Search tips
Search criteria 


Logo of scirepAboutEditorial BoardFor AuthorsScientific Reports
Sci Rep. 2017; 7: 44237.
Published online 2017 March 13. doi:  10.1038/srep44237
PMCID: PMC5346999

Activated full-length myosin-X moves processively on filopodia with large steps toward diverse two-dimensional directions


Myosin-X, (Myo 10), is an unconventional myosin that transports the specific cargos to filopodial tips, and is associated with the mechanism underlying filopodia formation and extension. To clarify the innate motor characteristic, we studied the single molecule movement of a full-length myosin-X construct with leucine zipper at the C-terminal end of the tail (M10FullLZ) and the tail-truncated myosin-X without artificial dimerization motif (BAP-M101–979HMM). M10FullLZ localizes at the tip of filopodia like myosin-X full-length (M10Full). M10FullLZ moves on actin filaments in the presence of PI(3,4,5)P3, an activator of myosin-X. Single molecule motility analysis revealed that the step sizes of both M10FullLZ and BAP-M101–979HMM are widely distributed on single actin filaments that is consistent with electron microscopy observation. M10FullLZ moves on filopodial actin bundles of cells with a mean step size (~36 nm), similar to the step size on single actin filaments (~38 nm). Cartesian plot analysis revealed that M10FullLZ meandered on filopodial actin bundles to both x- and y- directions. These results suggest that the lever-arm of full-length myosin-X is flexible enough to processively steps on different actin filaments within the actin bundles of filopodia. This characteristic of myosin-X may facilitate actin filament convergence for filopodia production.

Myosin-X is a member of the myosin superfamily and found in vertebrates and in filasterea1,2. Myosin-X exhibits a striking localization at the tips of filopodia1,3,4,5,6,7 which suggests that myosin-X moves towards the filopodial tips. Moreover, myosin-X overexpression leads to a dramatic increase in the number and length of filopodia1, while knockdown of the expression of endogenous myosin-X by small interference RNA led to the loss of filopodia4,5. These findings suggest that myosin-X plays a critical role in filopodia formation.

Myosin-X is composed of a conserved motor domain in its N-terminal region, a neck region consisting of three IQ motifs that serve as light chain binding sites, a predicted coiled-coil domain and a unique tail domain8. It is now known that myosin-X dimerizes via an antiparallel coiled-coil motif at the distal region of the predicted coiled-coil domain9. The tail contains a PEST domain, three pleckstrin homology domains, a myosin tail homology 4 (MyTH4) domain and a band 4.1/ezrin/radixin/moesin (FERM) domain10.

The tail domain was reported to bind to and transport specific cargo molecules, such as VASP and ß-integrins3,11, although the transportation of ß-integrins are not directly shown. Therefore it was assumed that myosin-X is a processive motor which is suitable for cargo transporter. Supporting this view, it was found that myosin-X is a high duty ratio motor12. Processive movement of myosin-X with exogenous parallel forced dimer motifs on various actin structures has been controversial. The tail-truncated myosin-X with exogenous forced dimerization motif at the C-terminal end of the endogenous coiled-coil can processively move on fascin-actin bundles but is less processive on single actin filaments13,14. On the other hand, Sun et al.15 showed that the tail-truncated forced dimer of myosin-X moves on single actin filaments as well as actin bundles without notable difference in processivity. The reason for the difference is still unclear, although Vavra et al. claim that these constructs could be structurally different16. The step size of myosin-X is also controversial. Ricca and Rock17 showed that the step size of the tail-truncated forced dimer of myosin-X (1–920 amino acids) is 18 nm on both actin filaments and bundles, while Sun et al.15 reported that the tail-truncated forced dimer of myosin-X (1–939 amino acids) moves on actin bundles predominantly with 34 nm steps and minor 18 nm steps while the step size is 34 nm on single actin filaments. Bao et al. revealed that the tail-truncated forced dimer of myosin-X (1–940 amino acids) takes ~31 nm steps in both single and actin bundles14. Takagi et al.18 recently reported that tail-truncated forced dimer of myosin-X (1–936 amino acids) takes 35 nm forward steps on single actin filaments in single molecule optical tweezers experiments. One of the key clues to solve the contradiction may be the dimerization of myosin-X at distal coiled-coil region which is reported to make anti-parallel coiled-coil9.

A critical finding is that full-length myosin-X is a monomer having a folded conformation, in which the tail domain folds back to attach the motor-neck domain, although it has a coiled-coil domain between the IQ domain and the PEST domain19. This finding indicates that the coiled-coil of myosin-X is not strong enough to produce a stable dimer, which is consistent with the dissociation constant (Kd) of myosin-X dimer, ~0.6 μM9. A question is how myosin-X can be a cargo transporting motor if it is a monomeric myosin. Answering this question, it was found that phosphatidylinositol 3,4,5-trisphosphate (PI(3,4,5)P3) binding at the PH domain of the tail activates the motor activity and induces dimer formation19. These results suggest that full-length myosin-X becomes active upon binding PI(3,4,5)P3 and that the endogenous coiled-coil domain can form dimer although the stability of the dimer is weak.

It was found previously that an active myosin-X dimer without the cargo binding tail can induce actin filament convergence at the cell’s leading edge to initiate filopodia formation5. However, produced filopodia is very short and unstable, and it is thought that the tail domain is important for producing long and stable filopodia. The tail domain can interact with actin regulatory proteins such as VASP and β-integrin, and it is plausible that these actin-interacting proteins are involved in the myosin-X induced filopodia formation process3,11. Alternatively, the full-length myosin-X dimer might be more effective at converging actin filaments than the artificially forced dimer. These findings suggest that it is important to study the motor activity of the full-length myosin-X to clarify the innate motor characteristic of myosin-X.

In this study, we attempted to clarify the motor properties of full-length myosin-X. To facilitate the dimer formation at the endogenous coiled-coil domain, we introduced a leucine zipper at the C-terminal end of the tail. We also produced and studied a myosin-X construct without artificial coiled-coil to ensure the formation of dimer with innate coiled-coil. Both constructs moved processively on actin filaments with broad step size distribution with major step of 34–38 nm.


