Reversible phosphorylation of smooth muscle myosin, heavy meromyosin, and platelet myosin

Smooth muscle myosin was purified from turkey gizzards with the 20,000-dalton light chains in the unphosphorylated state. The actin-activated MgATPase activity was 4 nmol/min/mg at 25 degrees C. When the myosin was phosphorylated to 2 mol of Pi/mol of myosin using purified myosin light chain kinase, calmodulin, and ATP, the actin-activated MgATPase activity rose to 51 nmol/min/mg. Complete dephosphorylation of the same myosin by a purified phosphatase lowered the activity to 5 nmol/min/mg, and complete rephosphorylation of the myosin following inhibition of the phosphatase raised it again to 46 nmol/min/mg. Human platelet myosin could be substituted for turkey gizzard myosin, with similar results. A chymotryptic fragment of smooth muscle myosin which retains the phosphorylated site on the 20,000-dalton light chain of myosin was prepared. Using the same scheme for reversible phosphorylation, this smooth muscle heavy meromyosin was found to show the same positive correlation between phosphorylation of the myosin light chain and the actin-activated MgATPase activity. The results with smooth muscle heavy meromyosin show that the effect of phosphorylation on the actin-activated MgATPase activity can be separated from the effects of phosphorylation on myosin filament assembly.     

Protein kinase C modulates in vitro phosphorylation of the smooth muscle heavy meromyosin by myosin light chain kinase

Protein kinase C phosphorylates different sites on the 20,000-Da light chain of smooth muscle heavy meromyosin (HMM) than did myosin light chain kinase (Nishikawa, M., Hidaka, H., and Adelstein, R. S. (1983) J. Biol. Chem. 258, 14069-14072). Although protein kinase C incorporates 1 mol of phosphate into 1 mol of 20,000-Da light chain when either HMM or the whole myosin molecule is used as a substrate, it catalyzes the incorporation of up to 3 mol of phosphate/mol of 20,000-Da light chain when the isolated light chains are used as a substrate. Threonine is the major phosphoamino acid resulting from phosphorylation of HMM by protein kinase C. Prephosphorylation of HMM by protein kinase C decreases the rate of phosphorylation of HMM by myosin light chain kinase due to a 9-fold increase of the Km for prephosphorylated HMM compared to that of unphosphorylated HMM. Prephosphorylation of HMM by myosin light chain kinase also results in a decrease of the rate of phosphorylation by protein kinase C due to a 2-fold increase of the Km for HMM. Both prephosphorylations have little or no effect on the maximum rate of phosphorylation. The sequential phosphorylation of HMM by myosin light chain kinase and protein kinase C results in a decrease in actin-activated MgATPase activity due to a 7-fold increase of the Km for actin over that observed with phosphorylated HMM by myosin light chain kinase but has little effect on the maximum rate of the actin-activated MgATPase activity. The decrease of the actin-activated MgATPase activity correlates well with the extent of the additional phosphorylation of HMM by protein kinase C following initial phosphorylation by myosin light chain kinase.     

Effect of phosphorylation on the binding of smooth muscle heavy meromyosin X ADP to actin

Relaxation of both smooth and skeletal muscles appears to be caused primarily by inhibition of the step associated with Pi release in the actomyosin ATPase cycle, rather than by a block in the binding of the myosin X ATP and myosin X ADP X Pi complexes to actin. In skeletal muscle, troponin-tropomyosin not only causes marked inhibition of Pi release, but it also markedly inhibits the binding of myosin subfragment-1 X ADP to actin, raising the possibility that the two phenomena are coupled in some way. In the present study we determined whether phosphorylation of smooth muscle heavy meromyosin (HMM) also affects both the binding of HMM X ADP to actin and the Pi release step. This was done by having phosphorylated and unphosphorylated HMM X ADP compete for sites on F-actin. At mu = 30 mM, phosphorylation increased the affinity of the HMM molecule for actin about 12-fold and at mu = 170 mM, there was less than a 3-fold increase in the affinity of HMM. If phosphorylation affects the binding of each head of HMM to the same extent, then phosphorylation caused about a 4- and 2-fold increase in the affinity of each head of HMM for actin at mu = 30 and 170 mM, respectively. In contrast, at both ionic strengths, phosphorylation caused more than 100-fold actin activation of the ATPase activity of smooth muscle HMM. Therefore, the marked activation of Pi release in the acto X HMM ATPase cycle upon phosphorylation of HMM is not accompanied by a comparable increase in the affinity of HMM X ADP for actin. We have also found that phosphorylation increases by only 4-fold the rate of Pi release from HMM alone. These results suggest that in smooth muscle, phosphorylation accelerates the step associated with the release of Pi both in the forward and the reverse direction without correspondingly affecting the binding of myosin X ADP to actin.     

