print ncb1102
13/2/04
3:07 PM
Page 232
LETTERS
Correcting improper chromosome–spindle
attachments during cell division
Michael A. Lampson1, Kishore Renduchitala1, Alexey Khodjakov2 and Tarun M. Kapoor1,3
For accurate segregation of chromosomes during cell division,
microtubule fibres must attach sister kinetochores to opposite
poles of the mitotic spindle (bi-orientation). Aurora kinases are
linked to oncogenesis1 and have been implicated in the
regulation of chromosome–microtubule attachments2.
Although loss of Aurora kinase activity causes an accumulation
of mal-orientated chromosomes in dividing cells3,4, it is not
known how the active kinase corrects improper chromosome
attachments. The use of reversible small-molecule inhibitors
allows activation of protein function in living vertebrate cells
with temporal control. Here we show that by removal of smallmolecule inhibitors, controlled activation of Aurora kinase
during mitosis can correct chromosome attachment errors by
selective disassembly of kinetochore–microtubule fibres,
rather than by alternative mechanisms involving initial release
of microtubules from either kinetochores or spindle poles5–7.
Observation of chromosomes and microtubule dynamics with
real-time high-resolution microscopy showed that malorientated, but not bi-orientated, chromosomes move to the
spindle pole as both kinetochore–microtubule fibres shorten,
followed by alignment at the metaphase plate. Our results
provide direct evidence for a mechanism required for the
maintenance of genome integrity during cell division.
Detection of errors in chromosome attachment to the mitotic spindle
has been studied extensively. Mechanisms have been proposed that
respond to tension at kinetochores or to accumulation of microtubule
attachments at kinetochores (reviewed in ref. 8). However, it is not
known how improper chromosome attachments are corrected.
Chromosome mal-orientations have been observed after inhibition of
Aurora kinase activity3,4, but it has not been possible to activate the
kinase and observe the correction process directly. Two unrelated
small-molecule inhibitors of Aurora kinase activity have been used in
recent studies: hesperadin3,9, and one referred to here as AKI-1
(Aurora kinase inhibitor-1)10,11 from a chemical structural class previously described12 (see Supplementary Information, Fig. S1). We used
both inhibitors in our experiments, as they target Aurora kinases but
are unlikely to have overlapping off-target activities. We first tested
whether inhibition of Aurora kinase activity by these small molecules
is reversible, which would allow activation of the kinase in living cells
by removal of the inhibitor.
To assay for Aurora kinase activity in mitotic cells, we used a known
substrate, histone H3 (ref. 13), whose phosphorylation during mitosis
has been previously used to demonstrate inhibition of Aurora kinase
activity3,12. In cells treated with either of two Aurora kinase inhibitors,
chromosomes failed to align at the metaphase plate (Fig. 1c, e; compare with control cells, Fig.1a, b), consistent with Aurora kinase inhibition by other methods14,15. As expected, histone H3 phosphorylation
was reduced by over 70% relative to control cells (Fig. 1d, f, k).
Phosphorylation was fully restored 30 min after removal of hesperadin (Fig. 1h, k). Recovery from AKI-1 treatment required ~60 min
in this assay (Fig. 1j–k). These data demonstrate that both inhibitors
are reversible in living cells at the level of kinase activity.
To examine the effects of Aurora kinase activation, we developed a
method to accumulate chromosomes with improper attachments to
the spindle. In the presence of the small molecule monastrol, cells arrest
in mitosis with monopolar spindles in which kinetochore–microtubule
fibres (K-fibres) extend in a radial array from the un-separated spindle
poles and chromosomes are positioned at the microtubule plus-ends,
away from the pole16. Many of these chromosomes are mal-orientated,
as K-fibres attach both sister kinetochores to the same pole (syntelic
mal-orientation)17. After removal of monastrol, spindles become bipolar and chromosomes align (Fig. 2a)18. Cells were incubated with
monastrol to accumulate monopolar mitoses and then released into
either of the two Aurora kinase inhibitors. Examination of fixed cells
showed that after spindle bipolarization in the presence of the kinase
inhibitor, many chromosomes remain mal-orientated with K-fibre
pairs attached to the same spindle pole (syntelic; Fig. 2b, arrows; also
see Supplementary Information, Fig. S2). As the K-fibres extend away
from the spindle body, both chromosomes and K-fibres could be
clearly distinguished. This method, which was confirmed for both
inhibitors (data not shown), allows us to examine mal-orientated chromosomes after Aurora kinase activation.
