Polyphosphoinositide micelles and polyphosphoinositide-containing vesicles dissociate endogenous gelsolin-actin complexes and promote actin assembly from the fast-growing end of actin filaments blocked by gelsolin

The Ca2+-activated actin-binding protein gelsolin regulates actin filament length by severing preformed filaments and by binding actin monomers, stabilizing nuclei for their assembly into filaments. Gelsolin binds to phosphatidylinositol 4,5-bisphosphate (PIP2), with consequent inhibition of its filament severing activity and dissociation of EGTA-resistant complexes made with rabbit macrophage or human plasma gelsolin and rabbit muscle actin. This study provides evidence for an interaction of gelsolin with phosphatidylinositol monophosphate (PIP) as well as PIP2 and further describes their effects on gelsolin's function. Both phosphoinositides completely dissociate EGTA-insensitive rabbit macrophage cytoplasmic gelsolin-actin complexes and inhibit gelsolin's severing activity. The magnitude of inhibition depends strongly on the physical state of the phosphoinositides, being maximal in preparations that contain small micelles of either purified PIP or PIP2. Aggregation of PIP or PIP2 micelles by divalent cations or insufficient sonication or their incorporation into vesicles containing other phospholipids decreases but does not eliminate the inhibitory properties of the polyphosphoinositides. The presence of gelsolin partly inhibits the divalent cation-induced aggregation of PIP2 micelles. PIP2 in combination with EGTA inactivates gelsolin molecules that block the fast-growing end of actin filaments, thereby accelerating actin polymerization. Regulation of gelsolin by the intracellular messengers Ca2+ and polyphosphoinositides allows for the formation of several different gelsolin-actin intermediates with distinct functional properties that may be involved in changes in the state of cytoplasmic actin following cell stimulation.     

Isolation of a Ca2+-Dependent Actin-Fragmenting Protein from Brain, Spinal Cord, and Cultured Neurones

Extracts of ox spinal cord and chicken brain were fractionated by ion-exchange chromatography and assayed for their ability to reduce the viscosity of muscle F-actin solutions. Two distinct peaks of activity were obtained, one of which was further purified by affinity chromatography on a DNAase-actin Sepharose column. Following molecular exclusion chromatography, the actin component appeared as a complex of 1 molecule of a protein with molecular weight 90,000 and 2 molecules of actin (42,000). This tightly bound complex was resistant to most methods of protein separation, but was resolvable into its component proteins by sodium dodecyl sulphate acrylamide gel electrophoresis. The protein of molecular weight 90,000 could be eluted from such a gel in a fully active form. The activity of the protein from ox spinal cord was closely similar to that of gelsolin, an actin-fragmenting protein originally isolated from rabbit lung macrophages. Like gelsolin, the protein from ox spinal cord produced fragmentation of muscle F-actin filaments at Ca2+ concentrations greater than 10(-7) M, and had a nucleating effect on the polymerisation of muscle actin; the latter was measured most easily by the enhancement of fluorescence of muscle actin conjugated to N-(1-pyrenyl)iodoacetamide. Nucleation was more effective in the presence of Ca2+, but also occurred in its absence, and the same was true of complex formation between the 90,000 protein and muscle G-actin. On the basis of its actin-fragmenting activity, we estimate that the 90,000 molecular weight protein constitutes 0.2% of the protein initially extracted from ox spinal cord. A very similar protein, indistinguishable in its action on actin but containing variable amounts of a protein of molecular weight 85,000 as well as 90,000, was isolated from chicken brain. A similar protein was also detected in pure cultures of sympathetic neurones by enrichment on a DNAase-actin affinity column and by immune blotting and by immunofluorescence. We conclude that a protein similar, if not identical to macrophage gelsolin is present in neurones and that it probably plays a part in the actin-based movements of these cells.     

