Syndecan-4 and focal adhesion function.

Two groups have now reported the viability of mice that lack syndecan-4. These mice have wound healing/angiogenesis problems, and fibroblasts from these animals differ in adhesion and migration from normal. This is consistent with recent in vitro data indicating a need for signaling via syndecan-4 for focal adhesion formation, and reports that overexpression of proteins that bind syndecan-4 can modify cell adhesion and migration.

Syndecans: synergistic activators of cell adhesion.

Cell-surface proteoglycans participate in cell adhesion, growth-factor signalling, lipase activity and anticoagulation. Until recently, only the roles of the glycosaminoglycan chains were investigated. Now, with molecular characterization of several core proteins, the roles of each individual proteoglycan species in cellular signalling pathways are being determined. This review describes some of the recent advances in our understanding of the major transmembrane group of heparan sulfate proteoglycans, the syndecans, including evidence that they play an important role as accessory signalling molecules modulating integrin-based adhesion.

Heterogeneous distribution of a basement membrane heparan sulfate proteoglycan in rat tissues.

A heparan sulfate proteoglycan (HSPG) synthesized by murine parietal yolk sac (PYS-2) cells has been characterized and purified from culture supernatants. A monospecific polyclonal antiserum was raised against it which showed activity against the HSPG core protein and basement membrane specificity in immunohistochemical studies on frozen tissue sections from many rat organs. However, there was no reactivity with some basement membranes, notably those of several smooth muscle types and cardiac muscle. In addition, it was found that pancreatic acinar basement membranes also lacked the HSPG type recognized by this antiserum. Those basement membranes that lacked the HSPG strongly stained with antisera against laminin and type IV collagen. The striking distribution pattern is possibly indicative of multiple species of basement membrane HSPGs of which one type is recognized by this antiserum. Further evidence for multiple HSPGs was derived from the finding that skeletal neuromuscular junction and liver epithelia also did not contain this type of HSPG, though previous reports have indicated the presence of HSPGs at these sites. The PYS-2 HSPG was shown to be antigenically related to the large, low buoyant density HSPG from the murine Engelbreth-Holm swarm tumor. It was, however, confirmed that only a single population of antibodies was present in the serum. Despite the presence of similar epitopes on these two proteoglycans of different hydrodynamic properties, it was apparent that the PYS-2 HSPG represents a basement membrane proteoglycan of distinct properties reflected in its restricted distribution in vivo.

cDNA cloning of the basement membrane chondroitin sulfate proteoglycan core protein, bamacan: a five domain structure including coiled-coil motifs.

Basement membranes contain several proteoglycans, and those bearing heparan sulfate glycosaminoglycans such as perlecan and agrin usually predominate. Most mammalian basement membranes also contain chondroitin sulfate, and a core protein, bamacan, has been partially characterized. We have now obtained cDNA clones encoding the entire bamacan core protein of Mr = 138 kD, which reveal a five domain, head-rod-tail configuration. The head and tail are potentially globular, while the central large rod probably forms coiled-coil structures, with one large central and several very short interruptions. This molecular architecture is novel for an extracellular matrix molecule, but it resembles that of a group of intracellular proteins, including some proposed to stabilize the mitotic chromosome scaffold. We have previously proposed a similar stabilizing role for bamacan in the basement membrane matrix. The protein sequence has low overall homology, apart from very small NH2- and COOH-terminal motifs. At the junctions between the distal globular domains and the coiled-coil regions lie glycosylation sites, with up to three N-linked oligosaccharides and probably three chondroitin chains. Three other Ser-Gly dipeptides are unfavorable for substitution. Fusion protein antibodies stained basement membranes in a pattern commensurate with bamacan, and they also Western blotted bamacan core protein from rat L2 cell cultures. The antibodies could also specifically immunoprecipitate an in vitro transcription/translation product from a full-length bamacan cDNA. The unusual structure of this proteoglycan is indicative of specific functional roles in basement membrane physiology, commensurate with its distinct expression in development and changes in disease models.

Immunological characterization of a basement membrane-specific chondroitin sulfate proteoglycan.

