Single-myosin crossbridge interactions with actin filaments regulated by troponin-tropomyosin

Neil M. Kad*, Scott Kim*, David M. Warshaw*, Peter VanBuren, and Josh E. Baker*
Departments of *Molecular Physiology and Biophysics and Medicine, University of Vermont, Burlington, VT 05405 Edited by James A. Spudich, Stanford University School of Medicine, Stanford, CA, and approved October 1, 2005 (received for review July 28, 2005)

Striated muscle contraction is governed by the thin lament regulatory proteins troponin and tropomyosin. Here, we investigate the molecular mechanisms by which troponintropomyosin inhibits myosins interactions with the thin lament in the absence of calcium by using a laser trap. The displacement events for a single-myosin molecule interacting with a reconstituted thin lament were shorter (step size 5 nm) and prolonged (69 ms) compared with actin alone (11 nm and 26 ms, respectively). However, these changes alone do not account for the degree of inhibition of thin lament movement observed in an ensemble assay. Our investigations of single- and multiple-myosin molecules with regulated thin laments suggest the primary basis for this inhibition derives from a 100-fold decrease in the probability of myosin attaching to actin. At higher myosin concentrations, short bursts of motility are observed in a laser trap consistent with the strong binding of a single-myosin crossbridge, resulting in cooperative binding of other cycling crossbridges. We conrmed this cooperativity in the in vitro motility assay by observing thin lament translocation in the absence of calcium but at low [ATP], consistent with rigor activation. We have developed a simple mechanistic model that reproduces and provides insight into both the observed single-myosin molecule and ensemble data in the absence of Ca2 . These data support the hypothesis that thin lament inhibition in the absence of Ca2 is largely achieved by modulating the rate of attachment and or transition from the weakly to strongly bound state.
muscle regulation single molecule thin lament laser trap

result of tens to hundreds of myosin molecules interacting with the thin filament. In contrast, with a laser trap assay (1517) one can measure the interaction of a single-myosin molecule with a regulated thin filament to specifically assess how TnTm inhibits myosin binding in the absence of Ca2 and how myosin strong binding leads to cooperative activation of the thin filament. Here, we report that in the absence of Ca2 the apparent rate of myosin strong binding to a reconstituted thin filament is reduced 100-fold with a more modest effect ( 2-fold) on myosins step size and strong binding duration. Interestingly, laser trap studies indicate strong binding of a single-myosin molecule can, even in the absence of Ca2 , cooperatively accelerate the binding of neighboring myosin molecules. With the in vitro motility assay, we confirmed this cooperative behavior; thin filament motility was fully activated in the absence of Ca2 by reducing the MgATP concentration analogous to previous solution studies demonstrating rigor activation of actinmyosin ATPase activity (18). In agreement with several models of thin filament regulation (3, 5, 1923), our data indicate that thin filament regulation in the absence of Ca2 is largely achieved by modulating the rate of attachment and or transition from the weakly to strongly bound state. Materials and Methods
Proteins. Skeletal muscle myosin was prepared from chicken pec-

orce generation in striated muscle results from myosin cyclically interacting with actin, a Ca2 -regulated process mediated by the actin-associated regulatory proteins, troponin (Tn) and tropomyosin (Tm). Early studies suggested that the Ca2 dependent movement of TnTm on actin functioned as an on-off switch, regulating myosin binding to actin (1, 2). However, subsequent studies indicated that myosins interaction with the thin filament is graded, leading to the proposal that Tm equilibrates among three states (3, 4): blocked, closed, and open (5). These biochemical data were supported by more recent structural data (6). In the absence of Ca2 , Tm sterically prevents myosin from binding to actin by occupying the blocked state (7). Upon Ca2 binding to Tn, Tms equilibrium position shifts toward the closed state, exposing sites that allow myosin weak binding while still inhibiting isomerization to the strong binding state (6). Once bound, myosins weak-to-strong binding transition shifts Tms equilibrium position further toward the open state, permitting cooperative binding of additional myosins by exposing neighboring actin binding sites (8, 9). Thus, myosin strong binding to actin is required to fully activate the thin filament (10). Characterization of striated muscle thin filament regulation has relied extensively on the collective behavior of many myosin molecules, e.g., skinned muscle fibers and solution studies. By reconstituting fully regulated thin filaments from isolated proteins and using the in vitro motility assay, Ca2 dependent modulation of thin filament motility can be investigated at the molecular level (1114). In this simple system, motility is still the
16990 16995 PNAS November 22, 2005 vol. 102 no. 47

