Supramolecular Regulation of the Actin-activated ATPase Activity of Filaments of Acanthamoeba Myosin 11*

Acanthamoeba myosin I1 has three phosphorylation sites clustered near the end of the tail of each of its two heavy chains (six phosphorylation sites/molecule). Myosin I1 has little or no actin-activated ATPase activity when four to six of these sites are phosphorylated. Maximal actin-activated ATPase activity is obtained when all six sites are dephosphorylated. Under assay conditions, both phosphorylated and dephosphorylated myosin I1 form bipolar filaments. Filaments of dephosphorylated myosin I1 have larger sedimentation coef- ficients than filaments of phosphorylated myosin I1 but this difference does not explain the difference in their actin-activated ATPase activities. Heteropolymers, formed by mixing soluble dephosphorylated and phos- phorylated myosins and then diluting the mixture into low ionic strength buffer containing MgC12, have sedi- mentation coefficients close to those of the homopolymer of phosphorylated myosin. The actin-activated ATPase activities of heteropolymers are, under most conditions, lower than the equivalent mixtures of ho- mopolymers of dephosphorylated and phosphorylated myosins. It is concluded, therefore, that the phospho- rylation of myosin tails regulates the actin-activated ATPase activity of Acanthamoebu myosin I1 by affect- ing the myosin filament

(4) located within a peptide region no larger than about 9,000 daltons located very near the end of the tail of the heavy chain (5). At least four of these six sites can be phosphorylated in uiuo (5)(6)(7). Myosin I1 with four to six sites phosphorylated has little or no actinactivated ATPase activity while the maximally dephosphorylated enzyme has maximal actin-activated ATPase activity (4, 6, 7 ) . Dephosphorylation increases the Vmax of the actomyosin I1 ATPase with little, if any, effect on the affinity of myosin I1 for F-actin (8).
* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertkement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
It is difficult to imagine that the state of phosphorylation of sites at the end of the tail of the heavy chain could affect the catalytic site in the head region, which is at least 100,000 daltons away in linear sequence (5), unless the molecule were bent back upon itself, as has been shown for monomers of dephosphorylated smooth muscle myosin under specific circumstances (9). However, the strong positive cooperativity of the actin-activated ATPase activity of Acanthamoeba myosin I1 as a function of Mg2+ and myosin concentrations (7) suggested that myosin-myosin interactions are required for enzymatic activity. More recently, we observed (8) that both phosphorylated and dephosphorylated myosin I1 were present in the enzyme assay solutions as bipolar filaments (8). A myosin molecule could not be curled back upon itself in a bipolar filament.
These observations led us to suppose that regulation of the actin-activated ATPase of myosin I1 by phosphorylation might be exerted through intermolecular interactions within the filament. To test this hypothesis, we compared the actinactivated ATPase activities of homopolymers, of mixtures of homopolymers, and of heteropolymers formed from phosphorylated and maximally dephosphorylated myosin 11. The data reported in this paper strongly indicate that regulation of the actin-activated ATPase activity of Acanthamoeba myosin I1 by phosphorylation does, in fact, occur through some generalized conformational effect on the filament as a whole.

MATERIALS AND METHODS
Rabbit skeletal muscle F-actin (10) was obtained from Dr. Lois E. Greene (National Heart, Lung, and Blood Institute) and treated with Dowex-1 to remove free ATP. Acantharnoeba myosin 11 was prepared by a slight modifiction of previous procedures (3,7). Acanthmoeba F-actin was added to the myosin I1 fraction obtained by DEAEchromatography of the amoeba extract in the absence of ATP (1) and the myosin was recovered from the actomyosin precipitate as described (7). The purified myosin I1 had an actin-activated ATPase activity of 0.27 pmol/min.mg, indicating a phosphate content of about 2 mol/mol of myosin (6). One portion of the myosin was incubated with Acanthamoeba myosin I1 heavy chain kinase and [y-32P]ATP for 1 b at 35 "C as described (4), incorporating 1.9 additional mol of phosphate/mol of myosin. Therefore, the phosphorylated myosin contained about 4 mol of phosphate/mol. Its actin-activated ATPase activity was only 0.02 pmol/min.mg consistent with this level of phosphorylation (6). Another portion of the isolated myosin was incubated with potato acid phosphatase for 2 h at 35 "C to remove the protein-bound phosphate (4,6). Its actin-activated ATPase activity was 0.57 pmol/min.mg as expected (6) for fully dephosphorylated myosin 11. The phosphorylated and dephosphorylated myosins were chromatographed on Bio-Gel A-1.5m (7) to remove the kinase and phosphatase, respectively, dialyzed for 24 h against 10 mM imidazole, pH 7.5, containing 10% sucrose (w/v) and 1 mM dithiothreitol, and concentrated by dialysis against solid sucrose. The final solutions, which contained 60% (w/w) sucrose by refractometry, were stored at 2 "C.
Actin-activated ATPase activity was measured as the difference in the rate of release of 3zPi from [Y-~~PIATP (8) in the presence and absence of F-actin at 30 "C in buffer containing 10 mM imidazole, pH 7.0, 1 mM ATP, 0.1 mM CaC12, and other components as indicated in the individual experiments. At myosin concentrations of 0.2-10 fig/ ml, 2.1 mg of F-actin/ml were used and the samples were incubated at 30 "C for 30 min. At myosin concentrations of 60-240 pg/ml, 3.9 mg of F-actin/ml were used and the incubations were for only 3 min. The rate of ATP hydrolysis was linear over the entire time courses and proportional to the myosin concentration. All the components of the reaction mixture were mixed at 0 "C with the myosin and F-actin added last, in that order. To make myosin copolymers, the samples 6011 This is an Open Access article under the CC BY license.

