Cells and culture

Dictyostelium discoideum (strain NC4) cells were cultured with Escherichia coli (strain B/r) on nutrient agar that contained 5 g of peptone, 5 g of lactose and 16 g of agar in 1 l of distilled water (Yumura and Kitanishi-Yumura, 1992). After 40 hours of culture, vegetative cells were harvested and washed three times with cold distilled water by centrifugation at 300 g for 1.5 minutes. They were suspended in Bonner's salt solution (BSS; 10 mM NaCl, 10 mM KCl, 3 mM CaCl₂; actual pH was 6.2), spread on non-nutrient agar plates and incubated at 21°C until further use. Strain AX2 cells and myosin heavy chain-null cells were cultured in HL-5 medium [1.3% (w/v) bacteriological peptone, 0.75% (w/v) yeast extract, 85.5 mM D-glucose, 3.5 mM Na₂HPO₄.12H₂O, 3.5 mM KH₂PO₄, pH 6.4] as described previously (Yumura and Uyeda, 1997). All data presented as wild-type cells were obtained from the NC4 strain cells, but AX2 cells showed the same results as NC4 cells.

Preparation of silicone rubber film

The preparation of silicone substrate was performed by modifying the method of Lee et al. (Lee et al., 1994). About 4 μl of the silicone fluid (dimethylpolysiloxane; 1000 centistokes of viscosity; Sigma, St. Louis, MO) was spread in 10 square mm with the thickness of 10-30 μm onto a cleaned coverslip (18×18 mm, Matsunami Glass, Japan) by a handmade glass spreader. The coverslip was settled on a tile plate and heated at 450°C using a custom-made hot plate, which was retained 2 mm above the coverslip. In the preparation of silicone substrate for the experiments of live cells, the silicone was heated for 20 minutes for wild-type cells and 25 minutes for MHC-null cells, respectively. The stiffness of silicone substrate was controlled via the heating time.

Calibration of bead displacement

The mechanical properties of the film were characterized by modifying the method of Lee et al. (Lee et al., 1994). Highly flexible microneedles with sharply pointed tips were made using a pipette puller (Narishige, Japan), and calibration was done by plotting the amount of tip deflection generated by weights in the range of 15-100 μg made from strips of 15 μm thick aluminum sheets. The force required to displace a bead on a silicone substrate was determined as the force that reproduces the deflection of the tip of the highly flexible microneedle. Video images of the calibration experiments were used to know the extent of tip deflection for a given bead displacement. The percent recovery of a bead toward its original position was measured 15 seconds after the substrate was released from microneedle. The percentage recovery was defined as [(p—q)/p]×100, where p is the bead displacement and q is the bead displacement after recovery.

Microscopy and acquisition of images

The observation chamber was made from a solid silicone sheet (17×17×0.5 mm) with a hole of about 10×10 mm, which was placed on a silicone film spread on the coverslip. About 200 μl BSS containing 5% carboxylate latex beads (0.5 μm in diameter, Polyscience, Warrington, PA) was carefully poured into the chamber and pipetted gently to attach the beads to the silicone surface. After 20 minutes, the chamber was washed with BSS to remove unattached beads. The silicone substrate was coated with 500 μg/ml bovine serum albumin (BSA) because the surface was so sticky without coating that cell migration became much slower.

The cell suspension was placed in the chamber filled with BSS, and the chamber was sealed with a coverslip. The cells were monitored by an inverted microscope equipped with a confocal system (LSM 510; Carl Zeiss, Germany) with× 63C Apochromat objective lens (NA 1.2). Resolution of acquired images was 512×512 or 1024×1024 pixels. Sequential images were stored in a computer at intervals 2-6 seconds and analyzed by Scion Image software (Scion Corporation, Frederick, MD). Drift of silicone substrate was corrected in a series of images on the basis of the position of some beads that were far away from the cell and did not move. Each image was moved to adjust the position of these beads on the same software. All experiments were performed at 22°C.

Image analysis

To analyze cell movement, the gained area was calculated over time by subtracting the retraction area from the extension area. Here, the extension area is defined as the increase in area determined from cell contours taken from two successive images, and the retraction area is defined as the decrease in area between two successive images (Weber et al., 1995; Yumura and Fukui, 1998).

