Cell Motility and the Cytoskeleton 40:87–100 (1998) Plakoglobin Induces Desmosome Formation and Epidermoid Phenotype in N-Cadherin-Expressing Squamous Carcinoma Cells Deficient in Plakoglobin and E-Cadherin Henry R. Parker,1 Zhi Li,1 Hannah Sheinin,2 Gille Lauzon,2 and Manijeh Pasdar1* 1Department of Cell Biology and Anatomy, University of Alberta, Edmonton, Alberta, Canada 2Department of Dermatology, University of Alberta, Edmonton, Alberta, Canada Pg is a homologue of b-catenin and Armadillo, the product of the Drosophila segment polarity gene and has been shown to have both adhesive and signaling functions. It interacts with both classic and desmosomal cadherins. Pg interaction with the desmosomal cadherins is essential for desmosome assembly. Its precise role in the classic cadherin complexes is unclear, although Pg-E–cadherin interaction appears to be necessary for the formation of desmosomes. In addition to cadherins in adhesion complexes, Pg interacts with a number of proteins involved in regulation of cell differentiation and proliferation such as Lef-1/Tcf-1 transcription factors and the tumor suppressor protein APC. In this study, we have introduced Pg cDNA into SCC9 cells, a Pg- and E-cadherin-deficient squamous cell carcinoma line, which also lacks desmosomes. These cells have both a-catenin and b-catenin, display unusual expression of N-cadherin, and have the typical fibroblastic phenotype of transformed cells. Pg-expressing SCC9 cells (SCC9P) formed desmosomes. Desmosome formation coincided with the appearance of an epidermoid phenotype, with increased adhesiveness and a contact-dependent decrease in growth. Biochemical characterization of SCC9P cells showed an increase in the expression and stability of N-cadherin and a decrease in level and stability of b-catenin, without any apparent effects on a-catenin. These results show that, in the absence of E-cadherin, Pg can efficiently use N-cadherin to induce desmosome formation and epidermoid phenotype. They also suggest a role for Pg as one of the regulators of the intracellular b-catenin levels and underscore the pivotal role of this protein in regulating cell adhesion and differentiation. Cell Motil. Cytoskeleton 40:87–100, 1998. r 1998 Wiley-Liss, Inc. Key words: plakoglobin; b-catenin; cadherins; tumorigenesis; epidermoid; squamous carcinomas INTRODUCTION Plakoglobin (Pg, 83 kDa) is a multifunctional protein that was first studied as a cytoplasmic component common to both kinds of epithelial adhering junctions (desmosomes and adherens junctions) [Cowin et al., 1986]. Adherens junctions and desmosomes connect cells to each other and anchor the cytoskeleton to the plasma membrane [for reviews, see Geiger and Ayalon, 1992; r 1998 Wiley-Liss, Inc. Contract grant sponsor: Medical Research Council of Canada; Contract grant sponsor: Leo Laboratories Canada Ltd Dr. Parker is currently at the Department of Biochemistry, University of Alberta, Edmonton, Alberta, Canada. *Correspondence to: Manijeh Pasdar, Department of Cell Biology and Anatomy, University of Alberta, Edmonton, Alberta T6G 2H7, Canada; E-mail: email@example.com Received 9 December 1997; accepted 12 February 1998 88 Parker et al. Garrod et al., 1996; Schmidt et al., 1994]. The transmembrane domain of adherens junctions is enriched in E-cadherin, which interacts with the actin cytoskeleton via the cytoplasmic catenins [Aberle et al., 1996; Kemler, 1993]. b-Catenin is structurally similar to Pg and interacts directly with the classic cadherins. a-Catenin shows some structural similarity to vinculin, exhibits actinbinding activity, interacts with cadherins through b-catenin or Pg, and connects the cadherin–catenin complex to the actin cytoskeleton [Rimm et al., 1995; Knudsen et al., 1995]. g-Catenin is identical to Pg [Knudsen et al., 1992; Peifer et al., 1992]. Pg interacts with both classic and desmosomal cadherins [for a recent review, see Cowin, 1994]. Pg is an indispensable structural element of the desmosome. At this junction, desmosomal cadherins, desmogleins (Dsg) and desmocollins (Dsc) constitute the transmembrane domain and Pg is essential to mediate the interaction between these cadherins and the cytokeratin intermediate filaments via desmosomal plaque proteins, desmosplakins (DP) [Mathur et al., 1994; Troyanovsky, 1994; Bierkamp et al., 1996; Ruiz et al., 1996; Palka and Green, 1997; Kowalczyk et al., 1997]. Pg and b-catenin interact with classic cadherins in a mutually exclusive manner [Hinck et al., 1994]. The cytoplasmic domains of desmosomal and classic cadherins share sequence similarity in a small region which is the Pg site of interaction [Mathur et al.; 1994; Jou et al., 1995; Aberle et al., 1994]. Although Pg is not an essential structural component of adherens junction, its interaction with E- or P-cadherin in epithelial cells appears to be necessary for the formation of desmosomes [Lewis et al., 1997]. In addition to interaction with cadherins and its structural role in junctional complexes, Pg, as is also true of its closely related protein b-catenin, interacts with a number of proteins involved in pathways regulating cell fate determination and cellular growth and proliferation [for a review, see Peifer, 1996, and references therein]. One such protein is the product of the tumor suppressor gene APC, which has been shown to be the major regulator of the cytoplasmic b-catenin levels. APC binds to b-catenin and Pg and via either of these two proteins to acatenin. This protein is also a component of the wnt signaling pathway. The vertebrate wnt and its Drosophila homologue wingless (wg) are involved in the determination of embryonic segment polarity [Nusse, and Varmus, 1992; Parkin et al., 1993; Perrimon, 1994; Burrus, 1994; Moon et al., 1997]. In vertebrates, overexpression of wnt proteins causes duplication of dorsal axis. Certain tissue culture cells exposed to wnt show increased levels of Pg and b-catenin [Bradley et al. 1993; Hinck et al., 1994]. Increased expression of Pg by itself has also been shown to induce axis duplication in Xenopus [Karnovsky and Klymkowsky, 1994]. Pg also interacts with transcription factors Lef-1/Tcf-1 and, if overexpressed, Pg and b-catenin can be translocated into the nucleus [Karnovsky and Klymkowsky, 1994; Behrens et al., 1996; Huber et al., 1996b]. The adhesive and signaling functions of Pg are regulated by its interacting protein partners. To gain further knowledge into the functions of Pg, we have used a squamous carcinoma cell line (SCC9) that lacks both Pg and E-cadherin and does not assemble desmosomes. These cells have both a- and b-catenins but display unusual expression of N-cadherin and have typical fibroblastoid morphology of transformed cells. We have shown that transfection of SCC9 cells with L-CAM or E-cadherin cDNAs induces a morphologic transformation from fibroblast to epidermoid which coincides with downregulation of the endogenous N-cadherin and increased synthesis and stability of the catenins [Li et al., in press]. Here, we show that the introduction of Pg cDNA into SCC9 cells enables them to form desmosomes and induces a fibroblast to epidermoid transition, similar to that observed for L-CAM. However, unlike the effects of E-cadherin expression, these Pg-induced changes coincide with increased stability and steady-state level of N-cadherin and decreased level and stability of b-catenin without any significant effects on a-catenin. MATERIALS AND METHODS Cells