Oncogene (1997) 14, 2383 ± 2393
 1997 Stockton Press All rights reserved 0950 ± 9232/97 $12.00
Disregulation of mitotic checkpoints and regulatory proteins following
acute expression of SV40 large T antigen in diploid human cells
Ted Hung-Tse Chang1,4, F Andrew Ray2, D Alan Thompson1,5 and Robert Schlegel1,3
1
Department of Molecular and Cellular Toxicology, Harvard School of Public Health, 665 Huntington Avenue, Boston,
Massachusetts 02115; 2Department of Microbiology, Albany Medical College, 47 New Scotland Avenue, Albany, New York 12208,
USA; 3Chiron Diagnostics, 333 Coney Street, East Walpole, MA 02032-1597, USA
SV40 large T antigen (T) inactivates the tumor
suppressor proteins p53 and pRb, and can induce cells
to enter DNA replication at inappropriate times. We
show here that T also compromises three cell cycle
checkpoints that regulate the entry into and exit from
mitosis. Human diploid ®broblasts infected with a
retrovirus expressing T displayed an attenuated radiation-induced mitotic delay, were more susceptible to
chemical-induced uncoupling of mitosis from the
completion of DNA replication, and were more likely
to exit mitosis and rereplicate their DNA when mitotic
spindle assembly was inhibited. Consistent with altered
mitotic checkpoint control, cells expressing T displayed
elevated protein levels and/or associated activities of
the mitotic regulatory proteins cyclin A, cyclin B,
Cdc25C and p34cdc2. These changes in mitotic control
were evident within 5 ± 10 population doublings after
retroviral infection, indicating a direct e€ect of T
expression. Cells acutely infected with the T-expressing
retrovirus su€ered numerical and structural chromosome aberrations, including increases in aneuploidy,
dicentric chromosomes, chromatid exchanges and
chromosome breaks and gaps. These ®ndings indicate
that T rapidly disrupts mitotic checkpoints that help
maintain genomic stability, and suggest mechanisms by
which T induces chromosome aberrations and promotes
the immortalization and neoplastic transformation of
human cells.
Keywords: mitosis; cell cycle; checkpoints; SV40 T
antigen; chromosome aberrations
Introduction
Simian virus 40 (SV40) is a member of a class of DNA
tumor viruses that can immortalize primary human
cells in culture. The multifunctional large T antigen
protein of SV40 is necessary and sucient for
immortalization and is capable of transforming human
cells to tumorigenicity in athymic mice (Ray et al.,
1990; Ray and Kraemer, 1993). The mechanisms by
which T immortalizes cells is not completely understood, but it is believed to be dependent upon the
Correspondence: R Schlegel
Present addresses: 4DeÂpartment de Biologie MoleÂculaire, Institut
Pasteur, 25, rue du Dr Roux, F-75724 Paris CeÂdex 15, France.
5
Department of Pathology, Harvard Medical School, 220 Longwood
Avenue, Boston, MA 02115, USA
Received 30 July 1996; revised 10 April 1997; accepted 10 April 1997
ability of T to inactivate critical tumor suppressor
proteins as well as to induce chromosome aberrations
in the host cell (reviewed in Fanning and Knippers,
1992). Expression of SV40 T antigen (T) in normal
human cells extends their proliferative lifespan past the
point where they would normally senesce, leading to a
crisis period from which immortal clones emerge at a
low frequency (Stein, 1985; Wright et al., 1989). Tinduced genomic damage during precrisis replication
and crisis likely leads to the accumulation of mutations
in key growth regulatory genes, thereby facilitating
immortalization and neoplastic transformation (reviewed in Solomon et al., 1991).
The amino acid residues of T critical for transforming cells map to three domains: an N terminal domain
which may bind p300, an amino acid 105 ± 114 domain
which binds the retinoblastoma susceptibility protein
pRb, and an amino acid 398 ± 626 domain which binds
p53 (DeCaprio et al., 1988; Tack et al., 1989; Lin and
Simmons, 1991a, b; Zhu et al., 1991, 1992; Quartin et
al., 1994). T inactivates pRb and p53, well-characterized tumor suppressor proteins that regulate cell cycle
entry into S phase. pRb is normally phosphorylated in
G1 by the kinase complexes composed of D-type
cyclins and cyclin dependent kinases (cdk) 4 and 6,
thereby inactivating the growth suppressing activity of
pRb (reviewed in Sherr, 1993). p53 is the most
frequently mutated gene in common human cancers
and is required for mediating cell cycle arrest in G1
after exposure to ionizing radiation as well as for
activating apoptosis in some cell types (reviewed in
Zambetti and Levine, 1993; Meikrantz and Schlegel,
1995).
Evidence strongly implicates a role for p53 not only
in regulating entry into DNA replication, but also in
controlling entry into and exit from mitosis. Transient
expression of wild type p53 in normal human
®broblasts arrests cells in both G1 and G2 (Agarwal
et al., 1995; Stewart et al., 1995). In mouse embryo
®broblasts, p53 participates in a cell cycle checkpoint
preventing exit from mitosis in the absence of a
properly assembled mitotic spindle (Cross et al.,
1995). Cells from mice expressing an SV40 T antigen
transgene display multiple centrioles, multipolar
mitoses, and abnormal chromosome segregation,
which are probably due to the loss of p53 in these
cells (Levine et al., 1991; Fukasawa et al., 1996). Entry
into mitosis can be modulated by T antigen. Human
and rodent cells immortalized by a thermolabile T
construct cease to proliferate at the nonpermissive
temperature, arresting with an increased G2 cell
population (Jat and Sharp, 1989; Radna et al., 1989;
Mignotte et al., 1990). In addition, recent work by
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
2384
Kaufmann et al. (1995) has shown that precrisis human
®broblasts that express T antigen display a reduced
mitotic delay following DNA damage. It was unclear,
however, whether this e€ect was a direct consequence
of T antigen expression or was a phenotype acquired
during the extended proliferation of genetically
unstable T antigen-expressing cell clones. The human
papillomavirus E6 oncoprotein, which targets p53 for
degradation, was found to attenuate mitotic delay in
normal human ®broblasts only after an extended
period of cell replication (Paules et al., 1995).
