Advertisement
JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M314245200 on May 24, 2004

J. Biol. Chem., Vol. 279, Issue 30, 31533-31543, July 23, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/30/31533    most recent
M314245200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Witt, S.
Right arrow Articles by Illenberger, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Witt, S.
Right arrow Articles by Illenberger, S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Comparative Biochemical Analysis Suggests That Vinculin and Metavinculin Cooperate in Muscular Adhesion Sites*

Sebastian Witt, Anke Zieseniss, Ulrike Fock, Brigitte M. Jockusch, and Susanne Illenberger{ddagger}

From the Cell Biology, Zoological Institute, Technical University of Braunschweig, D-38092 Braunschweig, Germany

Received for publication, December 29, 2003 , and in revised form, May 4, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Metavinculin, the muscle-specific splice variant of the cell adhesion protein vinculin, is characterized by a 68-amino acid insert within the C-terminal tail domain. The findings that mutations within this region correlate with hereditary idiopathic dilated cardiomyopathy in man suggest a specific contribution of metavinculin to the molecular architecture of muscular actin-membrane attachment sites, the nature of which, however, is still unknown. In mice, metavinculin is expressed in smooth and skeletal muscle, where it co-localizes with vinculin in dense plaques and costameres, respectively, but is of conspicuously low abundance in the heart. Immunoprecipitates suggest that both isoforms are present in the same complex. On the molecular level, both vinculin isoforms are regulated via an intramolecular head-tail interaction, with the metavinculin tail domain having a lower affinity for the head as compared with the vinculin tail. In addition, metavinculin displays impaired binding to acidic phospholipids and reduced homodimerization. Only in the presence of phospholipid-activated vinculin tail, the metavinculin tail domain is readily incorporated into heterodimers. Mutational analysis revealed that the metavinculin insert significantly alters binding of the C-terminal hairpin loop to acidic phospholipids. In summary, our data lead to a model in which unfurling of the metavinculin tail domain is impaired by the negative charges of the 68-amino acid insert, thus requiring vinculin to fully activate the metavinculin molecule. As a consequence, microfilament anchorage may be modulated at muscular adhesion sites through heterodimer formation.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
The microfilament-associated protein vinculin is a highly conserved component of both cell-cell and cell-matrix adherens-type junctions, where it anchors actin filaments to the plasma membrane (1). It has an essential function in embryogenesis and is involved in the regulation of adhesion, motility, and spreading of cells at least in cell culture (2–5). Vinculin consists of a 90-kDa globular head domain (VH)1 and a 30-kDa tail domain (VT) that are connected by a proline-rich hinge region, conferring high flexibility to the molecule (6). It is a multiligand protein, and binding of VH to talin or {alpha}-catenin targets vinculin to microfilament-associated cell adhesions (7–10). Anchorage of microfilaments is mediated by VT that harbors two actin binding regions (11, 12). Through the recruitment of the vasodilator-stimulated phosphoprotein VASP and the Arp 2/3 complex to the hinge region, vinculin is also tightly linked to actin dynamics at cell adhesion sites (13–15). Ligand interactions are regulated by an intramolecular association of VH and VT, which masks the binding sites for most ligands (12, 15–19). Two different mechanisms were described to activate vinculin, both leading to an opening of the molecule, thus exposing the ligand binding sites of the protein. First it was shown that binding of acidic phospholipids, especially phosphatidylinositol 4,5-bisphosphate (PIP2) to VT, causes a release from VH (20, 21). In addition, talin peptides corresponding to each of the three vinculin binding domains in talin also prevent VH-VT interactions (22–24).

The structure of the whole vinculin molecule is still unknown. However, the crystal structure of the VT has been solved (25), and most recently, co-crystals of human VT (amino acids 879–1066) and a head fragment (amino acids 1–258) have been analyzed (23). The VT comprises five amphipathic helices that are connected by short flexible loop regions and form an antiparallel bundle. The C-terminal region extends across the base of the bundle and emerges as a flexible loop followed by a {beta} strand and a hydrophobic hairpin. It is believed that this "hydrophobic finger," which is surrounded by a basic collar formed by the base of the helical bundle, mediates the initial binding to acidic phospholipids and inserts into the lipid membrane (25). Furthermore, binding of acidic phospholipids unfurls the helical bundle, which leads to the exposition of cryptic oligomerization sites. Conserved residues in helices 3–5 of VT form a hydrophobic interface interacting with the VH. The latter contains seven amphipathic {alpha}-helices that are arranged as two helical bundles (23). Binding of a talin peptide triggers conformational changes that distort the VH-VT interface, thus displacing the tail from the head domain.

Taken together these data reveal vinculin to be a highly flexible structure in both head and tail domains, and conformational changes seem to be the key event in vinculin activation. Furthermore, they suggest a two-step activation model of vinculin in which the first step initiates the release of the head-tail interaction either by phospholipids or talin binding (20–22, 24), whereas the second step comprises a conformational change of the tail domain (25).

In vertebrate smooth and heart muscle, the larger splice variant metavinculin has been found co-expressed with vinculin, albeit at different molar ratios (26–28). Both proteins differ solely by an acidic insert of 68 amino acids in man and mouse that is positioned between helices I and II of the tail domain (29). Although metavinculin expression is generally high in smooth muscle (26, 27, 30), varying results have been published for the heart, ranging from only trace amounts to 18% of total vinculin protein expressed (26, 27). Immunofluorescence analyses indicated that both vinculin isoforms co-localize in muscular adhesive structures, such as dense plaques, intercalated discs, and costameres (31). However, the cellular function of metavinculin remains elusive. So far, no metavinculin-specific ligand has been identified, but there is strong evidence that the metavinculin-insert has a modulatory influence on the interaction with ligands of the tail domain. For example, the tail domains of vinculin and metavinculin (MVT) display different F-actin bundling properties (32), and a recently identified novel ligand protein, raver1, shows a higher affinity for metavinculin than for vinculin (33). Because neither binding to actin nor to raver1 directly involves the metavinculin insert, these data argue for a more complex effect of the insert, possibly on the conformation of the vinculin molecule. A specific function of metavinculin in the anchorage of actin filaments in muscle is suggested by the observation that mutations or complete loss of human metavinculin correlate with hereditary dilated cardiomyopathies (34, 35), concomitant with a disruption of intercalated discs in the afflicted patients (35).

The present study presents a comparative investigation of murine vinculin and metavinculin. We show that in contrast to previous reports, metavinculin is not only expressed in smooth and cardiac muscle but also in skeletal muscle, where it also co-localizes with vinculin in costameres, the adhesion sites that anchor sarcomeric Z-lines to the sarcolemma. In vitro data and analyses from muscle extract precipitates suggest the existence of heterodimers. However, vinculin and metavinculin significantly differ in those biochemical properties that regulate intramolecular interactions as well as intermolecular complex formation with different ligands. In summary, our data support a model for a differential activation of both vinculin isoforms in which metavinculin may require vinculin as a coactivator.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Cloning of Vinculin and Metavinculin Constructs—Cloning of human and murine tail domains of vinculin (VT) and metavinculin (MVT) comprising amino acids 858–1066 has already been reported (35). All further truncated tail constructs were generated accordingly by PCR using these constructs as a template. A list of all constructs is given in Fig. 1. Amplification primers introduced 5' EcoRI and 3' SalI restriction sites for further cloning. The two metavinculin mutants, MVT-2NQ and MVT-3N3Q, were generated by site-directed mutagenesis according to the manufacturer's instructions (QuikChange kit; Stratagene, Heidelberg, Germany). Glutamic and aspartic acid residues within the metavinculin insert were mutated to glutamine and asparagine, respectively. In the MVT-2NQ mutant 3 residues (Asp-29, Asp-33, and Glu-37 of the metavinculin insert) were replaced, whereas 6 residues were altered in the MVT-3N3Q mutant by an additional replacement of Glu-31, Asp-32, and Glu-35. For the generation of the MVT mutants the following primer pairs were used (5'-3'): 2NQ sense, TTCCCCTCTAACATGGAAGACAATTACGAACCTCAGCTGCTG; 2NQ reverse, CAGCAGCTGAGGTTCGTAATTGTCTTCCATGTTAGAGGGGAA; 3N3Q sense, TTCCCCCTCAACATGCAAAACAATTACCAACCTCAGCTGCTG; 3N3Q reverse, CAGCAGCTGAGGTTGGTAATTGTTTTGCATGTTAGAGGGGAA.