toralis and stored in glycerol at 20C as described (24). Before use, myosin was further purified to eliminate denatured myosin by centrifugation with equimolar actin and 1 mM MgATP in myosin buffer (see Buffers below) (25). N-ethylmaleimide-modified skeletal myosin was prepared as described (24) and was used to bind actin or reconstituted thin filaments to the polystyrene beads in a laser trap assay (17). Actin was isolated from chicken pectoralis muscle (26). Tn and Tm were isolated from bovine cardiac muscle as described (27). Thin filaments (i.e., actin, Tn, and Tm) were reconstituted as reported (11) and diluted into actin buffer (see below) with excess (100 nM) Tn and Tm. Under these conditions thin filaments remained fully regulated for hours as demonstrated by the lack of movement in an in vitro motility assay in the presence of ATP and absence of Ca2 [ log 10 Ca concentration (pCa) 8, data not shown]. Actin and thin filaments were labeled with equimolar tetramethyl-rhodamine-phalloidin before use (11, 12).
Buffers. Myosin buffer contained 0.3 M KCl, 25 mM imidazole, 1 mM EGTA, 4 mM MgCl2, and 10 mM DTT, adjusted to pH 7.4. Actin buffer contained 25 mM KCl, 25 mM imidazole, 1 mM EGTA, 4 mM MgCl2, 10 mM DTT, and oxygen scavengers (0.1 mg ml 1 glucose oxidase, 0.018 mg ml 1 catalase, and 2.3 mg ml 1

Conict of interest statement: No conicts declared. This paper was submitted directly (Track II) to the PNAS ofce. Abbreviations: Tn, troponin; Tm, tropomyosin; pCa,
To

log 10 Ca concentration.

whom correspondence should be addressed. E-mail: warshaw@physiology.med. uvm.edu. address: Department of Biochemistry, University of Nevada, Reno, NV 89557.

Present

2005 by The National Academy of Sciences of the USA

www.pnas.org cgi doi 10.1073 pnas.0506326102

glucose) adjusted to pH 7.4. Motility solutions containing varying calcium and or MgATP were prepared as described (12).
Laser Trap. Detailed protocols for the laser trap assay have been described (17, 25). Contractile proteins were added to the experimental flow cell chamber with the following series of solution incubations: (i) skeletal myosin (0.120 g ml) for 2 min; (ii) 20 l of 0.5 mg ml BSA in myosin buffer for 1 min; (iii) 3 20 l of actin buffer; and (iv) 3 20 l of actin buffer with 10 M MgATP, tetramethyl-rhodamine-phalloidin-actin or tetramethyl-rhodamine-phalloidin-thin filaments, Nethylmaleimide-modified skeletal myosin-coated beads, and 100 nM excess Tn and Tm to ensure stoichiometric binding when thin filaments were studied. Experiments were performed at 25C. By manipulating the microscope stage, an N-ethylmaleimidemodified skeletal myosin-coated bead was captured in each of the two laser traps, and the ends of a single actin or reconstituted thin filament were then attached to the beads. The filament was pretensioned to 4 pN, and then brought near a skeletal myosin-coated, 3- m-diameter silica microsphere serving as a pedestal. The bright-field image of one bead attached to the filament was projected onto a quadrant photodiode detector, and signals were acquired for bead displacement parallel to the filaments long axis. Signals were recorded for at least 120 s before moving the bead-filament-bead to another pedestal within the flow cell. These data records were digitized at 4 kHz after initial filtering at 2 kHz. Laser Trap Data Analysis. Myosin strongly binds to an actin filament, undergoes its powerstroke, and displaces the filament. The binding of myosin attenuates the Brownian motion of the bead-filament-bead (see Fig. 2a) and shifts its mean position (16, 17). Using mean-variance analysis (17, 28) we determined the displacement or step size, d, generated by the myosin powerstroke, the duration, ton, that the myosin remains strongly bound to actin, and the total number of events within a data record. From the number of events, a frequency of attachment can be readily calculated by dividing the total record time by this number.

Fig. 1. Thin lament motility versus pCa. (a) Regulated thin lament velocity versus pCa in an in vitro motility assay. (b) Thin lament motility data analyzed as the percentage of motile laments (F) and percentage moving smoothly ( ) as a function of pCa. The pCa50 for motile laments is 6.81 0.04, which was signicantly greater than the pCa for smoothly moving laments (6.43 0.06; P 0.001).

