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

The functional interaction of cells in the formation of tissues requires the establishment and maintenance of cell-cell contact by the junctional complex. However, little is known biochemically about the mechanism(s) that regulates junctional complex assembly. To address this problem, we have initiated a study of the regulation of assembly of one component of the junctional complex, the desmosome, during induction of cell-cell contact in cultures of Madin-Darby canine kidney epithelial cells. Here we have analyzed two major protein components of the desmosomal plaque, desmoplakins I (Mr of 250,000) and II (Mr of 215,000). Analysis of protein levels of desmoplakins I and II by immunoprecipitation with an antiserum that reacts specifically with an epitope common to both proteins revealed that desmoplakins I and II are synthesized and accumulate at steady state in a ratio of 3-4:1 (in the absence or presence of cell-cell contact). The kinetics of desmoplakins I and II stabilization and assembly were analyzed after partitioning of newly synthesized proteins into a soluble and insoluble protein fraction by extraction of whole cells in a Triton X-100 high salt buffer. In the absence of cell-cell contact, both the soluble and insoluble pools of desmoplakins I and II are unstable and are degraded rapidly (t1/2 approximately 8 h). Upon induction of cell-cell contact, the capacity of the insoluble pool increases approximately three-fold as a proportion of the soluble pool of newly synthesized desmoplakins I and II is titrated into the insoluble pool. The insoluble pool becomes relatively stable (t1/2 greater than 72 h), whereas proteins remaining in the soluble pool (approximately 25-40% of the total) are degraded rapidly (t1/2 approximately 8 h). Furthermore, we show that desmoplakins I and II can be recruited from this unstable soluble pool of protein to the stable insoluble pool upon induction of cell-cell contact 4 h after synthesis; significantly, the stabilization of this population of newly synthesized desmoplakins I and II is blocked by the addition of cycloheximide at the time of cell-cell contact, indicating that the coordinate synthesis of another protein(s) is required for protein stabilization.     

Regulation of desmosome assembly in epithelial cells: kinetics of synthesis, transport, and stabilization of desmoglein I, a major protein of the membrane core domain

Desmosomes are composed of two morphologically and biochemically distinct domains, a cytoplasmic plaque and membrane core. We have initiated a study of the synthesis and assembly of these domains in Madin-Darby canine kidney (MDCK) epithelial cells to understand the mechanisms involved in the formation of desmosomes. Previously, we reported the kinetics of assembly of two components of the cytoplasmic plaque domain, Desmoplakin I/II (Pasdar, M., and W. J. Nelson. 1988. J. Cell Biol. 106:677-685 and 106:687-699. We have now extended this analysis to include a major glycoprotein component of the membrane core domain, Desmoglein I (DGI; Mr = 150,000). Using metabolic labeling and inhibitors of glycoprotein processing and intracellular transport, we show that DGI biosynthesis is a sequential process with defined stages. In the absence of cell-cell contact, DGI enters a Triton X-100 soluble pool and is core glycosylated. The soluble DGI is then transported to the Golgi complex where it is first complex glycosylated and then titrated into an insoluble pool. The insoluble pool of DGI is subsequently transported to the plasma membrane and is degraded rapidly (t1/2 less than 4 h). Although this biosynthetic pathway occurs independently of cell-cell contact, induction of cell-cell contact results in dramatic increases in the efficiency and rate of titration of DGI from the soluble to the insoluble pool, and its transport to the plasma membrane where DGI becomes metabolically stable (t1/2 greater than 24 h). Taken together with our previous study of DPI/II, we conclude that newly synthesized components of the cytoplasmic plaque and membrane core domains are processed and assembled with different kinetics indicating that, at least initially, each domain is assembled separately in the cell. However, upon induction of cell-cell contact there is a rapid titration of both components into an insoluble and metabolically stable pool at the plasma membrane that is concurrent with desmosome assembly.     