Aurora kinase inhibition could accumulate mal-orientated chromosomes during mitosis by either increasing the formation, or decreasing
the correction, of improper chromosome attachments to the spindle.
It has not been possible to distinguish between these possibilities by
analysis of fixed cells, which shows only the steady-state situation3.
1Laboratory
of Chemistry and Cell Biology, Rockefeller University, New York, NY 10021, USA. 2Division of Molecular Medicine, Wadsworth Center, New York State
Department of Health, Albany, NY 12201, USA.
3Correspndence should be addressed to T.M.K. (e-mail: [email protected])
Published online: 8 February 2004; DOI: 10.1038/ncb1102
232
NATURE CELL BIOLOGY VOLUME 6 | NUMBER 3 | MARCH 2004
©2004 Nature Publishing Group
©2004 Nature Publishing Group
print ncb1102
13/2/04
3:07 PM
Page 233
LETTERS
b
c
d
e
g
f
i
h
j
k
Hesperadin
Phospho-H3 (mean, s.e.m.)
a
AKI-1
1.5
1.0
0.5
0.0
Inhibitor
Inhibitor
removed
Figure 1 Inhibition of Aurora kinase is reversible. (a–f) Cells were treated (1 h)
with no Aurora kinase inhibitor (a, b), 100 nM hesperadin (c, d) or 20 µM AKI1 (e, f) before processing for immunofluorescence microscopy. (g–j) In parallel
experiments, inhibitors were removed after a 1-h treatment. After an additional
30 min (hesperadin; g, h) or 60 min (AKI-1; i, j), cells were processed for
immunofluorescence microscopy. In the top panels, tubulin is shown in green
and DNA is shown in blue. Bottom panels show phospho-H3 staining. Scale
bar represents 5 µm. (k) Quantification of phospho-H3 immunofluorescence,
normalized to controls (no inhibitor), for conditions c–j. The proteasome
inhibitor MG132 was included in all experiments to prevent mitotic exit.
Both chromosomes and individual K-fibres were examined over time
in living cells expressing α-tubulin–green fluorescent protein (GFP)
using near-simultaneous, three-dimensional GFP fluorescence and
differential interference contrast (DIC) microscopy. In monopolar
spindles, K-fibres formed the expected radial array around the pole,
with chromosomes at their plus-ends (Fig. 2c, f). An Aurora kinase
inhibitor was added after release of the arrest, and spindle poles separated rapidly, as expected18 (Fig. 2d). Some chromosomes maintained
their positions away from the spindle body during bipolarization
(Fig. 2g, h), with both K-fibres clearly visible and attached to the same
spindle pole throughout the observation time (1 h; Fig. 2d, e; also see
Supplementary Information, Movie 1), demonstrating that without
Aurora kinase activity these mal-orientations are stable and fail to correct. These persistent mal-orientations provide a starting point from
which to examine their correction.
