Immunolocalization of meta-vinculin in human smooth and cardiac muscles

Meta-vinculin, a vinculin-related protein, has been isolated from human uterus smooth muscle. Specific antibodies to meta-vinculin, which distinguish between meta-vinculin and vinculin, were prepared by absorption of anti-meta-vinculin serum on vinculin coupled to nitrocellulose. Meta-vinculin specific antibody demonstrates only smooth and cardiac muscle specificity and is able to cross-react with a small 21-kD fragment of the meta-vinculin polypeptide chain. This antibody does not interact with protease resistant 95-kD core shared by vinculin and meta-vinculin. Meta-vinculin specific antibody was used for the localization of meta-vinculin in smooth and cardiac muscles by the indirect immunofluorescence method. At the light microscopy resolution level it was found that meta-vinculin and vinculin are localized in the same cellular adhesive structures. Meta-vinculin is present in membrane-associated microfilament-bound plaques of smooth muscle, in intercalated discs and costameres of cardiac muscle. In primary culture of smooth muscle cells from human aorta, meta-vinculin and vinculin were found to be present in focal contacts of the cells. During the cultivation of smooth muscle cells, the quantity of meta-vinculin decreased progressively and finally meta-vinculin completely disappeared from the focal contacts. The data show that in smooth and cardiac muscles meta-vinculin could be a structural component of microfilament-membrane attachment sites, defined earlier by the localization of vinculin.     

Vinculin is a component of an extensive network of myofibril-sarcolemma attachment regions in cardiac muscle fibers

Immunofluorescent staining of bovine and avian cardiac tissue with affinity-purified antibody to chicken gizzard vinculin reveals two new sites of vinculin reactivity. First, vinculin is organized at the sarcolemma in a striking array of rib-like bands, or costameres. The costameres encircle the cardiac muscle cell perpendicular to the long axis of the fiber and overlie the I bands of the immediately subjacent sarcomeres. The second new site of vinculin reactivity is found in bovine cardiocytes at tubular invaginations of the plasma membrane. The frequency and location of these invaginations correspond to the known frequency and distribution of the transverse tubular system in bovine atrial, ventricular, and Purkinje fibers. We do not detect tubular invaginations that stain with antivinculin in avian cardiocytes and, in fact, a transverse tubular system has not been found in avian cardiac fibers. Apparent lateral Z-line attachments to the sarcolemma and its invaginations have been observed in cardiac muscle by electron microscopy in the same regions where we find vinculin. On the basis of these previous ultrastructural findings and our published evidence for a physical connection between costameres and the underlying myofibrils in skeletal muscle, we interpret the immunofluorescence data of this study to mean that, in cardiac muscle, vinculin is a component of an extensive system of lateral attachment of myofibrils to the plasma membrane and its invaginations.     

Distribution of vinculin in the Z-disk of striated muscle: analysis by laser scanning confocal microscopy

Vinculin is a major cytoskeletal component in striated muscle, where it has been reported to form a rib-like structure between the cell membrane and the Z-disk termed a costamere. This arrangement of vinculin has been purported to be involved in the alignment of the myofibrils. However, the three-dimensional arrangement of vinculin in relation to the Z-disk of the myofibril was not known. In the present study, we examined the distribution of vinculin in striated muscle with monospecific antibodies using immunofluorescence and laser scanning confocal microscopy. Isolated cardiac and skeletal muscle cells from a variety of species, tissue sections, and neonatal myocytes with developing myofibrils were examined. Optical sectioning in the X-Y and X-Z planes demonstrated that vinculin immunoreactivity was heaviest at the periphery of the cell; however, the immunoreactivity was also distributed within the Z-disk although at a relatively reduced level. This distribution is potentially significant in understanding the physiological significance of vinculin in striated muscle function and in myofibrillogenesis.     

Localization of talin in skeletal and cardiac muscles

Antibodies to talin and vinculin were used for localization of these proteins in skeletal and cardiac muscles by the indirect immunofluorescence method. We have found that talin is localized in intercalated discs of cardiac muscle and in costameres of skeletal and cardiac muscles. It is suggested that in striated muscles talin and vinculin play an important role in interactions between actin filaments and membranes.     

