The epithelial polarity program: machineries involved and their hijacking by cancer.

The Epithelial Polarity Program (EPP) adapts and integrates three ancient cellular machineries to construct an epithelial cell. The polarized trafficking machinery adapts the cytoskeleton and ancestral secretory and endocytic machineries to the task of sorting and delivering different plasma membrane (PM) proteins to apical and basolateral surface domains. The domain-identity machinery builds a tight junctional fence (TJ) between apical and basolateral PM domains and adapts ancient polarity proteins and polarity lipids on the cytoplasmic side of the PM, which have evolved to perform a diversity of polarity tasks across cells and species, to provide 'identity' to each epithelial PM domain. The 3D organization machinery utilizes adhesion molecules as positional sensors of other epithelial cells and the basement membrane and small GTPases as integrators of positional information with the activities of the domain-identity and polarized trafficking machineries. Cancer is a disease mainly of epithelial cells (90% of human cancers are carcinomas that derive from epithelial cells) that hijacks the EPP machineries, resulting in loss of epithelial polarity, which often correlates in extent with the aggressiveness of the tumor. Here, we review how the EPP integrates its three machineries and the strategies used by cancer to hijack them.

Polarity and developmental regulation of two PDZ proteins in the retinal pigment epithelium.

PURPOSE: Identification of binding partners for ezrin, an actin-binding protein crucial for morphogenesis of apical microvilli and basolateral infoldings in RPE cells. METHODS: Rat eyes, rat primary RPE, the rat RPE-J cell line, and a clonal line of RPE-J cells transfected with human ezrin cDNA were analyzed by immunofluorescence microscopy and immunoblot. Immunofluorescence localization of two ezrin-binding proteins was performed in cryosections of rat eyes of various ages and in monolayers extracted with the detergent Triton X-100 and fixed in paraformaldehyde. The interaction of both proteins with ezrin and gluthathione-S-transferase (GST)-ezrin fusion proteins was analyzed by SDS-PAGE and immunoblot. RESULTS: Immunofluorescence microscopy of adult rat eyes detected a polarized distribution of ERM (ezrin, radixin, and moesin)-binding phosphoprotein of 50 kDa (EBP50) at the apical microvilli and synaptic-associated protein of 97 kDa (SAP97) at the basolateral surface of RPE cells, which overlapped with ezrin. These two PDZ (postsynaptic density protein [PSD-95]/disc large [DLG]-A/ZO-1) domain proteins had a similar polarized distribution and high resistance to detergent extractability, indicative of cytoskeletal association, both in primary cultures of rat RPE and in a clonal RPE-J cell line expressing high levels of transfected ezrin. RPE cell lysates from rat retinas of various postnatal ages revealed increasing levels of EBP50 and SAP97 compared with alphav integrin, a protein expressed at constant adult levels from birth. GST pull-down and immunoprecipitation experiments demonstrated a direct interaction between EBP50 and SAP97 and ezrin. CONCLUSIONS: The data indicate that EBP50 localizes at the apical microvilli, whereas SAP97 localizes at the basolateral surface of RPE cells, probably through a direct interaction with ezrin.

Selective control of basolateral membrane protein polarity by cdc42.

The rho GTPase cdc42 is implicated in several aspects of cell polarity. A recent study (Kroschewski R, Hall A, Mellman I. Nat Cell Biol 1999;1:8-13) demonstrated that a dominant negative mutant of cdc42 abolishes the polarity of basolateral membrane proteins in MDCK cells, but did not elucidate whether this effect was selective for basolateral proteins or nonselective for all secreted proteins. To answer this question, we analyzed the polarity of newly synthesized membrane and soluble proteins in MDCK cell lines previously induced to overexpress mutant forms of cdc42. GTPase-deficient and dominant negative cdc42 did not affect the apical targeting of a newly synthesized apical membrane protein, but reversed to apical the distribution of two exogenous basolateral membrane proteins. In striking contrast, GTPase-deficient cdc42 did not affect polarized exocytosis of endogenous soluble proteins, either apical or basolateral. The exquisitely selective regulation of polarized protein targeting by cdc42 may allow cells to fine-tune their membrane composition in response to extracellular signals during development, migration and in response to injury.

cdc42 regulates the exit of apical and basolateral proteins from the trans-Golgi network.

