Functional Significance of the Conserved Residues in the Flexible Hinge Region of the Myosin Motor Domain*

Analysis of the three-dimensional crystal structure of the Dictyostelium myosin motor domain revealed that the myosin head is required to bend at residues Ile-455 and Gly-457 to produce the conformation changes observed in the ternary complexes that resemble the pre- and post-hydrolysis states (Fisher, A. J., Smith, C. A., Thoden, J. B., Smith, R., Sutoh, K., Holden, H. M., and Rayment, I. (1995) Biochemistry34, 8960–8972). Asp-454, Ile-455, and Gly-457 of smooth muscle myosin were substituted by Ala, Met, and Ala, respectively, and the mechano-enzymatic activities were determined to study the role of these residues in myosin motor function. Whereas the basal steady-state Mg2+-ATPase activity of D454A was higher than that of the wild type, the rate of the hydrolytic step is reduced ∼2,000-fold and becomes rate-limiting. M-ATP rather than M-ADP-P is the predominant steady-state intermediate, and the initial Pi burst and the ATP-induced enhancement of intrinsic tryptophan fluorescence are absent in D454A. D454A binds actin in the absence of ATP but is not dissociated from actin by ATP. Moreover, actin inhibits rather than activates the ATPase activity; consequently, D454A does not support actin translocating activity. I455M has normal actin-activated ATPase activity, Pi burst, and ATP-induced enhancement of intrinsic tryptophan fluorescence, suggesting that the enzymatic properties are normal. However, the actin translocating activity was completely inhibited. This suggests that the side chain at Ile-455 is critical for myosin motor activity but not for relatively normal enzymatic function, which indicates an apparent uncoupling between enzymatic activity and motile function. Although G457A has normal ATP-dependent actin dissociation, ATP hydrolytic step is reduced by ∼105-fold in the presence or absence of actin; consequently, G457A does not have actin translocating activity. These results indicate the importance of these conserved residues at the hinge region for normal myosin motor function.

Myosin is a motor protein that translocates actin filaments using energy from ATP hydrolysis. Myosin consists of a large family of actin-based motors (1)(2)(3)(4) that play a role in diverse biological contractile events such as muscle contraction, cytokinesis, cell locomotion, and organelle movements. Two key functions for the motor activity of myosin, ATP hydrolysis and actin binding, reside in the conserved N-terminal domain of myosin (4).
A key question is how ATP hydrolysis is coupled with the mechanical motor activity of myosin, which is thought to be universal among all myosin family members. A critical feature of the chemo-mechanical coupling of myosin motor function is the cyclic association and dissociation of the myosin molecule and the Factin molecule during the hydrolysis of ATP, which leads the movement of actin filaments in cells. The ATP hydrolysis by actomyosin can be explained by the following mechanism.
slow fast Key features of this mechanism are: 1) actomyosin is rapidly dissociated upon ATP binding to myosin, which is followed by a rapid hydrolysis of ATP, 2) the product release is rate-limiting and is significantly accelerated by actin binding, and 3) the transitions between the weakly bound states and the strongly bound states are associated with the cross-bridge movement (5)(6)(7). Although the structural changes associated with each step of the ATP hydrolysis cycle of myosin are not completely understood, recent three-dimensional structural analysis of the myosin motor domain/nucleotide complex has provided important information for understanding of the molecular mechanism of myosin motor function (8 -10). The myosin head contains several clefts, which divide the motor domain into distinct subdomains. The cleft that splits the 50-kDa central segment of myosin S1 extends from the nucleotide binding pocket to the actin binding interface, and it is proposed that this cleft closes after ATP hydrolysis (8,11). This opening and closing process is thought to be coupled with weak and strong binding states, respectively. Furthermore, it is suggested that the closure of the nucleotide binding pocket triggers a conformational change to generate a bent configuration (12), which is coupled with cross-bridge movement. ATP is bound in a narrow tunnel formed by two loops composed of amino acid residues, which are highly conserved in the myosin superfamily (4). There are a number of interactions between the triphosphate moiety and the amino acid residues of myosin, which are thought to be important for tight ATP binding, rapid ATP hydrolysis, and the stabilization of the myosin⅐ADP⅐P i metastable ternary complex.
