The removal of extracellular fibronectin from areas of cell-substrate contact

Immunofluorescent labeling for fibronectin was largely excluded from sites of closest contact between spreading chicken gizzard fibroblasts and the substratum. This was observed by double immunofluorescent labeling of fixed cells for fibronectin and vinculin, a smooth muscle intracellular protein that is specifically associated with focal adhesion plaques, in conjunction with interference-reflection microscopy. When the cells were plated on a fibronectin-coated substratum they adhered to its surface and rapidly spread on it. The immunofluorescent labeling for fibronectin in those cultures (after fixation and triton permeabilization) was usually absent from the newly formed, vinculin-containing focal adhesion plaques. We have found, however, that the accessibility to the cell-substrate gap at the focal adhesion plaques is limited and therefore a more direct approach was adopted. We have found that cells spreading on a substrate coated with rhodamine-labeled fibronectin progressively removed the underlying protein from the substrate. The removal of fibronectin involved at least two distinct mechanisms. Part of the substrate-associated fibronectin was removed from small areas and displaced toward the cell center. The arrowhead-shaped areas from which fibronectin was removed often coincided with vinculin-rich focal contacts. We observed, however, many areas where focal contacts were found over unperturbed fibronectin carpet, as well as fibronectin-free areas with no overlapping focal contacts. The possibilities that fibronectin is actively displaced from areas of cell-substrate contact, that the focal adhesion plaques are transiently associated with these areas and their implications on the dynamics of cell spreading and locomotion are discussed. The second route of fibronectin removal from the substrate was endocytosis. The rhodamine-labeled fibronectin was found in the cells in a partial or transient association with clathrin-containing structures.     

Cell surface distribution of fibronectin and vitronectin receptors depends on substrate composition and extracellular matrix accumulation

We used antibodies against the alpha subunits of the human fibronectin receptor (FNR) and vitronectin receptor (VNR) to localize simultaneously FNR and VNR at major substrate adhesion sites of fibroblasts and melanoma cells with double-label immunofluorescence microscopy. In early (2-6-h) serum-containing cultures, both FNR and VNR coaccumulated in focal contacts detected by interference reflection microscopy. Under higher resolution immunoscanning electron microscopy, FNR and VNR were also observed to be distributed randomly on the dorsal cell surface. As fibronectin-containing extracellular matrix fibers accumulated beneath the cells at 24 h, FNR became concentrated at contacts with these fibers and was no longer detected at focal contacts. VNR was not observed at matrix contacts but remained strikingly localized in focal contacts of the 24-h cells. Since focal contacts represent the sites of strongest cell-to-substrate adhesion, these results suggest that FNR and VNR together play critical roles in the maintenance of stable contacts between the cell and its substrate. In addition, the accumulation of FNR at extracellular matrix contacts implies that this receptor might also function in the process of cellular migration along fibronectin-containing matrix cables. To define the factors governing accumulation of FNR and VNR at focal contacts, fibroblasts in serum-free media were plated on substrates coated with purified ligands. Fibronectin-coated surfaces fostered accumulation of FNR but not VNR at focal contacts. On vitronectin-coated surfaces, or substrata derivatized with a tridecapeptide containing the cell attachment sequence Arg-Gly-Asp, both FNR and VNR became concentrated at focal contacts. These observations suggest that the availability of ligand is critical to the accumulation of FNR and VNR at focal contacts, and that FNR might also recognize substrate-bound vitronectin.     

Localization of urokinase type plasminogen activator to focal adhesions requires ligation of vitronectin integrin receptors