Reichert's membrane, an extraembryonic membrane present in developing rodents, has been proposed as an in vivo model for the study of basement membranes. We have used this membrane as a source for isolation of basement membrane proteoglycans. Reichert's membranes were extracted in a guanidine/3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate buffer followed by cesium chloride density-gradient ultracentrifugation under dissociative conditions. The proteoglycans were subsequently purified from the two most dense fractions (greater than 1.3 g/ml) by ion-exchange chromatography. Mice were immunized with the proteoglycan preparation and four mAbs recognizing the core protein of a high-density, buoyant chondroitin sulfate proteoglycan were raised. Confirmation of antibody specificity was carried out by the preparation of affinity columns made from each of the mAbs. Chondroitin sulfate proteoglycans (CSPGs) were purified from both supernatant and tissue fractions of Reichert's membranes incubated in short-term organ culture in the presence of radiolabel. The resultant affinity-purified proteoglycan samples were examined by gel filtration, SDS-PAGE, and immunoblotting. This proteoglycan is of high molecular weight (Mr = 5-6 x 10(5)), with a core protein of Mr = approximately 1.5-1.6 x 10(5) and composed exclusively of chondroitin sulfate chains with an average Mr = 1.6-1.8 x 10(4). In addition, a CSPG was purified from adult rat kidney, whose core protein was also Mr = 1.6 x 10(5). The proteoglycan and its core protein were also recognized by all four mAbs. Indirect immunofluorescence of rat tissue sections stained with these antibodies reveal a widespread distribution of this proteoglycan, localized specifically to Reichert's membrane and nearly all basement membranes of rat tissues. In addition to heparan sulfate proteoglycans, it therefore appears that at least one CSPG is a widespread basement membrane component.

Adhesion, growth, and matrix production by fibroblasts on laminin substrates.

Human embryonic skin fibroblasts have been shown to attach and spread on laminin substrates in the absence of protein synthesis and presence of fibronectin-depleted serum and anti-fibronectin antibodies. Rates of attachment and the type of spreading are virtually identical on fibronectin and laminin-coated substrates with the development of microfilament bundles and focal adhesions. Antibodies to laminin, but not fibronectin, will prevent or reverse fibroblast adhesion to laminin, whereas antibodies to fibronectin but not laminin will give similar results on fibronectin-coated substrates. These and other results indicate that fibroblasts possess distinct receptors for laminin and fibronectin which on contact with suitable substrates promote adhesion through interaction with common intermediates. This type of adhesion is compatible with subsequent growth and extracellular matrix production.

Synthesis and localization of two sulphated glycoproteins associated with basement membranes and the extracellular matrix.

Two sulphated glycoproteins (sgps) of apparent molecular weight (Mr) 180,000 and 150,000, are synthesized by murine PYS and PF HR9 parietal endoderm and Swiss 3T3 cells. The Mr 150,000 sgp has a similar chemical structure to the sulphated glycoprotein, C, synthesized and laid down in Reichert's membrane by mouse embryo parietal endoderm cells (Hogan, B. L.M., A. Taylor, and A.R. Cooper, 1982, Dev. Biol., 90:210-214). Both the Mr 180,000 and 150,000 sgps are deposited in the detergent-insoluble matrix of cultured cells, but they do not apparently undergo any disulphide-dependent intermolecular interactions and are not precursors or products of each other. They contain asparagine-linked oligosaccharides, but these are not the exclusive sites of sulphate labeling. Antiserum raised against the Mr 150,000 sgp C of Reichert's membranes has been used in an immunohistochemical analysis of rat skin. In early foetal and adult skin the antigen is present only in basement membranes, but transiently before and after birth it is also found throughout the upper part of the dermis. This suggests that 150,000 sgp C is at times synthesized by nonepithelial cells and contributes to the extracellular matrix of mesenchymal tissues.

Fibronectin has a dual role in locomotion and anchorage of primary chick fibroblasts and can promote entry into the division cycle.