View larger version (33K):
[in this window]
[in a new window]
  		FIG. 1.
Bar diagrams of all recombinant VT and MVT proteins used in this study. The vinculin tail domain (VT) contains 5 helices that are represented by gray boxes and labeled 1–5. The C terminus symbolizes a hairpin structure (25). The metavinculin insert is positioned between the helices 1 and 2 of VT and is depicted as a black box. All constructs were equipped with an N-terminal histidine tag (black oval) and a FLAG tag (black circle). For simplification and better comparison, numbering of all deletion constructs was based on the positions of the first and last amino acids of VT, not counting the metavinculin-specific amino acids of the insert. MVT-2NQ and MVT3N3Q, two mutants with an exchange of three or six acidic residues, respectively, to asparagine or glutamine, as indicated by the corresponding number of white bars (for details see "Experimental Procedures"). The calculated pI values demonstrate the influence of the metavinculin-specific insert on the net charge of the vinculin tail.

 
All deletion constructs were numbered in their amino acid content with respect to vinculin. Metavinculin constructs containing the 68-amino acid insert are named MVT, and the vinculin tail constructs are labeled VT (e.g. VT858–1052 and MVT858–1052 for vinculin and metavinculin constructs, respectively, lacking the C-terminal hairpin). All constructs were sequenced before further cloning into the prokaryotic expression vector pQE30 (Qiagen, Hilden, Germany).

Protein Expression and Purification—All His-tagged proteins were purified from Escherichia coli strain M15 according to the manufacturer's instructions (Qiagen), with slight modifications (35). Protein was eluted with 150 mM histidine in the elution buffer (50 mM Na2HPO4, pH 6.7, 100 mM KCl, 0.5 mM EGTA, 25 mM EDTA, 0.1% Triton X-100, 1 mM 4-(2-aminoethyl)benzenesulfonyl fluoride (freshly made), 2 µM pepstatin A, 20 µM leupeptin, Trasylol® (2.8 mg/ml in H2O) 14.3 mM {beta}-mercaptoethanol). All fractions were analyzed by SDS-PAGE. Fractions containing the protein of interest were dialyzed against storage buffer (50 mM sodium phosphate buffer, pH 7.2, containing 0.2 mM EGTA, protease inhibitors, and {beta}-mercaptoethanol). Proteins were stored on ice. Rabbit skeletal muscle actin was prepared from acetone powder (36) with an additional gel filtration step as described (37).

Full-length vinculin and metavinculin were purified from chicken gizzard as described (30, 38, 39). The head domain (VH) was obtained from native chicken gizzard vinculin (in 10 mM Tris-Cl, pH 7.4, 100 mM KCl, 1 mM EDTA, 0.3 mM dithiothreitol) after cleavage with V8 protease coupled to agarose beads (Sigma). Using 2 units of V8 per mg of vinculin, cleavage was achieved after 2 h at 37 °C under agitation. Subsequent ion exchange chromatography (Mono Q, Amersham Biosciences) yielded pure vinculin head fractions.

In Vitro Chemical Cross-linking—Chemical cross-linking of recombinant VT, MVT, and VH was performed essentially as described (13). Purified (M)VT alone with VH or mixtures of both tail domains (100 pmol, 3.5 µM) were incubated in the presence or absence of lipids at 37 °C for 30 min. Cross-linking of proteins was achieved by incubating the proteins or protein mixtures with chemical cross-linkers at 30 °C for 5–30 min in NaH2PO4, pH 7.2, 0.2 mM EGTA, 7 mg/ml N-hydroxysulfosuccinimide (NHS, Pierce), and 1.5 mg/ml 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide (EDC, Pierce). The reactions were stopped by adding 4x SDS-PAGE sample buffer (375 mM Tris, pH 6.8, 2% SDS, 12% glycerol, 0.5 M {beta}-mercaptoethanol, 0.025% bromphenol blue). Samples were analyzed by SDS-PAGE and immunoblotting as described (13).

Cosedimentation Analysis—Co-sedimentation assays were performed essentially as described (35) with slight modifications of the buffer conditions. 30 µM actin was pre-polymerized in assay buffer (10 mM imidazole, pH 7.4, 2 mM CaCl2, 0.2 mM dithiothreitol, 1 mM ATP) supplemented with 100 mM KCl for 1 h at 37 °C. 3 µM F-actin was incubated for 2 h with 100 pmol or genuine full-length vinculin and metavinculin, respectively, dialyzed against assay buffer (final concentration 660 nM) in the absence or presence of a 10–100-fold molar excess of a talin peptide (VBS3) containing the third vinculin binding site of talin (22). After high speed centrifugation (100,000 x g, 60 min in an Airfuge; Beckman Instruments) pellets and supernatants were subjected to SDS-PAGE, and Coomassie Brilliant Blue-stained gels were analyzed densitometrically (E.A.S.Y. RH apparatus, E.A.S.Y. Image Plus Software, Herolab, Wiesloch, Germany).

Lipid Preparation and Sucrose-loaded Vesicle Pull-down Assay— Binding of the recombinant (M)VT proteins to phospholipids was analyzed in a sedimentation assay with sucrose-loaded lipid vesicles. Sucrose-loaded vesicles were prepared essentially as described (40, 41). Briefly, phosphatidylcholine (PC), phosphatidylserine (PS), and PIP2, or mixtures thereof were dried under liquid nitrogen and resuspended at 1 mg/ml in 40 mM Tris, pH 7.5, 100 mM NaCl, 3 mM MgCl2, 500 mM sucrose. To achieve sucrose enclosure, the mixtures were subjected to several freeze-thaw cycles. To ensure the formation of unilamellar lipid vesicles of similar size (pore size, 100 nm), the lipid preparations were extruded according to the manufacturer's prescription (Avanti Polar Lipids, Inc.). The sucrose loaded vesicles were centrifuged in an Airfuge (Beckman) at 100,000 x g for 30 min, and the pellet was resuspended in 40 mM Tris, pH 7.5, 100 mM NaCl, 3 mM MgCl2, yielding a concentration of 1 mg/ml. The proteins were diluted in the same buffer (5 µM final concentration) and incubated (15 min, 37 °C) with the phospholipid vesicles (0.5 mg/ml final concentration). After incubation, reaction mixes were centrifuged (20,000 x g for 15 min at room temperature). The supernatant was subjected to a methanol/chloroform precipitation. Pellet and supernatant fractions were analyzed by SDS-PAGE and densitometric analysis of the Coomassie-stained gels as described above.

Overlay Assay—Heterodimer formation was analyzed in dot overlays. Increasing amounts (1.6, 3.2, and 6.4 pmol) of recombinant VT were spotted onto a nitrocellulose membrane without or after preincubation with 10-fold molar excesses of PIP2. After blocking the membrane for 1 h at room temperature using Tris-buffered saline, pH 7.4, supplemented with 5% lowfat milk powder and 0.5% Tween 20, the membrane was incubated with 95 µM recombinant MVT, which in some assays had also been preincubated with a 10-fold molar excess of PIP2. Bound MVT was detected by a monoclonal antibody specific for metavinculin (see under "Antibodies and Immunochemistry") using a monoclonal peroxidase-conjugated secondary antibody.

Surface Plasmon Resonance Studies—Kinetic analyses of head-tail interactions were performed using a Biacore 2000 unit with CM5 sensor chips. Flow cells of the sensor chip surface (fc2) were coated with 1 ng (1000 resonance units) of the 90-kDa VH diluted in a 10 mM sodium acetate buffer, pH 4.5, using the amine coupling kit (Biacore). All experiments were conducted in running buffer (10 mM Hepes, pH 7.4, 100 mM KCl, 3 mM EDTA, and 0.005% surfactant P20). Before the analysis the protein stock solutions were equilibrated with running buffer. Analytes (recombinant VT or MVT) were subjected to the surface in dilutions of 0.3/0.1/0.03/0.01 mg/ml with a flow rate of 30 µl/min. Bound analytes (3min [PDB] /injection) were allowed to dissociate for 15 min and were finally removed by a 30-µl pulse of 1.5 M NaCl for 1 min. Evaluation of the data was performed with the BIAevalution software 3.2 following the reaction equation for a 1:1 reaction (A + B {Leftrightarrow} AB (dR/dt = ka x CA x (Rmax – R) – kd x R)) using a cross-linked, uncoated flow cell (fc1) for reference subtraction. Only measurements that could be fitted satisfactorily ({chi}2 values <10) were taken into account for calculation of association and dissociation rate constants as well as the dissociation constants for both tail domains. Three independent experiments for each tail domain were analyzed statistically (unpaired t test) using the StatView 5.0 software (SAS Institute Inc.).

