Dictyostelium Myosin-5b Is a Conditional Processive Motor*

Dictyostelium myosin-5b is the gene product of myoJ and one of two closely related myosin-5 isoenzymes produced in Dictyostelium discoideum. Here we report a detailed investigation of the kinetic and functional properties of the protein. In standard assay buffer conditions, Dictyostelium myosin-5b displays high actin affinity in the presence of ADP, fast ATP hydrolysis, and a high steady-state ATPase activity in the presence of actin that is rate limited by ADP release. These properties are typical for a processive motor that can move over long distances along actin filaments without dissociating. Our results show that a physiological decrease in the concentration of free Mg2+-ions leads to an increased rate of ADP release and shortening of the fraction of time the motor spends in the strong actin binding states. Consistently, the ability of the motor to efficiently translocate actin filaments at very low surface densities decreases with decreasing concentrations of free Mg2+-ions. In addition, we provide evidence that the observed changes in Dd myosin-5b motor activity are of physiological relevance and propose a mechanism by which this molecular motor can switch between processive and non-processive movement.

mechanoenzymatic properties, which characterize them either as processive or non-processive motors. Processive myosins, like vertebrate myosin-5a, are capable of taking successive steps along actin as single molecules before detaching (16). The overall movement has been described as a coordinated stepping process of both heads in a hand-over-hand mechanism that is driven by intramolecular strain (17,18). In contrast, non-processive myosins bind to the actin filament perform just one step and then dissociate rapidly. A notable difference between processive and non-processive myosins is displayed in the duty ratio, i.e. the fraction of the total ATPase cycle time a motor spends in the strong actin binding states. Processive myosins have a high duty ratio (Ͼ0.5), whereas non-processive myosins display a low duty ratio that is generally far below 0.5. Characteristic kinetic parameters contributing to a high duty ratio thus minimizing early detachment from actin include (i) a fast ATP hydrolysis rate, (ii) a high affinity for actin in the weak binding states, (iii) a high ADP affinity in the actin-bound states, (iv) a rate-limiting ADP dissociation rate, (v) an increased P i release rate, and (vi) a weak coupling between nucleotide and actin binding sites.
The relevance of a high duty ratio for processive movement has been shown by comparison of the kinetic properties of class 5 myosins from different subclasses and organisms. Accordingly, vertebrate myosin-5a is a high duty ratio motor that moves processively along actin filaments (19,20). Recently, a kinetic study of human myosin-5b revealed that this myosin is also characterized by a high duty ratio; however, direct observation of the predicted processivity has not been reported (21).
Not all class 5 myosins are high duty ratio motors. Homo sapiens myosin-5c (22,23), Drosophila melanogaster myosin-5 (24), and Saccharomyces cerevisiae myo2p and myo4p (25,26) display properties that are not compatible with those of a processive motor. It is assumed that these myosins need to function as ensembles for the efficient intracellular translocation of cargo. So far, there is limited information about the kinetic, structural, and mechanoenzymatic properties of class 5 myosins that belong to subclasses other than subclass 5a. Thus, it is difficult to define in detail the parameters and molecular mechanisms that distinguish processive class 5 myosins from nonprocessive ones. The members of the respective groups are assumed to use different ways to couple conformational changes at the nucleotide binding regions to changes that occur at the actin binding sites during the ATPase cycle.
This study provides a detailed kinetic and functional characterization of Dictyostelium discoideum myosin-5b (Dd myosin-5b), 2 a heavy chain dimer forming class 5 myosin, which previously has been referred to as MyoJ (27,28). We compare the results obtained for Dd myosin-5b with those previously reported for processive and non-processive members of the myosin-5 family (19,(21)(22)(23)(24). Our investigations reveal that under standard assay conditions the kinetic properties of Dd myosin-5b are similar to those of other processive myosins: ADP-release limiting the actomyosin ATPase cycle, a low degree of coupling between the nucleotide and actin binding sites, and a high duty ratio. We show that changes in the concentrations of free Mg 2ϩ -ions that lie in the physiological range modulate the ADP release kinetics of the motor and affect the duty ratio, which is a critical determinant for processivity. Our results show that this particular mechanism enables native Dd myosin-5b to switch between processive and non-processive motor activity in the context of the contractile vacuole.

EXPERIMENTAL PROCEDURES
Reagents-Standard chemicals, TRITC-phalloidin, and anti-His antibody were purchased from Sigma; restriction enzymes, polymerases, and DNA-modifying enzymes were purchased from MBI-Fermentas and Roche Applied Sciences.