and Culture Conditions The human squamous carcinoma cells SCC9, SCC15, and SCC25, all derived from carcinomas of the tongue (American Type Culture Collection [ATCC], Rockville, MD) were provided by Dr. A. Klein-Santoz (Fox Chase Cancer Center, Philadelphia, PA). These lines were maintained in minimum essential medium (MEM) (Sigma Chemical Co., St. Louis, MO) supplemented with 10% fetal bovine serum (FBS). SCC9 cells that were transfected with the Pg expression plasmid (hereafter referred to as SCC9P) were maintained in the same medium containing 200 µg/ml geneticin (G418, Life Technologies, Grand Island, NY). SCC9L, L-CAMexpressing SCC9 cells have been described previously [Li et al., in press] and were maintained under the same conditions as described for SCC9P cells. Primary cultures of keratinocytes from neonatal foreskins were established in low Ca21-containing keratinocyte medium (Life Technologies) supplemented with epidermal growth factor (EGF), pituitary extract, and 20% FBS. Before metabolic labeling (30 min–2 h), and to facilitate methionine uptake, all cultures were incubated in MEM, with calcium levels adjusted to 5 µM (low Ca21 concentration medium [LCM]). Plakoglobin Induces Fibroblast to Epidermoid Transition Plasmid Construction and Transfection A full-length cDNA for human Pg was produced using standard polymerase chain reaction (PCR) protocols based on the published sequences [Franke et al., 1989] with the incorporation of BamHI restriction sites in the primers. The authenticity of the PCR product was verified by sequencing. The entire coding sequence was digested with BamHI and cloned into the mammalian episomal expression vector, pREP9 (Invitrogen, San Diego, CA) creating pREP9Pg. Positive clones were selected for proper orientation relative to the vector’s promoter, Rous sarcoma virus (RSV), long terminal repeat enhancer/promoter. This plasmid was used to transfect SCC9 cells using LipofectAMINE (Gibco-BRL, Grand Island, NY) reagent, according to the manufacturer’s protocol. Briefly, SCC9 cells were grown in 60-mm culture dishes to 80% confluency, at which time the medium was replaced with LCM. Transfection was performed 12–16 h later using 25 µg (per dish) of either pREP9 or pREP9Pg. At 48 h post-transfection, medium was replaced with MEM containing 400 µg/ml G418. Resistant colonies were selected and screened for Pg expression, using immunofluorescence and immunoblotting assays (see below). Positive colonies (SCC9P) were subcultured by limiting dilution to obtain single-cell isolated colonies. Antibodies and cDNA Probes Rabbit polyclonal anti-a-catenin, b-catenin, and pan-cadherin were purchased from Sigma. We recently showed that the single cadherin detected by the pancadherin antibodies in SCC9 cells is N-cadherin [Li et al, in press]; therefore, it is referred to as such throughout this report. Rabbit polyclonal anti-Pg antibodies have been described [Pasdar et al., 1995a]. The secondary antibodies used in immunofluorescence studies were purchased from Boehringer-Mannheim Biochemicals (Indianapolis, IN). The N-cadherin cDNA probe was provided by Dr. John Hemperly (Becton Dickinson Research Center, Research Triangle Park, NC) and corresponds to the 38 EcoR1 fragment (1.1 kb) of the human N-cadherin cDNA [Reid and Hemperly, 1990]. A 0.6-kb BamH1 fragment of mouse a-catenin cDNA and a 1.3 -kb EcoR1 fragment of mouse b-catenin cDNA were provided by Dr. Lionel Larue (Curie Institute, Orsay, France). The Pg probe corresponds to the 1.5-kb Sac1 fragment of the human cDNA. Metabolic Labeling Turnover studies were performed as described [Pasdar, 1992]. Briefly, confluent cultures of SCC9 and SCC9P cells were grown in MEM in 35-mm culture dishes. The medium was removed, cells washed, and LCM added for 1 h, followed by removal, wash, and 89 addition of methionine-free LCM for another hour. Cultures were metabolically labeled with 125 µCi of 35S-methionine (1,189 Ci/mmol, ICN Biochemicals, Irvine, CA) in methonine-free LCM for 30 min, then chased for different times (0–24 h) in .10,000-fold excess of unlabeled methionine in MEM. For protein interaction studies, cells were grown on 60-mm culture dishes and processed as above except the labeling was with 250 µCi 35S-methionine for 1 h, at which time Ca21 was added to the media, bringing the concentration to 1.8 mM, and labeling continued for another 3 h.Upon completion of labeling or chasing, cells were kept at 4°C, rinsed twice with phosphate-buffered saline (PBS) and processed for extraction, followed by immunoprecipitation (see below). Cell Fractionation and Immunoprecipitation Cells were extracted in the culture dish with 1 ml of cytoskeleton extraction (CSK) buffer containing 300 mM sucrose, 10 mM Pipes, pH 6.8, 50 mM NaCl, 3 mM MgCl2, 0.5% (v/v) Triton X-100, 1.2 mM PMSF, 0.1 mg/ml DNase, and 0.1 mg/ml RNase [Pasdar, 1992]. This extraction buffer allows the separation of the cytoskeletonassociated (insoluble) from the soluble pool of cadherin and catenins. Cells were removed from the dishes with a rubber policeman and centrifuged at 48,000g for 10 min. The soluble fraction was removed from the insoluble pellet. The pellet was resuspended in 100 µl of sodium dodecyl sulfate (SDS) buffer (1% SDS, 10 mM Tris-HCl, pH 7.5, 2 mM EDTA, 0.5 mM DTT, 0.5 mM PMSF) and boiled for 10 min. Before immunoprecipitation, the SDS concentration was reduced to 0.1% by the addition of 900 µl CSK buffer. The soluble and insoluble fractions were precleared for 30 min by incubation with 10 µl of preimmune sera and 35 µl of Pansorbin cells (Calbiochem, La Jolla, CA) as described [Pasdar, 1992]. After a 7-min centrifugation at 14,000g, precleared extracts were divided into 4 equal aliquots and processed for immunoprecipitation with 2–3 µl of different antibodies, either Pg, pan-cadherin, a-catenin, or b-catenin. To ensure complete depletion, each aliquot was immunoprecipitated three times. Protein interactions were analyzed in the soluble fractions. These analyses did not include the insoluble fractions because the insoluble proteins have to be solubilized before immunoprecipitation and the solubilization conditions (1% SDS at 100°C) causes dissociation of the complex. The soluble proteins of metabolically labeled cells were first immunoprecipitated with anti-pancadherin under low stringency conditions as described [Pasdar et al., 1995a]. The resulting immune complexes were eluted from the protein A-Sepharose beads by boiling for 20 min followed by centrifugation at 14,000g for 5 min. Supernatants were then transferred to new tubes and the SDS concentration adjusted to 0.1% by the 90 Parker et al. addition of 900 µl CSK buffer. The second immunoprecipitations were performed under high stringency conditions with anti-a- and b-catenins, and Pg antibodies as described above. Immunoprecipitates (200 µg total protein for each cell line and/or time point, unless otherwise stated) were boiled in 65 µl SDS sample buffer and separated by electrophoresis on SDS 6% polyacrylamide gels, followed by fluorography. For protein turnover studies, the relative amount of radioactivity in the protein bands corresponding to Pg, N-cadherin, and a- and b-catenins was determined by scanning densitometry from multiple exposures of resulting fluorograms [Pasdar, 1992]. Each experiment was repeated 3–5 times; the results are presented for one typical experiment. Immunoblotting Total cell extracts were prepared by directly solubilizing cells