Review articles
Balanced regulation of microtubule
dynamics during the cell cycle:
a contemporary view
Søren S.L. Andersen
Summary
Assembly of mitotic and meiotic spindles into an elliptical bipolar shape is an
example of morphogenetic processes that involve local chromosomal regulation
of microtubule dynamics for proper spatial microtubule assembly. Global microtubule dynamics during the cell cycle and local microtubule dynamics during
spindle assembly are regulated by a balance between microtubule stabilizing and
destabilizing factors. How a chromosome-induced phosphorylation gradient may
be generated and modulate spindle microtubule assembly through balanced
regulation of the activity of microtubule-associated proteins and Stathmin/Op18 is
analyzed. BioEssays 1999;21:53–60. r 1999 John Wiley & Sons, Inc.
Introduction
Microtubules (MTs) represent one of the polymer systems of
the eukaryotic cytoskeleton. They are essential for a wide
variety of cellular functions, including mitosis, transport, cell
motility, cell shape, and polarity. How the cytoskeleton organizes itself into the dynamic asymmetric arrays required for
function is one of the most challenging questions in cell biology.
During interphase, MTs form a radial array emanating from
the centrosome, which is the principal cellular MT nucleator.(43) In addition to centrosomal MTs, free MTs not associated with the centrosome are observed during interphase. At
the onset of mitosis, there is a dramatic rearrangement of the
interphase MT array. The radial interphase MT array depolymerizes and repolymerizes into the elliptical bipolar shape of
the spindle essential for cell division(1,2) (as described later
and outlined in Fig. 2A). How MTs can undergo such distinct
three-dimensional rearrangements from interphase to mitosis
is largely unknown. In this review, we will examine the molecular
machinery that regulates this spectacular rearrangement of MTs
during the cell cycle, and in particular, during spindle assembly.
The building blocks of MTs are tubulin dimers. Tubulin dimers
polymerize into a 25-nm-diameter MT filament via the formation
of 13 protofilaments and with concomitant GTP hydrolysis (Fig.
1). MTs are assymmetric, or so-called ‘‘polar’’ polymers. They are
polar because each tubulin dimer is composed of an ␣- and a ␤-
Funding agencies: EMBO. (Grant number: ALTF-97721); Danish
Natural Science Research Council (Grant number: 9700570).
Correspondence to: Søren S.L. Andersen, University of California, San
Diego, 9500 Gilman Drive, La Jolla, CA92093–0357; E-mail:
[email protected]. Abbreviations: MAP, microtubule-associated
protein; MT, microtubule.
BioEssays 21:53–60, r 1999 John Wiley & Sons, Inc.
subunit (only 50% identical), and because of the uniform polar
arrangement of these heterodimers in the MT lattice. This polarity
is used by motor proteins to sort cargo and MTs spatially. The
energy for the movement of motor proteins comes from the
hydrolysis of ATP coupled with the translocation of the motor
along MTs. MT polarity also results in MTs having a fast-growing
and a slow-growing end, the plus- and minus-ends, respectively.
The plus-end extends from the spindle poles/centrosomes toward the chromosomes, and the minus-end is often associated
with the spindle pole/centrosome.(1,2)
An understanding of how the three-dimensional rearrangement of MTs is achieved during the cell cycle requires knowledge
about the basic molecular polymerization characteristics of MTs.
However, MTs have very complex polymerization characteristics.
The ‘‘dynamic instability’’ model(1,3) offers a satisfactory theoretical
framework for the description of many of the MT polymerization
phenomenons observed in vitro and in vivo(4–6); it is used here to
explain how MTs polymerize during the cell cycle. In the dynamic
instability model, a MT polymerizes with a certain ‘‘growth rate’’
(vg) and depolymerizes with a certain ‘‘shrinkage rate’’ (vs). The
transition from growth to shrinkage is called a ‘‘catastrophe’’(7) and
occurs with a certain frequency (fcat). The opposite event, from
shrinkage to growth, is called a ‘‘rescue’’ and occurs with a
frequency (fres)(8) (Fig. 1). Regulation of the value of these four
parameters seems sufficient to regulate MT polymerization during
the cell cycle(4–6) (A more comprehensive model would include
the frequency of MT nucleation, fnuc, as a fifth parameter).
During the past fifteen years, work on MT turnover has been
strongly dominated and guided by the dynamic instability model1.
However, MTs may also turn over by so-called ‘‘treadmilling’’,
where tubulin subunits continuously added to one end of MTs
‘‘treadmill’’ through the MTs and subsequently are lost by continuous depolymerization at the other end.(9) It is important to realize
BioEssays 21.1
53
Review articles
that MT length changes occur at a rate of up to 1 µm/min by
treadmilling and up to 20 µm/min by dynamic instability. Therefore
‘‘treadmilling’’ conveys slow plasticity to the MTs, whereas ‘‘dynamic instability’’, in particular, is required for more extensive and
fast rearrangements of the MT network. For example, the observed constant but relatively slow turnover of the seemingly
stable interphase MT network(10) and the flux of tubulin subunits in
the spindle11 requires treadmilling.(9) On the other hand, the
dramatic rearrangement of the MT network at the interphase to
mitosis transition is strictly dependent on dynamic instability.
Although treadmilling in the past few years has experienced a
renaissance and received renewed attention, it is not the topic of
this review and will not be further discussed (reviewed by Margolis
and Wilson (1998)(9)). In the next section we address how MT
dynamic instability may be regulated in a cellular context.
fcat.(5,6,12,13) This change is regulated by phosphorylation by
cdc2 kinase,(5,6,12,14,15) and probably other kinases,(16,17) as
well as by phosphatases.(6,18) However, the molecular targets
of these kinases and phosphatases are largely unknown and
have to be identified to understand how the MT rearrangement at the interphase to mitosis transition is achieved. We will
now discuss how MTs may become dynamic during mitosis.