Plasmid Construction-The oligonucleotides 5Ј-C GGA TCC ACC ACA TCA ACA ATT-3Ј and 5Ј-GT CTC GAG CAC TAC GAT CCA-3Ј were used to isolate a PCR fragment from Dictyostelium AX2 genomic DNA that encodes the 829 amino acids of the motor domain of Dictyostelium myosin-5b (28). The product was cloned into the expression vector pDXA-3H between restriction sites BamHI and XhoI (pDXA-J829) (29). The introduction of the extra XhoI site created mutation T829R in the protein. A motor domain construct fused to two D. discoideum ␣-actinin repeats (J829-2R) was obtained as the XhoI/SphI fragment from pM790-2R-eYFP (30) and inserted in the XhoI/SphI-digested pDXA-J829 motor domain expression plasmid. To produce full-length Dd myosin-5b fused to EYFP, base pairs 2692-6947 starting from the open reading frame of the Dd myosin-5b gene were amplified by PCR from genomic DNA and inserted into the pDXA-J829 expression vector using XhoI as a unique restriction site. This produced the plasmid pDXA-Dd myosin-5b encoding the full-length protein. It was digested with BamHI and SphI and the gene fragment was cloned into the vector pDXA-EYFP-MCS for the N-terminal fusion with EYFP (29). All plasmids were confirmed by sequencing.
Protein Production and Purification-Plasmids for the high level production of the Dd myosin motor domain constructs were transformed into AX3-Orf ϩ cells by electroporation as described earlier (31,32). The full-length EYFP-Dd myosin-5b plasmid was transformed for cell biological investigations in AX2 cells. Transformants were grown at 21°C in HL-5c medium and selected in the presence of 10 g/ml G418 and 100 units/ml penicillin/streptomycin. Screening for the production of the recombinant myosins and protein purification was performed as described (33). Rabbit skeletal muscle actin was purified as described by Lehrer and Kerwar (34) and pyrene-labeled actin was prepared as described by Criddle et al. (35).
Kinetic Measurements-ATPase activities were measured at 25°C with the NADH-coupled assay as described previously (36). Values for k cat and K app were calculated from fitting the data to the Michaelis-Menten equation. Transient kinetic experiments were performed at 20°C with either a Hi-tech Scientific SF-61 DX single mixing stopped-flow system or an Applied Photophysics PiStar 180 Instrument in MOPS buffer (25 mM MOPS, 100 mM KCl, 1 mM dithiothreitol, pH 7.0) supplemented with varying concentrations of MgCl 2 using procedures and kinetic models described previously (37). Free Mg 2ϩion concentrations were calculated using Maxchelator software as described (38). Kinetic parameters of nucleotide and actin interactions were analyzed in terms of the model shown in Scheme 1.
Direct Functional Assays-Actin-sliding motility was measured as described previously (30). The movement of more than 200 TRITC-phalloidin-labeled actin filaments was recorded for each individual concentration of free Mg 2ϩ -ions. Automated actin filament tracking was performed with the program DiaTrack 3.01 (Semasopht, Switzerland) and data analysis was performed with Origin 7.0 (Originlab, USA).
Landing assays were performed as described by Rock et al. (39) with the following modifications: Dd myosin-5b molecules were immobilized on nitrocellulose-coated coverslips via antipenta-His antibodies (concentration range 0.5 to 41 g/ml) to obtain surface densities between 50 and 4000 myosin molecules/m 2 . The assay was started by the addition of TRITCphalloidin-labeled actin (100 nM) to the motility buffer (described above) containing 1.5 mM Mg 2ϩ -ATP and varying concentrations of free Mg 2ϩ -ions. The landing events were recorded with an objective type TIRF microscope equipped with a 532 nm diode laser (150 milliwatts). An inverted micro-Scheme 1. Kinetic reaction scheme of the actomyosin ATPase cycle. A refers to actin, M to myosin, T to ATP, and D refers to ADP. Rate constants are referred to as k ϩn and k Ϫn is assigned to the corresponding forward and reverse reactions. An additional notation is used that distinguishes between the constants in the absence and presence of actin by italic type (k Ϫ1 , K 1 ) and bold (k ؊1 , K 1 ), respectively; subscript A refers to actin (K A ) and subscripts D (K D ) refers to ADP. scope was used (Olympus IX81) fitted with a 60 ϫ 1.49 NA oil immersion lens (ApoN, Olympus). The landing rate was measured by counting the number of actin filaments that landed and moved Ն0.5 m in an observation area of ϳ12,000 m 2 .
