Classical cadherins control survival through the gp130/Stat3 axis.

Stat3 (Signal Transducer and Activator of Transcription-3) is activated by a number of receptor and nonreceptor tyrosine kinases. We recently demonstrated that engagement of E-cadherin, a calcium-dependent, cell to cell adhesion molecule which is often required for cells to remain tightly associated within the epithelium, also activates Stat3. We now examined the effect of two other classical cadherins, cadherin-11 and N-cadherin, whose expression often correlates with the epithelial to mesenchymal transition occurring in metastasis of carcinoma cells, upon Stat3 phosphorylation and activity. Our results indicate that engagement of these two cadherins also, can trigger a dramatic surge in Stat3 activity. This activation occurs through upregulation of members of the IL6 family of cytokines, and it is necessary for cell survival, proliferation and migration. Interestingly, our results also demonstrate for the first time that, in sharp contrast to Stat3, the activity of Erk (Extracellular Signal Regulated kinase) was unaffected by cadherin-11 engagement. Further examination indicated that, although IL6 was able to activate Erk in sparsely growing cells, IL6 could not induce an increase in Erk activity levels in densely growing cultures. Most importantly, cadherin-11 knock-down did allow Erk activation by IL6 at high densities, indicating that it is indeed cadherin engagement that prevents Erk activation by IL6. The fact that the three classical cadherins tested so far, E-cadherin, N-cadherin and cadherin11, which are present in essentially all tissues, actually activate Stat3 regardless of their role in metastasis, argues for Stat3 as a central survival, rather than invasion factor.

Homophilic interactions between cadherin fragments at the single molecule level: an AFM study.

We report measurements of the adhesion forces between single E-cadherin fragments anchored on solid surfaces. These fragments consist of the two outermost extracellular domains of the protein. The specificity of the measured rupture forces was demonstrated by Ca2+ exchange experiments. Two series of experiments were performed using two linkers of different rigidity and length. We find that the pull-off force is distributed with a maximum value independent of the linker and logarithmically dependent on the velocity of separation of the two surfaces. Our dynamical results are compatible with previous flow chamber experiments performed with the same fragments and can be compared from a different perspective with previously reported AFM experiments on the full-length extracellular domain of the VE-cadherin. Interestingly, using a rigid linker, we have been able for the first time to evidence the deformation of the cadherin molecule under mechanical stress, a piece of information not accessible with more classical grafting strategies.

Trans-bonded pairs of E-cadherin exhibit a remarkable hierarchy of mechanical strengths.

Classical cadherins are primary mediators of calcium-dependent cell interactions in multicellular organisms. Organized in five tandemly repeated E-cadherin (EC) modules, the extracellular segments of these membrane-spanning glycoproteins interact homophilically between opposing cells to create highly regulated patterns of attachment stabilized by cytoskeletal elements inside the cells. Despite many structural and functional studies, a significant controversy exists in regard to the organization of cadherin binding in adhesion sites. Supported by considerable evidence, perhaps the most widely held view is that opposing N-terminal EC1-EC2 (EC12) domains form a "zipper" of bonds. However, immobilized on two atomically smooth surfaces and pushed to adhesive contact, opposing cadherins become fully interdigitated and unbind through three discrete jumps comparable with domain dimensions when pulled apart. So the question remains as to whether mechanical adhesion strength emanates solely from interactions between the peripheral N-terminal domains or involves multiple overlapping domains. It is also unclear whether a primary adhesion complex is formed by a single opposing pair of cadherins or whether the complex involves a more complicated network of cis-bonded multimers. To address these questions, we used a special jump/ramp mode of force spectroscopy to test isolated pairwise interactions between recombinant fragments of ECs. Besides the formation of strong trans-bonded dimers, we find a remarkable hierarchy of rupture strengths for bonds between the full five-domain fragments that suggests multiple mechanical functions for cadherins, perhaps providing distinct properties needed for transient-specific recognition as well as stable tissue formation.