in SDS sample buffer. Total and CSK soluble and insoluble cell extracts were resolved in SDS polyacrylamide gels, and the proteins electrophoretically transferred onto nitrocellulose membranes as described previously [Towbin et al., 1979; Pasdar and Nelson, 1988]. Membranes were incubated sequentially with 1/1,000 dilution of primary antibodies (Pg, pan-cadherin, a-catenin, and b-catenin) for 1.5 h, followed by a 30-min incubation with 125I-protein A (1 mCi/ml, ICN) and were exposed at 280°C to XAR-5 x-ray films (Kodak) using two intensifying screens. Unless otherwise stated equivalent of 50 mg total protein from each cell line were analyzed in immunoblots. Northern Blotting RNA was isolated from cells of various cell lines and primary keratinocytes using Trizol reagent (Gibco), following the manufacturer’s protocol. Approximately 25 µg of the resulting RNA was electrophoresed in formaldehyde-containing 1% agarose gels, and transferred to Hybond N membranes (Amersham, Arlington Heights, Ill.). Prehybridization and hybridization conditions were as described [Church and Gilbert, 1984]. After hybridization with 32P-labeled cDNA probes, membranes were processed for autoradiography. Immunofluorescence and Electron Microscopy SCC9 and SCC9P cells were grown in MEM on collagen-coated coverslips, extracted with CSK buffer, fixed with formaldehyde, and processed for indirect immunofluorescence and confocal microscopy as described [Pasdar et al., 1995a]. All antibodies were diluted in PBS, Pg (1/100), catenins, and pan-cadherin (1/250), secondary rhodamine-conjugated goat anti-rabbit IgG (1/100). Coverslips were mounted in elvanol containing 0.2% paraphenylene diamine (PPD), pH 8.6, viewed with a 1003 objective ,using a laser scanning confocal microscope (Leica Lasertechnic, Heidelberg). Optical sections through the middle of the cells were recorded on optical disks and viewed on a Silicon Graphics imager (SGI). For electron microscopy, cultures were grown to confluency on collagen-coated coverslips. Cells were fixed in a freshly made solution containing 1.25% glutaraldehyde, 1% OsO4 in 50 mM Pipes, pH 7.2, for 5 min at room temperature and processed for electron microscopy as described [Pasdar et al., 1995b] and examined using a Philips 410 transmission electron microscope. Cell Growth and Aggregation Assays To measure growth, 3 3 104 SCC9 and SCC9P cells were each plated in replicate 35-mm culture dishes. At 1, 3, 5, 7, 9, and 11 days after plating, cultures were trypsinized and cells counted. Aggregation assays were performed as described [Li et al., in press] and entailed triplicate confluent 60-mm culture dishes of SCC9 and SCC9P cells as stocks for single-cell suspensions. After trypsinization (0.04% trypsin in PBS containing 1 mM Ca21) of each culture, replicates of 2 3 106 single cells were resuspended into 2 ml aliquots. Cells were incubated at 37°C with constant shaking at 90 rpm for a total of 2 h. At 30-min intervals, 200-µl aliquots were withdrawn and the number of single cells counted using a hemocytometer. The percentage of aggregation represents the percent decrease in the number of single cells at each time point. RESULTS SCC9 Cells Lack Plakoglobin Figure 1A shows the immunofluorescence staining of Pg in control primary keratinocytes (Fig.1A; KC), the cells of SCC9 and two other, tongue-derived, squamous cell carcinoma lines; SCC15 and SCC25. Pg was absent in SCC9 and expressed in all the other cell lines. In keratinocytes, most of the Pg was present in a punctate peripheral distribution typical of desmosomal staining, whereas in SCC15 and SCC25 Pg was distributed both intracellularly and at the cell surface. The absence of Pg protein in SCC9 cells was supported by immunoblotting of the TX-100-soluble and -insoluble extracts of all cells with anti-Pg antibodies that detected Pg, in both the soluble and insoluble pools, in primary keratinocytes and SCC15 and SCC25, and not in SCC9 (Fig.1B). Furthermore, Northern blotting using a human Pg probe detected mRNAs of the appropriate size (,3.5 kb) in the primary keratinocytes and SCC15 and 25 and not in SCC9 cells (Fig.1C). These three independent assays showed that Pg is not expressed in SCC9 cells and, therefore, that this cell line can be used as a useful model to study the function of Pg. Plakoglobin Induces Fibroblast to Epidermoid Transition 91 Morphology, Growth, and Adhesive Properties of SCC9 and SCC9P Cells Fig. 1. SCC9 cells do not express plakoglobin. A: Primary cultures of neonatal foreskin keratinocytes and cells of human squamous carcinoma cell lines SCC9, SCC15, and SCC25 were established on coverslips, extracted with CSK buffer, and processed for indirect immunofluorescence with a rabbit anti-Pg antibody, as described under Materials and Methods. Bar 5 20 µm. B: The same cultures were grown to confluency and extracted with CSK buffer to separate the CSK soluble (S) and insoluble (P) fractions, followed by immunoblot analysis using the anti-Pg antibodies. C: A total of 25 µg of total RNA from the same cultures was processed for Northern blot analysis, using 32P-labeled cDNA probes for Pg, as described under Materials and Methods. First, SCC9 cells were subcloned by limiting dilution to develop homogeneous fibroblastoid lines. A number of lines were characterized for expression of the epithelial marker proteins, cytokeratins, and various desmosomal proteins, and were all positive. One such clone was used for transfection with pREP9Pg containing the full-length human Pg cDNA. A number of Pgexpressing clones were isolated and used for the studies reported here. All Pg-expressing clones had similar properties. Three clones were fully characterized and the results of the studies performed on one such clone are presented here. SCC9 cultures showed the typical fibroblastoid morphology of transformed cells (Fig. 2A,a). Electron microscopic examination of these cells showed the absence of any well-organized adhering junctions, although the plasma membranes of adjacent cells were closely opposed (Fig.2A,b). Growth rate analysis showed that at single-cell density, these cells grew slowly until they became subconfluent (5 days), after which the rate of growth increased significantly. Upon reaching confluency (day 7), the growth rate of these cells accelerated further, and they continued to grow at a high rate up to 11 days (Fig. 2B; growth, open squares). The aggregation assay showed about 35% decrease in the number of single cells within the first 30 min. There was a small increase in the number of aggregates within the next 90 min, so that by the end of the assay time, 45% of cells were present in aggregate forms (Fig. 2C; open squares). In contrast to the untransfected SCC9 cells, Pgexpressing cells became epidermoid, formed desmosomes and showed decreased growth rate and increased adhesiveness (Fig. 2A,c,d; Fig. 2B,C; solid squares). SCC9P cells were flat and developed tight epidermoid monolayers (Fig. 2A,c). Ultrastructural examination of these cells showed the presence of well formed desmosomes (Fig. 2A,d, arrows). Growth and aggregation characteristics of these cells also underwent changes. Pg expression in these cells led to contact inhibition of growth as once SCC9P cells reached confluency (7 days), they began to die (Fig. 2B; solid squares) The epidermoid morphology of SCC9P cells was associated with increased adhesivity of these cells (Fig. 2C; solid squares). There was a significant increase in aggregation of SCC9P cells from the beginning, which persisted throughout the assay. At 30 min, there were approximately twice as many aggregates in SCC9P as that detected in SCC9 cultures (68% vs. 35%). Within the next 60 min, there was a further increase (to 77%), which remained unchanged for the rest of the assay period (Fig. 2C; solid squares). Together, these results showed that Pg expression in SCC9 cells promoted desmosome formation which coincided with a fibroblast to epidermoid morphological transition associated with contact inhibition of growth and increased adhesiveness. Exogenously Expressed Pg Shows Biological Properties of Pg Expressed in Normal Epithelial Cells SCC9 and SCC9P cells were grown to confluency on collagen-coated coverslips, extracted with CSK buffer, fixed, and stained with anti-Pg antibodies, followed by rhodamine-conjugated goat-anti-rabbit IgG (Fig. 3A). 