Involvement of protein kinase C in translocation of desmoplakins from cytosol to plasma membrane during desmosome formation in human squamous cell carcinoma cells grown in low to normal calcium concentration

The intracellular signal transduction mechanism leading to desmosome formation in low-calcium-grown keratinocytes after addition of calcium to the medium was studied by immunofluorescence using antibodies to desmoplakins I and II (cytoplasmic desmosomal proteins) and by electron microscopy before and after addition of calcium; protein kinase C (PKC) activators 12-O-tetradecanoylphorbol-13-acetate (TPA), phorbol-12,13-dibutyrate (PDBu), and 1,2-dioctanoylglycerol (DOG); calcium ionophore A23187; selective PKC inhibitors 1-(5-isoquinolinylsulfonyl)-2-methylpiperazine (H-7) and staurosporine; and a Ca2+/calmodulin-dependent kinase inhibitor, N-(6-aminohexyl)-5-chloro-1-naphthalenesulfonamide (W-7). In previous studies using a low-calcium-grown human epidermal squamous cell carcinoma, we have shown that an increase in extracellular Ca2+ caused a four-fold increase in PKC activity and addition of TPA (10 ng/ml) induced a transient increase in membrane-bound PKC activity in association with cell-cell contact formation. The present study showed that TPA (10 ng/ml). PDBu (10 ng/ml), and DOG (1 mg/ml) induced a rapid cell-cell contact and redistribution of desmoplakins from cytoplasm to the plasma membrane with desmosome formation within 60-120 min, which was similar, although less marked, to the effect of increased Ca2+. The TPA-induced desmosome formation was inhibited by selective PKC inhibitors, H-7 (20 microM) or staurosporine (100 nM). On the other hand, calcium ionophore A23187 induced only a temporary increase in the number of desmoplakin-containing fluorescent spots in the cytoplasm and a temporary cell-cell attachment without desmosome formation. The calcium-induced desmosome formation was partially inhibited by 20-100 microM H-7 or 100 nM staurosporine; however, it was not inhibited by W-7 at a concentration of 25 microM, at which this agent selectively inhibits calmodulin-dependent protein kinase. These results suggest that PKC activation plays an important role in desmoplakin translocation from the cytoplasm to the plasma membrane as one of the processes of calcium-induced desmosome formation.     

Regulation of desmosome assembly in MDCK epithelial cells: coordination of membrane core and cytoplasmic plaque domain assembly at the plasma membrane

Desmosomes are major components of the intercellular junctional complex in epithelia. They consist of at least eight different cytoplasmic and integral membrane proteins that are organized into two biochemically and structurally distinct domains: the cytoplasmic plaque and membrane core. We showed previously that in MDCK epithelial cells major components of the cytoplasmic plaque (desmoplakin I and II; DPI/II) and membrane core domains (desmoglein I; DGI) initially enter a pool of proteins that is soluble in buffers containing Triton X-100, and then titrate into an insoluble pool before their arrival at the plasma membrane (Pasdar, M., and W. J. Nelson. 1988. J. Cell Biol. 106:677-685; Pasdar. M., and W. J. Nelson. 1989. J. Cell Biol. 109:163-177). We have now examined whether either the soluble or insoluble pool of these proteins represents an intracellular site for assembly and interactions between the domains before their assembly into desmosomes at the plasma membrane. Interactions between the Triton X-100-soluble pools of DPI/II and DGI were analyzed by sedimentation of extracted proteins in sucrose gradients. Results show distinct differences in the sedimentation profiles of these proteins, suggesting that they are not associated in the Triton X-100-soluble pool of proteins; this was also supported by the observation that DGI and DPI/II could not be coimmunoprecipitated in a complex with each other from sucrose gradient fractions. Immunofluorescence analysis of the insoluble pools of DPI/II and DGI, in cells in which desmosome assembly had been synchronized, showed distinct differences in the spatial distributions of these proteins. Furthermore, DPI/II and DGI were found to be associated with different elements of cytoskeleton; DPI/II were located along cytokeratin intermediate filaments, whereas DGI appeared to be associated with microtubules. The regulatory role of cytoskeletal elements in the intracellular organization and assembly of the cytoplasmic plaque and membrane core domains, and their integration into desmosomes on the plasma membrane is discussed.     