Laser Trap Studies Using Unregulated Actin Filaments. Vactin in the

motility assay is the result of many myosin molecules interacting simultaneously with a given thin filament. How regulation is achieved in terms of myosins ability to interact with the thin filament was investigated by limiting the number of myosins in a laser trap assay. Fig. 2a is a sample displacement record obtained when an unregulated actin filament was lowered onto a pedestal surface that had been incubated with 0.1 g ml myosin. At this concentration only a single-myosin molecule on average interacts with the actin filament (30). At 10 M MgATP, we observed isolated mechanical events occurring at a frequency, f, of 1.75 0.3 s 1. Events were analyzed by using mean-variance (see Materials and Methods) to give an average step size, d, of 10.9 1.2 nm and a step duration, ton, of 26 4 ms (n 11 experiments) (31). When the myosin density was increased by a factor of 100 (i.e., incubated with 10 g ml), relatively continuous actin filament movement was observed with an average velocity of 0.14 0.08 m s (Fig. 2b). This velocity is slower than that seen previously in the motility assay (Vactin 0.47 0.06 m s) with similar myosin density and [MgATP] (31). The slower velocity in the laser trap is likely the result of the resistive load imposed by the laser trap.
Laser Trap Studies Using Regulated Thin Filaments. Fig. 2c shows a

In Vitro Motility. The movement of actin and regulated thin

filaments over a myosin-coated surface (100 g ml myosin) was determined as described (12, 29). Specifically, thin filament motility can be characterized in terms of: (i) the mean velocity of movement; (ii) the percent of filaments moving in a smooth and continuous fashion (defined as filaments having a standard deviation of velocity less than half the mean velocity); and (iii) the percent motile filaments (defined as filaments moving 0.33 m s). Results
Regulated Thin Filament Motility. Reconstituted thin filaments were fully regulated as demonstrated by the effect of calcium on myosin-based thin filament sliding (Fig. 1). As reported (11, 12), the velocity, Vactin, of regulated thin filaments increased sigmoidally with increasing Ca2 concentrations in the presence of 2 mM MgATP (Fig. 1a). Fitting the velocity data to the Hill equation yields a pCa50 of 6.40 0.04, a Hill coefficient of 1.94 0.32, and a maximal velocity of 6.0 0.2 m s. These data were also analyzed in terms of the percent that were motile and percent of filaments moving smoothly (Fig. 1b). At high Ca2 concentrations all filaments are motile and most move smoothly (76%). At subsaturating Ca2 , the proportion of motile filaments decreases with a smaller percent of these filaments moving smoothly.
Kad et al.

typical laser trap displacement trace obtained from a regulated thin filament interacting with a myosin surface incubated with

Fig. 2. Sample data traces obtained in a laser trap, showing displacement events of unregulated actin laments (a and b) and regulated thin laments (c and d). (a and b) Displacement events obtained when unregulated actin laments interact with surface sparsely coated with myosin (0.1 g ml myosin incubation) (a) and surface densely coated with myosin (10 g ml myosin incubation) (b). (c and d) Displacement events obtained when regulated thin laments interact with surface sparsely coated with myosin (0.1 g ml myosin incubation) (c) and surface densely coated with myosin (10 g ml myosin incubation) (d).

PNAS

November 22, 2005

vol. 102

no. 47

16991

BIOPHYSICS

Fig. 3. Short bursts of motility with regulated actin. (a) A sample data trace showing both individual displacements and long staircase stepping events generated by multiple-myosin molecules. After each staircase event, the restoring force exerted by the laser trap returns the thin lament to its original position. (b) Expanded view of the rst two steps in a staircase event. To accentuate the steps in the raw data trace, a line is drawn through the data to guide the eye. Region 0 represents the baseline where myosin is detached and displacement variance is high. Step 1 reects attachment of the rst myosin as evidenced by the decrease in variance. Step 2 is the displacement associated with the second myosin. (c) The mean-variance histogram of the data in b showing the baseline (marked as b) and two distinct displacement populations corresponding to a 3-nm step from baseline and a subsequent 16-nm step (or 19 nm relative to baseline).

Fig. 4. Rigor activation in the motility assay. Thin lament velocity at activating ( Ca; pCa 4) and subactivating Ca2 ( Ca; pCa 8) concentrations versus [MgATP].