Processive movement of M10FullLZ on single actin filaments

We constructed recombinant bovine full-length myosin-X containing 1–2052 native myosin-X heavy chain (Fig. 1A, Supplementary Fig. 1A). We placed a GCN4 leucine zipper motif at the C-terminal end of the tail domain to combine the two heavy chains at the C-terminal end, thus facilitating the dimerization at the endogenous coiled-coil domain of myosin-X. We subcloned this recombinant myosin-X construct both in baculovirus and mammalian cell expression vectors. It has been known that myosin-X induces filopodia and localizes at the filopodial tips4,5. To check the functional authenticity, we examined whether this myosin-X construct facilitates filopodia formation in mammalian cells (Supplementary Fig. 2). The expression of the M10FullLZ in COS7 cells notably increases the number of filopodia and M10FullLZ showed distinct localization at the tip of filopodia, suggesting that M10FullLZ has an authentic function of myosin-X. To prepare the M10FullLZ protein, Sf9 cells were co-infected with the myosin-X heavy-chain-expressing virus with calmodulin-expressing virus. To activate M10FullLZ, PI(3,4,5)P3 diC8 was mixed with M10FullLZ. The multi-molecule in vitro motility assay showed that purified M10FullLZ moves Rhodamine-labeled actin filaments in PI(3,4,5)P3 dependent manner, similarly to how M10Full does19. Qdot was attached to the C-terminal end of an isolated M10FullLZ through anti-c-Myc antibodies as shown in Fig. 1B. From Poisson distribution, the ratio of M10FullLZ: Qdot of 1: 20 assures that ~97% of moving Qdot molecules have M10FullLZ single molecule.

Figure 1
Schematic drawing of the myosin-X construct used in this study.

We first observed the successive continuous movement of M10FullLZ on single actin filaments in the presence of 2 μM ATP using a TIRF microscope. M10FullLZ processively moves along single actin filaments as previously reported with the tail-truncated forced dimer of a myosin-X construct, in which the exogenous coiled-coil was introduced immediately after the endogenous coiled-coil domain15,17. By tracking the center position of myosin-X-Qdots using FIONA technique20, we determined the step sizes of M10FullLZ (Fig. 2A and B). The step size distribution of M10FullLZ was asymmetrical wide distribution with a notably long step size, in contrast to myosin Va HMM labeled with Qdots at the C-terminus, which showed a symmetrical step size distribution without long step size (Supplementary Fig. 3). The mean step size of M10FullLZ was 38.2 ± 17.5 nm (mean ± s.d., Nsteps/Qdots = 731/59) for forward step and −31.4 ± 14.4 nm (s.d., Nsteps/Qdots = 62/34) for backward step. The mean forward step was slightly longer than a half pitch of F-actin helix (~36 nm). It should be noted that back step was only single steps and we did not observe successive back steps. The obtained forward step size distribution was wide, and best explained with two different step sizes, i.e., 29.1 ± 10.9 nm (s.d.) and 55.3 ± 11.0 nm using the sum of two Gaussian distribution (R2 = 0.973). The former step size is consistent with the step sizes previously obtained for the tail-truncated forced dimer constructs15,17. The latter step size is close to 3/4 pitch of actin filaments. Since >50 nm steps are not seen in M101–939M5CC15 (Fig. 2B, inset) or myosin Va HMM (Supplementary Fig. 3), the latter step is considered to be unique to M10FullLZ. Note that while some >70 nm steps can be considered missing steps below the resolution of the measurement (estimated to be 3–6% in total steps), this cannot explain the observed 20–30% of the ~50 nm steps of M10Full LZ.

Figure 2
Processive movement of full-length myosin-X dimer (M10FullLZ) on single actin filaments at both low and high ATP concentrations.

Figure 2C is dwell time distribution of M10FullLZ at 2 μM ATP. The average waiting time is 0.49 ± 0.05 s (s.e.m., 0.45 ± 0.01 s from Kaplan-Meier survival curve, see also Supplementary Fig. 5Aa), which is consistent with the M101–939M5CC value (0.89 s at 1 μM MgATP15). The average waiting time gives the single turnover rate of ~2 s−1 at 2 μM ATP, which is 1/10–1/2 value of the Vmax of myosin-X (4–20 s−1)12,19,21. Run length distribution of M10FullLZ gave the average run length of 1.04 ± 0.08 μm (s.e.m.) at 2 μM ATP (Fig. 2D), which is similar to M101–939M5CC (~1 μm15), but much longer than M101–920LZ (~0.2 μm13,22). To test whether M10FullLZ moves processively on single actin filaments in physiological ATP concentration, we observed the movement at 2 mM ATP (Fig. 2E and F). M10FullLZ also processively moves at 2 mM ATP with average run length of 0.78 ± 0.03 μm (s.e.m.), slightly shorter than that in 2 μM ATP. This indicates that M10FullLZ moves 20–25 steps on average at the physiological concentration of ATP on single actin filaments. The velocity of the movement at 2 mM ATP is 0.33 ± 0.01 μm/s. This means that M10FullLZ takes 10 steps per second, i.e., ~10 s−1, which is much larger than the result of Vmax value of steady-state actin-activated ATPase activity of full-length myosin-X reported previously19, but similar to the Vmax of myosin-X S112. This may be because the motility assay measures only activated myosin-X and it is known that full-length myosin-X activity is regulated19.