Intermediate filaments and the initiation of desmosome assembly

The desmosome junction is an important component in the cohesion of epithelial cells, especially epidermal keratinocytes. To gain insight into the structure and function of desmosomes, their morphogenesis has been studied in a primary mouse epidermal (PME) cell culture system. When these cells are grown in approximately 0.1 mM Ca2+, they contain no desmosomes. They are induced to form desmosomes when the Ca2+ level in the culture medium is raised to approximately 1.2 mM Ca2+. PME cells in medium containing low levels of Ca2+, and then processed for indirect immunofluorescence using antibodies directed against desmoplakins (desmosomal plaque proteins), display a pattern of discrete fluorescent spots concentrated mainly in the perinuclear region. Double label immunofluorescence using keratin and desmoplakin antibodies reveals that the desmoplakin-containing spots and the cytoplasmic network of tonofibrils (bundles of intermediate filaments [IFB]) are in the same juxtanuclear region. Within 1 h after the switch to higher levels of Ca2+, the spots move toward the cell surface, primarily to areas of cell-cell contact and not to free cell surfaces. This reorganization occurs at the same time that tonofibrils also move toward cell surfaces in contact with neighboring cells. Once the desmoplakin spots have reached the cell surface, they appear to aggregate to form desmosomes. These immunofluorescence observations have been confirmed by immunogold ultrastructural localization. Preliminary biochemical and immunological studies indicate that desmoplakin appears in whole cell protein extracts and in Triton high salt insoluble residues (i.e., cytoskeletal preparations consisting primarily of IFB) prepared from PME cells maintained in medium containing both low and normal Ca2+ levels. These findings show that certain desmosome components are preformed in the cytoplasm of PME cells. These components undergo a dramatic reorganization, which parallels the changes in IFB redistribution, upon induction of desmosome formation. The reorganization depends upon both the extracellular Ca2+ level and the establishment of cell-to-cell contacts. Furthermore, the data suggests that desmosomes do not act as organizing centers for the elaboration of IFB. Indeed, we postulate that the movement of IFB and preformed desmosomal components to the cell surface is an important initiating event in desmosome morphogenesis.     

Desmoplakin expression and distribution in cultured rat bladder epithelial cells of varying tumorigenic potential

The expression and distribution of the desmosomal plaque proteins, desmoplakins (DPs) I and II, were studied in nontumorigenic (RBE-8) and a series of tumorigenic (AY34, R-4909, SS-24B, RBTCC-8, and 804G) rat bladder epithelial cell lines. These cell lines ranged from slow-growing papillary transitional cells (AY34) to rapidly metastatic carcinoma cells (RBTCC-8). DPs I and II were shown by immunoblotting and Northern analysis to be present in nontumorigenic RBE-8 cells as well as in all of the tumorigenic cell lines, albeit in differing amounts. Immunofluorescence microscopy revealed striking differences in DP distribution, corresponding in general with increases in tumorigenic potential. Whereas DPs of normal RBE-8 cells and less tumorigenic AY34 cells were localized predominantly at cell interfaces, the more tumorigenic lines exhibited a high proportion of DP in the form of cytoplasmic dots, a distribution reminiscent of that seen in epithelial cells maintained in low levels of extracellular calcium. In 804G cells, which represented the most extreme example of this phenomenon, the majority of DPs were organized as cytoplasmic dots. Electron microscopy revealed intermediate filament (IF)-associated spots in the cytoplasm as well as an elaborate array of IF-associated plaques at the cell-substratum interface. The IF-associated spots in the cytoplasm reacted with anti-DP antibody in immunogold labeling experiments while those at the cell-substratum did not react. In more dense cultures of 804G cells, certain cells stratified and expressed increased amounts of DP followed by the induction of new keratins including those of the skin type. Decreasing extracellular calcium resulted in a rearrangement of DP in each cell line; staining at cell-cell interfaces disappeared and was replaced with a pattern of cytoplasmic dots. These results demonstrate a possible relationship between desmosome assembly and/or maintenance and tumorigenic potential.     

Desmosome assembly in MDCK epithelial cells does not require the presence of functional microtubules.

