INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Cadherins are a family of structurally and functionally related molecules that mediate Ca²⁺-dependent cell-cell interactions in a homophilic manner (1). They are subdivided according to their primary structures into several groups, such as the classical cadherins and the protocadherins (2). The classical cadherins, including the E-, P-, and N-cadherins, are highly conserved transmembrane glycoproteins with similar domains for homophilic binding, Ca²⁺-binding, and interaction with intracellular proteins termed catenins (-, -, and -catenins) (1, 2).

The function of cadherins depends on their association with the actin cytoskeleton. Deletion or truncation of the cytoplasmic domain of cadherin results in a loss of function despite their continued expression on the cell surface (3). The interaction of cadherins with the cytoskeleton is mediated by -, -, and -catenins (4, 5). Consistent with the role of the catenins in the linking of cadherins to the cytoskeleton, a loss of the cadherin function has been correlated with alterations in catenins (6-8). Furthermore, cell-cell adhesion could be restored by transfection of these cells lacking - or -catenin with the respective cDNA (8-10). In vitro and in vivo experiments have shown that - and -catenin bind directly to the cytoplasmic domain of E-cadherin, whereas -catenin binds to - or -catenin (11-18). The -catenin-binding site of both - and -catenin is the amino-terminal part (11, 12, 15, 17). The central core region, consisting of thirteen copies of the so-called Armadillo repeat, is involved in the association of both - and -catenin with cadherin (11-14, 18, 19). The core of both is also involved in complex formation with the APC tumor suppressor protein (12, 13, 15) and, in the case of -catenin (plakoglobin), with desmogleins and desmocollins, the desmosomal cadherins (19-21). The region of -catenin responsible for - and -catenin binding was recently identified as the amino-terminal region of the molecule (22, 23). -Catenin mediates the interaction between the cadherin·catenin complex and the actin cytoskeleton through its association with -actinin (23, 24) and actin filaments (25). The complex formation is, however, a dynamic process including the exchange of catenins (26).

p120^cas is a protein that is tyrosine phosphorylated in cells transformed with Src and in response to growth factors such as epidermal growth factor.¹ Like - and -catenin, p120^cas possesses 11 copies of the Armadillo repeat (27) and also associates with E-cadherin·catenin complexes (28-30).

Cells have to regulate the strength of their adhesiveness, and post-translational modifications such as tyrosine phosphorylation may be one mechanism of regulation. In v-src transfected cells, -catenin in the cadherin·catenin complex is preferentially tyrosine-phosphorylated, and the increased tyrosine phosphorylation of -catenin is apparently associated with a dysfunction of cadherin (31-33). Treatment of adenocarcinoma cells with growth factors results in the accumulation of phosphorylated tyrosine residues on - and -catenin, concomitant with a loss of cadherin-mediated adhesion (34). These data suggest that the degree of tyrosine phosphorylation of -catenin is a critical parameter controlling cadherin function, possibly by regulating the association of cadherin with the cytoskeleton. In the case of Ras-transformed mammary cells, p120^cas seems to displace -catenin for cadherin binding upon elevated tyrosine phosphorylation of these proteins (35). The dissociation of tyrosine-phosphorylated -catenin from N-cadherin has also been reported (36). Our knowledge on the dysfunctional state of cadherin induced by the tyrosine phosphorylation of -catenin, however, remains fragmentary. In this study, we have examined the consequence of tyrosine phosphorylation of - and -catenin on cadherin function and the composition of the cadherin·catenin complex. We find that the tyrosine phosphorylation of - and -catenin induced by pervanadate treatment of cells does not result in the dissociation of these proteins, instead causing a significant decrease in the amount of -catenin in the E-cadherin·catenin complex.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Chemicals and Antibodies-- Sodium orthovanadate, okadaic acid, and cytochalasin D were purchased from Sigma. Monoclonal antibodies against -, -, and -catenins were purchased from Transduction Laboratories. DECMA-1, a monoclonal antibody to E-cadherin, and monospecific antibodies to -catenin were described previously (37, 38). A monoclonal antibody to phosphotyrosine (4G10) was obtained from Upstate Biotechnology.

