The role of phosphorylation and limited proteolytic cleavage of talin and vinculin in the disruption of focal adhesion integrity

SPARC mediates focal adhesion disassembly in endothelial cells through a follistatin-like region and the Ca(2+)-binding EF-hand

SPARC is a one of a group of extracellular matrix proteins that regulate cell adhesion through a loss of focal adhesion plaques from spread cells. We previously reported that SPARC reduced the number of bovine aortic endothelial (BAE) cells positive for focal adhesions [Murphy-Ullrich et al. (1991): J Cell Biol 115:1127-1136]. We have now characterized the effect of SPARC on the cytoskeleton of BAE cells. Addition of SPARC to spread BAE cells caused a dose-dependent loss of focal adhesion-positive cells, that was maximal at approximately 1 microgram/ml (0.03 microM). Consistent with the loss of adhesion plaques as detected by interference reflection microscopy, vinculin appeared diffuse and F-actin was redistributed to the periphery of cells incubated with SPARC. However, the distribution of the integrin alpha v beta 3 remained clustered in a plaque-like distribution. These data, and the observation that SPARC binds to BAE cells but not to the extracellular matrix, indicate that SPARC acts via interactions with cell surface molecules and not by steric/physical disruption of integrin-extracellular matrix ligands. To determine the region(s) of SPARC that mediate a loss of focal adhesions, we tested peptides from the four distinct regions of SPARC. The cationic, cysteine-rich peptide 2.1 (amino acids 54-73) and the Ca(2+)-binding EF-hand-containing peptide 4.2 (amino acids 254-273) were active in focal adhesion disassembly. Furthermore, antibodies specific for these regions neutralized the focal adhesion-labilizing activity of SPARC. These results are consistent with previous data showing that peptide 2.1 and 4.2 interact with BAE cell surface proteins and indicate that the loss of focal adhesions from endothelial cells exposed to SPARC is a receptor-mediated event.

Effects of microcystin-LR on actin and the actin-associated proteins alpha-actinin and talin in hepatocytes

Microcystin-LR (MCLR) is a commonly encountered blue-green algal hepatotoxin and a known inhibitor of cellular protein phosphatase types 1 and 2A. The toxin causes alterations in, and redistribution of, intermediate filaments, microtubules, and actin microfilaments (MFs) in affected cells. In this study, the effect of MCLR on the sequence of alterations in MFs and actin-associated proteins (AAPs) of isolated hepatocytes was examined in an effort to determine whether morphologic changes induced in MFs by microcystins are a result of prior dislocation of AAPs. We studied the effects of MCLR exposure on alpha-actinin and talin, two AAPs that play a role in the orientation of the MFs. Primary hepatocytes were incubated with 10 microns MCLR for 0-64 min. The distribution of actin, alpha-actinin, and talin were examined using fluorescence microscopy. MCLR induced similar changes in the distribution of actin and the AAPs. Actin filament redistribution was first observed after 12 min of MCLR exposure, and was characterized by detachment of MFs from the cell periphery, followed by condensation at distinct focal points and progressive collapse into the interior of affected cells. Changes in alpha-actinin and talin distribution were first observed after 20 min of toxin exposure. The AAPs appeared to detach from focal contacts on the cytoplasmic surface of the plasma membrane, condense into cytoplasmic aggregates, and ultimately collapse into a juxtanuclear bundle. The results of this study indicate that, in hepatocytes exposed to MCLR, the collapse of actin MFs occurs prior to the dislocation of alpha-actinin and talin. Changes in these actin associated proteins are not likely to account for the initial changes in actin MFs.

Disruption of cell adhesion in renal epithelium without cadherin loss

Loss of cell adhesion is a critical event in the development of tumour invasiveness and metastases. Although loss of cadherin expression has been demonstrated to be associated with increased invasiveness and metastatic potential in some tumours, others, including renal carcinoma, show no such correlation. The aim of this study was to demonstrate that cell adhesion could be lost in phorbol ester-treated renal epithelial cells and renal tumour cells without loss of A-CAM expression. The model used has been shown previously to mimic changes that occur in the progression of renal carcinoma. We found that A-CAM expression persists on the lateral surfaces of phorbol ester-treated cells even though these cells lose cell-cell adhesion. Similar findings were seen in renal carcinoma cells in culture. We conclude that loss of cell adhesion between tumour cells may other either by loss of cadherins or as a result of loss cadherin function occurring as a consequence of cell transformation.