Desmosomes, complex multisubunit structures that assemble at sites of cell-cell contact, are important components of the epithelial junctional complex. Desmosome assembly requires the coordinated interaction at the plasma membrane of at least 8 cytoplasmic and integral membrane proteins organized into two structurally and functionally distinct domains, the cytoplasmic plaque and membrane core. Previous studies (Pasdar et al., J. Cell Biol., 113:645-655) provided evidence that cytokeratin filaments and microtubules may regulate transfer and assembly of cytoplasmic plaque and membrane core proteins, respectively. To determine directly the role of microtubules in these processes, Madin-Darby canine kidney (MDCK) cells were treated with nocodazole or colchicine to disrupt the microtubular network. Biochemical analysis of the different components of the cytoplasmic plaque and membrane core domains revealed little or no effect of nocodazole or colchicine on the kinetics of synthesis, post-translational modifications, transfer of proteins to the plasma membrane or their metabolic stability in the presence or absence of cell-cell contact. Likewise, immunofluorescence analysis of desmosome formation demonstrated an apparently normal desmosome assembly in the presence of nocodazole or colchicine upon induction of cell-cell contact. These results indicate that an intact microtubular network is not necessary for the processing or transport of the desmosomal membrane core glycoproteins to the plasma membrane in the absence or presence of cell-cell contact. Furthermore, the integration of the cytoplasmic plaque and membrane core domains induced by cell-cell contact at the plasma membranes of adjacent cells does not require the presence of functional microtubules.     

Junctional uvomorulin/E-cadherin and phosphotyrosine-modified protein content are correlated with paracellular permeability in Madin-Darby canine kidney (MDCK) epithelia

Strains I and II of Madin-Darby canine kidney (MDCK) cells, which differ markedly in transepithelial resistance (R _t ) and paracellular permeability, have been used to investigate whether differences in the cellular content of uvomorulin/E-cadherin and phosphotyrosine may be correlated with junctional properties. Using immunocytochemistry, the strain I tight epithelia showed significantly stronger uvomorulin staining at regions of cell-cell contact compared with strain II leaky MDCK epithelia. In contrast, strain I MDCK cells showed a relatively faint phosphotyrosine staining, distributed evenly throughout the cytoplasm, while strain II MDCK cells displayed intense staining for phosphotyrosine residues in the junctional region and the lateral cell membrane with additional labelling of the cytoplasm. Exposure to vanadate in conjunction with H₂O₂ (which are potent inhibitors of protein tyrosine phosphatases) resulted in a dramatic increase in phosphotyrosine staining at the intercellular area and, concomitantly, induced changes in cell morphology, a significant decrease in R _t , increase in paracellular inulin permeability, and time-dependent disappearance of uvomorulin from the cell-cell contact sites. Moreover, the effects of vanadate/H₂O₂ treatment were more dramatic in strain II compared with strain I cells, consistent with greater generation of tyrosine-modified protein in strain II cells. An inverse relationship was demonstrated between membrane-associated uvomorulin/E-cadherin and cellular phosphotyrosine content, which varied between the two strains of MDCK cells and when phosphotyrosine was directly manipulated. These data support the hypothesis that regulation of paracellular permeability may result from specific tyrosine phosphorylation of protein components of the junctional complex.     

Modulation of fodrin (membrane skeleton) stability by cell-cell contact in Madin-Darby canine kidney epithelial cells

During growth of Madin-Darby canine kidney (MDCK) epithelial cells, there is a dramatic change in the stability, biophysical properties, and distribution of the membrane skeleton (fodrin) which coincides temporally and spatially with the development of the polarized distribution of the Na+, K+-ATPase, a marker protein of the basolateral domain of the plasma membrane. These changes occur maximally upon the formation of a continuous monolayer of cells, indicating that extensive cell-cell contact may play an important role in the organization of polarized MDCK cells (Nelson, W. J., and P. J. Veshnock, 1986, J. Cell Biol., 103:1751-1766). To directly analyze the role of cell-cell contact in these events, we have used an assay in which the organization of fodrin and membrane proteins is analyzed in confluent monolayers of MDCK cells in the absence or presence of cell-cell contact by adjusting the concentration Ca++ in the growth medium. Our results on the stability and solubility properties of fodrin reported here show directly that there is a positive correlation between cell-cell contact and increased stability and insolubility of fodrin. Furthermore, we show that fodrin can be recruited from an unstable pool of protein to a stable pool during induction of cell-cell contact; significantly, the stabilization of fodrin is not affected by the addition of cyclohexamide, indicating that proteins normally synthesized during the induction of cell-cell contact are not required. Together these results indicate that cell-cell contact may play an important role in the development of polarity in MDCK cells by initiating the formation of a stable, insoluble matrix of fodrin with preexisting (membrane) proteins at the cell periphery. This matrix may function subsequently to trap proteins targeted to the membrane, resulting in the maintenance of membrane domains.     

