2071
Journal of Cell Science 107, 2071-2079 (1994)
Printed in Great Britain © The Company of Biologists Limited 1994
Microtubules are required in amoeba chemotaxis for preferential stabilization
of appropriate pseudopods
Masahiro Ueda* and Satoshi Ogihara*,†
Department of Biology, College of General Education, Osaka University, Toyonaka, Osaka 560, Japan
*Present address: Department of Biology, Faculty of Science, Osaka University, Osaka 560, Japan
†Author for correspondence
SUMMARY
Amoebae of Physarum polycephalum exhibit chemotactic
responses to glucose and to cAMP. The chemotaxing
amoebae exhibit alternating locomotive movements: relatively linear locomotion and movements that change the
direction of the locomotion. Such locomotive activity is
tightly coupled with the changes in the number and the
positions of the pseudopods; cells have one pseudopod at
the leading edge during their linear locomotion, while they
have multiple pseudopods when they are changing the
direction of locomotion. Treatment of cells with microtubule-disrupting reagents inhibited the chemotaxis of the
cells. To characterize the role of the microtubule system in
chemotaxis, we quantitatively analyzed the relationship
between the positions of multiple pseudopods of the
amoebae and the relative stability of the pseudopods
during reorientation. No significant differences were
observed in the pseudopod dynamics between the
untreated and the treated amoebae. In both cases, one
pseudopod at the leading edge continued to expand during
linear locomotion. It then split into two to three pseudopods
in the reorientation phase, and the positions of the multiple
pseudopods were random. Among multiple pseudopods,
however, the pseudopods closer to the microneedle tip were
selectively stabilized more often than those distant from the
tip in the presence of the microtubule system. By contrast,
such preferential stabilization of the appropriate
pseudopods was completely abolished by microtubule
inhibitors. The microtubule-dependent selection of appropriately located pseudopods enables amoebae to turn
correctly at the reorientation step.
INTRODUCTION
the source. In the second model, the spatial sensing model,
amoebae sense the gradient simultaneously at different points
over their entire cell surface. As a result of integration of the
sensed information, one pseudopod is specifically formed on
the side of the amoeba that corresponds to the higher concentration of chemoattractant. The leading edge formed in this
way is maintained via the cell’s adoption of a polarized shape,
with one pseudopod and a tapered tail (Zigmond et al., 1981).
In addition to these mechanisms, Swanson and Taylor (1982)
tried to explain the chemotactic movements of the amoebae of
Dictyostelium discoideum by employing a term ‘global
response’. According to their hypothesis, chemotaxis is guided
by a local response to a gradient, i.e. extension of a pseudopod,
and is coordinated to achieve directed movements by a global
response, i.e. cytoplasmic flow in the direction of stimulation.
They suggest that the microtubule system has a crucial role in
the global response. On the other hand, previous studies of the
chemotaxis of polymorphonuclear leukocytes revealed that
microtubules are neither obligatory for the spatial detection of
gradients nor required for locomotion towards the source of an
attractant (Allan and Wilkinson, 1978; Devreotes and
Zigmond, 1988).
Chemotaxis of amoeboid cells is a phenomenon that depends
on amoeboid movements. Pseudopods are essential for such
amoeboid movements (Taylor and Condeelis, 1979), and
investigations of cellular aspects of chemotaxis in amoebae
have focused on the behavior of pseudopods (Gerisch et al.,
1974; Zigmond, 1974, 1978; Swanson and Taylor, 1982;
Varnum-Finney et al., 1987a,b; Stossel, 1989). To explain the
mechanism whereby cells sense a gradient of a chemoattractant, two general models have been proposed. In one model,
the temporal sensing model, amoebae assess the gradient by
sampling the concentrations of an attractant at different times,
via different pilot pseudopods, during the extension of
pseudopods (Gerisch et al., 1974). Among the pilot
pseudopods, only those that keep receiving a positive input are
stimulated to continue extension. Varnum-Finney et al.
(1987b) showed that Dictyostelium amoebae extend their
pseudopods towards the source of a chemoattractant as frequently as they extend them in the opposite direction. Those
pseudopods formed in the direction of the source more often
become a new leading edge than those extending away from
Key words: Physarum polycephalum, amoeba chemotaxis,
microtubule, pseudopod formation
2072 M. Ueda and S. Ogihara
We report here details of the chemotaxis of the amoebae of
Physarum polycephalum. We focused our attention on changes
in the positions and numbers of the pseudopods. As a consequence, we found that microtubules were necessary for the
preferential stabilization of the pseudopods that were closer
than others to higher concentrations of the chemoattractant.
The role of microtubules appeared to be essential for chemotaxis of Physarum amoebae.
MATERIALS AND METHODS
Cell culture
Haploid cells of Physarum polycephalum, strain J, were obtained from
a single spore (Wakasugi and Ohta, 1973). They were maintained in
culture, with Aerobacter aerogenes, on nutrient-agar plates that
contained 5.00 g/l glucose, 0.50 g/l yeast extract, 5.00 g/l bactopeptone, 2 mM MgSO4 and 25 mM potassium phosphate buffer (pH 6.5;
Wakasugi and Ohta, 1973). Under these culture conditions, cells
mostly formed cysts after amoebic growth. To obtain amoeboid cells,
the cysts were harvested and washed with several successive centrifugations (800 g, 2 min) in PDF buffer, which contained 20 mM
KCl, 40 mM NaPO4, 0.7 mM CaCl2 and 2 mM MgSO4, pH 6.3, at
22°C. This buffer was originally developed for experiments with the
amoebae of Dictyostelium discoideum (Bennett and Condeelis, 1984).