Phosphorylation of vertebrate nonmuscle and smooth muscle myosin heavy chains and light chains

In this article we review the various amino acids present in vertebrate nonmuscle and smooth muscle myosin that can undergo phosphorylation. The sites for phosphorylation in the 20 kD myosin light chain include serine-19 and threonine-18 which are substrates for myosin light chain kinase and serine-1 and/or-2 and threonine-9 which are substrates for protein kinase C. The sites in vertebrate smooth muscle and nonmuscle myosin heavy chains that can be phosphorylated by protein kinase C and casein kinase II are also summarized.Original data indicating that treatment of human T-lymphocytes (Jurkat cell line) with phorbol 12-myristate 13-acetate results in phosphorylation of both the 20 kD myosin light chain as well as the 200 kD myosin heavy chain is presented. We identified the amino acids phosphorylated in the human T-lymphocytes myosin light chains as serine-1 or serine-2 and in the myosin heavy chains as serine-1917 by 1-dimensional isoelectric focusing of tryptic phosphopeptides. Untreated T-lymphocytes contain phosphate in the serine-19 residue of teh myosin light chain and in a residue tentatively identified as serine-1944 in the myosin heavy chain.Abbreviations MLC myosin light chain - MHC myosin heavy chain - Tris tris(hydroxymethyl)aminomethane - EGTA [ethylenebis(oxyethylenenitrilo)]tetraacetic acid - EDTA ethylenediaminetetraacetate - TPCK N-tosyl-L-phenylalanine chloromethyl ketone - PMA phorbol 12-myristate 13-acetate     

Proteolysis of smooth muscle myosin by Staphylococcus aureus protease: preparation of heavy meromyosin and subfragment 1 with intact 20 000-dalton light chains

The proteolysis of gizzard myosin by Staphylococcus aureus protease produces both heavy meromyosin and subfragment 1 in which the 20 000-dalton light chains are intact, and conditions are suggested for the preparation of each. Cleavage of the myosin heavy chain to produce subfragment 1 is dependent on the myosin conformation. Proteolysis of myosin in the 10S conformation yields predominantly heavy meromyosin, and myosin in the 6S conformation yields mostly subfragment 1 and some heavy meromyosin. Two sites are influenced by myosin conformation, and these are located at approximately 68 000 and 94 000 daltons from the N-terminus of the myosin heavy chain. The latter site is thought to be located at the subfragment 1-subfragment 2 junction, and cleavage at this site results in the production of subfragment 1. The time courses of phosphorylation of both heavy meromyosin and subfragment 1 can be fit by a single exponential. The actin-activated Mg2+-ATPase activity of heavy meromyosin is markedly activated by phosphorylation of the 20 000-dalton light chains. From the actin dependence of Mg2+-ATPase activity the following values are obtained: for phosphorylated heavy meromyosin, Vmax approximately 5.6 s-1 and Ka (the apparent dissociation constant for actin) approximately 2 mg/mL; for dephosphorylated heavy meromyosin, Vmax approximately 0.2 s-1 and Ka approximately 7 mg/mL. The actin-activated ATPase activity of subfragment 1 is not influenced by phosphorylation, and Vmax and Ka for both the phosphorylated and dephosphorylated forms are 0.4 s-1 and 5 mg/mL, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)     

Phosphorylation-dependent structural changes in the regulatory light chain domain of smooth muscle heavy meromyosin.

Smooth muscle heavy meromyosin, a double-headed proteolytic fragment of myosin lacking the COOH-terminal two-thirds of the tail, has been shown previously to be regulated by phosphorylation. To examine phosphorylation-dependent structural changes near the head-tail junction, we prepared five well regulated heavy meromyosins containing single-cysteine mutants of the human smooth muscle regulatory light chain labeled with the photocross-linking reagent, benzophenone-iodoacetamide. For those mutants that generated cross-links, only one type of cross-linked species was observed, a regulatory light chain dimer. Irradiated mutants fell into two classes. First, for Q15C, A23C, and wild type (Cys-108), a regulatory light chain dimer was formed for dephosphorylated but not thiophosphorylated heavy meromyosin. These data provide direct chemical evidence that in the dephosphorylated state, Gln-15, Ala-23, and Cys-108 on one head are positioned near (within 8.9 A) the regulatory light chain of the partner head and that thiophosphorylation abolishes proximity. This behavior was also observed for the Q15C mutant on a truncated heavy meromyosin lacking both catalytic domains. For the actin-heavy meromyosin complex, cross-links were formed in both de- and thiophosphorylated states. S59C and T134C mutants were in a second mutant class, where regulatory light chain dimers were not detected in dephosphorylated or thiophosphorylated heavy meromyosin, suggesting positions outside the region of interaction of the regulatory light chains.     

