Effect of F-actin upon the binding of ADP to myosin and its fragments.

The effect of F-actin upon the binding of ADP to rabbit skeletal muscle myosin, heavy meromyosin, and subfragment 1 was studied by equilibrium dialysis, ultracentrifuge transport, and light scattering techniques. Both myosin and H-meromyosin (HMM) bind a maximum of approximately 1.6 mol of ADP/mol of protein, while S-1 binds approximately 0.9 mol of ADP/mol of protein. The affinity for ADP of all three preparations was similar at a given ionic strength (approximately 10(6) M-1 at 0.05 M KCl) and decreased with increasing ionic strength. Under conditions similar to those used for the measurement of ADP binding, the binding sites of myosin, HMM, and subfragment 1 (S-1) are saturated with actin at molar ratios of 2, 2, and 1 mol of actin monomer/mol of protein, respectively, as determined by light scattering, ultracentrifuge transport, and in the case of myosin by ATPase measurements. F-actin was found to inhibit ADP binding, but even at an actin concentration at least twice that required for saturation of myosin, HMM, or S-1, significant ADP binding remained. This ADP binding was inhibited by 10(-4) M pyrophosphate. The observations are consistent with the formation of an actomyosin-ADP complex in which actin and ADP are bound to myosin at distinct but interacting sites.

The effect of F-actin upon the binding of ADP to rabbit skeletal muscle myosin, heavy meromyosin, and subfragment 1 was studied by equilibrium dialysis, ultracentrifuge transport, and light scattering techniques. Both myosin and H-meromyosin (HMM) bind a maximum of approximately 1.6 mol of ADP/mol of protein, while S-1 binds approximately 0.9 mol of ADP/mol of protein. The affinity for ADP of all three preparations was similar at a given ionic strength (approximately lo6 M-' at 0.05 M KCl) and decreased with increasing ionic strength. Under conditions similar to those used for the measurement of ADP binding, the binding sites of myosin, HMM, and subfragment 1 (S-1) are saturated with actin at molar ratios of 2, 2, and 1 mol of actin monomer/mol of protein, respectively, as determined by light scattering, ultracentrifuge transport, and in the case of myosin by ATPase measurements. F-actin was found to inhibit ADP binding, but even at an actin concentration at least twice that required for saturation of myosin, HMM, or S-1, significant ADP binding remained. This ADP binding was inhibited by lo-' M pyrophosphate.
The observations are consistent with the formation of an actomyosin.ADP complex in which actin and ADP are bound to myosin at distinct but interacting sites.
During the contraction cycle, binding of ATP dissociates actomyosin followed by ATP hydrolysis and the formation of a myosin product complex (2). When the myosin head interacts with a new actin molecule, the products of the reaction are released from the active site and the crossbridges develop tension (2). In agreement with this scheme actomyosin in solution is readily dissociated by ATP (3), and F-actin decreases the affinity of ADP (4), PP, (5, 6), and ATP (7) for myosin. As 2 mol of actin (8), ADP (4,9), ATP (7, lo), and PP, (6, 11) are bound per mol of myosin, each of the myosin heads is likely to contain one actin and one closely related nucleotide binding site (6,8,12).
The functional significance of the duplicate set of sites is not known. Although Tokiwa and Morales (13) suggested that both heads are required for contraction, Margossian and Lowey (8)  usually interpreted in terms of a single set of noninteracting sites. Differences between the two nucleotide binding sites were inferred from the ultraviolet difference spectra of myosin induced by ATP, ADP,or PP,(14) and from the stoichiometry of the rapid initial burst of ATP hydrolysis (14, 15). Indirect indications of cooperativity between the two sets of sites came from kinetic data (16) and from electron spin resonance measurements (17). We have reported earlier that the binding of ADP to myosin is inhibited by actin (4). Surprisingly, even at actin to myosin ratios higher than that required for the saturation of myosin, significant ADP binding remained. As the precise relationship of the two ADP binding sites of myosin to the actin binding would have significant implications upon the mechanism of crossbridge movement during muscle contraction, the effect of actin upon the ADP binding of myosin was investigated in detail.

EXPERIMENTAL PROCEDURES
Myosin was isolated from white rabbit skeletal muscle as described earlier (4) Yount and Koshland (20) and stored at -20". HMM S-l fragments were prepared according to Young et al. (21)  The freshly prepared S-l fractions were used within 1 week. In some cases S-l was prepared according to Margossian and Lowey (8). The data obtained with the two types of preparations were similar.