Kinetic studies indicate that the release of P i occurs before the release of ADP from myosin⅐ADP⅐P i (13). The analysis of the three-dimensional crystal structure of myosin⅐ADP⅐ vanadate suggests that the phosphate cannot depart from the myosin ac-tive site via the same route as ATP enters, and it was proposed that the phosphate leaves the active site through the bottom of the active site pocket (14). The binding of actin was proposed to promote the movement of the P-loop to allow the phosphate to leave via the back door (14) in the 50-kDa fragment. In the presence of actin, the rate of product release is increased 100-fold for smooth muscle myosin S1 (15), and the rate of the associated hydrolysis step (AM⅐ATP 3 AM⅐ADP⅐P i ), which is rate-limiting at saturating actin is decreased 4 -5-fold by actin. It has been hypothesized that the movement of myosin relative to actin is achieved by a conformational change in myosin. A critical problem is the identification of such a conformational change coupled with actin sliding. Using small angle x-ray diffraction, Wakabayashi et al. (12) found that the myosin head forms a bent conformation after ATP hydrolysis, whereas it forms a rather extended conformation in the absence of ATP. This view was further elucidated at an atomic level by Fisher et al. (9) and Smith and Rayment (10) by analyzing the crystal structure of the three ternary complexes of myosin that are thought to be analogues of the pre-and post-hydrolytic myosin-substrate complexes. A comparison of the pre-and post-hydrolysis structures revealed that the myosin motor domain bends at Ile-455 and Gly-457 1 located at the ␤-strand, which lies between the upper part and the lower part of the central 50-kDa domain of the myosin head. This suggests that a flexible hinge region in the myosin motor domain has a critical role in the coupling of ATP hydrolysis to mechanical work.
In the present study, we have made three mutant myosins in which the residues in the flexible hinge region are altered, and we analyzed the mechano-enzymatic properties to better understand the mechanism of the coupling between ATP hydrolysis and mechanical work in myosin.

EXPERIMENTAL PROCEDURES
Materials-Restriction enzymes and modifying enzymes were purchased from New England Biolab (Beverly, MA). Smooth muscle myosin and myosin light chain kinase were prepared from turkey gizzards as described previously (Refs. 16 and 18, respectively). Actin was prepared from rabbit skeletal muscle according to Spudich and Watt (17). Recombinant calmodulin of Xenopus oocyte (19) was expressed in Escherichia coli (20).

Expression of the Recombinant Smooth Muscle Myosin Mutants-
The baculovirus transfer vectors containing the nucleotides encoding smooth muscle myosin heavy chain and light chains were produced as described previously (21,22). Recombinant baculoviruses for the heavy chain and the light chains were produced according to the protocols described by O'Reilly et al. (23). To express smooth muscle myosin mutants, Sf9 insect cells were co-infected with three separate viruses expressing the heavy chain and two light chains. The recombinant smooth muscle myosin was purified as described previously (22).
D454A was purified by a series of liquid chromatography steps because it does not dissociate from actin in the presence of ATP. The cell extract was applied to a hydroxyapatite column (1.3 ϫ 6 cm) equilibrated with Buffer A (10 mM Tris-HCl, pH 7.5, 0.2 M KCl, and 1 mM DTT 2 ). D454A was eluted at 0.25 M KP i by a linear KP i gradient (0.1-0.4 M). The fractions containing D454A were collected, dialyzed against Buffer A, and applied to a QAE-Sepharose column (1.3 ϫ 6 cm) equilibrated with the same buffer. D454A was eluted at 0.3 M KCl by a linear KCl gradient (0.2-0.6 M). The fractions containing D454A were dialyzed against Buffer B (15 mM MgCl 2 , 1 mM DTT, and 20 mM Tris-HCl, pH 7.5) to precipitate endogenous myosin. The sample was centrifuged at 10,000 ϫ g for 10 min, dialyzed against Buffer B, and stored on ice. The concentration of the myosin heavy chain was determined by the densitometric analysis of SDS-PAGE bands using NIH image software.