Washed cysts were replated at high density (3×106 to 5×106 cells/9
cm plate) with A. aerogenes on the same nutrient-agar plates as used
for the stock culture. After incubation for 24-48 hours at 22°C, cells
were harvested at the exponential phase of growth and suspended in
PDF buffer at 22°C. Cells cultured and harvested in this way were
100% amoeboid, and there were no cysts or flagellates.
To prepare amoebae for microscopy, cells suspended in PDF buffer
were allowed to settle on glass coverslips that had been coated with
1.5% agar or with 0.01% polyethyleneimine. After starvation in a
moist chamber at 22°C for 16-24 hours, cells were used in subsequent
experiments.
Video recording of live cells
The amoebae were observed with Nomarski optics under an Olympus
inverted microscope (IM-2) or a Nikon microscope (Optiphoto 2).
Images of locomoting cells were monitored and recorded with a TV
camera (WV-1550; Panasonic, Osaka, Japan) and a videocassette
recorder (NV-FS5; Panasonic). Images on the recorded tapes were
captured with a personal computer (Macintosh Quadra 950; Apple
Computer, Cupertino, CA, USA) equipped with an analog-digital
converter for the video images (Radius Inc., San José, CA, USA).
Chemotactic stimulation
Chemotactic stimulation was provided by solutions of 10 mM glucose
or 1 µM cAMP in PDF buffer, which were allowed to diffuse from
the tip of a glass microneedle (Gerisch et al., 1975; Swanson and
Taylor, 1982). The microneedle, which was connected to a microinjector (IM-6; Narishige, Tokyo, Japan), was manipulated by use of a
micromanipulator (MO-202; Narishige). There were no significant
differences in the chemotactic behavior stimulated with glucose and
that with cAMP, as described below. Formation of a stable gradient
of a chemotactic attractant was confirmed by examination of fluorescence, from a medium that contained 6 mM CaCl2, around the tip of
a microneedle that contained a calcium-sensitive fluorescent dye,
fura-2 (Dojindo Laboratories, Kumamoto, Japan) at 1 mM
(Grynkiewicz et al., 1985). The fluorescence image was captured by
a personal computer as described above, and the fluorescence intensities were quantitated using NIH Image 1.53. The concentration
gradient of the fluorescent dye was stable for at least 10 minutes.
Quantitative analysis of chemotactic tracks
The positions of the centroids of individual amoebae were determined
with the image-processor, Σ-III (Nippon Avionics Co., Tokyo, Japan).
In brief, a cell’s behavior was recorded by video microscopy as
described above. A frozen image of cells was displayed via the imageprocessor on a video screen (PVM-1442Q; SONY, Tokyo, Japan).
The screen had 480×640 pixels and each pixel was assigned x,y coordinates according to its position on the screen. For accuracy, we used
differential-interference contrast optics and a manual procedure for
the determination of cell contours. The contour of a cell was bounded
and fitted by a rectangle that was placed horizontally on the monitor
screen. The centroid of the rectangle was regarded as the apparent
centroid of the cell. This estimation of the location of the centroid
contains unavoidable errors. The apparent centroid occupies different
positions in a cell if the rectangle is placed at different orientations
on the screen. The distance between the apparent centroid and the true
centroid, which was obtained by calculation of the mean x and mean
y coordinates of the pixels located along the edge of the cell (Varnum
et al., 1986), was computed and found to have a mean value of 5.6±2.4
pixels (n=66) on the screen. This value is equivalent to 1.1±0.46 µm,
which is about 6% of the mean length of 10 amoebae with various
shapes; each cell had an area of approximately 5500-8000 pixels.
Based on these values of the errors, the symbols in Fig. 2 are
expressed so that their sizes correspond to those that incorporate 60%
of the maximum value of the error.
To obtain chemotactic tracks, centroids of chemotactically moving
amoebae were plotted at intervals of 5 seconds. These sampling
intervals were short enough to allow detection of the changes in the
locomotive activity of the amoebae.
To quantitate chemotactic orientation of cells towards the source of
the attractant, the centripetal components of the net movements were
calculated from the locomoting tracks of cells. The centripetal components were the vectors, directed towards the tip of a microneedle
that contained the attractant, of the net movements of the amoebae.
Briefly, first, the net movements were obtained by determining the
change in the x,y coordinates of the centroid of one cell over the
course of one minute. Second, the net movements were divided geometrically into two vectorial components, i.e. the centripetal components, directed towards the tip of the microneedle, and the vertical
components perpendicular to the centripetal components. The values
of the centripetal components were plotted against the distance from
the tip of the microneedle. When an amoeba moved away from the
tip of the microneedle, the value of the centripetal component was
negative. In these cases, the movement was centrifugal.
Calculation of cell area and pseudopod area
The original images of the amoebae on the recorded tapes were
processed for calculation of cell areas and pseudopod areas. Images
were captured at 5 second intervals with a personal computer
(Macintosh IIci; Apple Computer, Cupertino, CA, USA) equipped
with an analog-digital converter (RasterOps 364; Santa Clara, CA,
USA). The contours of cells and pseudopods were traced manually
on the digital images with a commercially available software system
(Studio/8; Electronic Arts, San Mateo, CA, USA). We defined
pseudopods as hyaline sheet-like extensions that were distinguishable
from the particulate regions of the cytoplasm in the main body of the
cell on the DIC images (Taylor and Condeelis, 1979). The areas of
both cells and pseudopods were computed by counting the numbers
of pixels contained within the traces.
