Mitosis is the critical time of the cell cycle, during which the
genetic material is faithfully distributed among daughter cells.
Errors during the mitotic division result in the uneven
segregation of chromosomes, yielding aneuploid or polyploid
cells. Such genomic imbalances are among the most common
hallmarks of cancer and are regarded as crucial in tumor
progression (Lengauer et al., 1998; Pihan and Doxsey, 1999).
Correct assembly and function of the mitotic apparatus are
therefore essential to ensure the balanced transmission of
genetic information through cell division.
The Ran GTPase network has attracted increasing interest
during the past 10 years as the major regulator of nucleo-
cytoplasmic transport in interphase cells. The directionality of
transport in and out of the nucleus has been shown to rely on
the different distributions of nucleotide-bound forms of Ran
in specific subcellular compartments: Ran-GTP is generated
essentially in the nucleus, where the RCC1 nucleotide
exchange factor resides, whereas factors activating GTP
hydrolysis (RanGAP1 and RanBP1) are largely cytoplasmic
(Clarke and Zhang, 2001; Hetzer et al., 2002; Dasso, 2002).
Nuclear RanGTP promotes the dissociation of import
complexes – and hence the release of nuclear proteins in the
nucleoplasm – as well as the assembly of export complexes,
which, conversely, mediate transport of cytoplasmic proteins
and RNAs to the cytoplasm.
More recent evidence also indicate that the Ran system
carries out mitotic regulatory functions after nuclear envelope
breakdown (NEB). In Xenopus-oocyte-extract-based in vitro
systems, RanGTP and RCC1 are required for the assembly of
mitotic microtubule (MT) arrays in spindle-like structures
(Kalab et al., 1999; Ohba et al., 1999; Wilde and Zheng, 1999;
Carazo-Salas et al., 1999). This is largely due to the ability of
GTP-bound Ran to regulate the release of active ‘aster-
promoting activities’ (APAs), including NuMA and TPX2
(Gruss et al., 2001; Nachury et al., 2001; Wiese et al., 2001).
In the presence of low concentrations of RanGTP, APAs are
sequestered in inactivating complexes with importin α and β;
APAs need be released in the free form in the presence of
RanGTP to promote spindle assembly. Thus, the functional
role of Ran in nucleo-cytoplasmic transport and in spindle
formation relies essentially on one same mechanism – the
ability of RanGTP locally to dissociate macromolecular
complexes formed by import vectors and their partners
(Melchior, 2001; Dasso, 2002). In this framework, the ability
of released NuMA and TPX2 to orchestrate spindle assembly
is essentially determined by the redistribution of nuclear
and cytoplasmic components after NEB. The underlying
biochemical basis of RanGTP activity in transport and in
mitosis is otherwise identical except for the different
localization of molecules that act as downstream targets of the
Ran system before and after NEB. Because RCC1 remains
largely chromatin-bound throughout mitosis in Xenopus
extract (Carazo-Salas et al., 1999) and in somatic cells
(Guarguaglini et al., 2000; Moore et al., 2002), GTP exchange
on Ran during mitosis is expected to take place near
The Ran GTPase plays a central function in control of
nucleo-cytoplasmic transport in interphase. Mitotic roles of
Ran have also been firmly established in Xenopus oocyte
extracts. In this system, Ran-GTP, or the RCC1 exchange
factor for Ran, drive spindle assembly by regulating the
availability of ‘aster-promoting activities’. In previous
studies to assess whether the Ran network also influences
mitosis in mammalian cells, we found that overexpression
of Ran-binding protein 1 (RanBP1), a major effector of
Ran, induces multipolar spindles. We now show that these
abnormal spindles are generated through loss of cohesion
in mitotic centrosomes. Specifically, RanBP1 excess induces
splitting of mother and daughter centrioles at spindle
poles; the resulting split centrioles can individually
organize functional microtubule arrays, giving rise to
functional spindle poles. RanBP1-dependent centrosome
splitting is specifically induced in mitosis and requires
microtubule integrity and Eg5 activity. In addition, we have
identified a fraction of RanBP1 at the centrosome. These
data indicate that overexpressed RanBP1 interferes with
crucial factor(s) that control structural and dynamic
features of centrosomes during mitosis and contribute to
uncover novel mitotic functions downstream of the Ran
Key words: RanBP1, Ran GTPase, Mitosis, Spindle pole, Centriole,
Mammalian RanBP1 regulates centrosome cohesion
Barbara Di Fiore1, Marilena Ciciarello1, Rosamaria Mangiacasale1, Antonella Palena1, Anne-Marie Tassin2,
Enrico Cundari1and Patrizia Lavia1,*
1CNR Institute of Molecular Biology and Pathology, Section of Genetics, c/o University ‘La Sapienza’, 00185 Rome, Italy
2Institut Curie, Section Recherche, UMR144-CNRS, 75248 Paris Cedex, France
*Author for correspondence (e-mail: email@example.com)
Accepted 16 April 2003
Journal of Cell Science 116, 3399-3411 © 2003 The Company of Biologists Ltd
JCS ePress online publication date 2 July 2003
chromosomes. Indeed, visual evidence for the bulk of RanGTP
being concentrated near mitotic chromosomes has recently
been provided (Kalab et al., 2002).
Different mechanisms underlie spindle assembly in
mammalian somatic cells and in meiotic Xenopus extracts,
despite of the high conservation of molecular components
(Merdes and Cleveland, 1997). One obvious difference in
spindle organization lies in the role played by centrosomes in
somatic cells but not in meiotic extract. Centrosomes act as the
major organizing centers for MT nucleation in somatic cells,
and hence their function is intimately connected with the
organization of spindle poles. Thus, specific aspects of Ran-
controlled processes during mitosis might differ in these
Evidence from living cells, albeit still fragmentary, clearly
implicate the Ran network in control of spindle organization
and function. Injection of anti-RanBP1 antibody in mitosis
perturbs MT dynamics to the point of impairing complete
chromosome segregation (Guarguaglini et al., 2000).
Microinjection of a deleted importin β protein, lacking the
Ran-binding domain, causes misassembly of the spindle and
chromosome misalignment (Nachury et al., 2001). Aberrant
chromosome alignment is also seen in cells overexpressing
RanBP1, associated with the formation of multipolar spindles
(Guarguaglini et al., 2000). Consistent with the absence of a
specific checkpoint that would detect multipolar spindles
(Sluder et al., 1997), these cells do not arrest at metaphase but
progress to ana-telophase and segregate uneven groups of
chromosomes. Similar defects have been reported following
expression of a RCC1 mutant that mislocalizes to the mitotic
cytoplasm (Moore et al., 2002). Thus, the Ran network, as well
as regulating spindle assembly in the proximity of chromatin
in the Xenopus system, also controls aspects of spindle
function in mammalian cells, in which spindle pole formation
and mitotic MT nucleation are directed from centrosomes. To
achieve these functions, components of the Ran network might
locally act at crucial mitotic locations in animal cells.
Here, we focus on mitotic functions of the RanBP1 protein
in mammalian cells. Expression of the mammalian RanBP1
gene varies during the cell cycle (Di Matteo et al., 1995; Di
Fiore et al., 1999), with highest protein levels in G2 and M
phases, and an abrupt decline in late telophase (Guarguaglini
et al., 2000). As recalled above, RanBP1 overexpression yields
abnormal mitotic spindles with multiple poles (Guarguaglini et
al., 2000). To date, this is one of the clearest phenotypes
visualized during the mammalian mitosis under alteration of
Ran network components. We have now sought to identify the
underlying defects of multipolar spindle formation.
Correct reproduction and structural organization of
centrosomes are crucial for the establishment of the spindle
bipolarity. Multipolar spindles that direct chromosome mis-
segregation often form in consequence of centrosome
overduplication during cell transformation (Lingle and
Salisbury, 2000; Brinkley, 2001; Doxsey, 2001). Here, we
report that RanBP1 does not influence the centrosome
duplication cycle but instead induces a specific and distinct
aberration (unscheduled splitting between mother and daughter
centrioles at spindle poles). This process is specifically induced
after NEB in a MT- and Eg5-dependent manner. Split
centrioles retain the ability to anchor functional MT arrays
and give rise to multipolar spindles that direct abnormal
chromosome segregation. We also show that a RanBP1 fraction
localizes to centrosomes. These results uncover a novel aspect
of mitotic centrosome cohesion, the maintenance of which is
important to ensure proper chromosome segregation, and
indicate that this function is sensitive to RanBP1 levels.