Diversity of vinculin/meta-vinculin in human tissues and cultivated cells. Expression of muscle specific variants of vinculin in human aorta smooth muscle cells

Microheterogeneity of different vinculin and meta-vinculin isoforms in adult human tissues and cultured cells was studied by two-dimensional gel electrophoresis and immunoblotting technique. Four isoforms of vinculin (alpha, alpha', beta, and gamma) and two isoforms of meta-vinculin (alpha and beta) were resolved. alpha-, alpha'-, and beta-isoforms of vinculin were found in all cell types and tissue samples analyzed in the present study. gamma-Isoform of vinculin and both alpha- and beta-isoforms of meta-vinculin were found in smooth (aorta wall and myometrium) and cardiac muscle, rather than in skeletal muscle, liver, foreskin fibroblasts, and macrophages. In the primary culture of human aorta smooth muscle cells, the fractional content of gamma-isoform of vinculin and meta-vinculin was dramatically reduced, and, by the onset of intensive cell division, the proteins could hardly be detected. Subcultured human aorta smooth muscle cells did not contain gamma-vinculin and meta-vinculin. We analyzed the microheterogeneity of vinculin and meta-vinculin in three smooth muscle layers of human aorta wall--media, muscular-elastic (adjacent to media) intima, and subendothelial (juxtaluminal) intima. It was shown that in media the fractional content of gamma-isoform of vinculin was 45% and meta-vinculin, 42%; in muscular-elastic intima the fractional content of gamma-vinculin was 42% and meta-vinculin, 36%. However, in subendothelial intima, the share of these proteins was significantly lower than in adjacent muscular-elastic intima and media. Isoactin pattern that is characteristic of smooth muscle was identical in all aortic layers, thus proving the smooth muscle origin of subendothelial intima cells. These findings demonstrate that human aortic smooth muscle cells in vivo and in vitro undergo coordinated differential expression of smooth muscle specific variants of vinculin, i.e. gamma-vinculin and meta-vinculin.     

A plasma membrane integral sialoglycoprotein (Sgp 130) molecularly distinguishes nonjunctional dense plaque sites of microfilament attachment

An integral sialoglycoprotein with Mr approximately 130,000 (Sgp 130) and highest expression in adult chicken gizzard smooth muscle has been recently identified as an excellent candidate for classification as a plasma membrane protein natively associated (directly or indirectly) with actin microfilaments (Rogalski, A.A., and S.J. Singer, 1985, J. Cell Biol., 101:785-801). In this study, the relative in situ distributions of the Sgp 130 integral species (a designation that also includes non-smooth muscle molecular forms) and the peripheral protein, vinculin, have been simultaneously revealed for the first time in selected cultured cells and tissues abundant in microfilament-membrane attachment sites, particularly, smooth and cardiac muscle. Specific antibody probes against Sgp 130 (mouse mAb 30B6) and vinculin (affinity-purified rabbit antibody) were used in double indirect immunofluorescent and immunoelectron microscopic experiments. In contrast to the widespread distributions of vinculin at microfilament-membrane attachment sites, Sgp 130 has been shown to exhibit striking site-specific variation in its abundancy levels in the plasma membrane. Sgp 130 and vinculin were found coincidentally concentrated at focal contact sites in cultured chick embryo fibroblasts and endothelial cells, membrane dense plaques of smooth muscle, and sarcolemma dense plaque sites overlying the Z line in cardiac muscle. However, at the fascia adherens junctional sites of cardiac muscle where vinculin is sharply confined, Sgp 130 was immunologically undetectable in both intact and EGTA-uncoupled tissue. This latter result was confirmed with immunoblotting experiments using isolated forms of the fascia adherens. The double immunolabeling studies of this report establish Sgp 130 as a major integral protein component of nonjunctional membrane dense plaque structures and raise the possibility that the 130-kD integral sialoglycoprotein (Sgp 130) and vinculin assume stable transmembrane associations at these particular microfilament-membrane attachment sites. Nonjunctional dense plaques are further suggested to be a molecularly distinct class of plasma membrane structures rather than a subgroup of adherens junctions. Our data also support a hypothesis that Sgp 130 is involved in plasma membrane force coupling events but not in junctional-related cell-cell coupling.     