Quantification of pseudopod position with respect to the
tip of a microneedle
The positions of pseudopods were determined as explained in the
scheme shown in Fig. 1. In brief, both the cell centroids and the
pseudopod centroids were obtained by calculation of the mean x and
mean y coordinates of the pixels located along the traced edges of the
cells and pseudopods (Varnum et al., 1986). The coordinates were set
for individual amoebae as shown in Fig. 1, such that the cell centroid
and the tip of the microneedle became the origin (0,0) and a site on
Role of microtubules in chemotaxis 2073
y
P2
x
P1
x
P3
C1
P1
y
y
P3
C2
P2
x
Fig. 1. Schematic diagram illustrating the way in which the positions
of pseudopods in individual cells were determined with respect to the
tip of a microneedle. The x and y axes are arranged so that the tip of
the microneedle (+) and the cell centroid (C1 and C2) fall on the y
axis. The centroids of pseudopods (P1, P2, and P3) in terms of these
coordinates are shown in the inset. Pseudopods are illustrated by
shading. Inset: the centroids of pseudopods found in the proximal
region (heavy stippling), the distal region (without stippling), and
lateral region (light stippling) were scored. Proximal and distal
regions of cells were defined as fanlike areas making an angle of
120° ahead of and behind the cell centroid with respect to the tip of
microneedle, respectively. The remaining areas were regarded as the
lateral region of the cells.
the y axis (y>0), respectively. Positions of pseudopod centroids were
then assigned by reference to these coordinates (inset). With few
exceptions, pseudopod centroids having positive values on the y axis
were located in the region of the cell that faced a higher concentration of attractant and those having negative values faced a lower concentration.
Immunofluorescence procedure
To determine whether or not microtubules were disrupted by nocodazole, we used indirect immunofluorescence to observe the subcellular distribution of tubulin in amoebae that had been pretreated with
5 µM nocodazole or 10 µM mebendazole (Sigma, St Louis, MO,
USA) for 10 minutes at 22°C. Fixation and staining of cells were
performed by the method of Carboni and Condeelis (1985) with some
modifications. A mouse monoclonal antibody against chicken tubulin
(Amersham, Buckinghamshire, England) was used at 10-20 µg/ml,
with subsequent staining with FITC-labeled sheep antibodies against
mouse IgG at a dilution of 1:100 (Amersham). Immunofluorescent
images of cells were observed with a Nikon Optiphoto 2 microscope
equipped with a ×100 objective lens (NA 1.25). Photographs were
taken on Kodak T-Max 3200 film.
RESULTS
Microtubules are necessary for chemotaxis of P.
polycephalum amoebae
Individual amoebae of P. polycephalum were placed in a stable
gradient of an attractant and each showed a chemotactic
response. Amoebae kept moving more or less towards the tip
of the microneedle during chemotactic stimulation (Fig. 2A,B).
Note that it was very rare for an amoeba to arrive at the tip of
the microneedle solely as a result of linear locomotion against
the gradient of an attractant, as described below in detail.
Fig. 2. Tracks of locomoting amoebae in the absence (A and B) and
in the presence (C and D) of the microtubule-disrupting reagent,
nocodazole. (A and B) Starved amoebae were stimulated with 1 µM
cAMP from a glass microneedle. The tip of the microneedle (+) was
introduced into the field from the left side. The concentric circles in
A represent the concentration gradient of a chemoattractant as
determined using fura-2. The numbers near each circle indicate the
relative concentration of a chemoattractant. The extreme distorted
signal pattern on the left side of the circles arisen from the intense
signal contained in the microneedle. (C and D) Amoebae were
treated with 5 µM nocodazole for 10 minutes and then stimulated in
the same way. Cell centroids of moving amoebae were plotted at
intervals of 5 seconds. The starting points are indicated by S, and the
direction of locomotion by arrows. Filled symbols indicate the cell
centroids for which the persistency factor (Euteneuer and Schliwa,
1984) had values >1.110. The persistency factor was calculated by
comparison of two locations of cell centroid at intervals of 30
seconds. The tracks indicate that chemotaxis of the amoebae was
inhibited by nocodazole. Bar, 20 µm.
The chemotactic response was demonstrated more clearly by
a quantitative analysis of the locomotive tracks of individual
amoebae (Fig. 3A-D). From the locomotive tracks, we calculated the centripetal components towards the tip of the
microneedle (see Materials and Methods for details). The centripetal components, which are the vectorial components up the
gradient of the net movements, are a measure of the chemotactic orientation of the amoebae. In the case of unstimulated
amoebae, centrifugal movements (negative values on the y
axis) were observed to the same extent as centripetal
movements. That is to say, migration of the amoebae was
totally nondirectional and random (Fig. 3A). Upon chemotactic stimulation, the amoebae showed a rapid decrease in centrifugal movements within one minute (Fig. 3B). Such
dominance of centripetal components over centrifugal components was maintained during chemotactic stimulation (Fig.
3C,D). As a result, most amoebae (79%, n=84) that had been
located within 80 µm of the tip of the microneedle before stim-
2074 M. Ueda and S. Ogihara
(n=54) of treated amoebae were able to travel successfully to
the tip of microneedle, they seemed to arrive there merely by
chance, since their centripetal movements occurred to the same
extent as their centrifugal movements in the presence of nocodazole (Fig. 3E-H). Therefore, it appears that the microtubule
system is necessary for chemotaxis of P. polycephalum
amoebae. Mebendazole at 10 µM, which is known to inhibit
the polymerization of tubulin (Laclette et al., 1980), also
disrupted the microtubules, as determined by immunofluorescence microscopy, and inhibited the chemotaxis of Physarum
amoebae in a similar manner (data not shown).