Adhesion between giant vesicles and supported bilayers decorated with chelated E-cadherin fragments.

Here, we present a study of adhesion between cadherin fragments using giant unilamellar vesicles and supported bilayers. These objects are partially made of nickel chelating lipids and are subsequently decorated with proteins bearing a 6His tag. Initially, we observed their fixation and correct orientation by using a fluorescent protein, the green fluorescent protein (GFP)-6His. The adhesive behavior of E-cadherin functionalized giant vesicles and supported bilayers was studied as a function of the calcium concentration and of the protein functionality by reflection interference microscopy. We show that such a system retains specific cadherin-mediated adhesion and could be used to study the statics and dynamics of adhesive plaques as well as to gain insight into the fundamental mechanisms of cellular adhesion at the mesoscopic scale.

Expression of laminin-2 by normal and neoplastic rat C cells during the development of medullary thyroid carcinoma.

Medullary thyroid carcinoma (MTC) originates from C cells, which secrete calcitonin (CT), their specific marker. C cells are located in contact with the basement membrane (BM) of the thyroid follicles, which is partly made up of the laminin-2 isoform synthesized by thyrocytes. During oncogenesis, proliferation of the C cells, invading the centre of the follicles, leads to a break in their normal contact with the BM. As specific interactions of cells with BM components, especially laminins, are important for proliferation and differentiation, we investigated the relationships of normal and neoplastic C cells with laminin in the Wag/Rij rat model of human MTC. Immunocytochemical studies showed a progressive loss of the laminin layer underlying the hyperplastic C cell nodules around the large dedifferentiated tumours. The alpha2, beta1 and gamma1 chains of the laminin-2 isoform were synthesized and secreted by rat MTC 6-23 cell cultures and the tumours induced by subcutaneous injection of these cells. In situ hybridization combined with anti-CT immunocytochemistry showed a low expression of alpha2 mRNA on differentiated C cells and thyrocytes, but an overexpression on immunonegative spontaneous MTC and induced intrathyroid tumours. The high level of alpha2 gene expression, together with tumour dedifferentiation, suggests a relationship with malignancy.

Expression of the cytoplasmic domain of E-cadherin induces precocious mammary epithelial alveolar formation and affects cell polarity and cell-matrix integrity.

Cadherins are cell adhesion molecules involved in cell-cell adhesion, signalling, and cellular proliferation and differentiation. E-cadherin is required for the formation of epithelium in vivo. We investigated the contribution of the cytoplasmic domain of E-cadherin to adhesion, signalling, and differentiation during murine mammary gland development, by in vivo expression of a gene encoding a truncated form of E-cadherin lacking the extracellular domain. The expression of this gene in mammary epithelial cells during pregnancy induced precocious lobular epithelial morphogenesis associated with morphological differentiation and the early synthesis of various molecules (advanced milk fat globule appearance and milk protein production). After delivery, when a fully differentiated and secretory epithelium is required for lactation, the cytoplasmic domain of E-cadherin had a dominant-negative effect on cell-cell adhesion and affected the structure and function of the epithelium. This also led to the partial loss of epithelial polarisation and changes in the basement membrane, both important in malignancy. Thus, the cytoplasmic domain of E-cadherin induces epithelial morphogenesis, but also alters the cohesiveness of the fully differentiated epithelium.

Experimental study of the interaction range and association rate of surface-attached cadherin 11.