Adenovirus E1A, another DNA tumor virus oncoprotein, can drive quiescent cells into both DNA
replication and mitosis (Zerler et al., 1987). The
domains of E1A required for these two functions are
separable, implying that E1A interacts with cellular
proteins that regulate entry into mitosis and that these
proteins are distinct from those regulating entry into S
phase.
The cyclin B/p34cdc2 and cyclin A/p34cdc2 heterodimeric complexes are both required for cells to enter
mitosis (reviewed in Norbury and Nurse, 1992). At the
onset of mitosis, cyclin B/p34cdc2 is dephosphorylated
and activated by the Cdc25C phosphatase. The
resulting active kinase complex (also known as
maturation promotion factor or MPF) phosphorylates
key structural substrates, including histones, nuclear
lamins and cytoskeletal proteins (reviewed in Nigg,
1993). The immortal state is often correlated with
elevated activities, protein levels and mRNA levels of
the mitotic regulatory proteins cyclin A, cyclin B, and
p34cdc2 (e.g. Buckley et al., 1993; Keyomarsi and
Pardee, 1993; Oshima et al., 1993; Rice et al., 1993;
Steinmann et al., 1994; Chang and Schlegel, 1996).
These changes have been documented in both rodent
and human cells and in cells that have undergone
either spontaneous or viral-induced immortalization.
Although elevated protein levels and kinase activities
associated with cyclin A, cyclin B and p34cdc2 have been
seen in precrisis human cells that express SV40 T
antigen (Kaufmann et al., 1995; Chang and Schlegel,
1996), these studies were conducted with stably
transfected single cell clones that had expressed T
antigen for the long period of time required for clonal
expansion. Therefore, it was not possible to determine
whether these cell cycle e€ects were a direct consequence of T antigen expression or were an indirect
e€ect produced by genetically altered clones that
acquired a growth advantage and dominated the cell
population during extended proliferation. To determine
whether disregulation of mitotic control proteins is a
direct result of T expression and whether this viral
oncoprotein directly alters mitotic checkpoint control,
we expressed T antigen in normal human diploid cells
using a retroviral vector. In this report, we show that
acute retroviral-mediated expression of SV40 T antigen
in human diploid ®broblasts increased the kinase
activities associated with mitotic regulatory proteins
within a few population doublings after retroviral
infection. This rapid disregulation of mitotic control
proteins was accompanied by increased karyotype
instability and by compromised integrity of three cell
cycle checkpoints regulating entry into and exit from
mitosis. Cells expressing T su€ered a loss in the ability
to prevent entry into mitosis in the presence of
unreplicated DNA, entry into mitosis in the presence
of damaged DNA, and exit from mitosis and
rereplication of DNA in the absence of a properly
assembled mitotic spindle.
Results
Retroviral infections
To study rapid changes in mitotic regulation caused by
SV40 T antigen, we constructed an amphotropic
retrovirus capable of infecting and expressing T in
normal human cells (Figure 1). All experiments used
intermediate passage IMR-90 (population doubling
*25) human diploid lung ®broblasts that were
infected in parallel with either the LTSN (T-containing) retrovirus or the control LXSN retrovirus and
selected with 300 mg/ml G418 for 5 ± 10 days prior to
analysis. Retroviral-mediated gene transfer was efficient, with greater than 50% of cells gaining G418
resistance. Immunoblots con®rmed that LTSN-infected
cells expressed T protein, and karyotype analysis
con®rmed that the retrovirally expressed T had
clastogenic and aneuploidogenic activity (see Table 1
and Figure 3). Cells infected with LXSN and LTSN
are referred to as IMR-90 LXSN and IMR-90 LTSN,
respectively.
T rapidly induces chromosome aberrations
Karyotype analysis was performed as described (Ray et
al., 1992) on IMR-90 LXSN and IMR-90 LTSN cells.
T induced statistically signi®cant increases in tetraploidy, dicentric chromosomes, chromatid exchanges,
chromosome breaks, chromosome gaps, and the
number of cells containing two or more aberrations
(Table 1). Chromatid exchanges are, in our experience,
an extremely rare event in normal human ®broblasts,
and the appearance of these aberrations is striking. The
high frequency of dicentric chromosomes in LTSN
infected cells is consistent with previous studies
Figure 1 Map of pLTSN retroviral vector. The SV40 T gene
from the plasmid pRSVEdl884 was inserted into the retroviral
vector pLXSN, via the plasmid pBR322, as described in Materials
and methods. Shown are the 5' LTR, 3' LTR, SV40 early
promoter (SV), Tn5 neomycin phosphotransferase gene (neo),
origin of replication, and ampicillin resistance gene (Ap) from
pLXSN. The SV40 T coding sequence is from pRSVEdl884
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
2385
Table 1 Karyotype analysis of IMR-90 cells after infection with
LXSN and LTSN
Chromatid gaps
Chromatid breaks
Chromatid exchanges
Chromosome gaps
Chromosome breaks
Minute chromosomes
Ring chromosomes
Dicentric chromosomes
Tetraploid cells
Cells with 1 aberration or more
Cells with 2 aberrations or more
LXSN
LTSN
5%
4%
0%
0%
4%
0%
0%
0%
3.9%
6%
4%
4%
8%
13%
0%
1%
15%
7.3%
27%
1%
58%
13%
X
T
X
T
cyclin A —
P50.05
P50.01
P50.05
P50.01
P50.01
P50.01
P50.01
100 cells of each group were examined, except for chromatid
exchanges, where 200 IMR-90 LXSN cells were examined, and
tetraploid analysis, where greater than 500 cells were examined from
each treatment. Signi®cance was determined by binomial statistics
reporting high frequencies of dicentric chromosomes in
human cells infected with SV40 virus (Koprowski et
al., 1962; Wolman et al., 1964). These observations
demonstrate that T antigen alone rapidly causes
numerical and structural chromosome aberrations
after expression in normal human cells, that this
activity requires no assistance from small t or any
other SV40 molecule, and that the aberrations
produced are similar to those seen with wild type
SV40 virus.