Fig. 3. Quantitative analysis of chemotactic orientation of the
amoebae toward the tip of a microneedle in the absence (A-D) and in
the presence (E-H) of nocodazole. Amoebae were treated with 5 µM
nocodazole and stimulated with 1 µM cAMP as described in the
legend to Fig. 2. Chemotactic responses of individual amoebae are
expressed as centripetal movement, and details are given in Materials
and Methods. (A and E) Centripetal components of locomotive
tracks of unstimulated amoebae; (B and F) 0 to 1 minute after
stimulation; (C and G) 5 to 6 minutes after stimulation; (D and H) 10
to 11 minutes after stimulation. The untreated amoebae exhibiting
random locomotion before stimulation began to migrate toward the
tip of the microneedle in the first minute, irrespective of their
distance from the tip (A-D). By contrast, centrifugal movements and
centripetal movements of the treated amoebae were observed to the
same extent before and after chemotactic stimulation, showing the
absence of chemotaxis by cells treated with nocodazole (E-H).
ulation reached the tip within 10 minutes after the start of
chemotactic stimulation.
To test whether or not amoebae need microtubules for
chemotaxis, we pretreated amoebae with 5 µM nocodazole for
10 minutes and stimulated them with an attractant in the
presence of nocodazole. The treated amoebae had shorter and
much smaller numbers of microtubules, as determined by
immunofluorescence microscopy (see Fig. 4). Nocodazole
inhibited the chemotactic responses (Fig. 2C,D). As shown in
Fig. 3, the treated amoebae continued to move with random
orientation after chemotactic stimulation. Although 20%
Microtubules have a pivotal role in the preferential
maintenance of specific pseudopods
It has been well established that a migrating amoeba extends
pseudopods at its leading edge and slides forward. Thus, the
position of pseudopods in an amoeba is closely related to the
direction of migration. As Fig. 2 shows, the Physarum
amoebae exhibited two different, alternating types of locomotive movement: relatively linear locomotion and less linear
movements that resulted in decreased net migration. When
amoebae were moving in almost straight lines, the centroids of
the amoebae moved at relatively high speeds (about 20-30
µm/minute) in a consistent direction. When the amoebae
reduced their migration speeds (about 3-10 µm/minute), they
often changed the direction of their locomotion. Such locomotive activity was tightly coupled with morphological changes;
cells adopted a polarized shape during linear locomotion (Fig.
5E,F; unipolar cells), and they had two to three pseudopods
during the reorientation phase (Fig. 5A,B,C,D,G; multipolar
cells). The morphology of the amoebae during reorientation
indicated that the amoebae reoriented themselves first by
formation of multiple pseudopods and then by selective maintenance of one pseudopod. The selection of one pseudopod
resulted in the formation of a new leading edge. Thus, the
stability of the pseudopods on multipolar cells varied significantly.
Such preferential stabilization of one pseudopod was more
clearly shown by analyzing the relationship between the
positions of the pseudopods, relative to the gradient of the
attractants, and the stability of pseudopods (Fig. 6; see
Materials and Methods for details). The stabilization of
pseudopods occurred irrespective of whether nocodazole was
absent (Fig. 6A) or present (Fig. 6B). The number of
pseudopods changed repeatedly from one (clear zone in Fig.
6) to two (shadow zone in Fig. 6); the repeats are shown for
five times in Fig. 6A and seven times in Fig. 6B. Selection of
pseudopods to be stabilized did not depend on the initial sizes
of the pseudopods when two new pseudopods appeared by
splitting one previous pseudopod. In Fig. 6A, the larger
pseudopods continued to grow in the first three turns, while the
smaller pseudopods were chosen in the next two turns. A
similar inclination is obvious in the presence of nocodazole
(Fig. 6B). The untreated amoeba, however, in all cases showed
correct selection of the pseudopod to be stabilized (thick lines).
The stabilized pseudopods were those formed on the cell side
that corresponded to the higher concentration of the attractant.
By contrast, such selection of appropriate pseudopods was
abolished by microtubule inhibitors. The treated amoebae
chose appropriate pseudopods in only three cases out of seven;
in four cases, the amoebae stabilized those pseudopods that
Role of microtubules in chemotaxis 2075
Fig. 4. Amoebae treated with nocodazole showed
a significant decrease of the numbers and lengths
of microtubules. (A and B) Control cells. (C and
D) Amoebae treated with 5 µM nocodazole for 10
minutes before fixation for immunofluorescence.
(A and C) DIC images. (B and D)
Immunofluorescence of tubulin. Bar, 10 µm.
were located on the cell side away from the source of the
chemoattractant.
The dependence on microtubules of the preferential stabilization of appropriate pseudopods during chemotaxis was
confirmed statistically by analyzing the positions and the
relative stability of over one thousand pseudopods as shown in
Fig. 7. In unipolar cells, the pseudopods were primarily located
in the proximal regions with respect to the tip of the microneedle (Fig. 7A, n=82). Pseudopods continued to be extended
preferentially at the cell margin that faced the higher concentration of attractant (y>0). By contrast, the pseudopods of drugtreated amoebae were distributed almost randomly with respect
to the tip of the microneedle (Fig. 7C, n=260). Though the
treated unipolar cells similarly had one continuously extending
pseudopod at the leading edge, the pseudopod region of the
treated unipolar cells tended to be wider, spreading slightly
toward the lateral sides of the cell, than the pseudopod region
of untreated unipolar cells.