0.1 g ml myosin. Unlike data obtained at this surface density with unregulated actin (Fig. 2a), the frequency of events was extremely low ( 0.01 s 1), preventing determination of the step size and duration. As expected for a completely reconstituted thin filament, the low event frequency was uniformly observed along the entire sampled length of the thin filament at both 25 and 100 mM KCl (data not shown). However, at 100-fold higher myosin density with 10 g ml incubation (Fig. 2d), the event frequency (0.5 0.3 s 1) resembled that of unregulated actin at 0.1 g ml incubation. Under these conditions, we determined the average step size, d 5.0 0.9 nm (n 11 experiments), and step duration, ton 69 33 ms (n 11 experiments). This d was 50% shorter (P 0.001; unpaired t test), whereas the ton was twice as long (P 0.001; unpaired t test) compared with unregulated actin (see above). When the myosin density was increased 2-fold (from 10 to 20 g ml incubation), both single binding and staircase stepping events were observed with an average distance moved during a staircase event of 43 2 nm (n 67) (Fig. 3a). Staircases are caused by multiple myosins simultaneously interacting with the thin filament until all myosins spontaneously detach. Interestingly, the frequency of steps within a staircase event ( f 11 1 s 1, determined as the inverse of the time between steps) is much higher than the frequency observed when only single steps predominate (see above), suggesting that the attachment of one myosin head to actin cooperatively accelerates the attachment rate of subsequent heads during a staircase. In addition, binding of the first myosin to actin affects the step size of subsequent myosin heads that bind (Fig. 3b). Using mean-variance analysis (Fig. 3c), the displacement of the first step in a staircase was 6.6 1.0 nm, whereas the second step was significantly larger (P 0.01, unpaired t test), 13.7 0.8 nm (n 40). The displacement of the first step is comparable to the step size observed when only a single myosin interacted with a regulated thin filament (see above). The second and subsequent steps in a staircase event were twice that of the first step, resembling the myosin step size with unregulated actin.
Regulated Thin Filament Motility at Low Ca2 with Low [MgATP].

analogous to the rigor activation of the actomyosin ATPase observed by Bremel and Weber (18). To fully characterize the apparent activation of the thin filament by the binding of rigor myosin heads to the thin filament, Vactin was measured over a range of [MgATP] between 2 and 125 M (Fig. 4). In the presence of Ca2 (pCa 4), Vactin increased with [MgATP], yielding a Km of 25 M consistent with values determined previously (31). Surprisingly, at low Ca2 (pCa 8) and [MgATP] below the Km, thin filament motility was observed at speeds equivalent to those observed at pCa 4. In contrast, complete inhibition of motility was observed at saturating MgATP and pCa 7 (Figs. 1 and 4). Discussion To characterize the molecular mechanisms of thin filament inhibition (see Fig. 1), we have determined the motion of thin filaments with high spatial and temporal resolution by using the laser trap under nominal Ca2 conditions (pCa 8). At myosin surface densities where only a single skeletal muscle myosin molecule can interact with an unregulated actin filament (i.e., fully on) (Fig. 2a), binding events are detected at a frequency (1.75 s 1), limited by both the inherent kinetics of the actomyosin ATPase cycle and the spatial constraints placed on the actomyosin interaction within the laser trap assay. With a regulated filament in the absence of Ca2 (i.e., fully off), the event frequency is significantly reduced (Fig. 2c). However, a 100-fold increase in the myosin concentration can restore the event frequency to that for unregulated actin (Fig. 2d). Consistent with prior solution biochemical studies (9, 32, 33), we demonstrate at the single-molecule level that TnTm dramatically reduces the frequency of actomyosin strong binding events at low Ca2 and thus is the primary cause for complete inhibition of continuous thin filament movement. The extent of inhibition within fibers may in fact be greater than observed here given the 1,000-fold difference in the heat liberated between resting and active muscle (34). Although reconstituted thin filaments fully inhibit in vitro motility, they may be less effective than thin filaments in muscle at regulating attachment kinetics because of spatial relationships and other factors. In addition to modulating the myosin attachment frequency, the presence of TnTm affects the inherent mechanics (i.e., step size, d) and kinetics (i.e., step duration, ton) of the actomyosin interaction. In the absence of Ca2 , a single-myosin molecule generates only half the displacement and remains strongly bound for twice as long compared with myosin binding to an unregulated actin filament. We have previously shown at 10 M ATP that ton is determined equally by the rate of ADP release from myosin and the rate of ATP binding to myosin (31). Therefore, TnTm in the absence of Ca2 could modulate one or both of these kinetic steps (35).
Kad et al.