After spindle bipolarization in the presence of an Aurora kinase
inhibitor to accumulate chromosome mal-orientations, the inhibitor
was removed to activate the kinase and cells were imaged by 3D GFP
fluorescence and DIC microscopy (Fig. 3a). Immediately after kinase
activation, multiple mal-orientated chromosomes were apparent
(Fig. 3b, f). A surprising observation was that the correction process
began with rapid movement of mal-orientated chromosomes to the
pole (Fig. 3b–r). Chromosomes began poleward movement an average
of 27 min (s.d = 7 min, n = 10) after removal of the inhibitor, consistent with the timing of kinase activation as determined by recovery of
histone H3 phosphorylation (Fig. 1k). From the pole, chromosomes
moved only in the direction of the opposite pole, resulting in alignment at the metaphase plate (Fig. 3n–r). As an example, one chromosome (Fig. 3, arrowheads) began poleward movement 28 min after
removal of the inhibitor, reached the pole by 30 min and aligned by 36
min (Fig. 3t). This behaviour was observed for seven syntelic mal-orientated chromosomes in the two cells shown (Fig. 3; and see
Supplementary Information, Movies 2, 3), and in additional experiments (n = 19 cells) using either inhibitor (data not shown).
This poleward movement is reminiscent of events that have been
described during early stages of mitosis, when chromosomes also make
direct movements to the spindle pole as a precursor to bi-orientation
and metaphase alignment19. For comparison with this process, we
examined the details of poleward chromosome movement after activation of Aurora kinase. Our analysis revealed that both K-fibres shortened as a syntelic chromosome moved to the pole (Fig. 3u–w, arrow),
indicating that kinetochores remained attached to the microtubule plus
ends. K-fibre pairs were observed attached to sister chromatids until
their length reduced by 72% (s.d. = 5%, n = 6). The proximity of
chromosomes to the spindle poles limited our ability to resolve individual K-fibres within 1.5 µm of a pole. The average overall velocity
during poleward movement was 1.2 µm min−1 (s.d. = 0.4 µm min−1,
n = 7), consistent with movement of chromosomes attached to
K-fibre plus ends during prometaphase and metaphase20, rather than
the more rapid movements (~18 µm min−1) along the lateral surface
of a microtubule19,20. Chromosomes that were already aligned at the
metaphase plate did not seem to be affected by Aurora kinase activation, indicating that K-fibre disassembly was selective for mal-orientated chromosomes.
K-fibres can in principle disassemble from either end, and Aurora
kinase family members have been shown to localize to both kinetochores (Aurora B) and centrosomes (Aurora A; reviewed in ref. 1),
where K-fibre plus and minus ends are attached. It is not known
whether Aurora A or B is specifically inhibited by the small molecules
at concentrations used in our experiments3,12. Although we cannot
exclude a role for Aurora A kinase activity in error correction,
expected functions of Aurora A in centrosome separation21–23 were
not inhibited in our experiments. Furthermore, the phenotypes we
observed in the presence of Aurora kinase inhibitors are consistent
with loss of Aurora B function at kinetochores, as shown previously
by RNA interference (RNAi) and other methods3,12,14,15,24. In addition to chromosome misalignment and checkpoint inactivation, we
found that BubR1, dynactin and MCAK (mitotic centromere-associated kinesin) fail to localize to kinetochores in the presence of the
inhibitors (data not shown). Our observation of selective disassembly
of K-fibres attached to mal-orientated, but not bi-orientated, chromosomes may be explained by a mechanism that responds to forces in
opposite directions at kinetochores (tension)25. Aurora B localizes to
kinetochores and may function as the tension sensor, as has been suggested for the yeast Aurora homolog, Ipl1 (ref. 26). One possibility is
that Aurora kinase is activated specifically at kinetochores of mal-orientated chromosomes. Alternatively, Aurora activation may occur at
all kinetochores, but tension would prevent K-fibre disassembly at biorientated chromosomes because forces act in opposite directions.
This possibility could be tested by estimating forces at kinetochores,
as has been done by measuring distance between kinetochores for biorientated chromosomes27. At syntelic mal-orientated chromosomes,
however, we found that inter-kinetochore distance varies widely (ref.
