Role of Microtubules in Stimulating
Cytokinesis in Animal Cells”
E. D. SALMON AND S. M. WOLNIAK
Department of Biologv
University of North Carolina
Chapel Hill, North Carolina 27599
and
Department of Botany
University of Maryland
College Park, Maryland 20742
Cytokinesis in animal cells is produced by actom yosin-based contraction in the cell
cortex which constricts the cell surface to form a furrow that cleaves the cell into
two.’-’ The circumferential concentration of cortical actin and myosin I1 filaments
seen at the leading edge of an established furrow has been termed the “contractile
ring” in symmetrically cleaving cells4 or the “contractile arc” for asymmetrical
~leavage.~
The dynamic, transient contractile ring or arc appears at the onset of
furrowing and disappears upon the completion of
Remarkably, any region
of the cell cortex usually is capable of forming a furrow after the onset of anaphase?
The pioneering studies by Dan, Rappaport, and Hiramoto of cytokinesis in developing marine embryos have shown that the microtubule arrays of the anaphase
mitotic spindle and asters stimulate and orient furrow formation.’s2s6
Furrowing usually
occurs along the equatorial plane of the spindle, thus insuring that chromosomes
segregated by the mitotic spindle become contained in different cells. In the large firstdivision cells of developing marine embryos, the distance between the spindle poles
and the cell surface can be greater than 50 pm. During most of the mitotic phase of
the cell cycle, microtubules rarely elongate beyond the spindle region, and few or no
microtubles extend from the spindle poles to the cell
(FIGS. 1 and 2).
Following anaphase onset, however, astral microtubules begin to elongate from the
spindle poles outwards to the cell surface, reforming the interphase cytoplasmic microtubule complex (CMTC) (FIGS.l and 2). No furrowing occurs if these microtubule
arrays are abolished before the onset of furrowing.’o’l’In contrast, once the furrow
apparatus is well formed, cytokinesis continues even in the absence of the microtubule
complex.’o~l’The microtubule-dependent factor that stimulates furrow formation is
unknown. The basic mechanism by which the furrow apparatus assembles is also
controversial,’ and opposing views have been presented by Rappap~rt:.~~.‘’White and
and other investigators in this volume.
In this paper, we describe and discuss the results of two experimental approaches
designed to modify the dynamics of microtubule assembly during mitosis and cytokinesis. In the first experimental series, hydrostatic pressure was used to suppress
‘This work was supported by NIH Grant GM-24364 t o E.D.S. and NSF Grant DCB
87-00422 to S.M.W.
88
SALMON & WOLNIAK: MICROTUBULE DYNAMICS
89
microtubule assembly in first division embryos of the sea urchin Lytechinus vnriegatus.
In the second series of experiments, taxol was used t o stabilize microtubule assembly
in first division embryos of the sand dollar Echinuruchnis purmu. Although the pressures and the concentrations of taxol in these experiments d o not significantly delay
the cell cycle, they both inhibit cytokinesis, but in different ways.
Complex (CMTC)
2-
CLEAVAGE
FIGURE 1. Rearrangements of microtubule arrays during the cell cycle in a typical animal
tissue cell. This diagram shows the major features inferred from immunofluorescence and electron
micrograph^.^^ It does not provide a comprehensive description of all the microtubules in the
cell. All polar microtubules are oriented with their “plus” or fast-growing ends distal from their
centrosomes or spindle poles. The cytoplasmic microtubule complex (CMTC) is a monopolar
microtubule array, and the mitotic spindle is a bipolar array. The great majority of microtubules
in both these arrays never grow to a constant length. Instead, they exhibit dynamic instability
where nucleation occurs mainly at the centrosome or spindle poles, tubulin association and
dissociation reactions occur at plus ends, and an end alternates between persistent, constant
velocity, phases of growth or shortening with abrupt, stochastic transitions between phases
(reviewed by Cassimeris et al. ”). In interphase newt tissue cells, the growth velocities are about
7 wm/min and shortening velocities are about 17 pm/min, based on direct observation of
individual microtubules in living cells.Js
RESULTS AND DISCUSSION
Hydrostatic Pressure Studies
Marsland” and co-workers showed many years ago that cytokinesis in marine
embryos can be inhibited reversibly by hydrostatic pressure in two ways: directly, by
90
ANNALS NEW YORK ACADEMY OF SCIENCES
SALMON & WOLNIAK MICROTUBULE DYNAMICS
91
pressure-induced disruption of the furrow apparatus, and indirectly, by pressureinduced disassembly of the microtubule arrays required to initiate the formation of
the furrow apparatus.