Compartmentalisation and localisation of the translation initiation factor (eIF) 4F complex in normally growing fibroblasts

Previous observations of association of mRNAs and ribosomes with subcellular structures highlight the importance of localised translation. However, little is known regarding associations between eukaryotic translation initiation factors and cellular structures within the cytoplasm of normally growing cells. We have used detergent-based cellular fractionation coupled with immunofluorescence microscopy to investigate the subcellular localisation in NIH3T3 fibroblasts of the initiation factors involved in recruitment of mRNA for translation, focussing on eIF4E, the mRNA cap-binding protein, the scaffold protein eIF4GI and poly(A) binding protein (PABP). We find that these proteins exist mainly in a soluble cytosolic pool, with only a subfraction tightly associated with cellular structures. However, this "associated" fraction was enriched in active "eIF4F" complexes (eIF4E.eIF4G.eIF4A.PABP). Immunofluorescence analysis reveals both a diffuse and a perinuclear distribution of eIF4G, with the perinuclear staining pattern similar to that of the endoplasmic reticulum. eIF4E also shows both a diffuse staining pattern and a tighter perinuclear stain, partly coincident with vimentin intermediate filaments. All three proteins localise to the lamellipodia of migrating cells in close proximity to ribosomes, microtubules, microfilaments and focal adhesions, with eIF4G and eIF4E at the periphery showing a similar staining pattern to the focal adhesion protein vinculin.     

Patterns of organelle distribution in mouse embryos during preimplantation development

We have evaluated the distribution of mitochondria and acidic organelles using, respectively, the specific vital fluorescent dyes rhodamine 123 and acridine orange during preimplantation embryonic development in the mouse. Under conditions used to visualize organelles in living embryos, staining with either dye was found to have no effect on either the rate or extent of in vitro development of five- to eight-cell embryos up to the blastocyst stage. Mitochondria were randomly distributed throughout the cytoplasm and located around nuclei in blastomeres of uncompacted embryos. During compaction, mitochondria initially reorganized to the blastomere cortex; however, these organelles were later confined to the perinuclear region in the trophectoderm (TE) of expanded blastocysts. Acidic organelles were randomly distributed in the cytoplasm of uncompacted embryos, but following compaction, they were concentrated in cortical and perinuclear locations. Moreover, in TE cells of expanded blastocysts, acidic organelles were found exclusively in a tight perinuclear pattern. Microtubules and microfilaments in TE cells were localized in fixed embryos stained with antitubulin antibodies and rhodamine phalloidin, respectively; these structures were found primarily in the cortical cytoplasm at areas of cell-cell contact and secondarily in a perinuclear location. Thus mitochondria and acidic organelles undergo stage-specific redistributions from a diffuse or cortical pattern at the eight-cell stage to a tight perinuclear localization in the TE. We conclude that the polarized distributions of some organelles and cytoskeletal proteins during compaction may not be reliable permanent markers of the mature TE.     