Elevated expression and associated kinase activities of
mitotic regulatory proteins following acute expression of
T antigen
If the karyotype instability induced in human cells by
SV40 T antigen is related to its ability to alter normal
mitotic control, one would expect mitotic regulatory
proteins to be a€ected shortly after T antigen is
expressed. To address this question, histone H1 kinase
assays were performed on asynchronously growing
IMR-90 LXSN and IMR-90 LTSN, allowing analysis
of e€ects within days of the gene transfer event. In
independent, matched infections, consistent elevations
in the histone H1 kinase activities associated with
cyclin A, cyclin B and p34cdc2 were observed in LTSNinfected cells when compared with LXSN-infected cells
(Figure 2). On average, cyclin A-associated activity was
elevated 4.1-fold, cyclin B-associated activity 2.4-fold,
and p34cdc2 activity 2.6-fold, as determined with a
PhosphorImager. An additional analysis of cyclin Bassociated activity (Figure 5b) revealed a 3.3-fold
increase following expression of T. These results
demonstrate that elevated mitotic kinase activities are
a direct result of T antigen expression.
To determine if increased histone H1 kinase
activities were accompanied by increased abundance
of the corresponding proteins, cell extracts were
prepared for immunoblotting concurrently with histone H1 kinase extracts. Cells were analysed for the
expression of T antigen, p53, cyclin A, cyclin B, p34cdc2,
Cdc25C, and the extracellular signal regulated kinases
ERK1 and ERK2 (Figure 3). As expected, T was
detected in LTSN-infected cells but not LXSN-infected
cells. p53 was also elevated in LTSN-infected cells,
con®rming the ability of the retrovirally expressed T to
bind to and stabilize p53 (Lane and Crawford, 1979;
cyclin B —
p34cdc2 —
Figure 2 Histone H1 kinase activities associated with cyclin A,
cyclin B, and p34cdc2 in IMR-90 cells infected with LXSN and
LTSN. IMR-90 cells were infected with LXSN (X) and LTSN (T)
in two independent and matched infections. 50 mg of protein from
each cell type was immunoprecipitated with antibodies against
cyclin A, cyclin B and p34cdc2 and assayed for histone H1 kinase
activity, as described in Materials and methods
X
T
X
T
T antigen —
—
p53 —
cyclin A —
cyclin B —
Cdc25C —
p34cdc2 —
ERK1 —
ERK2 —
Figure 3 Immunoblots of SV40 T, p53, cyclin A, cyclin B,
Cdc25C, p34cdc2 and ERK1/ERK2 in IMR-90 cells infected with
LXSN (X) and LTSN (T). 50 mg of protein from each cell type
was immunoblotted with antibodies against the appropriate
proteins, as described in Materials and methods
Linzer and Levine, 1979; Oren et al., 1981). Cyclin A
and p34cdc2 protein levels were elevated twofold and
2.3-fold, respectively, in IMR-90 LTSN cells relative to
IMR-90 LXSN cells, qualitatively similar to previous
®ndings in stably transfected cell clones (Chang and
Schlegel, 1996). Two closely-spaced bands were
speci®cally immunoreactive with our polyclonal cyclin
B antiserum. These same two bands are seen in extracts
of mitotic HeLa cells (data not shown). It is not known
whether this upper band represents a phosphorylated
form of cyclin B seen in active kinase complexes, but
the abundance of this immunoreactive species was
increased approximately threefold in T-expressing cells,
consistent with increased cyclin B/p34cdc2 activity seen
in these cells. The overall level of cyclin B, however,
was not a€ected by acute expression of T. 2.4-fold
increases in the protein level of Cdc25C, an activator
of p34cdc2 kinase activity, were also observed in T-
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
2386
expressing cells. Increased expression of this phosphatase may contribute to the elevated mitotic kinase
activities seen in IMR-90 LTSN cells. The level of
ERK2 was essentially una€ected, while that of ERK1
was slightly decreased (30% less than controls),
demonstrating that not all cell cycle-regulated kinases
are induced by T.
Growth curves and ¯ow cytometric analysis of IMR90 LXSN and IMR-90 LTSN cells were performed to
rule out the possibility that increased expression and/or
activities of mitotic regulatory proteins were a
consequence of accelerated growth or an increase in
the population of cells in G2/M phase, a phase where
some of these proteins are maximally expressed and
activated (Pines and Hunter, 1989, 1990). Growth
curves shown in Figure 4 indicate little di€erence in the
growth rate of IMR-90 LXSN and IMR-90 LTSN
cells, while ¯ow cytometric analyses (see upper two
panels of Figure 7) indicate essentially identical cell
cycle pro®les for these cells. It is evident, therefore,
that altered regulation of mitotic control proteins by
acute expression of T antigen can occur without
signi®cant changes in growth rate or cell cycle
distribution.
T antigen disrupts three mitotic checkpoints critical for
maintaining chromosome stability
Altered expression and activities of mitotic regulatory
proteins would be expected to have functional
consequences. Since expression of T antigen increases
karyotype instability (Ray et al., 1990, 1992; Stewart
and Bacchetti, 1991), we assessed the e€ects of acute
expression of T antigen on the integrity of three mitotic
checkpoints that are known to be important for
maintaining chromosome stability. Cells were assessed
for their ability to delay mitotic onset in the presence
of damaged or incompletely replicated DNA and to
delay mitotic exit in the absence of a properly formed
mitotic spindle.