Processive movement of M101–979 without LZ on single actin filaments

To examine the large steps derived from myosin-X dimer derived from innate coiled-coil, we observed the step size distribution of N-terminal Q-dot-tagged myosin-X HMM without LZ (BAP-M101–979HMM) (Fig. 3). Sf9 cells were co-infected with the myosin-X heavy chain-expressing virus and with the calmodulin-expressing virus as was done for M10FullLZ. Streptavidin-Qdot was attached to the N-terminal end of isolated M101–979HMM through BAP sequence to avoid the possible disruption of dimer formation via endogenous coiled-coil domain of myosin-X. Note that this construct provides the step size of each head, therefore the obtained step size is expected to be double of the construct with Q-dot at the C-terminal end. Using a TIRF microscope, we observed the successive continuous movement of BAP-M101–979 HMM on single actin filaments upon addition of 2 μM ATP to acto-BAP-M101–979HMM. The result suggests that BAP-M101–979HMM can form a dimer with endogenous coiled-coil domain. It was previously observed for myosin VI and myosin-X that the endogenous coiled-coil domain can form a dimer when the constructs associate with F-actin in the absence of ATP to facilitate a dimer formation15,23. The average waiting time (2τ) of BAP-M101–979HMM at 2 μM ATP was 0.70 ± 0.05 s (s.e.m., n = 84) (Fig. 3E), which yields the calculated velocity of ~0.1 μm/s. This is consistent with the actual velocity at 2 μM ATP (0.106 ± 0.003 μm (s.e.m., n = 93)). The waiting time of BAP-M101–979HMM was similar to, but slightly faster than that of M10FullLZ (τ = 0.35 s for BAP-M101–979HMM vs. 0.49 s for M10FullLZ at 2 μM ATP). The step size distribution of BAP-M101–979HMM was asymmetrical wide distribution with notably long step sizes (Fig. 3C and D) in contrast to BAP-M5HMM labeled with Qdots at the N-terminus, which showed a symmetrical step size distribution without long step size (Fig. 3B). The mean step size of BAP-M101–979HMM was 67 ± 19 nm (mean ± s.d., n = 210). This result indicates that each myosin-X head takes a ~34 nm step, and the wide distribution also supports the view that the myosin-X dimer with the innate coiled-coil could take the large >50 nm step, although the most of them seem to be from the split dimer, in which the two heavy chains are connected at the C-terminal end of the molecule with LZ motif.

Figure 3
Step size distributions of N-terminal Qdot-labeled myosin Va and native myosin-X HMM constructs.

Movement of M10FullLZ on filopodia

As mentioned above, myosin-X travels along filopodia in cells4,5. To study the movement of full-length myosin-X in physiological set-up, we examined the movement of M10FullLZ in filopodia of demembranated MEF3T3 cells. MEF3T3 cells were transfected with cdc42 expression vector to induce filopodia (Supplementary Fig. 4A). The cells were then treated with Triton-X100 to remove surface membrane, and stained with Alexa phalloidin as described in Methods. M10FullLZ -Qdot was then applied on the demembranated cells. After being washed with the blocking buffer in PBS, the motility buffer containing 2 μM ATP or 2 mM ATP was applied to the cells (Supplementary Fig. 4B). M10FullLZ-Qdot primarily moved on filopodia and we could not find them moving on other cytoskeletal structures (Supplementary Fig. 4B and Supplementary Movie 1). The FIONA analysis at 2 μM ATP revealed the successive multiple stepping of M10FullLZ in filopodia (Fig. 4A). The mean step size of forward and backward steps of x-projection were 35.8 ± 16.5 nm (mean ± s.d., Nsteps/Qdots/filopodia/cells = 660/90/70/35) and −23.9 ± 10.8 nm (s.d., Nsteps/Qdots/filopodia/cells = 56/33/28/21), respectively (Fig. 4B). The mean forward step size was smaller than those found in in vitro single actin filaments (Fig. 2B), which is likely to be due to the presence of the y-component of the stepping during the movement in filopodia (see Fig. 5). The step size distribution was asymmetrical and broad unlike the tail-truncated forced dimer construct of myosin-X (Fig. 4B inset), and was best explained with two different step sizes, i.e., ~27 nm and ~50 nm. Dwell time distribution in filopodia showed the average waiting time (τ) of 0.89 ± 0.06 s (s.e.m.), which is twice longer than that on single actin filaments at 2 μM ATP (Fig. 4C). This indicates that it takes a longer time to find the landing point on filopodia than on single actin filaments. The movement of M10FullLZ on filopodia was also tested at 2 mM ATP (Fig. 4D). The average run length was 0.77 ± 0.06 μm (s.e.m., 0.691 ± 0.006 μm from survival curve, see also Supplementary Fig. 5Bb), which is similar to that at 2 mM ATP on single actin filaments (Fig. 2E). The mean velocity of M10FullLZ on filopodia is 0.317 ± 0.002 μm/s (s.e.m., Fig. 4E), similar to the velocity on single actin filaments (Fig. 2F).

Figure 4
Processive movement of full-length myosin-X dimer on filopodia.
Figure 5
Two-dimensional stepping analysis of full-length myosin-X dimer on filopodia at 2 mM ATP.

Since filopodia is composed of F-actin filament bundles, we studied two-dimensional movement of M10FullLZ at 2 mM ATP (Supplementary Fig. 6). Figure 5 shows a Cartesian plot analysis of M10FullLZ movements (Nsteps/Qdots/filopodia/cells = 205/13/12/9). While M10FullLZ primarily steps towards the tip of filopodia (x-step size), the displacement to vertical or diagonal direction is also observed (Fig. 5A). For y-steps, 34 positive steps and 22 negative steps were observed (Fig. 5B). The mean step sizes of forward x-steps were 36.1 ± 1.5 nm (s.e.m., Fig. 5C). This value is similar to a half-pitch of F-actin helix (~36 nm) and ~34 nm step size of BAP-M101–979HMM on single actin filaments (Fig. 3B), although large step sizes (>50 nm) were observed. On the other hand, the mean step sizes of positive y-steps and negative y-steps were 26.2 ± 1.4 nm (s.e.m.), and −22.7 ± 1.2 nm (s.e.m.). These values coincide with the space between the two actin filaments ~24 nm apart (2 × 11–13 nm space between the adjacent filaments in fascin bundles24) from each actin (see Fig. 6Ab) in fascin-based actin bundles. The presence of y-steps in part explains the decrease in the overall mean step size of M10FullLZ on filopodia. We also found a significant number of diagonal displacements (Fig. 5D and E). The values in Fig. 5D and E are the x- and y- projection of the diagonal steps, respectively. These results suggest that M10FullLZ moves various direction in filopodia, which is consistent with the flexible nature of the myosin-X motor15. It should be noted that M10FullLZ continuously moved after y-stepping or diagonal stepping towards filopodial tips. Sometimes we observed successive multiple y-stepping and diagonal stepping prior to x-stepping towards filopodial tips.