Actin was prepared from acetone-dried muscle powder as described earlier (22 The nucleotide spots were cut out and the radioactivity counted by the method of Loftfield and Eigner (25). Some of the reported binding data were corrected for ADP decomposition which was usually less than 208 of the total ADP in an equilibrium dialysis experiment of 12 to 20 hours duration at 3-5". With the centrifuge transport method (see below) no correction was required because of the shorter incubation time.

Ultracentrifuge Transport
The protein solution containing 50 mM Tris-HCl buffer, pH 8.0, 1.0 mM MgCl,, and ['%]ADP in varying concentration was placed in an ultracentrifuge tube of 2-ml volume and centrifuged at 80,000 x g for 4 to 12 hours at 5" in a Spinco No. 40 1A) and decreases to approximately 1.7 x lo5 M-' at a KC1 concentration of 0.6 M (Fig. 1C). The maximum number of ADP binding sites is 1.4 at low ionic strength (Fig. 1A) and 1.60 and 1.70 at KC1 concentrations of 0.3 M and 0.6 M, respectively (Fig. 1, B and C). This is in essential agreement with previous reports (4,9) and is usually taken to indicate the existence of two ADP binding sites per mol of myosin. Myosin is in the form of filamentous aggregates at low ionic strength and forms true solution only at a KC1 concentration above 0.3 M. The marked deviation of the Scatchard plot from linearity at low ionic strength (Fig. 1A) may be related in part to the aggregated state of the proteins and in part to ADP decomposition which affects the free ADP especially at low ADP concentrations.
Addition of F-actin to myosin in actin monomer to myosin mole ratios ranging from 0 to 4 causes an inhibition of ADP binding ( Fig. 1 A to C). At low ionic strength the inhibition is largely noncompetitive as indicated by the marked change in the maximum number of ADP binding sites as the concentration of F-actin increases.
Even at an actin/myosin ratio of 4, a significant amount of ADP remained bound to myosin. At this actin concentration the actin binding sites of myosin as judged 7873 from the effect of actin upon the Mg2+-moderated ATPase activity of myosin are fully saturated (Fig. 2). At 0.3 M KC1 (Fig. 1B) or 0.6 M KC1 concentration (Fig. 1C) the inhibitory effect of F-actin upon the ADP binding appears competitive up to an actin to myosin ratio of 2 where on the basis of light scattering measurements (Fig. 3) myosin is fully saturated with actin.
Surprisingly, at an actin to myosin mole ratio of 2, under the conditions of these experiments a significant portion of the ADP binding sites remains occupied with ADP, and a further increase in the actin concentration to an actin to myosin ratio of 4 has little effect upon the residual binding to ADP to myosin. This is not attributable to the dissociation of actomyosin by ADP since ADP up to 100 HIM concentration had little effect upon the steady state rate of actin-activated ATP hydrolysis or upon the light scattering of actomyosin solutions measured at 25" in the presence of 0.6 M KCl, 50 mM Tris, pH  (Table I). ADP can be readily displaced from the actin-insensitive sites in the presence of 10e5 to lo-' M inorganic pyrophosphate (Fig. 4), in agreement with the previously demonstrated simple competition between these two substrate analogs (4, 6).
Effect of F-Actin upon Binding of ADP to H-Meromyosin-The tendency of myosin to form aggregates below 0.3 M KC1 concentration complicates the interpretation of the competition between actin and ADP at physiological ionic strengths. Hence the effect of F-actin on the ADP binding of H-meromyosin was investigated. Scatchard plots of ADP binding at 0.05 M (Fig. 5A), 0.3 M (Fig. 5B), and 0.6 YV (Fig. 5C) KC1 concentration indicate inhibition by F-actin with a decrease in the apparent affinity of ADP binding without a significant change in the maximum number of binding sites. The linearity of the Scatchard plots of H-meromyosin .ADP interaction at low ionic strength (Fig. 5A) is in contrast to the marked nonlinearity observed with myosin ( Fig. 1A). At an actin to H-meromyosin ratio of 2, significant ADP binding remains at free ADP concentrations as low as 1 to 15 pM (Fig. 5 A   The stoichiometry of actin. H-meromyosin interaction was investigated by light scattering (Fig. 6) and ultracentrifuge measurements (Fig. 7). Irrespective of ionic strength, clear-cut saturation of H-meromyosin with F-actin was obtained at an actin:H-meromyosin ratio of 2. The presence of ADP up to lo-' M concentration had little influence upon the actin .HMM interaction as judged by ultracentrifuge transport (Table I). Binding of Actin and ADP to Subfragment I-Myosin and H-meromyosin contain two nucleotide and actin binding sites located on the two head portions of the molecule. The insensitivity of some of the nucleotide binding to actin may indicate that F-actin lowers the affinity of ADP binding at two equivalent sites by a negative interaction between the actin and the nucleotide binding sites. Alternatively, chemical differences between the two sites or steric effects connected with the proximity of the two otherwise identical head portions of the myosin or H-meromyosin molecules may be considered. In order to distinguish between these alternatives the effect of actin upon the binding of ADP to S-l fragments was studied (Fig. 8).