Materials and Methods
The murine RanBP1 ORF was amplified by PCR from the pCMV-
RanBP1 construct (Battistoni et al., 1997) using the following
oligonucleotide sets: (i) CCGGAATTCATGGCTGCGCAGGG-
AGAG and CGCGGATCCCAGGTCATCATCCTCATCCG, for
ligation to the pEGFP-N1 and pDsRed1-N1 EcoRI/BamHI-digested
vectors (both from Clontech); and (ii) AGAATTCGTCGCGC-
GCGCCCCCATGGCGGCCGCCAA and CGCCTCGAGCTAA-
TTCTC, encoding the hemagglutinin (HA) epitope, for ligation to the
basic pCMV vector (previously named pX) (Battistoni et al., 1997)
after EcoRI/XhoI digestion. Ligation of the amplified products yielded
the pRanBP1-GFP, pRanBP1-RFP and pRanBP1-HA expression
constructs, carrying the chimaeric tags at the C-terminus of the
RanBP1 sequence. By densitometric analysis of western blots, the
tagged and untagged RanBP1 expression plasmids yield a similar
increase (over fourfold) in levels of total RanBP1 protein compared
to non-transfected or vector-transfected cells.
Cell culture and synchronization
Murine NIH/3T3 embryo fibroblasts (ATCC CRL-1658), murine
L929 lung epithelial cells (ATCC CCL-1) and derived cell lines stably
transfected with centrin 1-GFP (Piel et al., 2000) (kindly given by M.
Bornens, Institut Curie, Paris), human HeLa epithelial cells (ATCC
CCL-2), were all grown in DMEM (Dulbecco’s Modified Eagle
Medium, Euroclone) supplemented with 10% fetal calf serum (FCS;
Gibco BRL) at 37°C in the presence of 5% CO2. Centrin 1-GFP L929
cell lines were cultured with G418 (500 µg ml–1, Gibco BRL). For
cell cycle synchronization experiments, cell cultures were maintained
in low FCS (0.5%) for at least 48 hours to induce quiescence, and
subsequently stimulated to synchronously re-enter the cell cycle by
raising the FCS concentration to 15%. Cells were collected 9 hours,
15 hours and 22 hours after stimulation. To analyse G1-S progression
to mitosis, NIH/3T3 and L929 cell cultures were blocked in the
presence of thymidine (Sigma Aldrich, 2 mM for NIH/3T3 and 5 mM
for L929 cells) for 24 hours, then released in complete DMEM
supplemented with 30 µM deoxycytidine (Sigma Aldrich) and
harvested 6-8 hours after release from thymidine arrest, when the cell
population was mostly in G2-M by fluorescence-activated cell sorting
(FACS) analysis and the mitotic index was highest by microscope
scoring. Where indicated, cell cultures were released from thymidine
arrest for 4-6 hours and subsequently exposed to 0.1 µg ml–1
nocodazole (NOC; Sigma Aldrich) or 100 µM monastrol (MA;
Tocris) for 4 hours before harvesting. Cells were then fixed, or
released in drug-free medium for 45 minutes (NOC) or 30 minutes
(MA). For localization experiments, thymidine-arrested and released
cultures were exposed to 1 µM Taxol (Sigma Aldrich) for 4 hours. In
all cases cell cycle phase synchronization was analysed by FACS
(Beckton Dickinson) as described (Battistoni et al., 1997).
NIH/3T3 cells were seeded in 60 mm Petri dishes onto sterile glass
coverslips and transfected using Fugene (Roche Diagnostic, 3 µl µg–1
DNA). L929 cells were transfected by electroporation (950 µF, 310
V) and reseeded onto sterile glass coverslips. Six hours after
transfection, the medium was replaced with fresh medium. Cells were
Journal of Cell Science 116 (16)
3401 RanBP1 and mitotic centrosome splitting
routinely collected 36-48 hours after transfection (asynchronous cell
cultures). Where indicated, transfected cell cultures were submitted to
synchronization protocols starting 6-10 hours after transfection; the
overall duration of thymidine arrest and release, with or without
mitosis-arresting drugs, covered 36-42 hours of culture after
transfection (see above). Cells were then harvested and processed for
parallel FACS and indirect immunofluorescence (IF) assays.
Goat polyclonal anti-RanBP1 (M-19 for murine cells and C-19 for
human cells) antibodies were from Santa Cruz Biotechnology and
were used 0.5 µg ml–1in western blotting and 2 µg ml–1in IF
experiments. Anti-HA (Y-11; Santa Cruz Biotechnology) antibody
was used at 1:100 dilution. Monoclonal Ran antibody (clone 20;
Transduction Laboratories) was used at 0.25 µg ml–1. Goat polyclonal
anti-RCC1 (C-20) and anti-RanGAP1 (N-19) antibodies (Santa Cruz
Biotechnology) were used at 1 µg ml–1and 2 µg ml–1, respectively.
Monoclonal α-tubulin (clone B-5-1-2; Sigma Aldrich) antibody was
used at 1:1000 dilution. Monoclonal (GTU-88) and rabbit polyclonal
anti-γ-tubulin antibodies (both from Sigma Aldrich) were used at
1:5000 dilution for western blotting and 1:1000 for IF assays.
Monoclonal anti-GT335 antibody (used 1:3000) was kindly provided
by P. Denoulet (Université Pierre et Marie Curie, Paris); rabbit
polyclonal anti-centrin 2 antibody (used at 1:2000 dilution) was from
M. Bornens (Institut Curie, Paris). The monoclonal antibody CTR453
(IgG2b) was generated in M. Bornens’s laboratory and has previously
been characterized as specific for the centrosome (Bailly et al., 1989).
Horseradish peroxidase (HRP)-conjugated secondary antibodies were
from Santa Cruz Biotechnology. Secondary antibodies conjugated to
fluorescein-, AMCA- (Jackson ImmunoResearch Laboratories),
rhodamine (Santa Cruz Biotechnology), Texas Red (Vector) and Cy-
3 (Amersham) were chosen depending on the basis of species
specificity and used as recommended by the suppliers.
Cells were grown on sterile glass coverslips, washed in PBS and fixed
in methanol for 6 minutes at –20°C or in 3% PFA, 30 mM sucrose
for 10 minutes at room temperature. Where indicated, cells were
permeabilized for 30 seconds in 0.5% Triton X-100 in PHEM (45 mM
PIPES pH 6.9, 45 mM HEPES pH 6.9, 10 mM EGTA, 5 mM MgCl2,
1 mM PMSF) before fixation. Incubation with primary antibodies was
carried out for 1 hour at 37°C. Secondary antibodies were incubated
for 45 minutes. DNA was counterstained with DAPI (0.1 µg ml–1).
Coverslips were then mounted in Vectashield (Vector). IF was also
performed as above using purified centrosomes from the KE37 cell
line (see below), after sedimentation onto coverslips (at 20,000 g, 15
minutes, 4°C) and fixation in methanol for 6 minutes at –20°C.
Fixed cell preparations were examined under an upright Olympus
AX70 microscope equipped for epifluorescence and images were
taken (100× objective) using either a CoolSnap FX, or a Photometrics
CCD camera. Where indicated, fluorescence intensity was quantified
in arbitrary units using Adobe Photoshop software on CCD images of
single cells acquired under identical exposure and gain setting within
each experiment. Video recording of living mitotic cells was carried
out on an inverted fluorescence microscope (Leica DMIRBE)
controlled by Metamorph software; cells transfected with pRanBP1-
RFP were identified on the red channel and images were taken every
10 minutes (10× objective). Confocal images were taken (60×
objective) using a TCS-SP2 confocal microscope (Leica) with a 488
nm laser excitation line.
To assess the statistical significance of the results, each experiment
was repeated at least three times; means and standard deviations were
calculated to compare the same category in different experiments.
This procedure consistently gave extremely low, statistically
insignificant deviations within each experimental condition. Data
from different experiments were therefore pooled and P values were
calculated on pooled data using the χ2test.
Protein extraction from the centrosomal fraction and
Centrosomes were isolated from the KE37 cell line as described by
Moudjou and Bornens (Moudjou and Bornens, 1994). Pelleted
centrosomes were incubated for 1 hour at 4°C in extraction buffer (20
mM Tris-HCl pH 7.4, 2 mM EDTA) alone or in the presence of: (i)
0.5% NP40 (1D buffer); (ii) 0.5% NP40 and 0.5% deoxycholate
(DOC, 2D buffer); (iii) 0.5% NP40, 0.5% DOC and 0.1% SDS (3D
buffer); (iv) 8 M urea. Centrosome-associated and non-associated
proteins were recovered in the pellet and supernatant fractions,
respectively, by centrifugation at 10,000 g for 15 minutes. Proteins
were separated through SDS-PAGE and transferred onto
nitrocellulose filters (Schleicher & Schuell). Filters were saturated in
5% milk in TBS (10 mM Tris-HCl pH 7.4, 150 mM NaCl) containing
0.1% Tween 20, for 1 hour at 37°C. Primary and secondary antibodies
were incubated for 1 hour or 45 minutes, respectively, at room
temperature. HRP-conjugated secondary antibodies were revealed
with ECL plus (Amersham-Pharmacia).