Interaction between the heavy and the regulatory light chains in smooth muscle myosin subfragment 1.

The interaction between the heavy and the regulatory light chains within chicken gizzard myosin heads was investigated by using a zero-length chemical cross-linker, 1-ethyl-3-[3-(dimethylamino)-propyl]carbodiimide (EDC). The chicken gizzard subfragment 1 (S-1) used was treated with papain so that the heavy chain was partly cleaved into the NH2-terminal 72K and the COOH-terminal 24K fragments and the regulatory light chain into the 16K fragment. S-1 was reacted with EDC either alone or in the presence of ATP or F-actin. In all cases, the 16K fragment of the regulatory light chain formed a covalent cross-link with the 24K heavy chain fragment but not with the 72K fragment. The 38K cross-linked peptide, which was the product of cross-linking between the 16K light chain and the 24K heavy chain fragments, was isolated and further cleaved with cyanogen bromide and arginylendopeptidase. Smaller cross-linked peptides were purified by reverse-phase HPLC and then characterized by amino acid analysis and sequencing. The results indicated that cross-linking occurred between Lys-845 in the heavy chain and Asp-168, Asp-170, or Asp-171 in the regulatory light chain. The position of the cross-linked lysine was only three amino acid residues away from the invariant proline residue mapped as the S-1-rod hinge by McLachlan and Karn [McLachlan, A. D., & Karn, J. (1982) Nature (London) 299, 226-231]. We propose that the COOH-terminal region of the regulatory light chain is located in the neck region of myosin and that this region and the phosphorylation site of the regulatory light chain together may play a role in the phosphorylation-induced conformational change of gizzard myosin.     

Phosphorylation of smooth muscle myosin heavy and light chains. Effects of phorbol dibutyrate and agonists

A number of different protein kinases phosphorylate purified heavy chains or the 20-kDa light chain of smooth muscle myosin. The physiological significance of these phosphorylation reactions has been examined in intact smooth muscle. Myosin heavy chain was slightly phosphorylated (0.08 mol of phosphate/mol) under control conditions in bovine tracheal tissue. Treatment with carbachol, isoproterenol, or phorbol 12,13-dibutyrate resulted in no significant change. In contrast, heavy chain was phosphorylated to 0.30 mol of phosphate/mol of heavy chain in tracheal smooth muscle cells in culture. This value increased significantly with ionomycin treatment. In control tissues, 9% of the light chain was monophosphorylated with 32P in the serine site phosphorylated by myosin light chain kinase. Carbachol (0.1 microM) alone resulted in contraction and 42% monophosphorylated light chain with 32P only in the serine site phosphorylated by myosin light chain kinase. Similarly, stimulation with histamine, 5-hydroxytryptamine, or KCl resulted in 32P incorporation into only the myosin light chain kinase serine site. Phorbol 12,13-dibutyrate (1 microM) alone resulted in 22% monophosphorylated light chain. However, only 25% of the 32P was in the myosin light chain kinase serine site, whereas 75% was in a serine site phosphorylated by protein kinase C. Phorbol 12,13-dibutyrate plus carbachol resulted in 27% monophosphorylated light chain; 75% of the 32P was in the myosin light chain kinase serine site, with the remainder in the protein kinase C serine site. These results indicate that phorbol esters act to increase phosphorylation of myosin light chain by protein kinase C. However, receptor-mediated stimulation or depolarization leading to tracheal smooth muscle contraction results in phosphorylation of myosin light chain by myosin light chain kinase alone.     