Decreased plakophilin-1 expression promotes increased motility in head and neck squamous cell carcinoma cells

Desmosomes are prominent cell-cell adhesive junctions found in a variety of epithelial tissues, including the oral epithelium. The transmembrane core of the desmosome is composed of the desmosomal cadherins that interact extracellularly to mediate cell-cell adhesion. The cytoplasmic domain of desmosomal cadherins interact with plaque proteins that in turn interact with the keratin intermediate filament cytoskeleton. Plakophilin 1 is a major desmosomal plaque component that functions to recruit intermediate filaments to sites of cell-cell contact via interactions with desmoplakin. Decreased assembly of desmosomes has been reported in several epithelial cancers. We examined plakophilin-1 expression in an esophageal squamous cell carcinoma tissue microarray and found that plakophilin-1 expression inversely correlates with tumor grade. In addition, we sought to investigate the effect of plakophilin-1 expression on desmosome assembly and cell motility in oral squamous cell carcinoma cell lines. Cell lines expressing altered levels of plakophilin-1 were generated and the ability of these cells to recruit desmoplakin to sites of cell-cell contact was examined. Our results show that decreased expression of plakophilin-1 results in decreased desmosome assembly and increased cell motility and invasion. These data lead us to propose that loss of plakophilin-1 expression during head and neck squamous cell carcinoma (HNSCC) progression may contribute to an invasive phenotype.     

The desmoplakin carboxyl terminus coaligns with and specifically disrupts intermediate filament networks when expressed in cultured cells

Specific interactions between desmoplakins I and II (DP I and II) and other desmosomal or cytoskeletal molecules have been difficult to determine in part because of the complexity and insolubility of the desmosome and its constituents. We have used a molecular genetic approach to investigate the role that DP I and II may play in the association of the desmosomal plaque with cytoplasmic intermediate filaments (IF). A series of mammalian expression vectors encoding specific predicted domains of DP I were transiently expressed in cultured cells that form (COS-7) and do not form (NIH-3T3) desmosomes. Sequence encoding a small antigenic peptide was added to the 3' end of each mutant DP cDNA to facilitate immunolocalization of mutant DP protein. Light and electron microscopical observations revealed that DP polypeptides including the 90-kD carboxy-terminal globular domain of DP I specifically colocalized with and ultimately resulted in the complete disruption of IF in both cell lines. This effect was specific for IF as microtubule and microfilament networks were unaltered. This effect was also specific for the carboxyl terminus of DP, as the expression of the 95-kD rod domain of DP I did not visibly alter IF networks. Immunogold localization of COS-7 cells transfected with constructs including the carboxyl terminus of DP demonstrated an accumulation of mutant protein in perinuclear aggregates within which IF subunits were sequestered. These results suggest a role for the DP carboxyl terminus in the attachment of IF to the desmosome in either a direct or indirect manner.     

Decreased Plakophilin-1 Expression Promotes Increased Motility in Head and Neck Squamous Cell Carcinoma Cells

Desmosomes are prominent cell-cell adhesive junctions found in a variety of epithelial tissues, including the oral epithelium. The transmembrane core of the desmosome is composed of the desmosomal cadherins that interact extracellularly to mediate cell-cell adhesion. The cytoplasmic domain of desmosomal cadherins interact with plaque proteins that in turn interact with the keratin intermediate filament cytoskeleton. Plakophilin 1 is a major desmosomal plaque component that functions to recruit intermediate filaments to sites of cell-cell contact via interactions with desmoplakin. Decreased assembly of desmosomes has been reported in several epithelial cancers. We examined plakophilin-1 expression in an esophageal squamous cell carcinoma tissue microarray and found that plakophilin-1 expression inversely correlates with tumor grade. In addition, we sought to investigate the effect of plakophilin-1 expression on desmosome assembly and cell motility in oral squamous cell carcinoma cell lines. Cell lines expressing altered levels of plakophilin-1 were generated and the ability of these cells to recruit desmoplakin to sites of cell-cell contact was examined. Our results show that decreased expression of plakophilin-1 results in decreased desmosome assembly and increased cell motility and invasion. These data lead us to propose that loss of plakophilin-1 expression during head and neck squamous cell carcinoma (HNSCC) progression may contribute to an invasive phenotype.     