Fibronectin (FN), which is already known to be a natural factor for fibroblast spreading on substrata, has now been shown to be essential for two distinct types of adhesion with different biological functions in chick heart fibroblasts, namely adhesion directed toward locomotion and toward stationary anchorage for growth. Manipulation of culture conditions and the use of antisera of differing specificities has demonstrated that both exogenous and cell-derived FN are important in each process. The organization of the fibronectin-containing matrix differs between the two states. Immunoelectron microscopy with a colloidal gold marker reveals the presence of small membrane-associated plaques of fibronectin in motile cells with associated submembranous specialization. A fibrillar matrix containing fibronectin is dominant in nonmotile, growing fibroblasts. The development of focal adhesions for stationary anchorage can be dramatically enhanced by addition of cell-derived FN at an appropriate stage, and this promotes entry into the growth cycle. New macromolecular synthesis in addition to FN is necessary for focal adhesion development but not for locomotion.

A cell surface chondroitin sulfate proteoglycan, immunologically related to CD44, is involved in type I collagen-mediated melanoma cell motility and invasion.

The metastatic spread of tumor cells occurs through a complex series of events, one of which involves the adhesion of tumor cells to extracellular matrix (ECM) components. Multiple interactions between cell surface receptors of an adherent tumor cell and the surrounding ECM contribute to cell motility and invasion. The current studies evaluate the role of a cell surface chondroitin sulfate proteoglycan (CSPG) in the adhesion, motility, and invasive behavior of a highly metastatic mouse melanoma cell line (K1735 M4) on type I collagen matrices. By blocking mouse melanoma cell production of CSPG with p-nitrophenyl beta-D-xylopyranoside (beta-D-xyloside), a compound that uncouples chondroitin sulfate from CSPG core protein synthesis, we observed a corresponding decrease in melanoma cell motility on type I collagen and invasive behavior into type I collagen gels. Melanoma cell motility on type I collagen could also be inhibited by removing cell surface chondroitin sulfate with chondroitinase. In contrast, type I collagen-mediated melanoma cell adhesion and spreading were not affected by either beta-D-xyloside or chondroitinase treatments. These results suggest that mouse melanoma CSPG is not a primary cell adhesion receptor, but may play a role in melanoma cell motility and invasion at the level of cellular translocation. Furthermore, purified mouse melanoma cell surface CSPG was shown, by affinity chromatography and in solid phase binding assays, to bind to type I collagen and this interaction was shown to be mediated, at least in part, by chondroitin sulfate. Additionally we have determined that mouse melanoma CSPG is composed of a 110-kD core protein that is recognized by anti-CD44 antibodies on Western blots. Collectively, our data suggests that interactions between a cell surface CD44-related CSPG and type I collagen in the ECM may play an important role in mouse melanoma cell motility and invasion, and that the chondroitin sulfate portion of the proteoglycan seems to be a critical component in mediating this effect.

Selective immunoreactivities of kidney basement membranes to monoclonal antibodies against laminin: localization of the end of the long arm and the short arms to discrete microdomains.

To examine the ultrastructural distribution of laminin within kidney basement membranes, we prepared rat anti-mouse laminin mAbs to use in immunolocalization experiments. Epitope domains for these mAbs were established by immunoprecipitation, immunoblotting, affinity chromatography, and rotary shadow EM. One mAb bound to the laminin A and B chains on blots and was located to a site approximately 15 nm from the long arm-terminal globular domain as shown by rotary shadowing. Conjugates of this long arm-specific mAb were coupled to horseradish peroxidase (HRP) and intravenously injected into mice. Kidney cortices were fixed for microscopy 3 h after injection. HRP reaction product was localized irregularly within the renal glomerular basement membrane (GBM) and throughout mesangial matrices. In addition, this mAb bound in linear patterns specifically to the laminae rarae of basement membranes of Bowman's capsule and proximal tubule. This indicates the presence of the long arm immediately beneath epithelial cells in these sites. The laminae densae of these basement membranes were negative by this protocol. In contrast, the lamina rara and densa of distal tubular basement membranes (TBM) were both heavily labeled with this mAb. A different ultrastructural binding pattern was seen with eight other mAbs, including two that mapped to different sites on the short arms by rotary shadowing and five that blotted to a large pepsin-resistant laminin fragment (P1). These latter mAbs bound weakly or not at all to GBM but all bound throughout mesangial matrices. In contrast, discrete spots of HRP reaction product were seen across all layers of Bowman's capsule BM and proximal TBM. These same mAbs, however, bound densely across the full width of distal TBM. Our findings therefore show that separate strata of different basement membranes are variably immunoreactive to these laminin mAbs. The molecular orientation or integration of laminin into the three dimensional BM meshwork therefore varies with location. Alternatively, there may be a family of distinct laminin-like molecules distributed within basement membranes.

