Article Information

• PDF (487 kb)
• Full Text with Large Figures
• Supporting Material
• Cited by in Scopus (17)

PubMed

• Articles by Mitsuhiro Iwaki
• Articles by Toshio Yanagida

Related Articles

• A Chemically Reversible Brownian Motor: Application to Kines...

  A Chemically Reversible Brownian Motor: Application to Kinesin and Ncd Biophysical Journal, Volume 77, Issue 2, 1 August 1999, Pages 993-1002 R. Dean Astumian and Imre Derényi Abstract Kinesin and nonclaret disjunctional protein (ncd) are two microtubule-based molecular motors that use energy from ATP hydrolysis to drive motion in opposite directions. They are structurally very similar and bind with similar orientations on microtubule. What is the origin of the different directionality? Is it some subtle feature of the structure of the motor domains, not apparent in x-ray diffraction studies, or possibly some difference near the neck regions far from the microtubule binding site? Perhaps because the motors function as dimers, the explanation involves differences in the strength of the interaction between the two motor monomers themselves. Here we present another possibility, based on a Brownian ratchet, in which the direction of motion of the motor is controlled by the chemical mechanism of ATP hydrolysis and is an inherent property of a single head. In contrast to conventional power stroke models, dissociation of the individual heads is not obligatory in the chemomechanical cycle, and the steps during which motion and force generation occurs are best described as one-dimensional thermally activated transitions that take place while both heads are attached to the microtubule. We show that our model is consistent with experiments on kinesin in which the velocity is measured as a function of external force and with the observed stiochiometry of one ATP/8-nm step at low load. Further, the model provides a way of understanding recent experiments on the ATP dependence of the variance (randomness) of the distance moved in a given time. Abstract | Full Text | PDF (142 kb)

• Protein-Protein Ratchets: Stochastic Simulation and Applicat...

  Protein-Protein Ratchets: Stochastic Simulation and Application to Processive Enzymes Biophysical Journal, Volume 81, Issue 3, 1 September 2001, Pages 1333-1344 Charles J. Brokaw Abstract Interaction between a protein and a series of binding sites on a cytoskeletal substrate can create a resistance, or “protein friction,” as the protein is moved along the substrate. If attachment and detachment rates are specified asymmetrically, this resistance can depend on the direction of movement, and the binding interaction acts as a ratchet. Stochastic computer simulations have been used to examine this type of protein-protein interaction. The performance of a protein-protein ratchet in the piconewton and nanometer range is significantly limited by thermal fluctuations, which in experimental measurements with single molecules are evident as Brownian motion. Simulations with a two-component model combining a conventional motor enzyme model with a protein-protein ratchet confirm previous suggestions that the processive movement of a single motor enzyme molecule against a load, as seen in experiments with inner arm dynein molecules, might be made possible by an accessory protein interaction that prevents backward slippage. When this accessory protein interaction is defined so that it acts as a ratchet, backward slippage can be prevented with minimal interference with forward progression. Abstract | Full Text | PDF (241 kb)

• Fluctuations and Randomness of Movement of the Bead Powered ...

  Fluctuations and Randomness of Movement of the Bead Powered by a Single Kinesin Molecule in a Force-Clamped Motility Assay:Monte Carlo Simulations Biophysical Journal, Volume 83, Issue 5, 1 November 2002, Pages 2360-2369 Yi-der Chen, Bo Yan and Robert J. Rubin Abstract The motility assay of K. Visscher, M. J. Schnitzer, and S. M. Block (, 400:184–189, 1999) in which the movement of a bead powered by a single kinesin motor can be measured is a very useful tool in characterizing the steps of the mechanochemical cycle of kinesin motors, because in this assay the external force applied to the bead can be controlled (clamped) arbitrarily. However, because the bead is elastically attached to the motor and the response of the clamp is not fast enough to compensate the Brownian motion of the bead, interpretation or analysis of the data obtained from the assay is not trivial. In a recent paper (Y. Chen and B. Yan, . . 91:79–91, 2001), we showed how to evaluate the velocity of the bead and the motor in the motility assay for a given mechanochemical cycle. In this paper we extend the study to the evaluation of the or the of the velocity using a Monte Carlo simulation method. Similar to the mean, we found that the randomness of the velocity of the motor is also influenced by the parameters that affect the dynamic behavior of the bead, such as the viscosity of the medium, the size of the bead, the stiffness of the elastic element connecting the bead and the motor, etc. The method presented in this paper should be useful in modeling the kinetic mechanism of any processive motor (such as conventional kinesin and myosin V) based on measured force-clamp motility data. Abstract | Full Text | PDF (363 kb)