Platelet talin is phosphorylated by calyculin A

Calyculin A and okadaic acid, potent and cell permeable inhibitors of type 1 and type 2A protein phosphatases, inhibit platelet aggregation and secretion. However, the relationship between phosphatase inhibition and inhibition of platelet function is not well understood. We found that in unstimulated platelets, talin (P235) was phosphorylated at threonine residues by calyculin A. Furthermore, the extent of talin phosphorylation by calyculin A was closely correlated with its inhibition of thrombin-induced platelet aggregation. Since the binding of talin to platelet glycoprotein IIb/IIIa complex has been shown to be affected by its phosphorylation, these results suggest that type 1 and/or type 2A protein phosphatases may play a role in the regulation of membrane-cytoskeleton interaction through dephosphorylation of talin.

Stimulation of signal transduction pathways in osteoblasts by mechanical strain potentiated by parathyroid hormone

Second-messenger systems have been implicated to transmit mechanical stimulation into cellular signals; however, there is no information on how mechanical stimulation is affected by such systemic factors as parathyroid hormone (PTH). Regulation of adenylyl cyclase and phosphatidylinositol pathways in rat dentoalveolar bone cells by mechanical strain and PTH was investigated. Two different cell populations were isolated after sequential enzyme digestions from dentoalveolar bone (group I and group II) to study potential differences in response. Mechanical strain was applied with 20 kPa of vacuum intermittently at 0.05 Hz for periods of 0.5, 1, 5, 10, and 30 minutes and 1, 3, and 7 days using the Flexercell system. Levels of cAMP, measured by RIA, and levels of inositol 1,4,5-triphosphate (IP3) and protein kinase C activity (PKC), measured by assay systems, increased with mechanical strain. When PTH was added to the cells, there was a significant increase in levels of all the intracellular signals, which appeared to potentiate the response to mechanical strain. IP3 levels (0.5 minute) peaked before those of PKC activity (5 minutes), which in turn peaked before those of cAMP (10 minutes). Group II cells showed higher levels of cAMP and IP3 than the group I cells. This suggests that the former may ultimately play the predominant roles in skeletal remodeling in response to strain. Immunolocalization of the cytoskeleton proteins vimentin and alpha-actinin, focal contact protein vinculin, and PKC showed a marked difference between strained and nonstrained cells. However, the addition of PTH did not cause any significant effect in cytoskeleton reorganization. Staining of PKC and vimentin, alpha-actinin, and vinculin suggests that PKC participates actively in the transduction of mechanical signals to the cell through focal adhesions and the cytoskeleton, although only PKC seemed to change with short time periods of strain. In conclusion, dentoalveolar osteoblasts responded to mechanical strain initially through increases in levels of IP3, PKC activity, and later cAMP, and this response was potentiated when PTH was applied together with mechanical strain.

Talin distribution and phosphorylation in thrombin-activated platelets

We have previously demonstrated that the subcellular distribution of the adhesion plaque protein, talin, changes dramatically in human platelets in response to platelet activation (Beckerle et al., J. Cell Biol. 109, 3333–3346, 1989). Talin is uniformly distributed throughout the cytoplasm of resting platelets. However, when platelets are stimulated to become activated and adhesive, a significant amount of the talin population rapidly redistributes to a peripheral, submembranous location. In the present study we have examined talin phosphorylation and proteolytic cleavage as possible mechanisms by which talin's subcellular distribution could be regulated in platelets. We have found that thrombin activation of platelets leads to a fourfold increase in talin phosphorylation. Proteolytic cleavage of talin, however, is not detected in washed platelets activated with thrombin for as long as 30 minutes. Because talin moves to a submembranous location upon platelet activation and has been shown to interact with integrins in vitro, we also investigated whether the major platelet integrin, GPIIb-IIIa, is required for talin redistribution. Using Glanzmann thrombasthenic platelets, which are deficient in GPIIb-IIIa, we found that talin redistribution occurs even in the absence of GPIIb-IIIa. Collectively, our studies suggest that neither proteolytic cleavage of talin nor interactions between talin and GPIIb-IIIa is required for the regulated redistribution of talin in thrombin-activated platelets. Phosphorylation of talin in response to thrombin activation may, however, be one mechanism utilized by platelets to regulate talin distribution and function in human platelets.