At KC1 concentrations ranging from 0.05 to 0.6 M, S-l fragments bound close to 1 mol of ADP/lOO,OOO g of protein.
The affinity constant of ADP binding was 2.4 x lo6 Mm1 at 0.05 M KC1 (Fig. 8A) and decreased with increasing KC1 concentration to 5.4 x lo5 Me' (Fig. 8C). Maximum inhibition of ADP binding was obtained at an actin to S-l molar ratio of 1 where approximately one-half of the ADP bound by S-l in the absence of actin was displaced from the binding site (Fig. 8,  insets). Increasing the actin to S-l mole ratio from 1 to as high as 8 mol of actin/mol of S-l fragment did not cause a significant change in the amount of ADP bound to S-l at free ADP concentrations ranging from 1 to 15 pM (Fig. 8, insets). Direct measurement of the binding of F-actin to S-1 fragments in the analytical ultracentrifuge showed that at an actin to S-l mole ratio of 1 and above, no detectable free S-l Light scattering measurements were carried out as described under "Experimental Procedures" in a medium of 0.6 M KCl, 44 mM Tris, pH 8.0, and 1.5 mM M&l,. The concentration of H-meromyosin was 0.23 mg/ml with F-actin added in the molar ratios indicated on the abscissa. The points represent the relative intensity of the scattered light with solutions containing F-actin and H-meromyosin in the indicated molar ratios, after subtraction of the 90" scattering given by the same concentrations of actin and the H-meromyosin measured separately.
remained and therefore essentially all S-l is considered to be bound to F-actin (Fig. 9).
These observations follow the pattern described for myosin and H-meromyosin.
Saturation of S-1 fragments with F-actin causes the displacement of approximately one-half of the bound ADP from the active site. The remaining ADP binding is unaffected by actin even at high actin/S-l molar ratios. It is unlikely that the ADP binding observed with saturating concentrations of F-actin is due to the nonspecific interaction of ADP with S-l fragments since the ADP binding is inhibited by 10m6 to 10m5 M inorganic pyrophosphate, a substrate analog known to interact with the nucleotide binding site of myosin (Fig. 10). Scatchard plots of ADP binding in F-actin-free systems indicate a single set of noninteracting sites with an affinity constant of the order of lo6 to lo5 M which is much higher than the expected affinity for nonspecific anion binding sites.
The apparent insensitivity of a portion of the ADP binding of S-l to F-actin is not attributable to the dissociation of acto-S-l complex by ADP since 0.01 to 0.1 mM ADP had no significant effect upon the concentration of free S-l fragments in a system containing 0.8 mol of actin/mol of S-l protein (Table I).

DISCUSSION
The observations presented in this report confirm earlier findings (4,8,9) about the existence of two ADP and two actin binding sites in myosin or H-meromyosin and one of each in subfragment 1. The equilibrium constant of ADP binding at the two sites of myosin is similar.
F-actin inhibits the binding of ADP to myosin but significant ADP binding occurs even at actin to myosin ratios where myosin is fully saturated with actin. Similar observations were made with H-meromyosin and with subfragment 1, indicating that the results are not attributable to the formation of myosin filaments or to steric interference between the two heads of a myosin molecule.
The ADP binding observed in the presence of a saturating concentration of F-actin is inhibited by lo-5 to lo6 M inorganic pyrophosphate, and the inhibitor constant of pyrophosphate calculated from the competitive inhibition is similar to the previously determined (6) dissociation constant of pyrophosphate for the active site of myosin. Therefore, the F-actinresistant ADP binding presumably occurs at the active site. S-l with 10m5 M inorganic pyrophosphate; X-~-X, S-l with lo-" M inorganic pyrophosphate. Inset figures contain the same data arranged in double reciprocal plot. The inhibitor constant (K,) of inorganic pyrophosphate is approximately 3.1 x 10 -' M in essential agreement with the results obtained by direct measurement of PP, binding to myosin (6).
