HSP100, a 100-kDa heat shock protein, is a Ca2+-calmodulin-regulated actin-binding protein

The 100-kDa heat shock protein, HSP100, was purified from mouse lymphoma cells. Amino acid sequences of three peptide fragments which were obtained from the purified protein by lysylendopeptidase digestion were completely or nearly identical with those of a mouse endoplasmic reticulum protein, ERp99, of a hamster glucose-regulated protein, GRP94, and of a chicken heat shock protein, HSP108, all of which have been known to have strong homology with the 90-kDa heat shock protein, HSP90. HSP100 bound to actin filaments and an apparent Kd for the binding was determined to be 8 x 10(-7) M in 2 mM MgCl2 + 100 mM KCl. Calmodulin inhibited the binding in a Ca2+-dependent manner. Equilibrium gel filtration demonstrated that HSP100 has an ability to bind to calmodulin only in the presence of Ca2+. Moreover, HSP100 competed with HSP90 for binding to actin filaments. These results together with our previous findings that HSP90 and HSP100 have similar physicochemical properties (Koyasu, S., Nishida, E., Kadowaki, T., Matsuzaki, F., Iida, K., Harada, F., Kasuga, M., Sakai, H., and Yahara, I. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 8054-8058) and HSP90 is a calmodulin-regulated actin-binding protein (Nishida, E., Koyasu, S., Sakai, H., and Yahara, I. (1986) J. Biol. Chem. 261, 16033-16036), strongly suggest that HSP100 is structurally and functionally related to HSP90.     

Effect of the state of oxidation of cysteine 190 of tropomyosin on the assembly of the actin-tropomyosin complex

Tropomyosin, cross-linked at cysteine 190, was found to bind more weakly to actin filaments than uncross-linked tropomyosin. Cross-linking of tropomyosin can cause actin filaments nearly completely covered with tropomyosin to be uncovered almost completely. The critical monomer concentration of actin is not significantly changed by binding of cross-linked or uncross-linked tropomyosin to actin filaments. The binding curves were analyzed quantitatively, thereby taking into account the polar end-to-end contact of tropomyosin molecules bound by actin and the overlap of the seven subunit binding sites along the actin filament. Under the conditions of the experiment (80 mM KCl, 1 mM MgCl2, pH 7.5, 38-42 degrees C), the equilibrium constant for isolated binding of tropomyosin to actin filaments is in the range 1 x 10(3)-3 x 10(3) M-1. The equilibrium constants for binding of tropomyosin to binding sites along the actin filament with one or two neighbouring tropomyosin molecules are in the range of 10(6) or 10(8) to 10(9) M-1, respectively. The equilibrium constants for binding of tropomyosin to binding sites along the actin filament with one or two neighbouring tropomyosin molecules are in the range of 10(6) or 10(8) to 10(9) M-1, respectively. The equilibrium constants for cross-linked and uncross-linked tropomyosin differ by a factor of only about two. Owing to the highly cooperative binding, these differences are sufficient so that actin filaments nearly completely covered with uncross-linked tropomyosin are uncovered almost completely by cross-linking tropomyosin at cysteine 190.     

Rate of binding of tropomyosin to actin filaments

The decrease of the rate of actin polymerization by tropomyosin molecules which bind near the ends of actin filaments was analyzed in terms of the rate of binding of tropomyosin to actin filaments. Monomeric actin was polymerized onto actin filaments in the presence of various concentrations of tropomyosin. At high concentrations of monomeric actin (c1) and low tropomyosin concentrations (ct) (c1/ct greater than 10), actin polymerization was not retarded by tropomyosin because actin polymerization was faster than binding of tropomyosin to actin filaments. At low actin concentrations and high tropomyosin concentrations (c1/ct less than 5), the rate of elongation of actin filaments was decreased because actin polymerization was slower than binding of tropomyosin at the ends of actin filaments. The results were quantitatively analyzed by a model in which it was assumed that actin-bound tropomyosin molecules which extend beyond the ends of actin filaments retard association of actin monomers with filament ends. Under the experimental conditions (100 mM KCl, 1 mM MgCl2, pH 7.5, 25 degrees C), the rate constant for binding of tropomyosin to actin filaments turned out to be about 2.5 X 10(6) to 4 X 10(6) M-1 S-1.     