Pressurization of cleaving cells slows furrow propagation, and furrows regress at
pressures greater then 400-600 atm at normal physiologic temperature.” The inhibition
of furrowing, which occurs when cleaving cells are pressurized, correlates with disruption of cortical cytoskeletal structure.”
Lower pressures, which do not inhibit cytokinesis once furrowing has started, can
inhibit furrow formation when pressurization occurs during mitosis or early anaphase.
The assembly of the labile microtubules of the mitotic spindle and the interphase
CMTC is very sensitive to hydrostatic pressure.’sz’ Spindle microtubule assembly
decreases rapidly (10-20-s time constants) and reversibly in response to moderate
3, application of 200 atm pressure
pressurization.21In the experiment shown in FIGURE
at metaphase did not suppress microtubule assembly sufficiently to block chromosome
segregation, but it did block cytokinesis. The asters remained small and did not appear
to extend to the cell surface as they normally would after the onset of anaphase.
Furrow formation did not occur. Instead, at the normal time of cleavage, the cell
surface became wrinkled (FIG.3, 11.5 min). In contrast, pressurization of cells to
200 atm during cytokinesis slowed, but did not inhibit, furrow propagation or division.
These observations are explained most simply by the general concept (reviewed
by Rappaport’) that in the large dividing cells of embryos, furrow induction requires
microtubule growth to near the cell surface shortly after anaphase onset. If the extent
of microtubule growth is suppressed, as occurs under 200 atm pressure, then furrow
formation is inhibited.
In other experiments, pressure was returned to atmospheric at various times after
control cells began furrowing to permit normal microtubule assembly. If pressure was
released within about 10 minutes after the induction of furrows in control cells, the
asters extended rapidly towards the cell surface, furrows began to form, and the cells
divided normally into two (data not shown). After longer delays, cytokinesis was not
completed, and furrow induction became weaker.
These results indicate that furrow establishment can occur only during a limited
period following anaphase onset. Outside this period of the cell cycle, microtubules
grow outward to near the cell surface, but furrows are not initiated. Perhaps the
cortex becomes insensitive to the signal transmitted by the polar microtubule arrays,
or the signal is no longer active.
3, pressure was returned to atmospheric at 25
In the example shown in FIGURE
minutes after nuclear envelope breakdown (NEB) of the first division and 11 minutes
before NEB of the second mitosis. All cells in the chamber contained two well-separated
nuclei. At second mitosis, two normal second-division bipolar spindles formed in each
cell, and each spindle progressed through mitosis with normal pattern and timing of
microtubule assembly similar to that shown in FIGURE
2. At cytokinesis, furrows
formed initially in the plane of the spindle equators (FIG. 3, 42.5 min) and, a short
time later, in the plane between adjacent spindle asters. This resulted in division into
four cells and normal development through the subsequent early divisions (FIG.3,
75 min).
Could the signal for furrow induction come from the chromosomes as they enter
anaphase?” In support of this idea, division frequently occurs along a plane through
the spindle equator or metaphase plate, and the metaphase chromosomes are closer
to the equatorial cell surface than to the polar cell surface. However, we observed
that furrows can occur in sea urchin embryos between two adjacent asters that never
had a central spindle, metaphase plate, or chromosomes between them. A similar
result was seen by Rappaport in the now famous “Torus” experiment: by Sluder et
92
ANNALS NEW YORK ACADEMY OF SCIENCES
SALMON & WOLNIAK. MICROTUBULE DYNAMICS
93
01. 23 in dividing, but enucleated, sea urchin eggs, and during the initial cellularization
of the blastoderm layer of insect embry0s.2~Thus, in the large early division cells of
embryos, the microtubule-dependent signal for furrow induction is associated with
the spindle asters and not the chromosomes or the central spindle.