Cells and Transfection-- Human leukemia K562 cells (kindly provided by Dr. K. Sekiguchi, Research Institute, Osaka Medical Center for Maternal and Child Health) were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. The E-cadherin cDNA encoding the wild-type or mutant proteins, EC37 (5) or ED134A (39), was cloned into a mammalian expression vector, pCAGGS neo (40) (a gift from Dr. K. Yamamura, Kumamoto University). For the expression of an E-cadherin--catenin chimeric protein, the ClaI-EcoRV fragment of E-cadherin cDNA that encodes the 71 amino acid, including catenin-binding domain of E-cadherin, was replaced with a 2469-base pair Eco47III-EcoRV fragment that encodes the carboxyl-terminal two-thirds of -catenin (amino acids 301-906). K562 cells (5 × 10⁶) were transfected with the expression vectors (10 µg) by electroporation using a Bio-Rad Gene Pulser set at 280 V and 960 microfarads.

Immunoprecipitation and Immunoblotting-- For immunoblot analysis, cells (1 × 10⁵) were boiled for 5 min in Laemmli SDS gel sample buffer, run on 8% polyacrylamide gels, and then electroblotted onto nitrocellulose membranes. The membranes were blocked with 5% nonfat milk in phosphate-buffered saline and then incubated with monoclonal antibodies and finally peroxidase-conjugated antibodies (Jackson ImmunoResearch Laboratories). After washing with the buffer containing 0.1% Tween 20, the protein bands were visualized with an ECL detection kit (Amersham Corp.). Immunoprecipitation was carried out as described previously (4) with the following modifications. Cells were lysed in 10 mM Tris-HCl buffer, pH 7.6, containing 1% Triton X-100, 0.5% Nonidet P-40, 150 mM NaCl, 1 mM CaCl₂, 0.1 mM sodium orthovanadate, 1 mM phenylmethylsufonyl fluoride, 10 µg/ml leupeptin, and 25 µg/ml aprotinin. The E-cadherin·catenin complex was collected with either rabbit anti-E-cadherin or -catenin antibodies, both of which had been preabsorbed to protein A-Sepharose CL4B. The immune complex was washed with the same buffer four times and then boiled in the SDS-PAGE sample buffer.

Cell Aggregation and Dissociation-- The cell aggregation assay was performed as described previously (5) except that the cells were passed through Pasteur pipettes several times to obtain single cells. For the cell dissociation assay, cells (2 × 10⁶) were cultured overnight in 60-mm plastic dishes and then incubated for another 1 h after the addition of reagents dissolved in a minimum volume of an appropriate solvent.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Expression of E-Cadherin on K562 Leukemia Cells-- A commonly used method devised by Takeichi and co-workers (41) to examine potential homophilic adhesive properties of surface molecules includes the transfection of mouse L fibroblasts with the cDNA of interest and analysis of the adhesive properties of the transfectants. Although L cells have no endogenous cadherin activity, they exhibit adhesiveness and form aggregates after overnight culture in suspension, suggesting that they have cell-cell adhesion molecules other than cadherin. To determine the conditions that modulate the cadherin activity of the preformed cell-cell contacts, L cells were less suitable because of their cadherin-independent adhesiveness. A human leukemia cell line, K562, was chosen for analysis because it expresses no endogenous cadherin and grows in suspension as non-adhesive single cells. K562 cells transfected with an E-cadherin expression vector and selected for G418 resistance exhibited two different phenotypes. One type grows as aggregates and the other as single cells. On immunofluorescence staining with E-cadherin antibodies, the clones growing as aggregates were found to be positive for E-cadherin expression, whereas those growing as single cells were negative (not shown).

View larger version (138K): [in this window] [in a new window]   Fig. 1.   Aggregation of K562 cells expressing wild-type E-cadherin but not mutant E-cadherin polypeptides. K562 cells expressing the wild-type E-cadherin (EK cells, panels A and E); nonfunctional E-cadherins, i.e. either ED134A (ED134AK, panels B and F), with an amino acid substitution in the Ca2+-binding motif; or EC37 (EC37K, panels C and G), with a deletion in the cytoplasmic domain, or K562 cells transfected with the control vector (NK, D and H) were allowed to aggregate for 1 h in the absence (A, B, C, and D) or presence (E, F, G, and H) of an anti-E-cadherin monoclonal antibody. Bar, 50 µm.