Cell Imaging-The cellular localization of Dd myosin-5b was assayed by confocal microscopy with an inverted Leica TCS SP2 AOBS microscope. Cells transfected with EYFP-Dd myosin-5b were seeded on glass bottom Petri dishes (coverslip thickness: 160 -180 m), washed twice with Bonner's salts solution (10 mM NaCl, 10 mM KCl, and 3 mM CaCl 2 ), and kept in this medium during image acquisition. Images were recorded at 21°C at one frame per 10-s interval with a 63 ϫ 1.4 NA immersion oil objective. The excitation wavelength was 514 nm; fluorescence emission was detected from 528 to 600 nm. Image processing was done with the Leica Confocal Software.
Experiments with the fluorescent dye KMG-104AM were performed according to Ref. 40 using the TIRF setup described above. Cells were incubated in 10-ml flasks with shaking at 180 rpm in the presence of 40 M KMG-104AM. KMG-104AM was dissolved in 50 mM HEPES, pH 7.3. After 2 h cells were seeded on Petri dishes and immediately before imaging the solution containing KMG-104AM was replaced by 50 mM HEPES, pH 7.3, containing 10 mM MgCl 2 .

RESULTS
All kinetic experiments were performed with a single-headed Dd myosin-5b construct (J829) comprising 829 amino acids of the motor domain. Nucleotide and actin interactions were analyzed according to Scheme 1.
Steady-state ATPase Activity of Dd Myosin-5b-The steadystate ATPase activity of Dd myosin-5b was measured in the absence and presence of actin in the range from 0 to 60 M actin. Dd myosin-5b displays a basal ATPase rate (k basal ) of 0.069 s Ϫ1 . The maximum actin-activated ATPase activity (k cat ) is 12.4 s Ϫ1 and comparable with the steady-state ATPase rates reported for other class 5 myosins. Half-maximal activation of the ATPase (K app ) is reached at 21 M F-actin and the apparent second-order rate constant for actin binding (k cat /K app ) is 0.59. The obtained steady-state parameters are summarized in Table  1, together with published values of human myosin-5b (H. sapiens myosin-5b), chicken myosin-5a (Gallus gallus myosin-5a), human myosin-5c (H. sapiens myosin-5c), and Drosophila myosin-5 (Dm myosin-5).
ATP Binding to Dd Myosin-5b and ATP-induced Dissociation of Acto⅐Dd Myosin-5b-ATP binding to the Dd myosin-5b motor domain was monitored from the increase in intrinsic protein fluorescence following the addition of ATP. Fluorescence transients were best fit to single exponentials at all ATP concentrations examined. In the range from 5 to 25 M ATP, the observed rate constants were linearly dependent upon ATP concentration. The apparent second-order rate constant obtained from the slope corresponds to K 1 k ϩ2 ϭ 0.47 M Ϫ1 s Ϫ1 . At higher ATP concentrations the observed rate constants k obs followed a hyperbolic dependence (Fig. 1A, filled circles). At saturating ATP concentrations, k max defines the maximum rate of the conformational change that corresponds to the rate of ATP hydrolysis in the absence of actin (k ϩ3 ϩ k Ϫ3 ). In the case of Dd myosin-5b this rate is Ͼ300 s Ϫ1 (Table 2).
ATP binding to acto⅐Dd myosin-5b was followed by observing the exponential increase in fluorescence of pyrene-actin as the actomyosin complex dissociates following the addition of excess ATP. The mechanism of ATP-induced fluorescence enhancement was modeled according to Scheme 2, which describes a two-step mechanism for ATP binding to actomyosin.
At lower ATP concentrations the observed rate constants increased linearly up to 50 M ATP giving a second-order rate binding constant K 1 k ؉2 of 0.19 Ϯ 0.01 M Ϫ1 s Ϫ1 (Fig.  1A, filled squares). At higher ATP concentrations the increase of the observed rate constants (k obs ) was best described by a hyperbola, approaching a maximum value k ؉2 of ϳ75 s Ϫ1 and an apparent equilibrium constant for ATP binding of K 1 Ͼ 400 M. ADP Binding to Dd Myosin-5b in the Presence and Absence of Actin-Because binding of ADP to the motor domain of Dd myosin-5b did not result in a change of the fluorescence signal, neither in the absence nor presence of F-actin measurements were performed using the fluorescent analogue mantADP. Binding of mantADP was determined by monitoring the increase in mant-fluorescence upon the addition of increasing concentrations of the fluorescent analogue to the Dd myosin-5b motor domain construct. In the range from 1 to 30 M mantADP, time courses of mantADP binding to Dd myosin-5b and acto⅐Dd myosin-5b followed single exponentials with rates that were linearly dependent on the concentration of nucleotide (Fig. 1B). The apparent second-order rate constants for ADP binding to Dd myosin-5b (k ϩD ) and acto⅐Dd myosin-5b (k ؉AD ) were determined from the slopes of the straight lines fitted to the data. The ratio of k ؉AD ϭ 4.0 M Ϫ1 s Ϫ1 and k ϩD ϭ 0.17 M Ϫ1 s Ϫ1 indicates that the rate of ADP binding is more than 20-fold increased in the presence of actin (Table 2).