17; M.A.L. and T.M.K., unpublished observations), more likely
NATURE CELL BIOLOGY VOLUME 6 | NUMBER 3 | MARCH 2004
©2004 Nature Publishing Group
©2004 Nature Publishing Group
233
print ncb1102
13/2/04
3:07 PM
Page 234
LETTERS
a
b
1
2
1
2
c
d
e
−5:00
f
36:50
g
58:09
h
Figure 2 Inhibition of Aurora kinase activity stabilizes chromosome malorientations. (a, b) Spindles bipolarize as cells recover from monastrol
treatment. After removal of monastrol, cells were incubated with (b) or
without (a) 20 µM AKI-1 for 1 h and then processed for
immunofluorescence microscopy. Optical sections for boxed regions (1 and
2) highlight mal-orientations. (c–h) After removal of monastrol and addition
of 100 nM hesperadin (t = 0), GFP–tubulin (c–e) and chromosomes (f–h)
were imaged live by near-simultaneous 3D confocal fluorescence microscopy
and DIC, respectively. Selected images from timelapse recordings are
shown. Arrows and arrowheads indicate chromosome mal-orientations (b,
d–h). MG132 was included in all incubations (a–h). Time is shown in
minutes and seconds. Scale bars represent 5 µm. Stable chromosome malorientations were also observed before anaphase onset in the absence of
MG132 (see Supplementary Information, Fig. S3).
reflecting the unusual chromosome geometry rather than tension
across kinetochores. A further molecular understanding of the correction mechanism will require determination of which of the many
Aurora substrates (for example, ref. 28) have roles in regulating Kfibre dynamics.
Experiments in model systems have suggested that syntelic chromosome mal-orientations (Fig. 4a) are corrected by K-fibre release from
either the kinetochore or the pole (Fig. 4b, c), followed by reattachment until the correct orientation is stabilized4,7. We found that syntelic mal-orientations can be corrected by selective K-fibre
disassembly, coupled with movement to the pole (Fig. 4d).
Chromosomes spend variable amounts of time at the pole; when they
move, however, movement is invariably directed towards the opposite
pole of the bipolar spindle. This observation suggests that the K-fibre
disassembly mechanism is active, keeping the chromosome at the pole
until a microtubule attachment is formed to the opposite pole. After
this attachment, the K-fibre disassembly mechanism would be turned
off, allowing movement to the metaphase plate (Fig. 4e). It has previously been shown that even a single microtubule attachment from the
opposite pole is sufficient for movement29. We expect that such an
attachment will be difficult to detect near the pole, where microtubule
density is high.
234
NATURE CELL BIOLOGY VOLUME 6 | NUMBER 3 | MARCH 2004
©2004 Nature Publishing Group
©2004 Nature Publishing Group
print ncb1102
13/2/04
3:08 PM
Page 235
LETTERS
a
2h
1h
Monsastrol
Live imaging
Aurora kinase inhibitor
MG132 (proteasome inhibitor)
b
c
d
5:37
e
13:34
23:49
27:00
f
g
h
i
j
k
l
m
34:48
o
r
s
44:55
Distance to pole (µm)
31:13
n
u
v
13:12
15:12
42:15
43:16
p
q
Chromosome 1
(arrow)
t
Chromosome 2
(arrowhead)
Near pole
Far pole
20
10
0
20
Metaphase
alignment
Metaphase
alignment
25
30
35
Time (min)
19:26
25
30
35
40
Time (min)
x
w
Figure 3 Correction of chromosome mal-orientations after activation of
Aurora kinase. (a) A schematic representation of the experiment. Spindles
became polar in the presence of hesperadin (100–200 nM). Hesperadin was
then removed (t = 0), and GFP–tubulin (b–e, j–m and u–y) and
chromosomes (f–i, n–r) were imaged live by 3D confocal fluorescence
microscopy and DIC, respectively. (b–r) Two chromosomes and associated
45 20
40
24:56
y
26:04
K-fibres (arrows and arrowheads) are tracked until metaphase alignment. (s,
t) Chromosome position relative to spindle poles (marked by circles in DIC
images). (u–y) K-fibre disassembly after Aurora kinase activation in another
cell (arrow shows a K-fibre pair; chromosome not shown). Time is shown in
minutes and seconds. Scale bars represent 5 µm. Timelapse recordings for
these two cells are provided in Supplementary Information.