Subcellular localization of dystrophin and vinculin in cardiac muscle fibers and fibers of the conduction system of the chicken ventricle

The subcellular localization of dystrophin and vinculin was investigated in cardiac muscle fibers and fibers of the conduction system of the chicken ventricle by immunofluorescence confocal microscopy. In ventricular cardiac muscle fibers, strong staining with antibody against dystrophin appeared as regularly arranged transverse striations at the sarcolemmal surface, and faint but uniform staining was seen in narrow strips between these striations. In fibers of the ventricular conduction system, the sarcolemma was stained uniformly with this antibody, but strong staining was found as regular striations in many areas and as scattered patches in other areas of the sarcolemma. These intensely stained striations and scattered patches of dystrophin were colocalized with those of vinculin. Because dystrophin striations were located at the level of Z bands of the underlying myofibrils, they were regarded as the concentration of this protein at costameres together with vinculin. In fibers of the conduction system, myofibrils were close to the sarcolemma where dystrophin and vinculin assumed a striated pattern, at some distance from the cell membrane where these proteins exhibited a patchy distribution, and distant from the sarcolemma where dystrophin was uniformly distributed. These data suggest that the distribution patterns of dystrophin reflect the degree of association between the sarcolemma and underlying myofibrils.     

Immunoelectron microscopic studies of desmin (skeletin) localization and intermediate filament organization in chicken cardiac muscle

We studied the localization of desmin (skeletin), the major protein subunit of muscle-type intermediate filaments, in adult chicken cardiac muscle by high resolution immunoelectron microscopic labeling of ultrathin frozen sections of the intact fixed tissues. We carried out single labeling for desmin and double labeling for both desmin and either vinculin or alpha-actinin. In areas removed from the intercalated disk membranes, we observed desmin labeling between adjacent Z-bands in every interfibrillar space. Where these spaces were wide and contained mitochondria, convoluted strands of desmin labeling bridged between the periphery of neighboring Z-bands and the mitochondria. The intermediate filaments appeared to be organized in a more three-dimensional manner within the interfibrillar spaces of cardiac as compared to skeletal muscle. Near the intercalated disks, desmin labeling was intense within the interfibrillar spaces, but was completely segregated from the microfilament attachment sites (fascia adherens) where vinculin and alpha-actinin were localized. Desmin therefore appears to play no role in the attachment of microfilaments to the intercalated disk membrane. We discuss the role of intermediate filaments in the organization of cardiac and skeletal striated muscle in the light of these and other results.     

Ultrastructural and biochemical localization of N-RAP at the interface between myofibrils and intercalated disks in the mouse heart.

N-RAP is a recently discovered muscle-specific protein found at cardiac intercalated disks. Double immunogold labeling of mouse cardiac muscle reveals that vinculin is located immediately adjacent to the fascia adherens region of the intercalated disk membrane, while N-RAP extends approximately 100 nm further toward the interior of the cell. We partially purified cardiac intercalated disks using low- and high-salt extractions followed by density gradient centrifugation. Immunoblots show that this preparation is highly enriched in desmin and junctional proteins, including N-RAP, talin, vinculin, beta1-integrin, N-cadherin, and connexin 43. Electron microscopy and immunolabeling demonstrate that N-RAP and vinculin are associated with the large fragments of intercalated disks that are present in this preparation, which also contains numerous membrane vesicles. Detergent treatment of the partially purified intercalated disks removed the membrane vesicles and extracted vinculin and beta1-integrin. Further separation on a sucrose gradient removed residual actin and myosin and yielded a fraction morphologically similar to fasciae adherentes that was highly enriched in N-RAP, N-cadherin, connexin 43, talin, desmin, and alpha-actinin. The finding that N-RAP copurifies with detergent-extracted intercalated disk fragments even though beta-integrin and vinculin have been completely removed suggests that N-RAP association with the adherens junction region is mediated by the cadherin system. Consistent with this hypothesis, we found that recombinant N-RAP fragments bind alpha-actinin in a gel overlay assay. In addition, immunofluorescence shows that N-RAP remains bound at the ends of isolated, detergent-treated cardiac myofibrils. These results demonstrate that N-RAP remains tightly bound to myofibrils and fasciae adherentes during biochemical purification and may be a key constituent in the mechanical link between these two structures.     