92 Parker et al. Fig. 2. Induction of desmosome assembly and epidermoid phenotype by the exogenous Pg in SCC9 cells. A: Phase-contrast (left) and electron micrographs (right) of untransfected (SCC9; a and b) and Pg-transfected SCC9 cells (SCC9P, c and d). Arrows, desmosomes. Bars (a and c) 5 10 µm; (b) 5 0.5 µm; (d) 5 0.25 µm. B: Growth and aggregation properties of SCC9 and SCC9P cells. Growth measurements were taken by establishing replicate cultures at single cell density and counting cells at 2-day intervals after trypsinization as described under Materials and Methods. Each time point represents the average of three independent experiments. For the aggregation assays, single cell suspensions were prepared and cells allowed to aggregate under constant shaking as described under Materials and Methods. Aliquots were taken at 30-, 60-, 90-, and 120-min intervals and the number of single cells counted. The rate of aggregation was calculated as the percentage of decrease in the number of single cells. Each value is the average of at least three independent experiments; standard errors are indicated by error bars. The SCC9 cells were negative, whereas SCC9P cells showed Pg staining, most of which was detected at the cell surface in a punctate pattern typical of desmosomes. Immunoblot analysis of the steady-state distribution of Pg indicated a ratio of 40:60 for the CSK soluble (S) and insoluble (P) Pg expressed in SCC9P cells (Fig. 3B), in agreement with similar ratios reported for various cell types previously [Kapprell et al., 1987; Pasdar, 1995a]. A faster migrating band was also detected in the soluble Pg immunoblots (and immunoprecipitates), the identity of Plakoglobin Induces Fibroblast to Epidermoid Transition 93 Fig. 3. Characterization of the exogenously expressed Pg in SCC9 cells. A: (SCC9) and Pg-expressing SCC9 cells (SCC9P) were established on glass coverslips and processed for indirect immunofluorescence with Pg antibodies. Bar 5 10 µm. B: Immunoblot analysis of total (T) and CSK soluble (S) and insoluble (P) fractions of SCC9 and SCC9P cells using anti-Pg antibodies. C: Fate of the newly synthesized exogenously expressed Pg in SCC9 cells. SCC9P cells were grown to confluency, labeled with 35S-methionine for 30 min and chased up to 24 h, as described under Materials and Methods. After cell fractionation, the soluble (S) and insoluble (P) fractions were processed for immunoprecipitation with anti-Pg antibodies and analyzed by PAGE followed by fluorography. D: To estimate the Pg turnover rate, relative amounts of radioactivity in Pg bands from multiple exposures of the fluorograms in C were scanned using a scanning densitometer. E: The exogenously expressed Pg associates with the N-cadherin expressed in SCC9 cells. Confluent cultures of SCC9P cells were 35S-labeled for a total of 4 h, as described under Materials and Methods, and the soluble cell extracts were immunoprecipitated with pan-cadherin antibodies. In a duplicate aliquot, immune complexes were eluted and reprecipitated with Pg antibodies as described under Materials and Methods. The position of N-cadherin, a-catenin, b-catenin, and Pg are indicated in E. which is unclear. It is noteworthy that this protein has also been detected in other cell types [Piepenhagen and Nelson, 1993; Pasdar, 1995a]. Immunoprecipitation of the metabolically labeled SCC9P cell extracts with anti-Pg antibodies indicated a 40:60 ratio for newly synthesized soluble (S) and insoluble (P) Pg at the termination of the 30-min labeling period (Fig. 3C; 0S, 0P), which remained throughout the chase period. The soluble Pg was relatively stable (Fig.3D; open circles; half-life [t1/2] ,9 6 2 h). There was an initial rapid decrease in the amount of the insoluble Pg within the first 2 h of the chase (Fig. 3C; 2S, 2P); after this time point, the protein was degraded gradually with a t1/2 of ,16 6 1.8 h (Fig. 3C,3D, solid circles). Sequential immunoprecipitation showed that the exogenously expressed soluble Pg associated with the N-cadherin expressed in SCC9 cells (Fig. 3E, N-cad -. Pg), consistent with the previous reports on the associa- 94 Parker et al. Fig. 4. Expression of N-cadherin and catenins in SCC9P cells. A: SCC9 and Pg-expressing (SCC9P) cells were processed for indirect immunofluorescence with antibodies to pan-cadherin detecting Ncadherin, a-catenin, and b-catenin. Bar 5 10 µm. B: Immunoblot analysis of the total (T) or CSK soluble (S) and insoluble (P) fractions of SCC9 and SCC9P cells using pan-cadherin detecting N-cadherin and a-catenin, and b-catenin antibodies. C: Quantitation of the amounts of N-cadherin and catenins from three independent experiments, as represented in B. Error bars indicate standard errors. D: SCC9 and SCC9P cells were 35S-labeled and chased as described in Fig. 3C. After cell fractionation, the soluble (S) and insoluble (P) fractions were processed for N-cadherin immunoprecipitation. Immune complexes were analyzed by PAGE, followed by fluorography. Arrows (D), position of N-cadherin precursor. tions between Pg and N- as well as other classic cadherins [Knudsen and Wheelock, 1992; Hamagouchi et al., 1993; Sacco et al., 1995; Simcha et al., 1996; Hertig et al., 1996; Aberle et al., 1996]. Together, the results of these experiments indicated that the exogenously expressed Pg follows the same biosynthetic fate as the Pg normally expressed by epithelial cells. SCC9 and SCC9P cells fixed and processed for indirect immunofluorescence. Staining of SCC9 cells with a pan-cadherin antibody, which detects N-cadherin in SCC9 cells, and antibodies to a- and b-catenins detected both intracellular and cell surface distributions of these proteins (Fig. 4A). N-Cadherin and the catenins were distributed in a similar manner in SCC9 and SCC9P cells (Fig. 4A), although in SCC9P cells, the staining appeared more intense and/or more concentrated at the periphery (Fig. 4A; SCC9P). The redistribution and/or increase in staining was particularly apparent for N-cadherin (Fig. 4A; cf. SCC9P and SCC9 for N-cadherin). We then Characterization of the N-Cadherin and Catenins in SCC9P Cells SCC9 cells express both a- and b-catenins and N-cadherin (Fig. 4). Figure 4A shows CSK extracted Plakoglobin Induces Fibroblast to Epidermoid Transition 95 Fig. 5. Comparison of the fate of newly synthesized adhesion molecules in SCC9 and SCC9P cells. Cultures of SCC9 and SCC9P cells were grown to confluency, labeled with 35S-methionine for 30 min and chased up to 24 h, as described under Materials and Methods. After cell fractionation, the