It has been hypothesized that the increase in fcat at the onset of
mitosis is due to the specific mitotic activation of a so-called
‘‘catastrophe factor.’’(19,20) The definition of a bona fide catastrophe factor is a molecule that is specifically activated during mitosis
and induces MT catastrophes either directly or indirectly. However, no such factor has yet been identified, as further discussed
in the next section. An alternative hypothesis to the ‘‘catastrophe
factor hypothesis’’ is that the balance between MT stabilizing and
MT destabilizing factors at any time and spatial point during the
cell cycle controls the dynamic state of MTs. Stabilizing factors
consist of a large group of well-characterized molecules called
microtubule-associated proteins (MAPs). We know from previous
studies that MAPs show a phosphorylation-dependent decrease
in MT affinity during mitosis (ref. 21 and references therein).
Moreover, it has been shown that mitotic MAPs promote MT
assembly twofold less than interphase MAPs; purified phosphorylated MAPs are also known to promote MT assembly less than
the corresponding unphosphorylated MAPs (ref. 21 and references therein). Thus, because interphase MAPs are dephosphorylated, they promote MT assembly more than their mitotic counterparts. Knowledge about MT destabilizing factors and their
regulation is far less advanced than that of MAPs, and encompasses Stathmin/Op18(22) and the motor proteins XKCM1(23) from
Xenopus and Kar3p from Saccharomyces cerevisiae.(24)
The increase in MT dynamics during mitosis may be
explained if the destabilizing factors are constitutively active
during the cell cycle. If so, they may simply overcome the
residual MT stabilizing activity of mitotic MAPs and cause
MTs to become dynamic.(21) In fact, the available data on the
destabilizing factors Stathmin/Op18 and XKCM1 suggest that
these are indeed constitutively active during the cell cycle.
In summary, the present experimental evidence supports a
picture whereby a balance between the activity of dominant MT
stabilizing factors and constitutively active MT destabilizing factors regulates the dynamic state of MTs during the cell cycle (Fig.
2B). Future work will involve testing of this hypothesis, which
requires more insight into the nature, number, and cell cycle
regulation of MT destabilizing factors.
Microtubule-associated proteins as regulators of
microtubule dynamics
In interphase, MTs are stable and typically form a large radial
array. In mitosis, by contrast, MTs are very dynamic and only
polymerize in the area around the chromosomes (Fig. 2A).
This change in MT organization requires a change in MT
dynamics. The hallmark of the change in MT dynamics
between interphase and mitosis is a 10-fold increase in
How does Stathmin/Op18 regulate microtubule
dynamics?
Since its discovery in 1983,(25) knowledge about the function of
Stathmin/Op18 (for ‘‘relay’’ in Greek and Op18 for Oncoprotein of
18 kD)(Fig. 3) has been lacking, despite numerous studies.(26)
Given its long pre-history, it was exciting when Stathmin/Op18
was discovered to interact with tubulin and destabilize MTs in
vitro, apparently by increasing fcat 5- to 10-fold.(19,22) Recent
Figure 1. Microtubule dynamic instability. The growth and
shortening (shrinking) phases of MT dynamic instability are
persistent because thousands of tubulin heterodimers add
during a single growth phase, or dissociate during a single
shortening phase. The abrupt transition between growth and
shortening is termed catastrophe, while the switch from
shortening back to growth is termed rescue. Polymerization of
MTs require the tubulin heterodimers to have GTP bound,
which is hydrolyzed rapidly upon incorporation into the MT
lattice. (From ref. 82, with permission from the ASCB.)
54
BioEssays 21.1
Review articles
Figure 2. Balanced regulation of microtubule dynamics during the
cell cycle. A: Schematic representation of the spatial arrangement of
MTs (black lines) in interphase, prophase, metaphase, and anaphase A
to telophase; ‘‘anaphase A to telophase’’ summarizes: segregation of
chromosomes to the poles (anaphase A), spindle elongation and astral
MT growth (anaphase B; reformation of the nuclear envelope during
telophase is not shown); Note that the exit from mitois is reflected by an
increase in average MT length. MTs not attached to kinetochores are
called ‘‘interpolar MTs’’; Motor proteins are localized in this area; They
keep poles together during anaphase A and push poles apart during
anaphase B (ie. the bipolar KLP61F). Interphase nucleus (white circle)
and centrosomes (black dots). B: Integrated hypothetical model for the
regulation of MT dynamics during the cell cycle. The relative activities of
kinases (K) and phosphatases (P) during the cell cycle are shown by
the size and boldness of the letters; The phosphorylation level of a
given substrate is at any time determined by the activity of the kinase
and the opposing phosphatase (upper right corner). In interphase (Int),
MAPs (M) are dephosphorylated and stabilize MTs more than the
combined activity of the destabilizing factors (D). This results in long
centrosome associated and spontaneously assembled cytoplasmic
MTs. At the onset of mitosis (prophase shown here, Pro) the kinase
activities (K) increases dramatically and more than the corresponding
change in the phosphatase activities (P).(77) In addition to these
enzymatic activity changes, there is a precise subcellular localization of
both kinases and phosphatases (not shown).(18) The change in enzymatic activities pushes the balance toward the phosphorylation of
MAPs (M-P), which leads to dynamic MTs, as the destabilizing factors
are constitutively active. The first observable effects of this change in
the balance of kinase/phosphatase activities is the depolymerization of
cytoplasmic MTs during early prophase; the centrosome nucleated MTs
still are long and are required for the motor-dependent separation of the
duplicated centrosomes. At metaphase (Meta), MTs have become short
and are very dynamic. Although the MTs are very dynamic, a spindle
still forms because kinetochores (not shown) capture and stabilize MTs;
chromosomes also stabilize MTs through local dephosphorylation of MAPs.(63,64,78) and local phosphorylation-dependent inactivation(29)
of destabilizing factors (see also Fig. 4B). During ana-/telophase (Ana/Telo) the MTs return to the interphase state. This change involves
dephosphorylation of MAPs through the inactivation of kinases and perhaps activation of phosphatases.(6)
studies in vitro have, however, shown that Stathmin/Op18 does
not increase fcat; rather, it regulates MT polymerization by sequestering the pool of soluble/free tubulin.(27,28) What are the differences, then, between these studies(22,27,28)? The major difference between the former(22) and the more recent studies(27,28)