Calmodulin-regulated binding of the 90-kDa heat shock protein to actin filaments

We have found that the 90-kDa heat shock protein (HSP90) prepared from a mouse lymphoma exists in homodimeric form under physiological conditions and has the ability to bind to F-actin (Koyasu, S., Nishida, E., Kadowaki, T., Matsuzaki, F., Iida, K., Harada, F., Kasuga, M., Sakai, H., and Yahara, I. (1986) Proc. Natl. Acad. Sci. U.S.A., in press). Here we show that calmodulin regulates the binding of HSP90 to F-actin in a Ca2+-dependent manner. The binding of HSP90 to F-actin occurred optimally under physiological solution conditions, i.e. in 2 mM MgCl2 + 100 mM KCl. The binding was saturable in a molar ratio of about 1 HSP90 (dimer) to 10 actins. HSP90 was dissociated from F-actin by the binding of tropomyosin to F-actin. Calmodulin was found to inhibit the binding of HSP90 to F-actin in a Ca2+-dependent manner. Moreover, the equilibrium gel filtration demonstrated that calmodulin binds to HSP90 in the presence of Ca2+, but not in the absence of Ca2+. These data indicate that HSP90 complexed with Ca2+-calmodulin is unable to bind to F-actin. Ca2+-dependent interaction of HSP90 with calmodulin as well as calmodulin-regulated binding of HSP90 to F-actin revealed here may provide new insight into the function of HSP90 and the regulation of actin structure in cells.     

Characterization of 83-kilodalton nonmuscle caldesmon from cultured rat cells: stimulation of actin binding of nonmuscle tropomyosin and periodic localization along microfilaments like tropomyosin

Nonmuscle caldesmon purified from cultured rat cells shows a molecular weight of 83,000 on SDS gels, Stokes radius of 60.5 A, and sedimentation coefficient (S20,w) of 3.5 in the presence of reducing agents. These values give a native molecular weight of 87,000 and a frictional ratio of 2.04, suggesting that the molecule is a monomeric, asymmetric protein. In the absence of reducing agents, the protein is self-associated, through disulfide bonds, into oligomers with a molecular weight of 230,000 on SDS gels. These S-S oligomers appear to be responsible for the actin-bundling activity of nonmuscle caldesmon in the absence of reducing agents. Actin binding is saturated at a molar ratio of one 83-kD protein to six actins with an apparent binding constant of 5 X 10(6) M-1. Because of 83-kD nonmuscle caldesmon and tropomyosin are colocalized in stress fibers of cultured cells, we have examined effects of 83-kD protein on the actin binding of cultured cell tropomyosin. Of five isoforms of cultured rat cell tropomyosin, tropomyosin isoforms with high molecular weight values (40,000 and 36,500) show higher affinity to actin than do tropomyosin isoforms with low molecular weight values (32,400 and 32,000) (Matsumura, F., and S. Yamashiro-Matsumura. 1986. J. Biol. Chem. 260:13851-13859). At physiological concentration of KCl (100 mM), 83-kD nonmuscle caldesmon stimulates binding of low molecular weight tropomyosins to actin and increases the apparent binding constant (Ka from 4.4 X 10(5) to 1.5 X 10(6) M-1. In contrast, 83-kD protein has slight stimulation of actin binding of high molecular weight tropomyosins because high molecular weight tropomyosins bind to actin strongly in this condition. As the binding of 83-kD protein to actin is regulated by calcium/calmodulin, 83-kD protein regulates the binding of low molecular weight tropomyosins to actin in a calcium/calmodulin-dependent way. Using monoclonal antibodies to visualize nonmuscle caldesmon along microfilaments or actin filaments reconstituted with purified 83-kD protein, we demonstrate that 83-kD nonmuscle caldesmon is localized periodically along microfilaments or actin filaments with similar periodicity (36 +/- 4 nm) as tropomyosin. These results suggest that 83-kD protein plays an important role in the organization of microfilaments, as well as the control of the motility, through the regulation of the binding of tropomyosin to actin.     