Myosin assembly critical for the enzyme activity of smooth muscle myosin phosphatase: effects of MgATP, ionic strength, and Mg(2+)

We suggested that an assembled form of phosphorylated myosin (P-myosin) might exhibit higher affinity for smooth muscle myosin phosphatase (SMMP) than dissociated P-myosin on the basis of the effect of MgATP [Sato and Ogawa (1999) J. Biochem. 126, 787-797]. To further deepen our understanding, we examined the SMMP activity and P-myosin assembly with various ionic strengths and Mg(2+) concentrations, with and without MgATP, all of which are well known to be critical for myosin assembly. The structure of myosin molecules was directly observed by electron microscopy using a rotary shadowing procedure, which was found to be consistent with the sedimentation assay. We found that the SMMP activity was always high when P-myosin was assembled. MgATP, which disassembled P-myosin mostly into a folded conformation, in contrast, decreased the enzyme activity. We also found that glycerol had a dissociating action on P-myosin, primarily dissociating it into an extended conformation, resulting in reduced SMMP activity, and that increases in the ionic strength and Mg(2+) (>5 mM) inhibited SMMP. These results indicate that myosin assembly is essential for SMMP activity.     

The effect of phosphorylation of myosin light chains on the structural state of tropomyosin in thin filaments, decorated with heavy meromyosin

The structural state of tropomyosin (TM) modified by 5-(iodoacetamidoethyl)-aminonaphthalene-1-sulfonate (1.5-IAEDANS) upon F-actin decoration with myosin subfragment 1 (S1) and heavy meromyosin (HMM) in glycerinated myosin- and troponin-free muscle fibers was studied. HMM preparations contained native phosphorylated myosin light chains, while S1 preparations did not. The changes in the polarized fluorescence of 1.5-IAEDANS-TM during the F-actin interaction with S1 were independent of light chains phosphorylation and Ca2+ concentration, but were dependent on these factors during the F-actin interaction with HMM. The binding of myosin heads to F-actin is supposed to initiate conformational changes in TM which are accompanied by changes in the flexibility and molecular arrangement of TM. In the presence of light chains, the structural changes in TM depend on light chains phosphorylation and Ca2+ concentration. The conformational changes in TM seem to be responsible for the mechanisms of coupling of the myosin and tropomyosin modulation system during the actin-myosin interaction in skeletal muscles.     

Calcium-sensitive binding of heavy meromyosin to regulated actin requires light chain 2 and the head-tail junction

Sedimentation in a preparative ultracentrifuge was used to determine the affinity of heavy meromyosin, HMM, for regulated actin, F-actin plus troponin-tropomyosin, in the presence of MgATP at pH 7.0, 20 degrees C, and mu = 18 mM. HMM was prepared from vertebrate striated muscle myosin by a mild chymotryptic digestion. This HMM contained 85-90% intact 19 000-dalton light chains, LC2. In the presence of calcium, 90% of the HMM bound to regulated actin with an association constant of (2-4) X 10(4) M-1. In the absence of calcium, while one-third of the HMM bound with an affinity similar to that observed in the presence of calcium, the rest bound much more weakly. It was not possible to accurately determine the association constant for this weakly binding HMM, but a 20-fold reduction in affinity is consistent with the binding data. The binding of single-headed heavy meromyosin to regulated actin was similarly sensitive to the calcium concentration. Since removal of calcium inhibits the regulated actin-activated ATPase of HMM greater than 20-fold, troponin-tropomyosin must be capable of inhibiting both the binding of HMM to regulated actin and a step which occurs after binding but prior to product release. Removal of LC2 increased the fraction of HMM with calcium-insensitive binding, and adding LC2 back to this depleted HMM restored most of the calcium sensitivity. Chymotryptic cleavage of LC2 to a 17 000-dalton fragment destroyed the calcium-sensitive binding of HMM to regulated actin. Phosphorylation of LC2, however, had no detectable effect on this binding. Thus, the calcium-sensitive binding of HMM to regulated actin requires that both the head-tail junction and the N-terminal part of LC2 be intact. Binding studies with cross-linked regulated actins and kinetic measurements of the rates of change in turbidity demonstrate that this calcium sensitivity is due to calcium binding to troponin and not to LC2.     

Structures of smooth muscle myosin and heavy meromyosin in the folded, shutdown state