It is well established that Rho-GTPases regulate vesicle fusion and fission events at the plasma membrane through their modulatory role on the cortical actin cytoskeleton. In contrast, their effects on intracellular transport processes and actin pools are less clear. It was recently shown that cdc42 associates with the Golgi apparatus in an ARF-dependent manner, similarly to coat proteins involved in vesicle formation and to several actin-binding proteins. We report here that mutants of cdc42 inhibited the exit of basolateral proteins from the trans-Golgi network (TGN), while stimulating the exit of an apical marker, in two different transport assays. This regulation may result from modulation of the actin cytoskeleton, as GTPase-deficient cdc42 depleted a perinuclear actin pool that rapidly exchanges with exogenous fluorescent actin.

The C-terminal tail of the metabotropic glutamate receptor subtype 7 is necessary but not sufficient for cell surface delivery and polarized targeting in neurons and epithelia.

Complex neuronal functions rely upon the precise sorting, targeting, and restriction of receptors to specific synaptic microdomains. Little is known, however, of the molecular signals responsible for mediating these selective distributions. Here we report that metabotropic glutamate receptor subtype 7a (mGluR7a) is polarized at the basolateral surface when expressed in Madin-Darby canine kidney (MDCK) epithelial cells but is not polarized when expressed in cultured hippocampal neurons. Truncation of the mGluR7 cytoplasmic tail produces a protein that is restricted to a perinuclear intracellular compartment in both neurons and MDCK cells, where this protein colocalizes with a trans-Golgi network antigen. The mGluR7 cytoplasmic domain appended to the transmembrane portion of the vesicular stomatitis virus G protein and the ectodomain of human placental alkaline phosphatase is distributed over the entire cell surface in cultured neurons. When expressed in MDCK cells, this construct remains in an intracellular compartment distinct from endosomes or lysosomes. Thus, the cytoplasmic tail domain of mGluR7 is necessary but not sufficient for polarized targeting in MDCK monolayers, whereas in neurons the cytoplasmic tail is sufficient for cell surface expression but not polarization. Additional mechanisms are likely required to mediate mGluR7 neuronal polarization and synaptic clustering.

Saturation of, and competition for entry into, the apical secretory pathway.

To investigate mechanisms of apical sorting in the secretory pathway of epithelial cells, we expressed varying amounts of the 165 amino acid isoform of vascular endothelial growth factor (VEGF(165)) and transforming growth factor beta1 (TGF-beta1) via replication defective adenoviruses. Apical sorting of both proteins was efficient at low expression levels but saturated or was reversed at high expression levels. High expression levels of TGF-beta1 were effective at competing VEGF(165) out of the apical pathway; however, VEGF(165) did not compete out TGF-beta1. Tunicamycin inhibition experiments showed that the apical polarity of VEGF(165) was independent of N-glycosylation. We conclude that the apical sorting of these two molecules is a saturable, signal-mediated process, involving competition for apical sorting receptors. The sorting of the two proteins does not appear to involve N-glycans as sorting signals, or lectin sorters. The observations are particularly relevant to gene therapy because they demonstrate that overexpression of a transgene can result in undesirable missorting of the encoded protein.

Dynein- and microtubule-mediated translocation of adenovirus serotype 5 occurs after endosomal lysis.

Modified viruses are used as gene transfer vectors because of their ability to transfer genetic material efficiently to the nucleus of a target cell. To better understand intracellular translocation of adenovirus serotype 5 (Ad), fluorophores were covalently conjugated to Ad capsids, and movement of fluorescent Ad within the cytoplasm was observed during the first hour of infection of a human lung epithelial carcinoma cell line (A549). Ad translocation was characterized with respect to its ability to achieve nuclear envelope localization as well as directed movement in the cytoplasm. Whereas Ad achieved efficient nuclear localization 60 min after infection of A549 cells under control conditions, depolymerization of the microtubule cytoskeleton by addition of 25 microM nocodazole reversibly inhibited development of nuclear localization. In contrast, depolymerization of microfilaments by addition of 1 microM cytochalasin D had no effect on nuclear localization. Direct video observation of Ad motility showed that nocodazole, but not cytochalasin D, caused a reversible decrease in rapid linear translocations of Ad in the cytoplasm of A549 cells. Microinjection of function-blocking antibodies against the microtubule-dependent motor protein, cytoplasmic dynein, but not kinesin, blocked nuclear localization of Ad, consistent with net minus end-directed motility indicated by accumulation of Ad at mitotic spindles. Fluorescence ratio imaging revealed a neutral pH in the environment of translocating Ad, leading to a model in which the interaction of Ad with an intact microtubule cytoskeleton and functional cytoplasmic dynein occurs after escape from endosomes and is a necessary prerequisite to nuclear localization of adenovirus serotype 5.