NATURE CELL BIOLOGY VOLUME 6 | NUMBER 3 | MARCH 2004
©2004 Nature Publishing Group
©2004 Nature Publishing Group
235
print ncb1102
13/2/04
3:08 PM
Page 236
LETTERS
a
K-fibre pair
Chromosome
Spindle pole
K-fibre release
from kinetochore
~
d
K-fibre release
from pole
e
~
~
~
c
~
b
~
Aurora kinase activation
Selective K-fibre
disassembly
Figure 4 Mechanisms to correct syntelic chromosome mal-orientations
during cell division. (a) Syntelic mal-orientation arises during mitosis if
K-fibres attach sister kinetochores to the same spindle pole. (b, c) Initial
K-fibre release hypothesis: Aurora kinase activity may correct malorientation by K-fibre release from either the kinetochore (b) or the
spindle pole (c). (d) Mal-orientated chromosomes move to the pole as K
-fibres selectively disassemble. K-fibres initially remain attached to both
kinetochores and the spindle pole. We propose that capture by
microtubule(s) from the opposite pole results in alignment at the
metaphase plate (e).
Attachment of chromosomes to the mitotic spindle is a stochastic
and dynamic process, during which errors arise (recently reviewed in
ref. 30). These errors are corrected by a complex regulatory network
that ensures accurate chromosome segregation during cell division.
Based on direct observations of the dynamic responses of spindle
microtubules and chromosomes in living cells to reversible perturbations of Aurora kinases, we propose a model for the molecular basis of
correcting errors in chromosome–spindle attachments. Activation of
Aurora kinases selectively eliminates improper chromosome-spindle
attachments by disassembling the microtubule fibres (Fig. 4d). Such
spatially specific control over cytoskeletal polymerization is likely to be
a general mechanism for regulating intracellular transport and cell
migration.
chloride and 0.2% Triton X-100. Images were acquired on a Carl Zeiss Axiovert
200M microscope, with either a 63× 1.4 NA objective (for presentation) or a
40× 0.75 NA objective (for quantification). Phospho-H3 immunofluorescence
intensity was quantified in mitotic cells using Metamorph software (Universal
Imaging, Downington, PA) and normalized to the control (no inhibitor; n ≥ 20
cells for each data point). For each spindle, Hoechst fluorescence was used to
define the chromosome region, and phospho-H3 immunofluorescence was
measured in that region. Background fluorescence, estimated from regions with
no DNA, was subtracted.
For spindle bipolarization (Fig. 2a, b), cells were first arrested in mitosis for
4 h by monastrol treatment. Monastrol was then removed, and MG132 was
added together with hesperadin or AKI-1. 1 h after removal of monastrol, cells
were fixed with calcium by permeabilization for 90 s at 37 °C in 100 mM PIPES
at pH 6.8, 1 mM magnesium chloride, 0.1 mM calcium chloride and 0.1%
Triton X-100 to depolymerize non-kinetochore microtubules, followed by fixation in the same buffer supplemented with 3.7% formaldehyde. Images were
acquired as z-stacks on a DeltaVision Image Restoration Microscope (Applied
Precision Instruments, Issaquah, WA and Olympus, Melville, NY) using a 60×
objective and processed by iterative constrained deconvolution. Maximal intensity projections are shown, and optical sections (Fig. 2b, boxed regions) show
individual K-fibres more clearly.
METHODS
Inhibitors. Inhibitors were synthesized according to the procedures described
for hesperadin9 and AKI-1 (ref. 10). Chemical structures are shown in
Supplementary Information, Fig. S1. Hesperadin was used at 100–200 nM,
AKI-1 at 20 µM. Unless otherwise stated, the proteasome inhibitor MG132
(Sigma, St Louis, MO) was included in all incubations (20 µM) with Aurora
kinase inhibitors to prevent mitotic exit. The Eg5 inhibitor monastrol was used
at 100 µM.