Figure 4 Growth curves of IMR-90 cells infected with LXSN
and LTSN. Cell number was determined with a Coulter Counter,
using duplicate dishes for each time point. Error bars show the
standard error of the mean when error was larger than the ®gure
symbols
(i) Expression of T attenuates mitotic delay following
exposure to ionizing radiation Human cells typically
arrest in G2 after exposure to DNA damaging agents.
This delay of entry into mitosis is believed to facilitate
the repair of damaged DNA prior to chromosome
condensation and segregation, thereby preserving
genomic integrity. When this mitotic checkpoint is
abrogated by chemical agents (Busse et al., 1977, 1978;
Lau and Pardee, 1982; Fan et al., 1995; Powell et al.,
1995; Russell et al., 1995) and gene mutations (Weinert
and Hartwell, 1988; Brown et al., 1991; Weinert et al.,
1994; Walworth and Bernards, 1996), cells are sensitized
to DNA damaging agents, often su€ering dramatic
increases in chromosome aberrations and cell killing.
The mitotic delay pro®le of IMR-90 cells infected
with LXSN and LTSN was examined after exposure of
cells to 1.5 Gy gamma radiation (Figure 5a). In control
IMR-90 LXSN cells, the mitotic index dropped rapidly
to near zero for the ®rst 3 h after irradiation, recovered
slowly to about 30% of the unirradiated level within
6 h and by 8 h had approached control unirradiated
levels. This rapid drop in the mitotic index indicates an
intact G2 arrest point following irradiation. IMR-90
LTSN cells, however, displayed a greatly attenuated
G2 arrest following irradiation. The mitotic index
never fell below 30% of the unirradiated level, and had
fully recovered to control values within 4 h. The
relative mitotic indices for T-expressing cells were
signi®cantly higher than those for control cells
throughout the entire 6 h period where control cells
exhibited a mitotic delay.
The relative activity of cyclin B/p34 cdc2 has
previously been shown to correlate with the extent of
mitotic delay in human cell lines (Kaufmann et al.,
1995). To determine whether premature recovery of
IMR-90 LTSN cells from radiation-induced mitotic
delay was accompanied by a commensurate recovery of
cellular mitotic kinase activity, the cyclin B-associated
histone H1 kinase activities of IMR-90 LXSN and
IMR-90 LTSN cells were assayed at 2 h time intervals
after 1.5 Gy irradiation (Figure 5b). Relative histone
H1 kinase activities correlated well with the mitotic
indices for both cell populations (Table 2). By 4 h
postirradiation, the cyclin B-associated kinase activity
of IMR-90 LTSN had returned to unirradiated levels,
while the activity of IMR-90 LXSN cells was still
markedly suppressed. It is interesting to note that the
mitotic kinase activity of IMR-90 LTSN cells at the
time of lowest mitotic index (2 h post-irradiation) was
still higher than the maximum activity of IMR-90
LXSN cells. It is apparent that entry into mitosis is not
simply regulated by the absolute level of cyclin Bassociated kinase activity, but is more closely aligned
to changes in relative activity. Why higher kinase
activities are required for LTSN-infected cells to enter
mitosis is unknown, but it seems that these Texpressing cells somehow partially compensate for
their increased baseline activity.
The sharp drop in the mitotic index of irradiated
IMR-90 LXSN cells (Figure 5a) indicated that they
were delayed at a point in the cell cycle less than 1 h
before mitosis. To con®rm this G2 delay, ¯ow
cytometry was performed on IMR-90 LXSN and
LTSN cells synchronized by a double thymidine
block. At a point near the S/G2 boundary, and 5.8 h
after release from a thymidine-imposed G1/S arrest,
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
cells were irradiated with 1.5 Gy. The sizes of the G2/
M populations were assessed at subsequent time points
(Figure 5c). Irradiated IMR-90 LXSN cells exited from
a
b
0
2
4
6
8
LXSN —
LTSN —
c
Figure 5 Radiation-induced mitotic delay of IMR-90 cells infected
with LXSN and LTSN. Cells were irradiated with 1.5 Gy from a
60
Co source. (a) Mitotic indices were determined from methanol
®xed and Hoechst stained cells at the indicated times after
irradiation. Mitotic indices are shown as a percentage of the
appropriate unirradiated control cells. The mitotic indices for
unirradiated controls were 1.4% for both LXSN- and LTSNinfected cells. 2000 cells were scored for each time point, with
signi®cant di€erences between LXSN- and LTSN-infected cells
determined by binomial statistics (*P50.05; **P50.01). (b)
Histone H1 kinase activity associated with cyclin B was assayed
from cell extracts, each containing 50 mg of protein, at the indicated
times (hours) after irradiation, as described in Materials and
methods. Film exposure times were identical for all samples. (c) Cell
populations were exposed to 1.5 Gy, 5.8 h after release from double
thymidine synchronization. The percentages of cells in G2/M phase
were determined by ¯ow cytometry at the indicated time points
post-irradiation. At least 104 cells were analysed for each time point
G2/M roughly 2 h later than IMR-90 LTSN cells, with
a maximal di€erence in G2/M populations occurring
6 h post-irradiation. No such di€erence was seen in
unirradiated cells. The divergence between G2/M cell
populations of IMR-90 LXSN and LTSN began 3 h
after irradiation, consistent with the amount of time
required for the ®rst S/G2 cells to pass through mitosis
(data not shown).
