Interactions between actin, myosin, and an actin-binding protein from rabbit alveolar macrophages. Alveolar macrophage myosin Mg-2+-adenosine triphosphatase requires a cofactor for activation by actin.

The interactions were analyzed between actin, myosin, and a recently discovered high molecular weight actin-binding protein (Hartwig, J. H., and Stossel, T. P. (1975) J. Biol Chem.250,5696-5705) of rabbit alveolar macrophages. Purified rabbit alveolar macrophage or rabbit skeletal muscle F-actins did not activate the Mg2+ATPase activity of purified rabbit alveolar macrophage myosin unless an additional cofactor, partially purified from macrophage extracts, was added. The Mg2+ATPase activity of cofactor-activated macrophage actomyosin was as high as 0.6 mumol of Pi/mg of myosin protein/min at 37 degrees. The macrophage cofactor increased the Mg2+ATPase activity of rabbit skeletal muscle actomyosin, and calcium regulated the Mg2+ATPase activity of cofactor-activited muscle actomyosin in the presence of muscle troponins and tropomyosin. However, the Mg2+ATPase activity of macrophage actomyosin in the presence of the cofactor was inhibited by muscle control proteins, both in the presence and absence of calcium. The Mg2+ATPase activity of the macrophage actomyosin plus cofactor, whether assembled from purified components or studied in a complex collected from crude macrophage extracts, was not influenced by the presence of absence of calcium ions. Therefore, as described for Acanthamoeba castellanii myosin (Pollard, T. D., and Korn, E. D. (1973) J. Biol. Chem. 248, 4691-4697), rabbit alveolar macrophage myosin requires a cofactor for activation of its Mg2+ATPase activity by F-actin; and no evidence was found for participation of calcium ions in the regulation of this activity.In macrophage extracts containing 0.34 M sucrose, 0.5 mM ATP, and 0.05 M KCl at pH 7.0,the actin-binding protein bound F-actin into bundles with interconnecting bridges. Purified macrophage actin-binding protein in 0.1 M KCl at pH 7.0 also bound purified macrophage F-actin into filament bundles. Macrophage myosin bound to F-actin in the absence but not the presence of Mg2+ATP, but the actin-binding protein did not bind to macrophage myosin in either the presence or absence of Mg2+ATP.     

In vitro motility of skeletal muscle myosin and its proteolytic fragments

We have compared actin-activated myosin ATPase activity, myosin binding to actin, and the velocity of myosin-induced actin sliding in order to understand the mechanism of myosin motility. In our in vitro assay, F-actin slides at a constant velocity, regardless of length. The F-actin could slide over myosin heads at KCl concentrations below a critical value (60 mM with myosin and HMM, 100 mM with S-1), and the sliding velocities were quite similar below the critical KCl concentration. However, at KCl concentrations close to the critical value, the sliding F-actin is attached to only one or a few particular points on the surface, each of which perhaps consists of a single head of myosin. The KATPase values for actin-activated ATPase were approximately 300 microM for S-1 and approximately 200 microM with HMM below the critical KCl concentration, and approximately 5,000 microM above the critical KCl concentration. This increase in KATPase is due to a drastic reduction in the binding affinity of myosin heads to F-actin, as determined by a proteolytic digestion method and direct observation by fluorescence microscopy. We also show that the Vmax of actin-activated myosin ATPase activity decreases steadily with increasing KCl concentration, even though the velocity of F-actin sliding remains unchanged. This result provides evidence that the ATPase activity is not necessarily linked to motility. We discuss possible models that do not require a tight coupling between myosin ATPase and motility.     

Uncoupling of actin-activated myosin ATPase activity from actin binding by a monoclonal antibody directed against the N-terminus of myosin light chain 1.

The role of the N-terminal region of myosin light chain 1 (LC1) in actomyosin interaction was investigated using an IgG monoclonal antibody (2H2) directed against the N-terminal region of LC1. We defined the binding site of 2H2 by examining its cross-reactivity with myosin light chains from a variety of species and with synthetic oligopeptides. Our findings suggest that 2H2 is directed against the N-terminal region of LC1 which includes the trimethylated alanine residue at the N-terminus. In the presence of 2H2, the rate of actomyosin superprecipitation was reduced, although the extent was not. 2H2 caused a reduction in the Vmax of both myosin and chymotryptic S1(A1) actin-activated ATPase activity, while the Km appeared to be unaltered. The Mg(2+)-ATPase activity of myosin alone was also unaffected. Binding studies revealed that 2H2 did not prevent the formation of acto-S1 complex, either in the presence or in the absence of ATP, nor did it affect the ability of ATP to dissociate S1 from F-actin. Our findings suggest that the N-terminal region of LC1 is not essential for actin binding but is involved in modulating actin-activated ATPase activity of myosin.     