Exchange of the bound nucleotide of F-actin is not likely to account for the F-actin-insensitive ADP binding since the extent of such exchange is limited (24, 28,29). Furthermore, inorganic pyrophosphate nearly completely inhibited the F-actin-insensitive ADP binding of actomyosin although its affinity for F-actin is negligible (23). Hanson et al. (20) observed that actin forms lateral aggregates at high (70 mM) MgCl, concentration.
Aggregation of actin would explain the incomplete inhibition of ADP binding at high actin concentration, since myosin would interact only with actin molecules on the surface of these aggregates. However, at the MgCl, concentration used in these experiments (1 to 1.5 mM) such aggregation is negligible and the saturation of myosin or its fragments with F-actin shows clear stoichiometry.
The most likely explanation of the observed effects of F-actin upon the ADP binding is that F-actin and ADP are bound simultaneously to myosin leading to the formation of an actomyosin .ADP complex. The inhibitory effect of F-actin upon the ADP binding may be explained by the negative interaction between the actin binding sites and the distinct nucleotide binding sites, as indicated in the following scheme: F-actin + ADP + Myosin ctomyosin-ADP where K,, Kg, K',, and K', denote affinity constants. It is assumed that the two nucleotide binding sites located on the two myosin heads are identical and the same is true for the actin binding site. According to the data, binding of F-actin to myosin lowers the affinity of ADP for the nucleotide binding site by at least one order of magnitude (K, >> K',), while ADP has apparently little effect upon the binding of F-actin to myosin (5). The ADP binding observed in the presence of a saturating concentration of F-actin is defined by K',, the affinity constant of ADP binding to actomyosin. In this interpretation the relationship between the actin and the nucleotide binding sites is not truly competitive but involves some interaction between the two types of sites.
Simultaneous binding of actin and ADP to myosin was also observed in glycerinated muscle fibers (39). In the absence of ATP, glycerinated fibers are in rigor, i.e. the myosin crossbridges are linked to actin. Yet these fibers bind the same amount of ADP that can be bound during ATP hydrolysis. Maximum ADP binding was achieved without a significant change in the elastic modulus, indicating that ADP was bound to the fibers without dissociating actin-myosin links. This is in agreement with previous observations showing that ADP dissociates actomyosin only at very high concentrations (5). The existence of actin-insensitive ADP binding to myosin in living muscle remains to be established (40). As the bound ADP of actomyosin is readily displaced by ATP it is likely that in living muscle the actomyosin ADP complex is immediately converted into myosin ATP. 1.

5.
The nucleotide and actin binding sites apparently involve distinct functional groups (31)(32)(33)(34), and conditions exist for the selective modification of each set of sites. Treatment of actomyosin with --SH group reagents makes it insensitive to the dissociating effect of ATP without breaking actomyosin links (5). Myofibrils treated in a similar manner lose both their contractility and relaxing response (31). After treatment of actomyosin with a higher concentration of N-ethylmaleimide or iodoacetamide at pH 8.0, complete inhibition of ATPase activity was obtained without dissociation of actomyosin (32)(33)(34). The myosin component isolated from these preparations formed actomyosin which was resistant to the dissociating effect of ATP. Finally, N-ethylmaleimide in high concentration inhibits the Ca*+-or EDTA-moderated ATPase activity of myosin, but leaves the binding of ATP (35) and ADP (36,37) relatively unaffected. The selective inhibition of the different functions of myosin by -SH group reagents implies that the catalytic and actin binding sites are separate but related. Conditions may exist for the chemical modification of myosin which leaves the binding of actin and ADP relatively unaffected but eliminates the negative interaction between the actin and nucleotide binding sites.
The observed effects of F-actin upon the ADP binding may also be explained by models based upon differences in the intrinsic affinity and actin sensitivity of the ADP binding sites 22. 23. 24. 25. located on the two heads of myosin. Among early indications of 26. 7877 binding site heterogeneity are the stoichiometry of early phosphate burst of ATP hydrolysis (14, 15) and the relationship between the ultraviolet difference spectrum and the binding of ATP, ADP, and pyrophosphate to H-meromyosin (14, 38). These models require special assumptions to fit the data and therefore appear less plausible. In view of the large degree of interpretive freedom permitted by the binding data, the demonstration of two populations of S-l fragments, each capable of ADP binding but differing in actin sensitivity, would be required to substantiate this possibility.