(ii) T sensitizes human diploid ®broblasts to ca€eineinduced premature mitosis The onset of mitosis in all
eukaryotic cells is normally suppressed until DNA
replication is completed. Treatment of hamster cells
with chemical agents such as ca€eine, okadaic acid, 2aminopurine and staurosporine can uncouple S phase
and M phase, inappropriately activating mitotic
kinases prior to the completion of DNA replication
(Schlegel and Pardee, 1986; Schlegel et al., 1990;
Yamashita et al., 1990; Steinmann et al., 1991; Tam
and Schlegel, 1992). The resulting premature mitosis
phenotype is characterized by nuclear envelope breakdown and pulverized, condensed chromatin (Johnson
and Rao, 1970; Schlegel and Pardee, 1986). In contrast
to rodent cells, human cells very rarely undergo
chemically-induced premature mitosis. Since we have
shown previously that ectopic overexpression of cyclin
B in human HT1080 osteosarcoma cells can sensitize
the cells to ca€eine-induced premature mitosis (Tam et
al., 1995), and since T-expressing cells exhibit elevated
cyclin B-associated kinase activity, we asked whether
IMR-90 LTSN cells are more susceptible to spontaneous and ca€eine-induced premature mitosis than
IMR-90 LXSN cells.
IMR-90 cells infected with either LXSN or LTSN
were arrested in S phase with hydroxyurea for 15 h, the
last 8 h of which was in the presence or absence of
ca€eine (5 mM). Nocodazole was also added during the
last 8 h to capture all cells undergoing premature mitosis.
Chromosome spreads prepared from these cells indicated
that acute expression of T did not induce any detectable
spontaneous premature mitosis, but did signi®cantly
increase the frequency of ca€eine-induced premature
mitosis, with levels reaching approximately 1% of the
cell population (Figure 6a). Premature mitotic cells were
identi®ed microscopically on the basis of their pulverized
condensed chromatin (Figure 6b), which is easily
distinguished from normal nuclei. When nocodazole is
omitted during ca€eine treatment, prematurely condensed chromatin typically decondenses and forms
Table 2 Mitotic indices and cyclin B-associated histone H1 kinase
activities of IMR-90 cells infected with LXSN or LTSN, after 1.5 Gy
irradiation
Hours
after
irrad.
0
1
2
3
4
5
6
8
a
Mitotic index
(% of unirradiated)a
LXSN
LTSN
100
7
4
11
26
30
33
81
100
41
33
41
107
107
133
122
Cyclin B-assoc. H1
kinase activity
(% of LXSN unirrad.)
LXSN
LTSN
100
332
28
149
50
329
93
78
432
590
Mitotic indices for unirradiated IMR-90 cells infected with LXSN
and LTSN were both 1.4% (mean of two trials)
2387
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
2388
a
b
c
Figure 6 Spontaneous and ca€eine-induced premature mitosis
in IMR-90 cells infected with LXSN (X) and LTSN (T). (a)
Cells were arrested in S phase for 15 h with 2.5 m M
hydroxyurea treatment, with or without 5 mM ca€eine treatment during the ®nal 8 h. Premature mitotic cells were collected
in 0.1 mg/ml nocodazole during the ®nal 8 h. The frequency of
premature mitotic cells was determined by ¯uorescence
microscopy of Hoechst stained nuclei. 3000 ± 5000 cells were
examined for each treatment group. (b) Typical premature
mitotic cell, shown adjacent to normal nucleus, displaying
pulverized and condensed chromatin, and breakdown of the
nuclear envelope. (c) IMR-90 cells infected with LTSN were
arrested in S phase with hydroxyurea (2.5 mM) for 27 h, with
5 mM ca€eine treatment during the ®nal 20 h. Nocodazole was
not added, allowing premature mitotic cells to undergo
chromatin decondensation and reformation of the nuclear
envelope, thereby generating micronucleated cells (such as the
one shown)
micronucleated cells (Schlegel and Pardee, 1986; Schlegel
et al., 1990). Such multinucleate cells were observed in
IMR-90 LTSN cells following 20 h of ca€eine treatment
(Figure 6c). These ®ndings indicate that acute expression
of T in diploid human cells imparts a susceptibility to
chemically-induced premature mitosis and compromises
the cell cycle checkpoint coupling mitotic entry with the
completion of DNA replication.
(iii) Expression of T disrupts the mitotic spindle
assembly checkpoint in diploid human cells Fibroblasts
from p53 knockout mice embryos show a predisposition
to rereplicate their DNA in the presence of mitotic
spindle inhibitors such as nocodazole and colcemid
(Cross et al., 1995). Conversely, ®broblasts from wild
type p53-expressing mice respect this spindle checkpoint,
as assayed by ¯ow cytometry, and refrain from undergoing a new round of DNA replication in the presence of
an unassembled mitotic spindle. In a number of human
cell lines, however, there was no obvious role for p53 in
coupling the onset of DNA replication to the proper
completion of mitosis (Kung et al., 1990; Roberts et al.,
1990; Schimke et al., 1991; Lopes et al., 1993; Liebmann
et al., 1994; Cross et al., 1995). Since transformed human
cells were used in the above studies, we wanted to
determine whether acute expression of T antigen, which
binds to and inactivates wild type p53 (Lane and
Crawford, 1979; Linzer and Levine, 1979; Oren et al.,
1981), could disrupt the mitotic spindle checkpoint in
otherwise normal human diploid cells.
IMR-90 LXSN and IMR-90 LTSN cells were treated
with 50 ng/ml nocodazole or no nocodazole for 24 or
48 h, harvested by trypsinization into single cells, and
analysed by ¯ow cytometry (Figure 7). Asynchronously
growing IMR-90 LXSN and IMR-90 LTSN displayed
similar cell cycle distributions. After 24 h of nocodazole
Figure 7 Integrity of the mitotic spindle checkpoint in cells
infected with LXSN and LTSN. Asynchronously growing cells
were treated with either 50 ng/ml nocodazole or no nocodazole.