Actin Co-sedimentation Assays-Purified wild type and mutant HMM at concentrations of 0.1 mg/ml were incubated with rabbit skeletal actin (0.06 mg/ml) in a buffer consisting of 20 mM Tris-HCl, 50 mM KCl, 1 mM DTT, and 2 mM MgCl 2 , pH 7.5, at 0°C for 10 min and then centrifuged at 270,000 ϫ g for 10 min. The supernatant was removed, and the pellet was resuspended in 20 mM Tris-HCl, pH 7.5, 50 mM KCl, 2 mM MgCl 2 , 1 mM ATP, and 1 mM DTT. The suspension was then clarified by centrifugation at 270,000 ϫ g for 10 min. Again, the supernatant was removed, and the pellet was resuspended in SDS-gel sample buffer. Equal proportions of each supernatant and pellet fraction were then subjected to SDS-PAGE.
Single Turnover Measurement of ATP Hydrolysis-Chemical quench measurements were done using a computer-controlled stepper motordriven quench-flow apparatus used in pulse-quench mode as described previously (32).
Stopped-flow Measurements of Tryptophan Fluorescence Enhancement by ATP-Stopped-flow measurements of tryptophan fluorescence enhancement were done as described previously (25). Mutant HMM (1-2 M) was mixed with 200 M MgATP. The observed increases in tryptophan fluorescence were fit to one or two exponential terms by the method of moments (26). The rate of the hydrolysis step was estimated by the rate of the tryptophan fluorescence increase at saturating ATP concentration (27).
Photoaffinity Labeling of Myosin Mutants with ATP-Photoaffinity labeling was performed as described by Maruta and Korn (28), with some modification. Myosin mutants (0.1 mg/ml) were mixed with 40 M ATP (9 TBq/mmol) in 100 l of 30 mM Tris-HCl, pH 7.5, 2 mM MgCl 2 , 1 mM DTT, and 5% sucrose. The solution was irradiated at a distance of approximately 4 cm for 30 min with ultraviolet light (36 watts) at 254 nm. The protein was precipitated by the addition of cold 5% trichloroacetic acid containing 1% sodium pyrophosphate and collected by centrifugation. The pellets were washed once with the solution and dissolved in 50 l of SDS-loading buffer. SDS-PAGE was followed by autoradiography.
In Vitro Motility Assay-The in vitro motility assay was performed as described previously (22). After thiophosphorylation, myosin was attached to a coverslip using monoclonal antibody MM9, which recognizes the S2 portion (Ala-873-Ser-944) of chicken gizzard smooth myosin (29). Actin filament translocating velocity was calculated from the movement distance and the elapsed time in successive snapshots. Student's t test was used for statistical comparison of mean values. A value of p Ͻ 0.01 was considered to be significant.
Steady-state ATPase Activity Assay-Actin-activated ATPase activity was measured at 25°C in the assay mixture containing 0.05 mg/ml HMM, 0.2 mM CaCl 2 , 15 g/ml MLCK and 10 g/ml calmodulin, 1 mM MgCl 2 , 50 mM KCl, 0.2 mM ATP, and 30 mM Tris-HCl, pH 7.5, with various concentrations of F-actin. For dephosphorylated HMMs, 1 mM EGTA was added without MLCK and calmodulin. ATPase activity of HMM or acto-HMM was determined by measuring the liberated 32 P as described previously (16).

Expression and Purification of Smooth Muscle Myosin
Mutants-Three residues, Asp-454, Ile-455, and Gly-457, in the switch II loop of the smooth muscle myosin heavy chain were mutated. Smooth muscle myosin heavy chain cDNA that encoded Met-1-Ser-1110 was used as a template for site-directed mutagenesis because the expressed truncated heavy chain containing Met-1-Ser-1110 forms a stable double-headed structure and is thus regulated by regulatory light chain phosphorylation (21,22). Smooth muscle myosin heavy chain mutants were coexpressed in Sf9 cells with two light chains. The expressed truncated myosin mutants were well recovered in the extract, but a portion of the expressed heavy chain was insoluble ( Fig.  1). Repeated extraction did not increase the soluble myosin fraction, suggesting that the insoluble myosin was improperly folded. Sf9 cells expressed endogenous myosin, with a 200-kDa heavy chain, and most endogenous myosin was precipitated by dialysis against low ionic strength buffer (15 mM MgCl 2 , 1 mM DTT, and 20 mM Tris-HCl, pH 7.5) and removed by centrifugation as shown in Fig. 1.