Reversible dissociation of dog cardiac myosin regulatory light chain 2 and its influence on ATP hydrolysis

The regulatory light chains of dog heart myosin were removed by digestion with myopathic hamster neutral protease. The heavy chains were also cleaved to an extent of 15%, but a homogeneous, rod-free LC2-deficient myosin was obtained by ion-exchange chromatography. A similar approach was used to prepare LC2-deficient heavy meromyosin. Neither Ca2+- nor K+-EDTA-activated ATPases were affected by LC2 removal. The Lineweaver-Burk plots for actin-activated ATPase in 25 mM KCl were biphasic giving a Vmax of 1.54 s-1 for control and LC2-recombined myosins and 1.08 s-1 for LC2-deficient myosin at low actin concentrations. At high actin concentrations, the Vmax for control and recombined myosins was 2.33 s-1 and 1.39 s-1 for LC2-deficient myosin. Increasing the KCl concentration in the reaction mixtures resulted in more linear plots without suppressing the 35-45% decrease in Vmax that accompanied LC2 removal. The results from assays with control and LC2-deficient heavy meromyosin performed in the absence of KCl, paralleled those obtained with myosin. The latter was also assayed in the presence of equimolar concentrations of C-protein in 50 mM KCl: C-protein induced a significant increase in the actin-activated ATPase of both control and LC2-recombined myosins, with no effect on LC2-deficient myosin. The Vmax for actin-activation in the presence of C-protein was 2.38 s-1, 0.83 s-1, and 1.71 s-1 for control, LC2-deficient, and recombined myosins, respectively. The enhancement of actin-activation in both the control and LC2-recombined myosins represents a possible role for C-protein in a LC2-mediated potentiation of actomyosin ATPase.     

Calcium-sensitive modulation of the actomyosin ATPase by fodrin

Fodrin, a spectrin-like protein isolated from brain, is a long flexible molecule which binds calmodulin and cross-links F-actin. The effects of fodrin on the actin-activated ATPase of myosin have been examined. When added after ATP, fodrin inhibited the actomyosin ATPase. Two to three times as much fodrin was required for inhibition in the presence of Ca2+ as in its absence. Complete inhibition in the absence of Ca2+ occurred at about one fodrin to 200 actins. Inhibition does not appear to result from fodrin cross-linking F-actin, and, thereby, preventing the myosin filaments from reaching the actin filaments; but cross-linking may promote inhibition by trapping the myosin filaments within the cross-linked F-actin. When added before ATP, fodrin stimulated the actomyosin ATPase almost 3-fold in the presence of Ca2+ and by less than 50% in the absence of Ca2+. Stimulation is thought to result from fodrin cross-linking F-actin. After several minutes the stimulations in Ca2+ were greatly reduced, and in the absence of Ca2+ the actomyosin ATPases were substantially inhibited. Whether added before or after ATP, fodrin inhibited the actin-activated ATPase of myosin subfragment 1. This inhibition was also slightly Ca2+ sensitive.     

Elementary steps of the actin-activated ATPase reaction of cardiac muscle myosin subfragment-1

The rates of the elementary steps of the actomyosin ATPase reaction were measured using the myosin subfragment-1 of porcine left ventricular muscle. The results could be explained only by the two-route mechanism for actomyosin ATPase (Inoue, Shigekawa, & Tonomura (1973) J. Biochem. 74, 923-934), in which ATP is hydrolyzed via routes with or without accompanying dissociation of actomyosin. The dependence on the F-actin concentration of the rate of the acto-S-1 ATPase reaction in the steady state was measured in 5 mM KCl at 20 degrees C. The maximal rate, Vmax, and the dissociation constant for F-actin of the ATPase, Kd, were 3.0 s-1 and 2.2 mg/ml, respectively. The Kd value was almost the same as that determined from the extent of binding of S-1 with F-actin during the ATPase reaction. The rate of recombination of the S-1-phosphate-ADP complex, S-1ADPP, with F-actin, vr, was lower than that of the ATPase reaction in the steady state. Thus, ATP is mainly hydrolyzed without accompanying dissociation of acto-S-1 into S-1ADPP and F-actin. In the cardiac acto-S-1 ATPase reaction, the rate of the ATPase reaction in the steady state and that of recombination of S-1ADPP with F-actin were about 1/5 those of the skeletal acto-S-1 ATPase reaction.(ABSTRACT TRUNCATED AT 250 WORDS)     

Structure and function of the two heads of the myosin molecule. IV. Physiological functions of various reaction intermediates in myosin adenosinetriphosphatase, studied by the interaction between actomyosin and 8-bromoadenosine triphosphate.