We describe a method allowing quantitative determination of the interaction range and association rate of individual surface-attached molecules. Spherical beads (1.4 micro(m) radius) were coated with recombinant outer domains of the newly described classical type II cadherin 11, a cell adhesion molecule. Beads were driven along cadherin-coated surfaces with a hydrodynamic force of approximately 1 pN, i.e., much less than the mechanical strength of many ligand-receptor bonds. Spheres displayed periods of slow motion interspersed with arrests of various duration. Particle position was monitored with 50 Hz frequency and 0.025 micro(m) accuracy. Nearly 1 million positions were recorded and processed. Comparison between experimental and computer-simulated trajectories suggested that velocity fluctuations might be related quantitatively to Brownian motion perpendicular to the surface. The expected amplitude of this motion was of order of 100 nm. Theoretical analysis of the relationship between sphere acceleration and velocity allowed simultaneous determination of the wall shear rate and van der Waals attraction between spheres and surface. The Hamaker constant was estimated at 2.9 x 10(-23) J. The frequency of bond formation was then determined as a function of sphere velocity. Experimental data were consistent with the view that the rate of association between a pair of adhesion molecules was approximately 1.2 x 10(-3) s-1 and the interaction range was approximately 10 nm. It is concluded that the presented methodology allows sensitive measurement of sphere-to-surface interactions (with approximately 10 fN sensitivity) as well as the effective range and rate of bond formation between individual adhesion molecules.

Merosin is synthesized by thyroid cells in primary culture irrespective of cellular organization.

The in vitro synthesis and deposition of laminin family glycoproteins were studied using primary porcine thyroid cells cultured as monolayers or in follicles. The latter organization mimics the in vivo state of these polarized epithelial cells. In both cell systems a trimeric molecule was immunoprecipitated by using polyclonal antibodies against EHS-laminin. When the cells were fully polarized the protein was found at the basal pole of cells, irrespective of their organization. However, this molecule was different from laminin purified from a traditional source, the murine Engelbreth-Holm-Swarm (EHS) tumor. Thyroid cell laminin was composed of two light chains, analogous to EHS B1 and B2, and a disulfide-bonded heavy chain not found in EHS-laminin. The heavy chain was first synthesized as a 380 kDa polypeptide, then rapidly cleaved to a doublet of 350-380 kDa, which was subsequently found in both cell extracts and conditioned culture media. This thyroid laminin variant was compared with merosin, another variant found in the basement membranes of trophoblast, Schwann cells, striated muscle and liver. The heavy chain (M) of merosin shows homology to EHS-laminin heavy chain at the C-terminal domain, and is usually found as two polypeptides of 80 kDa and 300 kDa (Ehrig K., Leivo I., Argraves W. S., Ruoslahti E. and Engvall E. Proc. Nat. Acad. Sci. 87, 3264-3268, 1990). mRNA of the M chain was identified by RT-PCR in freshly isolated thyrocytes as well as in 6-day-old cultured thyroid cells. Furthermore, both the classical laminin heavy chain and the 350 kDa variant were detected by immunoblotting and immunofluorescence in the thyroid gland in vivo. All these results suggest strongly that merosin is a basement membrane component of thyroid cells in vivo and in vitro.

Molecular organization of the intestinal brush border.

The brush border of enterocytes represents one of the more specialized apical poles of epithelial cells. It is formed by particularly well-developed apical plasma membrane microvilli, whose shape is ensured by a highly organized cytoskeleton. The molecular organization of the cytoskeleton is described. Whereas several cytoskeleton proteins are ubiquitous, villin is highly specific for intestinal cells and can be used as a differentiation marker of these cells. The major glycoproteins, in particular hydrolases, of the brush border membrane have been characterized. They have many common structural features, in particular their mode of integration into the membrane by their N-terminal hydrophobic sequences that also plays the role of the 'signal peptide' responsible for their co-translational insertions into the endoplasmic reticulum. Studies on the biosynthesis and intracellular pathway of aminopeptidase N strongly suggest that sorting of apical and basolateral glycoproteins could occur after their integration into the basolateral domain.

Biogenesis of the rat hepatocyte plasma membrane in vivo: comparison of the pathways taken by apical and basolateral proteins using subcellular fractionation.