ATPase Activities of Mutant Myosins-The ATPase activities of the mutant myosins are summarized in Table I. D454A showed no actin-activated ATPase activity, although significant basal Mg 2ϩ -ATPase activity was observed, showing that D454A binds and hydrolyzes ATP in its active site. Whereas the Ca 2ϩ -ATPase activity of D454A was similar to that of the wild type, K ϩ -EDTA ATPase was completely abolished. However, F-actin significantly inhibited the Mg 2ϩ -ATPase activity of D454A at 0.2 mM ATP ( Fig. 2A). Similar inhibition of the ATPase activity by actin was also observed for phosphorylated D454A. To discover whether or not this is due to the decrease in the affinity for ATP, the effect of actin binding upon ATPase activity was measured as a function of ATP (Fig. 2C). It was found that the actin binding significantly reduced the affinity of D454A for ATP. However, the ATPase activity was inhibited by actin even at saturating concentrations of ATP. The results indicate that actin binding decreases the affinity of D454A for ATP, and that inhibits either ATP hydrolysis step or product release step. In contrast, I455M showed actin-activated ATPase activity with a V max of 40% of WT, which was regulated by light chain phosphorylation. Consistent with this result, I455M showed near normal Ca 2ϩ -ATPase and K ϩ -EDTA ATPase activities (Table I). For both mutants, K Mg 2ϩ -ATP was similar to that of the wild type (data not shown), suggesting that the affinity for ATP is not markedly altered by these mutations. Very low ATPase activities were detected for G457A using a single turnover experiment, suggesting that the hydrolysis step is reduced at least 10 5 -fold (Table II).
Binding of the Mutant Myosins to F-actin-The binding of mutant myosins to F-actin was monitored by co-sedimentation analysis. All myosin mutants examined in this study co-sedimented with F-actin in the absence of ATP, indicating that myosin mutants can form a rigor complex with actin (Fig. 3). In the presence of ATP, WT HMM binds very weakly to actin and is almost completely dissociated at physiological ionic strength. As shown in Fig. 3, I455M, like WT HMM, is dissociated from F-actin by ATP. However, D454A co-sedimented with F-actin in the presence of ATP, suggesting that D454A-HMM-ATP remains tightly bound. G457A was dissociated from actin by ATP, as measured by co-sedimentation. However, G457A does not decrease the fluorescence emission of pyrene-actin, which is normally observed upon the binding of myosin to actin in the absence of ATP (data not shown). This suggests that the apparent rigor binding observed by co-sedimentation may not be normal.
Photoaffinity Labeling of Myosin Mutants with 32 P-labeled ATP-UV irradiation induces photolabeling of myosin by ATP in the active site (28). Myosins were irradiated with [␣-32 P]ATP or [␥-32 P]ATP, and the incorporation of 32 P into myosin heavy chain was analyzed by autoradiography. For WT HMM, radioactivity is detected if myosin heavy chain is irradiated with [␣-32 P]ATP, but not with [␥-32 P]ATP (Fig. 4). WT HMM rapidly hydrolyzes ATP to form a stable steady-state mixture of myosin⅐ATP N myosin⅐ADP⅐P i . Acid quench experiments indicate that the noncovalently associated P i dissociates from myosin heavy chain upon acid denaturation of myosin. Thus, [␥-32 P]ATP cross-linked to the active site would be expected to lose ␥-32 P after hydrolysis. This is consistent with the notion that photoaffinity labeling occurs via the adenine moiety of the nucleotide and that the cross-linked enzyme is able to hydrolyze the covalently bound ATP.