Cell culture and fixed timepoint experiments. PtK2 cells expressing α-tubulin–GFP18 were used for all experiments. Cells were cultured in Ham’s F-12
medium (Invitrogen, Carlsbad, CA), supplemented with 10% foetal bovine
serum (Sigma) at 37 °C in a humidified atmosphere with 5% CO2. Antibodies
used for immunofluorescence microscopy were against phospho-histone H3
(Upstate Biotechnology, Lake Placid, NY) or α-tubulin (DM1α; Sigma). DNA
was labelled with Hoechst 33342 (Sigma).
For phospho-H3 immunofluorescence microscopy (Fig. 1), cells were incubated with MG132 together with either hesperadin or AKI-1. Cells were fixed
after 1 h, or the inhibitor was removed and cells were fixed after an additional
30–60 min in the presence of MG132. Cells were fixed at 37 °C with 3.7%
formaldehyde in 100 mM PIPES at pH 6.8, 10 mM EGTA, 1 mM magnesium
Live imaging. Cells were grown on 24 mm2 coverslips mounted in a Rose chamber for imaging using L-15 medium without phenol red. Washes were performed by flowing media through the chamber. Cells were incubated in
monastrol for 2 h or less (to prevent cells from rounding up). Monastrol was
removed and an Aurora kinase inhibitor was added with MG132. As indicated,
the Aurora kinase inhibitor was removed after 1 h and cells were kept in
MG132. Cells were imaged immediately after removal of monastrol (Fig. 2) or
after removal of the Aurora kinase inhibitor (Fig. 3).
Confocal GFP fluorescence time-lapse sequences were acquired on a Carl
Zeiss Axiovert 200M microscope equipped with a z-motor and a 100× 1.4 NA
objective. Experiments were performed in a temperature-controlled chamber
(Solent Scientific, Portsmouth, UK) maintained at 37 °C. z-stacks (six sections
separated by 1.0 µm) of GFP fluorescence were taken every 30 s with a
PerkinElmer Wallac (Boston, MA) UltraView confocal head with a 488-nm
236
NATURE CELL BIOLOGY VOLUME 6 | NUMBER 3 | MARCH 2004
©2004 Nature Publishing Group
©2004 Nature Publishing Group
print ncb1102
13/2/04
3:08 PM
Page 237
LETTERS
excitation filter and an argon ion laser (Melles Griot, Carlsbad, CA). A single
DIC image was taken immediately after each GFP stack. Images were acquired
with an Orca ER cooled CCD camera (Hamamatsu, Hamamatsu City, Japan)
with 2 × 2 pixel binning. Metamorph software (Universal Imaging) was used
for acquisition and analysis. Maximal intensity projections of GFP–tubulin zstacks and single DIC images are shown for selected timepoints. A 2 × 2 lowpass filter was applied to all images for presentation.
Note: Supplementary Information is available on the Nature Cell Biology website.
ACKNOWLEDGEMENTS
We acknowledge A. North and The Rockefeller University Bioimaging Facility. We
thank W. Brinkley for the generous gift of CREST antiserum. This work was
supported by National Institutes of Health grants GM65933 (T.M.K.) and
GM59363 (A.K.). M.A.L. is a Goelet fellow.
COMPETING FINANCIAL INTERESTS
The authors declare that they have no competing financial interests.
Received 19 December 2003; accepted 13 January 2004
Published online at http://www.nature.com/naturecellbiology.
1. Giet, R. & Prigent, C. Aurora/Ipl1p-related kinases, a new oncogenic family of mitotic
serine-threonine kinases. J. Cell Sci. 112, 3591–3601 (1999).
2. Biggins, S. et al. The conserved protein kinase Ipl1 regulates microtubule binding to
kinetochores in budding yeast. Genes Dev 13, 532–544 (1999).
3. Hauf, S. et al. The small molecule Hesperadin reveals a role for Aurora B in correcting
kinetochore-microtubule attachment and in maintaining the spindle assembly checkpoint. J. Cell Biol. 161, 281–294 (2003).