Gelsolin is expressed in early erythroid progenitor cells and negatively regulated during erythropoiesis

We have identified an approximately 85-kD protein in chicken erythrocytes which is immunologically, structurally, and functionally related to the gelsolin found in many muscle and nonmuscle cell types. Cell fractionation reveals a Ca2+-dependent partitioning of gelsolin into the soluble cytoplasm and the membrane-associated cytoskeleton of differentiating or mature erythrocytes. Depending on either the presence of Ca2+ during cell lysis or on the preincubation of the intact cells with the Ca2+-ionophore A23187, up to 40% of the total cellular gelsolin is found associated with the membrane skeleton. Expression of gelsolin shows a strong negative regulation during erythroid differentiation. From quantitations of its steady-state molar ratio to actin, gelsolin is abundant in early progenitor cells as revealed from avian erythroblastosis virus- and S13 virus-transformed cells which are arrested at the colony forming unit erythroid (CFU-e) stage of erythroid development. In these cells, which have a rudimentary and unstable membrane skeleton, gelsolin remains quantitatively cytoplasmic, irrespective of the Ca2+ concentration. During chicken embryo development and maturation, the expression of gelsolin decreases by a factor of approximately 10(3) in erythroid cells. This down regulation is independent from that of actin, which is considerably less, and is observed also when S13-transformed erythroid progenitor cells are induced to differentiate under conditions where the actin content of these cells does not change. In mature erythrocytes of the adult the amount of gelsolin is low, and significantly less than required for potentially capping of all membrane-associated actin filaments. We suggest that the gelsolin in erythroid cells is involved in the assembly of the actin filaments present in the membrane skeleton, and that it may provide for a mechanism, by means of its severing action on actin filaments, to extend the meshwork of the spectrin-actin-based membrane skeleton in erythroid cells during erythropoiesis.     

A re-evaluation of cytoplasmic gelsolin localization

Gelsolin is a 90,000-mol-wt Ca2+-binding, actin-associated protein that can nucleate actin filament growth, sever filaments, and cap barbed filament ends. Brevin is a closely related 92,000-mol-wt plasma protein with similar properties. Gelsolin has been reported to be localized on actin filaments in stress fibers, in cardiac and skeletal muscle I-bands, and in cellular regions where actin filaments are known to be concentrated. Previous localization studies have used sera or antibody preparations that contain brevin. Using purified brevin-free IgG and IgA monoclonal antibodies or affinity-purified polyclonal antibodies for gelsolin and brevin, we find no preferential stress fiber staining in cultured human fibroblasts or I-band staining in isolated rabbit skeletal muscle sarcomeres. Cardiac muscle frozen sections show no pronounced I-band staining, except in local areas where brevin may have penetrated from adjacent blood vessels. Spreading platelets show endogenous gelsolin localized at the cell periphery, in the central cytoplasmic mass and on thin fibers that radiate from the central cytoplasm. Addition of 3-30 micrograms/ml of brevin to the antibodies restores intense stress fiber and I-band staining. We see no evidence for large-scale severing and removal of filaments in stress fibers in formaldehyde-fixed, acetone-permeabilized cells even at brevin concentrations of 30 micrograms/ml. The added brevin or brevin antibody complex binds to actin filaments and is detected by the fluorescently tagged secondary antibody. Brevin binding occurs in either Ca2+ or EGTA, but is slightly more intense in EGTA suggesting some severing and filament removal may occur in Ca2+. The I-band staining is limited to the region where actin and myosin do not overlap. In addition, brevin does not appear to bind at the Z-line. A comparison of cells double-labeled with fluorescein-phallotoxin, exogenous brevin, and a monoclonal antibody, detected with a rhodamine-labeled secondary antibody, shows almost complete co-localization of F-actin with the brevin-gelsolin-binding sites. A major exception is in the area of the adhesion plaque. A quantitative comparison of the fluorescein-rhodamine fluorescence intensities along a stress fiber and into the adhesion plaque shows that the fluorescein signal, associated with F-actin, increases while the rhodamine signal decreases. We infer that exogenous brevin or endogenous gelsolin can bind to and potentially sever most actin filaments, but that actin-associated proteins in the adhesion plaque can prevent binding and severing.(ABSTRACT TRUNCATED AT 400 WORDS)     