Taxol Studies
Taxol promotes and stabilizes microtubule assembly in vivo and in vitro against a
wide variety of agents, including cooling and high calcium, which normally induce
rapid microtubule dep~lymerization?~.~~
Taxol binds to a site on tubulin when tubulin
polymerizes into the lattice of a microt~bule.2”~~
When we incubated sand dollar eggs with 10 pM taxol (plus 0.1% DMSO, which,
by itself, has no effect on development) at prophase before the first division, the eggs
made mitotic spindles that were noticeably larger and more birefringent than normal.
These spindles exhibited the normal changes in the pattern of microtubule assembly
during mitosis and cytokinesis, but the central spindle persisted for extended periods.
Chromosome condensation and segregation appeared to proceed normally and approximately on the same schedule as that of untreated eggs (data not shown). Furrows
formed, but cleavage was occasionally abortive.
At second division, spindle formation was very abnormal in the continued presence
of 10 p M taxol. Monasters or biasters formed (FIG.4) which were distinct from
normal anaphase spindle asters in at least three ways. First, they were very birefringent
(6-10 nm maximum retardation versus 1-3 nm retardation for untreated asters). The
high birefringent retardation was correlated with an abnormally high concentration
of microtubules (electron micrographs of thin sections through the large taxol asters
show a high density of parallel, radial arrays of microtubules and smooth endoplasmic
reticulum [data not shown]). Second, the asters appeared to be compact, almost
spherical structures 30-40 pm in diameter for biasters and larger for monasters. There
was a sharp boundary between the periphery of the aster and the cytoplasm (FIG.
4). The sharp periphery of the taxol asters was particularly apparent when the taxol
asters were freed from cells with detergent lysis buffer^*^,^^ (FIG. 5). Third, the
microtubules in the taxol asters did not appear to elongate toward the cell surface
during cytokinesis (FIG.4) as occurs in untreated cells (FIG.2).
Furrow formation in the presence of 10 pM taxol during second- and third-division
4. When the taxol biaster
cycles has several interesting features, as shown in FIGURE
was in the center of the cell, so that there was a great distance between the edges of
the asters and the periphery of the cell, furrow formation was poor. However, furrows
formed when the biaster was near the edge of the cell, but not in contact with the
lateral cell surface, as shown for the lower biaster in FIGURE
4. The equator of the
lower biaster is about 36 pm from the leading edge of the cell. The asters are about
31 pm in diameter, and from the flattened diameter of the cell in FIGURE
4a and a
cell volume of 1.5 x lo6 pm3 the height of the flattened cell is estimated to be 45
pm In this case, the furrow propagated inward at the spindle equator (FIG.4b and
c), pushing the taxol stabilized biaster towards the center of the cell and eventually
almost splitting the biaster into two (FIG.4c). Eventually the furrow regressed (FIG.
4d and e) before the time of third division when furrowing began again at the biaster
and between the biaster and another split monaster in the same cytoplasm (FIG.4f).
These results show that neither dynamic microtubule assembly nor microtubule
elongation is required for furrow formation. After their assembly in second mitosis,
94
ANNALS NEW YORK ACADEMY OF SCIENCES
the taxol asters and biasters did not change birefringence or size after anaphase onset.
Most likely, the elongation of astral microtubules, which normally occurs after anaphase onset (FIGS.l and 2), provide a mechanism for increasing the concentration
of the furrow stimulatory factors at the cell surface. Microtubules could serve as
polarized roadways for active translocation of stimulatory factors from the mitotic
centers to the cell surface. Kinesin may be involved in this movement, because it can
actively move organelles along taxol-stabilized microtubules toward their plus ends:'
and the plus ends of astral microtubules are oriented toward the cell surface?' An
important possibility is that microtubule-activated enzymes (e.g., kinases) generate
the active form of the stimulatory factors. Unfortunately, the nature of the stimulatory
factors is still unknown.
4 also shows that furrow formation at the lateral cell
The experiment in FIGURE
surface between the biasters occurs without direct contact between microtubule ends
and the lateral cell surface. The site of furrow initiation at the lateral cell surface was
greater than 20 pm away from the periphery of the asters and from the upper and
lower cell surfaces closest to the asters.