The positions of the anterior and posterior edges of the migrating cell were also measured over time. The anterior and posterior edges were defined as both of the extreme edges perpendicular to the vector of instantaneous migration of the cells, which is defined from the cyto-centers of three continuous images (Uchida and Yumura, 1999). If the cell turns more than 30°, the data was not counted. Image processing was performed by macro programs in Scion image (Scion Corp.) and improved by Photoshop (Adobe Systems Inc., Japan).

Position of beads in an image was measured at the center of the beads. If displacement of a bead was less than 0.2 μm (less than 3 pixels in the case of 1024×1024 pixel image), the bead displacement was not quantified. The vector of bead displacement, derived from the present position and the original position of the bead, was divided into perpendicular and parallel vector components to the direction of cell migration. We took the parallel component of the vector of the bead displacement as the substantial bead displacement around the anterior and the posterior regions of the cell and the perpendicular component as that around the side regions.

Characterization of silicone substrate

To examine the force vectors exerted against the substrate by Dictyostelium cells, we prepared an improved flexible silicone substrate. Silicone oil was spread on a coverslip, crosslinked by heating at 450°C, and the beads (0.5 μm in diameter) were embedded. The stiffness of the silicone substrate depends on the time of heating, and the absolute values of the stiffness were calibrated from the measurement of displacement of beads caused by a highly flexible microneedle. The silicone substrate had the stiffness of 5.74 nN/μm (10 minutes heating) and 8.59 nN/μm (25 minutes heating), respectively. Fig. 1a,b shows the mechanical properties of the softest (a) and the hardest (b) silicone substrates used in the present study. The correlation between the displacement of beads and the strength of applied force was almost linear. The percent recovery of bead displacement after removal of the microneedle from the silicone substrate was 87% on average in both cases (Fig. 1c,d), whereas the substrate crosslinked by glow discharge showed only 50% recovery at the stiffness of 5 nN/μm (Oliver et al., 1995). The higher recovery rate is an important improvement for estimating the traction force more accurately.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

Mechanical properties of the silicone substrate. The bead displacement in response to the force applied by a microneedle was plotted. The slope of the line represents the stiffness of silicone substrate. (a) The stiffness of the softest substrate is 5.74 nN/μm. (b) The stiffness of the hardest substrate is 8.59 nN/μm. (c,d) Percent recovery of the beads toward their original position of the softest (c) and the hardest (d) substrates. The average of percent recovery was 87% in both cases.

Alternate movements of beads during cell migration

Dictyostelium cells were allowed to settle on the silicone substrate, and the movement of beads around the cells was monitored and analyzed. The movement of beads reflected the distortion of the silicone substrate caused by the cell during its migration. The analysis was performed on straight moving cells. Fig. 2a shows a differential interference contrast image of a cell on the silicone substrate. The superimposed image of five successive images acquired at 6 second intervals clearly shows the movement of beads around the cell (Fig. 2b). Fig. 3 shows a typical example of two alternate patterns of bead movement around a migrating cell. The direction of bead movement reversed repeatedly. In Fig. 3a,b, the beads around the anterior and the posterior regions of the cell moved toward the cell body, and those around the side region moved away from the cell. Subsequently, the direction of bead movement was reversed (Fig. 3c,d). The beads around the anterior and the posterior regions moved away from the cell, and those around the side regions moved toward the cell. Interestingly, most of beads in the side and the posterior regions returned to their original positions, and the beads around the anterior region moved away from the cell over their original positions. The beads moved in similar patterns repeatedly. In addition, all 51 cells exhibited similar patterns. Hereafter the first pattern of bead movement is referred to as pattern 1 (inset in Fig. 3b) and the second as pattern 2 (inset in Fig. 3d).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Differential interference contrast images of a Dictyostelium cell on the silicone substrate. (a) A single image of a cell surrounded with beads. (b) Five successive images acquired at 6 second intervals were superimposed, which clearly shows the movement of beads around the cell. Bar, 5 μm. Movie 1 showing cell migration on the silicone surface is available at jcs.biologists.org/supplemental.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