Figure 6
Model of stepping pattern of myosin-X on filopodia.

Visualization of M10FullLZ molecules by electron microscopy

Electron microscopy was employed to clarify the molecular behaviors of the M10FullLZ molecules in the presence of the actin. Supplementary Figure 7A shows the general appearances of M10FullLZ molecules in single molecular state. According to appeared inter-head distances of two-headed molecules, there are two main conformations of M10FullLZ molecules visualized in the fields: shorter and longer lengths of the inter-head conformers, (inset image in Supplementary Fig. 7A). Similar observations of the inter-head distances of the single molecules were found in the presence of the actin (Supplementary Fig. 7B) and the inter-head lengths from the two main appearances of the molecules binding to actin filament were measured in the images (inset image in Supplementary Fig. 7B). Measured mean ± s.d. lengths of shorter and longer inter-head distances in the molecules are 35.7 ± 0.8 nm (21 particles) and 53.9 ± 1.4 nm (10 particles). Taken together with the step length measurements as shown in Fig. 5, this result demonstrates that M10FullLZ molecules can form the shorter and longer lengths of the inter-head conformers when interacted with the actin filaments.


Our aim of this study is to clarify the authentic nature of the movement of myosin-X on actin filaments. To address this question, we generated a full-length myosin-X construct and analyzed its movement at a single molecule level on both single actin filaments and filopodial actin bundles in cells. Since a full-length myosin-X is monomeric and inhibited19, we studied the motility in the presence of PI(3,4,5)P3, which activates the full-length myosin-X19. As we reported previously, a full-length myosin-X processively moves within filopodia in living cells at singe molecule level6,25,26. The result suggests that a full-length myosin-X forms a dimer in cells. However, the concentration of overexpressed myosin-X in living cells is much higher than that of in vitro single molecule experiment, in which pM concentration of myosin-X is used. In order to facilitate inter-heavy chain interaction at the coiled-coil domain of myosin-X, we introduced a LZ at the C-terminal end of the tail. The produced full-length myosin-X showed processive movement on actin filaments, however, unlike myosin Va, the step size distribution is broad and asymmetric for both single actin filaments and filopodia. Our result suggests that this is because there are multiple step sizes of myosin-X moving on actin filaments. The step size distribution was best explained with multiple step sizes. Electron microscopy revealed that M10FullLZ exhibits two configurations of different inter-head distance (Supplementary Fig. 7), which correspond to the multiple step sizes found in the present study.

An important novel point in this study is the observation of two-dimensional movement of single myosin-X molecules on filopodia (Fig. 5). The step size distribution along x-axis (mean = ~36 nm) was quite similar to the distributions on single actin filaments (Fig. 2) and on filopodia (Fig. 4). The distribution of y-step was broad, and the mean y-step was close to ~24 nm. If myosin-X steps along the y-axis at ~24 nm, the myosin-X leading head at the first step will land about 5 actin filaments apart from the trailing head (Fig. 6Ab). The subsequent ~24 nm y-step can be achieved by the inchworm stepping of the trailing head. Alternatively, ~24 nm y-steps can be due to the hand-over-hand stepping on every other actin filaments in the bundles. From the analysis of diagonal steps, we found that the y-step components were relatively constant (~24 nm), irrespective of the x-step sizes (Supplementary Table 1). Taken these findings together, the results suggest that the myosin-X leading head prefers to step on the fifth/or third actin filaments apart from the trailing head in filopodia.

Of interest is the production of >50 nm x-steps. The wide stepping distribution is also observed in artificially elongated myosin Va HMM27. The result suggests that the landing position of the myosin head can be the side of actin filaments. This suggests that the lever-arm of full-length myosin-X is flexible enough to step on various positions on actin filaments. We asked whether the >50 nm steps occur during the course of multiple stepping or it only occurs at the final step. The most >50 nm steps are observed between normal steps. As shown in Fig. 2A, the successive multiple >50 nm steps were also observed. The results suggest that the >50 nm steps are mechanically active steps. It is reported that the myosin-X lever arm with 3 IQ motifs is extending about 10.8 nm on F-actin filament15. Assuming that the postulated SAH domain of myosin-X is ~15 nm28, the entire lever arm of 2 myosin-X heads is expanded by about 51.6 nm when myosin-X is dimerized. When myosin-X forms an antiparallel dimer, an additional 4.7 nm is incorporated to the span between the two heads. Therefore, it is plausible that >50 nm steps are arisen from the myosin-X antiparallel dimer (Fig. 6Aa).

Several studies of single molecule analysis of a tail-truncated forced dimer of myosin-X have been reported, and the step size of myosin-X is controversial14,15,17,18,22. We previously reported15 that myosin-X step size is 34 nm on single actin filaments and 18 nm and 34 nm on fascin-actin bundles, while Nagy et al.22 and Ricca and Rock17 reported ~18 nm step size of myosin-X for both single actin filaments and fascin-actin bundles, but not 34 nm step size, and claimed that the difference is due to in or out of register with the heptad repeat of intrinsic coiled-coil and the connecting artificial coiled-coil. Their construct has 1–920 amino acids of native myosin-X with a leucine zipper motif right after the myosin-X’s coiled-coil with in-register while our HMM construct has 1–939 amino acids of native myosin-X, followed by myosin V’s coiled-coil. While the high score of coiled-coil probability ends around 920, we think that the forced dimer motif should be far from the intrinsic coiled-coil to avoid possible disruption of the intrinsic coiled-coil conformation of myosin-X. Lu et al.9 previously reported that myosin-X’s coiled-coil makes an anti-parallel dimer, and may alter the dimer structure when a forced dimer is introduced right after myosin-X’s intrinsic coiled-coil. Takagi et al.18 recently reported that the single molecule myosin-X 1–936 HMM followed by leucine zipper showed a ~17 nm lever arm displacement, which cause a ~35 nm forward step size in an optical tweezers assay. This study agrees well with our previous study15, and suggests that at least 16 amino acids sequence after 920th of myosin-X may affect the step size of myosin-X. One of the best approaches to circumvent this problem and address the innate motor property of myosin-X is the use of full-length myosin-X. The present study shows that a full-length myosin-X moves predominantly with 38 nm step size (Fig. 2B). Another approach is the use of myosin-X with only the innate myosin-X coiled-coil. Our study also shows that myosin-X HMM without an artificial dimer structure moves with ~34 nm, which is a half size of the 67 nm observed in single head-labeled BAP-M101–979HMM (Fig. 3C). These results indicate that the authentic step size of myosin-X is predominantly 34–38 nm.