The kinetic properties of the hydrolyses of 8-Br ATP and 8-SCH3 ATP by myosin [EC 3.6.1.3] and actomyosin were compared with those of ATP, and the following results were obtained. The Ca-NTPase activities of myosin using these two ATP analogs as substrates were smaller than that of ATPase, and the NTPase activities toward these analogs were strongly suppressed by EDTA. The Mg-NTPase activities toward these analogs were higher in a medium of high ionic strength than in a medium of low ionic strength, in contrast to the activity of Mg-ATPase. These analogs did not produce any initial burst of Pi liberation, activation of myosin NTPase by F-actin, or superprecipitation of actomyosin. The interactions between 8-Br ATP and HMM, acto-HMM, actomyosin, and myofibrils were studied in detail in the presence of Mg2+ in medium of low ionic strength. The Michaelis constant, Km, and the maximum rate, Vm, of 8-Br ATPase of HMM were 27 muM and 21 min-1, respectively. The fluorescence change of HMM induced by 8-Br ATP also followed the Michaelis-Menten equation, and the Michaelis constant, Kf1, was as low as 4 muM. Acto-HMM and acto-S-1 were fully dissociated by the addition of 8-Br ATP. The relation between the extent of dissociation of acto-HMM and the concentration of 8-Br ATP followed the Michaelis-Menten equation, and the apparent dissociation constant, Kd, was 22 muM. This Kd value is almost equal to the Km value of 8-Br ATPase of HMM described above. Myofibrillar contraction was not supported by 8-Br ATP. It was concluded that in the myosin NTPase reaction with 8-Br ATP as a substrate, M2NTP but not MNDPP is formed in route (1), while MNTP is formed in route (2). It was also concluded that the key intermediate for the actomyosin NTPase reaction is MNDPP, and that dissociation of acto-HMM is induced by the formation of M2NTP and MNTP in routes (1) and (2), respectively.     

Observations on the kinetics, subunit composition, and sulfhydryl reactivity of myosin from Physarum polycephalum.

A highly purified preparation of myosin from Physarum polycephalum has been shown by sodium dodecyl sulfate polyacrylamide gel electrophoresis to contain heavy chains and only one molecular weight class of light chains, of approx. 15 000 daltons. Kinetic investigations of the Ca2+-ATPase and Mg2+-ATPase (ATP phosphohydrolases, EC 3.6.1.3) at pH 8.0 gave Km and V values of 17.3 muM and 1.25 mumol Pi/min per mg, and 2.4 muM and 0.12 mumol Pi/min per mg, respectively. Adenylyl imidodiphosphate, a beta-gamma-imido ATP analog, inhibited the ATPase activity of Physarum myosin competitively with Ki values equal to 350 and 12 muM in the presence of Ca2+ and Mg2+, respectively. The ATPase activity of Physarum myosin was inhibited at a very low rate (t1/2 = 24 h) by the ATP analog, 6,6'-dithiobis(inosinyl imidodiphosphate), with concentrations of inhibitor previously shown to inactivate (t1/2 approximately 10 min) skeletal and cardiac myosins rapidly by reacting with key cysteines.     

Myosin and actin from ascidian smooth muscle and their interaction

Myosin and actin were purified from ascidian smooth muscle. Ascidian myosin contained two classes of light chains and the pH dependence of Ca2+-activated ATPase and the KCl dependence of actin-activated ATPase of ascidian myosin differed from those of vertebrate skeletal myosin. Troponin-tropomyosin complex from ascidian increased the ATPase activity of ascidian reconstituted actomyosin in a Ca2+-dependent manner. Ascidian myosin provided the reconstituted actomyosin with the responsiveness to calcium ions. Two actin isoforms were present in ascidian, which were distinguished by isoelectric points.     