soluble (S) and insoluble (P) fractions were processed for immunoprecipitation with anti-pan-cadherin detecting N-cadherin, a-catenin, and b-catenin antibodies. Immune complexes were analyzed by PAGE, followed by fluorography. The relative amount of radioactivity for each protein band were quantitated by scanning densitometry from multiple exposures of autoradiograms and graphs were plotted using Cricket Graph software. analyzed the steady-state levels of the total (T), and TX-100 soluble (S) and insoluble (P) N-cadherin, a- and b-catenins in SCC9 and SCC9P cells by immunoblotting, followed by quantitation by scanning densitometry (Fig. 4B,C). The results showed increased N-cadherin and decreased b-catenin levels in SCC9P cells, whereas there was very little or no difference between the two lines for a-catenin. (Fig. 4B,C). respectively). We then determined whetherthe Pg expression in SCC9 cells had any effects on the synthesis or metabolic stability of the catenins. Metabolic labeling and turnover analysis of a-catenin in SCC9 and SCC9P cells detected very little or no difference in the amount of newly synthesized soluble and insoluble a-catenin or their metabolic stability between the two lines (Fig. 5; a-catenin, (S) t1/2: 12 6 1.2 h and 10 6 2 h and (P): 5.3 1 0.5 h and 6 6 1.2 h for SCC9 and SCC9P, respectively). b-Catenin turnover studies showed both decreased synthesis and stability for newly synthesized soluble and insoluble b-catenin protein in SCC9P cells. In SCC9P cells, the soluble b-catenin degraded faster (t1/2 ,7 6 1.3 h vs. ,15 6 0.8 h in SCC9 cells; Fig. 5, b-catenin). Similarly, the insoluble b-catenin was degraded faster in SCC9P cells (t1/2 ,6 6 3 h vs. 4 6 1.75 h for SCC9 and SCC9P, respectively). The results of synthesis and turnover studies are consistent with the steady-state levels of N-cadherin and catenins in SCC9 and SCC9P cells. We then sought to determine whether the observed changes in newly synthesized N-cadherin and b-catenin were also reflected in the levels of mRNAs of these proteins in SCC9 and SCC9P cells. In Figure 6, total RNA from SCC9, SCC9P, and Pg Expression Changes the Level and Metabolic Stability of N-Cadherin and b-Catenin in SCC9P Cells SCC9 and SCC9P cells were metabolically labeled and processed for N-cadherin immunoprecipitation as described under Materials and Methods (Fig. 4D). The amount of newly synthesized N-cadherin was significantly more (,4-fold) in SCC9P relative to SCC9 cells, indicating increased N-cadherin synthesis in Pg-expressing cells. Turnover rate analysis showed that in SCC9P cells the soluble N-cadherin degraded faster (cf. SCC9 with SCC9P in Fig. 4D and, in Fig. 5, N-cadherin; (S) t1/2 ,14 6 3.5 and 8 6 1.5 h for SCC9 and SCC9P, respectively), whereas the insoluble N-cadherin became more stable (Figs. 4D and Fig. 5, N-cadherin; (P) t1/2 ,4 6 0.5 h and 12 6 1.6 h for SCC9 and SCC9P, 96 Parker et al. Fig. 6. Northern blot analysis of N-cadherin, a-catenin, and b-catenin mRNAs in SCC9 and SCC9P cells. Total RNA (,25 µg) isolated from SCC9 and SCC9P cells was processed for Northern blot analysis with 32P-labeled N-cadherin, a-catenin and b-catenin cDNA probes as described under Materials and Methods. MDCK (Madin-Darby canine kidney epithelial cells, used as control) cells were hybridized with cDNA probes for a-catenin, b-catenin, and N-cadherin. The results showed a slight decrease in b-catenin in Pg-expressing cells (ratio of b-catenin/total RNA, 0.7 and 0.9 for SCC9P and SCC9, respectively), and an approximately twofold increase in N-cadherin RNA levels in SCC9P cells (ratio of N-cadherin/total RNA, 1.1 and 0.6 for SCC9P and SCC9, respectively), whereas a-catenin RNA levels remained unchanged between the two lines. The Cadherin/Catenin Complex Is Not Altered in SCC9P Cells Figure 7 shows the results of the sequential immunoprecipitation of N-cadherin, followed by either a-catenin, b-catenin, or Pg. In SCC9 cells, N-cadherin, a-catenin, and b-catenin were detected (Fig. 7; SCC9; N-cad). The presence of the catenins in N-cadherin immunoprecipitates was further confirmed by the second immunoprecipitation with the respective antibodies, anti-a- or anti-b-catenin (Fig. 7; SCC9; N-cad -. a-catenin; N-cad -. b-catenin), indicating that N-cadherin formed complexes with both catenins. In these cells, no protein bands were detected in sequential immunoprecipitation of N-cadherin, followed by Pg (Fig.7; SCC9; N-cad -. Pg). In SCC9P cell extracts, the cadherin antibodies brought down 4 bands corresponding to N-cadherin, a-catenin, b-catenin, and Pg (Fig.7; SCC9P). The subse- Fig. 7. Intracellular interactions of newly synthesized adhesion molecules in SCC9 and SCC9P cells. Cultures of SCC9 and SCC9P cells were grown to confluency and 35S-labeled for a total of 4 h, as described under Materials and Methods. After cell fractionation, equal aliquots of the soluble fractions were processed for immunoprecipitation with the Pan-cadherin antibodies, which detect N-cadherin (Ncad). In duplicate samples, immune complexes were eluted and reprecipitated with a-catenin (N-cad -. a-cat), b-catenin (N-cad -. b-cat), or Pg (N-cad -. Pg). quent second immunoprecipitations with the anticatenins and anti-Pg antibodies, verified the interactions between these proteins and N-cadherin. Taken together, exogenous expression of Pg did not seem to alter the interactions significantly between the catenins and N-cadherin. DISCUSSION Pg is a member of the Arm family of proteins, which also includes b-catenin and the product of the Drosophila segment polarity gene Armadillo [Peifer and Wieschaus, 1990; Peifer et al., 1992, 1994].These proteins are involved in pathways regulating cell–cell adhesion, fate determination, and cell proliferation. The adhesive function of these proteins is mediated by Plakoglobin Induces Fibroblast to Epidermoid Transition interactions with cadherins. Whereas b-catenin interacts only with the classic cadherins (e.g., E-, N-, P-, M-), Pg associates with both classic and desmosomal cadherins [Knudsen and Wheelock, 1992; Cowin and Burke, 1995; Marrs and Nelson, 1996; Gumbiner, 1996]. The necessity of Pg–desmosomal cadherin interactions for the assembly of desmosomes is well documented and supported by the presence of the insoluble (cytoskeleton-associated) Pg– desmosomal cadherin complexes [Mathur et al., 1994; Troyanovsky et al., 1994; Kowalczyk et al., 1994; Pasdar et al., 1995a; Bornslaeger et al., 1996]. However, the precise role of Pg in complexes involving the classic cadherins is unclear. Two lines of evidence suggest that this Pg–cadherin complex may have regulatory functions: Pg-containing classic cadherin complexes are primarily present in soluble forms, indicating the lack of strong interactions with the cytoskeleton; and Pg and b-catenin interact with the classic cadherins in a mutually exclusive manner [Hinck et al., 1994]. The latter finding also suggests that b-catenin and Pg binding sequences are identical or overlapping. Arm is a component of the Drosophila Wg (wingless) signaling pathway, which regulates developmental patterning. Pg and b-catenin are components of the Wnt (mammalian homologue of Wg) pathway [for a recent review, see Peifer, 1996]. Within this pathway, Pg and b-catenin interact with the tumor suppressor protein APC and transcription factors Lef-1/Tcf-1 [Behrens et al., 1996; Huber et al., 1996b]. Pg and b-catenin interact with cadherins and signaling proteins in a