View larger version (56K): [in this window] [in a new window]   Fig. 2.   Stabilization of catenins by E-cadherin expression. K562 cells expressing wild-type E-cadherin (lanes 1), nonfunctional E-cadherins, i.e. either ED134A (lanes 2) or EC37 (lanes 3), or K562 cells transfected with the control vector (lanes 4), or parental K562 cells (lanes 5) were subjected to immunoblot analysis.

Complex Formation of E-Cadherin in K562 Cells-- Cadherin·catenin complex formation was studied by means of co-immunoprecipitation experiments. Cells were metabolically labeled with [³⁵S]methionine, and E-cadherin was immunoprecipitated from cell lysates with E-cadherin antibodies. Three proteins migrating to positions corresponding to 102, 88, and 82 kDa were coprecipitated with E-cadherin (120 kDa) in the case of EK cells (Fig. 3). These coprecipitated proteins were identified as -, -, and -catenin by subjecting the immunoprecipitates to immunoblot analysis with the respective antibodies (data not shown, but see below).

View larger version (56K): [in this window] [in a new window]   Fig. 3.   Complex formation of E-cadherin with catenins. Cells labeled overnight with [35S]methionine were lysed and then subjected to immunoprecipitation with an E-cadherin antibody. The immunoprecipitates were analyzed by SDS-PAGE and fluorography. L cells expressing E-cadherin (EL) were used as a positive control. The positions of E-cadherin and -, -, and -catenins are indicated on the right.

Dissociation of EK Cell Aggregates on Pervanadate Treatment-- The addition of pervanadate (0.5 mM sodium orthovanadate + 1.5 mM H₂O₂), an inhibitor of phosphotyrosine phosphatases, to EK cell aggregates formed on overnight culture resulted in almost complete dissociation of the aggregates into single cells within 1 h (Fig. 4). Okadaic acid (up to 500 nM), an inhibitor of serine-threonine phosphatase, induced morphological changes, including blebbing of individual cells, in the aggregates, but it had no effect on the aggregates of EK cells (not shown).

View larger version (87K): [in this window] [in a new window]   Fig. 4.   Dissociation of EK cell aggregates in the presence of pervanadate. Cells were allowed to aggregate overnight (A) and then incubated for 1 h in the presence of pervanadate (B). Bar, 50 µm.

Pervanadate Treatment Does Not Affect the Amounts of the Components of the cadherin·catenin Complex-- We examined whether the amounts of the components of the cadherin·catenin complex were changed after pervanadate treatment. Immunoblot analysis showed that there was almost no difference in the amounts of E-cadherin, -, -, -catenins, or p120^cas between the untreated and treated EK cell aggregates (Fig. 5). A small but reproducible shift was detected in the electrophoretic mobility of - catenin, -catenin, and p120^cas (Fig. 5).

View larger version (66K): [in this window] [in a new window]   Fig. 5.   Immunoblot detection of E-cadherin and catenins. EK cell aggregates treated with pervanadate (+), or untreated () were lysed in SDS sample buffer and then subjected to analysis.

Elevated Tyrosine Phosphorylation of -Catenin and -Catenin, and Reduced Association of -Catenin with the E-Cadherin·Catenin Complex after Pervanadate Treatment-- We next examined whether pervanadate treatment affected the association of E-cadherin with catenins by immunoblot analysis of immunoprecipitates collected with E-cadherin antibodies. In untreated EK cell aggregates, E-cadherin was associated with -, -, and -catenin, and no detectable tyrosine phosphorylation of these proteins was observed after blotting with anti-phosphotyrosine antibodies (Fig. 6). After pervanadate treatment, similar amounts of - and -catenin (with modified electrophoretic mobilities of 90 and 84 kDa, respectively) were found associated with E-cadherin. The amount of -catenin in the complex was, however, significantly decreased to less than one-half that of untreated aggregates (Fig. 6). The anti-phosphotyrosine antibody specifically reacted with three bands, at 120, 90, and 84 kDa, for the treated aggregates. Immunoprecipitation with E-cadherin antibodies under stringent conditions revealed that E-cadherin was tyrosine phosphorylated upon pervanadate treatment (data not shown). A similar analysis of -catenin collected with -catenin antibodies revealed that -catenin was not detectably phosphorylated (data not shown). Taken together, these results suggest that, upon increased tyrosine phosphorylation of components of the cadherin·catenin complex, -catenin is released from the complex.