Immunoblotting of contractile and cytoskeletal proteins of canine basilar artery in vasospasm

Vasospasm was produced in the canine basilar arteries by a two-hemorrhage method, and voltage- and receptor-dependent contractions of the normal canine basilar arteries were induced by local applications of potassium chloride (KCI) and serotonin, respectively, after transclival exposure. Actin, myosin, desmin, filamin, talin, vinculin, and alpha-actinin in the basilar artery were studied by immunoblotting. The immunoblots showed a decrease or loss in immunoreactivity of some native proteins and generation of protein fragments, smaller in size than native proteins, in spastic, KCI, and serotonin groups, indicating a proteolytic degradation. In the spastic group on Day 2, actin, desmin, and filamin were usually degraded slightly; myosin moderately; and talin and alpha-actinin substantially. Vinculin and metavinculin remained intact. In the spastic group on Day 7, actin and desmin were usually decomposed slightly; myosin, filamin, and vinculin substantially; and talin, metavinculin, and alpha-actinin markedly. In the KCI and serotonin groups, slight degradation was usually observed in filamin, often in alpha-actinin, and occasionally in actin, whereas desmin, vinculin, and metavinculin were not degraded. In addition, myosin was usually degraded moderately in the KCI group and slightly in the serotonin group, and talin was generally decomposed slightly in the KCI group and moderately in the serotonin group. The degraded fragments, although variable in number and immunoreactivity, were similar in size in the three groups. We suggest that the intracellular devices responsible for contraction of the basilar arteries are degraded more severely in the spastic group than in the KCI or serotonin group, probably by similar proteolytic mechanism and progressively with the passage of time after subarachnoid hemorrhage in vasospasm.

Adhesion of CHO cells to fibronectin is mediated by functionally and structurally distinct adhesion plaques

We have investigated the dynamics between free fibronectin receptors and clusters of them organized into adhesion plaques on CHO cells using the ability of these free integrins to be endocytosed and recycled to the plasma membrane. Indirect inhibition of the endocytic cycle by monensin resulted in the subsequent internalization of free receptors, which we followed by indirect immunostaining and confocal microscopy. Consequently, all the adhesive structures that were in equilibrium with free integrins became progressively disorganized. The cellular morphological changes were analyzed and correlated with the distribution of cell-substratum contacts viewed by confocal images obtained after immunostaining with antibodies raised against the fibronectin receptor, talin, vinculin and actin. After cell adhesion to fibronectin, blockage of the endocytic cycle induced disruption of the adhesion plaques that were mainly localized at the cell periphery, and disappearance of the stress fibers. However, the cells remained firmly attached to the substratum through focal contacts localized in the central part of the cell. These central focal contacts, but not the peripheral adhesion plaques, could form when the vesicular traffic was blocked prior to adhesion and they allowed the cells to attach and flatten onto the substratum. Whereas both adhesive structures contained the same receptors linked to talin and vinculin, the central adhesive structures were attached to a short stretch of actin but never permitted the organization of stress fibers.

Signal transduction pathways and cellular intoxication with Clostridium difficile toxins

In cultured cells the cytopathic effects (CPE) of Clostridium difficile toxins A and B are superficially similar. The irreversible CPEs involve a reorganization of the cytoskeleton, but the molecular details of the mechanism(s) of action are unknown. As part of the work to elucidate the events leading to the CPE, cultured cells were preincubated with agents known to either stimulate or inhibit some major signal transduction pathways, whereupon toxin was added and the development of the CPE was followed. Both toxin-induced CPEs were enhanced by phorbol esters and mezerein, which stimulate protein kinase C, while they were inhibited by the phospholipase A2 inhibitors quinacrine and 4-bromophenacylbromide. Agents affecting certain G-proteins, cGMP and cAMP levels, phosphatases, prostacyclin, lipoxygenase, and phospholipase C did not affect the development of the CPE of either toxin. Thus, the cytoskeletal effect induced by toxins A or B appears to require PLA2 activity and involves at least part of a protein kinase C-dependent pathway, but not pertussis toxin-sensitive G-proteins, cyclic nucleotides, eicosanoid metabolites, or phospholipase C activity. In addition, both toxins were shown to activate phospholipase A2.