mutually exclusive manner, thus regulating the balance between cell adhesion and cell proliferation. Several recent studies have characterized the regulation of b-catenin function through its interactions with APC and transcription factors [Peifer, 1996; Korinek et al., 1997; Rubinfeld et al., 1997; Morin et al., 1997], but the role of Pg in regulating cell growth and proliferation is not as clear. SCC9 cells are carcinoma cells with a transformed phenotype that lack Pg but express b-catenin and can therefore be a useful model to study the adhesive and signaling functions of Pg. These cells also lack E-cadherin but express N-cadherin, which is the primary cadherin of nonepithelial cells. We previously showed that the exogenous expression of L-CAM or E-cadherin in the absence of Pg could induce epidermoid phenotype in SCC9 cells; this phenotypic transition coincided with decreased synthesis and stability of the endogenous N-cadherin and increased synthesis and stability of both a- and bcatenins [Li et al., in press]. These results were consistent with the necessity of E-cadherin expression for the epidermoid phenotype. This study shows that, in the absence of E-cadherin, Pg expression enabled SCC9 cells to form desmosomes and induced the same phenotypic transition as that 97 induced by E-cadherin expression. Interestingly, unlike E-cadherin, Pg expression coincided with increased synthesis and stability of N-cadherin and decreased levels and stability of b-catenin. The Pg-induced epidermoid phenotype was associated with increased adhesiveness and decreased growth rate at confluency. In fact, the initial adhesion ability was significantly more (&mt;50%) in Pg-expressing cells than those expressing L-CAM or E-cadherin [Li et. al., in press] (Fig. 2C). Unlike L-CAMexpressing SCC9 cells, Pg-expressing cells are capable of forming desmosomes and therefore, the initial cell–cell adhesion may be more in these cells due to the presence of both desmosomes and adherens junctions. E-Cadherin expression and Pg-E–cadherin interaction have been known to be necessary for, and to precede, the formation of desmosomes [Gumbiner et al., 1988; Lewis et al., 1994, 1997; Jensen et al., 1997]. Our results suggest that E-cadherin can be replaced by N-cadherin and that N-cadherin–Pg complex is, presumably, as effective in initiating desmosome formation as Pg-E–cadherin complex. A similar fibroblast-to-epidermoid transition concurrent with desmosome assembly and upregulation of N-cadherin expression has been reported for HeLa cells transfected with cDNAs encoding P0 cell adhesion protein [Doyle et al., 1995]. The induction of epidermoid phenotype and decreased proliferation rate of SCC9P cells are consistent with the suggested tumor suppressor activity of Pg [Simcha et al., 1996]. The exogenous expression of Pg (with or without cadherin) in an SV40-transformed 3T3 fibroblast cell line lacking both Pg and cadherins led to suppression of tumorgenicity. When expressed alone, Pg did not modify the morphology of the transfected cells but decreased their tumorgenicity. By contrast, transfection with N-cadherin induced a fibroblast to epidermoid transition in the morphology but had no effect on their tumorgenicity. When expressed together with N-cadherin, Pg was much more effective in decreasing tumorigenicity. Simcha et al.  suggested two possible mechanisms for the effects of Pg on the ability of cells to form tumors. Either the Pg interacts with components of signal pathways that influence cell proliferation, morphology, and other cellular processes; or Pg enhances cellular adhesion, thereby inhibiting tumor-forming ability. We have not tested the tumorigenicity of SCC9 cells after expression of Pg, but the phenotypic transformation from fibroblast to epidermoid and decreased growth rate are consistent with the increased adhesivity and possible tumor suppressor activity of Pg. Furthermore, the upregulation of N-cadherin and its increased stability are consistent with the possibility that the tumor suppressor activity of Pg may be mediated by its involvement in adhesion complexes (see below). 98 Parker et al. The expression of N-cadherin by SCC9 cells was an unexpected finding, as this is a squamous cell carcinoma derived line and epidermoid cells generally express Eand P-cadherin [Nicholson et al., 1991; Johnson et al., 1993]. A recent study by Islam et al.  also reported N-cadherin expression in two other transformed squamous carcinoma cell lines expressing low levels of E- and P-cadherins. The authors suggested that the inappropriate N-cadherin expression contributed directly to the invasive phenotype because transfection of these carcinoma cells with N-cadherin anti-sense RNAs downregulated N-cadherin and induced a fibroblast–epidermoid phenotypic transition, coincident with the upregulation of Eand P-cadherin expression. We have not detected any E-cadherin in SCC9 cells, and the Pg-induced epidermoid morphology coincided with increased amounts of N-cadherin. Our data suggest that, at least in SCC9 cells, N-cadherin expression per se is not responsible for the transformed phenotype. In addition our results suggested that Pg could form an apparently functional complex with N-cadherin to induce desmosome assembly and epidermoid morphology. These observations underscore the pivotal role of Pg in regulating cell adhesive properties and epithelial morphology. What are the mechanisms by which Pg induces epidermoid phenotype? We suggest that the transformed phenotype of SCC9 cells could result from either decreased adhesive properties due to the absence of Pg, or excess b-catenin, or both. Several processes could result in excess b-catenin. For example, inefficient interactions between N-cadherin and b-catenin, due to alterations in b-catenin binding domain of N-cadherin, deficiency in APC, or abnormalities in APC–b–catenin interactions. It is well documented that, if not sequestered by APC, excess b-catenin could interact with Lef-1 transcription factors and translocate into the nucleus. Once in the nucleus, Lef-1–b-catenin complex can activate the target genes involved in cell proliferation and mesenchymal phenotype [Korinek et al., 1997; Huber et al., 1996b].This complex also has been shown to bind to the E-cadherin promoter in vitro, resulting in decreased E-cadherin expression [Huber et al., 1996a,b]. It has been suggested that Pg and b-catenin may have overlapping binding domains in classic cadherins [Hinck et al., 1994]. One could imagine that a small alteration in the nonoverlapping area of binding for one of the two proteins could result in inefficient binding of that molecule without affecting the other. Therefore, if the SCC9 N-cadherin is defective in the b-catenin binding region, this could affect the stability of the cadherin– catenin complex. This deficiency could lead to excess b-catenin and can be corrected by introducing either a ‘‘normal’’ cadherin molecule, to interact with the excess b-catenin or by expressing Pg to substitute for b-catenin and forming stable complexes. An exogenously expressed functional cadherin such as L-CAM or E-cadherin would remove the excess b-catenin by forming efficient and functional cadherin–catenin complexes. This effect is similar to the suppression of dorsal signaling activity of the soluble b-catenin by overexpression of cadherins in Xenopus embryo [Fagotto et al., 1996; Miller and Moon, 1997]. Similar results would be expected by the