Ezrin promotes morphogenesis of apical microvilli and basal infoldings in retinal pigment epithelium.

Ezrin, a member of the ezrin/radixin/moesin (ERM) family, localizes to microvilli of epithelia in vivo, where it bridges actin filaments and plasma membrane proteins. Here, we demonstrate two specific morphogenetic roles of ezrin in the retinal pigment epithelium (RPE), i.e., the formation of very long apical microvilli and of elaborate basal infoldings typical of these cells, and characterize the role of ezrin in these processes using antisense and transfection approaches. In the adult rat RPE, only ezrin (no moesin or radixin) was detected at high levels by immunofluorescence and immunoelectron microscopy at microvilli and basal infoldings. At the time when these morphological differentiations develop, in the first two weeks after birth, ezrin levels increased fourfold to adult levels. Addition of ezrin antisense oligonucleotides to primary cultures of rat RPE drastically decreased both apical microvilli and basal infoldings. Transfection of ezrin cDNA into the RPE-J cell line, which has only trace amounts of ezrin and moesin, sparse and stubby apical microvilli, and no basal infoldings, induced maturation of microvilli and the formation of basal infoldings without changing moesin expression levels. Taken together, the results indicate that ezrin is a major determinant in the maturation of surface differentiations of RPE independently of other ERM family members.

Macrophage and retinal pigment epithelium phagocytosis: apoptotic cells and photoreceptors compete for alphavbeta3 and alphavbeta5 integrins, and protein kinase C regulates alphavbeta5 binding and cytoskeletal linkage.

Noninflammatory monocyte macrophages use alphavbeta3 integrin to selectively bind apoptotic cells, initiating their phagocytic removal. In a related process, the retinal pigment epithelium (RPE) employs alphavbeta5 integrin to recognize spent photoreceptor outer segment particles (OS). Here, we show that apoptotic cells and OS compete for binding to these receptors, indicating that OS and apoptotic cells expose surface signals recognizable by alphavbeta3 and alphavbeta5. Particle binding to alphavbeta5 required protein kinase C (PKC) activation. In RPE, alphavbeta5 binding was maximally activated even before any phagocytic challenge and was reduced by PKC inhibitors. In macrophages, it was dormant but became activated upon PKC stimulation. PKC-activated alphavbeta5-mediated binding in macrophages differed from constitutive binding to the same integrin receptor in RPE cells in that the former followed much faster kinetics, similar to particle binding mediated by alphavbeta3. Activation of alphavbeta5 for particle binding correlated with its recruitment into a detergent-insoluble fraction, a process sensitive to pharmacological modulation of PKC in both types of phagocytes. Furthermore, alphavbeta5 but not alphavbeta3 particle binding required actin microfilaments. These data constitute the first evidence that noninflammatory phagocytes actively regulate the earliest phase of phagocytic clearance, particle binding, by controlling receptor activity.

Caveolin-2 localizes to the golgi complex but redistributes to plasma membrane, caveolae, and rafts when co-expressed with caveolin-1.