Analysis of multipolar cells indicated a random orientation
of the pseudopods irrespective of whether microtubules were
present or absent (Fig. 7B,D). The positions of pseudopods
with respect to the tip of the microneedle were scored by the
method illustrated in Fig. 1. In the absence of nocodazole, the
distribution of pseudopods in the multipolar cells was 39%
proximal, 29% lateral and 32% distal (Fig. 7B, n=517). In the
presence of nocodazole, the distribution was not very different,
being 38% proximal, 26% lateral and 36% distal (Fig. 7D,
n=519). Thus, disruption of microtubules did not disturb the
randomly oriented nature of pseudopod formation in multipolar cells.
The most stabilized pseudopods among the multiple
pseudopods are represented as filled circles in Fig. 7B. The
majority of such pseudopods (about 66%) were located in the
proximal region of the cell relative to the tip of the microneedle. The percentages of pseudopods in the lateral and the distal
regions of the cells were 31% and 2%, respectively. Thus, the
pseudopods on the side of the cell where there was a higher
concentration of an attractant were preferentially stabilized.
These results are consistent with observations of Dictyostelium
amoebae made previously by Varnum-Finney et al. (1987b).
In marked contrast, no such asymmetrical distribution of the
selected pseudopods was seen in amoebae treated with nocodazole (Fig. 7D). The distribution of the most stabilized
pseudopods was random, being 29% proximal, 32% lateral and
39% distal. Thus, microtubules obviously have a pivotal role
in the selection of the chemotactically important pseudopods.
DISCUSSION
Microtubules are necessary for the appropriate
selection of pseudopods during chemotaxis
The chemotactic response of Physarum amoebae was achieved
by repeated bouts of linear locomotion and changes in
direction. Irrespective of whether microtubule inhibitors were
present or absent, the amoebae exhibited reorientation behavior
(Fig. 2). The pseudopod dynamics were not affected significantly by inhibitors (Figs 6, 7). However, the treated amoebae
did not exhibit any chemotactic responses (Figs 2, 3). These
results indicate that the microtubule system enabled the
amoebae to select the appropriate direction during the reorientation process. Such an ability can be attributed to the appropriate selection of pseudopods during the reorientation process
because the changes in the direction of locomotion occurred in
the multipolar phase of the cells (Figs 5, 6, 7). Fig. 7C implies
that unipolar cells were forced to migrate mostly in incorrect
directions as a result of the loss of ability to select the appropriate pseudopods, caused by microtubule inhibitors. These
results show that microtubules are necessary to stabilize selectively the pseudopod on the side of the cell that is closest to
the chemoattractant more effectively than the other
pseudopods. Such preferential stabilization leads to selection
2076 M. Ueda and S. Ogihara
Fig. 5. Typical changes in shape of chemotactically moving
Physarum amoebae during the reorientation process. Sequential
video images (A-G) and the corresponding drawings (H-N) show an
amoeba alternating between the unipolar and the multipolar shapes.
The hyaline pseudopods are represented by the gray regions in H-N.
The numbers near the pseudopods represent the areas in µm2. The
centroids of the cell and the pseudopods are indicated by the black
and the white dots, respectively. The cross (+) represents the tip of
the microneedle. First, the amoeba exhibited two pseudopods that
kept extending in almost opposite directions, while the tail (one of
the multiple pseudopods in the last reorientation) was retracted
toward the cell body (A-C and H-J). Next, one pseudopod on the side
of cell with a higher concentration of the attractant was maintained
and the other began to be retracted (D-E and K-L). In consequence,
the cell had a polarized shape with one pseudopod at the leading
edge and moved toward the source of the attractant (E-F and L-M).
The cell began to have two pseudopods again by splitting the
pseudopod at the leading edge and resumed its multipolar shape (G
and N). Note that the pseudopods in the multipolar cell were formed
and extended almost equally on both surfaces of the cell facing the
source and away from the source of the chemoattractant. Bar, 20 µm.
Time is given as minutes:seconds.
of the correct direction of locomotion as a consequence. We
speculate that a chemoattractant provides information for the
microtubule system, which uses the information to bring about
preferential stabilization of the appropriate pseudopods.
There are previous reports that microtubules are not obligatory for chemotaxis in amoeboid cells. For example, cytokineplasts, which are motile fragments produced from neutrophils,
lack microtubules and still exhibit chemotaxis (Malawista and
de Boisfleury, 1982). Directed migration of neutrophils is
barely affected by microtubule-disrupting reagents (Allan and
Wilkinson, 1978; Devreotes and Zigmond, 1988). However,
disruption of microtubules renders the chemotactic tracks more
multi-directional than those of control cells (Allan and
Wilkinson, 1978). Devreotes and Zigmond (1988) interpreted
this observation as an effect of microtubules on the cell via
inhibition of the extension of pseudopods from the rear and the
sides, such that the pseudopod facing the higher concentration
of an attractant is preferentially maintained. Microtubules seem
to be functioning in exactly this way in Physarum amoebae in
the unipolar phase. The view of Devreotes and Zigmond may
explain the lateral spreading of the pseudopod region in
unipolar cells treated with nocodazole. Also, the unipolar cells
resemble neutrophils in terms of shape (Fig. 5E,F), the preferential maintenance of one pseudopod in the direction of the tip
of the microneedle (Fig. 7A), and the relatively linear locomotion (Fig. 2).
There are various reports of heterogeneous interactions of
microtubules with multiple pseudopods. Asymmetric distribution of microtubules has been found in the growth cones of
living neurons in Xenopus (Tanaka and Kirschner, 1991) and
grasshopper (Sabry et al., 1991). Before any obvious
asymmetry can be seen in the overall shapes of the growth
cones, microtubules selectively invade the lamella that has
been formed in the future direction of migration. In newt
eosinophils, which also show changes in direction during
chemotactic migration (Brundage et al., 1991), Fay et al.