Previous studies have shown that the adhesion protein, vitronectin, directs the localization of urokinase-type plasminogen activator (uPA) to areas of cell-substrate adhesion, where uPA is thought to regulate cell migration as well as pericellular proteolysis. In the present study, HT-1080 cell lines expressing either wild-type vitronectin or vitronectin containing a single amino-acid substitution in the integrin binding domain were used to assess whether ligation of the alphavbeta5 integrin was required for uPA localization to focal adhesions. The synthesis of wild-type vitronectin by HT-1080 cells adherent to either collagen or fibronectin resulted in the redistribution of both the alphavbeta5 integrin as well as uPA to focal adhesion structures. In contrast, cells synthesizing mutant vitronectin, containing the amino-acid substitution in the integrin binding domain, were unable to direct the redistribution of either alphavbeta5 or uPA to focal adhesions. Recombinant forms of wild-type and mutant vitronectin were prepared in a baculovirus system and compared for their ability to direct the redistribution of vitronectin integrin receptors as well as uPA on human skin fibroblasts. In the absence of vitronectin, fibroblast cells adherent to fibronectin assemble focal adhesions which contain the beta1 integrin but do not contain uPA. Addition of recombinant wild-type, but not mutant, vitronectin to fibroblasts adherent to fibronectin resulted in the redistribution of alphavbeta3, alphavbeta5, and uPA into focal adhesions. However, when cells were plated directly onto antibodies directed against either the alphavbeta3 or alphavbeta5 integrins, uPA was not localized on the cell surface. These data indicate that ligation of vitronectin integrin receptors is necessary but not sufficient for the localization of uPA to areas of cell matrix adhesion, and suggest that vitronectin may promote cell migration by recruiting vitronectin integrin receptors and components of the plasminogen activator system to areas of cell matrix contact.     

Cell-substrate contacts illuminated by total internal reflection fluorescence

A technique for exciting fluorescence exclusively from regions of contact between cultured cells and the substrate is presented. The technique utilizes the evanescent wave of a totally internally reflecting laser beam to excite only those fluorescent molecules within one light wavelength or less of the substrate surface. Demonstrations of this technique are given for two types of cell cultures: rat primary myotubes with acetylcholine receptors labeled by fluorescent alpha- bungarotoxin and human skin fibroblasts labeled by a fluorescent lipid probe. Total internal reflection fluorescence examination of cells appears to have promising applications, including visualization of the membrane and underlying cytoplasmic structures at cell-substrate contacts, dramatic reduction of autofluorescence from debris and thick cells, mapping of membranes topography, and visualization of reversible bound fluorescent ligands at membrane receptors.     

Plasminogen activator inhibitor type I stabilizes vitronectin-dependent adhesions in HT-1080 cells

Polyclonal antibodies against plasminogen activator inhibitor type-I (PAI-1) caused rapid retraction and rounding of substrate-attached HT- 1080 cells. The kinetics and extent of antibody-mediated cell rounding were not dependent on either urokinase or plasmin activity. Cells adherent to vitronectin-coated substrates detached within 2 h of antibody addition. Cells adherent to fibronectin were unaffected by the antibodies. Immunoblotting of substrate-attached material indicated that HT-1080 cells deposited PAI-1 into vitronectin, but not fibronectin, dependent contacts. These data suggest that the antibody- mediated cell rounding resulted from a steric disruption of vitronectin- dependent adhesions, indicating that the binding site on vitronectin for PAI-1 is near, but does not overlap, the binding site for vitronectin receptor. The accumulation of PAI-1 into vitronectin- dependent adhesion sites correlated temporally with the preferential degradation of fibronectin from the substrate. HT-1080 cells adherent to either fibronectin or vitronectin were able to activate exogenous plasminogen to plasmin. Plasmin levels were increased 200% on cells adherent to fibronectin and 100% on cells adherent to vitronectin. In the presence of a neutralizing antibody against PAI-1, vitronectin adherent cells activated plasminogen to the same extent as fibronectin adherent cells. Plasmin levels of 200% above baseline were associated with retraction of cells from the substrate. The ability of vitronectin adherent cells to activate exogenous plasmin was completely blocked in the presence of neutralizing antibodies against urokinase. These data represent the first demonstration that vitronectin-associated PAI-1 regulates urokinase in focal contact areas.     

Localization of Urokinase Type Plasminogen Activator to Focal Adhesions Requires Ligation of Vitronectin Integrin Receptors