It should be noted while our results do not show the apparent peak of ~18 nm step size in the step size distribution analysis, the data can be explained with the presence of 18 nm step size in addition to 36 nm step size and 54 nm step size (Fig. 6B). The mechanism for the ~18 nm step size of myosin-X is obscure, but hand-over-hand stepping is less likely on single actin filaments considering the long length of the lever-arm. It is rather plausible that ~18 nm steps are arisen from inchworm-like stepping, as seen in full-length myosin Va in EGTA condition29.

M10FullLZ increased the number of filopodia (Supplementary Fig. 2), and the filopodia length is much longer than that induced by the tail-truncated myosin-X forced dimer (data not shown). Furthermore, the effects of M10FullLZ on filopodia formation and elongation are similar to those of wild-type full-length myosin-X (data not shown), suggesting that M10FullLZ retains the authentic properties of myosin-X. The full-length myosin-X shows a ~38 nm mean step size on single actin filaments, and ~36 nm on filopodia. A slight decrease in the mean step size in filopodia is in part due to the presence of y steps (vertical to actin filaments). This is consistent with our previous report15, where in vitro fascin-actin bundles were used for the step size analysis. In order to study the movement of myosin-X in near physiological environment, we examined the myosin-X movement in filopodia using a demembranated cell system30,31. Our results show that the full-length myosin-X produces a significant number of y-directional stepping in addition to x-directional (towards filopodial tips) stepping in filopodia. These steppings are followed by x-directional, y-directional or diagonal stepping, indicating that these steppings represent actual movement of myosin-X. This finding suggests that myosin-X is a motor that can step to various angles in filopodia, which enables myosin-X to bridge different actin filaments in cells. This property is likely to be in part due to the flexibility of the lever-arm of myosin-X15, and it is plausible that this nature enables myosin-X to converge actin filaments to initiate filopodia. Schematic drawing of the possible conformations of M10FullLZ is shown in Fig. 6A, and the over-all step size distribution of M10FullLZ including y-stepping on filopodia is shown in Fig. 6B. The mean step size of myosin-X in all directions was calculated to be 36.6 ± 16.7 nm (s.d., n = 205). The step size distribution is widely distributed, which is likely to be due to the presence of multiple step sizes including 34–36 nm stepping14,15,18,32, 24 nm y-steppingThis study, >50 nm large steppingThis study and possibly ~18 nm stepping15,17,22. We think that the large >50 nm step size can be in part arisen from an antiparallel dimer of myosin-X with innate coiled-coil. The result of EM observation (Supplementary Fig. 7) is consistent with this view. However, while the step size observation of BAP-M101–979HMM without LZ shows wide step size distribution and notably more fractions of >90 nm size than that of BAP-M5HMM (Fig. 3B), the fraction of large step sizes is less than that of M10FullLZ. This may indicate that an open conformation of the molecule (see Fig. 6Aa) predominantly contributes to the >50 nm large steps.

During the preparation of this paper, Ropers et al.32 reported that full-length myosin-X moves on single actin filaments with ~36 nm step size, and on synthetic fascin-actin bundles with 19 nm, 38 nm and 52 nm steps. Consistent with the present work, full-length myosin-X showed the broad step size distribution on single actin filaments. However, the step size distribution on fascin-actin bundles is somewhat different from our data using filopodial actin. Moreover, the velocity of their full-length myosin-X is twice faster and the run-length is 2.4-fold longer in fascin-actin bundles, compared with those on single actin filaments. In this study, we could not find any substantial differences between the movements on single actin filaments and on filopodia. Although the reason of apparent difference remains to be elucidated, it is plausible that these differences are due to the difference of actin structures between the reconstituted fascin-actin bundles and filopodia.

By using synthetic fascin-actin bundles, it was reported that the run length of a tail-truncated forced dimer of myosin-X HMM on fascin-actin bundles is much longer than that on single actin filaments14,22,32. In this study, however, the mean run length of full-length myosin-X in filopodia (0.77 μm at 2 mM ATP, Fig. 4D) is similar to that on single actin filaments (0.78 μm at 2 mM ATP, Fig. 2E). This is consistent with the results of Sun et al.15, who reported that run length of myosin-X HMM is 0.95 μm on single actin filaments and 1.16 μm on fascin-actin bundles, respectively15. The apparent difference in the results can be in part due to the difference in the actin structure and it is plausible that the fascin-induced synthetic actin bundles are different from filopodial actin bundles presumably due to the difference in the structure with additional proteins in filopodial actin structure. Further studies are required to clarify the effect of additional filopodial components on intra-filopodial myosin movement.

Materials and Methods

Rhodamine-phalloidin and streptavidin conjugated Qdots were purchased from Invitrogen. Actin was purified from rabbit skeletal muscle according to Spudich et al.33.