Equilibrium of the actin-tropomyosin interaction

The actin-tropornyosin interaction was studied by means of light-scattering. The experimental data were analysed on the basis of the model of co-operative binding of large ligands to a one-dimensional lattice with overlapping binding sites. The affinity of tropomyosin for actin filaments was dependent on the magnesium concentration. A fivefold increase of the magnesium concentration (from 0·5 mm to 2·5 mm) enhanced the equilibrium constant twofold (from 700 to 1600 m^?1) for the isolated binding of tropomyosin molecules to actin filaments. At low magnesium concentrations (0·5 mm), tropomyosin molecules were bound to isolated binding sites on an actin filament about 600 times more weakly than to contiguous binding sites. At increased magnesium concentrations (2·5 mm), the tendency of tropomyosin to bind contiguously increased twofold. Due to the co-operative nature of the actin-tropomyosin interaction, a small change in the magnesium concentration may cause a great change of the structural organisation of the complex. A small enhancement of the magnesium concentration (from 1 mm to 1·5 mm) caused bare filaments to be covered almost completely with tropomyosin. The length of tropomyosin clusters and the number of gaps on actin filaments depended strongly on the magnesium concentration. From the values of the experimentally determined equilibrium constants, it was concluded that the end-to-end interaction of tropomyosin was not strong enough to bring about all-or-none behaviour, where actin filaments of physiological length (~1000 nm) are either completely covered with or completely free of tropomyosin.     

An actin subdomain 2 mutation that impairs thin filament regulation by troponin and tropomyosin

Striated muscle thin filaments adopt different quaternary structures, depending upon calcium binding to troponin and myosin binding to actin. Modification of actin subdomain 2 alters troponin-tropomyosin-mediated regulation, suggesting that this region of actin may contain important protein-protein interaction sites. We used yeast actin mutant D56A/E57A to examine this issue. The mutation increased the affinity of tropomyosin for actin 3-fold. The addition of Ca(2+) to mutant actin filaments containing troponin-tropomyosin produced little increase in the thin filament-myosin S1 MgATPase rate. Despite this, three-dimensional reconstruction of electron microscope images of filaments in the presence of troponin and Ca(2+) showed tropomyosin to be in a position similar to that found for muscle actin filaments, where most of the myosin binding site is exposed. Troponin-tropomyosin bound with comparable affinity to mutant and wild type actin in the absence and presence of calcium, and in the presence of myosin S1, tropomyosin bound very tightly to both types of actin. The mutation decreased actin-myosin S1 affinity 13-fold in the presence of troponin-tropomyosin and 2.6-fold in the absence of the regulatory proteins. The results suggest the importance of negatively charged actin subdomain 2 residues 56 and 57 for myosin binding to actin, for tropomyosin-actin interactions, and for regulatory conformational changes in the actin-troponin-tropomyosin complex.     

Crosslinking of actin filaments and inhibition of actomyosin subfragment-1 ATPase activity by streptococcal M6 protein

M proteins are antiphagocytic molecules on the surface of group A streptococci having physical characteristics similar to those of mammalian tropomyosin. Both are alpha-helical coiled-coil fibrous structures with a similar seven-residue periodicity of nonpolar and charged amino acids. To determine if M protein is functionally similar to tropomyosin we studied the interaction of M protein with F-actin. At low ionic strength, M protein binds to actin weakly with a stoichiometry different from that of tropomyosin. M protein does not compete with tropomyosin for the binding to actin, indicating that it is functionally different from tropomyosin. M protein does compete with myosin subfragment-1 for binding to actin and induces the formation of bundles of actin filaments. The formation of actin aggregates is associated with a sharp reduction in the rate of ATP hydrolysis by subfragment-1. Intact streptococci having M protein on their surface are shown to bind to actin.     