After 48 h of treatment, cells were analysed for DNA content by
propidium iodide staining and ¯ow cytometry, as described in
Materials and methods
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
treatment, the majority of cells of both cell types were in
the G2/M peak (57% for IMR-90 LXSN and 74% for
IMR-90 LTSN, data not shown). After 48 h in
nocodazole, or roughly two cell cycles, almost all of the
IMR-90 LXSN cells in G2/M remained arrested. On the
other hand, 18% of IMR-90 LTSN cells had rereplicated
their DNA to 4C DNA content without segregating
chromosomes into daughter cells. Cells with DNA
content between 2C and 4C may be in the S-phase
following escape from the spindle checkpoint. 27% of
IMR-90 LTSN cells contained greater than 2C DNA
content, up to and including 4C, compared to 6.7% of
IMR-90 LXSN cells. Without nocodazole, 3.0% of
IMR-90 LXSN and 4.2% of IMR-90 LTSN cells were
found in this 42C population. A more prolonged
incubation of IMR-90 LTSN cells in nocodazole (72
and 96 h) did not generate a distinct population of cells
with 8C DNA content. These ®ndings indicate that T
antigen rapidly compromises the ability of human cells to
suppress a new round of DNA replication when a proper
mitotic spindle cannot be formed. Loss of this checkpoint
could certainly contribute to the elevated frequencies of
aneuploidy and tetraploidy commonly found in Texpressing cells.
Discussion
In this report, we have used retroviral-mediated gene
transfer to examine the acute e€ects of SV40 T antigen
on mitotic control in normal human diploid cells.
Earlier studies have established the rapid appearance of
karyotype defects following infection of rodent and
human cells with SV40 virus (Koprowski et al., 1962;
Wolman et al., 1964), and here we observe that large T
alone induces aberrations very shortly after its
expression. Other SV40 products, such as small t
antigen, are not necessary for the induction of early
defects. In some cell systems, small t antigen has been
reported to increase the eciency of immortalization
and transformation, although t is neither clastogenic
nor aneuploidogenic (Bikel et al., 1987; Choi et al.,
1988). The high frequency of cells containing dicentric
chromosomes (15 of 100 cells) within 10 days after T
expression is consistent with previous karyotype
analyses of human cells infected with SV40 virus
(Koprowski et al., 1962; Wolman et al., 1964). T also
caused four chromatid exchange events, extremely rare
events, in 100 normal diploid human cells. The
exchange events must have occurred in the G2 or S
phase that immediately preceded the metaphase when
the chromosomes were analysed, suggesting that a
chromatid exchange event may transpire in 4% of Texpressing cells during each cell cycle. The aberrations
seen in chromatids, as well as in chromosome number,
are consistent with the observed disregulation of cell
cycle checkpoints and regulatory proteins controlling
the onset of mitosis.
Elevations in cyclin A-, cylin B-, and p34cdc2associated kinase activities that were induced rapidly
after T expression are in agreement with our earlier
®ndings in single clones stably transfected with SV40
tumor antigens (Chang and Schlegel, 1996). The
magnitudes of the elevations in the current study are
not as great, on average, as those seen in the clones
stably transfected with T. This di€erence may be
explained by our observation that the protein
abundance and associated kinase activities of mitotic
regulatory proteins decline as normal diploid cells age
and approach senescence. In our previous studies using
single cell IMR-90 clones that were transfected with
SV40 tumor antigen-expressing constructs, the control
transfectants were considerably older at the time of
analysis than were the control IMR-90 cells in this
report that were retrovirally-infected (population
doubling 55 versus 30). This led to a higher baseline
level of protein abundance and kinase activities in the
current studies, and thus a lower overall T-induced
increase.
Acute expression of T did not result in an increase in
the overall levels of cyclin B protein, although cyclin Bassociated histone H1 kinase activity was elevated. It is
possible that the elevated cyclin B levels that we
observed previously in precrisis single cell clones
expressing transfected T (Chang and Schlegel, 1996)
were a secondary consequence of T expression.
Prolonged ectopic overexpression of cyclin B alone in
human cells is dicult to achieve (unpublished data),
perhaps due to a failure to exit mitosis (reviewed in
Murray, 1995). T antigen, however, was able to
increase cyclin B protein levels in precrisis single cell
clones without hindering cell growth, perhaps by
altering expression of other genes in a cooperative
and balanced fashion. The elevations in cyclin Bassociated kinase activity in the absence of increased
protein, seen in the present study, may be explained in
part by the elevations in Cdc25C, a positive activator
of the cyclin B/p34cdc2 kinase (Gautier et al., 1991;
Strausfeld et al., 1991; Lee et al., 1992). It is interesting
to note that simultaneous ectopic overexpression of
Cdc25C and cyclin B in hamster cells induces
premature mitosis and chromatin pulverization
(Heald et al., 1993). One can imagine that the
coordinated activation of the mitotic regulatory
machinery by T is capable of causing subtler
karyotype aberrations in human cells.
The increases in mitotic kinase activities are a direct
consequence of T expression, as they occur less than 10
days after retroviral infection, insucient time for these
e€ects to be secondary consequences of speci®c genetic
mutations that then overtake the cell population. The
rapid alteration in mitotic control by expression of T
suggests the possibility that disregulation of cyclin A,
cyclin B, and p34cdc2 may be a causative event in some
chromosome aberrations. Inappropriately regulated or
timed activation of MPF is often correlated with
catastrophic chromosome damage. Several mutants of
S pombe which prematurely activate p34cdc2 undergo
lethal mitosis (Russell and Nurse, 1986, 1987; Gould
and Nurse, 1989; Enoch and Nurse, 1990; Lundgren et
al., 1991; Rowley et al., 1992). For example, the S
pombe rum1 mutant activates p34cdc2 inappropriately in
the cell cycle, in conjunction with severe chromosomal
damage (Moreno and Nurse, 1994). Expression of T
alone in human diploid cells also induces increased
chromosome damage and cell death (Ray et al., 1990).