Spindle pole defects are induced by RanBP1
Overexpression of RanBP1 in asynchronously cycling
NIH/3T3 cell cultures was previously found to induce
multipolar spindles (Guarguaglini et al., 2000). Such
supernumerary poles might be originated through different
mechanisms involving abnormal centrosome duplication,
disruption of the centrosomal structure or centrosome mis-
segregation to daughter cells at cytokinesis. As a first step to
identify process(es) targeted by RanBP1 overexpression, we
transfected murine NIH/3T3 cell cultures with RanBP1
expression construct and analysed the pattern of centrosomes
in transfected cells. In a first set of experiments, asynchronous
cell cultures were transfected with pRanBP1-HA for 36 hours,
then fixed and processed for double IF to visualize centrosomal
markers in HA-expressing cells. We used antibodies against
centrin-2, a protein localized in the lumen of individual
centrioles; γ-tubulin, the major pericentriolar matrix (PCM)
protein required for MT nucleation; or GT335, an antibody
recognizing glutamylated tubulin, a typical modification of
centriole microtubules (Wolff et al., 1992). Many RanBP1-
overexpressing mitosis showed supernumerary spots reactive
to antibodies against centrosomal components: for example,
Fig. 1Aa shows a cell expressing high levels of HA-tagged
RanBP1 protein with four GT335-reactive spots at four distinct
locations instead of the two paired spots expected for a bipolar
spindle, strongly suggesting that the organization of centrioles
was affected. We next used combinations of antibodies to
detect pairs of centrosomal proteins in RanBP1-overexpressing
cells – GT335 and anti-centrin 2 (Fig. 1Ab) and/or GT335 and
anti-γ-tubulin (Fig. 1Ac). In this set of experiments, the spindle
was not stained but DAPI staining revealed a high frequency
of chromosome misalignment, consistent with the assembly of
abnormal spindles (compare, for example, DAPI images in
rows b and d). In abnormal mitoses, all analysed combinations
of markers simultaneously labeled supernumerary poles (Fig.
1Ab,c), indicating that induction of multipolar spindles in
RanBP1-overexpressing mitoses involves the entire centriolar
structure rather than an abnormal recruitment of specific
components to the poles. To further assess the involvement of
centrioles in the altered structure of spindle poles, parallel
RanBP1-transfected cultures were processed for anti-α-tubulin
and anti-centrin 2, or anti-α-tubulin and anti-γ-tubulin
combinations, to simultaneously reveal the spindle structure
together with centrioles or with nucleating centers. Although
we could not stain RanBP1 in these experiments, we recorded
a high frequency of multipolar spindles in RanBP1- but not in
vector-transfected cultures. Each pole in those abnormal
spindles contains material reactive to anti-centrin 2 (for
example, see Fig. 1Ad) or anti-γ-tubulin (data not shown)
antibodies. We next wanted to establish whether centrosomal
abnormalities induced by RanBP1 overexpression consistently
gave rise to multipolar spindles, or whether part of them are
irrelevant to mitotic spindle organization. NIH/3T3 cell
cultures transfected with RanBP1-HA were seeded on twin
slides within the same culture dish, fixed and processed in
parallel IF assays to quantify mitotic cells that displayed either
abnormal spindles (by α-tubulin staining) or abnormal
centrosomes (revealed by anti-centrin 2 antibody) among
RanBP1-overexpressing cells, recognized by HA staining. We
considered as abnormal all cells that displayed abnormalities
in either centrosome number or arrangement. Histograms in
Fig. 1B show that RanBP1 transfection in NIH/3T3 cells
yielded a fivefold increase in multipolar spindles compared
with vector-transfected cells, and a 4.7-times increase in
mitotic cells displaying abnormal centrin spots compared with
controls. Thus, the induction of multipolar spindles parallels
that of centrosomal abnormalities in RanBP1-overexpressing
Centrosomal abnormalities in RanBP1-overexpressing
cells are induced during mitosis
Normal cells undergo only one round of centrosome
duplication, during which each of the two centrioles
composing the centrosome duplicates in a semiconservative
manner. Each centrosome eventually segregates to a daughter
cell at cytokinesis and becomes ‘licensed’ to undergo a novel
round of duplication in the next cell cycle. Loss of the spindle
bipolarity is often related to abnormal centrosome duplication
(Lingle and Salisbury, 2000; Brinkley, 2001; Doxsey, 2001).
The influence of specific factors on centrosome duplication can
be assessed after prolonged treatment of CHO cells with
hydroxyurea, which blocks DNA synthesis but not centrosome
duplication (Balczon et al., 1995). Ectopic expression of the
cyclinA and cdk2 genes in this system induces centrosome
overduplication, whereas pRb and p16 inhibit it, and E2F-1
overexpression rescues the inhibition (Meraldi et al., 1999).
Instead, overexpression of RanBP1 showed no additional effect
Journal of Cell Science 116 (16)
Fig. 1. Supernumerary poles in RanBP1-
overexpressing cells contain centrosomal
components. (A) RanBP1-HA-transfected NIH/3T3
cells were double-stained with antibodies to
glutamylated tubulin (GT335) and anti-HA-antibody
to identify transfected cells (a); with antibodies
coupled to centrosomal markers (b,c); and with anti-
α-tubulin to label the spindle (d) and anti-centrin 2 to
visualize centrioles. Experiments were carried out
using all combinations of coupled antibodies to HA,
α-tubulin and centrosomal markers, and selected
examples are shown. DNA was counterstained with
DAPI (third column from the left). Merged pictures
are shown on the rightmost column. Scale bar, 10
µm. (B) Quantification of RanBP1-induced
abnormalities in spindle polarity (visualized by α-
tubulin staining) and in centrosomes (shown as either
>4 or abnormally separated centrin spots). Data from
three independent experiments were pooled and 100 mitoses per group were scored in each experiment. Histograms show the proportion of
cells with abnormalities in vector-transfected (gray) and RanBP1-transfected (white) cultures. P values calculated using the χ2test were highly
3403 RanBP1 and mitotic centrosome splitting
on centrosome duplication compared with hydroxyurea alone,
nor did it overcome the block of centrosome duplication
imposed by pRb (P. Meraldi, personal communication). Thus,
RanBP1 has no direct effect on centrosome duplication.
We next sought to restrict the cell cycle window in which
centrosomal abnormalities are generated. RanBP1-transfected
cell populations were synchronized and centrosomal
components were analysed during synchronous progression
through the cell cycle phases. We first brought NIH/3T3 cell
cultures to G0/G1 arrest by serum starvation, then stimulated
cell cycle re-entry with high serum. Cell samples were fixed 9
hours, 15 hours and 22 hours after cell cycle re-stimulation to
obtain G1, S and G2/M phase enrichment, respectively, as
indicated by FACS analysis (data not shown). Centrosomal
abnormalities, revealed by staining centrioles with anti-centrin-
2 antibody, were quantified as for Fig. 1B: cells with one or
two paired dots (corresponding to one or two centrosomes,
respectively) were taken as normal, whereas cells with more
than two pairs of dots or with scattered dots were assumed to
reflect overduplication and abnormal splitting of centrosomes,
respectively, and considered to be abnormal. Serum re-
stimulation of quiescent cells induces per se a high frequency
of centrosome splitting in vector-transfected cells (Table 1), in
line with previous reports (Sherline and Mascardo, 1982;
Schliwa et al., 1982; Schliwa et al., 1983). RanBP1
overexpression had no additional effect on serum-induced
centrosomal abnormalities in interphase; however a significant
increase was recorded in RanBP1-overexpressing mitotic cells
compared to control cultures (Table 1). To analyse S-to-M
progression more accurately, cells were arrested at the G1/S
transition with thymidine, then released in thymidine-free
medium and centrosomes were analysed in cells that were
allowed to progress towards mitosis. Again, no difference
between vector- and RanBP1-transfected cells were observed
in S or G2 interphase cells, whereas a high proportion of
centrosomal abnormalities was recorded in mitoses from
RanBP1-transfected compared to vector-transfected cultures
(Table 1). Thus, centrosomal abnormalities induced by
RanBP1 overexpression are specifically generated in mitosis.
Quantification of RanBP1-associated fluorescence in CCD
images of single cells transfected with expression construct, or
with vector alone, indicated that the RanBP1 signal increased
by over fourfold, on average, in overexpressing cells: most
transfected cells (~55%) displayed a three- to fivefold increase,
and ~30% showed a five- to sevenfold increase in RanBP1
signal intensity compared to control cells. To assess whether
the induction of centrosomal abnormalities did correlate with
the level of exogenous RanBP1, we examined 100 mitotic cells
from cultures transfected with pRanBP1-HA, then processed
with anti-HA/FITC to visualize transfected cells, and
GT335/rhodamine to visualize centrioles. Cells were analysed
for the presence or absence of centrosomal abnormalities on
the red channel, and the intensity of the FITC signal, quantified
on the green channel. Among RanBP1-transfected mitoses that
displayed a normal phenotype (n=61), the mean fluorescence
scored 1.9 (±0.5), taking the faintest signals in the lowest-
expressing cells as 1; ~60% of them displayed relative
intensities below 2, and the remaining 40% fell between 2
and 3. Among RanBP1-transfected mitoses that developed
centrosomal abnormalities (n=39), the mean relative
fluorescence rose to 2.9 (±1.2); a minority (~23%) of these
abnormal mitoses displayed a fluorescence intensity below 2,
comparable to normal mitoses; all other cells had relative
intensities above 2, with a discrete cell population (~15%)
showing more than a fourfold increase in RanBP1 signal
intensity (Table 2). Thus, RanBP1-transfected cells that
develop mitotic centrosomal abnormalities tend to express the
highest levels of exogenous RanBP1.