Syndecan proteoglycan contributions to cytoskeletal organization and contractility.

Cells exert tension on the extracellular matrix through specific receptors that link to the actin cytoskeleton. The best characterized are the integrins, which, when activated and clustered, can link to the extracellular matrix at specialized adhesion zones, known as focal contacts or focal adhesions. However, other transmembrane receptors can also localize there, including one transmembrane proteoglycan, syndecan-4. This heparan sulfate proteoglycan can also link directly to the cytoskeleton through alpha-actinin, and can signal through protein kinase C. In turn, the pathway leads to RhoA and Rho kinases that control actomyosin contractility. Syndecan-4 may, therefore, be a sensor of tension exerted on the matrix. These processes are described here, their significance being potential roles in wound contraction, tumor-stroma interactions, fibrosis and the regulation of motility.

EphB/syndecan-2 signaling in dendritic spine morphogenesis.

We previously reported that the cell surface proteoglycan syndecan-2 can induce dendritic spine formation in hippocampal neurons. We demonstrate here that the EphB2 receptor tyrosine kinase phosphorylates syndecan-2 and that this phosphorylation event is crucial for syndecan-2 clustering and spine formation. Syndecan-2 is tyrosine phosphorylated and forms a complex with EphB2 in mouse brain. Dominant-negative inhibition of endogenous EphB receptor activities blocks clustering of endogenous syndecan-2 and normal spine formation in cultured hippocampal neurons. This is the first evidence that Eph receptors play a physiological role in dendritic spine morphogenesis. Our observations suggest that spine morphogenesis is triggered by the activation of Eph receptors, which causes tyrosine phosphorylation of target molecules, such as syndecan-2, in presumptive spines.

Syndecan 4 heparan sulfate proteoglycan is a selectively enriched and widespread focal adhesion component.

Focal adhesion formation in fibroblasts results from complex transmembrane signaling processes initiated by extracellular matrix molecules. Although a role for integrins with attendant tyrosine kinases has been established, there is evidence that cell surface heparan sulfate proteoglycans (HSPGs) are also involved with an associated role of protein kinase C. The identity of the proteoglycan has remained elusive, but we now report that syndecan 4 (ryudocan/amphiglycan) is present in focal adhesions of a number of cell types. Affinity-purified antibodies raised against a unique portion of the cytoplasmic domain of syndecan 4 core protein recognized an HSPG of similar characteristics to those of syndecan 4. These antibodies stained focal adhesions only after cell permeabilization and recognized differing mammalian species. Syndecan 4 was associated with focal adhesions that contained either beta 1 or beta 3 integrin subunits and those that formed on substrates of fibronectin, laminin, vitronectin, or type I collagen. No focal adhesions were found that were vinculin-containing but lacked syndecan 4. In contrast, syndecan 2, whose cytoplasmic domain is closely homologous to syndecan 4, does not appear to be a focal adhesion component. Thus, syndecan 4 represents a new transmembrane focal adhesion component, probably involved in their assembly.

A synthetic peptide from the COOH-terminal heparin-binding domain of fibronectin promotes focal adhesion formation.

Cell adhesion to extracellular matrix molecules such as fibronectin involves complex transmembrane signaling processes. Attachment and spreading of primary fibroblasts can be promoted by interactions of cell surface integrins with RGD-containing fragments of fibronectin, but the further process of focal adhesion and stress fiber formation requires additional interactions. Heparin-binding fragments of fibronectin can provide this signal. The COOH-terminal heparin-binding domain of fibronectin contains five separate heparin-binding amino acid sequences. We show here that all five sequences, as synthetic peptides coupled to ovalbumin, can support cell attachment. Only three of these sequences can promote focal adhesion formation when presented as multicopy complexes, and only one of these (WQPPRARI) retains this activity as free peptide. The major activity of this peptide resides in the sequence PRARI. The biological response to this peptide and to the COOH-terminal fragment may be mediated through cell surface heparan sulfate proteoglycans because treatment of cells with heparinase II and III, or competition with heparin, reduces the response. Treatment with chondroitinase ABC or competition with chondroitin sulfate does not.