Smooth muscle gelsolin and a Ca2+-sensitive contractile cell model

Smooth muscle gelsolin, termed smooth muscle 90-kDa protein in our previous paper (Kanno et al. FEBS Lett. 1985; 184:202-206), was purified from bovine aorta. Antibody prepared against smooth muscle gelsolin was used to detect the presence of gelsolin in human lung fibroblast MRC-5 cells permeabilized with Triton X-100 (MRC-5 cell models). These cells contracted in the presence of MgATP and Ca2+ in doses over 1 microM. Immunofluorescence microscopy using phalloidin and antigelsolin antibody showed that gelsolin was distributed along the stress fibers, except for a marginal bundle of cells, when MRC-5 cells were growth-arrested in serum-depleted medium. Making use of immunoblotting and indirect immunofluorescence techniques, we demonstrated that gelsolin is not retained in the MRC-5 cell models. We used purified smooth muscle gelsolin as a specific agent to sever the actin filaments. Preincubation of MRC-5 cell models with gelsolin led to a destruction of stress fibers, in a dose- and Ca2+ -dependent manner. The contractility was also lost, in the same manner described above, thereby indicating that a continuous distribution of actin filaments within the stress fibers is required for cell contraction. Treatment of MRC-5 cells with the Ca2+ ionophore A23187 induced an extracellular Ca2+ -dependent contraction but not a massive destruction of stress fibers, thereby indicating that most of the endogenous gelsolin was inactive under these conditions. Our interpretation of these results is that increases in cytoplasmic Ca2+ concentrations are sufficient for the contraction but may be too transient to activate endogenous gelsolin and thereby disrupt the stress fibers. Indeed, the inhibition of contraction of the MRC-5 cell, as induced by smooth muscle gelsolin, required preincubation in the presence of Ca2+, before the addition of MgATP. These results suggest that destruction of the stress fibers by endogenous gelsolin, which leads to inhibition of cell contraction, may occur if the cytoplasmic Ca2+ is maintained at high concentrations for a few minutes.     

Muscle is the major source of plasma gelsolin

Gelsolin, a Ca2+- and polyphosphoinositide-regulated actin-binding protein, is unique among vertebrate proteins in being both cytoplasmic and secreted. Plasma gelsolin, present at greater than 200 micrograms/ml in human plasma, may have a protective function by promoting the clearance of actin filaments released during tissue injury. Although there is evidence that smooth muscle tissues and HepG2 cells synthesize plasma gelsolin, the predominant secretory source is hitherto unknown. We report here that skeletal, cardiac, and smooth muscles have large amounts of plasma gelsolin mRNA and devote 0.5-3% of their biosynthetic activity to plasma gelsolin, whereas liver makes relatively little. Since skeletal muscle accounts for a large fraction of body mass and total protein synthesis, it is the major source of plasma gelsolin.     

Functional comparison of villin and gelsolin. Effects of Ca2+, KCl, and polyphosphoinositides

A family of homologous actin-binding proteins sever and cap actin filaments and accelerate actin filament assembly. The functions of two of these proteins, villin and gelsolin, and of their proteolytically derived actin binding domains were compared directly by measuring their effects, under various ionic conditions, on the rates and extents of polymerization of pyrene-labeled actin. In 1 mM Ca2+ and 150 mM KCl, villin and gelsolin have similar severing and polymerization-accelerating properties. Decreasing [Ca2+] to 25 microM greatly reduces severing by villin but not gelsolin. Decreasing [KCl] from 150 to 10 mM at 25 microM Ca2+ increases severing by villin, but not gelsolin, over 10-fold. The C-terminal half domains of both proteins have Ca2+-sensitive actin monomer-binding properties, but neither severs filaments nor accelerates polymerization. The N-terminal halves of villin and gelsolin contain all the filament-severing activity of the intact proteins. Severing by gelsolin's N-terminal half is Ca2+-independent, but that of villin has the same Ca2+ requirement as intact villin. The difference in Ca2+ sensitivity extends to 14-kDa N-terminal fragments which bind actin monomers and filament ends, requiring Ca2+ in the case of villin but not gelsolin. Severing of filaments by villin and its N-terminal half is shown to be inhibited by phosphatidylinositol 4,5-bisphosphate, as shown previously for gelsolin (Janmey, P.A., and Stossel, T.P. (1987) Nature 325, 362-364). The functional similarities of villin and gelsolin correlate with known structural features, and the greater functional dependence of villin on Ca2+ compared to gelsolin is traced to differences in their N-terminal domains.     