In addition, when mitotic delay following DNA
damage is prevented by genetic mutations (Weinert
and Hartwell, 1988, 1990; Brown et al., 1991; Weinert
et al., 1994) or by chemical treatments (Busse et al.,
1977, 1978; Lau and Pardee, 1982) that prematurely
activate cyclin B/p34cdc2, cells su€er signi®cantly more
2389
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
2390
cytogenetic damage. It is also possible that aberrant
mitotic events like improper spindle formation and
mitotic nondisjunction (Levine et al., 1991; Fukasawa
et al., 1996) may be incited by T antigen-induced
disregulation of mitotic cell cycle kinases.
Three checkpoints important for preserving genetic
integrity during passage through mitosis were compromised by expression of T. DNA damage-induced cell
cycle arrest prior to mitosis was shortened and less
complete in the presence of T. The loss of this
checkpoint in the S cerevisiae rad9 mutant results in
elevated levels of spontaneous aneuploidy and tetraploidy (Weinert and Hartwell, 1990), resembling the
nature of the T-induced aberrations seen in our
karyotype analysis. Several reports found an increased
sensitivity to a chemical-induced override of mitotic
delay in p53-impaired human and mouse cell types
(Fan et al., 1995; Powell et al., 1995; Russell et al.,
1995). Attenuation of the DNA damage-induced
mitotic delay in human cells without the assistance of
mitosis-inducing chemicals has been reported in one
cell clone stably transfected with T (Kaufmann et al.,
1995). We extend these ®ndings by observing
attenuated mitotic delay in a population of normal
human cells expressing T. Because these cells were
assayed about 1 week after retroviral infection, the
reduced mitotic delay must be a direct consequence of
T expression, rather than a result of T-induced
genomic alterations which are selected during the
clonal expansion of stable transfectants. Defective G2
checkpoint function has been seen in cells from
patients with the familial cancer syndrome ataxia
telangiectasia and in spontaneously immortalized, but
not preimmortal, Li-Fraumeni cells, linking wild type
p53 function with maintenance of the mitotic delay
checkpoint (Paules et al., 1995). However, the same
study found that human diploid foreskin ®broblasts
infected with a retrovirus expressing HPV-16 E6 did
not show an immediate attenuation of mitotic delay
following exposure to 3 Gy radiation, but rather a
gradual attenuation achieved 11 cell passages later,
implying that p53 is not an essential component of the
mitotic delay checkpoint. Our laboratory has seen, by
contrast, a nearly immediate attenuation of radiationinduced mitotic delay in IMR-90 human diploid lung
®broblasts infected with an LXSN-based, HPV-16 E6expressing retrovirus (unpublished data), suggesting
that the mitotic delay checkpoint across even very
similar cell types may show variable degrees of p53dependence.
Susceptibility to chemical-induced uncoupling of
mitosis from the completion of DNA replication was
imparted by acute expression of T antigen. Earlier work
from our laboratory showed that transient overexpression of cyclin B in HT1080 human osteosarcoma cells
increased the susceptibility of these cells to ca€eineinduced premature mitosis (Tam et al., 1995), and similar
elevations in the abundance and activities of mitotic
regulatory proteins may mediate this T-induced phenomenon in human diploid cells. It is interesting to note that
ectopic overexpression of cyclin B in this earlier study
increased the frequency of ca€eine-induced PCC from
approximately 0.3 to 2%, roughly the same increase as
that seen in the present study. Higher susceptibility to
chemical-induced premature mitosis implies a greater
degree of uncoupling between the mechanisms of DNA
replication and mitosis, which puts the genome at higher
risk of sustaining damage. Although the e€ect of T
antigen on the incidence of PCC was quite modest, one
can imagine that after numerous cell divisions slight
defects in mitotic control could have serious consequences on the cell population as a whole. For example, it
has been proposed that a basic di€erence between high
and low risk human papillomaviruses is the ability of the
former to compromise genome stability, even though this
instability is evident in less than one in 104 cells (White et
al., 1994). There is tremendous species-speci®c variability
in the response of cells to ca€eine-induced PCC. Human
and mouse cells are only slightly a€ected by ca€eine,
while hamster cells are very responsive, displaying PCC
levels of 50 ± 70% (Steinmann et al., 1991). At least part
of this species-speci®c di€erence is thought to be caused
by a less rigorous control of the timing of cyclin B
accumulation in hamster cells (Steinmann et al., 1991;
Tam et al., 1995).
Human diploid cells expressing T antigen displayed
a compromised mitotic spindle checkpoint, a defect
related to the spindle poison-induced polyploidy in
some, but not all, human tumor cell lines (Kung et al.,
1990; Roberts et al., 1990; Schimke et al., 1991; Lopes
et al., 1993; Liebmann et al., 1994). Although Cross et
al. (1995) found that ®broblasts from p537/7 mouse
embryos exhibited a defective mitotic spindle checkpoint, we have observed that osteosarcoma cell lines
expressing p53 (U2OS) and lacking p53 (SaOS-2) are
both susceptible to DNA rereplication in the presence
of 100 ng/ml nocodazole (unpublished data), indicating
that p53 status is not the sole determinant of spindle
checkpoint integrity in transformed human cells.
Results from the present study support the direct
participation of T antigen in disabling a spindle
checkpoint in human diploid cells and raise the
possibility that T may also contribute to spontaneous
DNA rereplication or improper chromosome segregation in the absence of spindle poisons. These ®ndings
are consistent with the gradual, spontaneous progression to tetraploidy observed in SV40 virus infected
human diploid ®broblasts as they approach crisis
(Lan et al., 1989).