Whereas D454A has an above normal basal Mg 2ϩ -ATPase activity, this mutant does not have significant actin-activated ATPase activity, nor is it dissociated from F-actin by ATP. This mutant was labeled by [␣-32 P]ATP and at a much reduced level by [␥-32 P]ATP (Fig. 4). Because there is not a rapid hydrolysis of ATP to form M⅐ADP⅐P i (Fig. 5) The liberated P i was determined as described previously (16). Actin-activated ATPase activity was measured at 25°C in 0.1 mg/ml myosin mutant, 0.3 mM ATP, 30 mM KCl, 2 mM MgCl 2 , 30 mM Tris-HCl, and various concentrations of F-actin. To measure the activity of phosphorylated myosin, 0.2 mM CaCl 2 , 15 g/ml myosin light chain kinase, and 10 g/ml calmodulin were added, whereas 1 mM EGTA was added for the dephosphorylated one. A computed nonlinear least squares curve fitting program was used to estimate the maximum actin-activated ATPase activity (V max ) and the apparent dissociation constant for actin ( (Fig. 2). G457A hydrolyzes ATP at a rate that is at least 10 5 slower than that of the WT (Table I) and is labeled by [␥-32 P]ATP to a significantly greater extent than the wild type (Fig. 4). This result suggests that the rate of hydrolysis of the ␤-␥ phosphoester bond is decreased and that the major intermediate is myosin⅐ATP. I455M was radiolabeled by [␣-32 P]ATP but not by [␥-32 P]ATP, which is consistent with its normal WT ATP hydrolysis mechanism. P i Burst of the Mutant Myosins-Upon the binding of ATP, myosin rapidly hydrolyzes ATP to form a metastable myosin⅐ADP⅐P ternary complex (P i burst), which is followed by slow product release. Formation of the myosin⅐ADP⅐P i complex is necessary for normal myosin motor function (15,30). Therefore, it is critical to test whether or not the switch II mutants exhibit P i burst. D454A and G457A have no initial P i burst, which indicates that these mutants do not bind rapidly and hydrolyze ATP to form myosin⅐ADP⅐P i (Fig. 5, B and D). For both D454A and G457A, the rates of hydrolysis of single turnovers of ATP hydrolysis are similar to the steady-state rate, and the hydrolytic step is therefore at least partially ratelimiting. The rate of the hydrolysis step is reduced ϳ2,000-fold for D454A and ϳ10 5 for G457A. This was consistent with the UV cross-linking data described above in which cross-linking to [␥-32 P]ATP could only be observed for the very slowly hydrolyzing G457A. On the other hand, I455M showed rapid initial hydrolysis. The data were fitted to two exponential rate processes (f burst and f steady state ). The equilibrium constant (K H ) of the enzyme bound hydrolysis step was calculated based on the ratio of the amplitudes of the two components, and the value is shown in Table II. The second order rate constant of ATP binding (k burst /sites) was estimated from the observed rate of the phosphate burst (Table II). The value for I455M was similar to that for the WT, indicating that the kinetics of ATP binding and hydrolysis are nearly normal (Fig. 5C) compared with the WT (Fig. 5A) (Table II).
Actin Translocating Activity of the Mutant Myosins-The motor activities of the three myosin mutants were determined directly by measuring the actin translocating velocity using an in vitro motility assay. Consistent with the fact that D454A showed neither actin-activated ATPase activity nor ATP-dependent dissociation from actin, there was no actin translocating activity for D454A. As expected, G457A, which has low ATP hydrolysis activity, also does not support actin translocation. Of particular interest is I455M, which showed a nearly normal actin-activated ATPase activity, ATP-dependent actin binding activity, and P i burst but did not show actin translocating activity.
Change in the Intrinsic Tryptophan Fluorescence upon ATP Binding-Upon ATP binding to the active site of myosin, the intrinsic fluorescence of tryptophan residues increases. The enhancement is due to an increased fluorescence of M⅐ATP and an additional enhancement that occurs upon hydrolysis to M⅐ADP⅐P i . At a saturating ATP concentration (Ͼ100 M), the rate of the fluorescence enhancement is equal to the rate of the bond splitting step (31), and it is assumed that the fluorescence change reflects the conformational change of the myosin motor domain critical for cross-bridge movement. The tryptophan fluorescence enhancement observed upon mixing 200 M ATP with I455M is shown in Fig. 6. The kinetics of the fluorescence increase are fit by two exponential values (33 s Ϫ1 and 8 s Ϫ1 ) that probably correspond to rapid ATP binding, which is followed by slower hydrolysis. This interpretation of the data would indicate that the hydrolysis step is 2-3-fold slower than that for WT HMM but is still much faster than the rate-limiting step in the presence or absence of actin. The I455M mutant is of particular interest because it has nearly normal enzymatic activities (i.e. steady-state ATPase activity, P i burst, and ATP-dependent dissociation from actin) and tryptophan fluorescence enhancement but does not support actin translocation activity. A previous mutation, such as S236A, that produced a similar small change in kinetic constants had nearly normal motility activity (32). In contrast, D454A (data not shown) and G457A (Fig. 6) do not have the tryptophan fluorescence enhancement associated with normal ATP binding and hydrolysis. In these two mutations, the inhibition of the hydrolysis step results in altered enzymatic and motility function.