RanBP1 overexpression disrupts cohesion of sister
centrioles in mitotic diplosomes
To resolve accurately the type of centrosomal abnormality
induced by RanBP1 overexpression, we made use of L929-
derived cell cultures stably transfected with centrin-1/GFP
chimera (Piel et al., 2000). The incorporation of GFP-
chimerized centrin in individual centrioles allows a higher
resolution of centrosomes than indirect immunofluorescence
techniques. This cell model therefore provides a particularly
useful tool to analyse the effects of RanBP1.
We initially characterized centrin-1/GFP L929 cells from
non-synchronized cultures and noticed that they spontaneously
develop a somewhat higher level of centrosomal abnormalities
(26.5% in 170 scored mitoses) compared with NIH/3T3
fibroblasts (10.7% in 400 mitoses). Of all centrosomal
abnormalities detected among L929 mitotic cells, nearly half
(12.3% of all mitoses) were represented by supernumerary,
structurally integral centrosomes (arrangement II in Fig. 2B).
The remaining abnormal mitoses showed diplosome splitting,
Table 1. RanBP1-dependent centrosomal abnormalities
during cell cycle progression
A, serum-starved (G0/G1) and restimulated cells harvested 9, 15 and 22 h
after cell cycle entry.
B, thymidine-arrested (G1/S) and released cells harvested 6, 7 and 8 h after
S phase resumption.
P values were calculated using the χ2test; ns, not significant.
Table 2. RanBP1 levels in transfected mitoses with normal
or abnormal phenotypes
Normal mitoses Centrosomal abnormalities
Total61 10039 100
aFluorescence intensities were measured on CCD images (see Materials
and Methods) and are expressed relative to the faintest intensity in the lowest-
expressing cell, taken as=1.
P values were calculated using the χ2test; ns, not significant.
either associated with a normal number of centrioles (i.e. four
centrin dots, arrangement III in Fig. 2B) or concomitant with
supernumerary centrosomes (i.e. more than four centrin dots,
arrangement IV in Fig. 2B). For comparison, the corresponding
phenotypes among NIH/3T3
(centrosomes overduplication) and 6.9% (diplosome splitting).
We next assessed the effect of transfected RanBP1-RFP
chimeras in L929-derived cell cultures stably expressing
centrin-1/GFP. Cells that reached mitosis after thymidine
synchronization and release were collected by the ‘shake off’
method, then immediately re-seeded on microscope slides, and
mitotic cells with supernumerary integral centrosomes or with
split diplosomes were examined by analysing the arrangement
of centrin-1/GFP centrioles (see scheme in Fig. 2B). In normal
mitoses, chromosomes were correctly aligned and centrioles
were arranged in typical diplosomes at each pole (Fig. 2A,
left corner in upper row, see magnification in a). RanBP1
overexpression did not significantly affect centrosome
duplication (Fig. 2B), consistent with results obtained in
NIH/3T3 cell lines (see above), but specifically induced sister
centrioles from single diplosomes to move apart from one
mitoses scored 3.8%
another (Fig. 2A, magnification in b and c). As shown in Fig.
2B, ~45% of RanBP1-overexpressing mitotic cells showed
split diplosome, compared with 20% in vector-transfected cells
(P<0.001). We also analysed cultures that remained adherent
during shaking off and were enriched in G2-phase cells:
RanBP1 overexpression in these cultures failed to increase the
frequency of abnormal centrosome numbers or splitting (data
not shown), as previously observed in NIH/3T3 cultures,
thereby confirming that RanBP1 specifically induces
diplosome splitting during mitosis.
Splitting of centrioles during mitosis was previously
reported to occur under induction of mitotic arrest (Sluder
and Rieder, 1985; Gallant and Nigg, 1992). RanBP1
overexpression actually causes some increase in the mitotic
index, as previously observed (Guarguaglini et al., 2000).
However, the extent of the induced delay in our experiments
was in the upper limit of the physiological range or just above
it (Table 3), different from that induced by MT drugs or failure
of cyclin-B degradation. Video recordings of cells transfected
with pRFP vector or pRanBP1-RFP depicted no dramatic delay
in the timing from prophase/prometaphase – indicated by
Journal of Cell Science 116 (16)
Fig. 2. RanBP1 overexpression induces centriole
splitting in mitosis. (A) L929 cell cultures stably
expressing a centrin 1-GFP chimera (Piel et al.,
2000) were transfected with RanBP1-RFP,
synchronized as described in the text, and mitotic
cells recovered by ‘shake-off’ were analysed. In
the upper row, the non-transfected cell (upper left
corner, negative for RFP emission) shows
correctly aligned chromosomes (DNA panel) and
centriole pairs in each centrosome, as shown in
the magnified insert (a). In RanBP1-transfected
cells (positive for RFP emission), single split
centrioles are visible: two examples are shown,
magnified in inserts b and c. Scale bar, 10 µm.
(B) Quantification of centrosome defects induced
by RanBP1 overexpression. Possible distributions
of centrioles in mitosis are: I, normal
arrangement; II, overduplicated centrosomes; III,
split centrioles; IV, overduplicated and split
centrosomes. Only tetrapolar spindles are
represented, for simplicity. Histograms in the left
panel show the frequency of centrosome
overduplication (gray), calculated by grouping
patterns II and IV (i.e. all cells with more than
four centrioles) as abnormal. The same samples
were re-analysed for the frequency of centriole
splitting (histograms in the right panel),
calculated by grouping patterns III and IV as
abnormal (i.e. all cells showing single centrioles,
regardless of total centriole number). 200 mitotic
cells from vector- and RanBP1-RFP-transfected
cultures were scored. The asterisks mark a highly
significant difference (P<0.001).
3405 RanBP1 and mitotic centrosome splitting
rounding-up of the cells – to anaphase in vivo: all video-
recorded control cells reached anaphase within 40 minutes
from mitosis onset, and most of them took 20-30 minutes.
RanBP1-transfected cells underwent some delay, with most of
them taking 30-40 minutes to execute the same stages. Thus,
RanBP1-dependent delay in early mitosis is well below that
induced by MT drugs or non-degradable cyclin B, which is in
the order of hours. Furthermore, progression through mitotic
substages was analysed in transfected cultures after IF to α-
tubulin: this revealed a higher proportion of ana/telophases
among RanBP1-overexpressing mitoses compared with
controls. Thus, the induction of mitotic delay by RanBP1
overexpression is essentially caused by prolonged duration of
ana/telophase stages, possibly reflecting hindrance in M exit
(Battistoni et al., 1997; Guarguaglini et al., 2000), whereas
earlier mitotic stages are not significantly affected. RanBP1
induction of centriole splitting is instead already visible in
prometaphase. Thus, RanBP1-dependent centriole splitting is
a specific phenomenon, not attributable to prolonged duration
We next wished to ascertain whether single split centrioles
were able to assemble functional spindle poles. Centrin-1/GFP
expressing L929 cultures were transfected with the RanBP1-
RFP chimera and spindle MTs were labeled with anti-α-tubulin
antibody, revealed with an AMCA-conjugated secondary
antibody. As shown in Fig. 3, microtubule arrays nucleating
from single centrioles are focused to form separate poles, hence
forming a multipolar spindle.
Diplosome splitting by RanBP1 requires integrity of
Cohesion between parental centrioles requires MT integrity
(Jean et al., 1999) and MT disruption by nocodazole favors
the separation of parental centrosomes (Mayor et al., 2000).
We wondered whether MTs are implicated in RanBP1-
induced splitting between centrioles. NIH/3T3 cultures
were transfected with RanBP1-HA or RanBP1-GFP and
subsequently synchronized in G2/M phases by thymidine
block and release as described above. Cells were then exposed
to nocodazole (NOC) and either collected after 4 hours, with
most cells arrested in prometaphase without spindle MTs, or
allowed to resume mitosis by removing NOC and fixed 45
minutes after release. Both FACS analysis and microscope
counting of mitotic cells (data not shown) were used to monitor
synchronization. Mitotic centrosomes were analysed using
either GT335 or anti-centrin-2 antibodies. In cultures exposed
to NOC, the overall centrosomal organization was altered in
interphase cells, with centrosomes being typically distanced
and displaced from the juxtanuclear region (data not shown),
consistent with the established role of MTs in anchoring
centrosomal structures to each other and to their subcellular
site (Jean et al., 1999). Fig. 4 shows the results obtained in
RanBP1-overexpressing cultures. In cells that underwent
mitosis after release from thymidine arrest, RanBP1
overexpression caused a highly significant increase in
diplosome splitting compared with cultures transfected with
vector. When NOC was added to G2 cultures to inhibit MT
polymerization, the effects of RanBP1 overexpression were
prevented, and the frequency of mitoses with split centrioles
was comparable in RanBP1-overexpressing and in vector-
transfected cultures. Thus, NOC per se does not affect the
organization of sister centrioles within diplosomes, in contrast
to its ability to induce separation of parental centrioles, yet
counteracts the disruptive effect caused by RanBP1 excess,
indicating that MT integrity is required for induction of
diplosome splitting. The specificity of this requirement was
further demonstrated by removing NOC from the culture
medium and allowing the cells to reform MTs in vivo: upon
resumption of mitosis, RanBP1-overexpressing mitoses again
underwent diplosome splitting (Fig. 4).