FIGURE 4. The effects of 10 p M taxol upon spindle assembly and cytokinesis as seen from
time-lapse micrographs of a living second-division mitotic cell of the sand dollar Echinuruchnis
purmu. The cells were incubated in taxol 10 minutes before first division as described in the
text. The cell is flattened between a slide and ~overslip.*~
Frames are sequential. Frames a, b, c,
e, and f a r e polarization micrographs taken as described in FIGURE
1 by Salmon and W~lniak.'~
Frame d is a conventional DIC micrograph. Between frames c and d the degree of flattening
was decreased. The onset of furrowing can be seen at about 5 o'clock on the circumference of
the cell in frame b. The leading edge of the furrow propagates inward, pushing the stable biaster
inwards and eventually propagating almost completely through the equatorial region of the
biaster. Scale = 50 pm.
SALMON & WOLNIAK MICROTUBULE DYNAMICS
95
FIGURE 5. Polarization (a) and DIC (b) micrographs of one of the large asters induced by
10 p M taxol in the second-division cells described in FIGURE
4. The aster was lysed from a cell
by perfusion of the preparation between slide and coverslip as described by Salmon and Wolniak,2s
using an EGTA perfusion bufferz9containing 10 pM taxol and 1% NP-40. Note that the diameter
of the aster is constant and that few fibers extend beyond the perimeter. Electron micrographs
show a high density of microtubules in a radial array (data not shown). Scale = 25 pm.
How is furrow formation initiated in the lateral cell cortex and aligned with the
equator of the taxol biaster when the lateral cell surface is 20 pm or more away from
the periphery of the taxol asters? One possibility is that furrow initiation in the very
flattened cell shown in FIGURE
4 occurs by the astral relaxation-cortical flow mechanism described by White and co-workers.1e’6 This model proposes that the microtubule-dependent signals initiate relaxation of cortical tension in regions of the cell
surface near the polar periphery of the asters. Subsequent flow of cortical contractile
elements from those regions to the equatorial lateral cell surface between the asters
produces the furrow apparatus.
However, Rappap~rt~.’~*’’
has shown in sand dollar eggs that astral microtubules
induce cortical contraction and not relaxation as required by the polar-relaxation,
cortical flow model. Most available evidence also supports some type of “equatorialstimulation” (and perhaps subsequent cortical flow) mechanism of furrow induction
rather than a “polar-relaxation” mechanism? If the microtubule-dependent signals
for furrow induction stimulate cortical contraction, then the furrow induced by the
4 cannot be explained by “equatorial-stimulation”
taxol biaster in the cell in FIGURE
models of cytokinesis which depend on close contact of microtubule ends with the
cell cortex in the furrow region?” Furrow induction must involve stirnulatory factors
that can move over 20 pm distances in the cytoplasm independently of microtubules.
Devore et al. 32 recently proposed a model for astral stimulation of cortical contraction and furrow formation in animal cells. This model currently requires that the
astral microtubule arrays are space filling and terminate close to the cell cortex. In
their scheme, the peripheral ends of astral microtubules provide a continuous flux of
stimulatory molecules that diffuse short distances through the cytoplasm to sites at
the cell cortex. Furrows form at sites on the cell surface where the highest steadystate concentrations of stimulatory factors occur.
It will be interesting to see how well this model predicts our taxol experimental
results using the computer simulation methods of Devore et al.” or Harris and
96
ANNALS NEW YORK ACADEMY OF SCIENCES
G e ~ a l tThe
. ~ ~20-pm distance between the periphery of the taxol asters and the lateral
cell surface is much greater than that proposed in the current Devore et aZ.” model.
A reasonable assumption is that the concentration of stimulatory factors leaving the
periphery of the astral microtubule arrays is proportional to microtubule density. The
density of microtubules in taxol asters is likely to be more than an order of magnitude
higher than that in normal asters, which could account for the ability of the taxol
biasters to induce furrow formation at great distances from the cell surface.
SUMMARY
The initiation of furrow formation is disrupted when microtubule elongation to
the cell surface is inhibited either by promoting microtubule disassembly with hydrostatic pressure or by stabilizing the mitotic astral microtubules with taxol. The pressure
studies confirmed Rappaport’s earlier observation that stimulation of furrow formation
is produced by a pair of asters and does not require chromosomes or a central spindle
in the large dividing cells of echinoderm embryos. The taxol studies showed that
furrow formation occurs between two stable asters when the asters are within 20 pm
of the lateral cell surface. Furrow formation at the lateral edge of the cell does not
appear to require microtubule dynamics, microtubule elongation, or contact of microtubule ends with the lateral cell surface. Microtubules may function to increase
the concentrations of the active forms of diffusible stimulatory factors that interact
with receptors at the inner cell surface to initiate the formation of the furrow apparatus
and activate contraction.