Regulation of the ATPase Activity
of Myosin Filaments of soluble dephosphorylated and phosphorylated myosin were mixed as concentrated solutions in 60% sucrose before dilution in the incubation buffers.
Sedimentation coefficients were obtained on myosin samples prepared exactly as for the enzyme assays. After preparing the samples at 0 "C, they were warmed to 30 "C in a water bath and transferred to cells warmed to 30 "C. Analyses were carried out in a Beckman Model E analytical ultracentrifuge equipped with UV optics. The rotor was spun at 30 'C at 16,000-20,000 rpm for samples containing filaments and at 34,000 rpm for samples containing myosin monomer. Scans were obtained at 3-to 12-min intervals for 1 to 2 h after reaching the desired speed. A wavelength of 290 mm was used because of the presence of ATP, and the samples were read against blanks prepared in the same way except that 60% sucrose replaced the myosin. Measurements were made from the midpoint of the boundary. At the end of the experiments, samples were analyzed by polyacrylamide gel electrophoresis to show that no degradation had occurred.

RESULTS
We first compared filament size and actin-activated AT-Pase activity of phosphorylated and dephosphorylated myosin I1 as a function of MgC1, concentration (Fig. 1). These experiments were carried out at a myosin concentration of 120 pg/ ml, which was near the minimal concentration at which reliable sedimentation coefficients could be obtained and near the maximal concentration a t which the enzymatic activity was proportional to the myosin concentration. Dephosphorylated myosin I1 was enzymatically inactive at 1 mM MgCl,, where it had a sedimentation coefficient of about 21 S , and fully active at 4 mM MgCl,, where it had a sedimentation coefficient of 31 S. With increasing concentrations of MgCl, up to 8 mM, the filaments of dephosphorylated myosin I1 increased in sedimentation coefficient to about 120 S with no increase in enzymatic activity. The slight decrease in ATPase activity observed at the higher concentrations of MgClz may be due to effects on the F-actin or on the myosin (8). In contrast, phosphorylated myosin I1 was enzymatically inactive over the entire range of MgCl, concentration although the filaments increased in sedimentation coefficient from 19 S in 1 mM MgC12 t o 57 S in 10 mM MgC12. Therefore, although dephosphorylated and phosphorylated myosin I1 form filaments of different sizes under identical conditions, this difference does not account for the difference in their actinactivated ATPase activities. We do not, however, know if the increase from 21 to 31 S for dephosphorylated myosin I1 was causally related to its increase in enzymatic activity between 1 and 4 mM MgC12. Monomeric myosin I1 in 0.6 M KCl, 5% sucrose was 4.8 S.
We next compared the enzymatic activities of 1:l copolymers of phosphorylated and dephosphorylated myosin I1 to the activities of homopolymers at several different concentrations of MgC12 and sucrose. In 4 mM MgCl, and 5 or 10% sucrose, dephosphorylated myosin 11, phosphorylated myosin 11, and their 1:1 copolymer each sedimented with a single symmetrical boundary. The sedimentation coefficients of the copolymers were much closer to the values for the homopolymers of phosphorylated myosin than to the homopolymer of dephosphorylated myosin (Table I , A and B). Under these conditions, the actin-activated ATPase activity of the 1:l copolymer, a t 120 pg/ml, was only about 44% of the sum of the activities of the homopolymers of dephosphorylated and phosphorylated myosins assayed separately, each a t 60 pg/ml (Table I , A and B). This is equivalent to a specific activity for the dephosphorylated myosin in the copolymer of about 43% of the specific activity of the dephosphorylated myosin homopolymer, if we assume that only the dephosphorylated myosin molecules were active in the copolymer. On the other hand, if we assume that both myosin species were equally active in the 1:1 copolymer, then the specific activity of the