Fimbrin localized to an insoluble cytoskeletal fraction is constitutively phosphorylated on its headpiece domain in adherent macrophages

The actin-bundling protein fimbrin is homologous to 1-plastin, a 65kD phosphoprotein expressed in leukocytes and transformed cells [de Arruda et al., J. Cell Biol. 111, 1069-1080]. Because fimbrin is present in cell adhesion sites, we studied the phosphorylation state of fimbrin and its distribution in macrophages sequentially extracted with Triton-X-100 (soluble fraction), Tween 40-deoxy-cholate (cytoskeletal fraction), and SDS (insoluble cytoskeletal fraction). The approximate distribution of fimbrin and actin among these fractions was found to be: 65% fimbrin/55% actin in the soluble fraction, 30% fimbrin/20% actin in the cytoskeletal fraction, and 5% fimbrin/25% actin in the insoluble cytoskeletal fraction. PMA did not alter this distribution. Fluorescence microscopy of acetone-extracted macrophages showed that actin is concentrated in podosomes at the substratum interface and is diffusely distributed throughout the remainder of the cell. Fimbrin colocalizes with actin in podosomes and also exhibits a punctate distribution in the cytoplasm that overlaps with actin. In Tween 40/DOC-extracted cells, podosomes remain, and fimbrin also exhibits a punctate distribution along actin filaments. Metabolic 32PO4 labeling revealed that fimbrin is constitutively phosphorylated and that phosphorylated fimbrin is concentrated in the insoluble cytoskeletal fraction. PMA increased the relative levels of fimbrin phosphorylation twofold but did not alter the pattern of fimbrin fluorescence or the distribution of phosphorylated fimbrin. Limited trypsin digestion and phosphoamino acid analysis demonstrated that phosphorylation occurs specifically on serine residues within the 10kD headpiece domain of fimbrin. Phosphorylation of the headpiece domain could regulate the actin binding and bundling properties of fimbrin, or it could regulate the interaction of fimbrin with other proteins.

Integrins and their ligands

Integrins are expressed on almost every cell type and are responsible for the linkage of the extracellular matrix with the cytoskeleton. In this review I have focused on the intra- and extracellular proteins that bind to integrins. Although many integrins bind to the same extracellular ligand, they mostly recognize different sites on these ligands. Some integrins interact with the same site but then there are requirements for different additional sequences to obtain high affinity. By modulating the expression and activity of integrins in the plasma membrane, cells can adapt their capacity of binding to the matrix. How integrins become activated is as yet not clear, but interaction with other proteins or lipids may be critical. Binding to ligands could also be modulated by alternative splicing of mRNAs for ligand binding sites in the extracellular domain. In Drosophila, the mRNA for the extracellular domain of the PS2 integrin is spliced near a site implicated in ligand binding. In humans, however, there are no indications that alternative splicing contributes to the regulation of function of the extracellular domain of integrins. The only splice variant of the extracellular domain of an integrin identified so far concerns are alpha subunit of the alpha IIb beta 3 complex, but the splicing occurs in a region that has not been implicated in cell adhesion. There is also no evidence as yet that integrin function can be modulated by alternative splicing of mRNA for the cytoplasmic domain of integrin subunits. However, the loss of function seen with some deletion mutants of the cytoplasmic domains of integrin subunits suggests that such a mechanism may well exist. In a different way the binding capacity of a given cell can be influenced by regulating the expression of its ligand or by alternative mRNA splicing of sequences encoding the cell binding domain in their ligands. In the case of fibronectin, the mRNA for one of the integrin binding sites is subject to alternative splicing. The mRNAs for the three chains of laminin appear not to be subject to alternative splicing but, by combining different variant chains of laminin, isoforms can be generated which may have different affinities for integrins. Binding of cells to the matrix therefore does not only depend on the expression and activity of the correct integrin but also of the correct variant of the ligand.(ABSTRACT TRUNCATED AT 400 WORDS)

Cellular interactions with the extracellular matrix are coupled to diverse transmembrane signaling pathways