We have characterized comparatively the subcellular distributions of caveolins-1 and -2, their interactions and their roles in caveolar formation in polarized epithelial cells. In Fischer rat thyroid (FRT) cells, which express low levels of caveolin-2 and no caveolin-1, caveolin-2 localizes exclusively to the Golgi complex but is partially redistributed to the plasma membrane upon co-expression of caveolin-1 by transfection or by adenovirus-mediated transduction. In Madin-Darby canine kidney (MDCK) cells, which constitutively express both caveolin-1 and -2, caveolin-2 localized to both the Golgi complex and to the plasma membrane, where it co-distributed with caveolin-1 in flat patches and in caveolae. In FRT cells, endogenous or overexpressed caveolin-2 did not associate with low density Triton insoluble membranes that floated in sucrose density gradients but was recruited to these membranes when co-expressed together with caveolin-1. In MDCK cells, both caveolin-1 and caveolin-2 associated with low density Triton-insoluble membranes. In FRT cells, transfection of caveolin-1 promoted the assembly of plasma membrane caveolae that localized preferentially (over 99%) to the basolateral surface, like constitutive caveolae of MDCK cells. In contrast, as expected from its intracellular distribution, endogenous or overexpressed caveolin-2 did not promote the assembly of caveolae; rather, it appeared to promote the assembly of intracellular vesicles in the peri-Golgi area. The data reported here demonstrate that caveolin-1 and -2 have different and complementary subcellular localizations and functional properties in polarized epithelial cells and suggest that the two proteins co-operate to carry out specific as yet unknown tasks between the Golgi complex and the cell surface.

Expression of caveolin-1 is required for the transport of caveolin-2 to the plasma membrane. Retention of caveolin-2 at the level of the golgi complex.

Caveolins-1 and -2 are normally co-expressed, and they form a hetero-oligomeric complex in many cell types. These caveolin hetero-oligomers are thought to represent the assembly units that drive caveolae formation in vivo. However, the functional significance of the interaction between caveolins-1 and -2 remains unknown. Here, we show that caveolin-1 co-expression is required for the transport of caveolin-2 from the Golgi complex to the plasma membrane. We identified a human erythroleukemic cell line, K562, that expresses caveolin-2 but fails to express detectable levels of caveolin-1. This allowed us to stringently assess the effects of recombinant caveolin-1 expression on the behavior of endogenous caveolin-2. We show that expression of caveolin-1 in K562 cells is sufficient to reconstitute the de novo formation of caveolae in these cells. In addition, recombinant expression of caveolin-1 allows caveolin-2 to form high molecular mass oligomers that are targeted to caveolae-enriched membrane fractions. In striking contrast, in the absence of caveolin-1 expression, caveolin-2 forms low molecular mass oligomers that are retained at the level of the Golgi complex. Interestingly, we also show that expression of caveolin-1 in K562 cells dramatically up-regulates the expression of endogenous caveolin-2. Northern blot analysis reveals that caveolin-2 mRNA levels remain constant under these conditions, suggesting that the expression of caveolin-1 stabilizes the caveolin-2 protein. Conversely, transient expression of caveolin-2 in CHO cells is sufficient to up-regulate endogenous caveolin-1 expression. Thus, the formation of a hetero-oligomeric complex between caveolins-1 and -2 stabilizes the caveolin-2 protein product and allows caveolin-2 to be transported from the Golgi complex to the plasma membrane.

Glycans in post-Golgi apical targeting: sorting signals or structural props?

A recent model proposed that N-glycans serve as apical targeting signals for soluble and membrane proteins in epithelial cells and neurons by interacting with lectin sorters in the trans-Golgi network. However, we believe that a number of experimental observations support an alternative hypothesis, that N-glycans play a facilitative role, by providing structural support or preventing aggregation of the proteins for example, thereby allowing interaction of proteinaceous apical sorting signals with the sorting machinery. This article discusses the experimental data currently available and how they relate to the proposed models.

Use of the ARPE-19 cell line as a model of RPE polarity: basolateral secretion of FGF5.