DISCUSSION
There are two conserved sequences present in many nucleotide-binding proteins, as originally found by Walker et al. (33). One is the P-loop that forms the triphosphate binding site, and the other corresponds to residues 454 -457 in the myosin motor domain. The corresponding region in G-protein is called switch II because this region changes conformation upon GTP hydrolysis (34,35). A comparison of the structure of the myosin⅐ADP⅐BeFx complex with myosin or the myosin⅐ ADP⅐AlF4 complex indicates that a partial closure of the narrow cleft in the central 50-kDa segment upon the hydrolysis of ATP results in the movement of the lower domain of the central 50-kDa segment (9,10). This requires flexibility in the peptide backbone of residues 454 -457 in the switch II sequence. Another conserved residue in switch II, Asp-454, forms a hydrogen bond to a water molecule, which is coordinated to the Mg 2ϩ that is also coordinated by the ␤-␥-phosphate of ATP and may play a key role in the mechanoenzymatic activity of myosin. In the present study, three smooth muscle myosin mutants with alterations in three residues in the switch II loop that might be critical for myosin motor function are expressed in Sf9 cells, and their mechanoenzymatic properties are characterized.
The substitution of Asp-454 with Ala abolishes the hydrogen bonding of the side chain of residue 454 to the Mg 2ϩ ion coordinated with the triphosphate moiety of ATP. This mutation severely disrupted normal myosin enzymatic and motor function. The results can be summarized as follows: D454A can form a rigor complex with actin, but ATP does not dissociate D454A from actin. Although the structure of the actin binding interface is preserved, the ATP-induced conformational change required to reduce actin affinity is disrupted. This is either because the affinity of ATP is reduced or because the normal coupling between the nucleotide and actin binding sites is severely disrupted. Although K ATP for D454A is increased in the presence of actin, the ATPase activity is concentrationindependent above 0.1 mM ATP. Because a high concentration of ATP (up to 10 mM) failed to dissociate D454A from actin, the latter possibility is more likely. Although D454A shows significant basal Mg 2ϩ -ATPase activity, it has neither actin-activated ATPase activity nor an initial P i burst.
Based upon the three-dimensional structure of the myosin motor domain (10), there is a water molecule that forms a hydrogen bond with the ␥-phosphate of ATP in the active site  a Parameters from quench-flow data in Fig. 5. b K H ϭ k ϩh /k Ϫh ϭ f burst /f ss . c Parameters from fluorescence data in Fig. 6. ( Fig. 7). This water molecule is expected to nucleophilically attack the ␥-phosphate of ATP for hydrolysis. Therefore, it is plausible that the disruption of the Asp-454-water-Mg 2ϩ interaction results in the displacement of ␥-phosphate from the proper position in the active site at which it is normally attacked by a water molecule.
Consistent with these enzymological results, the substitution of Asp-454 by Ala completely abolished the actin translocating activity. Asp-454 is completely conserved among the various myosins sequenced thus far, including myosin III, the most divergent member of myosin superfamily (4). This is reasonable because the substitution of this residue would be expected to cause severe disruption of myosin motor function.