Actin Binding Properties of Dd Myosin-5b-The rate of actin binding was measured by following the exponential decrease in pyrene fluorescence upon binding of excess pyrene-labeled actin to the Dd myosin-5b motor domain. The observed rate constants were linearly dependent upon F-actin concentration over the range studied (Fig. 1C, filled circles). The data were modeled as simple bimolecular reactions. The apparent second-order rate constant of pyrene-actin binding (k ؉A ) was obtained from the slope of the plot giving a value for k ؉A of 1.17 M Ϫ1 s Ϫ1 . The presence of 1 mM ADP did not significantly affect the second-order rate binding constant (k ؉DA ) of F-actin to Dd myosin-5b (Fig. 1C, filled squares).
Pyrene-actin dissociation from Dd myosin-5b was measured by competition with F-actin after mixing an equilibrated mixture of pyrene-acto⅐Dd myosin-5b with a 40-fold excess of unlabeled F-actin. In the absence and presence of ADP, the observed processes could be fit to single exponentials were k obs corresponds directly to k ؊A and k ؊DA , respectively. The rates of actin displacement in the absence and presence of 1 mM ADP (k ؊A ϭ 0.023 Ϯ 0.001 M Ϫ1 s Ϫ1 and k ؊DA ϭ 0.03 Ϯ 0.001 s Ϫ1 ) are very similar and indicate that the tight association of F-actin to Dd myosin-5b is not affected by the presence of excess amounts of ADP. Furthermore, the apparent acto⅐Dd myosin-5b affinities (K A and K DA ) were determined from the ratio of the rate constants for actin binding and dissociation. The parameters are summarized in Table 2.
ADP Dissociation from Dd Myosin-5b and Acto⅐Dd Myosin-5b-The rate of ADP dissociation was determined by monitoring the decrease in fluorescence upon displacement of mantADP from the myosin-mantADP and actomyosin-man-tADP complex by the addition of excess ADP. The observed processes could be fitted to single exponentials where k obs corresponds directly to the ADP dissociation rates k ϪD in the absence and k ؊AD in the presence of actin (Scheme 1). Man-tADP dissociation from Dd myosin-5b was ϳ25-fold increased by actin from k ϪD ϭ 0.92 s Ϫ1 to k ؊AD ϭ 21.6 s Ϫ1 (Table 2).
ADP Affinity of Dd Myosin-5b in the Absence and Presence of Actin-The affinity of ADP for Dd myosin-5b was determined by monitoring the reduction in the rate of pyrene-actin binding to the myosin motor domain as a function of ADP concentration. The decrease in pyrene fluorescence followed single exponentials at all ADP concentrations examined. The k obs values plotted against the ADP concentration are shown in Fig. 2A. High ADP concentrations decreased the observed rate of pyrene-actin binding ϳ2-fold. Fitting the data to a hyperbola gives an affinity constant of ADP for Dd myosin-5b (K D ) of 5.5 M. This value is consistent with the calculated affinity con- The affinity of ADP for the actomyosin complex (K AD ) was determined from the inhibition of the ATP-induced dissociation of acto⅐Dd myosin-5b by ADP. The observed rate of actin dissociation from Dd myosin-5b was reduced up to 8-fold when excess ATP was added to the actomyosin complex in the presence of varying concentrations of ADP. The dissociation reactions were monophasic and best described by single exponen-FIGURE 1. Transient kinetic analysis of the interaction of nucleotides with Dd myosin-5b and acto⅐Dd myosin-5b. A, ATP binding to Dd myosin-5b (F) and ATP-induced dissociation of acto⅐Dd myosin-5b (f). Hyperbolic dependence of the observed rate constant k obs on ATP concentrations in the range of 0.005 to 1 mM. At low ATP concentrations k obs were linearly dependent upon ATP concentration. The apparent second-order rate constant for ATP binding to myosin and actomyosin were determined from the slope of the straight lines. The rate constants for the ATP hydrolysis reaction k ϩ3 ϩ k Ϫ3 in the absence of actin and the isomerization step k ϩ2 , respectively, are given by the plateau values. B, mantADP binding to Dd myosin-5b (F) and acto⅐Dd myosin-5b (f) upon mixing 1 M Dd myosin-5b and 1 M acto⅐DdMyosin-5b, respectively, with increasing mantADP concentrations (0 -30 M). The observed rate constants were determined by fitting the time courses of the fluorescence change to single exponentials and plotted against the mantADP concentration. C, actin binding to Dd myosin-5b in the absence (F) and presence of ADP (f). The observed rate constants were plotted against the pyrene-actin concentration and second-order rate constants were obtained from linear fits to the data. All rate and equilibrium constants are summarized in Table 2.