We previously reported that multipolar spindles are similarly
induced by wild-type RanBP1 and by the RanBP1L186A/V188A
mutant, which carries inactivating mutations in the nuclear
export signal (NES) and hence is retained in nuclei throughout
interphase (Richards et al., 1996; Guarguaglini et al., 2000). If
multipolar spindles are generated through loss of diplosome
cohesion as a truly mitotic phenomenon, then similar effects to
Fig. 3. Split centrioles organize functional spindle
poles. (a) A RanBP1-RFP-transfected mitosis from
centrin 1-GFP stably expressing L929 cell cultures.
Centrin-1/GFP allows the visualization of centrioles
(b); the spindle is stained with anti-α-tubulin, revealed
with an AMCA-conjugated secondary antibody (c).
Merging of b and c produces d, which depicts single
split centrioles (green) at each spindle pole (AMCA-
stained MTs, in blue). Scale bar, 10 µm.
Table 3. Effect of RanBP1 overexpression on mitotic progression
Time from prometa to
anaphase (in vivo)b
% Mitoses in
aSimilar results were obtained with pRanBP1 untagged, pRanBP1-RFP and pRanBP1-HA.
bThe timing of early mitosis was calculated from video-recorded images taken with 10-min intervals.
those reported thus far are expected in cells overexpressing
the RanBP1L186A/V188Amutant, regardless of its abnormal
localization during interphase. Indeed, the NES-defective
RanBP1 mutant induced a highly significant increase in mitotic
diplosome splitting in NIH/3T3 cultures released from
thymidine arrest, similar to wild-type RanBP1 (Fig. 4). Parallel
analysis of L929 centrin-1/GFP cultures enabled us to establish
that the type of mitotic diplosome splitting induced by mutant
and wild-type RanBP1 was indistinguishable (data not shown).
The RanBP1L186A/V188Amutant failed instead to induce
diplosome splitting in nocodazole-exposed cells, similar to
wild-type RanBP1 (Fig. 4). Thus, the comparable ability of
export-defective and wild-type RanBP1 to disrupt centriole
cohesion in a MT-dependent manner further confirms that
diplosome splitting takes place after NEB.
Diplosome splitting by RanBP1 requires Eg5 activity
The Eg5 kinesin controls the establishment of the spindle
bipolarity by causing parental centrosome separation at the
onset of mitosis (Walczak et al., 1998) and Ran can modulate
Eg5 mobility on MTs (Wilde et al., 2001). Thus, we wondered
whether Eg5 activity influenced RanBP1-induced splitting
within mitotic diplosomes. Inhibition of Eg5 activity by
monastrol (MA) prevents centrosome separation, yielding
mitotic cells that typically arrest with monoastral spindles
(Kapoor et al., 2000). In our experiments, RanBP1- or vector-
transfected NIH/3T3 cell cultures were synchronized by
thymidine block and release as above, and, when in G2 as
judged by FACS analysis, MA was added for 4 hours. Cells
were then fixed and centrosomes were analysed. By γ-tubulin
staining of centrosomes and DAPI staining of chromosomes,
monoastral mitoses with unseparated centrosomes were
indistinguishable in RanBP1- and vector-transfected cells (Fig.
5A). Centrosome structure was more closely inspected using
anti-centrin-2 antibody. Although all mitoses had a monoastral
spindle, different arrangements could be appreciated at the
centrosome level: monoastral mitoses in which two sets of
paired centrioles were visible at the center of the spindle were
defined ‘normal’ (Fig. 5Ba); mitoses showing more than two
paired centrin spots (Fig. 5Bb) were assumed to reflect
overduplication, whereas clearly distanced centrioles in at least
one diplosome (Fig. 5Bc) were recorded as abnormal splitting.
By these criteria, centriole splitting occurred with similar
frequency in vector-transfected and RanBP1-overexpressing
monoastral mitoses (Fig. 5C). Eg5 inhibition by MA is
reversible and so cells released in MA-free medium readily
re-establish bipolarity. Under these conditions, diplosome
splitting was again appreciated in RanBP1-overexpressing
cells that progressed through mitosis 30 minutes after MA
removal (Fig. 5C). Thus, Eg5 function is required for induction
of diplosome splitting by overexpressed RanBP1.
Centrosomal localization of RanBP1
Because RanBP1 overexpression affects centriole cohesion, we
re-examined its localization relative to centrosomes. In
NIH/3T3 interphase cells fixed with paraformaldehyde (PFA),
RanBP1 is almost completely cytoplasmic; some enrichment
at the spindle can be appreciated in mitotic cells (Guarguaglini
et al., 2000). If a fraction of RanBP1 localizes at centrosomes,
such a fraction might be masked by the abundant soluble pool
and difficult to resolve. Indeed, partial permeabilization of
NIH/3T3 cells with Triton X-100 prior to methanol or PFA
fixation revealed a fraction of insoluble RanBP1 protein at the
centrosome, revealed by γ-tubulin, in both interphase (Fig.
6Aa) and mitotic cells (Fig. 6Ab,c). A small centrosomal
fraction of RanBP1 was also visualized in mouse L929 (Fig.
6Ad) and human HeLa cells (Fig. 6Ae) using independent
antibodies. The co-localization of RanBP1 signals with γ-
tubulin was confirmed by scanning NIH/3T3 cell spreads under
confocal microscopy (Fig. 6B).
To extend these results, we analysed preparations of purified
centrosomes isolated from the human lymphoblastic cell line
KE37. RanBP1 was retained on isolated centrosomes analysed
by IF (Fig. 6C) and showed a very similar labeling pattern to
that revealed using the CTR453 antibody, which specifically
recognizes the AKAP450 centrosomal matrix protein (Bailly
et al., 1989). Western immunoblotting was then used to assess
the strength of the interaction of RanBP1 with the KE37-
derived centrosomal fraction (Fig. 6D). Purified centrosome
preparations were treated with solubilizing detergents of
increasing strength, and the soluble (supernatant) and insoluble
(pellet) fractions were analysed with anti-RanBP1 antibody. As
shown in Fig. 6D, the association of a RanBP1 fraction with
centrosomes was resistant to strong solubilizing conditions:
centrosomal RanBP1 was not solubilized by NP40 alone (1D
buffer), nor by NP40 combined with DOC (2D buffer), nor
with DOC and SDS simultaneously (3D buffer). Treatment of
centrosomes with 8 M urea eventually solubilized centrosomal
Journal of Cell Science 116 (16)
Fig. 4. Centriole splitting requires
microtubule integrity. Centrosomal
abnormalities [i.e. overduplication (hatched
columns) and splitting (black columns)] were
recorded in cultures transfected with vector,
wild-type RanBP1 or RanBP1L186A/V188A
mutant (indicated as NES) during mitosis
following thymidine release, or after
treatment with nocodazole (NOC), or
released after NOC arrest. For each
condition, three to five experiments were carried out with wild-type RanBP1 and at least two with the RanBP1L186A/V188Amutant. Data were
pooled and analysed using the χ2test. *, P<0.05; **, P<0.001.
pX RanBP1 NESpXRanBP1 NES pXRanBP1 NES
% mitotic centrosomal
3407 RanBP1 and mitotic centrosome splitting
RanBP1. Thus, a RanBP1 fraction is actually involved in a
stable interaction with centrosomes.
To ascertain whether exogenously expressed RanBP1 also
reached centrosomes, IF experiments were performed in
cultures transfected with pRanBP1-HA. After solubilization
and fixation, exogenous RanBP1 was revealed by anti-HA
antibody and centrosomes were stained
for γ-tubulin. This showed that anti-HA
staining was concentrated
pericentrosomal region (Fig. 6E).
Because RanBP1 excess alters cohesion
within centrosomes in the presence of
intact MTs, we asked whether localization
of RanBP1 at the centrosome is influenced
by the status of mitotic MTs. When
thymidine-released cultures were exposed to NOC, under
conditions that prevent both MT polymerization and
RanBP1-dependent centriole splitting, a fraction of
RanBP1 was still detected at the centrosome (Fig. 7a). A
comparable localization was seen in cells exposed to Taxol
(Fig. 7b). Therefore, the association of a RanBP1 fraction
with centrosomes is independent of MT integrity or
dynamics. This result, together with the strength of the
association depicted in Fig. 6D, suggests that a fraction of
RanBP1 associates constitutively with centrosomes.