Identification of critical functional and regulatory domains in gelsolin

Gelsolin can sever actin filaments, nucleate actin filament assembly, and cap the fast-growing end of actin filaments. These functions are activated by Ca2+ and inhibited by polyphosphoinositides (PPI). We report here studies designed to delineate critical domains within gelsolin by deletional mutagenesis, using COS cells to secrete truncated plasma gelsolin after DNA transfection. Deletion of 11% of gelsolin from the COOH terminus resulted in a major loss of its ability to promote the nucleation step in actin filament assembly, suggesting that a COOH-terminal domain is important in this function. In contrast, derivatives with deletion of 79% of the gelsolin sequence exhibited normal PPI-regulated actin filament-severing activity. Combined with previous results using proteolytic fragments, we deduce that an 11-amino acid sequence in the COOH terminus of the smallest severing gelsolin derivative identified here mediates PPI-regulated binding of gelsolin to the sides of actin filaments before severing. Deletion of only 3% of gelsolin at the COOH terminus, including a dicarboxylic acid sequence similar to that found on the NH2 terminus of actin, resulted in a loss of Ca2+-requirement for filament severing and monomer binding. Since these residues in actin have been implicated as potential binding sites for gelsolin, our results raise the possibility that the analogous sequence at the COOH terminus of gelsolin may act as a Ca2+-regulated pseudosubstrate. However, derivatives with deletion of 69-79% of the COOH-terminal residues of gelsolin exhibited normal Ca2+ regulation of severing activity, establishing the intrinsic Ca2+ regulation of the NH2-terminal region. One or both mechanisms of Ca2+ regulation may occur in members of the gelsolin family of actin-severing proteins.     

The actin filament-severing domain of plasma gelsolin

Gelsolin, a multifunctional actin-modulating protein, has two actin-binding sites which may interact cooperatively. Native gelsolin requires micromolar Ca2+ for optimal binding of actin to both sites, and for expression of its actin filament-severing function. Recent work has shown that an NH2-terminal chymotryptic 17-kD fragment of human plasma gelsolin contains one of the actin-binding sites, and that this fragment binds to and severs actin filaments weakly irrespective of whether Ca2+ is present. The other binding site is Ca2+ sensitive, and is found in a chymotryptic peptide derived from the COOH-terminal two-thirds of plasma gelsolin; this fragment does not sever F-actin or accelerate the polymerization of actin. This paper documents that larger thermolysin-derived fragments encompassing the NH2-terminal half of gelsolin sever actin filaments as effectively as native plasma gelsolin, although in a Ca2+-insensitive manner. This result indicates that the NH2-terminal half of gelsolin is the actin-severing domain. The stringent Ca2+ requirement for actin severing found in intact gelsolin is not due to a direct effect of Ca2+ on the severing domain, but indirectly through an effect on domains in the COOH-terminal half of the molecule to allow exposure of both actin-binding sites.     