Extracellular matrix (ECM) glycoproteins such as laminin, fibronectin, or collagen IV play a major role in cell behavior regulation. The molecular mechanisms taking place at the interface between the ECM and the cell surface are now rather well defined; however, very little is known about intracellular signals induced by these interactions. In order to get insights into the transduction pathways involved in cell-ECM interactions we have investigated the effects of several intracellular kinase inhibitors. Calmodulin-dependent kinase inhibitors, W-7 and sphingosine, have negative effects on cell-matrix interactions. They inhibit adhesion of several cell lines to laminin (IC50 = 4-10 microM), fibronectin and collagen IV (IC50 = 7-25 microM). The effects are immediate, reversible, and also cell specific, certain combinations of cell line-substrate being irresponsive to these inhibitors. In contrast, two inhibitors, H-7 and staurosporine, for which protein kinase C is a common target, increase two- to fourfold the attachment of HT1080, OVCAR-4, and B16F10 cells to laminin but not to fibronectin. Another inhibitor, HA-1004, known to inhibit protein kinase A at low concentrations, has an activating effect only at high concentration (> 200 microM) when it becomes an inhibitor of protein kinase C. These inhibitors are without effect on RuGli and Saos-2 cell adhesion on the three substrates. Altogether these results suggest that calmodulin-dependent kinases and protein kinase C could be separately involved in ECM-induced cellular responses. However, the effects of kinase inhibitors are substrate-specific and cell type-specific, suggesting that the intracellular signals induced by the extracellular matrix vary with the nature of integrin involved in signal transmission.

Phosphorylation of vinculin in human platelets spreading on a solid surface

Vinculin is a cytoskeletal protein believed to be involved in linking microfilaments to the cell membrane. It is a substrate for the Ca(2+)- and phospholipid-dependent protein kinase C. We show here that when human platelets attach and spread on a solid surface, the alpha isoforms of vinculin become phosphorylated at serine and/or threonine residues. Phosphorylation is dependent on adhesion to a surface, since suspended, unattached platelets can produce filopodia but no phosphorylation of vinculin. Phosphorylation is also dependent on actin polymerization, as it does not occur when platelets had been pretreated with cytochalasin B. Most likely, protein kinase C is responsible for the phosphorylation of vinculin, since phosphorylation also occurs when platelets are treated with a phorbol ester, which activates protein kinase C, and is blocked by treatment with a staurosporine derivative which inhibits this enzyme. These results suggest that phosphorylation plays a role in anchoring vinculin at sites of microfilament-membrane interaction.

Further characterisation of the talin-binding site in the cytoskeletal protein vinculin

The cytoskeletal protein vinculin is a component of adherens-type junctions where it is one of a number of interacting proteins thought to link the cytoplasmic domain of adhesion receptors to F-actin. Vinculin has been shown to bind to at least three other cytoskeletal proteins, talin, paxillin and alpha-actinin. In this study, we further characterise the talin-binding domain in vinculin using a series of chick vinculin polypeptides expressed as glutathione-S-transferase fusion proteins in Escherichia coli. Thus 125I-talin bound to a fusion protein spanning residues 1–398, but not to those spanning residues 399–881 or 881–1066 in an SDS-PAGE gel-blot assay. We have previously characterised two chick vinculin cDNAs (2.89 kb cDNA and cVin5) which are identical in the region of overlap except that cVin5 lacks coding sequence for residues 167–207. Interestingly, a fusion protein spanning residues 1–398, but lacking residues 167–207, was unable to bind talin. However, further analysis showed that residues 167–207 are insufficient to support binding, and deletion of as few as 31 N-terminal residues abolished binding activity. The results of the gel-blot assay were essentially confirmed using purified fusion proteins adsorbed to glutathione-agarose beads. The smallest vinculin fusion protein able to bind talin contained residues 1–258. This fusion protein was as effective as whole vinculin in inhibiting the binding of 125I-vinculin to talin-coated microtitre wells. Interestingly, mutations which altered the charge characteristics of the highly conserved residues 178 and 181 abolished binding, whereas conservative substitutions were without effect. However, such mutations did not abolish the ability of mutant polypeptides spanning residues 1–398 to target to cell-matrix junctions in Cos cells. We have investigated the possible origin of the cDNA clone cVin5 by defining the structure of a 5′ portion of the chicken vinculin gene, and by analysing vinculin transcripts in a variety of adult tissues and embryonic fibroblasts using reverse transcriptase and polymerase chain reaction. Although residues 167–207 are encoded on a separate exon, we have been unable to identify a tissue where this exon is alternatively spliced.

