocytes.
Another interesting finding in this study is that ␣-catenin degradation is ubiquitination-independent. For typical proteasomal protein degradation, the rate-limiting step is the recruitment of the ubiquitin-protein isopeptide ligase (E3), which conjugates a polyubiquitin tree to the substrate (21). 26 S proteasomes preferentially degrade ubiquitinated proteins (22). However, we did not detect any evidence for the ubiquitination of ␣-catenin in articular chondrocytes, which suggests that ␣-catenin is degraded via the 26 S proteasome in a ubiquitination-independent manner. This is not unheard of; several other proteins are degraded by proteasomes in a ubiquitination-independent fashion. For instance, ornithine decarboxylase is degraded by the 26 S proteasome via a ubiquitin-inde-pendent pathway (23), and this phenomenon is conserved between animals and fungi (24). Other proteins such as calmodulin, troponin C (22), and ␣-synuclein (25) are also degraded by 26 S proteasomes without ubiquitination. The p53 protein can undergo proteasomal degradation by two alternative pathways; one is ubiquitin-dependent and regulated by Mdm-2, whereas the other is ubiquitin-independent and regulated by NAD(P)H quinone oxidoreductase 1 (26). Thus, it appears that ubiquitin-independent degradation by 26 S proteasomes may be more important than has generally been assumed. However, we do not yet know precisely how this occurs in the case of ubiquitin-independent degradation of ␣-catenin in chondrocytes. One possibility is that ␣-catenin degradation is dependent to the degradation of ␤-catenin and that ubiquitinated ␤-catenin accompanies ␣-catenin into proteasomal degradation. This is based on the observations that overexpression of S37A ␤-catenin or accumulation of ␤-catenin caused accumulation of ␣-catenin, whereas overexpression of ␣-catenin did not affect ␤-catenin accumulation. However, this is unlikely to happen in chondrocytes because ␣-catenin degradation is also observed in the ␤-catenin-negative cell line. Therefore, our experiments uncoupling the degradation of ␣-catenin from that of ␤-catenin using the ␤-catenin-negative cell line clearly indicate that ␣-catenin degradation is independent to the ␤-catenin degradation. Our results additionally indicate that accumulation of ␤-catenin inhibits degradation of ␣-catenin, as evidenced by the observation that ectopically expressed S37A ␤-catenin in ␤-catenin-negative cells binds to ␣-catenin and causes the increase in ␣-catenin protein level. This also explains the reason why the accumulation of ␤-catenin always causes accumulation of ␣-catenin in chondrocytes.
␣-Catenin Regulation of ␤-Catenin Signaling and Chondrocyte Differentiation-Our current results clearly indicated that ␣-catenin inhibits nuclear signaling of ␤-catenin and thereby blocks ␤-cateninor Wnt-induced dedifferentiation of chondrocytes. The observations that 1) ␣-catenin inhibits ␤-catenin signaling and that 2) Wnt signaling increased not only ␤-catenin but also ␣-catenin levels appear to be paradoxical for the maximum effects of Wnt signaling, i.e. stimulation of ␤-catenin-Tcf/Lef transcriptional activity and resulting dedifferentiation of chondrocytes. However, our results suggested that the increase in ␣-catenin by Wnt signaling is to assure fine-tuning of the transcriptional activity of ␤-catenin-Tcf/Lef complex. This conclusion is based on the observations that 1) knockdown of endogenous ␣-catenin by siRNA increased the basal level of ␤-catenin-Tcf/Lef transcriptional activity and increased Wnt-7a-induced activation of ␤-catenin-Tcf/Lef transcriptional ac- tivity and 2) binding of ␣-catenin to ␤-catenin inhibited degradation of ␣-catenin and thereby inhibited ␤-catenin signaling. Our previous study (16) indicated that chondrogenic differentiation of mesenchymal cells accompanied decreased expression of ␣and ␤-catenin. We also reported that down-regulation of ␤-catenin is necessary for chondrogenesis and accumulation of ␤-catenin inhibited chondrogenesis by stabilizing cell-cell adhesion (5). Therefore, ␤-catenin inhibits chondrogenesis by acting as a cytoskeletal component, whereas it causes dedifferentiation of chondrocytes by acting as a transcriptional coactivator. Therefore, it is likely that down-regulation of both ␣and ␤-catenin during chondrogenesis is necessary for the disruption of cell-cell adhesion, which is necessary for chondrocyte differentiation, and the increase of ␣-catenin during dedifferentiation is to assure fine-tuning of the ␤-catenin-Tcf/Lef transcriptional activity.
Consistent with our observations, several previous studies have indicated that the inhibition of ␤-catenin signaling by ␣-catenin occurs via direct binding to ␤-catenin (9,(12)(13)(14)27). It is thought that ␣-catenin overexpression inhibits ␤-catenin-Tcf/Lef-dependent transcription by sequestering ␤-catenin in the cytoplasm (9,12,13). In contrast, endogenous or exogenously expressed ␣-catenin is found in the nuclei of various cell types such as SW480, DLD-1, and Cos cells but not in the nuclei of HCT116 cells (14,27). The reason for this differential nuclear localization of ␣-catenin is presently unknown but may suggest that ␤-catenin is not the only factor involved in nuclear localization of ␣-catenin. It was further demonstrated that ␣-catenin binding to ␤-catenin does not affect entry of ␣-catenin into the nucleus but rather inhibits the interaction of the ␤-catenin-Tcf/Lef complex with target DNA (27). In articular chondrocytes, ␣and ␤-catenin complexes are localized in both the cytosol and the nucleus, indicating that sequestration of ␤-catenin in the cytoplasm is not the mechanism by which ␤-catenin signaling is inhibited by ␣-catenin. Instead, it is more likely that association of ␣-catenin with ␤-catenin blocks their interaction with target DNA. However, this hypothesis will await further study.
In summary, we have demonstrated that both ␣and ␤-catenins are actively degraded by 26 S proteasomes in articular chondrocytes. The degradation of ␤-catenin is ubiquitination-dependent, whereas ␣-catenin is degraded independent of detectable ubiquitination. We also demonstrated that accumulated ␣-catenin associates with ␤-catenin, resulting in the inhibition of ␤-catenin-Tcf/Lef transcriptional activity and subsequent block-ade of ␤-cateninor Wnt-induced the inhibition of type II collagen expression. Our results collectively suggest that ␣-catenin functions in chondrocytes not only as a cytoskeletal component but also as a modulator of ␤-catenin signaling.