Molecular biology of gelsolin, a calcium-regulated actin filament severing protein

Gelsolin is a Ca2+-binding protein of mammalian leukocytes, platelets and other cells which has multiple and closely regulated powerful effects on actin. In the presence of micromolar Ca2+, gelsolin severs actin filaments, causing profound changes in the consistency of actin polymer networks. A variant of gelsolin containing a 25-amino acid extension at the NH2-terminus is present in plasma where it may be involved in the clearance of actin filaments released during tissue damage. Gelsolin has two sites which bind actin cooperatively. These sites have been localized using proteolytic cleavage and monoclonal antibody mapping techniques. The NH2-terminal half of the molecule contains a Ca2+-insensitive actin severing domain while the COOH-terminal half contains a Ca2+-sensitive actin binding domain which does not sever filaments. These data suggest that the NH2-terminal severing domain in intact gelsolin is influenced by the Ca2+-regulated COOH-terminal half of the molecule. The primary structure of gelsolin, deduced from human plasma gelsolin cDNA clones, supports the existence of actin binding domains and suggests that these may have arisen from a gene duplication event, and diverged subsequently to adopt their respective unique functions. The plasma and cytoplasmic forms of gelsolin are encoded by a single gene, and preliminary results indicate that separate mRNAs code for the two forms. Further application of molecular biological techniques will allow exploration into the structural basis for the multifunctionality of gelsolin, as well as the molecular basis for the genesis of the cytoplasmic and secreted forms of gelsolin.     

Molecular cloning of human macrophage capping protein cDNA. A unique member of the gelsolin/villin family expressed primarily in macrophages.

Macrophage capping protein (MCP) is a Ca(2+)-sensitive protein which reversibly blocks the barbed ends of actin filaments but does not sever preformed actin filaments. The human cDNA for MCP has been cloned and sequenced. The derived amino acid sequence predicts a polypeptide of 38.4 kDa. Human MCP expressed in Escherichia coli using a pET12a vector was functionally identical to the native protein purified from rabbit alveolar macrophages with respect to Ca2+ sensitivity and ability to block monomer exchange at the barbed end of actin filaments. Sequence comparison with other actin-binding protein sequences indicates that MCP is a member of the gelsolin/villin family of barbed end blocking proteins. Unlike gelsolin, this protein has a limited tissue distribution being detected primarily in macrophages where it was abundant, representing 0.9-1% of the total cytoplasmic protein. Northern blot analysis of U937 and HL60 cells differentiated to macrophage-like cells demonstrated that MCP message increases to 2.6 and greater than 7 times initial levels, respectively. Human MCP displays a 93% amino acid sequence identity with two recently described mouse proteins, gCap39 and Mbh1. Its abundance in macrophages and the corresponding increases in mRNA levels upon promyelocyte and monocyte development into macrophages indicate that MCP may play an important role in macrophage function.     

Purification and characterization of a Ca2+-dependent actin filament severing protein from bovine adrenal medulla

We describe the purification of an actin regulatory protein from bovine adrenal medulla. This protein caused a dose-dependent decrease of the specific viscosity of actin solution within 30 s of its addition in a Ca2+-sensitive way. Sedimentation assays and the observation by electron microscopy showed that this effect was ascribable to the fragmentation of actin filaments. This protein apparently promoted nucleation of actin polymerization and increased the critical concentration of actin for polymerization nearly 5-fold, suggesting its binding to the barbed end of actin filaments. The inhibitory effect of this protein on the elongation of actin from the barbed end of the myosin subfragment S1-labeled actin seeds confirmed this suggestion. These properties are similar to those of gelsolin. However, the physicochemical properties of this protein having a single polypeptide chain with a molecular weight of 74,000, a Stokes radius of 3.9 nm, a sedimentation coefficient (s0(20),w) of 4.5 S, and an immunological characterization showed that this protein is different from gelsolin.     