Previous studies have shown that the adhesion protein, vitronectin, directs the localization of urokinase-type plasminogen activator (uPA) to areas of cell-substrate adhesion, where uPA is thought to regulate cell migration as well as pericellular proteolysis. In the present study, HT-1080 cell lines expressing either wild-type vitronectin or vitronectin containing a single amino-acid substitution in the integrin binding domain were used to assess whether ligation of the α_vβT₅ integrin was required for uPA localization to focal adhesions. The synthesis of wild-type vitronectin by HT-1080 cells adherent to either collagen or fibronectin resulted in the redistribution of both the α_vβT₅ integrin as well as uPA to focal adhesion structures. In contrast, cells synthesizing mutant vitronectin, containing the amino-acid substitution in the integrin binding domain, were unable to direct the redistribution of either α_vβT₅ or uPA to focal adhesions. Recombinant forms of wild-type and mutant vitronectin were prepared in a baculovirus system and compared for their ability to direct the redistribution of vitronectin integrin receptors as well as uPA on human skin fibroblasts. In the absence of vitronectin, fibroblast cells adherent to fibronectin assemble focal adhesions which contain the βT₁ integrin but do not contain uPA. Addition of recombinant wild-type, but not mutant, vitronectin to fibroblasts adherent to fibronectin resulted in the redistribution of α_vβT₃, α_vβT₅, and uPA into focal adhesions. However, when cells were plated directly onto antibodies directed against either the α_vβT₃ or α_vβT₅ integrins, uPA was not localized on the cell surface. These data indicate that ligation of vitronectin integrin receptors is necessary but not sufficient for the localization of uPA to areas of cell-matrix adhesion, and suggest that vitronectin may promote cell migration by recruiting vitronectin integrin receptors and components of the plasminogen activator system to areas of cell matrix contact.     

A GTPase controls cell-substrate adhesion in Xenopus XTC fibroblasts

Cell-substrate adhesion is crucial at various stages of development and for the maintenance of normal tissues. Little is known about the regulation of these adhesive interactions. To investigate the role of GTPases in the control of cell morphology and cell-substrate adhesion we have injected guanine nucleotide analogs into Xenopus XTC fibroblasts. Injection of GTP gamma S inhibited ruffling and increased spreading, suggesting an increase in adhesion. To further investigate this, we made use of GRGDSP, a peptide which inhibits binding of integrins to vitronectin and fibronectin. XTC fibroblasts injected with non-hydrolyzable analogs of GTP took much more time to round up than mock-injected cells in response to treatment with GRGDSP, while GDP beta S-injected cells rounded up in less time than controls. Injection with GTP gamma S did not inhibit cell rounding induced by trypsin however, showing that cell contractility is not significantly affected by the activation of GTPases. These data provide evidence for the existence of a GTPase which can control cell-substrate adhesion from the cytoplasm. Treatment of XTC fibroblasts with the phorbol ester 12-o-tetradecanoylphorbol-13-acetate reduced cell spreading and accelerated cell rounding in response to GRGDSP, which is essentially opposite to the effect exerted by non-hydrolyzable GTP analogs. These results suggest the existence of at least two distinct pathways controlling cell-substrate adhesion in XTC fibroblasts, one depending on a GTPase and another one involving protein kinase C.     

Myotubes differentiate optimally on substrates with tissue-like stiffness: pathological implications for soft or stiff microenvironments

Contractile myocytes provide a test of the hypothesis that cells sense their mechanical as well as molecular microenvironment, altering expression, organization, and/or morphology accordingly. Here, myoblasts were cultured on collagen strips attached to glass or polymer gels of varied elasticity. Subsequent fusion into myotubes occurs independent of substrate flexibility. However, myosin/actin striations emerge later only on gels with stiffness typical of normal muscle (passive Young's modulus, E approximately 12 kPa). On glass and much softer or stiffer gels, including gels emulating stiff dystrophic muscle, cells do not striate. In addition, myotubes grown on top of a compliant bottom layer of glass-attached myotubes (but not softer fibroblasts) will striate, whereas the bottom cells will only assemble stress fibers and vinculin-rich adhesions. Unlike sarcomere formation, adhesion strength increases monotonically versus substrate stiffness with strongest adhesion on glass. These findings have major implications for in vivo introduction of stem cells into diseased or damaged striated muscle of altered mechanical composition.     

Focal contacts: transmembrane links between the extracellular matrix and the cytoskeleton

The sites of tightest adhesion that form between cells and substrate surfaces in tissue culture are termed focal contacts. The external faces of focal contacts include specific receptors, belonging to the integrin family of proteins, for fibronectin and vitronectin, two common components of extracellular matrices. On the internal (cytoplasmic) side of focal contacts, several proteins, including talin and vinculin, mediate interactions with the actin filament bundles of the cytoskeleton. The changes that occur in focal contacts as a result of viral transformation are discussed.     