• …more

Copyright © 2006 The Biophysical Society. All rights reserved.
Biophysical Journal, Volume 90, Issue 10, 3643-3652, 15 May 2006

doi:10.1529/biophysj.105.075721

Muscles and Contractility

Cargo-Binding Makes a Wild-Type Single-Headed Myosin-VI Move Processively

Mitsuhiro Iwaki^*, †, Hiroto Tanaka^†, Atsuko Hikikoshi Iwane^‡, Eisaku Katayama^§, Mitsuo Ikebe^¶ and Toshio Yanagida^{*, †, ‡, ,}  

^* Department of Biophysical Engineering, Osaka University, Suita, Osaka, Japan
^† Formation of Soft Nanomachines, Core Research for Evolution Science and Technology, Japan Science and Technology Agency, Suita, Osaka, Japan
^‡ Soft Biosystem Group, Laboratories for Nanobiology, Graduate School of Frontier Biosciences, Osaka University, Suita, Osaka, Japan
^§ Division of Biomolecular Imaging, Institute of Medical Science, The University of Tokyo, Minato-ku, Tokyo, Japan
^¶ University of Massachusetts Medical School, Worcester, Massachusetts

Address reprint requests to Toshio Yanagida, Soft Biosystem Group, Laboratories for Nanobiology, Graduate School of Frontier Biosciences, Osaka University, 1-3,Yamadaoka, Suita, Osaka 565-871, Japan. Tel.: 81-6-6879-4632; Fax: 81-6-6879-4634.

During the last decade, a number of myosin-like molecules have been discovered, making it evident that myosin constitutes a diverse superfamily 1,2,3. In contrast to well-known filament-forming myosins, now classified as myosin-II, newly found myosins do not form filaments and are called unconventional myosins 4. One of the most important features of these myosins is that they function as intracellular cargo carriers, thus playing a critical role in vesicular trafficking in cells. Supporting this notion, several members of the myosin superfamily have been shown to be processive motors, in which myosin travels a long distance along actin filaments without dissociation 5,6,7,8,9. Myosin-Va was the first of many myosins shown to be processive 5. Because myosin-Va has a double-headed structure with a long neck, it is thought that myosin-Va moves processively along an actin filament with a hand-over-hand mechanism 10,11,12,13,14. Subsequently, it was shown that myosin-VI, which moves along actin filaments in the opposite direction to myosin-V 15, also shows processive movement 7,8 with a hand-over-hand mechanism 16,17. The myosin-VI constructs tested were double-headed by introducing a coiled-coil element at the C-terminal end of the molecule. However, it was reported quite recently that chicken wild-type myosin-VI is single-headed, casting a doubt on its processivity 18. Yet, in vivo, myosin-VI is implicated in moving endocytic vesicles into a cell 19,20,21.

We observed that wild-type single-headed myosin-VI without cargo did not move processively, consistent with a previous study 18. In cells, however, myosin-VI attaches a cargo to its tail end and transports it along actin filaments. The myosin head tethered to the cargo may undergo dissociation-reassociation with actin without diffusing away from the actin because of the relatively slow Brownian motion of the cargo. Therefore, we hypothesized that the cargo binding to myosin-VI may allow the myosin-VI motor to move processively. As a fact, we found in this study that myosin-VI processively moved along actin filaments with large (∼40-nm) steps upon binding to a cargo. Furthermore, we found that a cargo-associated single-headed wild-type myosin-VI processively moved in a solution with viscosity similar to that of a cellular environment.

Actin was obtained from rabbit skeletal muscle and purified as described 22. Biotinylated actin was prepared by linking Biotin-(AC₅)₂-Osu (Dojindo, Kumamoto, Japan) to the amino-terminus of actin as described 23. Actin filaments were obtained by mixing native actin and biotinylated actin at 10:1 in buffer (0.1M KCl, 10mM HEPES-KOH (pH 7.8)). To visualize under an optical microscope, actin filaments were labeled with phalloidin-tetramethylrhodamine (TMR). Preparation and expression for mouse full-length (wild-type) (M6WT) and engineered double-headed myosin-VI construct (M6DH) (fused with hexahistidine tag at the C-terminus) was as described previously 8. Purification was done by His-tag affinity chromatography.

Fluorescent latex beads (diameter 0.2μm, 2% solids, Molecular Probes, Eugene, OR) were incubated with 6×His monoclonal antibody (1mg/ml, BD Biosciences, Franklin Lakes, NJ) at the molar ratio of 1:200 for 1h at 0°C and sedimented at 21,800×g for 5min. The sedimented beads were suspended in buffer (25mM KCl, 20mM HEPES-KOH (pH 7.8 or 7.1), 5mM MgC_l2, and 1mM EGTA) and sedimented again. This process was repeated three times. The number of 6×His antibody molecules bound to a bead was estimated to be at least 50. 1μl of 6×His antibody-coated beads was incubated for 10min in assay buffer containing 10mg/ml bovine serum albumin (Sigma-Aldrich, St. Louis, MO) and His-tagged M6WT or M6DH. The beads were observed within 30min after preparation.