Effect of the filamentous structure of myosin on the actomyosin ATPase activity

The ATPase activities of acto-heavy meromyosin and of acto-myosin minifilaments have been compared under the same conditions at low ATP (0.1 mM) and at several KC1 concentrations. The activities, which are strongly salt-dependent in both systems, have been found to be similar at high ionic strength (about 0.16 M) but different at lower ionic strength (0.06-0.07 M). Under this last condition, the catalytic constants kcat and Km are lower for acto-myosin minifilaments than for acto-heavy meromyosin ATPase. In addition, at low ionic strength, any decrease in the concentration of any of the ionic species (ATP, citrate, etc.) induces an increase in the interaction strength between myosin and actin filaments, as revealed by the Km changes. The presence of the troponintropomyosin complex and of Ca2+ also enhances the strength of this interaction. On the other hand, the occurrence of particular interactions between F-actin and myosin minifilaments is further substantiated by the phenomenon of superprecipitation which occurs when the ATP concentration decreases. The favourable effect of the organized structure of the myosin minifilaments on the ATPase activity of actomyosin is discussed.     

Myosin-linked calcium regulation in vertebrate smooth muscle.

By the use of a new procedure, actomyosin may be extracted in high yield and purity from fowl gizzard which exhibits a calcium-dependent actin-activated ATPase activity comparable to that of the parent myofibril-like preparation. Studies of this vertebrate smooth muscle actomyosin show that the regulation of the actin-myosin interaction is effected, as in molluscan muscles, by the myosin molecule itself and not by an actin-linked regulatory system, as found in vertebrate skeletal muscle.Thus, calcium-sensitive smooth muscle actomyosin is composed of only myosin, actin and tropomyosin, any troponin-like components being absent. Myosin is the only component that binds significant amounts of calcium and shows a calcium-dependent actin-activated ATPase activity in the presence of F-actin from either gizzard or rabbit skeletal muscle.The cross-reaction of gizzard thin filaments with skeletal muscle myosin produces an actomyosin whose actin-activated ATPase is calcium-insensitive, showing that smooth muscle thin filaments do not serve a regulatory function.The effect of Mg²⁺ and pH, and evidence for the involvement of one of the myosin light chains in calcium regulation are described and discussed.     

Reaction of two heads of gizzard myosin with ATP

The reaction intermediates formed by the two heads of smooth muscle myosin were studied. The amount of myosin-phosphate-ADP complex, MPADP, formed was measured from the Pi-burst size over a wide range of ATP concentrations. At low concentrations of ATP, the Pi-burst size was 0.5 mol/mol myosin head, and the apparent Kd value was about 0.15 microM. However, at high ATP concentrations, the Pi burst size increased from 0.5 to 0.75 mol/mol myosin head with an observed Kd value of 15 microM. The binding of nucleotides to gizzard myosin during the ATPase reaction was directly measured by a centrifugation method. Myosin bound 0.5 mol of nucleotides (ATP and ADP) with high affinity (Kd congruent to 1 microM) and 0.35 mol of nucleotides with low affinity (Kd = 24 microM) for ATP. These results indicate that gizzard myosin has two kinds of nucleotide binding sites, one of which forms MPADP with high affinity for ATP while the other forms MPADP and MATP with low affinity for ATP. We studied the correlation between the formation of MPADP and the dissociation of actomyosin. The amount of Pi-burst size was not affected by the existence of F-actin, and when 0.5 mol of ATP per mol of myosin head was added to actomyosin (1 mg/ml F-actin, 5 microM myosin at 0 degrees C) most (93%) of the added ATP was hydrolyzed in the Pi-burst phase. All gizzard actomyosin dissociated when 1 mol of ATP per mol myosin head was added to actomyosin.(ABSTRACT TRUNCATED AT 250 WORDS)     

Presence of a unit for actin-myosin interaction during the superprecipitation of actomyosin.

The interaction of actin with myosin was studied in the presence of ATP at low ionic strength by means of measurements of the actin-activated ATPase activity of myosin and superprecipitation of actomyosin. At high ATP concentrations the ATPase activities of myosin, heavy meromyosin (HMM) and myosin subfragment 1 (S-1) were activated by actin in the same extent. At low ATP concentrations the myosin ATPase activity was activated about 30-fold by actin, whereas those of HMM and S-1 were stimulated only several-fold. This high actin activation of myosin ATPase was coupled with the occurrence of superprecipitation. The activation of HMM or S-1 ATPase by actin shows a simple hyperbolic dependence on actin concentration, but the myosin ATPase was maximally activated by actin at a 2:1 molar ratio of actin to myosin, and a further increase in the actin concentration had no effect on the activation. These results suggest the presence of a unit for actin-myosin interaction, composed of two actin monomers and one myosin molecule in the filaments.     