ACKNOWLEDGMENTS
We congratulate Gary Conrad and Tom Schroeder for an outstanding conference
on cytokinesis. We thank Albert Harris for sharing his thoughts about the mechanism
of cytokinesis. We also thank Nancy Salmon for editorial assistance, and Susan
Whitfield and Vicki Petrie for their help with the illustrations for this paper.
REFERENCES
1. MABUCHI,I. 1986. Biochemical aspects of cytokinesis. Int. Rev. Cytol. 101: 175-213.
2. RAPPAPORT,
R. 1986. Establishment of the mechanism of cytokinesis in animal cells. Int.
Rev. Cytol. 105: 245-281.
3. SALMON,
E. D. 1989. Cytokinesis in animal cells. Curr. Opinion Cell Biol. 1: 541-547.
4. SCHROEDER,
T. E. 1972. The contractilering. 11. Determining its brief existence, volumetric
changes, and vital role in cleaving Arbucia eggs. J. Cell Biol. 53: 419-434.
5. MABUCHI,
I., S. TSUKITA,
S. TSUKITA
& T. SAWAI.1988. Cleavage furrow isolated from
newt eggs: Contraction, organization of the actin filaments, and protein components of
the furrow. Proc. Natl. Acad. Sci. USA 85: 5966-5970.
6. HIRAMOTO,
Y. 1981. Mechanical properties of dividing cells. In Cellular Dynamics: Mi-
SALMON & WOLNIAK MICROTUBULE DYNAMICS
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
97
tosis/Cytokinesis. A. M. Zimmerman and A. Forer, eds.: 397-418. Academic Press,
New York, NY.
BALCZON,R. & G. SCHATTEN.1983. Microtubule-containing detergent-extracted cytoskeletons in sea urchin eggs, from fertilization through cell division: Antitubulin immunofluorescence microscopy. I. Cell Motility 3 213-226.
SCHROEDER,
T. E. 1987. Fourth cleavage of sea urchin blastomeres: Microtubule patterns
and myosin localization in equal and unequal cell divisions. Dev. Biol. 124 9-22.
VANDRE,D. D., P. KRONEBUSCH
& G. G. BORISY.1984. Interphase-mitosis transition:
Microtubule rearrangements in cultured cells and sea urchin eggs. In Molecular Biology
of the Cytoskeleton. G. G. Borisy, D. W. Cleveland & D. B. Murphy, eds.: 1-16. Cold
Spring Harbor Laboratory, Cold Spring Harbor, New York.
HAMAGUCHI,
Y. 1975. Microinjection of colchicine into sea urchin eggs. Dev., Growth &
Differ. 17: 111-117.
MULLINS,J. M. & J. A. SNYDER.1981. Anaphase progression and furrow establishment
in nocodozole arrested PtKl cells. Chromosoma 83: 493-505.
RAPPAPORT,
R. 1988. Surface behavior in artificially constricted sand dollar eggs and egg
fragments. J. Exp. Zool. 246: 253-257.
RAPPAPORT,
R. & B. N. RAPPAPORT.1988. Reversing cytoplasmic flow in nucleated,
constricted sand dollar eggs. J. Exp. Zool. 247: 92-98.
BRAY,D. & J. G. WHITE. 1988. Cortical flow in animal cells. Science 239 883-888.
WHITE, J. G. 1986. The astral relaxation theory of cytokinesis revisited. BioEssays 2:
267-272.
WHITE,J. G. & G. G. BORISY.1983. On the mechanism of cytokinesis in animal cells. J.
Theor. Biol. 101: 289-316.
MARSLAND,
D. 1970. Pressure-temperature studies on the mechanisms of cell division. In
High Pressure Effects on Cellular Processes. A. M. Zimmerman, ed.: 260-306. Academic
Press. New York, London.
BEGG,D. A., E. D. SALMON
& H. A. HYAIT. 1983. The changes in structural organization
of actin in the sea urchin egg cortex in response to hydrostatic pressure. J. Cell Biol.