is a moderate 0.7 pH units change in the buffer used. Thus,
contrary to expectation, time-lapse video microscopy of MTs
showed that Stathmin/Op18 simply reduced vg, due to the
lowered free tubulin concentration, without any effect on
fcat(28) (despite stronger tubulin sequestration at the pH used
by Curmi et al. (1997)(28) than at the pH used by Belmont and
Mitchison (1996)(22)). Furthermore, upon immunodepletion of
95% of Stathmin/Op18 from mitotic Xenopus egg extracts,
there was only a twofold reduction in the catastrophe frequency.(6)
Moreover, in Xenopus egg extracts, there is no differential
phosphorylation of Stathmin/Op18 between interphase and mitosis in the absence of chromosomes,(29) although MTs are dynamic in a mitotic Xenopus egg extract. Taken together, the
present data show that Stathmin/Op18 is not the long sought-
after bona fide catastrophe factor,(19,22) and most likely it simply
destabilizes MTs by sequestering tubulin.(27,28)
However, the mechanism by which Stathmin/Op18 destabilize
MTs in vivo remains unclear. Stathmin/Op18 is not essential for
viability(30) or for spindle assembly,(29) also indicating that, at most,
Stathmin/Op18 modulates MT dynamics in vivo, and that it is not
a major regulator of MT assembly. One hint as to the in vivo
function comes from the immunoprecipitation experiments previously mentioned.(6) These studies suggest that Stathmin/Op18
may have a minor direct effect on fcat, which may be envisioned if
the tubulin-Stathmin/Op18 complex (a ‘‘T2S complex’’)(27,28) can
disturb MT polymerization through competition with free tubulin
for polymerizing MT ends. However, additional studies in vivo
and in Xenopus extracts are required to clarify this issue (i.e.,
to resolve whether or not tubulin was also depleted during
these experiments(6)). Determination of the precise in vivo function is further complicated by the very complex regulation of
Stathmin/Op18 phosphorylation/activity (Fig. 3) in vivo, which is
under both intra- and extracellular control.(31,32) Phosphorylation
of Stathmin/Op18 turns off the MT destabilizing activity, and it
BioEssays 21.1
55
Review articles
Figure 3. Molecular organization of Stathmin/Op18. Stathmin/Op18 is a 149-amino acid-long molecule with an elongated
shape.(79) The human protein has four consensus Serine phosphorylation sites and the 79% identical Xenopus homologue has three
(not Ser63). Sequence comparison and in vitro studies have
determined the kinases that preferentially phosphorylate the 3–4
sites(19,31–35): CaMKIV(73) and PKA32 (Ser16), MAPK (Ser25),
cdc2 kinase (Ser38 is Ser39 in Xenopus(80)), and PKA (Ser63).
By circular dichroism, the protein has 45% of its sequence in
an ␣-helical conformation and computer analysis predicts the
region between amino acid 47 and 124 to be ␣-helical. This
region also contains characteristic heptad repeats of neutral
and hydrophobic residues, suggesting that this region is
involved in ‘‘coiled-coil’’ interactions with proteins presenting
similar structures.(79) The name Stathmin (‘‘relay’’ in Greek)
was given because of the presence of the 3–4 different serine
phosphorylation sites that are thought to ‘‘regulate/relay’’ coiledcoil interactions of the C-terminus with other proteins. In
neurons Stathmin/Op18 is localized to the soma whereas the
very close ‘‘SCG10’’ homolog is localized to the growth cone.(83)
becomes highly phosphorylated during mitosis.(28,31,33–35) Although more work is required, current in vitro data suggest that
Stathmin/Op18 primarily sequesters tubulin without significantly
regulating fcat.
If Stathmin/Op18 primarily sequesters tubulin, how could this
regulate MT dynamics in vivo? A rough calculation shows that at
an in vivo concentration of 6 µM(22,29) and a maximal stoichiometry
of binding between Stathmin/Op18 and tubulin heterodimers of
1:2 (the T2S complex)(27,28) Stathmin/Op18 could sequester almost 12 µM tubulin in vivo. Considering that about two-thirds of
the total 20 µM tubulin is polymerized into MTs in vivo,(36) a factor
like Stathmin/Op18 could completely regulate the amount of free
tubulin available for polymerization. This may be important, as the
tubulin concentration very strongly affects MT nucleation and MT
dynamics in vitro.(37) For example, regulation of the binding
between tubulin and Stathmin/Op18 may serve to fine-tune the
amount of free tubulin during the cell cycle and at any particular
spatial point. Since Stathmin/Op18 could serve to buffer the
amount of free tubulin available for MT polymerization and more
work should be aimed at analyzing how the free tubulin concentration influences MT dynamics in vivo.(13,38)
56
BioEssays 21.1
Regulation of microtubule dynamics during
spindle assembly
In the remainder of this review we will focus on how MTs
assemble around chromosomes during spindle assembly. As
shown in Figure 2A, the reorganization of the spatial arrangement
of MTs from the interphase array to the bipolar spindle is dramatic.
It is especially noteworthy that spindle MTs extend primarily from
the spindle poles toward the chromosomes, and that MTs are not
found elsewhere in the cytoplasm (Fig. 2A).At the molecular level,
it is known that the interphase to mitotic spindle MT rearrangement involves activation of MT severing factors,(39) and a significant increase in fcat caused by a regulated change in the activity of
MT stabilizing and destabilizing factors(2,5,12,36)(Fig. 2). It is unknown how the new highly dynamic centrosomal MTs elongate
preferentially toward the chromosomes. Assembly of the female
meiotic spindle occurs through a different pathway that involves
centrosome independent assembly of MTs around chromosomes. Therefore, assembly of both the meiotic and the mitotic
spindles involves selective local stabilization of MTs around the
chromosomes(2,40–43) (Fig. 4). But how is it that MTs grow preferentially around the chromosomes during mitosis, and not elsewhere
in the cytoplasm (Fig. 2A)?
Two models for spindle microtubule assembly
The model most frequently used to explain the preferential MT
polymerization toward chromosomes during spindle assembly is
called ‘‘search and capture.’’ This model proposes that dynamic
MTs nucleated by centrosomes randomly search in space. When
the MTs hit a kinetochore, they become captured and stabilized,
and a spindle forms with time.(1,44–46) This model has its roots in
the observations that MTs were preferentially found associated
with, and nucleated by, centrosomes and kinetochores (for older
literature, see Marek, 1978(47)). However, Marek’s classic experiments involving the removal of chromosomes from spindles,(47)
as well as the observation of the site of MT re-polymerization after
MT depolymerization in mitotic cells,(48) led De Brabander et al.