exogenous expression of Pg. In that case, Pg could potentially function in two capacities. It could replace b-catenin and form stable N-cadherin–Pg complexes that are necessary for, and that initiate, desmosome formation, leading to increased adhesiveness and epidermoid morphology. Alternatively or in addition, the exogenously expressed Pg could directly decrease b-catenin levels, consistent with decreased synthesis and faster degradation of b-catenin in Pg-expressing SCC9P cells (Fig.5) and the recent study by Salomon et al.  suggesting co-regulation of Pg and b-catenin. Finally, Pg could bind and sequester the transcription factors (Lef-1) in the cytoplasm inhibiting its interaction with b-catenin and the subsequent translocation of the complex into the nucleus, as has been suggested previously [Merriam et al., 1997]. In conclusion, this study shows that, in the absence of E-cadherin, Pg can induce desmosome formation and a transition from transformed to epidermoid morphology by interacting with the available cadherin, N-cadherin. This observation would suggest that the regulatory role of the Pg–classic cadherins complex may be specified by Pg. Furthermore, the data suggest that Pg may also function as one of a regulators of b-catenin levels in cells. ACKNOWLEDGMENTS We are grateful to Drs. A. Klein-Santoz, John Hemperly, and Lionel Larue for their gifts of cells and cDNAs, to Honey Chan for her valuable assistance in electron microscopy and to Drs. W.J. Gallin and E.K. Shibuya for their critical review of the manuscript . This research is supported by a grant from Medical Research Council of Canada (to M.P.). We also acknowledge financial assistance from Leo Laboratories Canada Ltd., as part of the Dermatology Foundation grant. The University of Alberta Faculty of Medicine Confocal Laser Scanning Microscopy Facility is supported in part by funds from the Medical Research Council and the Alberta Heritage Foundation for Medical Research (AHFMR). M. Pasdar is an AHFMR scholar. REFERENCES Aberle, H., Schwartz, H., and Kemler R. (1996): Cadherin–catenin complex: Protein interactions and their implications for cadherin function. J. Cell. Biochem. 61:514–523. Plakoglobin Induces Fibroblast to Epidermoid Transition Aberle, H., Butz, S., Stappert, J., Weissig, H., Kemler, R., and Hoschuetzky, H. (1994): Assembly of the cadherin–catenin complex in vitro with recombinant proteins. J. Cell. Sci. 107:3655–3663. Behrens, J., von Kries, JP., Kuhl, M., Bruhn, L., Wedlich, D., Grosschedl, R., and Birchmeier, W. (1996): Functional interaction of beta-catenin with the transcription factor LEF-1. Nature 382:638–642. Bierkamp, C., McLaughlin, K.J., Schwarz, H., Huber, O., and Kemler, R. (1996): Embryonic heart and skin defects in mice lacking plakoglobin. Dev. Biol. 180:780–785. Bornslaeger, E., Corcoran, C. M., Steppenbeck, T.S., and Green, K.J. (1996): Breaking the connection: Disruption of the desmosomal plaque protein desmoplakin from cell–cell interfaces disrupts anchorage of intermediate filament bundles and alters intercellular junction assembly. J. Cell Biol. 134:985–1001. Bradley, R.S., Cowin, P., and Brown, A.M.C. (1993): Expression of Wnt-1 in PC12 cells results in modulation of plakoglobin and E-cadherin and increased cellular adhesion. J. Cell Biol. 123: 1875–1865. Burrus, L.W. (1994): Wnt-1 as a short-range signaling molecule. BioEssays 16:155–157. Church, G.M., and Gilbert, W. (1984): Genomic sequencing. Proc. Natl. Acad. Sci. U.S.A. 81:1991–1995. Cowin, P. (1994): Unraveling the cytoplasmic interactions of the cadherin superfamily. Proc. Natl. Acad. Sci. U.S.A. 91:10759– 10761. Cowin, P., and Burke, B. (1995): Cytoskeleton–membrane interactions. Curr. Opin. Cell Biol. 8:56–65. Cowin, P., Kapprell, H.P., Franke, W.W., Tamkun, J., and Hynes, R.O. (1986): Plakoglobin: A protein common to different kinds of intercellular adhering junctions. Cell 46:1063–1073. Doyle, J.P., Stempak, J.G., Cowin, P., Colman, D.R., and D’Urso, D. (1995): Protein zero, a nervous system adhesion molecule, triggers epithelial reversion in host carcinoma cells. J Cell Biol. 131:465–482. Fagotto, F., Funayama, N., Gluck, U., and Gumbiner, B.M. (1996). Binding to cadherins antagonizes the signaling activity of b-catenin during axis formation in Xenopus. J. Cell. Biol. 132:1105–1114. Franke, W.W., Goldschmidt, M.D., Zimbelmann, R., Mueller, H.M., Schiller, D.L., and Cowin, P. (1989): Molecular cloning and amino acid sequences of human plakoglobin, the common junctional plaque. Proc. Natl. Acad. Sci. U.S.A. 86:4027–4031. Garrod, D.R., Chidgey, M., and North, A. (1996): Desmosomes: Differentiation, development, dynamics and disease. Curr. Opin. Cell Biol. 8:670–678. Geiger, B., and Ayalon, O. (1992): Cadherins. Annu. Rev. Cell Biol. 8:307–332. Gumbiner, B.M., Stevenson, B.R., and Grimaldi, A. (1988): Role of cell adhesion molecule uvomorulin in formation and maintenance of the epithelial junctional complex. J. Cell Biol. 107: 1575–1587. Gumbiner,B. (1996): Cell adhesion: The molecular basis of tissue architecture and morphogenesis. Cell 84:345–357. Hamaguchi, M., Matsuyoshi, N., Ohnishi, Y., Gotoh, B., Takeichi, M., and Nagai, Y. (1993): p60v-src causes tyrosine phosphorylation and inactivation of the N-cadherin cell adhesion system. EMBO J. 12:307–314. Hertig, C.M., Butz, S., Koch, S., Eppenberger-Eberhardt, M., Kemler, R., and Eppenberger, H.M. (1996): N-Cadherin in adult rat cardiomyocytes in culture. II. Spatio-temporal appearance of proteins involved in cell–cell contact and communication. 99 Formation of two distinct N-cadherin/catenin complexes. J. Cell Sci. 109:11–20. Hinck, L., Nathke, I.S., Papkoff, J., and Nelson, W.J. (1994): Dynamics of cadherin/catenin complex formation: Novel protein interactions and pathways of complex assembly. J. Cell Biol. 125:1327– 1340. Hinck, L., Nelson, W.J., and Papkoff, J. (1994): Wnt-1 modulates cell–cell adhesion in mammalian cells by stabilizing b-catenin binding to the cell adhesion protein cadherin. J. Cell Biol. 124:729–742. Huber, O., Bierkamp, C., and Kemler, R. (1996a): Cadherins and catenins in development. Curr. Opin. Cell Biol. 8:685–691. Huber, O. Korn, R. McLaughlin, J. Ohsugi, M. Herrmann, B.G., and Kemler, R. (1996b): Nuclear localization of beta-catenin by interaction with transcription factor LEF-1. Mech. Dev. 59:3–10. Islam, S., Carey, T.E., Wolf, G.T., Wheelock, M.J., and Johnson, K.R. (1996): Expression of N-cadherin by human squamous carcinoma cells induces a scattered fibroblastic phenotype with disrupted cell–cell adhesion. J. Cell Biol. 135:1643–1654. Jensen, P.J., Telegan, B., Lavker, R.M., and Wheelock, M.J. (1997): E-cadherin and P-cadherin have partially redundant roles in human epidermal stratification. Cell Tissue Res. 288:307–316. Johnson, K.R., Lewis, J.E., Li, D., Wahl, J., Soler, A.P., Knudsen, K.A., and Wheelock, M.J. (1993): P- and E-cadherin are in separate complexes in cells expressing both cadherins. Exp. Cell Res. 207:252–260. Jou, T.S., Stewart, R.S., Stappert, J., Nelson, W.J., and Marrs, J.A. (1995): Genetic and biochemical dissection of protein linkages in the cadherin–catenin complex. Proc. Natl. Acad. Sci. U.S.A. 92:5067–5071. Kapprell, H.-P, Cowin, P., and Franke, W.W. (1987): Biochemical characterization of the soluble form of the junctional protein, plakoglobin, from different cell types. Eur. J. Biochem. 