Identification and purification of a novel Mr 43,000 tropomyosin-binding protein from human erythrocyte membranes

A new Mr 43,000 tropomyosin-binding protein (TMBP) has been identified in erythrocyte membranes by binding of 125I-labeled Bolton-Hunter tropomyosin to nitrocellulose blots of membrane proteins separated by sodium dodecyl sulfate-gel electrophoresis. This protein is not actin, because 125I-tropomyosin does not bind to purified actin on blots. Binding of 125I-tropomyosin to this protein is specific because it is inhibited by excess unlabeled tropomyosin but not by F-actin or muscle troponins. This protein has been purified to 95% homogeneity from a 1 M Tris extract of tropomyosin-depleted erythrocyte membranes by DEAE-cellulose and hydroxylapatite chromatography, followed by gel filtration on Ultrogel AcA 44. The purified protein has a Stokes radius of 3.9 nm and a sedimentation coefficient of 2.8 S, corresponding to a native molecular weight of 43,000. Binding of 125I-tropomyosin to the purified TMBP saturates at one tropomyosin molecule (Mr 60,000) to two Mr 43,000 TMBPs, with an affinity of about 5 X 10(-7) M. The TMBP is associated with the membrane skeleton after extraction of membranes with the non-ionic detergent, Triton X-100, and is present with respect to tropomyosin at a ratio of about one for every two tropomyosin molecules. Because there is enough tropomyosin for two tropomyosin molecules to be associated with each of the short actin filaments in the membrane skeleton, the erythrocyte membrane TMBP, together with tropomyosin, could function to restrict the number of spectrin molecules attached to each of the short actin filaments and thus specify the hexagonal symmetry of the spectrin-actin lattice. Alternatively, this TMBP could be homologous to one of the muscle troponins and might function with tropomyosin to regulate erythrocyte actomyosin-ATPase activity and influence erythrocyte shape.     

Tropomyosins Regulate the Severing Activity of Gelsolin in Isoform-Dependent and Independent Manners

The actin cytoskeleton fulfills numerous key cellular functions, which are tightly regulated in activity, localization, and temporal patterning by actin binding proteins. Tropomyosins and gelsolin are two such filament-regulating proteins. Here, we investigate how the effects of tropomyosins are coupled to the binding and activity of gelsolin. We show that the three investigated tropomyosin isoforms (Tpm1.1, Tpm1.12, and Tpm3.1) bind to gelsolin with micromolar or submicromolar affinities. Tropomyosin binding enhances the activity of gelsolin in actin polymerization and depolymerization assays. However, the effects of the three tropomyosin isoforms varied. The tropomyosin isoforms studied also differed in their ability to protect pre-existing actin filaments from severing by gelsolin. Based on the observed specificity of the interactions between tropomyosins, actin filaments, and gelsolin, we propose that tropomyosin isoforms specify which populations of actin filaments should be targeted by, or protected from, gelsolin-mediated depolymerization in living cells.     

Study of the heteromeric structure of the untransformed glucocorticoid receptor using chemical cross-linking and monoclonal antibodies against the 90K heat-shock protein

The untransformed rat glucocorticoid receptor is assumed to be a hetero-oligomeric complex, containing a non-steroid binding component, the 90K heat-shock protein (HSP 90). Direct measurement of its molecular weight by chemical cross-linking provides new evidence for a trimeric structure with a Mr of ca. 270,000. Resorting to an anti HSP 90 probe (AC 88), we show that the native dimeric HSP 90 possess two accessible epitopes for this monoclonal antibody, while when bound to the steroid-binding subunit, only one epitope remains accessible. These data clearly suggest that the untransformed rat glucocorticoid receptor is an asymmetrical hetero-oligomeric complex.     

Characterization of f-actin tryptophan phosphorescence in the presence and absence of tryptophan-free myosin motor domain