(1981)(48) to suggest that the kinetochores perhaps were producing a diffusible factor that could locally induce MT assembly in the
vicinity of kinetochores.(48) This was a new idea because it
suggested that the nucleation of MTs could be enhanced by a
chromosome-derived diffusible factor, and thus that MT assembly
did not have to use a kinetochore or centrosome-based template.
Convincing evidence for such an active non-template role of
chromosomes in MT assembly came from the studies of Karsenti
et al. (1984)(49) showing that oligomerized ␭-phage DNA could
induce spindle-like MT structures when injected into Xenopus
eggs. Important was also the suggestion that chromosome areas
by themselves could induce MT polymerization independently of
the presence of kinetochores.(49) These experiments are the
origin of the second ‘‘local stabilization’’ model that proposes that
chromosomes actively promote spindle MT assembly(47–49) There
is now ample evidence for the local stabilization model. Experiments have shown that over short distances chromosomes can
induce the nucleation and assembly of MTs, as well as pro-
Review articles
Figure 4. Hypothetical model for local chromoMAPs and
somal regulation of Stathmin/Op18 activity and
Motors
(29,51,64)
microtubule assembly.
Chromosome(s)
(gray ellipse) somehow inhibit a diffusible type 2A
phosphatase (PP2A) (small blue ellipses, kinetochores). This inhibition results in a gradient (indicated by the increased thickness of the PP2Alabeled arrow) of PP2A activity around
chromosomes, if the enzyme reactivating PP2A
also is freely diffusible. The gradient in PP2A
activity then leads to a gradient in Stathmin/Op18
(18) phosphorylation/activity (red circle) if the
kinase that phosphorylates Stathmin/Op18, cdc2,
is freely diffusible. Note, it is also possible that
local PP2A inactivation leads to local PKA activation which subsequently hyperphosphorylates
Stathmin/Op18 (not shown). The local inactivation
of Stathmin/Op18 results in release of tubulin (not
shown) from phosphorylated Stathmin/Op18 (18P), which may then contribute to the preferential
MT nucleation and growth toward chromosomes
in systems devoid of centrosomes,(50,51,54) and
preferential growth of centrosome nucleated MTs
toward chromosomes in centrosome-containing
PP2A
systems.(43,51,62) Whether chromosomes or centrosomes become the nucleation site for MTs may be
Op18
Op18 + P
P
simply a kinetic issue, as centrosomes are kineticdc2
cally dominant over chromosomes for MT nucleation.(38,51,54,81) Because Stathmin/Op18 is not
hyperphosphorylated in the absence of chromosomes, it is suggested that, in the absence of chromosomes, PP2A dephosphorylates Stathmin/Op18 (Op18-P) faster (thick arrow) than
the corresponding phosphorylation rate by cdc2 kinase (thin arrow). Although the mechanism of Stathmin/Op18 hyperphosphorylation by
chromosomes is unknown, it involves the enzymatic activities of at least PP2A, cdc229 and perhaps PKA74. Curved arrow, illustrates the
likelihood that MAPs and motors are locally regulated in ways comparable to that of Stathmin/Op18. However, before MAPs and motors
can exert their function, MTs need to be nucleated. Stathmin/Op18 may play an important role in regulating this initial phase of MT
polymerization.(29,43)
mote their polymerization off centrosomes over longer distances(29,42,47,49–58) (Fig. 4). It is important to realize that these two
models are not mutually exclusive. Kinetochores are essential for
both the congression and segregation of chromosomes,(2,59) and
the efficiency of ‘‘search and capture’’ may be greatly enhanced
through ‘‘local MT stabilization’’ by the chromosomes. Thus, the
biggest difference between the local stabilization and the search
and capture models lies in the extent of active chromosomal
participation in MT formation and stabilization.
Interestingly, it appears that some systems favor one mechanism over the other, and that both mechanisms are probably
involved during spindle assembly to a variable extent. Thus, in
some systems, centrosomes and kinetochores may be kinetically
dominant over chromosomes and drive spindle assembly,
whereas chromosomes are necessary and sufficient in other
systems. For example, in Xenopus eggs in the absence of
centrosomes and kinetochores,(50,51) and in systems devoid of
centrosomes, such as plants(60) and mouse embryos,(61) the
major driving force regulating spindle assembly is ‘‘local stabilization.’’ In other systems, chromosome areas appear to play no role
at all and spindle assembly is completely dependent on centrosomes and kinetochores.(46,58,62)
Regardless of the extent of chromosomal contribution to
spindle assembly, nothing is known about how chromosomes
locally regulate the biochemical environment in favor of MT
assembly. In the remainder of this review, we will focus on this
affect of the local stabilization model.
How do chromosomes promote microtubule
assembly?
It has been proposed that an enzymatic activity localized to mitotic
chromosomes could modify the cytoplasm locally and create an
area of higher MT stability around chromosomes. The result of
this local chromosomal effect would be preferential growth of MTs
around chromosomes.(42,49,57) It was proposed that the local effect
was established by a type 1 phosphatase bound onto chromosomes and was counteracted by the action of a diffusible kinase
phosphorylating and inactivating a diffusible MT stabilizing molecule.(57) Since then, our understanding of MT dynamics regulation has increased. As outlined in Figure 2B, we now have knowledge about molecules that both stabilize and destabilize MTs as
BioEssays 21.1
57
Review articles
well as their regulatory enzymes.(2,6,21–23,63,64) We can therefore
approach the proposed model(42,57) with greater molecular detail.
Is there a morphogenetic gradient of microtubule
stabilizing and destabilizing factors during
spindle assembly?
Let us define the sum of MT polymerization destabilizing
(i.e., Stathmin/Op18 and XKCM1(22,23)) and stabilizing (i.e.,
MAPs(63–65)) activities as the ‘‘potential to polymerize MTs.’’ If one
then imagines that chromosomes create an area with a positive
potential for MT polymerization around them, this could contribute
to spindle assembly. Obviously, there are no strict boundaries in
the cytoplasm; such an area would consist of a region or gradient
going from a positive potential close to the chromosomes to a
negative potential at some distance from them. It seems likely that
such local regulation of MT dynamics around chromosomes
involves the same components and regulators required for global
regulation of MT dynamics during the cell cycle. Thus, proper
chromosome-induced local regulation of the activity of MAPs,
Stathmin/Op18, and XKCM1 would be important for proper
spindle assembly.