Differential trafficking of transforming growth factor-beta receptors and ligand in polarized epithelial cells

Epithelial cells in vivo form tight cell-cell associations that spatially separate distinct apical and basolateral domains. These domains provide discrete cellular processes essential for proper tissue and organ development. Using confocal imaging and selective plasma membrane domain activation, the type I and type II transforming growth factor-beta (TGFbeta) receptors were found to be localized specifically at the basolateral surfaces of polarized Madin-Darby canine kidney (MDCK) cells. Receptors concentrated predominantly at the lateral sites of cell-cell contact, adjacent to the gap junctional complex. Cytoplasmic domain truncations for each receptor resulted in the loss of specific lateral domain targeting and dispersion to both the apical and basal domains. Whereas receptors concentrate basolaterally in regions of direct cell-cell contact in nonpolarized MDCK cell monolayers, receptor staining was absent from areas of noncell contact. In contrast to the defined basolateral polarity observed for the TGFbeta receptor complex, TGFbeta ligand secretion was found to be from the apical surfaces. Confocal imaging of MDCK cells with an antibody to TGFbeta1 confirmed a predominant apical localization, with a stark absence at the basal membrane. These findings indicate that cell adhesion regulates the localization of TGFbeta receptors in polarized epithelial cultures and that the response to TGFbeta is dependent upon the spatial distribution and secretion of TGFbeta receptors and ligand, respectively.     

Involvement of the membrane-cytoskeleton in development of epithelial cell polarity

The generation of cell surface polarity in transporting epithelial cells occurs in three distinct stages that involve cell-cell recognition and adhesion, cell surface remodelling to form biochemically and functionally distinct cell surface domains, and development of vectorial function. A widely used model system to study mechanisms involved in these stages is the Madin-Darby canine kidney (MDCK) cell line. Under appropriate growth conditions, MDCK cells develop in similar stages into polarized, multicellular epithelial structures. Analysis of membrane-cytoskeletal proteins ankyrin and fodrin during development of MDCK cell surface polarity shows that they gradually assemble into an insoluble protein complex on the basal-lateral membrane domain upon cell-cell adhesion, concomitantly with the redistribution of Na+,K(+)-ATPase, a marker protein of the basal-lateral membrane. Biochemical analysis shows that ankyrin, fodrin occur in a complex with Na+,K(+)-ATPase and the cell adhesion molecule uvomorulin in MDCK cells. A model is presented in which assembly of membrane-cytoskeletal complexes at sites of uvomorulin-induced cell-cell contact causes a remodelling of the cell surface distribution of specific membrane proteins which, in turn, contributes to the generation of epithelial cell surface polarity.     

Phosphorylated human keratinocyte ornithine decarboxylase is preferentially associated with insoluble cellular proteins

Ornithine decarboxylase (ODC), the first enzyme in polyamine biosynthesis, is highly regulated by many trophic stimuli, and changes in its levels and organization correlate with cytoskeletal changes in normal human epidermal keratinocytes (NHEK). NHEK ODC exhibits a filamentous perinuclear/nuclear localization that becomes more diffuse under conditions that alter actin architecture. We have thus asked whether ODC colocalizes with a component of the NHEK cytoskeleton. Confocal immunofluorescence showed that ODC distribution in NHEK was primarily perinuclear; upon disruption of the actin cytoskeleton with cytochalasin D, ODC distribution was diffuse. The ODC distribution in untreated NHEK overlapped with that of keratin in the perinuclear but not cytoplasmic area; after treatment with cytochalasin D, overlap between staining for ODC and for keratin was extensive. No significant overlap with actin and minimal overlap with tubulin filament systems were observed. Subcellular fractionation by sequential homogenizations and centrifugations of NHEK lysates or detergent and salt extractions of NHEK in situ revealed that ODC protein and activity were detectable in both soluble and insoluble fractions, with mechanical disruption causing additional solubilization of ODC activity (three- to sevenfold above controls). Fractionation and ODC immunoprecipitation from [(32)P]orthophosphate-labeled NHEK lysates showed that a phosphorylated form of ODC was present in the insoluble fractions. Taken together, these data suggest that two pools of ODC exist in NHEK. The first is the previously described soluble pool, and the second is enriched in phospho-ODC and associated with insoluble cellular material that by immunohistochemistry appears to be organized in conjunction with the keratin cytoskeleton.     