The effect of binding the Trp-free motor domain mutant of Dictyostelium discoideum, rabbit skeletal muscle myosin S1, and tropomyosin on the dynamics and conformation of actin filaments was characterized by an analysis of steady-state tryptophan phosphorescence spectra and phosphorescence decay kinetics over a temperature range of 140-293 K. The binding of the Trp-free motor domain mutant of D. discoideum to actin caused red shifts in the phosphorescence spectrum of two internal Trp residues of actin and affected the intrinsic lifetime of each emitter, decreasing by roughly twofold the short phosphorescence lifetime components (tau(1) and tau(2)) and increasing by approximately 20% the longest component (tau(3)). The alteration of actin phosphorescence by the motor protein suggests that i), structural changes occur deep down in the core of actin and that ii), subtle changes in conformation appear also on the surface but in regions distant from the motor domain binding site. When actin formed complexes with skeletal S1, an extra phosphorescence lifetime component appeared (tau(4), twice as long as tau(3)) in the phosphorescence decay that is absent in the isolated proteins. The lack of this extra component in the analogous actin-Trp-free motor domain mutant of D. discoideum complex suggests that it should be assigned to Trps in S1 that in the complex attain a more compact local structure. Our data indicated that the binding of tropomyosin to actin filaments had no effect on the structure or flexibility of actin observable by this technique.     

Evidence for a direct but sequential binding of titin to tropomyosin and actin filaments

Titin is a giant molecule that spans half a sarcomere, establishing several specific bindings with both structural and contractile myofibrillar elements. It has been demonstrated that this giant protein plays a major role in striated muscle cell passive tension and contractile filament alignment. The in vitro interaction of titin with a new partner (tropomyosin) reported here is reinforced by our recent in vitro motility study using reconstituted Ca-regulated thin filaments, myosin and a native 800-kDa titin fragment. In the presence of the tropomyosin-troponin complex, the actin filament movement onto coated S1 is improved by the titin fragment. Here, we found that two purified native titin fragments of 150 and 800 kDa, covering respectively the N1-line and the N2-line/PEVK region in the I-band and known to contain actin-binding sites, directly bind tropomyosin in the absence of actin. We have also shown that binding of the 800-kDa fragment with filamentous actin inhibited the subsequent interaction of tropomyosin with actin, as judged by cosedimentation. However, this was not the case if the complex of actin and tropomyosin was formed before the addition of the 800-kDa fragment. We thus conclude that a sequential arrangement of contacts exists between parts of the titin I-band region, tropomyosin and actin in the thin filament.     

Cell-free synthesis of rat glucocorticoid receptor in rabbit reticulocyte lysate. In vitro synthesis of receptor in Mr 90,000 heat shock protein-depleted lysate

The glucocorticoid receptor is present in the cytosol of cell extracts as a large nonactivated (i.e. non-DNA-binding) approximately 9 S (Mr 300,000) complex. Experimental evidence indicates that the purified nonactivated glucocorticoid receptor contains a single steroid-binding protein and two approximately 90-kDa nonsteroid-binding subunits identified as heat shock protein (hsp) 90. Translation of the glucocorticoid receptor mRNA in vitro in reticulocyte lysates produces a large nonactivated glucocorticoid receptor complex similar to that found in cytosols. The cell-free synthesized glucocorticoid receptor is able to bind steroid and can be activated further to the DNA-binding form. To test the hypothesis of an active role played by hsp90 in the stabilization of a competent steroid-binding conformation of the glucocorticoid receptor, we have synthesized the receptor in a reticulocyte lysate that has been depleted of hsp90 by immunoadsorption with AC88 anti-hsp90. Although the translation capacity of the reticulocyte system was reduced considerably upon hsp90 removal, the glucocorticoid receptor was synthesized, and a significant number of molecules were found to bind [3H]triamcinolone acetonide. Chromatography on DEAE-cellulose showed that most of the receptor molecules synthesized in hsp90-depleted lysate had lost the capacity to form an oligomeric receptor complex. Addition of purified rat liver hsp90 to the hsp90-depleted lysate before translation did not increase steroid binding nor did it restore formation of the heteromeric receptor complex. Analysis of [35S] methionine-labeled glucocorticoid receptor molecules synthesized in the hsp90-depleted lysate showed the production of polypeptides differing from the expected chromatographic pattern on DEAE-cellulose. Upon addition of purified hsp90 to the hsp90-depleted lysate, before translation, the 35S-labeled synthesized receptor fractionated on DEAE-cellulose as an intermediate peak between activated and nonactivated receptor forms. The data suggest that hsp90 alone may not be sufficient for the formation of the nonactivated steroid receptor complex.     