Properties of porcine platelet myosin. I. Similarity between vertebrate smooth muscle and nonmuscle myosins in their binding properties with F-actin

The mechanism of the ATPase [EC 3.6.1.3] reaction of porcine platelet myosin and the binding properties of platelet myosin with rabbit skeletal muscle F-actin were investigated. The kinetic properties of the platelet myosin ATPase reaction, that is, the rate, the extent of fluorescence enhancement of myosin, the size of the initial P1 burst of myosin, and the amount of nucleotides bound to myosin during the ATPase reaction, were very similar to those found for other myosins. Strong binding of platelet myosin with rabbit skeletal muscle F-actin, as found for smooth muscle myosin, was suggested by the following results. The rate of the ATP-induced dissociation of hybrid actomyosin, reconstituted from platelet myosin and skeletal muscle F-actin, was very slow. The amount of ATP necessary for complete dissociation of hybrid actomyosin was 2 mol/mol of myosin, although skeletal muscle actomyosin is known to dissociate completely upon addition of 1 mol ATP per mol of myosin. Unlike skeletal muscle myosin, the EDTA(K+)-ATPase activity of platelet myosin was inhibited by skeletal muscle F-actin. These observations indicate that ATP hydrolysis by vertebrate nonmuscle myosin follows the same mechanism as with other myosins and that the binding properties of nonmuscle myosin with F-actin are similar to those of smooth muscle myosin but not to those of skeletal muscle myosin.     

Motor function and regulation of myosin X.

Myosin X is a member of the diverse myosin superfamily that is ubiquitously expressed in various mammalian tissues. Although its association with actin in cells has been shown, little is known about its biochemical and mechanoenzymatic function at the molecular level. We expressed bovine myosin X containing the entire head, neck, and coiled-coil domain and purified bovine myosin X in Sf9 cells. The Mg(2+)-ATPase activity of myosin X was significantly activated by actin with low K(ATP). The actin-activated ATPase activity was reduced at Ca(2+) concentrations above pCa 5 in which 1 mol of calmodulin light chain dissociates from the heavy chain. Myosin X translocates F-actin filaments with the velocity of 0.3 microm/s with the direction toward the barbed end. The actin translocating activity was inhibited at concentrations of Ca(2+) at pCa 6 in which no calmodulin dissociation takes place, suggesting that the calmodulin dissociation is not required for the inhibition of the motility. Unlike class V myosin, which shows a high affinity for F-actin in the presence of ATP, the K(actin) of the myosin X ATPase was much higher than that of myosin V. Consistently nearly all actin dissociated from myosin X in the presence of ATP. ADP did not significantly inhibit the actin-activated ATPase activity of myosin X, suggesting that the ADP release step is not rate-limiting. These results suggest that myosin X is a nonprocessive motor. Consistently myosin X failed to support the actin translocation at low density in an in vitro motility assay where myosin V, a processive motor, supports the actin filament movement.     

Differences between smooth and skeletal muscle myosins in their interactions with F-actin

Myosin and F-actin were prepared from bovine carotid arterial smooth muscle and the properties of the binding of myosin to F-actin were compared with those of the binding of skeletal muscle myosin to F-actin. The following differences were observed between skeletal and smooth muscle myosins. 1. The rate of ATP-induced dissociation of arterial actomyosin was equal to that of hybrid actomyosin reconstituted from arterial myosin and skeletal muscle F-actin, but was much lower than those of skeletal muscle actomyosin and of hybrid actomyosin reconstituted from skeletal muscle myosin and arterial F-actin. 2. The amount of ATP necessary for complete dissociation of arterial actomyosin was 2 mol/mol of myosin, although it is well known that skeletal muscle actomyosin is dissociated completely by the addition of 1 mol ATP per mol of myosin. 3. Arterial actomyosin and hybrid actomyosin reconstituted from arterial myosin and skeletal muscle F-actin did not dissociate upon addition of 0.1 mM PPi, while skeletal muscle actomyosin dissociated completely. 4. In the absence of Mg2+, neither dissociation by ATP nor ATPase [EC 3.6.1.3] activity was observed with arterial actomyosin and hybrid actomyosin reconstituted from arterial myosin and skeletal muscle F-actin. On the other hand, skeletal muscle actomyosin dissociated almost completely upon addition of ATP and showed a considerably high ATPase activity. These observations reveal marked differences between myosins from skeletal and smooth muscles in their binding properties to F-actin.     