Sample chambers with 10-μl capacity were made by placing a coverslip over a glass slide separated by two parallel slips of polyester film ∼25μm in thickness. Fluorescently labeled actin filaments containing biotinylated G-actin (10% of total G-actin) were bound to a glass surface that had been coated with avidin-biotinylated bovine serum albumin. Myosin tagged beads at 3–10 pM in assay buffer (25mM KCl, 20mM HEPES-KOH (pH 7.8 or 7.1), 5mM MgCl₂, 1mM EGTA, ATP, and oxygen scavenger system (0.11mg/ml glucose oxidase, 18μg/ml catalase, 2.3mg/ml glucose, and 0.5% 2-mercaptoethanol)) were introduced into the sample chamber. Myosin-VI tagged beads were optically trapped and brought into contact with actin filaments. Imaging and detection of bead positions were done as described 24,25. The response time of the 0.2-μm bead was 230μs at 0.25pN load, which was enough to detect stepping movement of myosin-VI at low load. All experiments were performed in assay buffer at 24°C.

Bead displacements were recorded at a sampling rate of 24kHz with a bandwidth of 10kHz. The load exerted on the beads was calculated from the bead displacement multiplied by the trap stiffness (∼8fN/nm), which was determined from the variance of the Brownian motion of a trapped bead by the equipartition theorem of energy. Interactions were detected as a stiffness increase (0.02–0.04pN/nm) followed by a long dwell (>50ms) from the raw trace, and the steps were detected as a rapid (<5-ms) positional change (larger than the standard deviation of the bead fluctuation, or 5–8nm) followed by a long dwell (50ms) from the Chebyshev-filtered trace. The step size was estimated from the difference between the mean prestep position and poststep position. To determine the precise step size at low load, we calculated the elastic component in our single trapping geometry and estimated the correction factor depending on load 26. At no load condition could we determine the size of each step directly without correction because of the absence of attenuation with optical trapping.

Single-molecule imaging of green-fluorescent-protein (GFP)-tagged myosin 27,28 and electron micrographs of myosin were performed as previously described 8. Single-molecule imaging of GFP-tagged myosin-VI with fluorescent bead was performed as follows. Emission wavelength of the fluorescent bead (E_x/E_m=600/610, Molecular Probes) was chosen not to overlay the GFP emission wavelength. Because the emission of both TMR-labeled actin filaments and the fluorescent beads are highly bright compared to that of single GFP molecules, we decided to use an excitation laser of 476-nm wavelength (which is not an optimized wavelength for TMR and fluorescent beads) to decrease these emission intensities. Finally, we simultaneously observed single GFP molecules and fluorescent beads and TMR-labeled actin filaments by choosing the dichroic mirror and filter carefully.

When we observed the GFP bound bead in a high viscous solution, the Brownian motion of the GFP molecule synchronized with that of the bead could be observed. By using this observation system, we analyzed the GFP-tagged myosin-VI binding ratio to the bead. Because of both translational and rotational Brownian motion of the bead, the fluorescence intensity of the GFP molecule was unstable. Therefore, to analyze the photobleaching process, we analyzed the GFP fluorescence on the bead stuck on the glass surface.

Methylcellulose (Sigma-Aldrich) was dissolved at 1% (w/v) in buffer solution (25mM KCl, 20mM HEPES-KOH (pH 7.8), 5mM MgCl₂, and 1mM EGTA) and stocked. The stocked methylcellulose buffer was diluted to 0.5% by adding assay buffer (25mM KCl, 20mM HEPES-KOH (pH 7.8), 5mM MgCl₂, 1mM EGTA, ATP, oxygen scavenger system, and 1 mg/ml Casein (Nacalai Tesque, Kyoto, Japan) and used in the experiment at 24°C.

Two-dimensional diffusion of a bead (220nm in diameter) was tracked (> 1μm) and the diffusion coefficient was calculated by plotting the mean-square displacement versus the time interval using the equation 〈d²(t)〉=4Dt. Here d is the displacement, t is the time, and D is the diffusion coefficient. The calculated diffusion coefficient in water (D=7.9±0.3×10²nm²ms⁻¹) was consistent with that obtained from the Power spectrum density of the Brownian motion of laser-trapped bead in water (see Supplementary Material ). In the condition of 0.5% (w/v) methylcellulose, the diffusion coefficient of the bead was 40 times smaller than that of water.