Distribution of the cell substratum attachment (CSAT) antigen on myogenic and fibroblastic cells in culture

Previous studies (Neff et al., 1982, J. Cell. Biol. 95:654-666; Decker et al., 1984. J. Cell. Biol. 99:1388-1404) have described a monoclonal antibody (CSAT Mab) directed against a complex of three integral membrane glycoproteins of 120,000-160,000 mol wt (CSAT antigen [ag]) involved in the cell matrix adhesion of myoblasts and fibroblasts. In localization studies on fibroblasts presented here, CSAT ag has a discrete, well-organized distribution pattern. It co-aligns with portions of stress fibers and is enriched at the periphery of, but not directly beneath vinculin-rich focal contacts. In this last location, it co-distributes with fibronectin, consistent with the suggestion that the CSAT ag participates in the mechanism by which fibroblasts attach to fibronectin. In prefusion myoblasts, which are rapidly detached by CSAT Mab, CSAT ag is distributed diffusely as are vinculin, laminin, and fibronectin. After fusion, myotubes become more difficult to detach with CSAT Mab. The CSAT ag and vinculin are organized in a much more discrete pattern on the myotube surface, becoming enriched at microfilament bundle termini and in lateral lamellae which appear to attach myotubes to the substratum. These results suggest that the organization of CSAT ag-adhesive complexes on the surface of myogenic cells can affect the stability of their adhesive contacts. We conclude from the sum of the studies presented that, in both myogenic and fibroblastic cells, the CSAT ag is localized in sites expected of a surface membrane mediator of cell adhesion to extracelluon of CSAT ag-adhesive complexes on the surface of myogenic cells can affect the stability of their adhesive contacts. We conclude from the sum of the studies presented that, in both myogenic and fibroblastic cells, the CSAT ag is localized in sites expected of a surface membrane mediator of cell adhesion to extracellular matrix. The results from studies that use fibroblasts in particular suggest the involvement of CSAT ag in the adhesion of these cells to fibronectin.     

Selective regulation of integrin--cytoskeleton interactions by the tyrosine kinase Src

Cell motility on extracellular-matrix (ECM) substrates depends on the regulated generation of force against the substrate through adhesion receptors known as integrins. Here we show that integrin-mediated traction forces can be selectively modulated by the tyrosine kinase Src. In Src-deficient fibroblasts, cell spreading on the ECM component vitronectin is inhibited, while the strengthening of linkages between integrin vitronectin receptors and the force-generating cytoskeleton in response to substrate rigidity is dramatically increased. In contrast, Src deficiency has no detectable effects on fibronectin-receptor function. Finally, truncated Src (lacking the kinase domain) co-localizes to focal-adhesion sites with alpha v but not with beta 1 integrins. These data are consistent with a selective, functional interaction between Src and the vitronectin receptor that acts at the integrin-cytoskeleton interface to regulate cell spreading and migration.     

Clusters of neural cell adhesion molecule at sites of cell-cell contact

I have examined the distribution of neural cell adhesion molecule (N-CAM) in cultured C2 myogenic cells and other cell lines to determine if N-CAM accumulates at sites of cell-cell contact. C2 cells growing in log phase display large clusters of neural cell adhesion molecule where they contact each other. These clusters are remarkably stable, do not form at cell-substrate contacts, and appear not to be enriched in a number of other cytoskeletal, membrane, or extracellular proteins. Thus, N-CAM clusters form preferentially in response to cell-cell contact and are specifically enriched in N-CAM. As C2 cultures mature and differentiate, clusters persist at contacts between aligning myoblasts and between myotubes, consistent with a role in myogenesis. N-CAM is also enriched at cell-cell contacts in cultures of PC12, NRK, and CHO cells. These cells have significant amounts of N-CAM as detected on immunoblots. Clusters are not seen in L929 cells, which do not have detectable amounts of N-CAM. Coculture of these cells with C2 cells results in the clustering of N-CAM at heterologous contacts between C2 cells and NRK, CHO, or PC12 cells, but not between C2 cells and L929 cells. These results suggest that N-CAM specifically accumulates where N-CAM-bearing cells contact one another. Clustering of N-CAM may be an important step in strengthening intercellular adhesion.     