Before we analyze in greater detail the evidence for local MT
dynamics regulation and a possible molecular gradient during
spindle assembly, it is worth noting that gradients are common in
biological, chemical, and physical systems. Specific examples of
molecular gradients in biology come from neurobiology,(66–68) cell
motility,(69) chemotaxis,(70) and the morphogenesis of fruit flies.(71,72)
It would not be completely unexpected if a gradient in the MT
polymerization potential were involved in forming the spindle.
Is there a gradient of Stathmin/Op18 and MAP
activity during spindle assembly?
The first specific molecular example of how chromosomes regulate MT assembly and may generate a gradient in the potential to
polymerize MTs comes from a recent paper on Stathmin/Op18.(29)
This paper reported that when increasing amounts of mitotic
chromosomes were added to mitotic Xenopus egg extract,
Stathmin/Op18 became increasingly phosphorylated. This was
interesting because it had previously been described that during
cellular mitosis Stathmin/Op18 becomes hyperphosphorylated.(34)
In the Xenopus egg extracts system, however, hyperphosphorylation required the presence of mitotic chromosomes.(29) These
experiments(29) led to the speculation that it is the presence of
mitotic chromosomes that induces Stathmin/Op18 hyperphosphorylation, and not simply the mitotic state of the cytoplasm.
Interestingly, the concentration of chromosomes required to
induce hyperphosphorylation of Stathmin/Op18(29) corresponds
to a situation in which each chromosome is surrounded by a
uniform cube of extract with a volume of 20–60 µm3, roughly the
length of the mitotic spindle in the Xenopus system. Moreover, the
observed dose-response effect of the chromosomes suggested
that Stathmin/Op18 is only locally phosphorylated around chromosomes. These experiments indicate that there may be a spatial
gradient in Stathmin/Op18 phosphorylation, and thus activity,
around chromosomes.
58
BioEssays 21.1
It is unknown exactly how chromosomes regulate Stathmin/
Op18 phosphorylation, and several models are possible. One
model put forward by Karsenti’s and Hyman’s groups(6,29) is
described here (Fig. 4). The model is based on the fact that
addition of okadaic acid, a chemical inhibitor of phosphatase 2A
(PP2A), to mitotic Xenopus egg extracts leads to a phosphorylation pattern identical to that observed in the presence of mitotic
chromosomes.(29) Thus, it was suggested that chromosomes
somehow inhibit diffusible PP2A(29) in a dose-dependent manner,
and that a gradient in PP2A activity around chromosomes is
established. This gradient in PP2A activity then leads to a gradient
in Stathmin/Op18 phosphorylation if its kinase, most likely cdc2,(29)
is freely diffusible and globally active. The gradient in Stathmin/
Op18 phosphorylation leads to a gradient in its binding activity
with tubulin subunits, and ultimately to a gradient in the free
tubulin concentration (highest close to the chromosomes; the
latter has also been proposed by Zhai et al., 1996(36)). This local
increase in tubulin concentration would contribute to the preferential nucleation and growth of MTs around chromosomes (Fig. 4).
Indeed, it has been shown that spindle size can be affected by the
amount of free tubulin,(47) an issue that, however, needs additional
investigation (see refs. 13 and 47 for discussion and references).
At this point, it is important to note that because Stathmin/Op18 is
not essential for viability(30) or for spindle assembly,(29) it should not
be considered the principal regulator of MT polymerization around
chromosomes. However, the increased MT nucleation observed
around chromosomes in the absence of Stathmin/Op18,(29) combined with the results from in vivo overexpression experiments,(34)
show that Stathmin/Op18 is involved in regulating MT dynamics
locally around mitotic chromosomes (Fig. 4).
One challenge to the proposed mechanism of chromosomal
regulation of Stathmin/Op18 phosphorylation (Fig. 4) comes from
recent reports by Gullberg’s group showing that the phosphorylation of two PKA sites (Ser16 and Ser63) are key to turning off the
tubulin binding activity of Stathmin/Op18, and not phosphorylation of the cdc2 (Ser38) or MAPK (Ser25) sites(33,73) (Fig. 3). The
model presented in Figure 4 is not incompatible with these
observations, however. Local inactivation of PP2A could cause
local activation of PKA,(74) which is known to phosphorylate Ser16
and Ser63, that would then inactivate Stathmin/Op18. Moreover,
preceding phosphorylation of Ser25 and in particular Ser38 is a
prerequisite(29,33) for subsequent phosphorylation on Ser16 and
Ser63. This suggests that phosphorylation of the cdc2 site may be
the master switch for activity regulation during the cell cycle.
Interestingly, a positive feedback loop between the extent of
Ser16 phosphorylation and MT assembly was reported.(84) This
implies that formation of spindle MTs may be an autocatalytic
process that involves local chromosomal regulation of a number
of different molecules, including Stathmin/Op18. It remains an
open question precisely how Stathmin/Op18 hyperphosphorylation is achieved in the presence of chromosomes during mitosis.
It will be exciting to learn how chromatin regulates the phosphorylation of Stathmin/Op18. That mechanism may be general
Review articles
and could be applicable to local regulation of MAPs and motor
activities (Fig. 4). For example, XMAP230, XMAP310, and E-MAP115 are localized to spindle MTs, despite their phosphorylationdependent reduced affinity for MTs during mitosis.(63,64,75) Thus, it
seems likely that chromosomes may reduce the phosphorylation
level of these MAPs locally, as previously discussed.(64)
A point of debate is the presence of a molecular MT polymerization gradient around chromosomes in vivo. Modelling work by
Leibler and colleagues supports the possibility of a gradient52,76.