PURPOSE: To determine the polarity of fibroblast growth factor 5 (FGF5) secretions from retinal pigment epithelium (RPE) cells and to examine the viability and utility of the ARPE-19 cell line as a model for the study of RPE polarity. METHODS: Influenza infection and adenovirus-mediated gene transfer were used to deliver and express genes encoding influenza hemagglutinin (HA), p75-NTR (a neurotrophin receptor), low-density lipoprotein (LDL) receptor (LDLR), and FGF5 in confluent monolayers of ARPE-19 cells. The localization of HA, p75-NTR, and LDLR was determined by confocal microscopy. Domain selective biotinylation assays were used to quantitatively determine the polarities of p75-NTR and LDLR. The secretion of FGF5 into the apical and basal media of ARPE-19 cultures was examined by immunoblot analysis of conditioned media. RESULTS: Hemagglutinin and p75-NTR were found to be localized on the apical surface of infected and transduced ARPE-19 cells. In contrast, LDLR was associated preferentially with the basolateral membrane of ARPE-19 cells. Biotinylation studies indicated that 84% of p75-NTR was present on the apical surface, and 79% of LDLR was basolaterally polarized. Over the course of 6 hours, more than 90% of the total secreted FGF5 protein accumulated in the basolateral media. CONCLUSIONS: ARPE-19 cells exhibit a polarized distribution of cell surface markers when examined by either confocal microscopy or surface-labeling assays. This indicates that the ARPE-19 cell line is a valid model for studies of RPE cell polarity. FGF5, a secreted protein normally produced by RPE cells, is accumulated preferentially in the basal media after only 6 hours, suggesting that it is vectorially secreted from the basolateral surface of ARPE-19 cells.

Apical polarity of N-CAM and EMMPRIN in retinal pigment epithelium resulting from suppression of basolateral signal recognition.

Retinal pigment epithelial (RPE) cells apically polarize proteins that are basolateral in other epithelia. This reversal may be generated by the association of RPE with photoreceptors and the interphotoreceptor matrix, postnatal expansion of the RPE apical surface, and/or changes in RPE sorting machinery. We compared two proteins exhibiting reversed, apical polarities in RPE cells, neural cell adhesion molecule (N-CAM; 140-kD isoform) and extracellular matrix metalloproteinase inducer (EMMPRIN), with the cognate apical marker, p75-neurotrophin receptor (p75-NTR). N-CAM and p75-NTR were apically localized from birth to adulthood, contrasting with a basolateral to apical switch of EMMPRIN in developing postnatal rat RPE. Morphometric analysis demonstrated that this switch cannot be attributed to expansion of the apical surface of maturing RPE because the basolateral membrane expanded proportionally, maintaining a 3:1 apical/basolateral ratio. Kinetic analysis of polarized surface delivery in MDCK and RPE-J cells showed that EMMPRIN has a basolateral signal in its cytoplasmic tail recognized by both cell lines. In contrast, the basolateral signal of N-CAM is recognized by MDCK cells but not RPE-J cells. Deletion of N-CAM's basolateral signal did not prevent its apical localization in vivo. The data demonstrate that the apical polarity of EMMPRIN and N-CAM in mature RPE results from suppressed decoding of specific basolateral signals resulting in randomized delivery to the cell surface.

Sec6/8 complex is recruited to cell-cell contacts and specifies transport vesicle delivery to the basal-lateral membrane in epithelial cells.

In budding yeast, the Sec6/8p complex is essential for generating cell polarity by specifying vesicle delivery to the bud tip. We show that Sec6/8 homologs are components of a cytosolic, approximately 17S complex in nonpolarized MDCK epithelial cells. Upon initiation of calcium-dependent cell-cell adhesion, approximately 70% of Sec6/8 is rapidly (t(1/2) approximately 3-6 hr) recruited to sites of cell-cell contact. In streptolysin-O-permeabilized MDCK cells, Sec8 antibodies inhibit delivery of LDL receptor to the basal-lateral membrane, but not p75NTR to the apical membrane. These results indicate that lateral membrane recruitment of the Sec6/8 complex is a consequence of cell-cell adhesion and is essential for the biogenesis of epithelial cell surface polarity.

Identification of the basolateral targeting determinant of a peripheral membrane protein, MacMARCKS, in polarized cells.