Antibodies and Immunochemistry—Monoclonal mouse antibodies specifically recognizing metavinculin were generated following standard protocols using the recombinant MVT or synthetic peptides corresponding to different regions of the metavinculin insert as immunogen. The antibodies obtained were tested in solid phase binding assays (enzyme-linked immunosorbent assay) and immunoblots. Of these antibodies the clone 6E3 was employed in the present study, which recognizes amino acids 46–51 of the metavinculin insert.2

For detection of both vinculin splice variants, the monoclonal antibody hVin-1 (Sigma) was used, which recognizes the head domain of both proteins. Actin was detected with a polyclonal {alpha}-actin serum (Sigma). FLAG-tagged proteins were stained with a monoclonal FLAG-antibody (M2, Sigma) using peroxidase-coupled secondary antibodies.

For immunoprecipitation, murine skeletal muscle tissue was pulverized in liquid nitrogen immediately after dissection.1gof tissue powder was lysed and sonicated in 4 ml of ice-cold buffer A (20 mM Tris, pH 8.0, 100 mM NaCl, 0.5 mM EDTA) containing protease inhibitors. Insoluble material was removed by centrifugation (13,000 x g for 1 h at 4 °C). The cleared lysate was incubated for 4 h with a metavinculin-specific antibody (6E3) and an additional 2 h with protein G-Sepharose. Precipitated proteins were resolved by SDS-PAGE and analyzed by immunoblotting using the vinculin head-specific antibody hVin1 (Sigma) recognizing both vinculin isoforms.

Cryosectioning and Immunofluorescence Labeling—Freshly isolated murine muscle samples were embedded in Polyfreeze tissue-freezing medium (Polysciences Inc.) and frozen on dry ice. 5–15 µM longitudinal and transverse cryostat sections were cut at –20 °C. Sections were mounted on glass slides and dried for at least 30 min. The sections were washed with PBS, incubated with 1% Triton X-100 for 10 –20 s, and fixed with 4% formaldehyde for 20 min. After washing with PBS the sections were permeabilized with 0.2% Triton X-100 for 30 min at room temperature. Unspecific background labeling was reduced by incubation with 5% bovine serum albumin (filtrated) or with 0.25 mg/ml Fab fragments (goat anti-mouse IgG, Dianova) for 1 h at room temperature. For indirect immunofluorescence, the sections were washed in PBS and incubated with the monoclonal antibody 6E3 or the monoclonal {alpha}-vinculin antibody hVin1 (Sigma, 1:1000) for 1 h at room temperature. Sections were washed 3 times in PBS and then incubated with Alexa 568-conjugated goat {alpha}-mouse IgG (diluted 1:300 or 1:100 for conventional and confocal fluorescence microscopy, respectively). Samples from skeletal muscle were also counterstained for F-actin with fluorescein isothiocyanate-labeled phalloidin (Sigma). The stained samples were extensively washed in PBS, mounted in Mowiol, and examined with a Zeiss Axiophot microscope equipped with epifluorescence optics. Confocal microscopy was performed with a CLSM-510META confocal laser-scanning microscope (Carl Zeiss, Göttingen, Germany) using a Plan-Neofluar 40x/1.3 oil differential interference contrast objective. Excitation wavelengths were 488 and 543 nm for fluorescein isothiocyanate and Alexa 586, respectively, and a primary beam splitting mirror (UV/488/543/633) was used. Emitted light was detected with 505–530-nm and 560–651-nm band pass filters. Samples were analyzed with the multitracking mode. The bright field of samples was taken using the transmitted light photomultiplier.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Metavinculin and Vinculin Are Differentially Expressed in Smooth and Striated Mouse Muscle—There are only few quantitative data on vinculin and/or metavinculin expression in the literature. Although all studies agree on a high level of metavinculin in smooth muscle from various organisms (26–28), quite different expression levels have been published for metavinculin in heart and skeletal muscle for man and chicken (26, 27). Hence, we first determined the vinculin and metavinculin levels in different mouse muscles using quantitative immunoblotting (Fig. 2). Tissue extracts from uterus and heart (Fig. 2A) fast (musculus gastrocnemius) and slow (musculus soleus) skeletal muscles (Fig. 2B) were standardized with respect to actin and probed with a monoclonal vinculin antibody, which recognizes both vinculin isoforms. The results of three independent experiments are shown as bar diagrams in Fig. 2C. As expected, metavinculin expression in smooth muscle was high, representing about 30–40% of the total vinculin detected in uterine muscle. In cardiac muscle, the metavinculin content was surprisingly low, comprising less than 3% of the total vinculin. In contrast to previous reports, metavinculin was also detected in both fast and slow skeletal muscle, albeit at different expression levels. In fast skeletal muscle, metavinculin accounted for up to 15–25% of the total vinculin, whereas the expression in slow skeletal muscle was significantly lower (5%). It is noteworthy that although smooth and cardiac muscle differ substantially in the metavinculin content, the total amount of vinculin in relation to actin was fairly similar and found three times as high compared with skeletal muscle.



View larger version (28K):
[in this window]
[in a new window]
  		FIG. 2.
Differential expression of metavinculin in smooth and striated murine muscle. Vinculin, metavinculin, and actin contents were analyzed by quantitative immunoblots in tissue extracts from different murine muscles using standard curves obtained with purified vinculin and actin. S1–S4, aliquots of purified vinculin (25/50/100/150 ng) or actin (10/20/40/60 ng), respectively. These standards, both vinculin isoforms, and actin in the extracts were detected with the vinculin head-specific antibody (upper rows) or with a polyclonal actin antiserum (bottom rows). Densitometry revealed that the immunoblots of the purified proteins (lanes 1–4 in A and B) yielded values within the linear region of a standard curve and could, thus, be used for quantitation of the relevant protein amounts in the extracts. A, analysis of uterine smooth muscle (lanes 5, 7, and 7') and heart muscle (lanes 6, 8, and 8') extracts. For both extracts two different amounts of proteins were loaded (lanes 5 and 7 for uterus, lanes 6 and 8 for heart), so that the signals for vinculin as well as actin bands were within the standard curves. Lanes 7' and 8', images obtained for lanes 7 and 8 after longer exposure, revealing the high expression of metavinculin in mouse uterus and the low expression in heart. B, corresponding analysis of these proteins in extracts of fast (musculus gastrocnemius) and slow (musculus soleus) skeletal muscle. Longer exposure of the samples in lanes 7 and 8 reveals a weak expression of metavinculin in slow (lane 7') but a high expression in fast (lane 8') skeletal muscle. C, bar diagrams representing relative amounts of vinculin (white bars) and metavinculin (gray bars) with respect to the actin content, as obtained from quantitative immunoblot analyses as demonstrated in A and B from three independent experiments. The bars represent mean values with S.D. indicated. Note that the total amount of vinculin is much higher in smooth and cardiac muscle than in skeletal muscle and that metavinculin accounts for 20–30% of total vinculin in smooth and fast skeletal muscle, whereas it is only weakly expressed in cardiac and slow skeletal muscle.

 
Metavinculin Is a Structural Component of Adherens Junctions in Smooth and Striated Mouse Muscle—We next analyzed the subcellular localization of both vinculin isoforms in different muscle types. To distinguish between the two vinculin isoforms, a monoclonal antibody was employed that is specifically directed against amino acids 46–51 of the metavinculin insert. Reactivity and specificity of this antibody are described elsewhere.2 The combined distribution of vinculin and metavinculin was visualized by a monoclonal antibody recognizing both isoforms. In smooth muscle of the urinary bladder, metavinculin colocalized with vinculin in dot-like structures that probably represent the smooth muscle-specific microfilament membrane adhesion structures called dense plaques, which are organized in rib-like arrays (Fig. 3A, upper and lower panels) as had previously been described for human smooth muscle (42, 43). In cardiac muscle, the reactivity seen with the antibody directed against both vinculin isoforms was confined to intercalated disks and to punctate delineations of the cell membrane, reflecting the peripheral portion of the Z-lines (Fig. 3B, upper panels). In contrast, the monoclonal antibody specific for metavinculin yielded only very low signals (Fig. 3B, lower panels), consistent with the notion that metavinculin is a minor component in the murine heart (Fig. 2). In skeletal muscle, metavinculin-specific reactivity was observed to colocalize with the Z-line of the muscle fiber (Fig. 4A). Confocal sectioning revealed that in contrast to {alpha}-actinin, both vinculin isoforms were absent from the inner part of the Z-disks (Fig. 4B) but restricted to the peripheral costameres, which are engaged in anchoring the Z-line structures to the sarcolemma (Fig. 4C). These findings are consistent with previous reports on vinculin (31, 44, 45).