Ca2+ control of actin gelation. Interaction of gelsolin with actin filaments and regulation of actin gelation

We elucidated the mechanism by which gelsolin, a Ca2+-dependent regulatory protein from lung macrophages, controls the network structure of actin filaments. In the presence of micromolar Ca2+, gelsolin bound Ca2+. The Ca2+-gelsolin complex reduced the apparent viscosity and flow birefringence of F-actin and the lengths of actin filaments viewed in the electron microscope. However, concentrations of gelsolin causing these alterations did not effect proportionate changes in the turbidity of actin filament solutions or in the quantity of nonsedimentable actin as determined by a radioassay. From these findings, we conclude that gelsolin shortens actin filaments without net depolymerization. Such an effect on the distribution of actin filament lengths led to the prediction that low concentrations of gelsolin would increase the critical concentration of actin-binding protein required for incipient gelation of actin filaments in the presence of Ca2+, providing an efficient mechanism for controlling actin network structure. We verified the prediction experimentally, and we estimated that the Ca2+-gelsolin complex effectively breaks the bond between actin monomers in filaments with a stoichiometry of 1:1. The effect of Ca2+-gelsolin complex on actin solation was rapid, independent of temperature between 0 degrees and 37 degrees C, and reversed by reducing the free Ca2+ concentration.     

Identification of a polyphosphoinositide-modulated domain in gelsolin which binds to the sides of actin filaments

Gelsolin is a Ca2+- and polyphosphoinositide-modulated actin-binding protein which severs actin filaments, nucleates actin assembly, and caps the "barbed" end of actin filaments. Proteolytic cleavage analysis of human plasma gelsolin has shown that the NH2-terminal half of the molecule severs actin filaments almost as effectively as native gelsolin in a Ca2+-insensitive but polyphosphoinositide-inhibited manner. Further proteolysis of the NH2-terminal half generates two unique fragments (CT14N and CT28N), which have minimal severing activity. Under physiological salt conditions, CT14N binds monomeric actin coupled to Sepharose but CT28N does not. In this paper, we show that CT28N binds stoichiometrically and with high affinity to actin subunits in filaments, suggesting that it preferentially recognizes the conformation of polymerized actin. Analysis of the binding data shows that actin filaments have one class of CT28N binding sites with Kd = 2.0 X 10(-7) M, which saturates at a CT28N/actin subunit ratio of 0.8. Binding of CT28N to actin filaments is inhibited by phosphatidylinositol 4,5-bisphosphate micelles. In contrast, neither CT14N nor another actin-binding domain located in the COOH-terminal half of gelsolin form stable stoichiometric complexes with actin along the filaments, and their binding to actin monomers is not inhibited by PIP2. Based on these observations, we propose that CT28N is the polyphosphoinositide-regulated actin-binding domain which allows gelsolin to bind to actin subunits within a filament before serving.     

Comparison between the gelsolin and adseverin domain structure.

Adseverin (74-kDa protein, scinderin) is a calcium- and phospholipid-modulated actin-binding protein that promotes actin polymerization, severs actin filaments, and caps the barbed end of the actin filament, with its NH2-terminal half retaining these properties (Sakurai, T., Kurokawa, H., and Nonomura, Y. (1991) J. Biol. Chem. 266, 4581-4585). Further proteolysis of this NH2-terminal half generated five fragments, and two of them (Mr 15,000 and 31,000) showed Ca(2+)-dependent binding to monomeric actin. The Mr 31,000 fragment especially caused actin filament fragmentation, although its severing activity was also inhibited by several acidic phospholipids as was found in adseverin and its NH2-terminal half. Amino acid sequencing demonstrated that the two fragments' NH2 terminus were blocked in the same manner as the NH2 terminus of adseverin, and thus these two fragments are possibly located at the NH2-terminal of the adseverin molecule. This would then indicate that NH2-terminal fragments had a Ca(2+)-sensitive actin-binding function that relates to actin severing. The other two fragments' NH2-terminal sequencing showed a similar homology to the amino acid sequences of gelsolin and villin. Based on these observations, we propose that adseverin has a functional domain structure similar to that of the gelsolin and villin core.     