We have used pulse-chase metabolic radiolabeling with L-[35S]methionine in conjunction with subcellular fractionation and specific protein immunoprecipitation techniques to compare the posttranslational transport pathways taken by endogenous domain-specific integral proteins of the rat hepatocyte plasma membrane in vivo. Our results suggest that both apical (HA 4, dipeptidylpeptidase IV, and aminopeptidase N) and basolateral (CE 9 and the asialoglycoprotein receptor [ASGP-R]) proteins reach the hepatocyte plasma membrane with similar kinetics. The mature molecular mass form of each of these proteins reaches its maximum specific radioactivity in a purified hepatocyte plasma membrane fraction after only 45 min of chase. However, at this time, the mature radiolabeled apical proteins are not associated with vesicles derived from the apical domain of the hepatocyte plasma membrane, but instead are associated with vesicles which, by several criteria, appear to be basolateral plasma membrane. These vesicles: (a) fractionate like basolateral plasma membrane in sucrose density gradients and in free-flow electrophoresis; (b) can be separated from the bulk of the likely organellar contaminants, including membranes derived from the late Golgi cisternae, transtubular network, and endosomes; (c) contain the proven basolateral constituents CE 9 and the ASGP-R, as judged by vesicle immunoadsorption using fixed Staphylococcus aureus cells and anti-ASGP-R antibodies; and (d) are oriented with their ectoplasmic surfaces facing outward, based on the results of vesicle immunoadsorption experiments using antibodies specific for the ectoplasmic domain of the ASGP-R. Only at times of chase greater than 45 min do significant amounts of the mature radiolabeled apical proteins arrive at the apical domain, and they do so at different rates. Approximate half-times for arrival are in the range of 90-120 min for aminopeptidase N and dipeptidylpeptidase IV whereas only 15-20% of the mature radiolabeled HA 4 associated with the hepatocyte plasma membrane fraction has become apical even after 150 min of chase. Our results suggest a mechanism for hepatocyte plasma membrane biogenesis in vivo in which all integral plasma membrane proteins are shipped first to the basolateral domain, followed by the specific retrieval and transport of apical proteins to the apical domain at distinct rates.

The establishment of hepatocyte cell surface polarity during fetal liver development.

Antibodies to six glycoproteins present in different domains of the hepatocyte plasma membrane were used to study the establishment of cell surface polarity during rat fetal liver development. The proteins were immunoprecipitated from fetal liver homogenates between 14 and 21 days of gestation and quantified by immunoblotting. Aminopeptidase N, CE 9, and HA 321, which reside in the apical, basolateral, and lateral plasma membrane in the adult hepatocyte, respectively, were present in high concentrations at 14 days of gestation and remained high until birth. In contrast, two apical proteins (HA 4 and dipeptidyl peptidase IV) and two basolateral proteins (ASGP receptor and EGF receptor) were first detected between 16 and 18 days of gestation and increased linearly until birth. HA 4 was the only molecule for which the fetal and adult forms differed, with the former having a faster mobility on SDS-PAGE, due to differences in N-linked oligosaccharides. With two exceptions, the localization of the molecules from earliest detection was restricted to the same domain as that in the adult. At 15 days of gestation, HA 321 and a small portion of aminopeptidase were detected on the basolateral membrane. By 21 days both molecules had assumed their adult localization pattern. Our results indicate that the biogenesis of cell surface polarity is an early event, implying that the mechanisms for sorting plasma membrane molecules are functional very early in development. Furthermore, the different patterns of appearance of the six molecules, irrespective of domain, indicate that the biochemical composition of the cell surface changes dramatically during fetal liver development.

Evidence for the transit of aminopeptidase N through the basolateral membrane before it reaches the brush border of enterocytes.

In vivo pulse-chase labeling of rabbit jejunum loops was used in conjunction with subcellular fractionation and quantitative immunoprecipitation to determine whether or not the newly synthesized aminopeptidase N transits through the basolateral membrane before it reaches the apical brush border, its final localization. The kinetics of the arrival of the newly synthesized enzyme in the Golgi complex, basolateral and brush border membrane fractions strongly suggest that on leaving the Golgi aminopeptidase N is transiently integrated into the basolateral domain before reaching the brush border.