To confirm if mouse recombinant wild-type myosin-VI used in this study is also monomeric, we observed M6WT by electron microscopy (Figure 1A). A field of rotary shadowed myosin-VI shows globular molecules followed by an ∼30-nm thin extension. We observed >300 molecules and confirmed that mouse M6WT is also a monomer and single-headed. To further confirm that mouse M6WT is a monomer, we observed single GFP-tagged M6WT in motility assay solution (see Materials and Methods) by total internal reflection microscopy 27,28. In all observations (187 spots), the GFP fluorescence photobleached in a single step, suggesting that M6WT is a monomer in the motility assay, too. To test for processivity of M6WT, we observed the single-molecule interaction between GFP-tagged M6WT and an actin filament. In all observations (200 spots), the GFP-tagged M6WT did not show processive movement over a 60-nm spatial resolution (Figure 1B) as shown previously 18, although the GFP-tagged M6DH did show processive movement (Figure 1C). Figure 1D shows durations of GFP-tagged M6WT (black bars) and GFP-tagged M6DH (gray bars) attachment on an actin filament. Both distributions fit well by a single-exponential curve (M6WT, τ=0.085s; M6DH, τ=0.42s). ATP turnover time, ∼0.11s, is almost the same as the duration time of attachment for M6WT 29, which also supports that the movement of M6WT without cargo is not processive.

Display large version of this figure

Figure 1

Structure and single-molecule imaging of M6WT. (A) Schematic drawing of M6WT. Myosin-VI consists of a motor domain (red), two calmodulin binding domains (gray), an ∼30-nm-long α-helix and a cargo-binding domain (black). At right is an electron micrograph of rotary-shadowed M6WT. Scale bar, 50nm. (B and C) Movement of GFP-tagged single-headed M6WT and M6DH, each without cargo. Time courses for GFP spot movements of single-headed M6WT (B) and M6DH (C) on an actin filament. (D) Duration of attachment of GFP-tagged single-headed M6WT (black bars) and GFP-tagged M6DH (gray bars) to an actin filament.

Although M6WT without cargo did not show processive movement, myosin-VI functions as a cargo transporter in cells, suggesting that myosin-VI in cells moves cargos processively. We hypothesized that the cargo binding to myosin-VI may allow the myosin-VI motor to move processively. Because myosin-VI transports 100- to 200-nm-diameter vesicles attached at its tail globular domain in cells 19, we attached a 200-nm-diameter polystyrene bead to this region and observed the bead movement by using an optical trapping nanometry system which enabled us to detect clearly the small and fast movement with nanometer accuracy and high temporal (submillisecond) resolution at low load 24,25. (Figure 2A and Materials and Methods).

Display large version of this figure

Figure 2

Evidence for cargo transport by a single M6WT molecule. (A) Measurement system. A myosin-coated bead was optically trapped and brought into contact with an actin filament attached to a glass surface. The bead image was projected onto a quadrant photodiode and the movement was determined with nanometer accuracy. (B) Typical photobleach of GFP on a bead stuck on the glass surface at M6WT/bead mixing stoichiometry of 2:1. (Upper) GFP-tagged M6WT on a bead. (Lower) GFP-tagged M6DH on a bead. (C) Poisson statistics for actin filament-binding probability. (Solid line) The probability, 1−exp(−λ_mx), that a bead carries one or more active M6WT molecules, fit by a least-squares method (λ_m=0.082, reduced χ²=0.03). Data are shown as mean±n^1/2N⁻¹, where n is the number of beads that moved and N (>79) is the number of trials. (Dashed line) The probability that a bead carries two or more active M6WT molecules (reduced χ²=0.98).

We examined whether the observed bead movement was indeed due to single monomeric myosin-VI molecules. We first determined the fraction of fluorescent GFP tagged to myosin-VI heads by monitoring the photobleach reaction of the fluorescence of the GFP-tagged M6DH attached to a glass surface 30, since it is known that GFP molecules are not always fluorescently active. Based upon the ratio of single-step and double-step photobleach events, the fraction of fluorescently active GFP was determined to be 83% in our assay system. The experimental data described below was corrected based upon this result.

Because myosin-VI was bound to a bead via an antibody for His-tag that has two binding sites for myosin-VI (see Materials and Methods), we first confirmed that myosin-VI bound to the two binding sites of the antibody randomly and not cooperatively through dimerization. GFP-tagged myosin-VI molecules were mixed with antibodies at a molar ratio of 1:10 in solution and incubated for 30min. This ratio was fivefold higher than that of bead to antibody (1:50) (see Materials and Methods). The fluorescence of complexes spread to a glass surface was observed at the single molecular level 8,27,28. Of the 135 fluorescent spots of GFP observed, the number of spots showing single-step bleaching was 131 (97.1%) and the number showing double-step bleaching was 4 (2.9%). These fractions are in agreement with the Poisson statistics for random binding of myosin-VI to the antibody.