Purification of a 175-kDa membrane protein, its localization in smooth and cardiac muscles. Interaction with cytoskeletal protein - vinculin

A new 175-kDa membrane protein was isolated from chicken gizzard smooth muscle. Antibodies to 175-kDa protein were used for localization of this protein in smooth and cardiac muscles. In both types of muscle 175-kDa protein was localized near plasma membrane. 175-kDa protein was able to interact specifically with vinculin immobilized on polysterene surface. It is suggested that this 175-kDa protein may be involved in physical connection between microfilaments and cell membrane.     

Localization of the actin-binding protein fesselin in chicken smooth muscle

This report compares cellular localization of fesselin in chicken smooth, skeletal and cardiac muscle tissues using affinity purified polyclonal fesselin antibodies. Western blot analyses revealed large amounts of fesselin in gizzard smooth muscle with lower amounts in skeletal and cardiac muscle. In gizzard, fesselin was detected by immunofluorescence as discrete cytoplasmic structures. Fesselin did not co-localize with talin, vinculin or caveolin indicating that fesselin is not associated with dense plaques or caveolar regions of the cell membrane. Immunoelectron microscopy established localization of fesselin within dense bodies. Since dense bodies function as anchorage points for actin and desmin in smooth muscle cells, fesselin may be involved in establishing cytoskeletal structure in this tissue. In skeletal muscle, fesselin was associated with desmin in regularly spaced bands distributed along the length of muscle fibers suggesting localization to the Z-line. Infrequently, this banding pattern was observed in heart tissue as well. Localization at the Z-line of skeletal and cardiac muscle suggests a role in contraction of these tissues.     

Vinculin Phosphorylation at Tyr1065 Regulates Vinculin Conformation and Tension Development in Airway Smooth Muscle Tissues

Vinculin localizes to membrane adhesion junctions in smooth muscle tissues, where its head domain binds to talin and its tail domain binds to filamentous actin, thus linking actin filaments to the extracellular matrix. Vinculin can assume a closed conformation, in which the head and tail domains bind to each other and mask the binding sites for actin and talin, and an open activated conformation that exposes the binding sites for talin and actin. Acetylcholine stimulation of tracheal smooth muscle tissues induces the recruitment of vinculin to the cell membrane and its interaction with talin and actin, which is required for active tension development. Vinculin phosphorylation at Tyr¹⁰⁶⁵ on its C terminus increases concurrently with tension development in tracheal smooth muscle tissues. In the present study, the role of vinculin phosphorylation at Tyr¹⁰⁶⁵ in regulating the conformation and function of vinculin during airway smooth muscle contraction was evaluated. Vinculin constructs with point mutations at Tyr¹⁰⁶⁵ (vinculin Y1065F and vinculin Y1065E) and vinculin conformation-sensitive FRET probes were expressed in smooth muscle tissues to determine how Tyr¹⁰⁶⁵ phosphorylation affects smooth muscle contraction and the conformation and cellular functions of vinculin. The results show that vinculin phosphorylation at tyrosine 1065 is required for normal tension generation in airway smooth muscle during contractile stimulation and that Tyr¹⁰⁶⁵ phosphorylation regulates the conformation and scaffolding activity of the vinculin molecule. We conclude that the phosphorylation of vinculin at tyrosine 1065 provides a mechanism for regulating the function of vinculin in airway smooth muscle in response to contractile stimulation.     

Expression of meta-vinculin associated with differentiation of chicken embryonal muscle cells

We confirmed the existence of meta-vinculin, which cross-reacts immunologically with vinculin and has a slightly higher molecular weight than the latter. The immunological cross-reactivity between meta-vinculin and vinculin was confirmed with monoclonal antibodies against vinculin. Furthermore, we found that this protein is present only in either smooth or striated muscle, but is absent in non-muscular tissues and that the expression of this protein is associated with a differentiation of muscle cells either in vivo or in vitro. Microinjection experiments of fluorescently labelled vinculin and/or meta-vinculin into the cytoplasm of cultured myotubes suggested that meta-vinculin may be localized at sites similar to vinculin in muscle cells.     

Muscle differentiation in Xenopus laevis is characterized by particular states of vinculin phosphorylation.

Vinculin is a 130 kD cytoskeletal protein which is involved in the anchorage of actin microfilaments to the plasma membranes at sites of cell-cell and cell-matrix contacts. In this paper we prove that smooth and cardiac muscles of Xenopus laevis contain a specific isoform of vinculin not present in any other tissue including skeletal muscle and epithelia and we demonstrate that this form of the molecule is characterized by a specific state of phosphorylation. These data are discussed in view of the importance of posttranslational modifications of structural proteins, such as vinculin, in the determination of cellular behaviour during differentiation and development.     