Changes in microfilament and focal adhesion distribution with loss of androgen responsiveness in cultured mammary tumor cells.

The actin-containing microfilaments, microtubules, and fibronectin expression of Shionogi 115 mouse mammary tumor cells were visualized by indirect immunofluorescence microscopy. Also studied was the focal adhesion distribution as revealed by interference reflection microscopy and the ability of the cells to grow in suspension culture. All these parameters were documented for androgen-responsive and -unresponsive cells grown in culture, as well as the transition of androgen-responsive to -unresponsive cells when deprived of androgen. The androgen-unresponsive cells had extensive and prominent microfilament bundles together with focal adhesions on the lower cell surface and also showed strict anchorage dependence for growth. In contrast, microfilament bundles and focal adhesions were absent from androgen-responsive cells, which furthermore had the ability to grow in suspension culture. Differences were also apparent in fibronectin expression, the androgen-unresponsive cells having more of this glycoprotein detectable on their surfaces than the androgen-responsive cells. The androgen-responsive and -unresponsive cells had similar microtubule arrays, however. During the transition from the androgen-responsive to the androgen-unresponsive phenotype, the androgen-responsive cells gradually took on the characteristics of androgen-unresponsive cells as judged by cellular morphology or the presence of focal adhesions and microfilament bundles. At intermediate stages in this process, characteristics of both cell types were visible in the cell populations. However, at the stage where all androgen-responsive characteristics were lost, the cells were no longer androgen sensitive. The loss of androgen responsiveness in Shionogi 115 mouse mammary tumor cells is correlated with changes at the cell membrane and the microfilament component of the cytoskeleton.

Syndecans and cell adhesion.

Now that transmembrane signaling through primary cell-matrix receptors, integrins, is being elucidated, attention is turning to how integrin-ligand interactions can be modulated. Syndecans are transmembrane proteoglycans implicated as coreceptors in a variety of physiological processes, including cell adhesion, migration, response to growth factors, development, and tumorigenesis. This review will describe this family of proteoglycans in terms of their structures and functions and their signaling in conjunction with integrins, and indicate areas for future research.

Organelle-cytoskeleton relationships in fibroblasts: mitochondria, Golgi apparatus, and endoplasmic reticulum in phases of movement and growth.

When fibroblasts first emerge from explants of embryonic chick heart, most mitochondria are clustered tightly around the nucleus, with very few extending towards the leading lamella. Although cytoplasmic microtubules are well displayed, mitochondria do not obviously codistribute with them. As the fibroblasts cease locomotion and adapt to growth, however, the mitochondria become dispersed through the cytoplasm and clearly codistributed with the microtubules and centrioles but not with microfilament bundles or 10 nm filaments. This rearrangement occurs in concert with the other changes we have shown previously - including the development of pronounced microfilament bundles and of stable and well-defined focal adhesions - and appears to be related to changes in the motility status of the cells rather than to alterations in growth or synthetic capability. Mitochondrial mobility is strongly reduced by the actions of both colchicine and dihydrocytochalasin B showing that orientation and translocation depend on a co-ordinate interaction of microtubules and microfilamentous meshwork around the centrioles as origin. The Golgi apparatus and endoplasmic reticulum do not rearrange dramatically during the phenotypic conversion, although the locomotory cells are characterized by a zone close to the leading lamella, which is completely free from these organelles. The form and distribution of the Golgi apparatus, but not the endoplasmic reticulum, was sensitive to microtubule disruption but was also shown to have direct functional associations with the centriolar region. The relative distributions of the three types of organelle during the phases of cell movements and cell growth, are consistent with their biochemical functions in cellular activity.

Coupling of cytoskeleton functions for fibroblast locomotion.