(1989) observed similar behavior of microtubules. That is,
microtubules in eosinophils seemed to move close to the cell
Pseudopod area (×10−3 pixels)
Role of microtubules in chemotaxis 2077
Time (seconds)
Fig. 6. Dynamics of the pseudopod area in the absence (A) and in the presence (B) of nocodazole. Amoebae were treated with 5 µM
nocodazole and stimulated with 10 mM glucose as described in the legend to Fig. 2. The areas of pseudopods were quantitated as described in
Materials and Methods. The thick lines and thin lines represent the areas of those pseudopods that were located either on the cell side closer to
the microneedle tip, or away from it, respectively. Amoebae had a unipolar shape during the period represented by the clear zones, and had a
multipolar shape in the shaded zones. In the lighter shaded zone in A, one pseudopod on the cell side away from the microneedle tip was split
into two, both of which decayed rapidly. During that time, the other pseudopod, which was initially located further from the chemoattractant,
continued to grow, even after the two small pseudopods on the other side disappeared. Note that those pseudopods closer to the microneedle tip
were preferentially stabilized in 5 cases out of 5 turns in A, and in only 3 cases out of 7 turns. One thousand pixels correspond to 37.0 µm2.
periphery prior to the formation of a pseudopod. These observations suggest that such asymmetric or localized distribution
of microtubules may be the key event in controlling the
stability of a pseudopod.
By what mechanism do microtubules control the stability of
the pseudopods? Microtubules may interact directly with
microfilaments and affect the dynamics of the pseudopods.
Interaction of microtubules with microfilaments has been
observed at the biochemical level (Nishida et al., 1981; Pollard
et al., 1984; Sattilaro, 1986; Itano and Hatano, 1991), but there
is little evidence for any functional interactions in vivo. Uyeda
and Furuya (1989) found that thick bundles of microfilaments
slide over bundles of microtubules in permeabilized flagellates
of P. polycephalum. It is also possible that the disruption of
the microtubule network has important consequences for
pseudopod dynamics via motor proteins. It has been shown that
the elimination of one of the myosin I genes from Dictyostelium amoebae results in an alteration in pseudopod
dynamics (Wessels et al., 1991). Fukui et al. (1990) reported
changes in the distribution of microtubules in multinucleated
cells induced by disruption of the myosin II gene in Dictyostelium amoebae, which may suggest mediation of the interaction between microtubules and pseudopods by myosin II.
Kinesin was shown to regulate pseudopodial activity in fibroblast (Rodionov et al., 1993). Alternatively, microtubules may
affect the adhesiveness of the pseudopods. Rinnerthaler et al.
(1988) reported close apposition of one end of a microtubule
to the site of cell-substratum contact at the leading edge of
chick fibroblasts. They suggested that microtubules may select,
stabilize and potentiate such contacts. It is quite possible that
stable contacts are favorable for stabilization of a pseudopod.
Therefore, if such contacts are formed beneath one of the
multiple pseudopods in the Physarum amoebae, it would be
maintained preferentially. There is also a possibility that the
selective stabilization of pseudopods may result from the
polarized transport of chemoattractant receptor to the appropriate pseudopod surface. Ojakian and Schwimmer (1992)
reported that the insertion of a glycoprotein into the cell’s
apical membrane is dependent on microtubules.
Two representative models of chemotaxis by
amoebae and the relevance of our observations
In our study, during chemotaxis, the amoebae of P. polycephalum formed multidirectional pseudopods (Figs 5, 6, 7),
which are apparently inconsistent with both the temporal
sensing model proposed by Gerisch et al. (1974) and the spatial
sensing model proposed by Zigmond (1974, 1978). Our observation can, however, be incorporated into either of the two
models if the models are slightly modified. The temporal
sensing model could incorporate the multipolar cells if the
putative functioning pilot pseudopods (Gerisch et al., 1974) are
equivalent to the well developed lamellipodia in the amoebae
in P. polycephalum. The spatial sensing model also seems to
provide an explanation for the multipolar shape if we make the
following assumptions: the spatial comparisons of the concentration of the attractant over the surface of the cell are suppressed for a time or, alternatively, an orchestrated response to
the chemotactic stimulation is postponed for a while in
Physarum amoebae. In either case, the chemotactic response
of the cell would be postponed, with resultant formation of
randomly directed multiple pseudopods.
Chemotaxis of the amoebae of P. polycephalum can be
compared with that of other amoebae as follows. Reorientation
of locomotion occurs during chemotaxis in a repeated manner.
As a result of each correct reorientation, the Physarum
amoebae move little by little up the gradient of the chemoattractant. The locomotion of single amoebae has been reported
to be rhythmic in polymorphonuclear leukocytes (Murray et
2078 M. Ueda and S. Ogihara
y (µm)
attractant could be assessed by either a spatial or a temporal
mechanism, as described above. Swanson and Taylor (1982)
claim that the extension of pseudopods from the cell surface
towards cAMP is only a local response by the cell and that the
amoeba’s cytoplasm then has to be coordinated in the cell as
a whole to achieve the directed movements. They suggest that
such cytoplasmic coordination plays a prominent role, at least
in situations where more than two pseudopods sense the
chemotactic stimulus, and they speculate that this coordination
is dependent on microtubules. The conclusion of the present
study is consistent with their view that Dictyostelium amoeba
chemotaxis is achieved by these two sequential events, a local
cell surface response followed by a microtubule-dependent
global response. Although the underlying molecular
mechanism for the dependence of the selective stabilization of
the appropriate pseudopod on microtubules remains to be
clarified, this work provides evidence that microtubules are
necessary for a process related to both the establishment of
directionality and the sensing mechanism in amoeba chemotaxis.