A typical movement of beads around a migrating wild-type cell. The gray and black lines represent a superimposed image of two cell contours at 0 seconds and 21 seconds in a, at 21 seconds and 42 seconds in b, at 42 seconds and 63 seconds in c, at 63 seconds and 84 seconds in d, respectively. Large closed arrows represent the direction of cell migration. The gray and black dots represent positions of the beads at 0 seconds and 21 seconds in a, at 21 seconds and 42 seconds in b, at 42 seconds and 63 seconds in c, at 63 seconds and 84 seconds in d, respectively. The crosses represent the original position of beads before the cells were placed on the silicone substrate. Small open arrows of red, green and blue represent the vectors of bead movements around anterior, side and posterior regions, respectively. Black open arrows represent the bead movement underneath the cell body. The length of the open arrows is twice as long as the displacement of the beads. Between 0 and 42 seconds (a,b), the beads around the anterior and posterior regions moved toward the cell body, and those around the side region moved away from the cell. Subsequently, between 42 and 84 seconds (c,d), the bead movements were reversed. Insets in b and d, schematic representations of the directions of the movement of beads surrounding the cell are shown. Large black arrows represent the direction of the cell migration. Closed arrows of red, green and blue represent the directions of bead movements around anterior, side and posterior regions, respectively. The pattern of bead movement between 0 and 42 seconds is referred to as pattern 1 (inset of 3b) and the pattern between 42 and 84 seconds as pattern 2 (inset of 3d). The beads underneath the cell body occasionally moved in a different manner and direction. In many cases, this feature could not be expressed in these figures because the directional change occurred in a short time. The stiffness of this silicone substrate is 5.97 nN/μm. Bar, 5 μm.

Correlation between specific patterns of bead movement and cyclic behavior of cells during migration

The cyclic change in the pattern of bead movement observed in Dictyostelium cells may be correlated with cyclic behavior of cells during migration, which has been described previously (Soll et al., 1987; Weber et al., 1995). Then, contours of a cell taken at two successive times were superimposed, and the difference between the extension and the retraction areas (gained area) was plotted over time. If the difference is more than zero, the extension is more prominent than the retraction (the extension phase). If the difference is less than zero, the retraction is more prominent than the extension (the retraction phase). The cell alternated between the extension and the retraction phases as shown in Fig. 4a. The average duration of a cycle was 70.4±28.3 seconds (n=22).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Cyclic behavior of cells corresponds with specific patterns of bead movement. (a) A representative time course of the gained area of a migrating wild-type cell. Two successive contours of a cell were superimposed, and the gained area was calculated by subtraction of the retraction area from the extension area. Note that extension and retraction phases alternated. The average time of a cycle was 70.4±28.3 seconds (n=22). Two patterns of bead movement (pattern 1, circled number 1; pattern 2, circled number 2) were well correlated with the extension and the retraction phases, respectively. The vertical lines are drawn on the basis of the transition points of the two patterns. (b) The displacement of the anterior (squares) and posterior (triangles) edges of the cell was plotted over time. The definition of the anterior and the posterior edges is described in Materials and Methods. When the speed of the advance of the posterior edge reached its maximum value, the bead movement switched from pattern 1 to pattern 2 (arrows in b). Note that the anterior edge advanced significantly after the peak of the forward movement of the posterior edge (arrowheads in b).

Interestingly, the two alternate patterns of bead movement (insets in Fig. 3b,d) were well correlated with the extension and the retraction phases of cell behavior during migration, respectively (circled numbers in Fig. 4a). During the extension phase, beads around the anterior and the posterior regions moved toward the cell body, and those around the side region moved away from the cell (pattern 1). During the retraction phase, beads around the anterior and the posterior regions moved away from the cell, and those around the side region moved toward the cell (pattern 2). All the beads around the cell simultaneously changed their direction of movement at the transition between the extension and the retraction phases.

The positions of the anterior and the posterior edges of the migrating cell were also monitored, and the displacement of both edges was plotted over time (squares and triangles in Fig. 4b). When the bead movement switched from `pattern 1' to `pattern 2', the peaks of forward movement of the posterior edge appeared at these transitions (arrows in Fig. 4b). After the peaks of the displacement of the posterior edge, the anterior edge began to move forward significantly (arrowheads in Fig. 4b).