In contrast, the mutation of Ile-455 to Met did not significantly alter the enzymatic activities of myosin, i.e. actin-activated ATPase activity, tryptophan fluorescence enhancement, P i burst, and ATP-induced rapid dissociation from actin. Nevertheless, the actin translocating activity was also completely abolished in I455M. Because normal P i burst and the actinactivated ATPase activity were retained in this mutant, the f is the fraction of the phosphate hydrolyzed in the initial rapid phase (the phosphate burst). k burst and k ss are the rate constants of the burst and of the slow (steady-state) phase of hydrolysis. The values used to calculate the theoretical curves are listed in Table II. myosin⅐ADP⅐P i metastable ternary complex is produced with nearly normal kinetics, and the change in the intrinsic tryptophan fluorescence of myosin upon ATP binding was also nearly normal. Previously, it was suggested by using energy transfer results (36) and by three-dimensional structural analysis (10) that the tryptophan residues responsible for the change in fluorescence induced by ATP binding to myosin are Trp-501 and either Trp-113 or Trp-131. Trp-501 is surrounded by the hydrophobic residues Tyr-494, Phe-503, Phe-692, and Gly-691 in the hydrophobic environment of the myosin⅐ADP⅐Vi complex. In nucleotide-free myosin, one edge of Trp-501 is exposed to solvent. Although the mechanism for tryptophan fluorescence enhancement is not completely understood, the present results suggest that the segment of random coil containing Trp-501 experienced similar environmental changes in I455M and in the wild type. According to the three-dimensional structural analysis, it was suggested that the myosin motor domain bends at Gly-457 and Ile-455 upon ATP hydrolysis (9). Therefore, it is plausible that the substitution of Ile-455 by Met disrupts the proper change in the torsional angle at residue 455 that is necessary for actin sliding upon product release. There are two possible explanations for the apparently normal ATP hydrolysis mechanism observed for I455M and its lack of motility: 1) the actual mechanism for the loss of motility in I455M could be due to a defect in the enzymatic pathway that has not yet been identified, and 2) ATP hydrolysis and motility are uncoupled by this mutation. Thus far, we are unable to distinguish between the two possibilities.
Ile-455 is well conserved among various myosins in different classes (I, II, IV, V,VI, VII, VIII, IX, X, XIII, and XV) in the superfamily (consensus sequence of DIXG), except for myosin III, in which Ile is substituted by Met. This suggests that myosin III might not function properly as a motor protein.
Interestingly, Ile in this motif is not conserved in other nucleotide-binding protein families, such as the G-proteins, in which DXXG has been found to be a conserved motif. In a G-protein, this region (switch II) also changes in conformation upon GTP hydrolysis, but in a different manner from that of myosin.
The mutation of another amino acid in the flexible loop, G457A, severely inhibits the mechano-enzymatic function of myosin. The substitution of Gly-457 by Ala abolished the actin-activated ATPase activity, P i burst, and ATP-induced intrinsic tryptophan fluorescence change, and, as a result, the actin translocating activity. The results are consistent with a previous report (37), which showed that the substitution of Gly-457 by Ala abolished the P i burst and markedly decreased the steady-state ATPase activity. The observation that G457A was photoaffinitylabeled with [␥-32 P]ATP indicates that G457A binds ATP and hydrolyzes it very slowly. As expected, the ATP hydrolysis rate was markedly reduced with this mutant (Table II). These results suggest that the inhibition of the motility in G457A is a result of the inhibition of the bond splitting step. According to the threedimensional structure of the myosin active site, the carbonyl oxygen of Gly-457 forms a short hydrogen bond with the proton of a water molecule, which in turn forms a hydrogen bond with a ␥-phosphate oxygen and is likely to be the water involved in catalysis. The amide of Gly-457 forms a hydrogen bond with a second phosphate oxygen. It is therefore likely that the position of Gly-457 is critical for catalysis. The region behind Gly-457 is very tightly packed, and the addition of the methyl group in G457A would clash with the conserved residues Asn-475 and either Gly-180 or Thr-179. It is therefore understandable that a mutation in this position has severe affects on the enzymatic and motile properties of the myosin. FIG. 7. Three-dimensional structure of the switch II region of the myosin⅐MgADP⅐Vi complex. This shows a ribbon representation of the switch II region of myosin along with the bound ADP/Vi. The figure also shows the interactions of the conserved amino acid residues in the loops with the triphosphate moiety of ATP. The N-terminal, central, and C-terminal segments of the heavy chain are colored in green, red, and blue, respectively. ADP, Vi, and Mg 2ϩ are colored in magenta, gray, and yellow, respectively.