Subcellular fractionation and subcellular localization of aminopeptidase N in the rabbit enterocytes.

A fast and easy procedure is proposed for preparing concomitantly from the same sample of intestinal mucosa of A+ rabbits, four fractions high enriched in the brush-border and basolateral plasma membrane domains, rough endoplasmic reticulum, and smooth endoplasmic reticulum plus Golgi apparatus membranes, respectively. This is the first time the technique of flow fluorometry has been applied to characterize the brush-border and basolateral membrane fractions using polyclonal or monoclonal antibodies against antigens common to or specific for these two plasma membrane domains. This technique definitely proves the presence of aminopeptidase in at least 60% of the basolateral membrane vesicles, where its level is about 4.5% of that in the brush-border membrane vesicles. The endoglycosidase H-sensitive intermediate of glycosylation of aminopeptidase N in the steady state is accumulated in both the rough and smooth endoplasmic reticulum membranes. Although the rough membrane is more extensive it contains only about 40% of this transient form.

Biosynthesis and intracellular pool of aminopeptidase N in rabbit enterocytes.

A papain treatment at 15 degrees C and pH 7.3 of a microsomal fraction from rabbit enterocytes quantitatively releases the aminopeptidase N integrated in the plasma membranes without solubilizing the enzyme integrated in the intracellular membranes. Working on A+ rabbits, characterized by the presence on the brush-border hydrolases of glycans corresponding to the human blood group A-determinant structure, it was possible to separate the intracellular aminopeptidase into two major molecular forms with or without these determinants. The molecular form devoid of human blood group A antigenicity corresponds to the only stable intermediate of glycosylation, bearing N-linked high mannose oligosaccharides. This endoglycosidase H-sensitive form is fully active and represents in the steady state about 1% of the total cellular aminopeptidase. It contains a cytoplasmic sequence of about 3000 daltons that has not yet been detected in the mature form. The A antigenicity is acquired simultaneously with processing of high mannose glycans to complex glycans. Pulse chase labeling of jejunum loops with [35S]-methionine showed that the complete processing of the transient form synthesized during 10 min takes 1 hr. During the last 30 min of processing, all the newly transformed molecules are transported to the plasma membrane.

Human blood group A-like determinants as marker of the intracellular pools of glycoproteins in secretory and absorbing of A+ rabbit jejunum.

Human blood group A antigenicity of glycoproteins is retained on epon-embedded jejunum sections after glutaraldehyde fixation and osmium treatment. The intracellular location of molecules bearing these determinants was visualized in the four types of epithelial cells of A+ rabbit jejunum sections with immuno-colloidal gold labeling. The brush border membrane and in particular the glycocalyx of absorbing cells as well as the secretory granules of goblet and Paneth cells were heavily labeled. In enteroendocrine cells, the membrane of secretory granules and not their content was lightly labeled. The differential labeling of secretory or membrane bound glycoproteins is accompanied by different labels of the Golgi complex as expected if labeling of the Golgi saccules was due to the presence of glycoproteins in transit. In all cases the label is primarily concentrated in only half the cisternae on the trans side of the Golgi stacks. In absorbing cells, structures have been revealed in the terminal web that could be related to the brush border membrane and consequently implicated in its biogenesis. The fibrillar material of the glycocalyx appears as highly labeled tangled structures which apparently proceed from densely stained "carrier" vesicles arising from the Golgi apparatus. Vesicles fusing at the lower part of microvilli could result of integration of this material into the lightly labeled vesicles strictly found in the terminal web. These last vesicles could also contain newly synthesized brush border hydrolases.

Aminopeptidases and proteolipids of intestinal brush border.