Their work shows that a molecular gradient the size of a spindle
can be generated by immobilizing a phosphatase on a spherical
bead, while a kinase and the respective substrate freely diffuse
(this model also works with a reversed configuration of the
regulators).(52,76)
It may be possible to use chimeric fluorescent proteins
with a phosphorylation-sensitive emission spectrum (similar
to that described by Post et al., 1995(69)) to demonstrate a
gradient experimentally. Alternatively, it may be possible to
show a gradient using antibodies that recognize phosphoepitopes(71) on nondiffusible spindle-associated molecules, like
MAPs.(63,64) New more creative approaches may also be
needed. It will be a considerable challenge in the future to test
experimentally whether a gradient in the potential to polymerize MTs exists around chromosomes during spindle assembly.
Conclusions
The current understanding of MT dynamics regulation in vivo
suggests that the dynamic state of MTs during the cell cycle is
controlled by a balance between MT stabilizing and destabilizing factors rather than by a mitotically activated catastrophe
factor (Fig. 2). Interestingly, the activity balance between
these same factors also regulates local MT dynamics during
spindle assembly. Most exciting is that we now start to have a
molecular understanding of how the mitotic chromosomes
locally regulate MT-stabilizing and MT-destabilizing factors in
favor of local MT assembly (Fig. 4).
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
Acknowledgments
Thanks are due to S.Clark, K. Ganguly, J.Großhans, Y. Jiang,
A. Merdes, M. Ruchhoeft, D. Sullivan, R.Wordel, M. Adams,
and A.Wilkins, as well as to anonymous reviewers for comments on the manuscript. Thanks are also due to former
colleagues at the EMBL for inspiration and sessions on the
topics discussed. This work was partially supported by an
EMBO long-term fellowship (ALTF-97721), and a fellowship
from the Danish Natural Science Research Council (9700570).
24.
References
27.
1. Kirschner M, Mitchison T. Beyond self-assembly: from microtubules to
morphogenesis. Cell 1986;45:329–342.
2. Hyman AA, Karsenti E. Morphogenetic properties of microtubules and
mitotic spindle assembly. Cell 1996;84:401–411.
3. Mitchison TJ, Kirschner MW. Dynamic instability of microtubule growth.
Nature 1984;312:237–242.
4. Gliksman NR, Skibbens RV, Salmon ED. How the transition frequencies of
microtubule dynamic instability (nucleation, catastrophe, and rescue)
23.
25.
26.
28.
29.
regulate microtubule dynamics in interphase and mitosis: analysis using a
Monte Carlo computer simulation. Mol Biol Cell. 1993;4:1035–1050.
Verde F, Dogterom M, Stelzer E, Karsenti E, Leibler S. Control of
microtubule dynamics and length by cyclin A and cyclin B dependent
kinases in Xenopus egg extracts. J. Cell Biol. 1992;118:1097–1108.
Tournebize R, Andersen SSL, Verde F, Dorée M, Karsenti E, Hyman AA.
Distinct roles of PP1 and PP2A-like phosphatases in control of microtubule
dynamics during mitosis. EMBO J 1997;16:5537–5549.
McIntosh JR. Microtubule catastrophe. Nature 1984;312:196–197.
Walker RA, O’Brien ET, Pryer NK, Sobeiro MF, Voter WA, Erickson HP, Salmon
ED. Dynamic instability of individual microtubules analysed by video light microscopy: rate constants and transition frequencies. J Cell Biol 1988;107:1437.
Margolis RL, Wilson L. Microtubule treadmilling: what goes around comes
around. BioEssays 1998;20:830–836.
Vorobjev IA, Svitkina TM, Borisy GG. Cytoplasmic assembly of microtubules in cultured cells. J Cell Sci 1997;110:2635–2645.
Desai A, Maddox PS, Mitchison TJ, Salmon ED. Anaphase A chromosome
movement and poleward spindle microtubule flux occur at similar rates in
Xenopus extract spindles. J Cell Biol 1998;141:703–713.
Belmont LD, Hyman AA, Sawin KE, Mitchison TJ. Real-time visualization of
cell cycle-dependent changes in microtubule dynamics in cytoplasmic
extracts. Cell 1990;62:579–589.
Parsons SF, Salmon ED. Microtubule assembly in clarified Xenopus egg
extracts. Cell Motil Cytoskeleton 1997;36:1–11.
Ookata K, Hisanaga S-I, Sugita M, Okuyama A, Murofushi H, Kitazawa H, Chari
S, Bulinski JC, Kishimoto T. MAP4 is the in vivo substrate for cdc2 kinase in
HeLa cells: identification of an M-phase specific and a cell cycle-independent
phosphorylation site in MAP4. Biochemistry 1997;36:15873–15883.
Verde F, Labbé JC, Dorée M, Karsenti E. Regulation of microtubule
dynamics by cdc2 protein kinase in cell-free extracts of Xenopus eggs.
Nature 1990;343:233–238.
Drewes G, Ebneth A, Preuss U, Mandelkow E-M, Mandelkow E. MARK, a
novel family of protein kinases that phosphorylate microtubule-associated
proteins and trigger microtubule disruption. Cell 1997;89:297–308.
Gotoh Y, Nishida E, Matsuda S, Shiina N, Kosako H, Shiokawa K, Akiyama
T, Ohta K, Sakai H. In vitro effects on microtubule dynamics of purified
Xenopus M phase-activated MAP kinase. Nature 1991;349:251–254.
Andreassen PR, Lacroix FB, Villa-Moruzzi E, Margolis RL. Differential
subcellular localization of protein phosphatase-1 alpha, gamma1, and
delta isoforms during both interphase and mitosis in mammalian cells. J
Cell Biol 1998;141:1207–1215.
Belmont L, Mitchison T, Deacon HW. Catastrophic revelations about
Op18/stathmin. Trends Biochem Sci 1996;21:197–198.
McNally FJ. Modulation of microtubule dynamics during the cell cycle.
Curr Opin Cell Biol 1996;8:23–29.
Andersen SSL. Xenopus interphase and mitotic microtubule-associated
proteins differentially suppress microtubule dynamics in vitro. Cell Motil
Cytoskeleton 1998;41:202–213.
Belmont LD, Mitchison TJ. Identification of a protein that interacts with
tubulin dimers and increases the catastrophe rate of microtubules. Cell
1996;84:623–631.