4. Tanaka, T. U. et al. Evidence that the Ipl1–Sli15 (Aurora kinase–INCENP) complex
promotes chromosome bi-orientation by altering kinetochore-spindle pole connections. Cell 108, 317–329 (2002).
5. Nicklas, R. B. How cells get the right chromosomes. Science 275, 632–637 (1997).
6. Ault, J. G. & Rieder, C. L. Chromosome mal-orientation and reorientation during mitosis. Cell Motil. Cytoskel. 22, 155–159 (1992).
7. Nicklas, R. B. & Ward, S. C. Elements of error correction in mitosis: microtubule capture, release, and tension. J. Cell Biol. 126, 1241–1253 (1994).
8. Rieder, C. L. & Salmon, E. D. The vertebrate cell kinetochore and its roles during
mitosis. Trends Cell Biol. 8, 310–318 (1998).
9. Walter, R. et al. in PCT Int. Appl. 112 ((Boehringer Ingelheim Pharma K.-G.,
Germany, WO, 2002).
10. Mortlock, A. A. & Jung, F. H. in PCT Int. Appl. 62 ((Astrazeneca AB, Sweden;
Astrazeneca UK Limited, WO, 2001).
11. Straight, A. F. et al. Dissecting temporal and spatial control of cytokinesis with a
myosin II Inhibitor. Science 299, 1743–1747 (2003).
12. Ditchfield, C. et al. Aurora B couples chromosome alignment with anaphase by targeting
BubR1, Mad2 and Cenp-E to kinetochores. J. Cell Biol. 161, 267–280 (2003).
13. Hsu, J. Y. et al. Mitotic phosphorylation of histone H3 is governed by Ipl1/aurora
kinase and Glc7/PP1 phosphatase in budding yeast and nematodes. Cell 102,
279–291 (2000).
14. Murata-Hori, M. & Wang, Y. L. The kinase activity of aurora B is required for kinetochore–microtubule interactions during mitosis. Curr. Biol. 12, 894–899 (2002).
15. Kallio, M. J., McCleland, M. L., Stukenberg, P. T. & Gorbsky, G. J. Inhibition of aurora
B kinase blocks chromosome segregation, overrides the spindle checkpoint, and perturbs microtubule dynamics in mitosis. Curr. Biol. 12, 900–905 (2002).
16. Mayer, T. U. et al. Small molecule inhibitor of mitotic spindle bipolarity identified in
a phenotype-based screen. Science 286, 971–974 (1999).
17. Kapoor, T. M., Mayer, T. U., Coughlin, M. L. & Mitchison, T. J. Probing spindle assembly mechanisms with monastrol, a small molecule inhibitor of the mitotic kinesin,
Eg5. J. Cell Biol. 150, 975–988 (2000).
18. Khodjakov, A., Copenagle, L., Gordon, M. B., Compton, D. A. & Kapoor, T. M. Minusend capture of preformed kinetochore fibers contributes to spindle morphogenesis. J.
Cell Biol. 160, 671–683 (2003).
19. Rieder, C. L. & Alexander, S. P. Kinetochores are transported poleward along a single
astral microtubule during chromosome attachment to the spindle in newt lung cells.
J. Cell Biol. 110, 81–95 (1990).
20. Skibbens, R. V., Skeen, V. P. & Salmon, E. D. Directional instability of kinetochore
motility during chromosome congression and segregation in mitotic newt lung cells: a
push-pull mechanism. J. Cell Biol. 122, 859–875 (1993).
21. Glover, D. M., Leibowitz, M. H., McLean, D. A. & Parry, H. Mutations in aurora prevent centrosome separation leading to the formation of monopolar spindles. Cell 81,
95–105 (1995).
22. Hannak, E., Kirkham, M., Hyman, A. A. & Oegema, K. Aurora-A kinase is required for centrosome maturation in Caenorhabditis elegans. J. Cell Biol. 155, 1109–1116 (2001).