Using a chick cell phenotype specialised for locomotion with morphometric measurements made possible by modern instrumentation technology, we have reinvestigated motile functions in fibroblast locomotion. Quantitative analysis of rapid fluctuations in cell form and organelle distribution during locomotion showed many significant correlations between different parts of the cell despite much irregularity in individual displacements over the time scale of the order of one second. These broke down when external perturbations caused changes in shape or direction. Partial energy deprivation caused the cells to lose control of shape and organelle distribution even though forward protrusion continued unaffected. Cytoplasmic displacements shown by marker mitochondria correlated with adjacent fluctuations at the leading edge, and drug treatments which increased the amplitude of mitochondrial movements caused visible protrusions in projected positions at the leading edge. We conclude that fibroblast locomotion may be driven coordinately by a common set of motility mechanisms and that this coordination may be lost as a result of physical or pharmacological disturbance. Taking our evidence with results from other Laboratories, we propose the following cytoskeleton functions. (i) Protrusive activity, probably based on solation--gelation cycles of the actin based cytoskeleton and membrane recycling which provides cellular and membrane components for streaming through the cell body to the leading edge; this is Ca++ sensitive but relatively energy insensitive. (ii) Constraining activity on the cell membrane and on certain organelles to maintain shape and so facilitate directionality and the drawing along of the trailing body; this is Ca++ insensitive but relatively energy sensitive. (iii) Channeling function of microtubules to direct the flow towards multiple foci on the leading edge, and so determine cell polarity. Such a mechanism of locomotion for fibroblasts has many features consistent with evidence for other cell types, especially amoebae and leukocytes.

Fibronectin in cultured rat keratinocytes: distribution, synthesis, and relationship to cytoskeletal proteins.

The aim of this study was to investigate whether epidermal cells can synthesise fibronectin and whether the distribution of this glycoprotein is related to the adhesion and cytoskeletal organisation of these cells. The production of fibronectin by newborn rat epidermal cells was shown by indirect immunofluorescence staining of cultures grown in the absence of a feeder layer using an antiserum which had been cross-adsorbed with foetal calf serum proteins to remove antibodies which recognised serum fibronectin. The distribution of fibronectin in areas of cell-cell and cell-substratum contact, characteristically in the form of short radial stitches, was examined in more detail using immunoelectron microscopy with colloidal gold as marker. This showed the close proximity of fibronectin to the cell membrane, with the ventral surface and fine cellular processes showing the heaviest labelling, and also revealed evidence of a relationship between external fibronectin and internal structure in epidermal cells. Immunofluorescence showed that tonofilaments (keratin) and microtubules were present as fibrillar arrays but were not related to fibronectin distribution. Vimentin and desmin were absent. Actin was distributed as a circumferential bundle of filaments, with finer stands running radially to the edge. The latter were reminiscent of the radial fibronectin stitches and a spatial correspondence between fibronectin and actin was confirmed by double-label immunofluorescence which revealed many instances of overlap and colinearity of actin and fibronectin filaments. The ability of keratinocytes to produce fibronectin suggests that these cells can contribute to the formation of the basement membrane in skin. The localisation of fibronectin and its close association with actin also suggests that it is involved in keratinocyte adhesion and is related to the internal organisation of these cells.

Rat hair follicle dermal papillae have an extracellular matrix containing basement membrane components.

Dermal papillae are small mesenchymally derived zones at the bases of hair follicles which have an important role in hair morphogenesis in the embryo and control of the hair growth cycle in postnatal mammals. The cells of the papilla are enmeshed in a dense extracellular matrix which undergoes extensive changes in concert with the hair cycle. Here it is shown that this matrix in anagen pelage follicles of postnatal rats contains an abundance of basement membrane components rather than dermal components such as interstitial collagens. In particular, type IV collagen, laminin, and basement membrane types of chondroitin sulfate and heparan sulfate proteoglycans are present as matrix components. Dermal papilla cells, when initially isolated from adult rat vibrissae and cultured in vitro, retained the potential to synthesize this spectrum of matrix components, but this was gradually lost, to be replaced by synthesis of other components including type I and III collagens. It seems likely therefore that the dermal papilla cells in vivo synthesize a basement membrane type of extracellular matrix, although a contribution from epithelial, and in some cases capillary endothelial, cells cannot be ruled out.