Diffusion of H-meromyosin in F-actin plus ATP solution at a very low electrolyte concentration

The translational diffusion coefficient (D) of H-meromyosin in actin (F-actin) and ATP solution was measured under conditions wherein the actin-activated ATPase activity is close to its maximal value at a very low electrolyte concentration. The results were compared with similar data obtained with 0.1 M KCl, where H-meromyosin and actin were almost completely dissociated. With 0.1 M KCl, it was found that there was no dependence of the D of H-meromyosin on actin concentration. On the other hand, at a very low electrolyte concentration, it was found that the D of H-meromyosin did depend on actin concentration; at a rather high actin concentration (and activation of ATPase), it was slightly larger than at low or zero actin concentrations. This behavior of D at a low electrolyte concentration is interpreted on the assumption that even in solution, H-meromyosin molecules can actively slide on actin filaments due to the ATPase activity.     

Extraction of an actin-like protein from the plasmodium of a myxomycete and its interaction with myosin A from rabbit striated muscle

An actin-like protein was obtained from the plasmodia of a myxomycete, Physarum polycephalum. It forms a complex with muscle myosin A which behaves similarly to the actomyosin from rabbit striated muscle. On the addition of ATP the complex of this protein with myosin A shows a viscosity drop at high concentrations of KCl (～0.5 M). At low concentrations of KCl (～0.05 M) this complex superprecipitates from solutions containing 1 mM MgCl₂ and shows Mg-activated ATPase activity. That is, the actin-like protein converts the ATPase of myosin A to the actomyosin type.     

Effects of phosphorylated and unphosphorylated C-protein on cardiac actomyosin ATPase

C-protein, a component of the thick filaments of striated muscles, is reversibly phosphorylated and dephosphorylated in heart. It has been hypothesized that C-protein may be involved in regulating contraction, because the extent of C-protein phosphorylation correlates with the rate of cardiac relaxation. To test this hypothesis, the effects of phosphorylated and unphosphorylated C-protein on the actin-activated ATPase activity of myosin filaments prepared from DEAE-Sephadex-purified myosin were examined. Unphosphorylated C-protein (0.1 microM to 1.5 microM) stimulated actin-activated myosin ATPase activity in a dose-dependent manner. With a myosin: C-protein molar ratio of approximately 1, actin-activated myosin ATPase activity was elevated up to 3.2 times that of the control. Phosphorylated C-protein (2.5 mol PO4/mol C-protein) stimulated the activity somewhat less (2.5 times that of control). The stimulation of ATPase activity by C-protein was due to an increase in the Vmax value (from 0.25/second to 0.62/second) and a decrease in the Km value (from 11.9 microM to 6.7 microM). The addition of C-protein to actomyosin solutions produced an increase in the light-scattering of the actomyosin solution and a distinct precipitation of the actomyosin with time. Phosphorylated C-protein had a smaller effect on light-scattering than dephosphorylated C-protein. C-protein had a negligible effect on Ca-ATPase, EDTA-K-ATPase, or Mg-ATPase activities in the absence of actin. C-protein had only small effects on the actin-activated ATPase of heavy meromyosin. These results suggest that C-protein stimulates actin-activated myosin ATPase activity by enhancing the formation of stable aggregates between actin and myosin filaments.     

A simple method for the isolation of actin from myxomycete plasmodia.

A new, simple method for the isolation of actin from myxomycete plasmodia has been developed. Plasmodium myosin B was incubated at 55 degrees C for 15 min in the presence of ATP or was treated with 90% acetone. By this treatment myosin was denatured completely. Actin was then extracted with a dilute ATP and cysteine solution from the heat- or acetone-treated myosin B. The method is simple and almost pure actin was obtained in high yield. The purified G-actin polymerized to F-actin on addition of 0.1 M KCl or 2 mM MgCl2. The viscosity of the purified F-actin was 8-10 dl/g. The F-actin activated muscle myosin ATPase, and actomyosin synthesized from the F-actin and muscle myosin showed superprecipitation on addition of ATP.