Bovine myosin-X cDNA fragments were kindly provided by Dr. D. P. Corey (Harvard University). Baculovirus expressing recombinant full-length bovine myosin-X (M10Full) (including 1–2052, amino acids) was made as previously described19. The stop codon of M10Full was removed and converted to Mlu 1 site by PCR site-directed mutagenesis34 using Pfu Ultra II (Agilent Technologies), and an oligo DNA containing c-Myc sequence (amino acid: EQKLISEEDL) was inserted between Mlu1 sites. A GCN4 leucine zipper sequence35 after a linker motif sequence (amino acid: GGGSGGGSGGGS) was then inserted before the c-Myc tag motif. The full-length myosin-X with leucine zipper (M10FullLZ) was subcloned into a modified pFastBacHT baculovirus transfer vector (Life technologies) having a FLAG sequence between His tag and Tev protease site. For experiments of myosin-X HMM, we constructed myosin-X HMM (1–939 residues) with myosin Va coiled-coil5,15, and inserted biotin acceptor peptide (BAP) sequence (GGGLNDIFEAQKIEWHE) at the C-terminal ends. Myosin-X native HMM construct (BAP-M101–979HMM) was also made by inserting BAP sequence instead of GFP at N-terminal side of myosin-X HMM15. For myosin Va experiments, we produced mouse myosin Va HMM (amino acids, 1–1105) with BAP sequence at the C- or N-terminal sides. M10 proteins were expressed in Sf9 cells by co-infection of myosin-X heavy chain with CaM viruses, and prepared using anti-FLAG antibodies affinity column chromatography as described36.

Labeling of myosin-X and F-actin

For M10FullLZ experiments, either Qdot 525 (Q11022MP, Invitrogen) or Qdot 655 (Q11041MP) conjugated with goat F(ab’)2 anti-mouse IgG (H + L) were mixed with anti-c-Myc monoclonal antibodies (Clontech) at the Qdot/antibody ratio of 0.85: 1, and incubated at 4 °C for >1 day. M10FullLZ was then labeled with the Qdot–antibody by mixing and incubating them on ice for >1 hour before experiments. Biotinylated M101–939M5CC was prepared and mixed with streptavidin–conjugated Qdot 525 (Q110141MP, Invitrogen) or Qdot 655 (Q10121MP) as described previously15. To ensure single molecule conditions, the myosin-X/Qdot ratio was usually adjusted to 1: 20 unless otherwise indicated. Under this condition, it is calculated from Poisson statistics that ~9% and ~90% of the myosin-X would be occupied with one and two Qdots at the C-terminus, respectively. We didn’t distinguish one or two Qdots/myosin-X in the present study. Instead, we didn’t adopt the data from extra bright fluorophores, which is probably composed of 3 or more Qdots from multi-heads of myosin-X. For M10FullLZ-Qdot 655 experiments, we adopted myosin-X/Qdot ratio of 1:5 by comparing first-power landing rates between myosin-X-Qdot 525 and myosin-X-Qdot 655 to maintain single molecule conditions as previously reported17. The proximity-induced dimerization of BAP-M101–979HMM (myosin-X HMM with native coiled-coil) was carried out according to Sun et al.15. The dimerized BAP-M101–979HMM was labeled with streptavidin-conjugated Qdot 655 with 1:4 ratio. F-actin was sparsely labeled with rhodamine-phalloidin as previously described36. For myosin Va experiments, myosin Va HMM biotinylated at C-terminus and streptavidin–conjugated Qdot 655 were mixed at the ratio of 1:20 before observations. The myosin Va HMM biotinylated at N-terminus was mixed with streptavidin-conjugated Qdot 655 at 1:4 ratio as described37.

Cell preparation

For preparation of extracted cells, MEF3T3 cells were treated with extraction buffer as described previously31 with slight modifications. In brief, the cells were treated with extraction buffer (30 mM imidazole, pH 7.5, 70 mM KCl, 1 mM EGTA, 2 mM MgCl2, 0.5% Triton X-100, 4% polyethylene glycol (mol wt 8,000), and 250 nM Alexa Fluor 568 phalloidin (invitrogen)) for 4 min on ice and then washed twice with ice cold PBS. All of the samples were incubated on ice with blocking solution (1% BSA, 0.2 mg/ml streptavidin and 1 mg/ml casein in PBS) before use in single molecule assays.

Single molecule TIRF motility assay

The movement of myosin-X was observed with a home-built TIRF setup using an inverted Olympus IX71 microscope, an objective lenz (PlanApo-N × 60, 1.45 numerical aperture, oil), 473 nm diode–pumped solid state laser (model BL-9040, PHOTOP) or 488 nm from an Argon/Krypton ion laser (model 35 LDL 840-208, Melles-Griot), and a back-illuminated electron multiplier CCD camera (model iXON DU-897E or DU-860E, Andor Technology). For the experiments of single actin filaments, flow chambers were prepared by using No. 1.5 glass coverslips and glass slides (Fisher Scientific). Alpha-actinin (A9776, SIGMA) was used to immobilize F-actin, and casein from bovine milk (07319-82, Nakarai Tesque, Japan) and/or streptavidin (Thermo Scientific) was used for glass-surface blocking. The movement of myosin-X was observed in a solution containing 25 mM KCl, 20 mM HEPES (pH 7.5), 5 mM MgCl2, 1 mM EGTA, 5 mM DTT, 12 μM calmodulin, and an O2 scavenging system containing glucose oxydase, catalase and glucose in the presence of indicated concentration of ATP. For M10FullLZ experiments, 5 μM PI(3,4,5)P3 diC8 (P-3908, Echeron) was added to the assay solution to activate myosin-X motor activity. The PI(3,4,5)P3 diC8 does not make micelles at 5 μM (CMC >3 mM). All the experiments were done at 24 °C.

DATA analysis

The captured images were analyzed by using in-house 2D Gaussian fitting software6. If necessary, a function for the recorded stage drift was calculated and subtracted from the raw data of each runs for each movie. Run length and velocity of myosin-X were concurrently determined by measuring the length and time between on and off the tracks. Step size and dwell time were determined by using a step-fitting software based on an algorithm described in Kerssemakers et al.38. The average number of frames in each step was 16.3 for M10FullLZ (2 μM ATP) on single actin filaments, and the typical position of stationary myosin-X-Qdot on actin filaments was 0.0 ± 2.5 nm (mean ± s.e.m., ς = 10.2), indicating that the tracking precision was <3 nm under the typical experimental condition. For the step size analysis in the presence of 2 mM ATP, we chose myosin-X molecules with a slow speed (0.1–0.2 μm/s). For the Cartesian coordinate plot, the x step size and y step size were independently determined and plotted to the quadrant planes as described previously17.