View larger version (52K): [in this window] [in a new window]   Fig. 6.   Immunoblot analysis of catenins co-precipitating with E-cadherin. EK cell aggregates were incubated for 1 h in the absence () or presence (+) of pervanadate and then lysed. E-cadherin and the associated proteins were collected with an E-cadherin antibody. After SDS-PAGE and transfer to nitrocellulose membranes, the proteins were stained with antibodies either E-cadherin, -, -, or -catenin, p120cas, or phosphotyrosine. Under the conditions used, p120cas was not detected in the immunoprecipitates.

View larger version (44K): [in this window] [in a new window]   Fig. 7.   Immunoblot analysis of the proteins co-precipitating with -catenin. EK cell aggregates were incubated for 1 h in the absence () or presence (+) of pervanadate and then lysed. The proteins associated with -catenin were collected by precipitation with an -catenin antibody. After SDS-PAGE and transfer to nitrocellulose membranes, the proteins were stained with either E-cadherin or -, -, or -catenin antibodies.

Aggregates of K562 Cells Expressing an E-Cadherin--Catenin Chimera Resist Dissociation by Pervanadate Treatment-- To test our hypothesis that the dissociation of -catenin from the E-cadherin·catenin complex underlies the dysfunction of E-cadherin, we constructed a cDNA for an E-cadherin--catenin chimeric molecule consisting of (a) the entire extracellular and transmembrane domains of E-cadherin as well as the first 80 amino acids of its cytoplasmic domain, excluding the region shown to associate with - or -catenin (5), and (b) amino acids 301-906 of -catenin, which include the domains necessary for association with -actinin and actin (23, 25), but not the domain essential for association with - and -catenins (22, 23) (Fig. 8A). This chimeric cDNA was transfected into K562 cells. Clones expressing the chimeric protein on their surface (E-CNK) were confirmed by immunofluorescence staining (data not shown), and by immunoblotting with anti-E-cadherin (Fig. 8B) and anti--catenin antibodies (not shown), both reacted with a protein of 170 kDa, the expected molecular mass of the chimera. The cells form aggregates (Fig. 8C) in an E-cadherin-dependent manner in that the aggregation was inhibited by the E-cadherin antibody (data not shown). In contrast to EK cell aggregates, the preformed aggregates were not dissociated upon treatment with pervanadate (Fig. 8D).

View larger version (77K): [in this window] [in a new window]   Fig. 8.   Resistance to pervanadate of aggregates of K562 cells expressing an E-cadherin--catenin chimeric protein. Schematic representation of an E-cadherin--catenin chimeric protein (A). The sequence derived from -catenin was shaded. K562 cells expressing wild-type E-cadherin (lane 1) or the chimeric molecule (lane 2) were subjected to immunoblot analysis with E-cadherin antibodies (B). Samples were run on 7% polyacrylamide gels. Cells expressing the chimeric protein were allowed to aggregate overnight (C) and then incubated for 1 h in the presence of pervanadate (D). Bar, 50 µm.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

The functions of cadherins are regulated from the cytoplasmic side, and these regulations are thought to be of great importance in the development of organs and in tumor metastasis. To understand the mechanisms regulating the cadherin function, especially to identify intrinsic factors involved in the inactivation of cadherin, we expressed E-cadherin on K562 leukemia cells. Since the formation of aggregates by the transfected K562 cells was solely mediated by E-cadherin, the system seems to be suitable for the analysis of cellular factors or of changes that lead to inactivation of cadherin.