Chimeric and truncated gCap39 elucidate the requirements for actin filament severing and end capping by the gelsolin family of proteins.

gCap39 is an actin filament end-capping protein which has a threefold repeated domain structure similar to the N-terminal half of gelsolin. However, unlike gelsolin, gCap39 does not sever actin filaments and dissociates completely from filament ends after calcium removal. We have capitalized on these differences to explore the structural basis for actin filament capping, severing, and their regulation. Using truncated gCap39, generated by limited proteolysis or deletion mutagenesis, we found that actin filament capping requires multiple gCap domains, and almost the entire molecule is necessary for optimal activity. gCap39 domain I, like the equivalent domain in gelsolin, contains an actin monomer binding site. gCap39 domains II-III are, however, different from gelsolin in that they do not bind to the side of actin filaments. Since filament side binding is hypothesized to be the first step in severing, lack of side binding may explain why gCap39 does not sever. This is confirmed directly by swapping gCap39 domains II-III for the side-binding gelsolin domains to generate a chimera which severs actin filaments. The chimera is Ca2+ independent in actin filament severing and capping, although gCap39 domain I itself is regulated by Ca2+.     

Kinetic analysis of F-actin depolymerization in the presence of platelet gelsolin and gelsolin-actin complexes

Platelet gelsolin (G), a 90,000-mol-wt protein, binds tightly to actin (A) and calcium at low ionic strength to form a 1:2:2 complex, GA2Ca2 (Bryan, J., and M. Kurth, 1984, J. Biol. Chem. 259:7480-7487). Chromatography of actin and gelsolin mixtures in EGTA-containing solutions isolates a stable binary complex, GA1Ca1 (Kurth, M., and J. Bryan, 1984, J. Biol. Chem. 259:7473-7479). The effects of platelet gelsolin and the binary gelsolin-actin complex on the depolymerization kinetics of rabbit skeletal muscle actin were studied by diluting pyrenyl F-actin into gelsolin or complex-containing buffers; a decrease in fluorescence represents disassembly of filaments. Dilution of F- actin to below the critical concentration required for filament assembly gave a biphasic depolymerization curve with both fast and slow components. Dilution into buffers containing gelsolin, as GCa2, increased the rate of depolymerization and gave a first order decay. The rate of decrease in fluorescence was found to be gelsolin concentration dependent. Electron microscopy of samples taken shortly after dilution into GCa2 showed a marked reduction in filament length consistent with filament severing and an increase in the number of ends. Conversely, occupancy of the EGTA-stable actin-binding site by an actin monomer eliminated the severing activity. Dilution of F-actin into the gelsolin-actin complex, either as GA1Ca1 or GA1Ca2, resulted in a decrease in the rate of depolymerization that was consistent with filament end capping. This result indicates that the EGTA-stable binding site is required and must be unoccupied for filament severing to occur. The effectiveness of gelsolin, GCa2, in causing filament depolymerization was dependent upon the ionic conditions: in KCI, actin filaments appeared to be more stable and less susceptible to gelsolin, whereas in Mg2+, actin filaments were more easily fragmented. Finally, a comparison of the number of kinetically active ends generated when filaments were diluted into gelsolin versus the number formed when gelsolin can function as a nucleation site suggests that gelsolin may sever more than once. The data are consistent with a mechanism where gelsolin, with both actin-binding sites unoccupied, can sever but not cap F-actin. Occupancy of the EGTA-stable binding site yields a gelsolin-actin complex that can no longer sever filaments, but can cap filament ends.     

Identification of gelsolin, a Ca2+-dependent regulatory protein of actin gel-sol transformation, and its intracellular distribution in a variety of cells and tissues