In vitro multi-molecule assay

In vitro motility assay was carried out according to the method previously described using IX71 microscope with a motorized filter wheel (ProScan II, Prior Scientific) and a CCD camera (Sensicam, PCO/Cooke)19,36.

Electron Microscopy

For negative staining, M10FullLZ (300 nM) was mixed with 6 μM of calmodulin (ratio of 1 to 20 to enhance calmodulin binding) and then 10-fold diluted with low salt buffer containing 50 mM Na acetate, 1 mM EGTA (1.1 mM CaCl2), 2 mM MgCl2, 10 mM MOPS, 200 μM ATP, and pH7.5. To visualize M10FullLZ molecules attached to F-actin, specimens were made by mixing 300 nM of M10FullLZ molecules and 6 μM of calmodulin with an equal volume of 10 μM of F-actin, then 10-fold diluted with above low salt buffer. After dilution, 5 μl of the final mixture were applied to a carbon-coated grid that had been glow-discharged (Harrick Plasma, Ithaca, NY) for 3 min in air, and the grid was immediately (~5 sec) negatively stained using 1% uranyl acetate39. Grids were examined in a Technai 10 TEM (FEI, USA) operated at 100 keV and images were recorded at a magnification of 34,000 (0.32 nm per pixel).

Confocal light microscopy

Localization of myosin-X in COS7 cells was observed by using a laser scanning confocal microscope, SP2 system (Leica Microsystems, Heidelberg, Germany).

Additional Information

How to cite this article: Sato, O. et al. Activated full-length myosin-X moves processively on filopodia with large steps toward diverse two-dimensional directions. Sci. Rep. 7, 44237; doi: 10.1038/srep44237 (2017).

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Material

Supplementary Information:
Supplementary Movie 1:


Step fitting software made by Mr. Tanaka and Dr. Mizutani (Hokkaido University, Japan) and kindly provided by them. We thank Dr. Yale Goldman for giving critical advice for the manuscript. This work is supported by the TEXAS STARS PLUS award and Korea Basic Science Institute Grant. The EM research was supported by the Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Science, ICT & Future Planning (NRF-2015R1C1A1A01053611), the framework of international cooperation program managed by National Research Foundation of Korea (NRF-2016K2A9A1A01951927) and the 2015 Research Grant from Kangwon National University.


The authors declare no competing financial interests.

Author Contributions O.S. performed single molecule experiments and analyses; H.-S.J. performed electron microscopic analysis; S.K. performed cell biological experiments; Y.T. & K.H. helped to set up TIRF microscope; T.M.W. developed 2D Gaussian fitting software; M.I. supervised the project; O.S. and M.I. conceived the study, designed the experiments, and wrote the manuscript with input from the other authors.