Second, we confirmed that GFP-tagged M6WT bound to a bead randomly, not cooperatively. When GFP-tagged M6WT molecules were mixed with antibody-coated beads at the stoichiometry of 2:1, 106 out of 152 beads (70%) did not show fluorescence and 46 beads did, indicating that ∼30% beads bound fluorescent GFP-tagged M6WT (Figure 2B). Thirty-seven (24%), 8 (5.2%), and 1 (0.7%) beads out of 152 beads showed single-, double- and triple-step photobleaching, respectively. Considering the fraction of nonfluorescent GFP, the fraction of M6WT bound successfully to the beads is calculated as (37×1+8×2+1×3)/0.83/(152×2)=0.22. If GFP-tagged M6WT binds randomly to the bead, the number of GFP-tagged M6WT bound to a bead can be given by the Poisson statistics, P(n)=(λ_bN)ⁿ/n!exp(−λ_bN), where n is the number of GFP-tagged M6WT molecules bound to a bead, N is the mixing molar ratio of M6WT to bead, and λ_b is the fraction for M6WT to successfully bind to the bead. According to these statistics, when λ_b=0.22 and N=2, the fractions of beads showing nonfluorescence, single-step photobleaching, and double-step photobleaching are calculated as P(0)+P(1)×(1−0.83)=69%, P(1)×0.83+P(2)×2×0.83×(1−0.83)=25%, and P(2)×0.83²+P(3)×3×0.83² ×(1−0.83)=4.6%, respectively. These values are in agreement with the observed values. Thus, the results show that GFP-tagged M6WT bound randomly to antibody-coated beads.

Based on the result that M6WT bound randomly to antibody-coated beads, we used Poisson statistics for actin-filament-binding probability 11,31 to evaluate whether the observed movement of beads was due to single M6WT molecules. The beads coated with antibodies for His-tag were mixed with GFP-tagged M6WT at various ratios of M6WT molecule to bead. The fraction of beads moving continuously (>50nm) when beads were brought into contact with actin filaments was recorded against various M6WT molecule/bead ratios (Figure 2C). The data were fit well to the probability function, 1-exp(−λ_mx), describing a bead carrying one or more active M6WT molecules, where x is the M6WT molecule/bead ratio in the mixture solution and λ_m is the fraction of myosin that bound to and moved the bead, 0.082 (Figure 2C). This result indicates that single monomeric molecules of M6WT are sufficient to move a bead processively. In this study, we observed the movement of M6WT at low M6WT/bead stoichiometries of 1:1 or 2:1. At the M6WT: bead ratio of 2:1, ∼6% of beads bound two M6WT molecules. However the percentage of beads that are moved by two monomeric M6WT molecules is calculated as 6%×(the fraction of active M6WT on the bead, λ_m/λ_b=37%)×q (<0.5)<1.1%, where q is the probability that two M6WT molecules (each 50nm in length) will randomly bind to one bead (200nm in diameter) and simultaneously interact with an actin filament, and is estimated to be <0.5 (refer to Svoboda and Block 31). The fraction of beads binding more than two M6WT molecules is negligibly small. Furthermore, the fraction of beads containing two M6WT bound to one antibody molecule when mixed at the stoichiometry of myosin/bead=2:1 is calculated as (the fraction of beads with M6WT bound, 30%)×(the probability for two M6WT binding to one antibody, <2.9%)<0.87%. Thus, these probabilities are too small to explain the fraction of beads moving, 15% (Figure 2C).