Processive Dictyostelium Myosin-5b OCTOBER 3, 2008 • VOLUME 283 • NUMBER 40 tials. The determined rate constants were plotted against the ADP concentration and the data were fitted to a hyperbola (Fig.  2B) yielding a dissociation equilibrium constant (K AD ) of 8 M. At high ADP concentrations the dissociation rate constant of the acto⅐Dd myosin-5b complex by 2 mM ATP decreased to 17.4 Ϯ 1.9 s Ϫ1 . Because ADP release from the A⅐M⅐D complex limits the rate of the ATP-induced dissociation, the rate of 17.4 s Ϫ1 corresponds directly to the ADP dissociation rate from acto⅐myosin (k ϪAD ). F-actin has minimal effects on the affinity of ADP to Dd myosin-5b, although both binding and dissociation rates are affected by the presence of actin. In addition the association constant of actin for Dd myosin-5b in the presence of ADP (K DA ) was calculated as follows: The resulting affinity for actin in the presence of ADP (K DA ) is 56 nM and comparable with the value calculated from of K DA ϭ k ؊DA /k ؉DA ϭ 38.5 nM ( Table 2).
Effect of Free Mg 2ϩ Ions on ADP Binding Kinetics to Acto⅐Dd Myosin-5b-We examined the kinetics of ADP binding to acto⅐Dd myosin-5b by mixing nucleotide-free acto⅐Dd myosin-5b with increasing concentrations of mantADP in the presence of 0.2, 1, 3, and 5 mM free Mg 2ϩ -ions. The resulting fluorescence increase followed single exponential functions at all conditions. The observed rate constants increased linearly with increasing ADP concentration (Fig. 3A). The second-order rate constant (k ؉AD ) determined from the slope of the linear fit to the data ranged from 4.0 Ϯ 0.7 to 6.7 Ϯ 0.4 M Ϫ1 s Ϫ1 . In contrast, the ADP dissociation rates (k ؊AD ) as obtained from the y intercepts of the straight lines decreased with increasing con-centrations of free Mg 2ϩ -ions. The Mg 2ϩ -ion dependence of the ADP dissociation from the actomyosin complex was further confirmed by directly measuring the rate of ADP dissociation from acto⅐Dd myosin-5b using the fluorescent analogue man-tADP. In the presence of 0.2, 1, 3, 4.3, and 5 mM free Mg 2ϩ -ions, the time courses for the observed fluorescence change after mixing acto⅐Dd myosin-5b-mantADP with 1 mM ADP follow single exponentials (Fig. 3B, inset). The apparent rate constant for mantADP release (k ؊AD ) dropped from 187 Ϯ 25 s Ϫ1 at 0.2 mM free Mg 2ϩ to 17.4 Ϯ 1.9 s Ϫ1 at 5 mM Mg 2ϩ . The observed inverse hyperbolic dependence of the rate of mantADP on the free Mg 2ϩ concentration is described by Equation 1, where k ؊AD is the observed rate constant, k max is the rate constant at 0 M free Mg 2ϩ -ions, k min the rate constant at saturating free Mg 2ϩ -ion concentrations, and K i is the apparent inhibition constant for free Mg 2ϩ -ions, which is 0.44 Ϯ 0.1 mM (Fig. 3B).

Influence of Free Mg 2ϩ Ions on the Motile Activity of Dd
Myosin-5b-The motor activity of Dd myosin-5b was studied using the in vitro motility assay with a motor domain construct fused to an artificial lever-arm consisting of two ␣-actinin repeats. The ATP concentration was kept constant at 4 mM and the free Mg 2ϩ -ion concentration was varied in the range from 0.01 to 20 mM. A maximum sliding velocity of 1.25 Ϯ 0.12 m/s was obtained at concentrations of free Mg 2ϩ -ions between 10 and 50 M. Increasing concentrations of free Mg 2ϩ -ions that are within the physiological range (0.1 to 1 mM) reduce the motile activity of Dd myosin-5b in a sigmoidal dependence up to 3-fold with an apparent inhibition constant for Mg 2ϩ -ions (K i ) of 0.43 Ϯ 0.05 mM (Fig. 3C) that is well consistent with the K i obtained from Fig. 3B (0.44 Ϯ 0.1 mM).