The formation of multipolar spindles predisposes mitotic
cells to undergo chromosome mis-segregation. Frequent causes
of multipolar spindle assembly include errors in centrosome
duplication or segregation to daughter cells, which can lead to
genomic imbalance and favor cell transformation and tumor
progression (Lingle and Salisbury, 2000; Brinkley, 2001;
Doxsey, 2001). Here, we have followed up previous indications
Fig. 5. Centriole splitting requires Eg5 activity.
(A) Examples of MA-arrested mitoses from
NIH/3T3 cultures transfected with GFP vector
or with RanBP1-GFP and stained with antibody
to γ-tubulin (right): no obvious difference in the
centrosomal pattern is observed. Chromosomes
are stained with DAPI and merged pictures are
shown on the left. Scale bar, 10 µm.
(B) Patterns of centrioles in monoastral mitoses
revealed by anti-centrin-2 antibody: (a) normal
arrangement with two pairs of centrioles;
(b) supernumerary centriole pairs; (c) split
centrioles. The rightmost column shows a
magnification of the centrin 2 panels. Scale bar,
10 µm. (C) Frequency of centrosomal
arrangements in NIH/3T3 MA-arrested and
MA-released mitoses transfected with vector or
RanBP1-GFP. Pooled data from four
experiments were analysed using the χ2test. *,
multipolar spindles, and have sought to
pinpoint the underlying defect. This is
relevant in view of the fact that the
RanBP1 gene is a regulatory target of E2F-
and retinoblastoma-related factors (Di
Matteo et al., 1995; Di Fiore et al., 1999;
Ishida et al., 2001), and hence can be
expected to be deregulated in tumors in
which this pathway is disrupted. Actually,
both RanBP1 and RCC1 were recently
identified as downregulated target genes of
a novel anticancer drug (Damm et al.,
2001), suggesting that either or both of
these genes can actually be deregulated in
Ran is an abundant GTPase (107
molecules cell–1in HeLa cells) (Bischoff
and Ponstingl, 1991), and is estimated to
be present in a 25-fold excess over
endogenous RCC1 and fivefold excess
over endogenous RanBP1 (Bischoff et al.,
1995). In our transfection experiments, we
RanBP1 overexpression yields
Journal of Cell Science 116 (16)
Fig. 6. A fraction of RanBP1 localizes at the
centrosome. (A) Centrosomal RanBP1 in
interphase (a), metaphase (b) and anaphase (c)
NIH/3T3 cells. Examples of L929 (d) and
HeLa (e) mitotic cells are also shown.
Endogenous RanBP1 (second column) and γ-
tubulin (third column) were revealed with
FITC- and rhodamine-conjugated secondary
antibodies, respectively. DNA was
counterstained with DAPI (first column on the
left). Signals are merged in the rightmost
column. Scale bar, 10 µm. (B) Confocal
signals for RanBP1 (FITC) (left) and γ-tubulin
(rhodamine, middle) in a typical NIH/3T3
metaphase. Merged images are shown on the
right. 15 focal planes of 2.28 µm thickness
were scanned. (C) Anti-RanBP1 antibody
(bottom) labels isolated KE37 centrosomes,
stained by CTR453 (top). Scale bar, 50 µM.
(D) RanBP1 is tightly associated with the
centrosome fraction. Isolated centrosomes
were extracted with buffers of increasing
strength and analysed by western
immunoblotting with the indicated antibodies.
Abbreviations: p, pellet containing
centrosome-associated proteins; s, supernatant
containing solubilized proteins. The interaction
of RanBP1 with centrosomes (bottom) is more
resistant to detergents than that of γ-tubulin, a
major PCM-recruited component (top).
(E) Overexpressed RanBP1 localizes at spindle
poles. An example of NIH/3T3 metaphase is
shown. Anti-HA antibody, directed against
exogenous RanBP1, is revealed with a
rhodamine-conjugated secondary antibody.
Centrosomes are stained with anti-γ-tubulin
antibody revealed with an AMCA-conjugated
secondary antibody. The merged image is
shown in the right panel. Scale bar, 10 µm.
3409RanBP1 and mitotic centrosome splitting
recorded an average fourfold increase in RanBP1 levels;
furthermore, RanBP1-transfected cells that displayed mitotic
centrosomal abnormalities typically showed higher than
average levels of overexpression. That range of increase is
expected to produce a significant shift in the balance of
nucleotide hydrolysis and exchange on Ran. We previously
reported that RanBP1 overexpression induces cell cycle
abnormalities (Battistoni et al., 1997; Guarguaglini et al., 2000)
comparable to those observed in the presence of Ran mutants
(Ren et al., 1993; Ren et al., 1994; Moore et al., 2002),
supporting the idea that RanBP1 acts by altering the Ran
network. In addition, we have now sought to quantify the
intracellular RanGTP levels using an antibody (AR12, a kind
gift from I. Macara) that preferentially – although not
exclusively – recognizes the GTP-bound conformation of Ran
(Richards et al., 1995). Although these experiments do not
allow us to draw a precise quantitative estimate, they do
indicate that RanGTP levels are lowered in RanBP1-
overexpressing compared with normal cells (data not shown).
Induction of multipolar spindles by RanBP1 excess reflects
the aberrant splitting of single centrioles within diplosomes in
mitosis. None of duplication of centrosomes, recruitment of γ-
tubulin or glutamylation of centriole MTs are affected instead.
Furthermore, no defects were recorded in focusing of MT
arrays to the poles. Split centrioles retain their functional
integrity and can organize polarized MT arrays, thereby giving
rise to spindles with multiple poles. This is a novel finding and
begins to identify aspects of centrosome organization and
function that are influenced by members of the Ran network.
Cohesion and dynamics of centrosomes are highly
regulated processes. After duplication, centrosomes remain
tethered together throughout most of interphase, then separate
in late G2 and eventually migrate to form the spindle poles.
MTs contribute to the link between centrosomes (Jean et al.,
1999). Cohesion in G2 and separation in mitosis are also
regulated by a network of specific factors (Meraldi and Nigg,
2001), including the centrosomal C-Nap1 protein (Mayor et
al., 2000), its upstream kinase Nek2 (Meraldi and Nigg,
2001) and the Inh2 regulator of Nek2 (Eto et al., 2002).
Deregulated activity of these factors induces unscheduled
centrosome separation but the integrity within centrosomes is
not affected and so neither spindle assembly nor the mitotic
division are necessarily perturbed (Mayor et al., 2002).
RanBP1 overexpression influences neither the timing nor the
extent of parental centrosome separation in interphase, but
selectively perturbs cohesion of centrioles within diplosomes
telophase, accompanied by
repositioning of the mother centriole to the mid-body in
preparation of cytokinesis (Piel et al., 2001). In early
interphase, split centrioles act as duplication templates. Under
abnormal circumstances, however, centrioles can split during
mitosis, as observed during (for example) mitotic arrest
induced by non-degradable cyclin B (Gallant and Nigg, 1992).
Indeed, induction of mitotic delay by mercaptoethanol or
colcemid was used as an experimental tool to study the
functional relationship between centrioles and spindle poles
(Sluder and Rieder, 1985). In RanBP1-overexpressing cultures,
we recorded some increase in the mitotic index, but this was
essentially due to prolonged of telophase. The timing of early
mitotic progression was instead not dramatically perturbed in
RanBP1-overexpressing cells, whereas centriole splitting could
already be detected in prometaphase, as soon as the nuclear
envelope disappeared – a stage that was not prolonged by
RanBP1 overexpression. These observations support the
conclusion that RanBP1-induced centrosomal abnormalities
are not a consequence of abnormally prolonged mitosis.
The RanBP1L186A/V188Aconstruct, which has a different
localization from wild-type RanBP1 throughout interphase,
has similar disruptive effects than wild-type on mitotic
centrosomes. This is paralleled by the ability of this mutant to
induce multipolar spindles as effectively as wild-type RanBP1
(Guarguaglini et al., 2000). These data are consistent with the
view that overexpressed RanBP1 interferes with crucial
factor(s) implicated in centrosome organization specifically
during mitosis. Such factor(s) might be activated, and/or be
capable of establishing crucial interactions at the centrosomal
level, specifically after NEB in a manner that is similarly
sensitive to NES-defective and wild-type RanBP1. The mitotic
nature of the splitting phenomenon induced by RanBP1 excess
was further evidenced in cells that resumed mitotic progression
after NOC-induced arrest. NOC prevented the disruptive effect
of RanBP1 excess, yet centrosome splitting was again
appreciated after as little as 45 minutes after NOC removal and
resumption of MT reconstitution in vivo. This experiment
further strengthens the conclusion that centriole cohesion is
sensitive to RanBP1 levels during mitosis and, furthermore,
implicates MTs. Induction of diplosome splitting by high
RanBP1 is also dependent on Eg5 activity, suggesting that
Fig. 7. RanBP1 retains its centrosomal localization
after exposure to either nocodazole (a) or Taxol (b).