Figure 3A shows a typical trace of the time course for the stepping movement of M6WT tagged with a bead at no external load. The mean step size was 40nm. M6WT showed steps mostly in one direction, which we define as the forward direction (Figure 4A, top panel, black bars). M6WT moved processively even when the optical tweezers was switched on and force was exerted on the beads (Figure 3B). The mean step size decreased from 40 to 20nm at maximum load and was independent of ATP concentration ranging from 25μM to 4mM (Figure 4B). The number of backward steps increased with load and was equal to that of forward steps at maximum load (Figure 4A, panels 2–4). In contrast, the M6DH produced mostly forward steps (Figure 4A, gray bars) and the step size was almost constant, independent of the load (Figure 4B), consistent with a previous report 32. Figure 4C shows histograms of maximum forces for M6WT (black bars) and M6DH (gray bars). The mean maximum force of single-headed M6WT (1.0±0.02pN) was approximately one-half that of M6DH (1.9±0.04pN). We think that the force reaches maximum when the frequency of forward steps balances with that of backward steps. M6DH undergoes more stably forward steps at high load than single-headed M6WT because in the case of M6DH, the bead tethered to an actin filament by the forward head remains longer in the vicinity of the actin filament, allowing more time for the detached rear head to reach the forward actin target. Therefore, the M6DH could develop a maximum force twice as large as the single-headed M6WT. Figure 4D shows a histogram of run lengths of beads moved by single-headed M6WT at zero load. The mean run length (91nm) was ∼2.5-fold shorter than that of M6DH (∼220nm) 7,8. When the optical tweezers was switched on (trap stiffness; ∼8fN/nm), the mean run length was similar to that without load. Since the step size decreased with load, more steps were observed (Figure 3B). When we applied the load, the bead was trapped by optical tweezers and so the thermal motions of the bead were suppressed. Therefore, the myosin tagged with the bead could run a longer distance along an actin filament without diffusing away when the load was applied by the optical tweezers.

Display large version of this figure

Figure 3

Processive steps of cargo-carrying myosin-VI. Stepwise movements of single-headed M6WT with a bead at no load, when the optical trap was off (A), and at load, when the trap was on (B). In both panels, the upper trace shows displacement and the lower shows stiffness. In panel A, interaction occurred just before 0s, after which the optical trap was turned off. In panel B, the interaction started at 2s, after which stiffness increased. (Inset) Trace where trapped bead was stalled and pulled back to the trap center. (B) An expanded view of the boxed area in the inset. ATP concentration, 100μM. (Gray dots) raw data; (black line) same data passed though a low-pass filter of 20-Hz bandwidth.

Display large version of this figure

Figure 4

Mechanical property of M6WT and M6DH. (A) Histograms of step sizes at various loads. (Black bars) Single-headed M6WT; (gray bars) M6DH. Solid and open arrowheads show mean step size for single-headed M6WT and M6DH, respectively, at various loads as stated in upper right corners. (B) Mean step sizes at various loads. Symbols indicate data points (mean±SE) for M6WT in the presence of 4mM ATP (■), 100μM ATP (●), and 25μM ATP (▴), and for M6DH in the presence of 4mM ATP (□). (Solid line) Fitting line for mean step-size plots of M6WT. (Dashed line) Fitting line for mean step-size plots of M6DH. (C) Histograms of maximum force caused by M6WT (black bars) and M6DH (gray bars). (D) A histogram of run length at zero load. ATP concentration is 25–100μM.

To investigate the coupling between the ATP turnover and step, we analyzed dwell times of steps at low loads (<0.5pN) (Supplementary Material, Fig. S1 ). Stepping rate, defined as the inverse of dwell time, became smaller as ATP concentration decreased. Stepping rate versus ATP concentration fit well to Michaelis-Menten kinetics. The measured maximum stepping rate and Michaelis constant were 8.8s⁻¹ and 250μM, respectively, consistent with the value from biochemical measurements 29,33, suggesting that each step corresponds to a single cycle of the ATP hydrolysis.

It is thought that myosin-VI transports vesicles in cells for micron-order distance 21. The present results support the idea that the rate of diffusion of myosin-VI from actin filaments influences the processive movement of myosin-VI. This raises a notion that the run length of myosin-VI may be increased in viscous solution environment in cells in which the viscosity is ∼50 times larger than that of water 34. A dense actin meshwork structure is one of the major factors of a high viscosity in cells 35.

To examine the cargo transport by myosin-VI in such a viscous meshwork structure, we examined the movement of bead-bound myosin-VI in a high viscous solution containing methylcellulose (see Materials and Methods for detail), a high-polymer chain composed of trimethyl glucose forming a network structure 36.

At 0.5% (w/v) methylcellulose in the motility buffer, the 200-nm bead underwent slow Brownian motion (40 times slower than that in water) (see Materials and Methods). Figure 5A shows typical movement of a bead with GFP-tagged M6WT along an actin filament. Green spots indicate fluorescence of GFP-tagged M6WT on the bead (white arrow). Figure 5BC, shows time courses of bead movements along actin filaments (x direction) and deviations of the beads in the perpendicular (y) direction, respectively. Myosin-VI on beads underwent dissociation and association with the actin filaments and occasionally backward movement, but on average they moved in one direction. We did not observe such long travel distances in the absence of methylcellulose (Figure 4D). However, the mean travel distance was still hundreds of nanometers. The mean velocity was decreased to be ∼100nm/s compared to that without methylcellulose (40nm×8.8s⁻¹=352nm/s).