Properties of smooth muscle vinculin

Vinculin, isolated from turkey gizzard smooth muscle, was purified by chromatography on CM-cellulose after isolation from a DEAE-cellulose column. Two-dimensional gel electrophoretic analysis of crude muscle fractions demonstrated that: 1) much of the approximately 130,000-dalton protein present in smooth muscle did not co-isoelectrically focus with the purified 130,000-dalton vinculin and 2) the purified vinculin consisted of three major, closely spaced isoelectric variants that were present only in small amounts in the original smooth muscle sample. Purified vinculin sedimented as a single peak with a sedimentation coefficient S0 20,w of 5.9. Circular dichroism spectra of purified vinculin indicated a considerable degree of secondary structure, with an alpha-helical content of approximately 50% as measured at 208 nm. The ultraviolet absorption spectrum of vinculin gave a measured E1%(278) of 4.64. Digestion of vinculin, much of which is located at the cytoplasmic surface of the cell membrane, with Ca2+-activated neutral protease purified from skeletal muscle yielded major fragments with molecular weights determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis of 98,000, 85,000, and 26,000. The factor(s) in DEAE-cellulose-purified vinculin responsible for decreasing the low shear viscosity of actin was removed and found in a crude fraction isolated by CM-cellulose chromatography. The purified vinculin had a small, but positive effect on the MgCl2-induced polymerization of actin as measured by low shear viscometry.     

Meta-vinculin distribution in adult human tissues and cultured cells

Meta-vinculin distribution in adult human tissue was studied by immunoblotting technique. Meta-vinculin was found in smooth (aorta wall and myometrium) and cardiac muscle, rather than in skeletal muscle, liver, kidney and cultured cells - macrophages, foreskin fibroblasts, peripheral blood lymphocytes and vascular endothelial cells. In the primary culture of smooth muscle cells from human aorta the meta-vinculin/vinculin ratio was reduced, and on the onset of cell division meta-vinculin could hardly be detected. Subcultured smooth muscle cells from human aorta did not contain meta-vinculin. The data show that the presence of meta-vinculin is characteristic of 'contractile' smooth muscle cells rather than of proliferating in vitro.     

The β1 Integrin Subunit is Not a Specific Component of the Costamere Domain in Human Myocardial Cells

Studies on altered integrin receptor expression during cardiac hypertrophy and heart failure requires accurate knowledge of the distributional pattern of integrins in myocardial cells. At present the general consensus is that in cardiac muscle the β₁ integrin receptor is mainly localized to the same sarcolemmal domain as vinculin at Z-band levels (‘costamere’). Since most previous studies have been focusing on myocardial integrin distribution in lower mammals, the myocardial localization of the β₁ integrin subunit was investigated in biopsies collected from the auricle of patients undergoing a coronary bypass operation. Non-invasive serial optical sectioning was carried out by immuno-laser scanning confocal microscopy. Double-labelling for vinculin/α-actinin, and the cytoplasmic domain for the β₁ integrin subunit, showed that β₁ integrin is deposited throughout both the vinculin/α-actinin domains and the non-vinculin/α-actinin domains. These results were supported by a semi-quantitative analysis in extended focus images of the latter preparations. Higher magnification views at the electron microscopical levels of the large, extracellular domain of the β₁ integrin subunit disclosed a pronounced labelling in the form of a dense, irregular punctuate pattern that was distributed at Z-disc domains as well as along the entire sarcolemmal area between Z-discs. Our findings show that in human, myocardial cells, the β₁ integrin receptor does not only localize to the surface membrane at the Z-disc level (‘costamere’ in cardiac muscle), but has a widespread distribution along the sarcolemma.     