Association of intermediate filaments with vinculin-containing adhesion plaques of fibroblasts

Double immunofluorescence staining of quail embryo fibroblasts with rabbit antibody to vinculin and mouse monoclonal antibody to vimentin revealed a coincidence between fluorescence patterns for cell-substrate focal contacts and intermediate filaments. Most of the vinculin-containing adhesion plaques coincided with the ends of vimentin-positive fibrils. This association was further corroborated by immunoelection microscopic observations of the cytoskeletons of quail and mouse fibroblasts using a platinum replica technique. The intermediate filaments were identified either by direct treatment with antivimentin IgM or by an indirect immunogold staining method. Colcemid treatment of the cells caused a collapse of intermediate filaments and destroyed their association with focal contacts. During the early stages of the colcemid-induced collapse of the intermediate filaments, single vimentin fibrils appeared to retain their association with focal contacts. The possible role of the intermediate filaments in the formation and maintenance of focal contacts is discussed.     

Cell-substrate interactions during amoeboid locomotion of neutrophil leukocytes

Cell-substrate interactions between human blood neutrophils moving on a glass substrate in serum-free medium have been investigated using reflexion interference microscopy, high voltage and scanning electron microscopy (SEM). The contact pattern with the substrate differed considerably from that found in fibroblasts and the amoeba Naegleria. Discrete focal contacts could not be detected but large broad areas of very close contact (accounting for about 30% of the total contact area) could be found particularly associated with the uroid. Considerable loss of membrane material occurred as a result of breakdown of the uroid during locomotion.     

Actin-independent association of vinculin with the cytoplasmic aspect of the plasma membrane in cell-contact areas

We investigated the mode of association of vinculin with areas of contact between the termini of microfilament bundles and the cell membrane in sites of focal contact with the substrate by selective removal of actin from these areas. Opened-up substrate-attached membranes of chick fibroblasts as well as detergent-permeabilized cells were treated with fragmin from Physarum in the presence of Ca+2. This treatment removed actin filaments from the cytoplasmic faces of the membranes, along with several actin-associated proteins (alpha-actinin, tropomyosin, myosin, and filamin). Vinculin distribution was not affected by treatment. Moreover, rhodamine- or fluorescein-conjugated vinculin, when added to these preparations, became specifically associated with the focal contacts regardless of whether the latter were pretreated with fragmin or not. We conclude that the association of vinculin with focal contacts is largely actin-independent. We discuss the implications of these findings in the molecular mechanisms of microfilament membrane association in areas of cell contact.     

The localization of acetylcholine receptor clusters in areas of cell-substrate contact in cultures of rat myotubes

We have used interference reflection and fluorescence microscopy to investigate the relationship between cell-substrate contact and the location of clusters of acetylcholine receptors (AChRs) in cultures of rat myotubes. We have found that AChR clusters on the ventral myotube surfaces are always located within broad regions of close cell-substrate contact. Detailed analysis of the fine structure of the AChR cluster and its associated contact region showed that AChRs within a cluster are concentrated between the points of closest cell-substrate apposition. Vinculin, a recently discovered intracellular smooth muscle protein, is also concentrated in broad regions of close contact, interdigitating with AChRs within the clusters.     

Association of fibronectin and vinculin with focal contacts and stress fibers in stationary hamster fibroblasts

We have recently observed a transmembrane association between extracellular fibronectin (FN) fibers and elongated focal patches or fibers of vinculin (VN) in G1-arrested stationary Nil 8 hamster fibroblasts, with double-label immunofluorescence microscopy (Singer and Paradiso, 1981, Cell. 24:481-492). We hypothesized that these FN-VN complexes might correspond to focal contacts, the membrane sites that are probably mainly responsible for attaching cells to their substrata, because vinculin is often localized in focal contacts. However, because fibronectin-vinculin associations may not be restricted to the substrate adhesive surface of the cell, it became necessary to determine whether some or all of the various kinds of FN-VN complexes which we described are in proximity to the substrate. Using interference reflection optics and double-label immunofluorescence microscopy for fibronectin and vinculin, many elongated (up to 38 micrometer) FN-VN associations were found to be strikingly coincident with focal contacts in the perinuclear area of extremely flattened arrested Nil 8 fibroblasts in 0.3% fetal bovine serum (FBS). In addition, the long FN-VN adhesion complexes were precisely aligned with the major phase-dense stress fibers observed at the ventral surfaces of these stationary cells with phase contrast microscopy. Fibronectin was neither associated with vinculin-containing focal contacts of Nil 8 cells cultured in medium with 5% FBS nor with vinculin-negative focal contacts located at the extreme edges of stationary cells arrested in 0.3 FBS. Our time-course experiments suggest that early FN-VN lacking- focal contacts, which form at the cellular margins, develop into mature substrate adhesion complexes containing both fibronectin and vinculin, localized in the major stress fibers at the centers of sessile fibroblasts.     