Reconstituted thin filament motion could be restored under inhibitory conditions (i.e., pCa 8) if the [MgATP] was lowered to subsaturating concentrations 75 M. This phenomenon is
16992 www.pnas.org cgi doi 10.1073 pnas.0506326102

Fig. 5. Simple model of thin lament regulation. (Top) For unregulated actin, the myosin attachment rate is limited by the weak-to-strong binding transition rate (kws). (Middle) For a regulated thin lament reconstituted with Tn and Tm in the absence of Ca2 , the myosin attachment rate is reduced by an inhibition constant (kws reg). The ability of myosin to bind even in the absence of Ca2 suggests that thermal uctuations of Tm on actin (indicated by a bent Tm and dashed arrow) expose potential binding sites. (Bottom) However, once a myosin binds strongly to actin, neighboring TnTm regulatory units, spanning three units, are activated, allowing myosins to bind at an accelerated rate [kws (acc reg)].

Fig. 6. Monte Carlo simulation of the interaction of myosin with unregulated (a and b) and regulated (c and d) actin. These data represent simulations of experimental data similar to that in Fig. 2, which is included as Insets. (a and b) Simulated laser trap displacement data at low surface density (0.1 g ml myosin) (a) and high surface density (10 g ml myosin) (b) where continuous velocity is observed. (c and d) Simulated data with regulated thin laments at low surface density (0.1 g ml myosin the event shown was the only event in a 20-s record) (c) and high myosin surface density (10 g ml myosin) (d). See Appendix for simulation conditions.

In the absence of Ca2 , TnTm prevents myosins optimal interaction with actin (4, 36) and could account for the step size being half that of whole skeletal myosin (see Results). Myosins two heads are required to generate a 10-nm displacement because single-headed myosin generates only half the displacement and remains attached to actin for a longer period after the step (37, 38). Thus, TnTm at low Ca2 may allow only one of myosins two heads to interact with actin, disrupting intrahead communication that is required to generate a 10-nm step. However, binding of the first myosin molecule appears to overcome the effect of TnTm because subsequent myosins are capable of generating their normal displacement as evidenced by comparing the step size of the first and second steps in a staircase stepping event (see Results and Fig. 3b). This scenario is not the case for unregulated actin where the first and subsequent steps are 10 nm (37). Therefore, binding of the first head is critical to the cooperative activation of the thin filament (see below).
A Simple Model: TnTm Inhibits Myosin Strong Binding. Experiments

to myosin strong binding. Geeves and coworkers (39) suggest that in the absence of Ca2 the regulatory state of the thin filament is 70% blocked, 25% closed, and 5% open. Because we cannot distinguish between detached or weakly bound myosin in the laser trap, our prediction of 99% actin site inaccessibility is consistent with the combined percentage of blocked and closed states (95%) estimated by Geeves and coworkers (5, 39). Because strong binding events still occur in the laser trap assay under conditions where myosin should be sterically prevented from binding, thermal fluctuations of the Tm on actin (Fig. 5) must temporarily expose actin binding sites (40), which could also account for the 5% open-state probability predicted by Geeves and coworkers (5).
Cooperative Regulation Experimentally and in Silico. In a laser trap, 20 g ml of skeletal muscle myosin is sufficient to support continuous movement of unregulated actin (see Fig. 2b). However, with regulated filaments, at the same myosin concentration, continuous motion was not seen; instead, short bursts of motility (i.e., staircase events) and long pauses were observed (Fig. 3a). To model these short bursts (Fig. 7a), we introduced a cooperativity constant (acc 10, see Eq. 3 in Appendix) that accelerated the strong binding rate 10-fold for myosin binding subsequent to the first head (Fig. 5). Such cooperativity of binding has been observed in earlier solution biochemistry studies (9, 32). From a structural viewpoint, binding of the first head may perturb Tms equilibrium position on actin so that other heads can bind to nearby myosin binding sites more readily. However, these additional heads still do not bind at a rate equivalent to that with unregulated actin; instead their binding is still inhibited 10-fold (in our model, acc reg 0.1). Thus, we predict that the myosin binding site accessibility for a regulated thin filament in the absence of Ca2 is reduced 99%, but with the attachment of a single head the thin filament is activated 10-fold so that 90% of the sites remain inaccessible to strong binding. With limiting numbers of myosin molecules interacting with the thin filament in the laser trap, it is possible to estimate the distance over which cooperative activation occurs. For a nitrocellulose surface incubated with 20 g ml myosin, where bursts of activity were observed, we previously determined based on NH4-ATPase measurements that 18 heads or nine myosin molecules are available per m of actin filament length (30), i.e., 111 nm between myosin molecules. Given this intermolecular
PNAS November 22, 2005 vol. 102 no. 47 16993