23. Roghi, C. et al. The Xenopus protein kinase pEg2 associates with the centrosome in a
cell cycle-dependent manner, binds to the spindle microtubules and is involved in
bipolar mitotic spindle assembly. J. Cell Sci. 111, 557–572 (1998).
24. Murata-Hori, M., Tatsuka, M. & Wang, Y. L. Probing the dynamics and functions of
aurora B kinase in living cells during mitosis and cytokinesis. Mol. Biol. Cell 13,
1099–1108 (2002).
25. Li, X. & Nicklas, R. B. Mitotic forces control a cell-cycle checkpoint. Nature 373,
630–632 (1995).
26. Biggins, S. & Murray, A. W. The budding yeast protein kinase Ipl1/Aurora allows the
absence of tension to activate the spindle checkpoint. Genes Dev. 15, 3118–3129
(2001).
27. Waters, J. C., Chen, R. H., Murray, A. W. & Salmon, E. D. Localization of Mad2 to
kinetochores depends on microtubule attachment, not tension. J. Cell Biol. 141,
1181–1191 (1998).
28. Cheeseman, I. M. et al. Phospho-regulation of kinetochore-microtubule attachments
by the Aurora kinase Ipl1p. Cell 111, 163–172 (2002).
29. McEwen, B. F., Heagle, A. B., Cassels, G. O., Buttle, K. F. & Rieder, C. L.
Kinetochore fiber maturation in PtK1 cells and its implications for the mechanisms of
chromosome congression and anaphase onset. J. Cell Biol. 137, 1567–1580 (1997).
30. Cleveland, D. W., Mao, Y. & Sullivan, K. F. Centromeres and kinetochores: from epigenetics to mitotic checkpoint signaling. Cell 112, 407–421 (2003).
NATURE CELL BIOLOGY VOLUME 6 | NUMBER 3 | MARCH 2004
©2004 Nature Publishing Group
©2004 Nature Publishing Group
237
S U P P L E M E N TA R Y I N F O R M AT I O N
1
1
2
2
Figure S2. Kinetochore staining demonstrates syntelic chromosome malorientations upon inhibition of Aurora kinase. Spindles bipolarized as cells
recovered from monastrol treatment. After monastrol removal, cells were
incubated 1 hr with 100 nM hesperadin and processed for
immunofluorescence. Optical sections for boxed regions (1 and 2) highlight
two chromosomes with syntelic mal-orientations. MG132 was included with
hesperadin to prevent mitotic exit. Tubulin (green), DNA (blue), CREST
(red) labels kinetochores. Scale bars 3 µm.
Figure S1. Chemical structures of Aurora kinase inhibitors.
WWW.NATURE.COM/NATURECELLBIOLOGY
1
©2004 Nature Publishing Group
S U P P L E M E N TA R Y I N F O R M AT I O N
a
b
c
10:00
d
20:01
e
33:54
f
Figure S3. Syntelic attachments are persistent in the absence of MG132, if
Aurora kinase is inhibited. Spindles bipolarized as cells recovered from
monastrol treatment, in the presence of an Aurora kinase inhibitor. Arrows
show persistent syntelic attachments in live-cell recordings of chromosome
and tubulin dynamics. As expected, after 30 minutes cells enter anaphase
with improper chromosome attachments due to inhibition of Aurora kinase.
Scale bar 5 µm.
Movie 1. Movie corresponds to images shown in Fig 2c-h. Timestamp
(min:s) starts ~5 min after monastrol washout. GFP-tubulin (a), DIC (b).
Movie 2. Movie corresponds to images shown in Fig 3b-r. Timestamp
(min:s) starts ~5 min after hesperadin washout. GFP-tubulin (a), DIC (b).
Movie 3. Movie corresponds to images shown in Fig 3u-y. Timestamp
(min:s) starts ~5 min after hesperadin washout. GFP-tubulin (a), DIC (b).
2
WWW.NATURE.COM/NATURECELLBIOLOGY
©2004 Nature Publishing Group