Changes in the cytoskeleton of 3T3 fibroblasts induced by the phosphatase inhibitor, calyculin-A

Addition of the protein phosphatase inhibitor, calyculin-A, to 3T3 fibroblasts causes a marked change in cell morphology. Initially the cells become rounded, develop surface blebs and then detach from the substratum. In the detached cells an unusual ball-like structure is observed. This study focuses on the cytoskeleton during these calyculin-A-induced morphological changes. Stress fibres disappear as the cells begin to round and aggregates of actin are formed towards the apical surface of the cell. These aggregates condense, in the detached cells, to form the ball structure of approximately 3 microns diameter. Between the ball and the nucleus are cables of intermediate filaments that appear to be attached to the surface of the ball and to the nuclear lamina. Using a procedure designed for the isolation of nuclei the nucleus-ball complex can be obtained. Analysis of the nucleus-ball preparation by immunofluorescence and electron microscopy demonstrate that the ball contains actin and that intermediate filaments are located between the ball and the nucleus. In this preparation, the intermediate filaments also appear to attach to the surfaces of the ball and the nucleus. Electrophoretic analysis of the nucleus-ball preparation indicates that, in addition to actin, a major component of the ball is myosin. It is suggested that the formation of the ball is caused by an actin-myosin-based contractile process, initiated by the phosphorylation of myosin. The aggregation of the actomyosin draws together the intermediate filaments into the area between the ball and nucleus. This hypothesis requires that vimentin binds both to the nucleus and to some component of the ball.

A tumour promoter induces alterations in vinculin and actin distribution in human renal epithelium

Loss of cell-substratum adhesion is an important factor during tumour progression. We have previously described reduced focal contact components and poorly organized cytoskeletal actin in renal cell carcinomas. In this study, we have used the potent tumour promoter TPA on cultured human renal epithelium to mimic neoplastic transformation. The morphological changes induced by TPA were examined by phase contrast and fluorescence microscopy. TPA treatment caused rounding up of cells and loss of adhesion to either fibronectin or laminin substrata. Cytoskeletal actin was redistributed from orientated stress fibre bundles to a perinuclear circumferential arrangement. This was accompanied by a progressive reduction in the number of vinculin-containing contacts with accumulation of vinculin in punctate spots in the perinuclear region. These altered membrane-cytoskeletal interactions induced by TPA are entirely reversible and mimic epigenetic changes which occur during tumour progression.

Sphingosine inhibits attachment of murine Lewis lung carcinoma cells to laminin and type IV collagen

The effect of sphingosine (SPH) on the adhesive properties of Lewis lung carcinoma (3LL) cells was investigated using plastic precoated with the extracellular matrix proteins, laminin, fibronectin, or type IV collagen. Treatment of 3LL cells with SPH (0.5-10 microM) resulted in a dose-dependent decrease in the ability to bind to laminin and type IV collagen but had little or no effect on attachment to fibronectin. Phorbol 12-myristate 13-acetate (PMA) selectively enhanced attachment of 3LL cells to laminin and collagen. The inhibitory effect of SPH on attachment to both proteins was competitively antagonized by PMA. These results suggest that SPH acts as a negative effector for cell attachment to laminin and collagen, and that the cell attachment process to both proteins might be regulated in part by protein kinase C.

Direct interactions between talin and actin

Talin was purified from chicken gizzard by a modification of the method of L. Molony et al. [J. Biol. Chem.(1987) 262, 7790-7795]. Unlike the talin purified by the previous method, the talin purified by the new method was found to bind to both F- and G-actin: Talin cosedimented with F-actin. On gel filtration of a mixture of talin and G-actin, a complex of talin and action was obtained. Talin stimulated the polymerization rate of G-actin. A major proteolytic fragment of talin that retained the binding ability to F-actin was also identified. These results indicate that talin can bind directly to actin and suggest that talin plays a key role in the organization of actin filaments at the actin-membrane attachment sites in vivo also.

Vinculin

Vinculin is clearly a key element in the transmembrane assemblages that link cells to each other or to the substrate. However, despite all the studies that have been done on the protein, we still do not know its function within these assemblages. The bulk of the biochemical and cell biological evidence suggests that, in some unknown way, its presence in the junctions may be involved in the stable association of actin with the membrane, yet vinculin by itself does not appear to interact with actin. In the future, identification of additional junctional molecules that interconnect actin and vinculin may resolve this dilemma. Alternatively, studies with vinculin that is phosphorylated or acylated may yield clues to its function. Perhaps the complexity of the protein composition of microfilament-containing junctions suggests that protein assemblages rather than individual proteins provide novel functions. As new proteins belonging to these junctions are discovered, it will be important to assess their interaction with already known components such as vinculin and to ask if the protein combination has a particular function.