Pase activities of dephosphorylated and phosphorylated
Acanthamoeba myosin I1 as a function of MgCla concentration. Samples were prepared and analyzed as described under "Materials and Methods." All samples contained 120 pg of myosin/ml and the indicated concentration of MgCl,. For the ATPase assays, Factin was added at 3.9 mg/ml. Because of the high concentration of myosin, the samples contained 5% sucrose derived from the myosin stock solutions which were 60% sucrose. At lower concentrations of myosin, this concentration of sucrose would be highly inhibitory (7).  Sedimentation coefficients have not been corrected to standard conditions of 20 "C in H20.
ND, not determined. Specific activity of the dephosphorylated molecules in the copolymer assuming the specific activity of the phosphorylated molecules are the same in the copolymer and the homopolymer.
Specific activity of the copolymer assuming phosphorylated and dephosphorylated molecules in the copolymer are equally active. myosin in the copolymer is 22% that of the dephosphorylated myosin homopolymer. Very similar results were obtained in an experiment in 7 mM MgCl, and 10% sucrose ( Table I, C).

The results of similar experiments carried out in 7 mM
MgCI, and 5% sucrose (Table I, D) suggest that it may be more correct to calculate the specific activities of the copolymers based on their total myosin concentrations. Under these incubation conditions, the 1:l copolymer was again much nearer in size to the phosphorylated homopolymer than to the dephosphorylated homopolymer, although all of the filaments were larger than in the previous conditions. However, actin-activated ATPase activity of the copolymer was greater in 7 mM MgC1, than in 4 mM MgCl, and, in fact, was equal to or higher than the sum of the activities of the homopolymers separately, 11% higher in the experiment reported in Table I, D. If the activity of the copolymer was due entirely to its content of dephosphorylated myosin, these molecules would have had an 18% higher specific activity in the copolymer than in their homopolymer, which is possible but quite unlikely. But, if we assume that all of the myosin molecules in the copolymer were equally active, they would have had only about 56% of the specific activity of the molecules in the homopolymer of dephosphorylated myosin.
We next compared the actin-activated ATPase activities of heteropolymers over a range of compositions to mixtures of homopolymers at the same ratios of dephosphorylated to phosphorylated myosin I1 and to the individual homopolymers at their concentrations in the heteropolymers (Fig. 2). The activity of the homopolymers of dephosphorylated myosin I1 was not affected by the presence in the incubation mixture of the homopolymers of phosphorylated myosin I1 but the heteropolymer was less active than the mixture of homopolymers  Fig. 2) or may reflect a true decrease in specific activity due to the low concentration of myosin. These solutions contained 0.5% sucrose.
at all ratios of phosphorylated myosin to dephosphorylated myosin. This establishes that the lower activities of the heteropolymers were not due to inhibitory substances in the solution of phosphorylated myosin 11. The enzymatic data for the homopolymers and heteropolymers in Fig. 2 are replotted as specific activities in Fig. 3.

DISCUSSION
Because myosin I1 is filamentous under assay conditions (7,8), the phosphorylation sites at the end of the tails are far removed from the catalytic site in the same polypeptide chain (4,5), and it, therefore, seemed unlikely that phosphorylation could regulate the actin-activated ATPase activity intramolecularly, i.e. that the ATPase activity of a myosin head was a function of the state of phosphorylation of its own tail.
Although phosphorylated and dephosphorylated myosin I1 do form different size filaments under a variety of conditions (Ref. 8 and Fig. l), the data in Fig. 1 show that a difference in filament size cannot explain their different enzymatic activities. The data in Table I and Figs. 2 and 3 establish that the regulation is, in fact, intermolecular, i.e. fully phosphorylated myosin molecules inhibit the ATPase activity of fully dephosphorylated myosin molecules when they are contained within the same filament. From the structure of the myosin I1 filaments (11) and the positions of the phosphorylation and catalytic sites within the myosin molecule (4, 5), however, it seems unlikely that the phosphorylated myosin molecules inhibit the activity of dephosphorylated myosin molecules by direct interaction. More probably, regulation by phosphorylation is exerted through some general conformational effect on the filament as a whole. In this case, it might well be that all the myosin heads in the heteropolymers have the same specific activity irrespective of whether they are attached to phosphorylated or to dephosphorylated tails. On this basis, the specific activity of each myosin molecule in the 1:l copolymer in Fig. 3 would be about