Remodelling the contractile apparatus within smooth muscle cells allows effective contractile activity over a wide range of cell lengths. Thick filaments may be redistributed via depolymerisation into inactive myosin monomers that have been detected in vitro, in which the long tail has a folded conformation. Using negative stain electron microscopy of individual folded myosin molecules from turkey gizzard smooth muscle, we show that they are more compact than previously described, with heads and the three segments of the folded tail closely packed. Heavy meromyosin (HMM), which lacks two-thirds of the tail, closely resembles the equivalent parts of whole myosin. Image processing reveals a characteristic head region morphology for both HMM and myosin, with features identifiable by comparison with less compact molecules. The two heads associate asymmetrically: the tip of one motor domain touches the base of the other, resembling the blocked and free heads of this HMM when it forms 2D crystals on lipid monolayers. The tail of HMM lies between the heads, contacting the blocked motor domain, unlike in the 2D crystal. The tail of whole myosin is bent sharply and consistently close to residues 1175 and 1535. The first bend position correlates with a skip in the coiled coil sequence, the second does not. Tail segments 2 and 3 associate only with the blocked head, such that the second bend is near the C-lobe of the blocked head regulatory light chain. Quantitative analysis of tail flexibility shows that the single coiled coil of HMM has an apparent Young's modulus of about 0.5 GPa. The folded tail of the whole myosin is less flexible, indicating interactions between the segments. The folded tail does not modify the compact head arrangement but stabilises it, indicating a structural mechanism for the very low ATPase activity of the folded molecule.     

Regulatory light chain-a myosin kinase (aMK) catalyzes phosphorylation of smooth muscle myosin heavy chains of scallop, Patinopecten yessoensis

Regulatory light chain-a myosin kinase (aMK), which phosphorylates one of the myosin regulatory light chains, RLC-a, contained in the catch muscle of scallop, was also found to phosphorylate heavy chains of scallop myosin. After incubation of myosin isolated from the opaque portion of scallop smooth muscle (opaque myosin) with aMK in the presence of [gamma-32P]ATP, about 2 mol of 32P was incorporated per mol of the myosin. The radioactivity was mostly found in the heavy chain at 0.26 M KCl. The pH-activity curve and MgCl2 requirement for the heavy chain phosphorylation were similar to those for RLC-a phosphorylation. In contrast, the dependency of activity on KCl concentration was different from that for RLC-a. The heavy chain phosphorylation activity decreased with increase in KCl concentration up to 0.06 M, and then increased at concentrations over 0.06 M to a maximum at around 0.26 M KCl. This complicated profile probably reflects the solubility of myosin, and the phosphorylation site may be located in the rod portion insoluble at low KCl concentrations. Phosphorylation of heavy chain did not change the solubility of the opaque myosin molecule at all. The acto-opaque myosin ATPase activity in the presence of Ca2+ was found to be decreased to less than one-fourth by the heavy chain phosphorylation.     

The binding of smooth muscle heavy meromyosin to actin in the presence of ATP. Effect of phosphorylation

Phosphorylation of the 20,000-dalton light chains of smooth muscle heavy meromyosin (HMM) from turkey gizzards results in a large increase in the actin-activated MgATPase activity over that observed with unphosphorylated HMM. In an attempt to define which step in the kinetic cycle is affected by phosphorylation, we have measured the binding of both unphosphorylated and phosphorylated HMM to actin in the presence of ATP using sedimentation. There was only a 4-fold difference in the actin binding constants of unphosphorylated HMM (5.35 x 10(3) M-1) and fully phosphorylated HMM (2.35 x 10(4) M-1). In contrast, the maximum rate of the actin-activated MgATPase activity (Vmax) of phosphorylated HMM was 25 times greater than that for unphosphorylated HMM. These data rule out a mechanism whereby the unphosphorylated light chain of myosin regulates actin-myosin interaction by directly or indirectly blocking the binding of HMM to actin. This implies that some step in the kinetic cycle other than the binding of HMM to actin must be regulated. We have also measured the rate constant for ATP hydrolysis (the initial phosphate burst) under the same conditions and found that this step was very fast compared to the steady state ATPase rate and was unaffected by phosphorylation. This suggests that the step which is regulated by phosphorylation is either phosphate release or a step preceding phosphate release but following ATP hydrolysis.     

Effects of MgATP, MgADP, and Pi on actin movement by smooth muscle myosin.

To test the idea that the in vitro motility assay is a simplified model system for muscle contraction, the MgATP-dependent movement of actin filaments by thiophosphorylated smooth muscle myosin was characterized in the presence of the products MgADP and inorganic phosphate. The dependence of actin filament velocity on MgATP concentration was hyperbolic with a maximum velocity of 0.6 micron/s and an apparent Km = 40 microM (30 degrees C). MgADP competitively inhibited actin movement by MgATP with a Ki = 0.25 mM. Inorganic phosphate did not affect actin filament velocity in the presence of 1 mM MgATP, but competitively inhibited movement in the presence of 50 microM MgATP with a Ki = 9.5 mM. The effects of ADP and Pi on velocity agree with fiber mechanical studies, confirming that the motility assay is an excellent system to investigate the molecular mechanisms of force generation and shortening in smooth muscle. The rate at which rigor cross-bridges can be recruited to move actin filaments was observed by initiating cross-bridge cycling from rigor by flash photolysis of caged MgATP. Following the flash, which results in a rapid increase in MgATP concentration, actin filaments experienced a MgATP-dependent delay prior to achieving steady state velocity. The delay at low MgATP concentrations was interpreted as evidence that motion generating cross-bridges are slowed by a load due to a transiently high percentage of rigor cross-bridges immediately following MgATP release.     