View larger version (119K):
[in this window]
[in a new window]
  		FIG. 3.
Localization of vinculin and metavinculin in smooth and cardiac muscle. A, immunofluorescence images of cryosections of murine urinary bladder, as obtained with the metavinculin-specific antibody (MV, top row) or the antibody recognizing both isoforms (V + MV, bottom row). Transversal (left panels) and longitudinal sections (right panels) are shown. Insets, higher magnifications of the boxed regions. Both antibodies react with specific dots ("dense plaques") arranged in a rib-like configuration, which is characteristic for smooth-muscle. B, analogous examples obtained for cardiac muscle showing fluorescence (left panels) and corresponding phase contrast (right panels) images. The signals for total vinculin are confined to intercalated disks and dense peripheral dots, probably representing cardiac costameres (top panels). In contrast, the metavinculin-specific antibody yields very low signals (lower panels). Bars, 10 µM.

 



View larger version (37K):
[in this window]
[in a new window]
  		FIG. 4.
Metavinculin and vinculin localize to costameres in skeletal muscle. A, confocal section of skeletal muscle stained for MV (left panel) and the corresponding bright field image, revealing that MV localizes to the Z-lines. Bar, 10 µm. B, confocal analysis of the subcellular localization of both vinculin isoforms as detected with the antibody hVin1. Upper panel, representative optical sections (1-µm intervals) of a skeletal muscle fiber oriented at an oblique angle to the optical plane. Note that a striation is only seen in the left half of the first two images, whereas in later sections only the periphery is stained. Lower panel, Z-axis scan of the same confocal Z-series (images taken at 0.33-µm intervals). The different planes (1–4) at which the analyses were performed are indicated by white lines on the large panel (representing a section (1 µm) from the upper panel). These data confirm that both vinculin isoforms localize to costameres and are absent from the inner part of the Z-disk. Bars, 10 µm. C, higher magnification of skeletal muscle stained for MV (upper left panel) and both vinculin isoforms (V + MV; lower left panel). Images were taken at a central plane of the fibers, revealing peripheral dot-like staining where the Z-disks attach to the sarcolemma. Conspicuously, phalloidin staining (middle panels) is absent from costamere staining as is demonstrated by the merged images (right panels). Bar, 5 µm.

 
In summary, using a metavinculin-specific antibody, we confirm previous reports on the colocalization of both vinculin isoforms in smooth muscle on the level of light microscopic analysis (26, 28, 46). In addition, we show for the first time that metavinculin is a regular component of skeletal muscle costameres, emphasizing the role of metavinculin in muscular microfilament anchorage.

Metavinculin and Vinculin Are Both Regulated by Intramolecular Head-Tail Interactions, but the Affinities of the Tail Domains for the Head Considerably Differ—As mentioned above, the initial step in vinculin activation is a conformational switch in which VH is released from the VT, thus unmasking binding sites for both head and tail ligands (12, 20, 21, 24). To elucidate whether the large acidic insert might alter the ability to change conformational shape, we analyzed the intramolecular association of VH with MVT and VT (Fig. 5). We first investigated the reconstitution of the whole molecules from proteolytically derived, isolated head domains and recombinant, tagged MVT or VT. The domain interaction was analyzed in the presence or absence of PIP2. Complex formation between heads and tails was revealed after cross-linking with EDC/NHS and SDS-PAGE in immunoblots with a monoclonal tag antibody (Fig. 5A). Reconstitution was observed by the appearance of bands at the Mr expected for full-length proteins. Both tail domains readily associated with the head domain, and complex formation of both vinculin isoforms was sensitive to the presence of PIP2. At a 10-fold excess of PIP2, complex formation was considerably reduced and, in the presence of a 20-fold molar excess, nearly abolished (Fig. 5A). These results suggest that vinculin and metavinculin head-tail interactions are regulated by PIP2 in a qualitatively similar manner despite the large acidic insert in MVT.



View larger version (22K):
[in this window]
[in a new window]
  		FIG. 5.
Both metavinculin and vinculin are regulated by head-tail interactions. A, complex formation between the isolated VH and tail fragments is PIP2-sensitive. 1 µM VH, obtained from vinculin after V8 cleavage was incubated with equal amounts of either tagged MVT (lanes 1, 3, and 5) or tagged VT (lanes 2, 4, and 6) in the absence (0) or presence (10 or 20 µM) of PIP2. The mixtures were then subjected to cross-linking with EDC/NHS and SDS-PAGE and immunoblotting with the tag antibody. Reconstitution of the full-length proteins is seen by the relevant position for both tail domains in the absence of PIP2, whereas increasing amounts of PIP2 inhibit head-tail interactions to a similar extent for both vinculin isoforms. B, F-actin co-sedimentation analysis by SDS-PAGE. Full-length V (upper panel) and MV (lower panel) at a final concentration of 0.66 µM each were incubated with prepolymerized muscle actin (A, 3 µM) in the absence (0) or presence of the talin peptide VBS3 (Ref. 22; 10–100-fold molar excess over metavinculin/vinculin) and subjected to centrifugation. Pellets (P) and supernatants (S) were analyzed by SDS-PAGE. In the presence of VBS3 the proportion of actin-bound pelletable metavinculin and vinculin increases substantially; it is slightly higher for MV at similar peptide concentrations (e.g. 32% for MV versus 20% for V at a 100-fold molar excess of the peptide). C, kinetic analyses of the head-tail association of vinculin and metavinculin using surface plasmon resonance. Isolated VH (1 ng) was immobilized on the chip surface. Complex formation was monitored using recombinant-tagged MVT or VT as analytes, applied in dilutions of 0.3, 0.1, 0.03, or 0.01 mg/ml with a flow rate of 30 µl/min. Representative fitted data are shown for a 1:1 complex according to the Langmuir surface adsorption model for VT (upper panel) and MVT (lower panel). The deduced association and dissociation rate constants and dissociation constants derived from three independent experiments (including S.E., see Table I) are shown in insets, revealing that MVT has a lower affinity to VH as compared with VT. RU, resonance units; Resp. Diff., response difference.

 
To investigate the influence of a head-specific ligand on the opening of the metavinculin/vinculin isoforms, we employed a talin peptide (VBS3) harboring the third VH binding site and capable of opening the closed vinculin molecule (22, 23). For this assay (Fig. 5B), we used full-length metavinculin and vinculin and probed their binding to actin filaments as an indicator of the open conformation, with the actin binding site in the tail (11, 12, 20, 21) accessible. Cosedimentation with F-actin was performed in the presence of increasing amounts of the talin peptide. Pellet and supernatant fractions obtained after high speed centrifugation (100,000 x g) were subsequently analyzed by SDS-PAGE. In the absence of the talin peptide, no significant amounts of metavinculin or vinculin were recovered in the pellet fraction, suggesting that indeed both molecules were in the closed, inactive conformation. In the presence of increasing amounts (molar excess 0–100) of the talin peptide, up to ~20 and 32% of vinculin and metavinculin, respectively, cosedimented with F-actin (Fig. 5B and data not shown), indicating that a significant fraction of both vinculin and metavinculin molecules had been opened by the peptide, in agreement with previous reports for vinculin (22).

Both the phospholipid and the actin binding assays had suggested that the principal activity regulation is similar for both isoforms. However, metavinculin seemed to be more easily activated by the talin peptide, as indicated by the higher percentage of bound protein present in the pellet as compared with vinculin. Because this could at least in part be due to different affinities of the tail domains for the head, we analyzed complex formation of VH with VT and MVT by surface plasmon resonance (Fig. 5C). Proteolytically obtained VH was immobilized to the carboxymethylated sensor chip surface and incubated with recombinant MVT or VT in solution. Best fit ({chi}2 < 10) was obtained when a 1:1 complex, according to the Langmuir surface adsorption model, was assumed for the experimental data set. Three independent measurements for each tail domain were performed, and association (ka) and dissociation (kd) rate constants as well as the dissociation constants (KD) were calculated, as shown in Table I. These values show that the association MVT to VH (ka = 1.50 ± 0.5 x 104) was about 9 times slower as compared with VT (ka = 1.34 ± 0.5 x 105), whereas the dissociation rates were nearly equal (kd = 4.13 ± 1.0 x 10–3 for MVT and 5.06 ± 0.1 x 10–3 for VT). Consequently, the apparent affinity of MVT for VH (KD = 3.36 ± 0.9 x 10–7) was significantly lower (p = 0.035) than the apparent affinity of VT (5.07 ± 1.7 x 10–8), the latter in good agreement with previous analyses (16). Hence, the metavinculin insert significantly affects the intramolecular association with VH.