Display large version of this figure

Figure 5

Movement of a GFP-tagged M6WT with a bead bound in 0.5% methylcellulose solution. (A) Time course of the movement. The bead and an actin filament were labeled with red fluorescent dyes and M6WT was labeled with GFP. Both fluorescences were simultaneously observed by total internal reflection microscopy 27 equipped with double-view optics 52. In the figure, they were artificially colored red and green, respectively. The arrowhead indicates the center of the bead. Frame interval, 4s; scale bar, 1μm. (B) Typical time courses of the movement of beads by M6WT along actin filaments (x direction). Each colored trace indicates the movement of one bead by M6WT. (C) Time courses of the movement of beads in B perpendicular to the actin filaments (y direction). Colors indicate corresponding traces in panels B and C.

We showed that a single-headed M6WT with a bead bound moved processively along an actin filament with large steps. A critical issue is whether the observed processive movement is due to single monomeric M6WT molecules but not multiple or dimerized molecules bound to the bead (because the predicted coiled-coil element exists in the tail domain of M6WT). The results showed that 1), M6WTs were single-headed in electron micrographs; 2), fluorescence from GFP-tagged M6WT bound to a glass surface photobleached in a single step; 3), GFP-tagged M6WT bound to antibody randomly; 4), GFP-tagged M6WT randomly bound to antibody-coated beads; 5), the fraction of beads on which two GFP-tagged M6WT bound to the two binding sites of an antibody was negligibly small; and 6), the Poisson statistics for actin filament-binding probability fit to the probability function, 1−exp(−λ_mx) (see Results). These results provide sufficient evidence that the observed processive movement is indeed due to a single one-headed M6WT molecule with a bead bound. Furthermore, the mechanical properties of M6WT were different from those of myosin-VI engineered to form a dimer (M6DH): 1), the step size depended on the load, whereas that of M6DH was almost independent; 2), the fraction of backward steps at higher load is larger; and 3), the maximum force was approximately one-half that of M6DH.

The results show that a single one-headed M6WT moves processively along an actin filament with large steps upon binding a bead. How does such processive movement occur? The conventional model for an actomyosin motor postulates that the neck domain (calmodulin binding domain) (Figure 1A) of the myosin head acts as a lever arm, and movement is caused by swinging this lever arm 6,37,38. Myosin-VI, however, has no rigid lever arm long enough to move the bead 40nm in one step 39,40. Thus, neither the hand-over-hand mechanism nor the conventional lever arm swinging mechanism applies to the single-headed myosin-VI.

The myosin head is attached at its tail end to the bead via a flexible α-helix, ∼30nm long as estimated from the electron micrograph (Figure 1A). Because the myosin head is much smaller than the bead, it undergoes rapid Brownian motion relative to the bead. If we assume a myosin head diffuses 40nm at zero load, the time to diffuse this distance is estimated to be 14μs (see note in Supplementary Material ). The bead diffuses ∼5nm in this time (14μs). The time it takes for a myosin head to diffuse 20nm at maximum load (1.25pN) is estimated to be 3.5μs. The bead diffuses ∼2.7nm in this time even when it is pulled backward by a force of 1.25pN (see Supplementary Material ). The myosin head may not release Pi soon after the myosin head reaches the forward actin target and may be in a weak binding state for a while 29. However, the myosin head should undergo rapid association and dissociation with actin in the micriosecond time range in the weak binding state 41, so the myosin head tethered to the large bead would stay at the forward actin target until the Pi release. Therefore, the myosin head tethered to the bead via a long (∼30-nm) flexible α-helix could detach from actin and reattach to the next target zone without diffusing away from the actin.

Along with being a clathrin-coated vesicle transporter, myosin-VI is also thought to function as an anchor to stabilize membrane tension 42,43. These functions are determined by a tension sensor mechanism caused by strain-dependent regulation of the ATP and ADP binding rates to the myosin head 32. These rates are thought to determine rate-limiting for attachment to and detachment from actin 29.

Based on these facts, we propose processivity is caused by slower Brownian motion of the cargo and preferential landing of the myosin head to the actin filament (Fig. 6). The bead tethered to an actin filament via myosin-VI undergoes Brownian motion. When the bead diffuses toward the minus end of the actin filament (forward direction) a sufficient distance and the myosin head is pulled in the forward direction or the backward strain exerted on the head is relaxed, ATP binding is accelerated (Figure 6A), consistent with a previous report 32. When ATP binds to the head, the head dissociates from actin and undergoes rapid Brownian motion (Figure 6B). Then, the head preferentially lands to the next actin target in the forward direction (Figure 6C), and strongly binds to the actin, accompanied by isomerization of the myosin head (-ADP-Pi) or Pi release (Figure 6D), shifting the bead 40nm. The strain dependence of product release of myosin-VI is directionaly reversed compared to that of mysin-V 30, which means that the myosin-VI moves in the opposite direction.