Steady-state kinetic studies on the actin activation of skeletal muscle heavy meromyosin subfragments: Effects of skeletal,smooth and non-muscle tropomyosins

The addition of either smooth muscle or brain tropomyosin to skeletal muscle actoheavy meromyosin (HMM) or acto-myosin subfragment-1 (SF1) produces an activation of the actin-activated ATPase activity up to 100%. This contrasts with the opposite, inhibitory effect produced by skeletal muscle tropomyosin. The degree of activation or inhibition depends on the ionic conditions, which influence the affinities of tropomyosin and HMM or SF1 for actin as well as on the molar ratio of actin to myosin.Enzyme kinetic analysis indicates that the inhibitory effect of skeletal muscle tropomyosin results from an approximately six- to tenfold increase in the apparent affinity (K_app) of the myosin head for the F-actin-tropomyosin complex with a concomitant six- to tenfold reduction in the maximal turnover rate (V_max). Thus, there is no direct competition of skeletal muscle tropomyosin and myosin for the same site on actin. Brain tropomyosin has an opposite effect, decreasing the apparent affinity with concomitant increase in the V_max.The effect of smooth muscle tropomyosin is more complex. At high ratios of myosin to actin this tropomyosin produces the same change in the K_app as skeletal muscle tropomyosin but yields a value of V_max that is about twofold higher. At lower molar ratios (below about 1 to 5 myosin subfragments to actin) the activating effect of this tropomyosin remains unchanged while the apparent affinity decreases to that observed for pure F-actin.On the basis of these data as well as from experiments carried out at fixed actin and varying SF1 concentrations, it is concluded that tropomyosins act in general as allosteric un-competitive inhibitors or activators of actomyosin by increasing or reducing the co-operative activation of myosin by actin at the level of product release.     

The apparently negatively cooperative phosphorylation of smooth muscle myosin at low ionic strength is related to its filamentous state

The correlation curve between phosphorylation and MgATPase activity suggests that the 20,000-dalton light chain of both heads of a smooth muscle myosin or heavy meromyosin (HMM) molecule must be phosphorylated before the MgATPase activity of either head can be activated by actin. The two heads of HMM appear to be phosphorylated randomly at equal rates, while those of myosin are phosphorylated in a negatively cooperative manner (Persechini, A., and Hartshorne, D.J. (1981) Science, 213, 1383-1385; Ikebe, M., Ogihara, S., and Tonomura, Y. (1982) J. Biochem. 91, 1809-1812). We have investigated the cause of this difference between HMM and myosin. We find that if myosin is first phosphorylated at high ionic strength (0.6 M KCl), where it is monomeric, and then assayed for MgATPase activity (in 0.05 M KCl), the data support a model where the two heads are phosphorylated randomly with equal rates (i.e. similarly to HMM). The correlation curves between MgATPase activity and dephosphorylation of fully phosphorylated myosin, both in a filamentous and monomeric state, are also best explained by a model where dephosphorylation of one head is sufficient to deactivate the entire molecule. With monomeric myosin, the dephosphorylation appears to occur randomly with equal rates, whereas with filamentous myosin the dephosphorylation appears to be negatively cooperative. The correlation between dephosphorylation of HMM and its MgATPase activity is more complex and is consistent with a positively cooperative dephosphorylation. Direct analyses of the time courses of phosphorylation of HMM and monomeric myosin show that a single exponential is sufficient to fit the data through greater than 90% of the reaction. However, when phosphorylation is carried out at low ionic strength (0.02 M KCl), where myosin is present as filaments, the time course consists of two exponential functions where the rate constant for the phosphorylation of one myosin head is 6-10 times greater than that for the other head which is located on the same molecule. This suggests that when myosin is polymerized into filaments the two previously indistinguishable heads either become nonequivalent or are subject to head-head interactions leading to a negatively cooperative phosphorylation reaction.