View this table:
[in this window]
[in a new window]
  		TABLE I
Kinetics of complex formation of VH with VT and MVT

 
Metavinculin and Vinculin Tails Differ Quantitatively in PIP2 Binding and Oligomerization Ability—The phospholipid binding properties of MVT and VT fragments were compared in a pull-down assay using sucrose-loaded phospholipid vesicles (Fig. 6A). To mimic the phospholipid composition of the plasma membrane, sucrose-loaded vesicles were prepared with PC, PS, and PIP2 at a molar ratio of 6 PC:3 PS:1 PIP2. We also tested vesicles with a higher relative content of acidic phospholipids (4.5 PC:4.5 PS:1 PIP2) and neutral phospholipid vesicles containing only PC. Recombinant MVT and VT were incubated with the vesicles and subjected to centrifugation. Supernatants and pellets were analyzed by SDS-PAGE. As expected from previous reports (20, 21, 47, 48), VT only bound to lipid vesicles containing acidic phospholipids but not to PC-vesicles. Furthermore, increasing the relative amount of acidic phospholipids markedly improved VT binding to the vesicles (Fig. 6A, upper panel). In contrast, no binding of MVT was observed for any vesicle type in this assay (Fig. 6A, lower panel), showing that phospholipid binding by metavinculin was significantly impaired.



View larger version (38K):
[in this window]
[in a new window]
  		FIG. 6.
MVT and VT differ in their binding to acidic phospholipids and oligomerization properties. A, association of MVT and VT with phospholipids. SDS-PAGE analysis of MVT and VT (5 µM final concentrations) incubated with sucrose-filled lipid vesicles. Lipid composition was varied as indicated (PC, PS, PIP2). After centrifugation, pellets (P) and supernatants (S) were analyzed in SDS-PAGE. Note that VT co-sedimented with lipid vesicles containing acidic phospholipids, whereas MVT remained in the supernatant irrespective of the lipid composition used. Neither construct bound to PC vesicles. B, immunoblots obtained with the FLAG-tag antibody after SDS-PAGE, revealing oligomerization of tagged MVT (right) and VT (left, 5 µM concentration each) after cross-linking in the absence (0) or presence of increasing amounts of PIP2 (given as molar excess (mol. exc.)). Note the reduction in PIP2-induced dimers of MVT (MVT2) as compared with VT (VT2). In addition, there are no MVT oligomers, whereas for VT complexes at the position expected for trimers could be detected (VT3, open arrowhead).

 
In addition to releasing the head-tail interaction in vinculin, the binding of acidic phospholipids to VT has been suggested to unfurl the helical bundle of the tail, thereby exposing cryptic oligomerization sites (25). In vitro, oligomerization of vinculin via its tail domain requires the presence of acidic phospholipids, and in cells such oligomers are supposed to multiply the interaction sites with other adhesion proteins, thus enforcing cellular attachment (13). To obtain information on the putative role of metavinculin in muscular attachment sites, we therefore compared the oligomerization capacity of both isoforms (Fig. 6B). Identical amounts of recombinant MVT or VT were incubated in the presence of increasing amounts of PIP2. After chemical cross-linking with NHS/EDC, samples were resolved by SDS-PAGE and analyzed by immunoblotting, again using the tag antibody. Although VT dimerization was readily observed at a low to moderate molar excess of PIP2, MVT dimer formation was greatly impaired, and only a fraction of the molecules appeared at the position of the dimer even in the presence of a 20-fold molar excess of the phospholipid. In addition, trimers were only observed for VT but never for MVT.

The Acidic Tail Insert Causes Reduced Phospholipid Binding of Metavinculin—To analyze whether the impaired phospholipid binding of MVT is related to the negative charges within the metavinculin insert, two mutants were constructed in which aspartic and glutamic acid residues were replaced by asparagine and glutamine, respectively. The structure of the metavinculin insert remains unknown, but computational sequence analysis predicts three {alpha}-helical regions (35). All mutations were positioned in the putative helix 2, where the acidic residues could form clusters on two sides of the helix. In the mutant MVT-2NQ, three residues were replaced, whereas six residues in total were altered in the mutant MVT-3N3Q (see "Experimental Procedures"). The recombinant proteins were incubated with PIP2 and subjected to cross-linking and SDS-PAGE (Fig. 7), similar to the experiments described above for the wild type MVT and VT. Both mutants showed a significantly improved dimerization in the presence of PIP2 (Fig. 7, A and B). Although in the MVT-2NQ mutant dimerization was moderate (Fig. 7A), the self-association of the MVT-3N3Q mutant was almost comparable with that of VT (compare Fig. 7B with 6B, left panel), including trimer formation at higher PIP2 concentrations. Lipid vesicle pull-down assays revealed that the replacement of acidic residues also resulted in an improved binding to phospholipid vesicles (data not shown), suggesting that the diminished binding of MVT to phospholipids is to a significant extent based on weakening electrostatic interactions.



View larger version (52K):
[in this window]
[in a new window]
  		FIG. 7.
The reduced oligomerization capacity of MVT depends on the negative charge of the metavinculin insert. Immunoblots of tagged MVT mutant proteins and C-terminal-truncated deletion fragments of MVT and VT after incubation with PIP2 and cross-linking as indicated followed by SDS-PAGE. Protein bands were detected by the FLAG antibody. A and B, dimer and oligomer formation of the mutant MVT-2NQ, where three acidic residues of the metavinculin insert were replaced (A), and of the mutant MVT-3N3Q, where six acidic residues were mutated (B). mol. exc., molar excess. Aspartate was replaced by asparagine, whereas glutamate was substituted by glutamine (for details, see "Experimental Procedures"). Note that PIP2-dependent MVT oligomerization capacity was markedly improved by neutralizing such acidic patches within the metavinculin insert, as compared with intact MVT (cf. Fig. 5B). C, MVT and VT mutants lacking the C-terminal hairpin structure (MVT858–1052 and VT858–1052) were cross-linked in the absence or presence of a 10-fold molar excess of PIP2. Both tail domains formed dimers only after incubation with PIP2, indicating that phospholipid binding still occurs, presumably to helix 3. The latter is sufficient for dimerization and not impaired by the metavinculin insert. D, shorter deletion constructs comprising helices 1–3 only (VT858–975– and MVT858–975–) were analyzed under similar conditions as in C. Both fragments readily formed dimers even in the absence of PIP2, suggesting that PIP2 binding to helix 3 does not directly mediate dimerization but induces conformational changes necessary for oligomer formation.

 
The Metavinculin Insert Impairs PIP2 Binding to the C-terminal Hairpin of the Tail—Phospholipid binding to VT is believed to be a two-step procedure. The initial interaction with acidic phospholipids is apparently mediated by the C-terminal hairpin. Binding of phospholipids releases the hairpin from the base of the helical bundle, causing the latter to unfurl (25). Additional phospholipid binding has been reported for the {alpha}-helix 3 of VT (13, 48). To determine which of these interactions is influenced by the metavinculin insert and which phospholipid binding site confers oligomerization, deletion constructs of MVT and VT without the C-terminal hairpin (MVT858–1052 and VT858–1052, see Fig. 1) were analyzed for their oligomerization ability in the presence of PIP2. As shown in Fig. 7C, both deletion constructs displayed strong dimerization in the presence of a 10-fold molar excess of PIP2, reminiscent of wild type VT. Dimerization was also observed for deletion constructs comprising only the first three {alpha}-helices and was even stronger in the absence of PIP2 (Fig. 6D). These data reveal that the metavinculin insert does not prevent PIP2 binding to the site located in helix 3, which seems to induce conformational changes necessary for oligomer formation. Hence, oligomerization depends on PIP2 occupying this site. In contrast, the initial binding of PIP2 via the hydrophobic finger is significantly impaired by the metavinculin insert, preventing unmasking of the oligomerization sites.

The Metavinculin Tail Domain Forms Heterodimers with the Vinculin Tail Domain—As described above (Fig. 6B), MVT displays a reduction in PIP2-induced oligomerization as compared with VT. Because metavinculin seems always co-expressed with vinculin, it was of interest to analyze whether PIP2-activated VT could form heterodimers with MVT (Fig. 8). Increasing amounts of recombinant VT either untreated or after preincubation with PIP2 were membrane-immobilized and incubated with increasing amounts of untreated or PIP2-incubated MVT, and the bound MVT partner was detected by the metavinculin-specific antibody (Fig. 8A). In the absence of PIP2, no significant interaction occurred (Fig. 8A, left panel). Only weak signals were obtained when untreated VT was overlaid with PIP2-treated metavinculin tail (Fig. 8A, middle panel), suggesting that MVT retained some PIP2 binding capacity in this assay. In contrast, when the membrane-adsorbed VT had been preactivated with PIP2, enhanced MVT binding was observed, and the strength of the signal increased with the amount of adsorbed VT (Fig. 8A, right panel). An analogous result was obtained when both tail domains had been preincubated with PIP2 (data not shown). These data demonstrate a direct interaction of both tail domains requiring activation of VT by PIP2.