Role of N-cadherin- and integrin-based costameres in the development of rat cardiomyocytes

Costameres, vinculin-containing structures found in skeletal and cardiac muscle, are thought to anchor the Z-discs of the peripheral myofibrils to the sarcolemma. Several lines of evidence indicate that two different sets of costameres, integrin- and N-cadherin-based, are present in cardiac muscles. In this study, immunoblot analysis was used to study the expression of N-cadherin, alpha-catenin, beta-catenin, vinculin, talin, and laminin in rat cardiac muscles at embryonic days 15 and 19, the day of birth (postnatal day 0), postnatal weeks 1, 2, 3, and 4, and in the adult. Double immunofluorescence microscopy was performed to study the spatial and temporal distribution of these two sets of costameres in rat cardiomyocytes. Costameric staining for N-cadherin, codistributed with beta-catenin, was strong from embryonic day 15 up to postnatal week 2, gradually decreased after postnatal week 3, and was undetectable at postnatal week 4 and in the adult. Confocal microscopy showed that N-cadherin colocalized with alpha-actinin at cortical myofibrils. Double-labeling of beta-catenin and talin indicated the coexistence of N-cadherin/catenin- and integrin/talin-based costameres in rat cardiac muscle. Although beta-catenin and vinculin were co-localized at the costamere of cardiomyocytes from embryonic day 15 to postnatal week 3, staining for beta-catenin or talin was mutually exclusive at all stages examined. These results demonstrate the simultaneous, but mutually exclusive, existence of N-cadherin/catenin- and integrin/talin-based costameres in rat cardiomyocytes between late embryonic stages and postnatal week 3, while only integrin/talin-based costameres were found in adult rats. The N-cadherin/catenin-based costameres in rat cardiac muscles may play a role in myofibrillogenesis similar to that of their counterparts in cultured cardiomyocytes.     

Trypanosoma cruzi infection disrupts vinculin costameres in cardiomyocytes

Chagas' disease cardiomyopathy is an important manifestation of Trypanosoma cruzi infection, leading to cardiac dysfunction and serious arrhythmias. We have here investigated by indirect immunofluorescence assay the distribution of vinculin, a focal adhesion protein with a major role in the transmission of contraction force, during the T. cruzi-cardiomyocyte infection in vitro and in vivo. No change in vinculin distribution was observed after 24 h of infection, where control and T. cruzi-infected cardiomyocytes displayed vinculin localized at costameres and intercalated discs. On the other hand, a clear disruption of vinculin costameric distribution was noted after 72 h of infection. A significant reduction in the levels of vinculin expression was observed at all times of infection. In murine experimental Chagas' disease, alteration in the vinculin distribution was also detected in the infected myocardium, with no costameric staining in infected myocytes and irregular alignment of intercalated discs in cardiac fibers. These data suggest that the disruption of costameric vinculin distribution and the enlargement of interstitial space due to inflammatory infiltration may contribute to the reduction of transmission of cardiac contraction force, leading to alterations in the heart function in Chagas' disease.     

Activation of vinculin induced by cholinergic stimulation regulates contraction of tracheal smooth muscle tissue

Vinculin localizes to membrane adhesion junctions where it links actin filaments to the extracellular matrix by binding to the integrin-binding protein talin at its head domain (Vh) and to actin filaments at its tail domain (Vt). Vinculin can assume an inactive (closed) conformation in which Vh and Vt bind to each other, masking the binding sites for actin and talin, and an active (open) conformation in which the binding sites for talin and actin are exposed. We hypothesized that the contractile activation of smooth muscle tissues might regulate the activation of vinculin and thereby contribute to the regulation of contractile tension. Stimulation of tracheal smooth muscle tissues with acetylcholine (ACh) induced the recruitment of vinculin to cell membrane and its interaction with talin and increased the phosphorylation of membrane-localized vinculin at the C-terminal Tyr-1065. Expression of recombinant vinculin head domain peptide (Vh) in smooth muscle tissues, but not the talin-binding deficient mutant head domain, VhA50I, inhibited the ACh-induced recruitment of endogenous vinculin to the membrane and the interaction of vinculin with talin and also inhibited vinculin phosphorylation. Expression of Vh peptide also inhibited ACh-induced smooth muscle contraction and inhibited ACh-induced actin polymerization; however, it did not affect myosin light chain phosphorylation, which is necessary for cross-bridge cycling. Inactivation of RhoA inhibited vinculin activation in response to ACh. We conclude that ACh stimulation regulates vinculin activation in tracheal smooth muscle via RhoA and that vinculin activation contributes to the regulation of active tension by facilitating connections between actin filaments and talin-integrin adhesion complexes and by mediating the initiation of actin polymerization.