Endogenous RanBP1 (second column) and γ-tubulin
(third column) were revealed with FITC- and
rhodamine-conjugated secondary antibodies,
respectively. DNA was counterstained with DAPI
(first column on the left). The rightmost column
shows the merged signals of RanBP1, γ-tubulin and
DAPI staining. Scale bar, 10 µm.
either centrosome separation is required or that some timely
regulated interaction that is physiologically dependent upon
Eg5 is required. It is noteworthy that RanBP1 was detected at
centrosomes even in NOC-exposed cells. This observation and
the ability of a RanBP1 fraction to localize at centrosomes
already in interphase and to interact with centrosomes in a
stable, detergent-resistant manner, converge to suggest that
a small RanBP1 fraction constitutively associates with
centrosomes. It has recently been found that a fraction of Ran
also localizes at centrosomes, in the presence or absence of
NOC (Keryer et al., 2003). Thus, the suppressive effect of NOC
is not due to failure of RanBP1 or Ran to localize at
centrosomes. Rather, MTs themselves or motor proteins might
play a role in cohesion within diplosomes in a manner that is
sensitive to high RanBP1 levels. We previously found that
inactivation of mitotic RanBP1 by antibody microinjection
impairs dynamics of the spindle MTs. RanBP1 excess might
influence MT dynamics at spindle poles, and altered dynamics
might in turn favor the aberrant separation of mother and
daughter centrioles. An alternative – but not necessarily
mutually exclusive – possibility is that one or more factor(s)
that regulate the organization and/or the intrinsic dynamic
features of mitotic centrosomes is transported to spindle poles
in a MT-dependent manner after NEB and, once there, is
sensitive to elevated levels of RanBP1. In the presence of NOC,
the hypothetical protein(s) would not be transported to poles
and so would not be in a position to modulate the behavior of
centrioles, regardless of RanBP1 levels. Based on these
observations, the mitotic role of Ran network components
might be critically dependent on their ability to associate with
specific mitotic structures. In mitosis, Ran members might
reorganize in ‘local factories’ at specific locations and act on
local downstream targets in the mitotic apparatus. The recent
observation that spindle pole defects and chromosome
misalignment are caused by a RCC1 mutant that mislocalizes
to the cytoplasm but not by wild-type RCC1 (Moore et al.,
2002) is consistent with this view.
Interestingly, while this work was in progress, disruption of
spindle pole organization was observed in mammalian cells
under interfering RNA-mediated inactivation of an important
Ran target, TPX2 (Garrett et al., 2002). Although, in other
studies, the major outcome of TPX2 inactivation was failure of
MT connections between spindle poles, probably because of
differences in the experimental conditions that yielded partial
inactivation of TPX2 (Gruss et al., 2002), in the study by
Garrett et al. (Garrett et al., 2002) multipolar spindles formed
as a consequence of spindle pole fragmentation. Remarkably,
these abnormalities are MT and Eg5 dependent, similar to
those reported here under RanBP1 excess. The authors suggest
that multipolar spindles induced in their conditions might
reflect an imbalance between TPX2-dependent structural
support and motor-driven force: when TPX2 is inactivated, the
force would be exerted freely and cause spindle pole
disruption. By analogy of reasoning, it is tempting to speculate
that defective RanGTP formation caused by RanBP1 excess
causes insufficient release of factor(s) that provide structural
support to sister centrioles during spindle assembly.
In summary, a fraction of the RanBP1 protein is present at
centrosomes throughout the cell cycle, where it can interact
with Ran. At this location, RanBP1 can act on factor(s) that
reach the centrosomes after NEB to contribute to the
organization of mitotic centrosomes. The presence of excess
RanBP1 favors the aberrant separation of individual centrioles
in mitosis, giving rise to multipolar spindles. Further
understanding the mechanisms through which Ran network
components act locally in mitosis and control downstream
targets in the assembly of mitotic structures will be a major
field to disentangle in the near future.
We are indebted with M. Bornens, in whose laboratory some of the
experiments reported here were performed. We are grateful to P.
Denoulet and I. Macara for the gift of antibodies, to C. Celati for
providing isolated centrosomes, and to E. Marchetti for help with
confocal images. We also thank P. Meraldi for communicating
unpublished results and G. Guarguaglini and M. Casenghi for helpful
comments on this manuscript. This work was supported by grant
ARC599 to AMT and by grants from Consiglio Nazionale delle
Ricerche (CNR), Associazione Italiana per la Ricerca sul Cancro
(AIRC) and Agenzia Spaziale Italiana (ASI) to PL. BDF and MC were
supported by MIUR/CNR Doctoral fellowships, and RM was
supported by a CNR research fellowship.
Bailly, E., Dorée, M., Nurse, P. and Bornens, M. (1989). p34cdc2is located
in both nucleus and cytoplasm; part is centrosomally associated at G2/M
and enters vesicles at anaphase. EMBO J. 8, 3985-3995.
Balczon, R., Bao, L., Zimmer, W. E., Brown, K., Zinkowski, R. P. and
Brinkley, B. R. (1995). Dissociation of centrosome replication events from
cycles of DNA synthesis and mitotic division in hydroxyurea-arrested
Chinese hamster ovary cells. J. Cell Biol. 130, 105-115.
Battistoni, A., Guarguaglini, G., Degrassi, F., Pittoggi, C., Palena, A., di
Matteo, G., Pisano, C., Cundari, E. and Lavia, P. (1997). Deregulated
expression of the RanBP1 gene alters cell cycle progression in murine
fibroblasts. J. Cell Sci. 110, 2345-2357.
Bischoff, F. R. and Ponstigl, H. (1991). Mitotic regulator protein RCC1 is
complexed with a nuclear Ras-related polypeptide. Proc. Natl. Acad. Sci.
USA 88, 10830-10834.
Bischoff, F. R., Krebber, H., Smirnova, E., Dong, W. and Ponstingl, H.
(1995). Co-activation of RanGTPase and inhibition of GTP dissociation by
RanGTP binding protein RanBP1. EMBO J. 14, 705-715.
Brinkley, B. R. (2001). Managing the centrosome numbers game: from chaos
to stability in cancer cell division. Trends Cell Biol. 11, 18-21.
Carazo-Salas, R. E., Guarguaglini, G., Gruss, O. J., Segref, A., Karsenti, E.
and Mattaj, I. W. (1999). Generation of GTP-bound Ran by RCC1 is required
for chromatin-induced mitotic spindle formation. Nature 400, 178-181.
Clarke, P. R. and Zhang, C. (2001). Ran GTPase: a master regulator of
nuclear structure and function during the eukaryotic cell division cycle?
Trends Cell Biol. 11, 366-371.
Damm, K., Hemmann, U., Garin-Chesa, P., Hauel, N., Kauffmann, I.,
Priepke, H., Niestroj, C., Daiber, C., Enenkel, B., Guilliard, B. et al.
(2001). A highly selective telomerase inhibitor limiting human cancer cell
proliferation. EMBO J. 20, 6958-6968.
Dasso, M. (2002). The Ran GTPase: theme and variations. Curr. Biol. 12,
Di Fiore, B., Guarguaglini, G., Palena, A., Kerkhoven, R. M., Bernards,
R. and Lavia, P. (1999). Two E2F sites control growth-regulated and cell
cycle-regulated transcription of the Htf9-a/RanBP1
functionally distinct mechanisms. J. Biol. Chem. 274, 10339-10348.
Di Matteo, G., Fuschi, P., Zerfass, K., Moretti, S., Ricordy, R., Cenciarelli,
C., Tripodi, M., Jansen-Durr, P. and Lavia, P. (1995). Transcriptional
control of the Htf9-a/RanBP-1 gene during the cell cycle. Cell Growth
Differ. 6, 1213-1224.
Doxsey, S. J. (2001). Centrosomes as command centres for cellular control.
Nat. Cell Biol. 3, E105-E108
Eto, M., Elliott, E., Prickett, T. D. and Brautigan, D. L. (2002). Inhibitor-
2 regulates protein phosphatase-1 complexed with NimA-related kinase to
induce centrosome separation. J. Biol. Chem. 277, 44013-44020
Gallant, P. and Nigg, E. A. (1992). Cyclin B2 undergoes cell cycle-dependent
nuclear translocation and, when expressed as a non-destructible mutant,
causes mitotic arrest in HeLa cells. J. Cell Biol. 117, 213-224.
Garrett, S., Auer, K., Compton, D. A. and Kapoor, T. M. (2002). hTPX2 is
Journal of Cell Science 116 (16)
3411 RanBP1 and mitotic centrosome splitting
required for normal spindle morphology and centrosome integrity during
vertebrate cell division. Curr. Biol. 12, 2055-2059
Gruss, O. J., Carazo-Salas, R. E., Schatz, C. A., Guarguaglini, G., Kast,
J., Wilm, M., le Bot, N., Vernos, I., Karsenti, E. and Mattaj, I. W. (2001).