Walczak CE, Mitchison TJ, Desai A. XKCM1: a Xenopus kinesin-related
protein that regulates microtubule dynamics during mitotic spindle assembly. Cell 1996;84:37–47.
Huyett A, Kahana J, Silver P, Zeng X, Saunders WS. The Kar3p and Kip2p
motors function antagonistically at the spindle poles to influence cytoplasmic microtubule numbers. J Cell Sci. 1998;111:295–301.
Sobel A, Tashjian JAH. Distinct patterns of cytoplasmic protein phosphorylation related to regulation of synthesis and release of prolactin by GH
cells. J Biol Chem 1983;258:10312–10324.
Sobel A. Stathmin: a relay phosphoprotein for multiple signal transduction? Trends Biochem Sci 1991;16:301–305.
Jourdain L, Curmi P, Sobel A, Pantaloni D, Carlier M-F. Stathmin: A
tubulin-sequestering protein which forms a ternary T2S complex with two
tubulin molecules. Biochemistry 1997;36:10817–10821.
Curmi PA, Andersen SSL, Lachkar S, Gavet O, Karsenti E, Knossow M,
Sobel A. The Stathmin/tubulin interaction in vitro. J Biol Chem 1997;272:
25029–25036.
Andersen SSL, Ashford AJ, Tournebize R, Gavet O, Sobel A, Hyman AA,
Karsenti E. Mitotic chromatin regulates phosphorylation of Stathmin/
Op18. Nature 1997;389:640–643.
BioEssays 21.1
59
Review articles
30. Schubart UK, Yu J, Amat JA, Wang Z-Q, Hoffmann MK, Edelmann W.
Normal development of mice lacking metablastin (P19), a phosphoprotein
implicated in cell cycle regulation. J Biol Chem 1996;271:14062–14066.
31. Lawler S. Microtubule dynamics: if you need a shrink try Stathmin/Op18.
Curr Biol 1998;8:R212–R214.
32. Gradin HM, Larsson N, Marklund U, Gullberg M. Regulation of microtubule dynamics by extracellular signals: cAMP-dependent protein kinase
switches off the activity of oncoprotein 18 in intact cells. J Cell Biol
1998;140:131–141.
33. Larsson N, Melander H, Marklund U, Osterman O, Gullberg M. G2/M
transition requires multisite phosphorylation of oncoprotein 18 by two
distinct protein kinase systems. J Biol Chem 1995;270:14175–14183.
34. Marklund U, Larsson N, Gradin HM, Brattsand G, Gullberg M. Oncoprotein 18. is a phosphorylation-responsive regulator of microtubule dynamics. EMBO J 1996;15:5290–5298.
35. Larsson N, Marklund U, Melander HG, Brattsand G, Gullberg M. Control of
microtubule dynamics by oncoprotein 18: dissection of the regulatory role of
multisite phosphorylation during mitosis. Mol Cell Biol 1997;17:5530–5539.
36. Zhai Y, Kronebusch PJ, Simon PM, Borisy GG. Microtubule dynamics at
the G2/M transition: abrupt breakdown of cytoplasmic microtubules at
nuclear envelope breakdown and implications for spindle morphogenesis. J Cell Biol 1996;135:201–214.
37. Fygenson DK, Braun E, Libchaber A. Phase-diagram of microtubules.
Phys Rev Ed 1994;50:1579–1588.
38. Hyman A, Karsenti E. The role of nucleation in patterning microtubule
networks. J Cell Sci 1998;111:2077–2083.
39. Hartman JJ, Mahr J, McNally K, Okawa K, Iwamatsu A, Thomas S,
Cheesman S, Heuser J, Vale RD, McNally FJ. Katanin, a microtubulesevering protein, is a novel AAA ATPase that targets to the centrosome
using a WD40-containing subunit. Cell 1998;93:277–287.
40. Waters JC, Salmon ED. Pathways of spindle assembly. Curr Opin Cell Biol
1997;9:37–43.
41. McKim KS, Hawley RS. Chromosomal control of meiotic cell division.
Science 1995;270:1595–1601.
42. Karsenti E. Mitotic spindle morphogenesis in animal cells. Semin Cell Biol
1991;2:251–260.
43. Andersen SSL. Molecular characteristics of the centrosome. Int Rev Cytol
1999;187:51–109.
44. Nicklas RB. Chromosomes and kinetochores do more in mitosis than previously thought. In: Gustafson JP, Appels R, editors. Chromosome structure and
function: the impact of new concepts. New York: Plenum Press; 1988. p 53–74.
45. Holy TE, Leibler S. Dynamic instability of microtubules as an efficient way
to search in space. Proc Natl Acad Sci USA 1994;91:5682–5685.
46. Nicklas RB. How cells get the right chromosomes. Science 1997;275:632–637.
47. Marek LF. Control of spindle form and function in Grasshopper spermatocytes. Chromosoma 1978;68:367–398.
48. De Brabander M, Geuens G, De Mey J, Joniau M. Nucleated assembly of
mitotic microtubules in living PTK2 cells after release from nocodazole
treatment. Cell Motil Cytoskeleton 1981;1:469–483.
49. Karsenti E, Newport J, Kirschner M. Respective roles of centrosomes and
chromatin in the conversion of microtubule arrays from interphase to
metaphase. J Cell Biol 1984;99:47s–57s.
50. Heald R, Tournebize R, Blank T, Sandaltzopoulos R, Becker P, Hyman A,
Karsenti E. Self-organization of microtubules into bipolar spindles around
chromatin beads in Xenopus egg extracts. Nature 1996;382:420–425.
51. Heald R, Tournebize R, Habermann A, Karsenti E, Hyman A. Spindle
assembly in Xenopus egg extracts: respective roles of centrosomes and
microtubule self-organization. J Cell Biol 1997;138:615–628.
52. Dogterom M, Felix MA, Guet CC, Leibler S. Influence of M-phase chromatin on
the anisotropy of microbule asters. J. Cell Biol 1996;133:125–140.
53. Zhang D, Nicklas RB. Chromosomes initiate spindle assembly upon
experimental dissolution of the nuclear envelope in grasshopper spermatocytes. J Cell Biol 1995;131:1125–1131.