View larger version (46K):
[in this window]
[in a new window]
  		FIG. 8.
The formation of MVT/VT heterodimers is induced by binding of vinculin to PIP2. A, dot overlay with increasing amounts (1.6, 3.2, and 6.4 pmol) of recombinant VT immobilized on nitrocellulose membrane with or without preincubation with a 10-fold molar excess of PIP2. Spotted proteins were incubated with 3 µg/ml recombinant MVT in solution, again with or without preincubation with excess PIP2. Complex formation was monitored with the metavinculin-specific antibody. In the absence of PIP2, no interaction between both proteins was detected (left panel). Preincubation of MVT with PIP2 (center panel) resulted in weak signals from all samples. In contrast, preincubation of VT with PIP2 before membrane adsorption induced significant, concentration-dependent binding of MVT, indicating that a PIP2-induced conformational change in VT triggered heterodimer formation. B, immunoblots obtained with the tag antibody after SDS-PAGE of mixtures of tagged MVT and VT protein after NHS-EDC cross-linking to reveal oligomerization as a consequence of different MVT/VT ratios in the presence of a 10-fold molar excess of PIP2. The lower part of the figure represents the same sector marked in the upper blot after longer exposure. In the absence of VT, PIP2-induced dimerization of MVT is weak (lane 1, cf. also Fig. 6B). Heterodimers are formed upon the addition of increasing amounts of VT, starting with 20–30% VT in the mixture. At equimolar concentrations (lane 6) VT/MVT heterodimers are as prominent as VT dimers (VT2) but decline with decreasing amounts of MVT. Note that VT trimers (VT3) are only observed in the absence of MVT (lane 11). C, binding of MVT/VT heterodimers to sucrose-loaded lipid vesicles. Recombinant tail domains (2.5 µM each) were incubated with lipid vesicles of varying composition as indicated (PC, PS, PIP2). After 20,000 x g centrifugation of the protein-vesicle mixture pellets (P) and supernatants (S) were analyzed in SDS-PAGE. Note that the heterodimers co-sedimented with lipid vesicles containing acidic phospholipids in a similar fashion as VT alone (compare Fig. 6A) and were also unable to bind to PC vesicles. D, immunoblot analysis of precipitates from skeletal muscle extracts. Metavinculin was precipitated from skeletal muscle extracts with the metavinculin-specific antibody. Untreated muscle lysate (CE, left lane), the metavinculin precipitate (6E3, center lane), and a control where the precipitating antibody had been omitted (right lane) were analyzed for the presence of both vinculin isoforms with the hVin1 antibody. Note that approximately equal amounts of V are coprecipitated with MV.

 
The MVT/VT heterodimer formation was analyzed with respect to the ratio between both partners (Fig. 8B). Various mixtures similar to the different ratios of the corresponding proteins determined in muscle (cf. Fig. 2, A and B) were incubated with PIP2 in a 10-fold molar excess over the total amount of recombinant, tagged MVT + VT and subjected to chemical cross-linking and SDS-PAGE. The respective bands were visualized by the tag antibody in immunoblots and identified by their position in the gel. Again, no significant amount of MVT homodimers was observed, but with increasing amounts of VT, heterodimers appeared when the mixture contained at least 20% VT, and they were even observed with mixtures of molar ratios MVT:VT of 1:9 (Fig. 8B). These data confirmed the physical interaction of both isoforms and the preference of MVT for hetero- rather than homodimer formation as seen in overlay assays (Fig. 8A). Heterodimer formation was also suggested by the results of vesicle pull-down assays, carried out as described for Fig. 6A. Only in the presence of VT approximately equal amounts of MVT co-sedimented with lipid vesicles containing acidic phospholipids, as seen after SDS-PAGE of pellet and supernatant fractions (Fig. 8C).

First evidence of heterodimer formation in vivo was obtained from immunoprecipitation experiments with murine skeletal muscle extracts (Fig. 8D). When endogenous metavinculin from such extracts was precipitated with the metavinculin-specific antibody, vinculin was co-precipitated, as revealed by Western blot analysis. In such precipitates, both isoforms were present in approximately equal amounts, whereas in the extracts before to precipitation, vinculin was clearly predominant (Fig. 8D, left two lanes). This suggests the presence of heterodimers in muscle tissue.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
The present study provides a comparative analysis of the two microfilament-associated adhesion proteins vinculin and metavinculin, which derive from a single gene by alternative splicing. Although vinculin has been extensively studied in the past, little is known about the cellular function of the larger splice variant metavinculin that is characterized by an additional insert of 68 amino acids in man and mice that is positioned between helices 1 and 2 of the vinculin tail domain. Our data reveal that this insert significantly alters the ligand binding properties of the metavinculin tail domain, suggesting different modes of activation for both vinculin isoforms.

Our biochemical studies demonstrate that both vinculin and metavinculin are regulated by an intramolecular head-tail interaction, as had already been suggested from yeast two-hybrid studies (32). However, the binding kinetics of the respective tail domains to the head domain differ to a considerable extent. The 68-amino acid insert specifically lowers the rate of association of MVT with VH when compared with VT, whereas the dissociation rates are similar once the complexes have formed. This results in a significant decrease in affinity of MVT for VH, which possibly facilitates conformational changes between the open and the closed state in metavinculin. As can be concluded from recent structural analyses (23), the metavinculin insert, positioned between helices 1 and 2 of VT, is probably not directly involved in binding to VH. It may, however, affect the architecture of the helical bundle and, thus, slightly alter the VH-VT interface, which has been shown to be sensitive to conformational changes in VH (23, 24). Although the reduced affinity of MVT for VH is not sufficient to constantly keep the metavinculin molecule in the open conformation, as shown by the cosedimentation experiments (Fig. 5B), it may still contribute to metavinculin activation, possibly by counteracting the reduced binding to acidic phospholipids (see below).

The vinculin molecule is activated probably in two discrete steps involving initial release of VT from VH and subsequent changes in tail conformation. First, recruitment of vinculin to adhesion sites may be mediated either by binding of VH to talin (24) or through association of the C-terminal hydrophobic finger with acidic phospholipids (PIP2) in the plasma membrane (25), causing the release of the intramolecular association VH and VT, thus allowing binding of VH to talin, {alpha}-actinin, or {alpha}-catenin of adhesion complexes (9, 10, 17, 20). Second, the C terminus of VT, acting as a clamp at the base of the helical bundle, may undergo a conformational change upon binding to PIP2, releasing the helical bundle to unfold (25). This PIP2-induced conformational activation of VT is shown in the model depicted in Fig. 9A, upper panel. As a consequence, sites for actin binding as well as oligomerization that were previously mapped to helices 3 and 5 (13, 49) are unmasked.



View larger version (28K):
[in this window]
[in a new window]
  		FIG. 9.
Hypothetical model for the modulation of vinculin-mediated microfilament anchorage by metavinculin through heterodimer formation. A, differential conformational activation of vinculin and metavinculin tail domains. In VT, PIP2 binding to the hydrophobic finger of the C-terminal hairpin structure causes a release of the latter from the base of the helical bundle. This induces unfurling of the helical bundle and allows additional interaction of PIP2 with helix 3, leading to oligomer formation. In MVT the initial binding of PIP2 to the hydrophobic finger is impaired, with the hydrophobic finger and the helical bundle held in place by the metavinculin insert. A release of the helical bundle in MVT can be induced via heterodimerization with PIP2-activated VT. The five helices comprising the helical bundle and the hairpin structure including the hydrophobic finger at the very C terminus are depicted as solid lines. The metavinculin insert is shown as a black triangle at the base of the helical bundle where helices 1 and 2 connect. (B) Modulation of vinculin-mediated anchorage in muscle by metavinculin (1). In the absence of metavinculin, cytosolic V (head domain in gray) is recruited to and activated at the plasma membrane via talin or PIP2, leading to the incorporation of vinculin via the head domain to {alpha},{beta}-integrin-linked adhesion complexes (AC). After release of the tail domain from the membrane (possibly by phosphorylation) actin filaments are anchored to the AC via oligomerized tail domains. In the presence of MV (head domain in white) both vinculin isoforms may simultaneously be activated at the plasma membrane and could be incorporated as heterodimers to adhesion complexes. Additionally, metavinculin may be recruited to preexisting adhesion complexes by vinculin and remodel microfilament anchorage through heterodimer formation and inhibition of vinculin self-association into higher oligomers (3).