x (µm)
Fig. 7. Positions of pseudopods in unipolar cells (A and C) and
multipolar cells (B and D). The y axis is set so that both the tip of the
microneedle and the cell centroid fall on the y axis. The positions of
the centroids of individual pseudopods are shown as open and filled
circles. The most stabilized pseudopods (see text) among the many
pseudopods are represented by filled circles (B and D). (A and B)
Positions of pseudopods in the normal amoebae. Unipolar cells
mainly have a pseudopod in the proximal region of the cell (A),
while multipolar cells extend multiple pseudopods in random
directions. The selectively stabilized pseudopods are mostly in the
proximal region (B). (C and D) Positions of pseudopods in amoebae
after treatment with 5 µM nocodazole or 10 µM mebendazole.
Unipolar cells mostly have one pseudopod located in the lateral
region (C). The stabilized pseudopods are distributed randomly
around the cell centroid (D), indicating that microtubules are
necessary for preferential stabilization of pseudopods. The positions
of pseudopods with respect to the tip of the microneedle were scored
by the method illustrated in Fig. 1. This scoring process contains
some errors that arise from variation in the gradient of
chemoattractants from true concentric circles as shown in Fig. 2A.
To minimize error in the data interpretation, however, we narrowed
the angles of both the proximal and the distal regions to 120° instead
of 180°. There were no significant differences in the distribution of
these plots after stimulation with glucose or with cAMP.
al., 1992) and Dictyostelium discoideum (Wessels et al., 1994).
In each cycle of the rhythm, two events seem necessarily to
occur successively to achieve chemotaxis of Physarum
amoebae. First, the pseudopods in the multipolar cells are
formed in random directions relative to the gradient of the
chemoattractant. This may have relevance to the finding by
Varnum-Finney et al. (1987b) that pseudopods are formed as
frequently on the surface of a Dictyostelium amoeba facing the
source as they are on the surface away from the source of
chemoattractant, although the Dictyostelium amoebae did not
assume a multipolar shape in the gradient. The ensuing and the
most crucial step in the chemotaxis of Physarum amoebae is
the correct selection of the appropriate pseudopod from the
many pseudopods in the reorientation process. The latter event
was shown here to be microtubule-dependent. During this
decision-making process, the concentration gradient of the
We thank Dr Kimiko Murakami-Murofushi of Ochanomizu University for a gift of amoebae of Physarum polycephalum, strain J. We
thank Mr Yasuaki Oku and Mr Ryo Masuda for invaluable help with
the computer program used for calculation of cell areas. We are most
grateful to Mr Hiromi Sesaki and Ms Miki Matsumura for valuable
discussions. M. Ueda is supported by a predoctoral fellowship from
the Japan Society for the Promotion of Science for Japanese Junior
Scientists.
REFERENCES
Allan, R. B. and Wilkinson, P. C. (1978). A visual analysis of chemotactic and
chemokinetic locomotion of human neutrophil leucocytes. Use of a new
chemotaxis assay with Candida albicans as gradient source. Exp. Cell Res.
111, 191-203.
Bennett, H. and Condeelis, J. (1984). Decoration with myosin subfragment-1
disrupts contacts between microfilaments and the cell membrane in isolated
Dictyostelium cortices. J. Cell Biol. 99, 1434-1440.
Brundage, R. A., Fogarty, K. E., Tuft, R. A. and Fay, F. S. (1991). Calcium
gradients underlying polarization and chemotaxis of eosinophils. Science
254, 703-706.
Carboni, J. M. and Condeelis, J. S. (1985). Ligand-induced changes in the
location of actin, myosin, 95k(α-actinin), and 120k protein in amebae of
Dictyostelium discoideum. J. Cell Biol. 100, 1884-1893.
Devreotes, P. N. and Zigmond, S. H. (1988). Chemotaxis in eukaryotic cells:
a focus on leukocytes and Dictyostelium. Annu. Rev. Cell Biol. 4, 649-686.
Euteneuer, U. and Schliwa, M. (1984). Persistent, directional motility of cells
and cytoplasmic fragments in the absence of microtubules. Nature 310, 5861.
Fay, F. S., Carrington, W., Lifshitz, L. M. and Fogarty, K. (1989). Threedimensional analysis of molecular distribution in single cells using the
digital imaging microscope. In SPIE Proceedings Series, vol. 1161 (ed.
J. E. Wampler), pp. 12-23. Soc. Photoptical Instrumentation Engineers,
Washington.
Fukui, Y., De Lozanne, A. and Spudich, J. A. (1990). Structure and function
of the cytoskeleton of a Dictyostelium myosin-defective mutant. J. Cell Biol.
110, 367-378.
Gerisch, G., Malchow, D. and Hess, B. (1974). Cell communication and
cyclic-AMP regulation during aggregation of the slime mold, Dictyostelium
discoideum. In Biochemistry of Sensory Functions (ed. L. Jaenicke), pp. 279298. Springer-Verlag, New York.
Gerisch, G., Hulser, D., Malchow, D. and Wick, U. (1975). Cell
communication by periodic cyclic-AMP pulses. Phil. Trans. R. Soc. B. Biol.
Sci. 272, 181-192.
Grynkiewicz, G., Poenie, M. and Tsien, R. Y. (1985). A new generation of
calcium indicators with greatly improved fluorescence properties. J. Biol.