Phalloidin and tropomyosin do not prevent actin filament shortening by the 90 kD protein-actin complex from brain

Previously we reported the purification from bovine brain of the 90 kD protein-actin complex that shortens actin filaments. In the present work we study the effect of this complex on actin polymerized in the presence of phalloidin (PL) or tropomyosin (TM) which are known to stabilize actin filaments. The effect of the complex has been compared with that of cytochalasin D (CD), a fungal metabolite that also shortens actin filaments. Low shear viscosimetry and electron microscopy showed that PL or TM could not prevent the shortening of actin filaments in the presence of 90 kD protein-actin complex whereas they effectively protected actin filaments from shortening by CD. We conclude that the 90 kD protein-actin complex is a more potent filament-shortening factor than CD.     

Three-dimensional reconstruction of thin filaments containing mutant tropomyosin

Interactions of the components of reconstituted thin filaments were investigated using a tropomyosin internal deletion mutant, D234, in which actin-binding pseudo-repeats 2, 3, and 4 are missing. D234 retains regions of tropomyosin that bind troponin and form end-to-end tropomyosin bonds, but has a length to span only four instead of seven actin monomers. It inhibits acto-myosin subfragment 1 ATPase (acto-S-1 ATPase) and filament sliding in vitro in both the presence and absence of Ca(2+) (, J. Biol. Chem. 272:14051-14056) and lowers the affinity of S-1.ADP for actin while increasing its cooperative binding. Electron microscopy and three-dimensional reconstruction of reconstituted thin filaments containing actin, troponin, and wild-type or D234 tropomyosin were carried out to determine if Ca(2+)-induced movement of D234 occurred in the filaments. In the presence and absence of Ca(2+), the D234 position was indistinguishable from that of the wild-type tropomyosin, demonstrating that the mutation did not affect normal tropomyosin movement induced by Ca(2+) and troponin. These results suggested that, in the presence of Ca(2+) and troponin, D234 tropomyosin was trapped on filaments in the Ca(2+)-induced position and was unable to undergo a transition to a completely activated position. By adding small amounts of rigor-bonded N-ethyl-maleimide-treated S-1 to mutant thin filaments, thus mimicking the myosin-induced "open" state, inhibition could be overcome and full activation restored. This myosin requirement for full activation provides support for the existence of three functionally distinct thin filament states (off, Ca(2+)-induced, myosin-induced; cf.;, J. Mol. Biol. 266:8-14). We propose a further refinement of the three-state model in which the binding of myosin to actin causes allosteric changes in actin that promote the binding of tropomyosin in an otherwise energetically unfavorable "open" state.     

Intrinsic capability of budding yeast cofilin to promote turnover of tropomyosin-bound actin filaments

The ability of actin filaments to function in cell morphogenesis and motility is closely coupled to their dynamic properties. Yeast cells contain two prominent actin structures, cables and patches, both of which are rapidly assembled and disassembled. Although genetic studies have shown that rapid actin turnover in patches and cables depends on cofilin, how cofilin might control cable disassembly remains unclear, because tropomyosin, a component of actin cables, is thought to protect actin filaments against the depolymerizing activity of ADF/cofilin. We have identified cofilin as a yeast tropomyosin (Tpm1) binding protein through Tpm1 affinity column and mass spectrometry. Using a variety of assays, we show that yeast cofilin can efficiently depolymerize and sever yeast actin filaments decorated with either Tpm1 or mouse tropomyosins TM1 and TM4. Our results suggest that yeast cofilin has the intrinsic ability to promote actin cable turnover, and that the severing activity may rely on its ability to bind Tpm1.     

Isolation from bovine brain of tropomyosins that bind to actin filaments with different affinities

Tropomyosin was isolated from bovine brain using mild conditions thereby avoiding heat precipitation. Separation by DEAE ion exchange chromatography yielded a 33 kDa tropomyosin and a mixture of 30 and 32 kDa tropomyosin. Binding of the tropomyosins to actin filaments was measured by a newly developed method. The binding was assayed by the retarding effect of tropomyosin on actin polymerization. The 33 kDa tropomyosin was found to bind to actin filaments with considerably higher affinity than the 30 and 32 kDa tropomyosin.