 
The data presented here suggest that PIP2-mediated activation of MVT differs from this model (Fig. 9A, lower panel). Both overall PIP2 binding and dimerization are reduced (Fig. 6). However, the metavinculin insert has no influence on phospholipid binding to helix 3 and permits oligomerization of metavinculin deletion constructs lacking the C-terminal hairpin (Fig. 7, C and D). Hence, we suggest that the negative charges of the insert mainly affect initial binding of PIP2 to the hydrophobic finger. Both hydrophobic and electrostatic interactions seem to contribute to PIP2 binding (25, 48); hydrophobic residues of the hydrophobic finger may insert into the lipid bilayer, whereas basic residues at the base of the helical bundle ("basic collar") are thought to interact with the PIP2 head groups (25). Interestingly, the loop between helices 1 and 2, where the metavinculin insert is positioned, seems to be intimately involved in both initial phospholipid binding and stabilization of the helical bundle. The metavinculin insert may, thus, either impair binding of the C terminus to phospholipids or confer additional stability to the helical bundle even in the presence of PIP2. In any case this depends at least in part on ionic interactions, since neutralizing negative charges in an acidic patch of the insert (amino acids 29–37) restores the phospholipid binding and oligomerization capacity of MVT comparable with that of VT (Fig. 7, A and B).

Although PIP2 cannot directly unfold MVT, it may nonetheless activate metavinculin indirectly through the induction of heterodimers by PIP2-activated VT (Fig. 8). The formation of MVT/VT heterodimers in vitro and in immunoprecipitation experiments from skeletal muscle extracts provide the first evidence for a possible direct interaction of both vinculin isoforms in metavinculin-expressing tissues. Because the presence of small amounts of MVT also prevents the formation of higher VT oligomers in vitro (Fig. 8B), one can imagine that heterodimerization in vivo may affect both the size and molecular architecture of vinculin/metavinculin-mediated adhesion complexes. Based on our data, we propose a model in which metavinculin is either recruited to nascent adhesion sites together with vinculin or may alternatively be incorporated into existing complexes through heterodimer formation with vinculin (Fig. 9B, see the figure legend for a detailed explanation).

The requirement for vinculin as a co-activator would explain why metavinculin is always co-expressed with vinculin. Furthermore, altering metavinculin expression levels in different muscle types may allow fine-tuning of microfilament organization, whereas the total amount of vinculin in relation to actin remains rather constant and is three times higher in murine smooth and cardiac muscle as compared with fast and slow skeletal muscle. In vitro, changing the molar ratio of VT:MVT significantly alters the organization and anchorage of actin filaments even if MVT is by far the minor component,3 but the relevance of this observation for the in vivo situation remains to be investigated.

Notably, in cells and tissues the metavinculin level is not an absolute value but responds to external stimuli. Thus, it has been shown recently that a change in mechanical load in rat skeletal muscle results in increased metavinculin expression (50). Moreover, it is known that smooth muscle cells in culture down-regulate the expression of metavinculin very rapidly (51), possibly because of the lack of external stimuli, such as mechanical stretching. Differential expression in response to mechanical stress has also been reported for other structural proteins such as filamin (52). On the other hand, the data available today point to species-specific differences in metavinculin expression. In cardiac muscle of man, metavinculin is apparently an important structural component in intercalated discs and costameres (31), and complete loss of or mutations in this vinculin isoform correlates with hereditary-dilated cardiomyopathy (34, 35), whereas metavinculin expression in cardiac muscle of mice and rats is extremely low (Fig. 2 and data not shown), and analogous data were reported for chicken heart (28). Such differences might correlate with a different mechanical challenge of cardiac muscle in various species related to differences in life span. In this context it is noteworthy that the metavinculin-linked cardiomyopathies in man occur as late-onset cardiomyopathies in adulthood (34, 35), suggesting that a deficiency in metavinculin is tolerated for quite some time but is not compatible with the mechanical stress a human heart is exposed to over decades. Our finding that metavinculin is a substantial component of the costameres of skeletal muscle, which is also exposed to severe although not continuous mechanical stress, is consistent with this assumption of an essential role for metavinculin in microfilament anchorage.

In summary, our data reveal that significant differences in the biochemical properties of vinculin and metavinculin exist that alter ligand binding and argue for a differential activation of both isoforms before incorporation into microfilament attachment sites. We provide biochemical evidence that PIP2-activated vinculin may act as a co-activator of metavinculin, consistent with the co-expression of both isoforms in muscle. It seems conceivable that metavinculin cannot be integrated into nascent or mature adhesion complexes without vinculin. The identification of heterodimeric complexes suggests that both isoforms cooperate in muscular adhesion sites, allowing fine tuning of microfilament attachment according to specific cellular demands.


   FOOTNOTES
 
* This study was supported by the German Research Council (to S. I. and B. M. J.), the Akademie Deutscher Naturforscher Leopoldina (to U. F.), and the Fonds der Chemischen Industrie (to B. M. J.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

{ddagger} To whom correspondence should be addressed: Cell Biology, Zoological Institute, Technical University of Braunschweig, Biocenter, Spielmannstrasse 7, D-38092 Braunschweig, Germany. Tel.: 49-531-391-3191; Fax: 49-531-391-8203.

1 The abbreviations used are: VH, vinculin head; VT, vinculin tail; PIP2, phosphatidylinositol 4,5-bisphosphate; EDC, 1-ethyl-3-[3-dimethylaminopropyl]carbodiimide; MV, metavinculin; MVT, metavinculin tail; NHS, N-hydroxysulfosuccinimide; PC, phosphatidylcholine; PS, phosphatidylserine; V, vinculin; PBS, phosphate-buffered saline. Back