Display large version of this figure

Figure 6

A possible mechanism for the processive movement of single-headed cargo carrying myosin-VI. (A) A cargo diffuses thermally against the load toward the (−) end of an actin filament (forward direction). Next, ATP binds to the head and dissociates it from actin. (B) The dissociated head undergoes rapid Brownian motion over the actin filament. (C) The head preferentially lands on the forward actin target before the bead can diffuse away. (D) The isomerization of the head or Pi release changes the conformation from a weak to strong bound state. The cargo is moved by 12nm by power stroke and/or the hopping of the head in this process (see text for details).

We have previously proposed a strain-sensor model 30,44 to describe the mechanism of preferential landing on the actin target by the myosin head. In this model, a portion of the myosin head, e.g., the neck domain or converter domain 40,45, acts as a strain sensor. This model assumes that when the backward strain is exerted on the head undergoing association and dissociation in the weakly binding state, the isomerization of the myosin or Pi release is accelerated, resulting in strong binding to actin. According to this model, if the dissociated head diffuses to the forward actin target and binds there, a backward strain should be exerted on the head and the head should associate tightly with actin, accompanied by isomerization of the myosin head or Pi release (Figure 6CD). The isomerization or Pi release may be accompanied by conformational changes (power stroke) in the head 6,37,38 and/or hopping of the head along actin momomers 46,47 to force the cargo forward (Figure 6D). This conformational change would promote forward preferential landing in the next stepping cycle 12,48. The compatibility between the directions of binding sites of myosin and actin may also be important for the preferential landing to the forward actin target 46.

To examine whether our mechanism is consistent with the actual mechanical features of myosin-VI, we analyzed the dwell times at various loads (Fig. 7) according to Kramer’s first-passage time 48,49,50. Taking into account that the force-generating step is fast and that ADP release is the rate-liming step at saturating ATP concentration 29, the mean dwell time is approximately given by

(1)

Display large version of this figure

Figure 7

Dwell times of steps at various loads. Solid squares indicate data points (mean±SE) in the presence of 4mM ATP (n=57–132). (See text for details.)

Our model is consistent with the 1957 Huxley model 51, in which random diffusive motion of the myosin head is rectified in one direction by an asymmetric binding affinity of the head for actin. Rock et al. 48 also explained the processive movement of M6DH based on the 1957 Huxley model (see Supplementary Material, Fig. S2 ). We tried to explain our data by their model in which the proximal tail acts as a spring with the constant of ∼0.25pN/nm and the power stroke size is 12nm. Based on the model of Rock et al., the time required for the proximal tail (spring) to diffuse forward to the next actin binding site when the bead was pulled back by the maximum force, 1.25pN (Figure 4C), was calculated to be ∼30μs (assuming one-dimensional Brownian motion of the head and a 13-nm stretch of the proximal tail). The bead would diffuse 11nm toward the backward direction in this time and the myosin head could not reach the forward actin target zone. Thus, it takes a long time to stretch the proximal tail (spring) for a sufficient distance, so we adopted an additional movement of the cargo in our model (Fig. 6).

In cells, myosin-VI transports endocytic vesicles in a dense actin meshwork structure. This meshwork structure causes high viscosity, ∼50 times higher than in water 34. Can our model be applied to the movement of myosin-VI with a cargo bound at such a high viscous environment? The high viscosity caused by a dense actin meshwork structure could affect the diffusion of the cargo but do little to the myosin head because the head is smaller than that of the meshwork structure 35. In such a high viscous condition, the cargo bound myosin-VI frequently detached from actin. But unlike in low viscosity, myosin reattached to the same actin filament due to the diffusion-limited slow Brownian motion of the cargo. This resulted in longer travel distances and slower transport velocity (∼100nm/s) caused by a detached phase. This slower velocity is comparable to that observed in cells 21. The viscous drag force would impose load on the bead, but the load should be negligible (F=−γv∼10fN, where γ is a drag coefficient and v is velocity) at the observed velocity (∼100nm/s). The results indicate that our model works successfully at high viscous environments in cells.

The results reported here indicate that single monomeric myosin-VI could transport large cargos such as intracellular vesicles in viscous environment in cells. However, it is likely that multiple molecules are present in single vesicles and this would further facilitate the processive transportation of the cargos (Supplementary Fig. S3 ). Alternatively, a certain population of myosin-VI assembles to form a dimer in cells, thus facilitating the processive movement of myosin-VI/cargo complex. However, it should be emphasized that the processive nature of myosin-VI/cargo complex observed in this study is the base to explain the physiological function of myosin-VI as a cargo transporter.