in the laser trap are able to detect only two actomyosin states: (i) the attached state when myosin is strongly bound, and (ii) the detached or weakly bound state. With this simple view, we have developed a Monte Carlo model of thin filament regulation in which TnTm modulates the attachment frequency ( f ) for a given number of myosin molecules (nM) by either inhibiting (reg) or accelerating (acc) the rate that myosin transitions from the weak to the strongly bound state (kws) (see Fig. 5 and Appendix, which is published as supporting information on the PNAS web site, for details): Attachment rate f Ca2 nM k ws acc reg .

In the absence of , the presence of TnTm reduced the frequency of actomyosin interactions ( f ) 100-fold (Fig. 6 a vs. c), such that kws was reduced by an inhibition constant (reg 100, see Fig. 5 and Eq. 2 in Appendix). This finding implies that in the absence of Ca2 99% of the actin binding sites are inaccessible
Kad et al.

BIOPHYSICS

Fig. 7. Cooperative activation underlies the short bursts of motility, which in turn leads to rigor activation at low [ATP]. (a) With a 10-fold activation of strong binding for heads at 20 g ml myosin (acc 10, see Appendix) short bursts of motility equivalent to Fig. 3 are predicted. (b) Simulated thin lament velocities at activating ( Ca; pCa 4) and subactivating Ca2 ( Ca; pCa 8) concentrations versus [MgATP] were simulated at 100 g ml myosin using the same parameters as in Fig. 6 but with varying [MgATP] (see Appendix for details). Fig. 8. Predicting the mode of lament motion. (a and b) Velocities were simulated from individual model lament runs at 10 M ATP (a) and 60 M ATP (b). (c) At 10 M ATP thin lament motion is continuous, leading to a high percentage of laments moving smoothly, which compares well with the experimental data. At 60 M ATP the model predicts fewer laments moving smoothly because of pauses between detachment limited runs, which is also consistent with the experimentally observed data.

spacing, the accelerated binding of heads subsequent to the first on average occurs within three regulatory units distal to the initial binding event, a distance equal to the next accessible actin target zone along the TnTm strand. This simple calculation assumes a homogeneous surface density, and there are no significant differences in the attached lifetime between the first and subsequent heads. Therefore, thin filament activation is a cooperative process in which the strong binding of a myosin head sets up an activation wave over long distances that is propagated through adjacent TnTm units (6, 4144).
Cooperative Rigor Activation. To enhance the detection of myosin

binding events in the laser trap, experiments were performed at limiting [MgATP] (i.e., 10 M), which results in a myosin head attached to actin waiting for ATP to bind (31). Therefore, consistent with rigor activation of actomyosin ATPase activity, the model predicted that the short bursts of activity at low [MgATP] (Fig. 3) resulted from rigor activation of the thin filament (10, 12, 18). To confirm this prediction, we measured Vactin of regulated thin filaments in the motility assay at varying [MgATP]. At saturating MgATP with minimal Ca2 ( pCa7), Vactin is completely inhibited (see Figs. 1 and 4); however, at low [MgATP] ( 75 M), when the rigor state is significantly populated, thin filament motility occurs with a Vactin comparable to that for a fully Ca2 -activated thin filament (Fig. 4). The observed rigor activation of event frequency in a laser trap (Fig. 3), rigor activation of velocity in the motility assay (Fig. 4), and rigor activation of ATPase activity in solution studies (18) all are consistent with a simple model in which Ca2 and rigor heads both activate actinmyosin binding (Fig. 4 and see Appendix and Fig. 9, which is published as supporting information on the PNAS web site, for details). If so, the proposed model described above can provide a mechanistic insight to the dependence of Vactin on [MgATP] for thin filaments in the presence and absence of Ca2 . In the presence of Ca2 Vactin is proportional to d ton (31, 45), so that Vactin is limited by the rate of myosin detachment (g 1 ton), which is determined by both the rates of ADP release and the MgATP binding to myosin (31). Thus, as the [MgATP] is lowered, Vactin slows as myosin spends an increasing amount of
16994 www.pnas.org cgi doi 10.1073 pnas.0506326102