To assay the effect of Mg 2ϩ -ions on the ability of Dd myosin-5b to bind and move actin filaments at very low surface densities, we performed landing assays at free Mg 2ϩ -ion concentrations corresponding to 5 and 0.28 mM. The number of landing events decreased as the surface density of myosin molecules decreased. At 5 mM free Mg 2ϩ -ions the landing rate was best fit to the equation: L() ϭ Z(1 Ϫ e Ϫ/0 ) n , according to the model by Hancock and Howard (41) with n ϭ 1.25 Ϯ 0.2 (Fig.  3D, solid circles) indicating that one Dd myosin-5b molecule is sufficient to bind and initiate movement. At 0.28 mM free Mg 2ϩ -ions the number of landing events was drastically reduced and the data were best fit with n ϭ 18 Ϯ 2 (Fig. 3D, open  circles). At Mg 2ϩ -ion concentrations Ͻ0.2 mM and myosin surface densities Ͻ3000 molecules/m 2 no landing events could be observed.
Influence of Free Mg 2ϩ -Ions on the Population of the Strong Actin Binding States and Duty Ratio of Dd Myosin-5b-The influence of free Mg 2ϩ -ions on the population of strong actin binding states was assayed by two different experimental approaches. First, the dissociation reaction of the acto⅐myosin-5b complex by excess ATP was monitored at different free Mg 2ϩ concentrations (Fig. 4A). A complex of 2 M Dd myosin-5b and 15 M pyrene-actin was dissociated by 50 M ATP at 5 and 0.2 mM free Mg 2ϩ -ion concentrations. The amplitudes were normalized and set in relation to the maximum change in fluorescence of Dd myosin-5b binding to pyrene-actin. An initial fast fluorescence decrease describing the dissociation of the complex is followed by a second, slower phase of re-association. At 5 mM free Mg 2ϩ concentration the dissociation rate constant equals 4.2 s Ϫ1 , which is ϳ3 times slower than the rate observed at 0.2 mM free Mg 2ϩ -ions (k obs ϭ 11.9 s Ϫ1 ). As this dissociation reaction describes the conformational transition of myosin from high to low actin affinity, the slow dissociation rate at 5 mM free Mg 2ϩ -ions indicates that Dd myosin-5b stays for a longer fraction of time strongly bound to actin than at 0.2 mM free Mg 2ϩ -ions.
To further evaluate whether Dd myosin-5b is a high duty ratio motor at high Mg 2ϩ concentrations and a low duty ratio motor at low Mg 2ϩ concentrations, we performed sequential mixing experiments. A pre-equilibrated mixture of 4 M myosin-5b and 20 M pyrene-actin was rapidly mixed with 400 M ATP, aged for 50 ms to allow ATP-binding, hydrolysis, and population of the weakly bound states (A⅐M⅐D⅐P i ), and then quenched with 2 mM ADP to prevent ATP binding to myosin (Fig. 4B). The final concentrations of pyrene-actin and myosin-5b after double mixing are 5 and 1 M, respectively. The observed fluorescence quench is assumed as the transition from the high fluorescence weak binding states to the low fluorescence strong binding states (Scheme 1). At high concentrations of free Mg 2ϩ -ions (5 mM, upper trace) the observed rate of the weak-to-strong transition is 2.0 Ϯ 0.3 s Ϫ1 and the amplitude is 0.68. At 0.2 mM free Mg 2ϩ -ions k obs equals 1.5 Ϯ 0.2 s Ϫ1 and the amplitude is decreased to 0.22 (Fig. 4B). Because both rate constants are similar to the steady-state ATPase rate (1.4 s Ϫ1 ) that was obtained at 5 M actin in the presence of 0.2 and 5 mM free Mg 2ϩ -ions, we conclude that ATP turnover is limited by the weak-to-strong transition in both cases. Thus, free Mg 2ϩ -ions have no influence on the actin activated steady-state ATPase rates. However, the differences in the amplitude of the pyreneactin fluorescence at 0.2 and 5 mM free Mg 2ϩ -ions indicate that 68 and 22% of the Dd myosin-5b molecules, respectively, are strongly bound to actin. These values provide direct estimates for the duty ratio (19).