Alternate movements of beads do not occur in MHC-null cells

Myosin II, conventional myosin, is considered to play an important role in the force generation during cell movement. However, the Dictyostelium MHC-null mutant cells can still migrate slowly on a substrate (Wessels et al., 1988). To investigate the role of myosin II in the generation of traction force, bead movement caused by MHC-null cells was analyzed. The analysis was performed on straight moving cells. Fig. 5a-d shows a typical bead movement during migration of a MHC-null cell. Interestingly, these cells showed only `pattern 1' observed in wild-type cells. Fig. 5e shows the gained area in a MHC-null cell as a function of time. The MHC-null cells still repeated the two phases, the extension and the retraction phases. The average duration of a cycle was 151.7±38.0 seconds (n=15), which was twice as long as that of wild-type cells. These results suggest that the alternate biphasic patterns of bead movement observed in wild-type cells is not directly related to the extension and the retraction phases of cell locomotory behavior.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5.

Alternate movements of beads did not occur in MHC-null cells. Representative movement of the beads around a migrating MHC-null cell (a-d). The gray and black lines represent a superimposed image of two cell contours at 0 seconds and 120 seconds in a, at 120 seconds and 240 seconds in b, at 240 seconds and 360 seconds in c, at 360 seconds and 480 seconds in d, respectively. Large closed arrows represent the direction of cell migration. The gray and black dots represent positions of the beads at 0 seconds and 120 seconds in a, at 120 seconds and 240 seconds in b, at 240 seconds and 360 seconds in c, at 360 seconds and 480 seconds in d, respectively. The bead movement did not reverse. Insets in c and d, schematic images of the direction of movement of beads surrounding the cell. The bead movements showed only the pattern 1 observed in wild-type cells. (e) A representative time course of the gained area of a migrating MHC-null cell. Note that MHC-null cells also repeated two phases, extension and retraction phases. The average time of a single cycle was 151.7±38.0 seconds (n=15). (f) The displacement of the posterior and the anterior edges of the cell was plotted over time. Note that the position of the posterior edge moved roughly at a constant rate when compared with the case of wild-type cells (see Fig. 4b), whereas the rate of the displacement of the anterior edge had obvious peaks (arrows in f). The cell shown in e and f was different from the cell in a-d. The stiffness of silicone substrate was 5.97 nN/μm. Bar, 5 μm.

Posterior contraction mediated by myosin II generates a pushing force of the anterior extension

Analysis of the displacement of the anterior and the posterior edges of the migrating MHC-null cells explained the loss of the pattern 2 in their bead movement. In Fig. 5f, the posterior edge of the MHC-null cell moved forward almost at a constant rate, in contrast to wild-type cells showing apparent peaks at the beginning of the retraction phases (arrows in Fig. 4b). Only the rate of the movement of the anterior edge changed cyclically in migrating MHC-null cells (arrows in Fig. 5f), and this cyclic anterior extension mainly contributed to the biphasic changes of their gained area. At all time, beads around the anterior and the posterior regions moved toward the cell body (insets of Fig. 5c,d). After the cell moved far enough away, the beads returned to their original positions owing to the relaxation of the substrate.

Comparing the results in Fig. 4b and Fig. 5f, we found that the considerable forward movement of the posterior edge during the retraction phase in wild-type cells is caused by myosin-II-dependent contraction. After the forward movement of the posterior edge, the anterior edge also began to move forward significantly during the retraction phase (arrowheads in Fig. 4b). This anterior extension pushed the beads forward in wild-type cells (Fig. 3c,d, Fig. 6a). These observations strongly suggest that during the retraction phase, the `pushing force' of the anterior extension is generated by the posterior contraction mediated by myosin II, which localizes at the rear cortex of the migrating cells (Yumura et al., 1984). This is the first convincing evidence to show the role of myosin II in the anterior extension of migrating cells.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6.