97: 1795- 1805.
BOURNS,B., S. FRANKLIN,
L. CASSIMERIS
& E. D. SALMON.1988. High hydrostatic
pressure effects in vivo: Changes in cell morphology, microtubule assembly and actin
organization. Cell Motil. & Cytoskel. 1 0 380-390.
OTTER,T., B. BOURNS,S. FRANKLIN,
C. RIEDER& E. D. SALMON.1986. Hydrostatic
pressure effects on cytoskeletal organization and ciliary motility: A calcium hypothesis.
In Current Perspectives in High Pressure Biology. H. W. Jannasch, A. M. Zimmerman
& R. E. Marquis, eds.: 75-93. Academic Press. New York.
SALMON,
E. D. 1975. Spindle microtubules: Thermodynamics of in vivo assembly and role
in chromosome movement. Ann. N.Y. Acad. Sci. 2 5 3 383-406.
COOKE,C. A., M. M. S. HECK& W. C. EARNSHAW.
1987. The inner centromere protein
(INCENP) antigens: Movement from inner centromere to midbody during mitosis. J.
Cell Biol. 105 2053-2067.
SLUDER,G., F. J. MILLER& C. L. RIEDER. 1986. The reproduction of centrosomes:
Nuclear verses cytoplasmic controls. J. Cell Biol. 103: 1873- 1881.
FULLILOVE,
S. L. & A. G. JACOBSON.
1971. Nuclear elongation and cytokinesis in Drosophila montana. Dev. Biol. 2 6 560-577.
SALMON,
E.D. & S. WOLNIAK.1984. Taxol stabilization of mitotic spindle microtubules:
Analysis using calcium-induced depolymerization. Cell Motil. 4 155- 167.
SCHAITEN, G., H. SCHATTEN,T. H. BESTOR& R. BALCZON.1982. Taxol inhibits the
nuclear movements during fertilization and induces asters in unfertilized sea urchin eggs.
J. Cell Biol. 9 4 455-465.
MANFREDI,
J. J., J. PARNESS& S. B. HORWITZ.1981. Taxol binds to cellular microtubules.
J. Cell Biol. 9 4 688-696.
PARNESS,
J. & S. B. HORWITZ.1981. Taxol binds to polymerized tubulin in vitro. J. Cell
Biol. 91: 479-487.
SALMON,
E. D. & R. R. SEGALL.1980. Calcium-labile mitotic spindles isolated from sea
urchin eggs (Lytechinus variegatus). J. Cell Biol. 8 6 355-365.
98
ANNALS NEW YORK ACADEMY OF SCIENCES
30. VALE,R. D. 1987. Intracellular transport using microtubule-based motors. Ann. Rev. Cell
Biol. 3: 347-378.
31. CASSIMERIS,L. U., R. A. WALKER,N. K. PRYER& E. D. SALMON.1987. Dynamic
instability of microtubules. BioEssays 7: 149- 154.
32. DEVORE,J. J., G. W. CONRAD& R. RAPPAPORT.1989. A model for astral stimulation
of cytokinesis in animal cells. J. Cell Biol. 2225-2232.
33. HARRIS,A. K. & S. L. GEWALT.1989. Simulation testing of mechanisms for inducing
the formation of the contractile ring in cytokinesis. J. Cell Biol. 2215-2224.
34. SALMON,
E. D. & P. WADSWORTH.
1986. Fluorescence studies of tubulin and microtubule
dynamics in living cells. In Applications of Fluorescence in the Biomedical Sciences. D.
L. Taylor, A. S. Waggoner, F. Lanni, R. F. Murphy & R. R. Birge, eds.: 377-403. Alan
R. Liss, Inc. New York.
35. CASSIMERIS,L., N. K. PRYER& E. D. SALMON.1988. Real-time observations of microtubule dynamic instability in living cells. J. Cell Biol. 107: 2223-2231.
36. SALMON,E. D. & G. W.ELLIS.1975. A new miniature hydrostatic pressure chamber for
microscopy. Strain-free optical glass windows facilitate phase-contrast and polarizedlight microscopy of living cells. Optional fixture permits simultaneous control of pressure
and temperature. J. Cell Biol. 6 5 587-602.