Free Mg 2ϩ -Ion Distribution in Dictyostelium Cells-To further elucidate the physiological relevance of the observed functional changes upon variation of the free Mg 2ϩ -ion concentra-FIGURE 2. ADP affinity of Dd myosin-5b in the presence and absence of actin. A, inhibition of the rate of pyrene-actin binding to myosin head fragments (F). The ADP dissociation constant (K D ) of the Dd myosin-5b-ADP complexes were determined by fitting the plot k obs /k ؉A versus ADP concentrations with a hyperbola. B, affinity of ADP to acto⅐Dd myosin-5b (K AD ). K AD was determined from the ADP inhibition of the ATP-induced dissociation of the actomyosin complex (f). Monophasic dissociation reactions were observed as the complex was dissociated with 1 mM ATP in the absence and presence of ADP. The observed rate constants were plotted against the ADP concentration, and the data were fitted with a hyperbola. All rate and equilibrium constants are summarized in Table 2. tion, we followed the spatial and temporal changes in free Mg 2ϩ -ion distribution in Dictyostelium cells by TIRF microscopy using the Mg 2ϩ -sensitive dye KMG-104AM that was previously described as a potent marker for measuring Mg 2ϩ -ion concentrations within living cells (40). The series of pictures in Fig. 5B show extensive fluctuations of fluorescence in the millisecond to seconds range at tubulovesicular structures that encompass the entire cellular surface (see supplemental Movie 2).
Cellular Localization of Dd Myosin-5b-Dd myosin-5b is produced in Dictyostelium cells throughout development. Expression levels are high during the first 18 h followed by a marked reduction during later stages of the development cycle (28). We determined the cellular localization of Dd myosin-5b during vegetative growth and early development in cells producing full-length protein tagged at the N terminus with YFP by confocal microscopy. The images revealed that Dd myosin-5b concentrates at highly dynamic tubular and ring-like structures (Fig. 5A). This pattern is characteristic for the contractile vacuole system of the organism (43). The contractile vacuole system in Dictyostelium is a dynamic organelle with an osmoreg-ulatory function, collecting fluid in a network of cisternae and tubular structures and releasing it out of the cell through transient, plasma membrane-associated pores (44).

DISCUSSION
Dd myosin-5b is a dimeric motor that consists of two heavy chains, each with a molecular mass of 258 kDa and six associated light chains. Dd myosin-5b is assumed to localize at the contractile vacuole (28), a specialized organelle for the regulation of osmotic pressure (44). Recent phylogenetic analyses of more than 1700 myosins from various species places Dd myosin-5b and the closely related smaller isoenzyme Dd myosin-5a (previously referred to as Myo H) among the class 5 myosins (45,46). Notable kinetic features of Dd myosin-5b emphasizing its close relation to processive class 5 myosins are summarized in Tables 1 and  2 and include slow rates of ADP release (k ؊AD ), ATP binding to actomyosin (K 1 k ؉2 ), and ATP-induced actin dissociation (k ؉2 ). Additionally, ATP hydrolysis is fast (k ϩ3 ϩ k Ϫ3 ), ADP has only a minor effect on actin binding and dissociation (k ؉DA and k ؊DA ), and coupling between actin and nucleotide binding (K AD /K D ) is low. The simultaneous occurrence of these kinetic properties is indicative for Dd myosin-5b being a high duty ratio motor that populates predominantly the strong actin binding states A⅐M and A⅐M⅐ADP (Scheme 1). The high duty ratio derives primarily from the slow rate of ADP release, which becomes rate-limiting for steady state turnover in the presence of actin. A slow rate of ATPbinding to actomyosin and a fast rate of ATP hydrolysis (Ͼ300 s Ϫ1 ) contribute further to a predominant population of strong binding states. However, particular steps in the ATPase cycle of Dd myosin-5b do critically depend on the concentration of free Mg 2ϩ -ions. We have reported earlier that changes in the physiological concentration of free Mg 2ϩ -ions modulate the kinetic properties of Dd myosin-1E (38) and Dd myosin-1D (47). High free Mg 2ϩ concentrations reduce the motile activity of these myosin motors by inhibiting the rate of ADP dissociation from actomyosin. As a consequence, high concentrations of free Mg 2ϩ -ions stabilize the tension-bearing acto⅐myosin⅐ADP state and shift the actomyosin system from the production of rapid movement toward the generation of tension. In an independent study with vertebrate myosin-5a Sweeney and co-workers (48) reported that free Mg 2ϩ -ions have an influence on the rate and order of product release. In accordance to these studies, we