Chem. 260, 3440-3450.
Role of microtubules in chemotaxis 2079
Itano, N. and Hatano, S. (1991). F-actin bundling protein from Physarum
polycephalum: purification and its capacity for co-bundling of actin filaments
and microtubules. Cell Motil. Cytoskel. 19, 244-254.
Laclette, J. P., Guerra, G. and Zetina, C. (1980). Inhibition of tubulin
polymerization by mebendazole. Biochem. Biophys. Res. Commun. 92, 417423.
Malawista, S. E. and de Boisfleury, A. (1982). The cytokineplast: purified,
stable, and functional motile machinery from human blood
polymorphonuclear leukocytes. J. Cell Biol. 95, 960-973.
Murray, J., Vawter-Hugart, H., Voss, E. and Soll, D.R. (1992). Threedimensional motility cycle in leukocytes. Cell Motil. Cytoskel. 22, 211-223.
Nishida, E., Kuwaki, T. and Sakai, H. (1981). Phosphorylation of
microtubule-associated protein (MAPs) and pH of the medium control
interaction between MAPs and actin filaments. J. Biochem. 90, 575-578.
Ojakian, G. K. and Schwimmer R. (1992). Antimicrotubule drugs inhibit the
polarized insertion of an intracellular glycoprotein pool into the apical
membrane of Madin-Darby canine kidney (MDCK) cells. J. Cell Sci. 103,
677-687.
Pollard, T. D., Selden, S. C. and Maupin, P. (1984). Interaction of actin
filaments with microtubules. J. Cell Biol. 99, 33s-37s.
Rinnerthaler, G., Geiger, B. and Small, J. V. (1988). Contact formation
during fibroblast locomotion: involvement of membrane ruffles and
microtubules. J. Cell Biol. 106, 747-760.
Rodionov, V. I., Gyoeva, F. K., Tanaka, E., Bershadsky, A. D., Vasiliev, J.
M. and Gelfand, V. I. (1993). Microtubule-dependent control of cell shape
and pseudopodial activity is inhibited by the antibody to kinesin motor
domain. J. Cell Biol. 123, 1811-1820.
Sabry, J. H., O’Connor, T. P., Evans, L., Toroian-Raymond, A., Kirschner,
M. and Bentley, D. (1991). Microtubule behavior during guidance of
pioneer neuron growth cones in situ. J. Cell Biol. 115, 381-395.
Sattilaro, R. F. (1986). Interaction of microtubule-associated protein 2 with
actin filaments. Biochemistry 25, 2003-2009.
Stossel, T. P. (1989). From signal to pseudopod. How cells control cytoplasmic
actin assembly. J. Biol. Chem. 264, 18261-18264.
Swanson, J. A. and Taylor, D. L. (1982). Local and spatially coordinated
movements in Dictyostelium discoideum amoebae during chemotaxis. Cell
28, 225-232.
Tanaka, E. M. and Kirschner, M. W. (1991). Microtubule behavior in the
growth cones of living neurons during axon elongation. J. Cell Biol. 115,
345-363.
Taylor, D. L. and Condeelis, J. S. (1979). Cytoplasmic structure and
contractility in amoeboid cells. Int. Rev. Cytol. 56, 57-144.
Uyeda, T. Q. P. and Furuya, M. (1989). Evidence for active interactions
between microfilaments and microtubules in myxomycete flagellates. J. Cell
Biol. 108, 1727-1735.
Varnum, B., Edwards, K. B. and Soll, D. R. (1986). The developmental
regulation of single-cell motility in Dictyostelium discoideum. Dev. Biol.
113, 218-227.
Varnum-Finney, B., Edwards, K. B., Voss, E. and Soll, D. R. (1987a).
Amoebae of Dictyostelium discoideum respond to an increasing temporal
gradient of the chemoattractant cAMP with a reduced frequency of turning:
evidence for a temporal mechanism in amoeboid chemotaxis. Cell Motil.
Cytoskel. 8, 7-17.
Varnum-Finney, B. J., Voss, E. and Soll, D. R. (1987b). Frequency and
orientation of pseudopod formation of Dictyostelium discoideum amoebae
chemotaxing in a spatial gradient: further evidence for a temporal
mechanism. Cell Motil. Cytoskel. 8, 18-26.
Wakasugi, M. and Ohta, J. (1973). Studies on the amoebo-flagellate
transformation in Physarum polycephalum. Bot. Mag. Tokyo 86, 299-308.
Wessels, D., Murray, J., Jung, G., Hammer, J. A. III and Soll, D. R. (1991).
Myosin IB null mutants of Dictyostelium exhibit abnormalities in motility.
Cell Motil. Cytoskel. 20, 301-315.
Wessels, D., Vawter-Hugart, H., Murray, J. and Soll, D. R. (1994). Threedimensional dynamics of pseudopod formation and the regulation of turning
during the motility cycle of Dictyostelium. Cell Motil. Cytoskel. 27, 1-12.
Zigmond, S. H. (1974). Mechanisms of sensing chemical gradients by
polymorphonuclear leukocytes. Nature 249, 450-452.
Zigmond, S. H. (1978). Chemotaxis by polymorphonuclear leukocytes. J. Cell
Biol. 77, 269-287.
Zigmond, S. H., Levitsky, H. I. and Kreel, B. J. (1981). Cell polarity: an
examination of its behavioral expression and its consequences for
polymorphonuclear leukocyte chemotaxis. J. Cell Biol. 89, 585-592.
(Received 11 November 1993 - Accepted 18 April 1994)