  • Berg J. S. & Cheney R. E. Myosin-X is an unconventional myosin that undergoes intrafilopodial motility. Nat. Cell Biol. 4, 246–250, doi: (2002).10.1038/ncb762 [PubMed] [Cross Ref]
  • Petersen K. J. et al. . MyTH4-FERM myosins have an ancient and conserved role in filopod formation. Proc. Natl. Acad. Sci. USA 113, E8059–E8068, doi: (2016).10.1073/pnas.1615392113 [PubMed] [Cross Ref]
  • Tokuo H. & Ikebe M. Myosin X transports Mena/VASP to the tip of filopodia. Biochem. Biophys. Res. Commun. 319, 214–220, doi: (2004).10.1016/j.bbrc.2004.04.167 [PubMed] [Cross Ref]
  • Bohil A. B., Robertson B. W. & Cheney R. E. Myosin-X is a molecular motor that functions in filopodia formation. Proc. Natl. Acad. Sci. USA 103, 12411–12416, doi: (2006).10.1073/pnas.0602443103 [PubMed] [Cross Ref]
  • Tokuo H., Mabuchi K. & Ikebe M. The motor activity of myosin-X promotes actin fiber convergence at the cell periphery to initiate filopodia formation. J. Cell Biol. 179, 229–238 (2007). [PMC free article] [PubMed]
  • Watanabe T. M., Tokuo H., Gonda K., Higuchi H. & Ikebe M. Myosin-X induces filopodia by multiple elongation mechanism. J. Biol. Chem. 285, 19605–19614 (2010). [PMC free article] [PubMed]
  • Kerber M. L. & Cheney R. E. Myosin-X: a MyTH-FERM myosin at the tips of filopodia. J. Cell Sci. 124, 3733–3741, doi: (2011).10.1242/jcs.023549 [PubMed] [Cross Ref]
  • Sousa A. D. & Cheney R. E. Myosin-X: a molecular motor at the cell’s fingertips. Trends Cell Biol. 15, 533–539, doi: (2005).10.1016/j.tcb.2005.08.006 [PubMed] [Cross Ref]
  • Lu Q., Ye F., Wei Z., Wen Z. & Zhang M. Antiparallel coiled-coil-mediated dimerization of myosin X. Proc. Natl. Acad. Sci. USA 109, 17388–17393, doi: (2012).10.1073/pnas.1208642109 [PubMed] [Cross Ref]
  • Berg J. S., Derfler B. H., Pennisi C. M., Corey D. P. & Cheney R. E. Myosin-X, a novel myosin with pleckstrin homology domains, associates with regions of dynamic actin. J. Cell Sci. 113, 3439–3451 (2000). [PubMed]
  • Zhang H. et al. . Myosin-X provides a motor-based link between integrins and the cytoskeleton. Nat. Cell Biol. 6, 523–531, doi: (2004).10.1038/ncb1136 [PubMed] [Cross Ref]
  • Homma K. & Ikebe M. Myosin X is a high duty ratio motor. J. Biol. Chem. 280, 29381–29391 (2005). [PubMed]
  • Nagy S. et al. . A myosin motor that selects bundled actin for motility. Proc. Natl. Acad. Sci. USA 105, 9616–9620 (2008). [PubMed]
  • Bao J., Huck D., Gunther L. K., Sellers J. R. & Sakamoto T. Actin structure-dependent stepping of myosin 5a and 10 during processive movement. PLoS One 8, e74936, doi: (2013).10.1371/journal.pone.0074936 [PMC free article] [PubMed] [Cross Ref]
  • Sun Y. et al. . Single-molecule stepping and structural dynamics of myosin X. Nat. Struct. Mol. Biol. 17, 485–491 (2010). [PMC free article] [PubMed]
  • Vavra K. C., Xia Y. & Rock R. S. Competition between Coiled-Coil Structures and the Impact on Myosin-10 Bundle Selection. Biophys. J. 110, 2517–2527, doi: (2016).10.1016/j.bpj.2016.04.048 [PubMed] [Cross Ref]
  • Ricca B. L. & Rock R. S. The stepping pattern of myosin X is adapted for processive motility on bundled actin. Biophys. J. 99, 1818–1826 (2010). [PubMed]
  • Takagi Y. et al. . Myosin-10 produces its power-stroke in two phases and moves processively along a single actin filament under low load. Proc. Natl. Acad. Sci. USA 111, E1833–1842, doi: (2014).10.1073/pnas.1320122111 [PubMed] [Cross Ref]
  • Umeki N. et al. . Phospholipid-dependent regulation of the motor activity of myosin X. Nat. Struct. Mol. Biol. 18, 783–788 (2011). [PubMed]
  • Yildiz A. et al. . Myosin V walks hand-over-hand: single fluorophore imaging with 1.5-nm localization. Science 300, 2061–2065 (2003). [PubMed]
  • Kovacs M., Wang F. & Sellers J. R. Mechanism of action of myosin X, a membrane-associated molecular motor. J. Biol. Chem. 280, 15071–15083, doi: (2005).10.1074/jbc.M500616200 [PubMed] [Cross Ref]
  • Nagy S. & Rock R. S. Structured post-IQ domain governs selectivity of myosin X for fascin-actin bundles. J. Biol. Chem. 285, 26608–26617, doi: (2010).10.1074/jbc.M110.104661 [PMC free article] [PubMed] [Cross Ref]
  • Park H. et al. . Full-length myosin VI dimerizes and moves processively along actin filaments upon monomer clustering. Mol. Cell 21, 331–336, doi: (2006).10.1016/j.molcel.2005.12.015 [PubMed] [Cross Ref]
  • Bryan J. & Kane R. E. Actin gelation in sea urchin egg extracts. Methods Cell Biol. 25 Pt B, 175–199 (1982). [PubMed]
  • Kerber M. L. et al. . A novel form of motility in filopodia revealed by imaging myosin-X at the single-molecule level. Curr. Biol. 19, 967–973, doi: (2009).10.1016/j.cub.2009.03.067 [PMC free article] [PubMed] [Cross Ref]
  • Baboolal T. G., Mashanov G. I., Nenasheva T. A., Peckham M. & Molloy J. E. A Combination of Diffusion and Active Translocation Localizes Myosin 10 to the Filopodial Tip. J. Biol. Chem. 291, 22373–22385, doi: (2016).10.1074/jbc.M116.730689 [PMC free article] [PubMed] [Cross Ref]
  • Sakamoto T., Yildez A., Selvin P. R. & Sellers J. R. Step-size is determined by neck length in myosin V. Biochemistry 44, 16203–16210 (2005). [PubMed]
  • Knight P. J. et al. . The predicted coiled-coil domain of myosin 10 forms a novel elongated domain that lengthens the head. J. Biol. Chem. 280, 34702–34708, doi: (2005).10.1074/jbc.M504887200 [PubMed] [Cross Ref]
  • Armstrong J. M. et al. . Full-length myosin Va exhibits altered gating during processive movement on actin. Proc. Natl. Acad. Sci. USA 109, E218–224, doi: (2012).10.1073/pnas.1109709109 [PubMed] [Cross Ref]
  • Brawley C. M. & Rock R. S. Unconventional myosin traffic in cells reveals a selective actin cytoskeleton. Proc. Natl. Acad. Sci. USA 106, 9685–9690 (2009). [PubMed]
  • Sivaramakrishnan S. & Spudich J. A. Coupled myosin VI motors facilitate unidirectional movement on an F-actin network. J. Cell Biol. 187, 53–60 (2009). [PMC free article] [PubMed]
  • Ropars V. et al. . The myosin X motor is optimized for movement on actin bundles. Nat. Commun. 7, 12456, doi: (2016).10.1038/ncomms12456 [PMC free article] [PubMed] [Cross Ref]
  • Spudich J. A. & Watt S. The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of the tropomyosin-troponin complex with actin and the proteolytic fragments of myosin. J. Biol. Chem. 246, 4866–4871 (1971). [PubMed]
  • Weiner M. P. et al. . Site-directed mutagenesis of double-stranded DNA by the polymerase chain reaction. Gene 151, 119–123 (1994). [PubMed]
  • Hinnebusch A. G. Evidence for translational regulation of the activator of general amino acid control in yeast. Proc. Natl. Acad. Sci. USA 81, 6442–6446 (1984). [PubMed]
  • Sato O. et al. . Human deafness mutation of myosin VI (C442Y) accelerates the ADP dissociation rate. J. Biol. Chem. 279, 28844–28854 (2004). [PubMed]
  • Warshaw D. M. et al. . Differential labeling of myosin V heads with quantum dots allows direct visualization of hand-over-hand processivity. Biophys. J. 88, L30–32, doi: (2005).10.1529/biophysj.105.061903 [PubMed] [Cross Ref]
  • Kerssemakers J. W. et al. . Assembly dynamics of microtubules at molecular resolution. Nature 442, 709–712 (2006). [PubMed]
  • Jung H. S. et al. . Conservation of the regulated structure of folded myosin 2 in species separated by at least 600 million years of independent evolution. Proc. Natl. Acad. Sci. USA 105, 6022–6026, doi: (2008).10.1073/pnas.0707846105 [PubMed] [Cross Ref]

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group