2 S. Illenberger, S. Witt, S. Buchmeier, R. Frank, and B. M. Jockusch, manuscript in preparation. Back

3 S. Witt, A. Zieseniss, U. Fock, B. M. Jockusch, and S. Illenberger, unpublished observation. Back


   ACKNOWLEDGMENTS
 
We thank T. Messerschmidt and S. Buchmeier for expert technical assistance and Dr. W. Ziegler for helpful discussions (Technical University of Braunschweig) and Dr. D. Critchley (University Leicester, Leicester, UK) for the talin peptide. We also thank Dr. J. Wehland (German Research Centre for Biotechnology (GBF) Braunschweig, Germany) for kind permission to use the Biacore facility at the GBF Braunschweig and Dr. R. Hänsch (Technical University of Braunschweig) for help with the confocal laser scanning microscope.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Geiger, B., Tokuyasu, K. T., Dutton, A. H., and Singer, S. J. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 4127–4131[Abstract/Free Full Text]
  2. Xu, W., Baribault, H., and Adamson, E. D. (1998) Development 125, 327–337[Abstract]
  3. Rodriguez Fernandez, J. L., Geiger, B., Salomon, D., and Ben-Ze'ev, A. (1992) Cell Motil. Cytoskeleton 22, 127–134[CrossRef][Medline] [Order article via Infotrieve]
  4. Rodriguez Fernandez, J. L., Geiger, B., Salomon, D., and Ben-Ze'ev, A. (1993) J. Cell Biol. 122, 1285–1294[Abstract/Free Full Text]
  5. Coll, J. L., Ben-Ze'ev, A., Ezzell, R. M., Rodriguez Fernandez, J. L., Baribault, H., Oshima, R. G., and Adamson, E. D. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 9161–9165[Abstract/Free Full Text]
  6. Coutu, M. D., and Craig, S. W. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 8535–8539[Abstract/Free Full Text]
  7. Moulder, G. L., Huang, M. M., Waterston, R. H., and Barstead, R. J. (1996) Mol. Biol. Cell 7, 1181–1193[Abstract]
  8. DePasquale, J. A., and Izzard, C. S. (1991) J. Cell Biol. 113, 1351–1359[Abstract/Free Full Text]
  9. Watabe-Uchida, M., Uchida, N., Imamura, Y., Nagafuchi, A., Fujimoto, K., Uemura, T., Vermeulen, S., van Roy, F., Adamson, E. D., and Takeichi, M. (1998) J. Cell Biol. 142, 847–857[Abstract/Free Full Text]
  10. Weiss, E. E., Kroemker, M., Rudiger, A. H., Jockusch, B. M., and Rudiger, M. (1998) J. Cell Biol. 141, 755–764[Abstract/Free Full Text]
  11. Huttelmaier, S., Bubeck, P., Rudiger, M., and Jockusch, B. M. (1997) Eur. J. Biochem. 247, 1136–1142[Medline] [Order article via Infotrieve]
  12. Johnson, R. P., and Craig, S. W. (1995) Nature 373, 261–264[CrossRef][Medline] [Order article via Infotrieve]
  13. Huttelmaier, S., Mayboroda, O., Harbeck, B., Jarchau, T., Jockusch, B. M., and Rudiger, M. (1998) Curr. Biol. 8, 479–488[CrossRef][Medline] [Order article via Infotrieve]
  14. Brindle, N. P., Holt, M. R., Davies, J. E., Price, C. J., and Critchley, D. R. (1996) Biochem. J. 318, 753–757[Medline] [Order article via Infotrieve]
  15. DeMali, K. A., Barlow, C. A., and Burridge, K. (2002) J. Cell Biol. 159, 881–891[Abstract/Free Full Text]
  16. Johnson, R. P., and Craig, S. W. (1994) J. Biol. Chem. 269, 12611–12619[Abstract/Free Full Text]
  17. Kroemker, M., Rudiger, A. H., Jockusch, B. M., and Rudiger, M. (1994) FEBS Lett. 355, 259–262[CrossRef][Medline] [Order article via Infotrieve]
  18. Johnson, R. P., and Craig, S. W. (1995) Biochem. Biophys. Res. Commun. 210, 159–164[CrossRef][Medline] [Order article via Infotrieve]
  19. Laine, R. O., Zeile, W., Kang, F., Purich, D. L., and Southwick, F. S. (1997) J. Cell Biol. 138, 1255–1264[Abstract/Free Full Text]
  20. Gilmore, A. P., and Burridge, K. (1996) Nature 381, 531–535[CrossRef][Medline] [Order article via Infotrieve]
  21. Weekes, J., Barry, S. T., and Critchley, D. R. (1996) Biochem. J. 314, 827–832[Medline] [Order article via Infotrieve]
  22. Bass, M. D., Patel, B., Barsukov, I. G., Fillingham, I. J., Mason, R., Smith, B. J., Bagshaw, C. R., and Critchley, D. R. (2002) Biochem. J. 362, 761–768[CrossRef][Medline] [Order article via Infotrieve]
  23. Izard, T., Evans, G., Borgon, R. A., Rush, C. L., Bricogne, G., and Bois, P. R. (2004) Nature 427, 171–175[CrossRef][Medline] [Order article via Infotrieve]
  24. Izard, T., and Vonrhein, C. (2004) J. Biol. Chem. 279, 27667–27678[Abstract/Free Full Text]
  25. Bakolitsa, C., de Pereda, J. M., Bagshaw, C. R., Critchley, D. R., and Liddington, R. C. (1999) Cell 99, 603–613[CrossRef][Medline] [Order article via Infotrieve]
  26. Belkin, A. M., Ornatsky, O. I., Kabakov, A. E., Glukhova, M. A., and Koteliansky, V. E. (1988) J. Biol. Chem. 263, 6631–6635[Abstract/Free Full Text]
  27. Feramisco, J. R., Smart, J. E., Burridge, K., Helfman, D. M., and Thomas, G. P. (1982) J. Biol. Chem. 257, 11024–11031[Abstract/Free Full Text]
  28. Saga, S., Hamaguchi, M., Hoshino, M., and Kojima, K. (1985) Exp. Cell Res. 156, 45–56[CrossRef][Medline] [Order article via Infotrieve]
  29. Koteliansky, V. E., Ogryzko, E. P., Zhidkova, N. I., Weller, P. A., Critchley, D. R., Vancompernolle, K., Vandekerckhove, J., Strasser, P., Way, M., Way, M., and Gimona, M. (1992) Eur. J. Biochem. 204, 767–772[Medline] [Order article via Infotrieve]
  30. Gimona, M., Furst, D. O., and Small, J. V. (1987) J. Muscle Res. Cell Motil. 8, 329–341[CrossRef][Medline] [Order article via Infotrieve]
  31. Belkin, A. M., Ornatsky, O. I., Glukhova, M. A., and Koteliansky, V. E. (1988) J. Cell Biol. 107, 545–553[Abstract/Free Full Text]
  32. Rudiger, M., Korneeva, N., Schwienbacher, C., Weiss, E. E., and Jockusch, B. M. (1998) FEBS Lett. 431, 49–54[CrossRef][Medline] [Order article via Infotrieve]
  33. Huttelmaier, S., Illenberger, S., Grosheva, I., Rudiger, M., Singer, R. H., and Jockusch, B. M. (2001) J. Cell Biol. 155, 775–786[Abstract/Free Full Text]
  34. Maeda, M., Holder, E., Lowes, B., Valent, S., and Bies, R. D. (1997) Circulation 95, 17–20[Abstract/Free Full Text]
  35. Olson, T. M., Illenberger, S., Kishimoto, N. Y., Huttelmaier, S., Keating, M. T., and Jockusch, B. M. (2002) Circulation 105, 431–437[Abstract/Free Full Text]
  36. Spudich, J. A., and Watt, S. (1971) J. Biol. Chem. 246, 4866–4871[Abstract/Free Full Text]
  37. Giehl, K., Valenta, R., Rothkegel, M., Ronsiek, M., Mannherz, H. G., and Jockusch, B. M. (1994) Eur. J. Biochem. 226, 681–689[Medline] [Order article via Infotrieve]
  38. Feramisco, J. R., and Burridge, K. (1980) J. Biol. Chem. 255, 1194–1199[Abstract/Free Full Text]
  39. Siliciano, J. D., and Craig, S. W. (1986) Methods Enzymol. 134, 78–85[Medline] [Order article via Infotrieve]
  40. Darszon, A., Vandenberg, C. A., Schonfeld, M., Ellisman, M. H., Spitzer, N. C., and Montal, M. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 239–243[Abstract/Free Full Text]
  41. Saito, Y., Hirashima, N., and Kirino, Y. (1988) Biochem. Biophys. Res. Commun. 154, 85–90[CrossRef][Medline] [Order article via Infotrieve]
  42. Yamada, K. M., and Geiger, B. (1997) Curr. Opin. Cell Biol. 9, 76–85[CrossRef][Medline] [Order article via Infotrieve]
  43. Small, J. V., and Gimona, M. (1998) Acta Physiol. Scand. 164, 341–348[CrossRef][Medline] [Order article via Infotrieve]
  44. Pardo, J. V., Siliciano, J. D., and Craig, S. W. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 1008–1012[Abstract/Free Full Text]
  45. Danowski, B. A., Imanaka-Yoshida, K., Sanger, J. M., and Sanger, J. W. (1992) J. Cell Biol. 118, 1411–1420[Abstract/Free Full Text]
  46. Gimona, M., Small, J. V., Moeremans, M., Van Damme, J., Puype, M., and Vandekerckhove, J. (1988) EMBO J. 7, 2329–2334[Medline] [Order article via Infotrieve]
  47. Ito, S., Werth, D. K., Richert, N. D., and Pastan, I. (1983) J. Biol. Chem. 258, 14626–14631[Abstract/Free Full Text]
  48. Johnson, R. P., Niggli, V., Durrer, P., and Craig, S. W. (1998) Biochemistry 37, 10211–10222[CrossRef][Medline] [Order article via Infotrieve]
  49. Johnson, R. P., and Craig, S. W. (2000) J. Biol. Chem. 275, 95–105[Abstract/Free Full Text]
  50. Chopard, A., Pons, F., and Marini, J. F. (2002) Pfluegers Arch. Eur. J. Physiol. 444, 627–633[CrossRef][Medline] [Order article via Infotrieve]
  51. Gimona, M., Herzog, M., Vandekerckhove, J., and Small, J. V. (1990) FEBS Lett. 274, 159–162[CrossRef][Medline] [Order article via Infotrieve]
  52. Kainulainen, T., Pender, A., D'Addario, M., Feng, Y., Lekic, P., and McCulloch, C. A. (2002) J. Biol. Chem. 277, 21998–22009[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Cell Sci.Home page
T. Hijikata, A. Nakamura, K. Isokawa, M. Imamura, K. Yuasa, R. Ishikawa, K. Kohama, S. Takeda, and H. Yorifuji
Plectin 1 links intermediate filaments to costameric sarcolemma through {beta}-synemin, {alpha}-dystrobrevin and actin
J. Cell Sci., June 15, 2008; 121(12): 2062 - 2074.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Cell. Biol.Home page
A. E. Zemljic-Harpf, J. C. Miller, S. A. Henderson, A. T. Wright, A. M. Manso, L. Elsherif, N. D. Dalton, A. K. Thor, G. A. Perkins, A. D. McCulloch, et al.
Cardiac-Myocyte-Specific Excision of the Vinculin Gene Disrupts Cellular Junctions, Causing Sudden Death or Dilated Cardiomyopathy
Mol. Cell. Biol., November 1, 2007; 27(21): 7522 - 7537.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
279/30/31533    most recent
M314245200v1
Right arrow Submit a Letter to Editor
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Witt, S.
Right arrow Articles by Illenberger, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Witt, S.
Right arrow Articles by Illenberger, S.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
 All ASBMB Journals   Molecular and Cellular Proteomics 
 Journal of Lipid Research   ASBMB Today 
Copyright © 2004 by the American Society for Biochemistry and Molecular Biology.
Advertisement
spacer
Advertisement
Advertisement