Identification of an epithelial protein related to the desmosome and intermediate filament network

Using a mAb, referred to as 08L, we have identified a protein, of M(r) approximately 140,000, associated with desmosomes of epithelial cells. The 08L antibody stained the intracellular side of lateral cell margins of monolayer epithelial cells but did not stain cell margins free of cell contact. Immunoelectron microscopy revealed that the 08L antigen was localized to the cytosolic surface of the desmosomal plaque near points of intermediate filament convergence with apparently little staining of the desmosomal plaque proper. Western blots revealed the 08L antigen to be a protein, of M(r) approximately 140,000, found in the Triton-X 100 insoluble pellet. High salt-containing buffers extracted the 08L antigen from the insoluble material. Examination of the assembly of 08L to the desmosome complex, in cells grown in low confluent culture or in calcium-switch assays, by double immunofluorescence with 08L and anti-desmoplakin antibody, revealed that 08L was recruited to morphologically identifiable desmosomes. 08L antigen may exist in a cytosolic pool prior to assembly to the cell surface. The solubility of 08L in low calcium and normal calcium conditions, however, was similar. 08L association to the desmosome was correlated with increased organization of the intermediate filament network. We suggest that the 08L antigen may be involved in the organization and stabilization of the desmosome-IF complexes of epithelia.     

Localization of the tight junction protein, ZO-1, is modulated by extracellular calcium and cell-cell contact in Madin-Darby canine kidney epithelial cells

Using the monoclonal antibody R26.4, we have previously identified a approximately 225-kD peripheral membrane protein, named ZO-1, that is uniquely associated with the tight junction (zonula occludens) in a variety of epithelia including the Madin-Darby canine kidney (MDCK) epithelial cell line (Stevenson, B. R., J. D. Siliciano, M. S. Mooseker, and D. A. Goodenough. 1986. J. Cell Biol. 103:755-766). In this study we have analyzed the effects of cell-cell contact and extracellular calcium on the localization and the solubility of ZO-1. In confluent monolayers under normal calcium conditions, ZO-1 immunoreactivity is found exclusively at the plasma membrane in the region of the junctional complex. If MDCK cells are maintained in spinner culture under low calcium conditions, ZO-1 is diffusely organized within the cytoplasm. After the plating of suspension cells at high cell density in medium with normal calcium concentrations, ZO-1 becomes localized to the plasma membrane at sites of cell-cell contact within 5 h in a process that is independent of de novo protein synthesis. However, if suspension cells are plated at high density in low calcium medium or if suspension cells are plated at low cell density in normal calcium growth medium, ZO-1 remains diffusely organized. ZO-1 localization also becomes diffuse in monolayers that have been established in normal calcium medium and then subsequently switched into low calcium medium. These results suggest that both extracellular calcium and cell-cell contact are necessary for normal localization of ZO-1 to the plasma membrane. An analysis of the solubility properties of ZO-1 from suspension cells and monolayers revealed that high salt, nonionic detergent, and a buffer containing chelators were somewhat more effective at solubilizing ZO-1 from suspension cells than from monolayers.     

The expression of desmosomal and corneodesmosomal antigens shows specific variations during the terminal differentiation of epidermis and hair follicle epithelia.

Using five monoclonal antibodies (MAb), we studied by indirect immunofluorescence the desmosomes and a junctional structure specific to cornified layers, the corneodesmosome, in normal and plantar epidermis and in the various sheaths of the anagen hair follicle. The monoclonal antibodies DP1&2.2-15, PG5.1, and DG3.10, specific for desmoplakins I/II, plakoglobin, and desmoglein I, respectively, were used to study the desmosome antigens, and G36-19 and G20-21 to study the corneodesmosome antigens. The distribution and sequence of expression of the five antigens allowed the nine epithelial differentiation pathways studied to be merged into four distinct families: non-plantar epidermis, characterized by the absence of desmosome and corneodesmosome antigens in the stratum corneum; the outer root sheath of the hair follicle, which behaves like the viable layers of the epidermis with regard to the desmosome antigens but does not express the corneodesmosome antigens; plantar epidermis and the three components of the inner root sheath in which the corneodesmosome antigens are present up to the desquamating layer; and the three components of the hair shaft, which are characterized by the absence of expression of both the desmosome and the corneodesmosome antigens in its mature portion.