time, ton, attached to actin waiting for ATP to bind (Fig. 7b). In contrast, in the absence of Ca2 and at high [MgATP], we have shown that the rate of myosin attachment limits Vactin (see Fig. 4). However, at low [MgATP], the prolonged attached lifetime of the rigor state cooperatively activates the thin filament (Fig. 5), resulting in an increased actinmyosin binding frequency and detachment limited thin filament movement even in the absence of Ca2 (see Figs. 3 and 7a). This mechanistic view of rigor activation is predicted by our model, as demonstrated by the striking similarity between the simulated and actual Vactin data (Fig. 7b vs. Fig. 4). In addition, the model predicts that the similarity in Vactin observed at 10 and 60 M MgATP in the absence of Ca2 (Fig. 4) results from different underlying motile mechanisms (Fig. 8). At 10 M MgATP the attached myosin lifetime is sufficiently long to result in multiple myosins binding to and steady motion of the actin filament (Fig. 8a), with myosin detachment limiting Vactin. At 60 M MgATP, myosins shorter attached lifetime reduces the probability of rigor activation, resulting in discontinuous thin filament motility comprised of bursts of motility and pauses (Fig. 8b). During the bursts, the velocity is limited by myosins detachment rate; however, the pause duration is limited by myosins attachment rate. When the average velocity over time is determined at 60 M MgATP, including pauses, its value is nearly identical to that at 10 M MgATP even though the dynamics of filament motion are substantially different. In fact, the model predicts that the percent of filaments moving smoothly is far greater at 10 M versus 60 M MgATP, which agrees with the experimental data (Fig. 8c).
Conclusion. The laser trap data presented here suggests that although the inherent mechanics and kinetics of a single-myosin
Kad et al.

molecule are altered 2-fold by the presence of TnTm, these alterations are not large enough to completely inhibit thin filament velocity at low Ca2 . Therefore, TnTm regulates actin in the absence of Ca2 by reducing the probability of myosins first encounter with actin 100-fold. However, upon myosin binding, the thin filament is activated 10-fold from its inhibited state because of activation of neighboring regulatory units. Because strong binding of the first head can only partially activate the thin filament, we infer that regulation must be at least a three-state process where further activation is associated with the binding of Ca2 , akin to the blocked, closed, and open states (5). At higher [Ca2 ], regulation may have additional modulatory effects on actomyosin kinetics as evidenced in motility studies where regulated thin filament velocity and force are enhanced
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Haselgrove, J. C. (1973) Cold Spring Harbor Symp. Quant. Biol. 37, 341352. Huxley, H. E. (1973) Cold Spring Harbor Symp. Quant. Biol. 37, 361376. Lehrer, S. S. & Morris, E. P. (1982) J. Biol. Chem. 257, 80738080. Phillips, G. N. J., Fillers, J. P. & Cohen, C. (1986) J. Mol. Biol. 192, 111131. McKillop, D. F. & Geeves, M. A. (1993) Biophys. J. 65, 693701. Vibert, P., Craig, R. & Lehman, W. (1997) J. Mol. Biol. 266, 814. Lehman, W., Rosol, M., Tobacman, L. S. & Craig, R. (2001) J. Mol. Biol. 307, 739744. Craig, R. & Lehman, W. (2001) J. Mol. Biol. 311, 10271036. Trybus, K. M. & Taylor, E. W. (1980) Proc. Natl. Acad. Sci. USA 77, 72097213. Swartz, D. R. & Moss, R. L. (1992) J. Biol. Chem. 267, 2049720506. Homsher, E., Kim, B., Bobkova, A. & Tobacman, L. S. (1996) Biophys. J. 70, 18811892. Gorga, J. A., Fishbaugher, D. E. & VanBuren, P. (2003) Biophys. J. 85, 24842491. Gordon, A. M., LaMadrid, M. A., Chen, Y., Luo, Z. & Chase, P. B. (1997) Biophys. J. 72, 12951307. Fraser, I. D. & Marston, S. B. (1995) J. Biol. Chem. 270, 78367841. Finer, J. T., Simmons, R. M. & Spudich, J. A. (1994) Nature 368, 113119. Molloy, J. E., Burns, J. E., Kendrick-Jones, J., Tregear, R. T. & White, D. C. (1995) Nature 378, 209212. Guilford, W. H., Dupuis, D. E., Kennedy, G., Wu, J., Patlak, J. B. & Warshaw, D. M. (1997) Biophys. J. 72, 10061021. Bremel, R. D. & Weber, A. (1972) Nat. New Biol. 238, 97101. Hill, T. L., Eisenberg, E. & Greene, L. (1980) Proc. Natl. Acad. Sci. USA 77, 31863190. Tobacman, L. S. & Butters, C. A. (2000) J. Biol. Chem. 275, 2758727593. Brenner, B. (1988) Proc. Natl. Acad. Sci. USA 85, 32653269. Chalovich, J. M. (1992) Pharmacol. Ther. 55, 95148. Zou, G. & Phillips, G. N., Jr. (1994) Biophys. J. 67, 1128. Warshaw, D. M., Desrosiers, J. M., Work, S. S. & Trybus, K. M. (1990) Prog. Clin. Biol. Res. 327, 815826.