The slope of each plot defines an apparent second-order rate constant (k ؉AD ) for ADP binding to myosin and the y intercept defines a dissociation rate constant (k ϪAD ). B, Mg 2ϩ dependence of mantADP dissociation from acto⅐Dd myosin-5b. The displacement of mantADP from the ternary mantADP⅐acto⅐Dd myosin-5b complex by excess unlabeled ADP at concentrations of free Mg 2ϩ -ions from 0.2 to 5 mM was monitored from the resulting decrease of mant-fluorescence (inset). The data traces represent the averages of at least eight individual transients. The transients follow single exponentials and the observed dissociation rate constants (k ؊AD ) decrease in an inverse hyperbolical manner with increasing free Mg 2ϩ concentrations. Excess free Mg 2ϩ concentrations lead to a more than 10-fold decrease in the mantADP dissociation rate constant. have observed that high free Mg 2ϩ -ion concentrations lead to a reduction in the coupling ratio of actin and ADP binding of Dd myosin-5b (K AD /K D Ϸ 1). Under these conditions Dd myo-sin-5b is able to bind both F-actin and ADP with high affinity. However, in the presence of low free Mg 2ϩ -ion concentrations, K AD is 5-fold increased (27.9 M), K D is almost unaffected, and the coupling ratio (K AD /K D ) is ϳ5-fold increased. Our in vitro motility experiments show that Mg 2ϩ -dependent alterations in this order of magnitude directly affect the motile activity of the motor. At ϳ50 M free Mg 2ϩ -ions Dd myosin-5b displays its highest sliding velocity, whereas free Mg 2ϩ -ion concentration Ͼ1 mM reduce the velocity up to 3-fold. This behavior is due to the reduced ADP release rate from actomyosin, which in turn results in an increased accumulation of the strong binding intermediate state A⅐M⅐ADP. From this experiment we conclude that high physiological concentrations of free Mg 2ϩ can extend the time the motor spends in strongly bound states up to 3-fold. We investigated this behavior by determining the duty ratio of the motor at low and high free Mg 2ϩ concentrations with the aid of the Equation 2, where T strong defines the time the motor spends in the strong actin binding states and T total the overall ATPase cycle time.
Using the parameters k ϪAD ϭ 21.6 s Ϫ1 , k ϩ2 ϭ 75 s Ϫ1 , and k cat ϭ 12.4 s Ϫ1 that were obtained at 5 mM free Mg 2ϩ -ions according to Tables 1 and 2, the calculated duty ratio of Dd myosin-5b equals 0.74. This value is in good agreement with the experimentally determined duty ratio of 0.68 that was obtained under the same conditions at 5 M F-actin concentrations (Fig. 4B).
Because the duty ratio depends on the actin concentration, it is expected that higher actin concentrations would lead to a further increase of the duty ratio, thus approaching the calculated value of 0.74, which is representative for high actin concentrations. In contrast, at low concentrations of free Mg 2ϩ -ions (0.2 mM), the rate of ADP release from acto⅐Dd myosin-5b increases ϳ10-fold (from 21.6 to 187 s Ϫ1 ), ATP binding becomes 3-fold faster (Fig. 4A), and the fraction of strongly bound states decreases more than 3-fold (Fig. 4B). The calculated and experimentally determined duty ratio is identical under these conditions (duty ratio ϭ 0.23). The modulation of velocity and duty ratio suggests that Dd myosin-5b can switch between working either as a stepping motor or as a motor that is adapted for tension bearing. The results from the landing assays confirm this hypothesis and indicate that Dd myosin-5b is a processive motor in the higher physiological range of free Mg 2ϩion concentrations and a non-processive in the lower range. We conclude that this kind of conditional adaptation of the mechanoenzymatic mechanism is important in cellular processes that require a cyclical switching between fast contractile activity and slow tension bearing. In agreement with this assumption, our results show that Dd myosin-5b localizes at the contractile vacuole (see Fig. 5 and supplemental Movie 1), where it appears to be actively involved in the fast contraction phase and the subsequent phase when expansion of the vacuole volume is suppressed. The concentration of free Mg 2ϩ -ions inside the cell varies in the range from 0.1 to 1 mM (42). The use of a new generation of Mg 2ϩ -sensitive fluorescent dyes allows the direct observation  of these changes with greater special and temporal resolution. Here, we have shown that similar changes as observed for HL60, HC11, and PC12 cells (49,50) do occur in Dictyostelium (see Fig. 5B and supplemental Movie 2). The observed fluctuations take place in the millisecond range and thus within a time domain where the consequences for myosin motor activity are of physiological relevance.