54. de Saint Phalle B, Sullivan W. Spindle assembly and mitosis without centrosomes in parthenogenetic sciara embryos. J Cell Biol 1998;141:1383–1391.
55. Church K, Nicklas BR, Lin H-PP. Micromanipulated bivalents can trigger
mini-spindle formation in Drosophila melanogaster spermatocyte cytoplasm. J Cell Biol 1986;103:2765–2773.
56. Steffen W, Fuge H, Dietz R, Bastmeyer M, Muller G. Aster-free spindle
poles in insect spermatocytes: evidence for chromosome-induced spindle
formation? J Cell Biol 1986;102:1679–1687.
60
BioEssays 21.1
57. Karsenti E, Verde F, Félix MA. Role of type 1 and type 2A protein
phosphatases in the cell cycle. Adv Protein Phosph 1991;6:453–482.
58. Nicklas RB. Chromosomes and kinetochores do more in mitosis than
previously thought. New York: Plenum Press; 1988. p 53–74.
59. Wood KW, Sakowicz R, Goldstein LS, Cleveland DW. CENP-E is a plus
end-directed kinetochore motor required for metaphase chromosome
alignment. Cell 1997;91:357–366.
60. Lambert A-M. Microtubule-organizing centers in higher plants. Curr Opin
Cell Biol 1993;5:116–122.
61. Maro B, Johnson MH, Webb M, Flach G. Mechanism of polar body formation in
the mouse oocyte: an interaction between the chromosomes, the cytoskeleton
and the plasma membrane. J Embryol Exp Morphol 1986;92:11–32.
62. Sluder G, Rieder CL, Miller F. Experimental separation of pronuclei in
fertilized sea urchin eggs: chromosomes do not organize a spindle in the
absence of centrosomes. J. Cell Biol 1985;100:897–903.
63. Andersen SSL, Buendia B, Domı́nguez JE, Sawyer A, Karsenti E. Effect on
microtubule dynamics of XMAP230, a microtubule-associated protein present
in Xenopus laevis eggs and dividing cells. J Cell Biol 1994;127:1289–1299.
64. Andersen SSL, Karsenti E. XMAP310: a Xenopus rescue promoting factor
localized to the mitotic spindle. J Cell Biol 1997;139:975–983.
65. Gard D, Kirschner M. A microtubule-associated protein from Xenopus
eggs that specifically promotes assembly at the plus end. J Cell Biol
1987;105:2203–2215.
66. Zheng JQ, Felder M, Connor JA, Poo M-M. Turning of nerve growth cones
induced by neurotransmitters. Nature 1994;368:140–144.
67. Mandell JW, Banker GA. A spatial gradient of tau protein phosphorylation
in nascent axons. J Neurosci 1996;16:5727–5740.
68. Tessier-Lavigne M, Placzek M. Target attraction: are developing axons
guided by chemotropism? Trends Neurosci 1991;14:303–310.
69. Post PL, Debiasio RL, Taylor DL. A fluorescent protein biosensor of myosin
II regulatory light chain phosphorylation reports a gradient of phosphorylated myosin II in migrating cells. Mol Biol Cell 1995;6:1755–1768.
70. Berg HC. Chemotaxis in bacteria. Annu Rev Biol Bioeng 1975;4:119–136.
71. Gabay L, Seger R, Shilo BZ. In situ activation pattern of Drosophila EGF
receptor pathway during development. Science 1997;277:1103–1106.
72. Gurdon JB, Dyson S, St Johnston D. Cells’ perception of position in a
concentration gradient. Cell 1998;95:159–162.
73. Gradin HM, Marklund U, Larsson N, Chatila TA, Gullberg M. Regulation of
microtubule dynamics by Ca2⫹/calmodulin-dependent kinase IV/Gr-dependent phosphorylation of oncoprotein 18. Mol Cell Biol 1997;17:3459–3467.
74. Grieco D, Porcellini A, Avvedimento EV, Gottesman ME. Requirement for
cAMP-PKA pathway activation by M phase-promoting factor in the
transition form mitosis to interphase. Science 1996;271:1718–1723.
75. Masson D, Kreis TE. Binding of E-MAP-115 to microtubules is regulated by cell
cycle-dependent phosphorylation. J Cell Biol 1995;131:1015–1024.
76. Dogterom M. Physical aspects of microtubule growth and mitotic spindle
formation. Thesis. University of Paris-Sud, France; 1994. p 140–146.
77. Karsenti E, Bravo R, Kirschner M. Phosphorylation changes associated
with the early cell cycle in Xenopus eggs. Dev Biol 1987;119:442–453.
78. Ookata K, Hisanaga S, Bulinski JC, Murofushi H, Aizawa H, Itoh TJ, Hotani
H, Okumura E, Tachibana K, Kishimoto T. Cyclin B interaction with
microtubule-associated protein 4 (MAP4) targets p34cdc2 kinase to
microtubules and is a potential regulator of M-phase microtubule dynamics. J Cell Biol 1995;128:849–862.
79. Curmi PA, Maucuer A, Asselin S, Lecourtois M, Chaffotte A, Schmitter JM,
Sobel A. Molecular characterization of human stathmin expressed in
Escherichia coli: site-directed mutagenesis of two phosphorylatable
serines (Ser-25 and Ser-63). Biochem J 1994;300:331–338.
80. Maucuer A, Moreau J, Mechali M, Sobel A. Stathmin gene family:
phylogenetic conservation and developmental regulation in Xenopus. J
Biol Chem 1993;268:16420–16429.
81. Compton DA. Focusing on spindle poles. J Cell Sci 1998;111:1477–1481.
82. Inoué, Salmon ED. Force generation by microtubule assembly/disassembly in mitosis and related movements. Mol Biol Cell 1995;6:1619–
1640.
83. Riederer BM, Pellier V, Antonsson B, Paolo GD, Stimpson SA, Lütjens R,
Catsicas S, Grenningloh G. Regulation of microtubule dynamics by the
neuronal protein SCG10. Proc Natl Acad Sci USA 1997;94:741–745.
84. Küntziger T, Gavet O, Manceau V, Sobel A, Bornens M. Phosphorylation of
the tubulin-sequestering protein stathmin is regulated by microtubule
assembly. Mol Biol Cell 1998;9s:257a.