BACKGROUND: Although the molecular determinants that specify the targeting of transmembrane proteins to the apical or basolateral membrane domains within polarized epithelial cells have been well characterized, very little is known about the targeting of peripheral membrane proteins within these cells. MacMARCKS is a member of the MARCKS family of protein kinase C (PKC) substrates. This myristoylated protein regulates actin structure at cell membranes and is essential for the morphogenic movement of neuroepithelial cells during the formation of the neural tube. RESULTS: MacMARCKS was specifically targeted to sites of cell-cell contact in the basolateral domain of polarized Madin-Darby canine kidney (MDCK) epithelial cells and was displaced from this location upon activation of PKC. We defined the basolateral targeting determinant of MacMARCKS to be the effector domain, a basic region spanning 24 amino acids and containing the PKC phosphorylation sites as well as binding sites for calmodulin and actin. This domain, in conjunction with a myristoyl moiety, was sufficient to target a non-membrane-associated protein--green fluorescent protein--specifically to the basolateral surface of polarized MDCK cells. CONCLUSIONS: This is the first description of a specific amino acid sequence that specifies targeting of a peripheral membrane protein to the basolateral membrane in polarized epithelial cells.

Caveolin transfection results in caveolae formation but not apical sorting of glycosylphosphatidylinositol (GPI)-anchored proteins in epithelial cells.

Most epithelial cells sort glycosylphosphatidylinositol (GPI)-anchored proteins to the apical surface. The "raft" hypothesis, based on data mainly obtained in the prototype cell line MDCK, postulates that apical sorting depends on the incorporation of apical proteins into cholesterol/glycosphingolipid (GSL) rafts, rich in the cholesterol binding protein caveolin/VIP21, in the Golgi apparatus. Fischer rat thyroid (FRT) cells constitute an ideal model to test this hypothesis, since they missort both endogenous and transfected GPI-anchored proteins to the basolateral plasma membrane and fail to incorporate them into cholesterol/glycosphingolipid clusters. Because FRT cells lack caveolin, a major component of the caveolar coat that has been proposed to have a role in apical sorting of GPI-anchored proteins (Zurzolo, C., W. Van't Hoff, G. van Meer, and E. Rodriguez-Boulan. 1994. EMBO [Eur. Mol. Biol. Organ.] J. 13:42-53.), we carried out experiments to determine whether the lack of caveolin accounted for the sorting/clustering defect of GPI-anchored proteins. We report here that FRT cells lack morphological caveolae, but, upon stable transfection of the caveolin1 gene (cav1), form typical flask-shaped caveolae. However, cav1 expression did not redistribute GPI-anchored proteins to the apical surface, nor promote their inclusion into cholesterol/GSL rafts. Our results demonstrate that the absence of caveolin1 and morphologically identifiable caveolae cannot explain the inability of FRT cells to sort GPI-anchored proteins to the apical domain. Thus, FRT cells may lack additional factors required for apical sorting or for the clustering with GSLs of GPI-anchored proteins, or express factors that inhibit these events. Alternatively, cav1 and caveolae may not be directly involved in these processes.

Morphogenesis of the retinal pigment epithelium: toward understanding retinal degenerative diseases.

The phenotype of an epithelial cell is defined by a unique combination of morphology, gene and protein expression, and protein localization. Results indicate that the terminal differentiation of the RPE cell can be described in part by changes in the polarity of its surface proteins alpha v beta 5 integrin, Na,K-ATPase, N-CAM, and EMMPRIN. Changes in protein/gene expression and protein localization in late stages of RPE development identify alpha v beta 5 integrin as a key player in RPE phagocytosis, and N-CAM and EMMPRIN as potentially important molecules in other RPE functions necessary for photoreceptor survival. By studying the trafficking of the later two proteins it is shown that entry into an apical or basolateral pathway in RPE cells cannot be predicted by the distribution of a given protein in other epithelial cells, and that this distribution may change through the course of RPE development. The mechanisms used by RPE and other epithelia to establish and maintain their specific polarity properties are fundamental to the formation and maintenance of their specific epithelial phenotype. The ability to therapeutically direct molecules incorporated into RPE by gene therapy into apical or basal surfaces requires an understanding of protein localization and expression. Furthermore, evidence is provided that assays capitalizing on changes in gene/protein expression and protein localization during the late stages of RPE development can prove a productive way of identifying proteins used by RPE for photoreceptor support. This approach can continue to be exploited to identify other proteins essential for the mission of the RPE cell, that may thus be likely candidates for participation in retinal degenerative disease.