Ran induces spindle assembly by reversing the inhibitory effect of importin
alpha on TPX2 activity. Cell 104, 83-93.
Gruss, O. J., Wittmann, M., Yokoyama, Hm, Pepperkok, R., Kufer, T.,
Sillje, H., Karsenti, E., Mattaj, I. W. and Vernos, I. (2002). Chromosome-
induced microtubule assembly mediated by TPX2 is required for spindle
formation in HeLa cells. Nat. Cell Biol. 4, 871-879.
Guarguaglini, G., Renzi, L., D’Ottavio, F., di Fiore, B., Casenghi, M.,
Cundari, E. and Lavia, P. (2000). Regulated Ran-binding protein 1 activity
is required for organization and function of the mitotic spindle in
mammalian cells in vivo. Cell Growth Differ. 11, 455-465.
Hetzer, M., Gruss, O. J. and Mattaj, I. W. (2002). The Ran GTPase as a
marker of chromosome position in spindle formation and nuclear envelope
assembly. Nat. Cell Biol. 4, E177-E184.
Ishida, S., Huang, E., Zuzan, H., Spang, R., Leone, G., West, M. and
Nevins, J. R. (2001). Role for E2F in control of both DNA replication and
mitotic functions as revealed from DNA microarray analysis. Mol. Cell.
Biol. 21, 4684-4699.
Jean, C., Tollon, Y., Raynaud-Messina, B. and Wright, M. (1999). The
mammalian interphase centrosome: two independent units maintained
together by the dynamics of the microtubule cytoskeleton. Eur. J. Cell Biol.
Kalab, P., Pu, R. T. and Dasso, M. (1999). The Ran GTPase regulates mitotic
spindle assembly. Curr. Biol. 9, 481-484.
Kalab, P., Weis, K. and Heald. R. (2002). Visualization of a RanGTP gradient
in interphase and mitotic Xenopus egg extracts. Science 295, 2452-2456.
Kapoor, T. M., Mayer, T. U., Coughlin, M. L. and Mitchison, T. J. (2000).
Probing spindle assembly mechanisms with monastrol, a small molecule
inhibitor of the mitotic kinesin, Eg5. J. Cell Biol. 150, 975-988.
Keryer, G., Di Fiore, B., Celati, C., Lechtreck, K. F., Mogensen, M.,
Delouvée, A., Lavia, P., Bornens, M. and Tassin, A. M. (2003). Part of
Ran is associated with AKAP450 at the centrosome. Involvement in
microtubule organizing activity. Mol. Biol. Cell (in press).
Lengauer, C., Kinzler, K. W. and Vogelstein, B. (1998). Genetic instabilities
in human cancers. Nature 396, 643-649.
Lingle, W. L. and Salisbury, J. L. (2000). The role of the centrosome in the
development of malignant tumors. Curr. Top. Dev. Biol. 49, 313-329.
Mayor, T., Stierhof, Y. D., Tanaka, K., Fry, A. M. and Nigg, E. A. (2000).
The centrosomal protein C-Nap1 is required for cell cycle-regulated
centrosome cohesion. J. Cell Biol. 151, 837-846.
Mayor, T., Hacker, U., Stierhof, Y. D. and Nigg, E. A. (2002). The
mechanism regulating the dissociation of the centrosomal protein C-Nap1
from mitotic spindle poles. J. Cell Sci. 115, 3275-3284.
Melchior, F. (2001). Ran GTPase cycle: one mechanism – two functions. Curr.
Biol. 11, R257-R260.
Meraldi, P. and Nigg, E. A. (2001). Centrosome cohesion is regulated by a
balance of kinase and phosphatase activities. J. Cell Sci. 114, 3749-3757.
Meraldi, P., Lukas, J., Fry, A. M., Bartek, J. and Nigg, E. A. (1999).
Centrosome duplication in mammalian somatic cells requires E2F and
Cdk2-cyclin A. Nat. Cell Biol. 1, 88-93.
Merdes, A. and Cleveland, D. W. (1997). Pathways of spindle pole formation:
different mechanisms; conserved components. J. Cell Biol. 138, 953-956.
Moore, W., Zhang, C. and Clarke, P. R. (2002). Targeting of RCC1 to
chromosomes is required for proper mitotic spindle assembly in human
cells. Curr. Biol. 12, 1442-1447.
Moudjou, M. and Bornens, M. (1994). Isolation of centrosomes from
cultured animal cells. In Cell Biology: A Laboratory Handbook (ed. J. E.
Celis), pp. 595-604. Academic Press.
Nachury, M. V., Maresca, T. J., Salmon, W. C., Waterman-Storer, C. M.,
Heald, R. and Weis, K. (2001). Importin beta is a mitotic target of the small
GTPase Ran in spindle assembly. Cell 104, 95-106.
Ohba, T., Nakamura, M., Nishitani, H. and Nishimoto, T. (1999). Self-
organization of microtubule asters induced in Xenopus egg extracts by GTP-
bound Ran. Science 284, 1356-1358.
Piel, M., Meyer, P., Khodjakov, A., Rieder, C. L. and Bornens, M. (2000).
The respective contributions of the mother and daughter centrioles to
centrosome activity and behavior in vertebrate cells. J. Cell Biol. 149, 317-
Piel, M., Nordberg, J., Euteneuer, U. and Bornens, M. (2001). Centrosome-
dependent exit of cytokinesis in animal cells. Science 291, 1550-1553.
Pihan, G. A. and Doxsey, S. J. (1999). The mitotic machinery as a source of
genetic instability in cancer. Semin. Cancer Biol. 9, 289-302.
Ren, M., Drivas, G., D’Eustachio, P. and Rush, M. G. (1993). Ran/TC4: a
small nuclear GTP-binding protein that regulates DNA synthesis. J. Cell
Biol. 120, 313-323.
Ren, M., Coutavas, E., D’Eustachio, P. and Rush, M. G. (1994). Effects of
mutant Ran/TC4 proteins on cell cycle progression. Mol. Cell. Biol. 14,
Richards, S. A., Lounsbury, K. M. and Macara, I. G. (1995). The C
terminus of the nuclear RAN/TC4 GTPase stabilizes the GDP-bound state
and mediates interactions with RCC1, RAN-GAP, and HTF9A/RANBP1. J.
Biol. Chem. 270, 14405-14411.
Richards, S. A., Lounsbury, K. M., Carey, K. L. and Macara, I. G. (1996).
A nuclear export signal is essential for the cytosolic localization of the Ran
binding protein, RanBP1. J. Cell Biol. 134, 1157-1168.
Schliwa, M., Pryzwansky, K. B. and Euteneuer, U. (1982). Centrosome
splitting in neutrophils: an unusual phenomenon related to cell activation
and motility. Cell 31, 705-717.
Schliwa, M., Pryzwansky, K. B. and Borisy, G. G. (1983). Tumor promoter-
induced centrosome splitting in human polymorphonuclear leukocytes. Eur.
J. Cell Biol. 32, 75-85.
Sherline, P. and Mascardo, R. N. (1982). Epidermal growth factor induces
rapid centrosomal separation in HeLa and 3T3 cells. J. Cell Biol. 93, 507-512.
Sluder, G. and Rieder, C. L. (1985). Centriole number and the reproductive
capacity of spindle poles. J. Cell Biol. 100, 887-896.
Sluder, G., Thompson, E. A., Miller, F. J., Hayes, J. and Rieder, C. L.
(1997). The checkpoint control for anaphase onset does not monitor excess
numbers of spindle poles or bipolar spindle symmetry. J. Cell Sci. 110, 421-
Walczak, C. E., Vernos, I., Mitchison, T. J., Karsenti, E. and Heald, R.
(1998). A model for the proposed roles of different microtubule-based motor
proteins in establishing spindle bipolarity. Curr. Biol. 8, 903-913.
Wiese, C., Wilde, A., Moore, M. S., Adam, S. A., Merdes, A. and Zheng,
Y. (2001). Role of importin-beta in coupling Ran to downstream targets in
microtubule assembly. Science 291, 653-656.
Wilde, A. and Zheng, Y. (1999). Stimulation of microtubule aster formation
and spindle assembly by the small GTPase Ran. Science 284, 1359-1362.
Wilde, A., Lizarraga, S. B., Zhang, L., Wiese, C., Gliksman, N. R.,
Walczak, C. E. and Zheng, Y. (2001). Ran stimulates spindle assembly by
altering microtubule dynamics and the balance of motor activities. Nat. Cell
Biol. 3, 221-227.
Wolff, A., de Nechaud, B., Chillet, D., Mazarguil, H., Desbruyeres, E.,
Audebert, S., Edde, B., Gros, F. and Denoulet, P. (1992). Distribution of
glutamylated alpha and beta-tubulin in mouse tissues using a specific
monoclonal antibody, GT335. Eur. J. Cell Biol. 59, 425-432.