Novel function for beta 1 integrins in keratinocyte cell-cell interactions.

We have examined the expression, localization, and function of beta 1 integrins on cultured human epidermal keratinocytes using polyclonal and monoclonal antibodies against the beta 1, alpha 2, alpha 3, and alpha 5 integrin subunits. The beta 1 polypeptide, common to all class 1 integrins, was localized primarily in areas of cell-cell contacts of cultured keratinocytes, as were alpha 2 and alpha 3 polypeptides, suggesting a possible role in cell-cell adhesion for these integrin polypeptides. In contrast, the fibronectin receptor alpha 5 subunit showed no such accumulations in regions of cell-cell contact but was more diffusely distributed in the keratinocyte plasma membrane, consistent with the absence of fibronectin at cell-cell contact sites. Colonies of cultured keratinocytes could be dissociated by treatment with monoclonal antibody specific to the beta 1 polypeptide. Such dissociation of cell-cell contacts also occurred under conditions where the monoclonal antibody had no effect on cell-substrate adhesion. Therefore, beta 1 integrin-dependent cell-cell adhesion can be inhibited without affecting other cell-adhesive interactions. Antibody treatment of keratinocytes maintained in either low (0.15 mM) or high (1.2 mM) CaCl2 also resulted in the loss of organization of intracellular F-actin filaments and beta 1 integrins, even when the anti-beta 1 monoclonal antibody had no dissociating effect on keratinocyte colonies at the higher calcium concentration. Our results indicate that beta 1 integrins play roles in the maintenance of cell-cell contacts between keratinocytes and in the organization of intracellular microfilaments. They suggest that in epithelial cells integrins can function in cell-cell interactions as well as in cell-substrate adhesion.     

Local calcium transients contribute to disappearance of pFAK, focal complex removal and deadhesion of neuronal growth cones and fibroblasts

Cell adhesion is crucial for migration of cells during development, and cell-substrate adhesion of motile cells is accomplished through the formation and removal of focal complexes that are sites of cell-substrate contact. Because Ca2+ signaling regulates the rate of axon outgrowth and growth cone turning, we investigated the potential role of Ca2+ in focal complex dynamics. We describe a novel class of localized, spontaneous transient elevations of cytosolic Ca2+ observed both in Xenopus neuronal growth cones and fibroblasts that are 2-6 mum in spatial extent and 2-4 s in duration. They are distributed throughout growth cone lamellipodia and at the periphery of fibroblast pseudopodia, which are regions of high motility. In both cell types, these Ca2+ transients lead to disappearance of phosphorylated focal adhesion kinase (pFAK) and deadhesion from the substrate as assessed by confocal and internal reflection microscopy, respectively. The loss of pFAK is inhibited by cyclosporin A, suggesting that these Ca2+ transients exert their effects via calcineurin. These results identify an intrinsic mechanism for local cell detachment that may be modulated by agents that regulate motility.     

Marching at the front and dragging behind: differential alphaVbeta3-integrin turnover regulates focal adhesion behavior.

Integrins are cell-substrate adhesion molecules that provide the essential link between the actin cytoskeleton and the extracellular matrix during cell migration. We have analyzed alphaVbeta3-integrin dynamics in migrating cells using a green fluorescent protein-tagged beta3-integrin chain. At the cell front, adhesion sites containing alphaVbeta3-integrin remain stationary, whereas at the rear of the cell they slide inward. The integrin fluorescence intensity within these different focal adhesions, and hence the relative integrin density, is directly related to their mobility. Integrin density is as much as threefold higher in sliding compared with stationary focal adhesions. High intracellular tension under the control of RhoA induced the formation of high-density contacts. Low-density adhesion sites were induced by Rac1 and low intracellular tension. Photobleaching experiments demonstrated a slow turnover of beta3-integrins in low-density contacts, which may account for their stationary nature. In contrast, the fast beta3-integrin turnover observed in high-density contacts suggests that their apparent sliding may be caused by a polarized renewal of focal contacts. Therefore, differential acto-myosin-dependent integrin turnover and focal adhesion densities may explain the mechanical and behavioral differences between cell adhesion sites formed at the front, and those that move in the retracting rear of migrating cells.