Antiserum prepared against gelsolin, a major Ca2+-dependent regulatory protein of actin gel-sol transformation in rabbit lung macrophages, was used to detect the presence of proteins immunologically related to gelsolin in a variety of cells and tissues. Cell extracts were electrophoresed on polyacrylamide gels, and replicas of the gels on cellulose nitrate paper were stained by an indirect immunohistochemical technique. A single band of crossreactive material which comigrates with macrophage gelsolin is found in at least nine different kinds of cells and tissues derived from rabbits and humans and in four lines of cultured cells from humans and rats. Gelsolin was also identified in human serum and plasma, raising the possibility that it may contribute to the clearance of actin from the circulatory system. Using this antiserum, we demonstrated, by indirect immunofluorescent staining of acetone-fixed macrophages and polymorphonuclear leukocytes, that gelsolin resides in the cortical cytoplasm and that during phagocytosis it is concentrated in pseudopodia engulfing particles to be ingested, an area of the cytoplasm actively engaged in movement. In longitudinal cryostat sections of contracted rabbit skeletal muscle, antigelsolin staining was associated with the I-band of the myofibril, suggesting that it may be involved, by an as yet undefined mechanism, in skeletal muscle function. In rabbit intestinal epithelial cells, gelsolin was associated with the cytoplasm and the terminal web region of the brush border, a localization distinct from that previously reported for villin, a structurally and functionally similar protein isolated from the brush borders of chicken intestinal epithelial cells. In conclusion, our findings support the idea that gelsolin is involved in the regulation of movement and suggest that gelsolin-mediated Ca2+- regulation of actin cytoskeletal structure, first characterized in macrophages, may be of general importance.     

Interactions of gelsolin and gelsolin-actin complexes with actin. Effects of calcium on actin nucleation, filament severing, and end blocking

Gelsolin is a calcium binding protein that shortens actin filaments. This effect occurs in the presence but not in the absence of micromolar calcium ion concentrations and is partially reversed following removal of calcium ions. Once two actin molecules have bound to gelsolin in solutions containing Ca2+, one of the actins remains bound following chelation of calcium, so that the reversal of gelsolin's effect cannot be accounted for simply by its dissociation from the ends of the shortened filaments to allow for elongation. In this paper, the interactions with actin of the ethylene glycol bis(beta-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA) stable 1:1 gelsolin-actin complexes are compared with those of free gelsolin. The abilities of free or complexed gelsolin to sever actin filaments, nucleate filament assembly, bind to the fast growing (+) filament ends, and lower the filament size distribution in the presence of either Ca2+ or EGTA were examined. The results show that both free gelsolin and gelsolin-actin complexes are highly dependent on Ca2+ concentration when present in a molar ratio to actin less than 1:50. The gelsolin-actin complexes, however, differ from free gelsolin in that they have a higher affinity for (+) filament ends in EGTA and they cannot sever filaments in calcium. The limited reversal of actin-gelsolin binding following removal of calcium and the calcium sensitivity of nucleation by complexes suggest an alternative to reannealing of shortened filaments that involves redistribution of actin monomers and may account for the calcium-sensitive functional reversibility of the solation of actin by gelsolin.     

Isolation and properties of two actin-binding domains in gelsolin

Gelsolin is a Ca2+-sensitive 90-kDa protein which regulates actin filament length. A molecular variant of gelsolin is present in plasma as a 93-kDa protein. Functional studies have shown that gelsolin contains two actin-binding sites which are distinct in that after Ca2+-mediated binding, removal of free Ca2+ releases actin from one site but not from the other. We have partially cleaved human plasma gelsolin with alpha-chymotrypsin and identified two distinct actin-binding domains. Peptides CT17 and CT15, which contain one of the actin-binding domains, bind to actin independently of Ca2+; peptides CT54 and CT47, which contain the other domain, bind to actin reversibly in response to changes in Ca2+ concentration. These peptides sequester actin monomers inhibiting polymerization. Unlike intact gelsolin, neither group of peptides nucleates actin assembly or forms stable filament end caps. CT17 and CT15 can however sever actin filaments. Amino acid sequence analyses place CT17 at the NH2 terminus of gelsolin and CT47 at the carboxyl-terminal two-thirds of gelsolin. Circular dichroism measurements show that Ca2+ induces an increase in the alpha-helical content of CT47. These studies provide a structural basis for understanding the interaction of gelsolin with actin and allow comparison with other Ca2+-dependent actin filament severing proteins.