over actin alone (12, 46) and by the Ca2 -dependent changes in the rate of tension redevelopment, ktr, in fiber studies (47). Future studies will likely provide greater molecular insight to the structural dynamics of TnTm movement on actin because the development of single-molecule fluorescence detection techniques may allow simultaneous assessment of TnTm motion coupled with mechanical measurements of the regulated actomyosin interaction.
We thank Amy Armstrong, Kelly Begin, and Joe Gorga for assistance with experiments; Guy Kennedy for technical support; and Jeffrey Moore and Edward Debold for helpful discussions. This study was funded by National Institutes of Health Grant HL59408 (to D.M.W.), National Institutes of Health Grant HL077637 (to P.V.), and National Institutes of Health Grant P20RR018751-01 (to J.E.B.).
25. Palmiter, K. A., Tyska, M. J., Haeberle, J. R., Alpert, N. R., Fananapazir, L. & Warshaw, D. M. (2000) J. Muscle Res. Cell Motil. 21, 609620. 26. Pardee, J. D. & Spudich, J. A. (1982) Methods Enzymol. 85, 164181. 27. Potter, J. D. (1982) Methods Enzymol. 85, 241263. 28. Patlak, J. B. (1993) Biophys. J. 65, 2942. 29. VanBuren, P., Alix, S. L., Gorga, J. A., Begin, K. J., LeWinter, M. M. & Alpert, N. R. (2002) Am. J. Physiol. 282, H1665H1671. 30. Harris, D. E. & Warshaw, D. M. (1993) J. Biol. Chem. 268, 1476414768. 31. Baker, J. E., Brosseau, C., Joel, P. B. & Warshaw, D. M. (2002) Biophys. J. 82, 21342147. 32. Greene, L. E. & Eisenberg, E. (1980) Proc. Natl. Acad. Sci. USA 77, 26162620. 33. Williams, D. L. J. & Greene, L. E. (1983) Biochemistry 22, 27702774. 34. Woledge, R. C., Curtin, N. A. & Homsher, E., eds. (1985) Energetic Aspects of Muscle Contraction (Academic, New York), pp. 167271. 35. Rosenfeld, S. S. & Taylor, E. W. (1987) J. Biol. Chem. 262, 99949999. 36. Lehman, W., Vibert, P., Uman, P. & Craig, R. (1995) J. Mol. Biol. 251, 191196. 37. Tyska, M. J., Dupuis, D. E., Guilford, W. H., Patlak, J. B., Waller, G. S., Trybus, K. M., Warshaw, D. M. & Lowey, S. (1999) Proc. Natl. Acad. Sci. USA 96, 44024407. 38. Kad, N. M., Rovner, A. S., Fagnant, P. M., Joel, P. B., Kennedy, G. G., Patlak, J. B., Warshaw, D. M. & Trybus, K. M. (2003) J. Cell Biol. 162, 481488. 39. Schaertl, S., Lehrer, S. S. & Geeves, M. A. (1995) Biochemistry 34, 1589015894. 40. Squire, J. M. & Morris, E. P. (1998) FASEB J. 12, 761771. 41. Geeves, M. A. & Lehrer, S. S. (1994) Biophys. J. 67, 273282. 42. Morris, C. A., Tobacman, L. S. & Homsher, E. (2001) J. Biol. Chem. 276, 2024520251. 43. Metzger, J. M. & Moss, R. L. (1991) J. Gen. Physiol. 98, 233248. 44. Gordon, A. M., Homsher, E. & Regnier, M. (2000) Physiol. Rev. 80, 853924. 45. Huxley, H. E. (1990) J. Biol. Chem. 265, 83478350. 46. Fujita, H., Sasaki, D., Ishiwata, S. & Kawai, M. (2002) Biophys. J. 82, 915928. 47. Brenner, B., Kraft, T. & Chalovich, J. M. (1998) Adv. Exp. Med. Biol. 453, 177184.

Kad et al.

PNAS

November 22, 2005

vol. 102

no. 47

16995

BIOPHYSICS