The presence of human blood group A determinants has been shown on the A+ rabbit intestinal brush border glycoproteins, particularly hydrolases. Sugar compositions of aminopeptidases N from A+ and A- rabbits were compatible with the presence in these molecules of eight N-linked glycans and of two O-linked glycans bearing the A determinants in A+ animals. The exact relative molecular masses of hydrophobic domain(s) of aminopeptidases N and A from pig and rabbit intestinal brush border have been determined by an isotopic dilution technique. The values obtained were compatible with the anchorage in the membrane of the monomeric rabbit enzymes, or of each subunit of the dimeric pig enzymes, by their N-terminal sequences, composed of 20-25 hydrophobic amino acids. This N-terminal hydrophobic sequence (14 residues) has been determined for rabbit aminopeptidase N. Short peptides containing approximately 60% hydrophobic amino acids have been extracted by chloroform-methanol from purified brush border and basolateral membranes of pig enterocytes. Their molecular properties were very similar to those of the aminopeptidase anchors released by trypsin treatment of detergent-extracted enzymes. However, several lines of evidence failed to support the assumption that these free hydrophobic peptides can be identified with anchors left inside the bilayer after proteolytic cleavage of surface hydrolases.

Localization by immunofluorescence and histochemical labeling of aminopeptidase N in relation to its biosynthesis in rabbit and pig enterocytes.

Antibodies raised against highly purified rabbit intestinal brush border aminopeptidase N were found in certain rabbits or pigs to crossreact with human blood-grouplike substances present in goblet cells and at the surface of the basolateral membrane of enterocytes. To obtain antibodies strictly specific for aminopeptidase, depletion of antibodies with a water-insoluble, mucus-rich fraction or human A erythrocytes was necessary. By using ultrathin frozen sections of rabbit and pig jejunum mucosa, both the brush border and a structure at the upper pole of the nucleus were labeled by the specific antiaminopeptidase antibodies. This structure, tentatively identified as the Golgi apparatus in which aminopeptidase is glycosylated, escaped detection by histochemistry, thus suggesting that the enzyme is not yet active at this stage of its intracellular processing. By contrast, histochemistry revealed active aminopeptidase molecules in another structure located under the terminal web and in the upper part of the lateral membrane.

The amino acid sequence of the hydrophobic anchor of rabbit intestinal brush border aminopeptidase N.

The N-terminal sequence (14 residues) of the detergent form of rabbit intestinal aminopeptidase N was shown to be different from that of the protease form of the same enzyme and to be mostly hydrophobic. This finding is fully consistent with a previous assumption according to which this class of enzymes may be anchored to the brush border membrane by their N-terminus. This special mode of assembly may be facilitated by a positively charged lysine residue near the beginning of the sequence (Lys 4) just before an uninterrupted stretch of hydrophobic amino acids.

Kinetic studies of the localization of aminopeptidase N in monolayer and in follicle-associated cultures of porcine thyroid cells.

The fate of aminopeptidase N (APN) was investigated by means of immunofluorescence in porcine thyroid-cell cultures. At day 1, control cells cultured without TSH formed flat, cylinder-like aggregates in which APN was mainly located on the lateral surface. At day 2, the fluorescence completely disappeared. At days 3 and 4, APN appeared in the apical pole of monolayer cells directed toward the medium and in intracellular organelles. The cilium in each cell was intensely labelled and located in the center of the apical pole. In thyrotropin-stimulated cell cultures, the fluorescence at days 1 and 2 was located over the entire plasma membrane. At days 3 and 4, morphological polarization of cells, reorganized into follicles, occurs, APN being located either in the apical or basolateral regions and in intracellular vesicles. Between day 4 and 6, segregation of the enzyme to the apical membrane occurred. Between day 6 and 14, fluorescence was generally retained in the apical region and in an intracellular pool. In growing cell cultures, after 2 to 5 subcultures, the cells exposed APN in a pattern similar to primary monolayers. After conversion of monolayer cells into follicles, the antigen was recovered in the apical pole of cells facing the follicular lumen.