166:505– 517. Karnovsky, A., and Klymkowsky, M.W. (1994): Anterior axis duplication in Xenopus induced by the over-expression of the cadherinbinding protein plakoglobin. Proc. Natl. Acad. Sci. U.S.A. 92:4522–4526. Kemler, R. (1993): From cadherins to catenins: Cytoplasmic protein interactions and regulation of cell adhesion. Trends Genet. 9:317–321. Knudsen, K.A., Soler, A. P., Johnson, K.R., and Wheelock, M.J. (1995): Interaction of a-actinin with the cadherin/catenin cell– cell adhesion complex via a-catenin.J. Cell.Biol. 130:67–77. Knudsen, K.A., and Wheelock, M.J. (1992): Plakoglobin, or an 83-kD homologue distinct from b-catenin, interacts with E-cadherin and N-cadherin. J. Cell Biol. 118:671–679. Korinek, V., Barker, N., Morin, P.J., van Wichen, D., de Weger R., Kinzler, KW., and Vogelstein, B. Clevers, H. (1997): Constitutive transcriptional activation by a beta-catenin–Tcf complex in APC2/2 colon carcinoma. Science 275:1784–1787. Kowalczyk, A. P., Bornslaeger E.A., Borgwardt, J.E., Palka, H.L., Dhaliwal, A.S., Corcoran, C.M., Denning, M.F., and Green, K.J. (1997): The amino terminal domain of desmoplakin binds to plakoglobin and clusters desmosomal cadherin–plakoglobin complexes. J. Cell Biol. 139:773–784. Kowalczyk, A.P., Palka, H., Luu, H.H., Nilles, L.A., Anderson, J.E., Wheelock, M.J., and Green, K.J. (1994): Posttranslational regulation of plakoglobin expression; Influence of the desmosomal cadherins on plakoglobin metabolic stability. J. Biol. Chem. 269:31214–31223. Lewis, J.E., Jensen, P.J., and Wheelock, M.J. (1994): Cadherin function is required for human keratinocytes to assemble desmosomes and stratify in response to calcium. J. Invest. Dermatol. 102:870–877. 100 Parker et al. Lewis, J.E., Wahl, J.K. III, Sass, K.M., Jensen, P.J., Johnson, K.R., and Wheelock, M.J. (1997): Cross-talk between adherens junctions and desmosomes depends on plakoglobin. J. Cell Biol. 136:919– 934. Li, Z., Gallin, W. J., Lauzon, G., and Pasdar, M. L-CAM expression induces fibroblast–epidermoid transition in squamous carcinoma cells and down regulates the endogenous N-cadherin. In press. Marrs, J.A., and Nelson, W.J. (1996): Cadherin cell adhesion molecules in differentiation and embryogenesis. Int. Rev. Cytol. 165:159– 205. Mathur, M., Goodwin, L., and Cowin, P. (1994): Interactions of the cytoplasmic domain of the desmosomal cadherin Dsg1 with plakoglobin. J. Biol. Chem. 269:14075–14080. Merriam, J.M., Rubenstein, A.B., and Klymkowsky, M.W. (1997): Cytoplasmically anchored plakoglobin induces a WNT-like phenotype in Xenopus. Dev. Biol. 185:67–81. Miller, J.R., and Moon, R.T. (1997): Analysis of the signaling activities of localization mutants of b-catenin during axis specification in Xenopus. J. Cell Biol. 139:229–243. Moon, R.T., Brown, J.D., and Torres, M. (1997): Wnts modulate cell fate and behaviour during vertebrate development. Trends Genet. 13:157–162. Morin, P. J., Sparks, A.B., Korinek, V., Barker, N., Clevers, H., Vogelstein, B., and Kinzler, KW. (1997): Activation of b-cateninTcf signaling in colon cancer by mutations in b-catenin or APC. Science 275:1787–1790. Nicholson, L.J., Pei, X.F., and Watt, F.M. (1991): Expression of E-cadherin, P-cadherin and involucrin by normal and neoplastic keratinocytes in culture. Carcinogenesis 12:1345–1349. Nusse, R., and Varmus, H. E. (1992): Wnt genes. Cell 69:1073–1087. Palka, H.L., and Green, K.J. (1997): Roles of plakoglobin end domains in desmosome assembly. J. Cell Sci. 110:2359–2371. Parkin, N.T., Kitajewski, J., and Varmus, H.E. (1993): Activity of Wnt-1 as a transmembrane protein. Genes Dev. 7:2181–2193. Pasdar, M. (1992): In: Biochemical approaches for analyzing de novo assembly of epithelial junctional components. In Stevenson, B.R., Gallin, W.J., and Paul, D.L. (eds.): ‘‘Cell–Cell Interactions, A Practical Approach.’’ pp. 203–226. IRL Press, Oxford. Pasdar, M., and Nelson, W.J. (1988): Kinetics of desmosome assembly in Madin-Darby canine kidney epithelial cells: Temporal and spatial regulation of desmoplakin organization and stabilization upon cell–cell contact. I. Biochemical analysis. J. Cell Biol. 106:677–685. Pasdar, M., Li, Z., and Chlumecky, V. (1995a): Plakoglobin: Kinetics of synthesis, phosphorylation, stability, and interactions with desmoglein and E-cadherin. Cell Motil. Cytoskeleton 32: 258–272. Pasdar, M., Li, Z., and Chan, H. (1995b): Desmosome assembly and disassembly are regulated by reversible protein phosphorylation in cultured epithelial cells. Cell Motil. Cytoskeleton 30:108– 122. Peifer, M. (1996): Regulating cell proliferation: As easy as APC. Science 272:974–975. Peifer, M., and Wieschaus, E. (1990): The segment polarity gene armadillo encodes a functionally modular protein that is the Drosophila homologue of human plakoglobin. Cell. 63:1167– 1178. Peifer, M., McCrea, P.D., Green, K.J., Wieschaus, E., and Gumbiner, B.M. (1992): The vertebrate adhesive junction proteins betacatenin, and plakoglobin and the Drosophila segment polarity gene armadillo form a multigene family with similar properties. J. Cell Biol. 118:681–691. Peifer, M., Berg, S., and Reynolds, A.B. (1994): A repeating amino acid motif shared by proteins with diverse cellular roles. Cell 76:789–791. Perrimon, N. (1994): The genetic basis of patterned baldness in Drosophila. Cell 76:781–784. Piepenhagen, P.A., and Nelson, W.J. (1993): Defining E-cadherin associated protein complexes in epithelial cells: Plakoglobin, b-catenin and g-catenin are distinct components. J. Cell Sci. 104:751–762. Rimm, D.L., Koslov, E.R., Kebriaei, P., Cianci, C.D., and Morrow, J.S. (1995): a1-(E) catenin is an actin-binding and -bundling protein mediating the attachment of F-actin to the membrane adhesion complex. Proc. Natl. Acad. Sci. U.S.A. 92:8813–8817. Rubinfeld, B., Robbins, P., E.-Gamil, M., Albert, I., Porfiri, E., and Polakis, P. (1997): Stabilization of b-catenin by genetic defects in melanoma cell lines. Science 275:1790–1793. Ruiz, P., Brinkmann, V., Ledermann, B., Behrend, M., Grund, C., Thalhammer, C., Vogel, F., Birchmeier, C., Gunthert, U., Franke, W.W., and Birchmeier, W. (1996): Targeted mutation of plakoglobin in mice reveals essential functions of desmosomes in the embryonic heart. J. Cell Biol. 135:215–225. Sacco, P.A., McGranahan, T.M., Wheelock, M.J., and Johnson, K.R. (1995): Identification of plakoglobin domains required for association with N-cadherin and a-catenin. J. Biol. Chem. 270:20201–20206. Salomon, D., Sacco, P.A., Roy, S.G., Simcha, I, Johnson, K.R., Wheelock, M.J., and Ben-Ze’ev, A. (1997): Regulation of b-catenin levels and localization by overexpression of plakoglobin and inhibition of the proteasome–ubiquitin system. J. Cell Biol. 139:1325–1336. Schmidt, A., Heid, H.W., Schafer, S., Nuber, U.A., Zimbelman, R., and Franke, W.W. (1994): Desmosomes and cytoskeletal architecture in epithelial differentiation: Cell type-specific plaque components and intermediate filament anchorage. Eur. J. Cell Biol. 65:229–245. Simcha, I., Geiger, B., Yehuda-Levenberg, S., Salomon, D., and Ben-Ze’ev, A. (1996): Suppression of tumorigenecity by plakoglobin: An augmenting effect of N-cadherin. J. Cell Biol. 133:199–209. Towbin, H., Staehlin, T., and Gordon, G. (1979): Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedures and some applications. Proc. Natl. Acad. Sci. U.S.A. 76:4350–4354. Troyanovsky, S.M., Troyanovsky, R.B., Eshkind, L.G., Leube, R.E., and Franke, W.W. (1994): Identification of amino acid sequence required for plakoglobin binding and plaque formation. Proc. Natl. Acad. Sci. U.S.A. 91:10790–0794.