The aggregates dissociate into single cells on the addition of pervanadate. In pervanadate-treated cells, - and -catenins exhibit increased tyrosine phosphorylation and conformational changes, and these changes are correlated with reduced association of -catenin to the E-cadherin·catenin complex. Since the increased tyrosine phosphorylation of - and -catenins and E-cadherin did not affect their interactions with each other, one of the molecular events resulting in the inactivation of E-cadherin is apparently the dissociation of -catenin from the complex. Consistent with this idea, the E-cadherin chimeric molecule covalently linked with -catenin was not inactivated by pervanadate treatment. However, in these cells, numerous proteins exhibit increased tyrosine phosphorylation, and it will be important in the future to determine if any other factors contribute to the inactivation of E-cadherin. With elevated tyrosine phosphorylation, there is a rapid reorganization of the actin cytoskeleton and a modulation of intercellular adherens-type junctions in epithelial and endothelial cells (46, 47). Disruption of the actin cytoskeleton by cytochalasin D did not, however, affect aggregation.

The correlation between the presence of phosphorylated tyrosine residues on - catenin and the loss of cadherin-mediated adhesion is consistent with reports from other laboratories. The dissociation of -catenin from the cadherin·catenin complex has not, however, been reported previously. Instead, the displacement of -catenin in the complex by p120^cas (35), the dissociation of -catenin from the complex (36), and no apparent changes in the composition of the complex upon tyrosine phosphorylation of -catenin have been reported (31-34). Differences in the protein kinases involved in the phosphorylation of these proteins could explain this discrepancy. BCR-ABL oncogene product exhibiting increased tyrosine kinase activity in K562 cells (48) is a candidate for the kinases responsible for the tyrosine phosphorylation of -catenin and -catenin in the pervanadate-treated cells. Association with the actin cytoskeleton and possible involvement in the cell adhesion of BCR-ABL protein has been reported (49). Further studies are needed to determine the nature of the protein-tyrosine kinases responsible for the elevated tyrosine phosphorylation of E-cadherin and - and -catenins in pervanadate-treated EK cell aggregates.

Besides protein-tyrosine phosphatase 1B-like phosphatase, which has been shown to be involved in regulation of N-cadherin function in chicken retina cells (36), several other protein tyrosine phosphatases have been shown to interact with E-cadherin or -catenin, including receptor-type protein tyrosine phosphatase µ (50), a member of the leukocyte antigen-related protein-related transmembrane tyrosine phosphatase family (51), and protein-tyrosine phosphatase  (52). The protein-tyrosine phosphatases involved in keeping the tyrosine phosphorylation of - and -catenins at low levels also should be identified.

The mechanism by which the tyrosine phosphorylation of - and -catenins results in the dissociation of -catenin from these proteins is unknown at present. In the case of tyrosine phosphorylation of -catenin induced by epidermal growth factor, the phosphorylated tyrosine residues are located in either the amino- or the carboxyl-terminal domain of the protein (53). As mentioned above, the -catenin-binding site of both - and -catenins has been localized to the amino-terminal domain of these proteins (11, 12, 14, 15, 17). Analysis by alanine mapping of the -catenin binding site of the proteins has revealed that hydrophobic amino acids are indispensable for the interaction with -catenin, suggesting that hydrophobic interactions stabilize the heterodimeric complex (17). The amino acid substitution of the tyrosine residue in the site of - and -catenin with alanine (Y142A and Y133A, respectively) also strongly reduces the ability of these proteins to bind to -catenin (17). The amino acid sequences surrounding this tyrosine residue seem, however, not to conform to the consensus sequence of previously identified tyrosine kinases.

An alternative possibility to be considered is that phosphorylation of tyrosine residues not in the -catenin binding site induces a conformational change, which in turn weakens the interaction with -catenin or masks the binding site. Upon pervanadate treatment resulting in tyrosine phosphorylation, - and -catenins show a slower electrophoretic mobility, suggesting a phosphorylation-induced conformational change of these proteins. It has been reported that after the treatment of cells with serine-threonine phosphatase inhibitors (okadaic acid and calyculin A), -catenin was phosphorylated on serine and to a lesser extent on threonine residues, and showed a slower electrophoretic mobility (54).