Correlation between bead displacement and phase transition of cell locomotory behavior. Solid squares on the line graphs represent the bead displacement around three regions, anterior (a,d), side (b,e) and posterior (c,f) of a wild-type (a-c) and a MHC-null (d-f) cell, respectively. The method for the measurement of bead displacements was described in Materials and Methods. Zero at the y axis indicates the original position of the beads before the cells were placed on the substrate. Negative values indicate displacement toward the cell, and positive values indicate displacement away from the cell. In wild-type cells, the bead displacement around the anterior region reversed and increased to positive values during the retraction phase (a), representing a pushing force in this region. The bead displacement around the side and the posterior regions reached zero during the retraction phase (b,c), suggesting that the adhesion of the cell body to the substrate in both regions was lost. Note that the bead displacement in MHC-null cells was not reversed but enhanced throughout the extension and the retraction phases. All graphs include results of several beads (multiple plotted lines).

Myosin II cancels the traction force at the side and the posterior of the cell

Where and how does the cell exert force against the substrate? The time course of displacements of beads from their original positions was investigated around three divided regions (anterior, side and posterior regions) of the migrating cells (Fig. 6). The vector of bead displacement was divided into components that were perpendicular and parallel to the direction of cell migration. The bead displacement around the anterior and the posterior regions was represented by the parallel component and that around the side region was represented by the perpendicular component. The bead displacement around the anterior region of a wild type cell changed from -1 μm during the extension phase to +1 μm during the retraction phase (Fig. 6a). Here, negative values indicate the movement of beads toward the cell body, and positive values indicate the movement away from the cell body. In the case of MHC-null cells, the bead displacement around the anterior region decreased from -1μ m to -3 μm (Fig. 6d), and these cells never pushed the beads forward during the retraction phase. These results suggest that the `pushing force' for anterior extension depends on myosin II during the retraction phase.

The bead displacement around the side region in wild-type cells was +1μ m during the extension phase, decreased to zero during the retraction phase, and it never turned negative (Fig. 6b). Here, zero indicates the original position of beads before the cells were placed on the substrate. The bead displacement around the posterior region was -1.5 μm during the extension phase and it reached zero during the retraction phase (Fig. 6c). Similar results were obtained in analyses of 50 different wild-type cells. The bead displacement around the side and the posterior regions returned to zero during the retraction phase, suggesting that the traction force of these regions was cancelled.

In the case of MHC-null cells, however, the bead displacements around all of three regions were enhanced both during the extension and the retraction phases (Fig. 6d-f). The bead displacement around the side region increased from +1 μm to +4 μm (Fig. 6e). The bead displacement around the posterior region decreased from -1 μm to -4 μm (Fig. 6f). Similar results were obtained in 20 different cells. Taken together with the results of wild-type cells, myosin II is required for the cancellation of the traction force, probably by detaching the adhesion of middle and posterior part of the cell body from the substrate.

Wild-type cells migrate more efficiently than MHC-null cells via the posterior contraction

The magnitude of the traction force was calculated by multiplying the stiffness of the substrate by the bead displacements (Table 1), which showed that the calculated magnitude of the traction force of MHC-null cells was larger than that of wild-type cells. However, we could explain these contradictory values by the lack of the cancellation process of the traction force in the MHC-null cells. Furthermore, the comparison of the locomotory behavior between a wild-type cell and a MHC-null cell under the pressure of overlaid agar sheet (Yumura and Fukui, 1985; Yumura, 2001) showed that the MHC-null cell extended long cell processes, but could not migrate (Fig. 7b), whereas the wild-type cell could easily migrate. These observations suggested that the posterior contraction mediated by myosin II is required for the cells to migrate under the pressure of the agar, and the wild-type cells could utilize the traction force more efficiently than the MHC-null cells, with the contraction of the posterior region.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 7.

Cell migration under the pressure of overlaid thin agar sheet. Wild-type (a) and MHC-null (b) cells were compressed between a thin agar sheet and a glass coverslip. Arrows represent the direction of the anterior extension. The wild-type cell was able to migrate under the pressure of the agar sheet. This cell showed both the posterior retraction and the anterior extension. However, the MHC-null cell was not able to migrate. Asterisks represent the position of the nucleus of this cell. The MHC-null cell did not show the posterior retraction, although it showed anterior extensions. Bar, 10 μm. Movies 2 and 3 showing cell migration are available at jcs.biologists.org/supplemental.