Findings presented in this report suggest that the
disregulation of cell cycle events governing entry into
and exit from mitosis is one mechanism by which T
antigen destabilizes the genome and immortalizes
human cells. The biochemical properties of T antigen
responsible for disruption of mitotic control are at
present unknown. Cells which have lost wild type p53
function can develop over time an incomplete DNA
damage-induced mitotic delay (Kaufmann et al., 1995;
Paules et al., 1995), or become more sensitive to
chemicals that can override mitotic delay (Fan et al.,
1995; Powell et al., 1995; Russell et al., 1995). In
addition p53-de®cient mouse embryo ®broblasts display a compromised mitotic spindle checkpoint (Cross
et al., 1995), implying a role for p53 in several mitotic
checkpoints. These studies, however, did not ®nd a
requirement for wild type p53 in the mitotic spindle
checkpoint in human cells (Cross et al., 1995) or in
mitotic delay following DNA damage (Paules et al.,
1995). It appears, therefore, that T antigen disrupts
mitotic control through both p53-dependent and p53independent mechanisms, and we are currently
assessing the relative contributions of these pathways.
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
Materials and methods
Cell culture
IMR-90 human diploid lung ®broblasts (catalog #I90-P04)
(Nichols et al., 1977) were obtained from the National
Institute on Aging Cell Repository (Camden, NJ). Cells
were grown in Dulbecco's modi®ed Eagle medium (Gibco
BRL, Gaithersburg, MD) containing 10% iron-supplemented calf serum (Hyclone, Logan, UT) and 3.5 m M
glutamine in a humidi®ed 5% CO2 atmosphere at 378C.
Plasmids
The plasmids used were pLXSN, a retroviral expression
vector (Miller and Rosman, 1989), and pLTSN, a pLXSNbased vector containing SV40 large T antigen. To construct
pLTSN, the open reading frame of T antigen was
transferred from the plasmid pRSVEdl884 (Ray et al.,
1990), which is a mutant deleted for small t antigen
sequences (4890-4643 SV40 numbering), to pLXSN. SV40
viral DNA sequence from HindIII site 5171 to BamHI 2533
was excised from pRSVEdl884 and inserted into pBR322 at
the unique HindIII and BamHI sites, resulting in a plasmid
which we call pBR322-T. The T antigen-containing
fragment from the unique EcoRI site of pBR322 to the
HpaI 2666 site of SV40 was then excised from pBR322-T
and inserted into the retroviral vector pLXSN at the
unique EcoRI and HpaI sites of the multiple cloning site.
The resulting plasmid, which we call pLTSN (Figure 1),
contains the coding sequence of T antigen and excludes
promoter sequences from pRSVEdl884 and the SV40 T
antigen polyadenylation sequence, thereby promoting the
expression of full length viral transcripts and ecient
retroviral packaging.
Retroviral packaging and infection of cells
Plasmids were transfected into PA317 murine amphotropic
packaging cells by the calcium phosphate method described
by Donahue and Stein (1988). Single cell clones were
selected and expanded in 500 mg/ml G418 antibiotic (Gibco
BRL, Gaithersburg, MD). The viral titres of individual
clones were estimated by the eciency with which their
supernatants were able to confer G418 resistance to
NIH3T3 target cells. The supernatants from clones
producing a high titre of LXSN and LTSN virus were
sterile-®ltered and stored at 7808C. To infect IMR-90
®broblasts with LXSN or LTSN, 56105 cells plated on a
75 cm2 ¯ask were incubated with 4 ml of viral supernatant
and 8 mg/ml polybrene. Two hours later, 5 ml of normal
medium was added to the viral supernatant, and the next
day the medium was changed. On the second day after
infection, cells were selected in 300 mg/ml G418 for 7 to 10
days before initiating experiments.
Karyotype analysis
Cells were arrested with 0.1 mg/ml Colcemid for no more
than 4 h and metaphase spreads were prepared and stained
with 10% Giemsa. For each cell type, 100 metaphase spreads
were counted and scored for aberrations, except for
chromatid exchanges, where up to 200 cells were scored,
and tetraploidy, where no fewer than 500 cells were scored.
Growth curves
Cells (26104) were plated on 35 mm dishes. Twenty-four
hours later, the medium was replaced. Cell number was
determined from duplicate plates at the time of medium
replacement and on the 2 days thereafter by trypsinizing
cells and counting cell suspensions in a Coulter Electronics
ZM counter.
Immunoblotting
Cellular proteins were extracted by boiling cell pellets for
5 min in Laemmli sample bu€er (10 mM Tris pH 6.8, 2%
SDS, 4% glycerol, 5% b-mercaptoethanol) and separated
by electrophoresis through 10% or 7.5% denaturing
polyacrylamide gels (Laemmli, 1970). Each lane contained
50 mg of protein, as estimated by a modi®ed Lowry protein
assay (Peterson, 1977) using g-globulin as the standard.
After electrophoresis, proteins were transferred onto a
polyvinylidene di¯uoride membrane (Immobilon-P from
Millipore, Bedford, MA) in a tank blotting system at 60 V
for 5 h, in 25 mM Tris pH 8.3/192 mM glycine/20% (v/v)
methanol.
Proteins were probed with antiserum raised against the Cterminal peptide of human p34cdc2 (LDNQIKKM) (Oshima et
al., 1993), with antisera raised against full-length human
cyclin A and cyclin B proteins (Oshima et al., 1993), with
antiserum 692 recognizing the extracellular signal-regulated
kinases ERK1 and ERK2 (Boulton and Cobb, 1991), with
monoclonal antibody recognizing the carboxy terminus of
SV40 large T antigen, partially puri®ed by ammonium sulfate
precipitation from the supernatant of the PAb 423
hybridoma, previously designated as L23 (Harlow et al.,
1981), with Cdc25C monoclonal antibody (provided by Dr
Helen Piwnica-Worms), and with p53 antibody from the
DO1 hybridoma (Santa Cruz Biochemicals). Cdc25C
immunoblots were performed using the Amersham ECL kit,
according to manufacturer's instructions. For other antibodies, probing and washing procedures are as described
(Oshima et al., 1993). Brie¯y, for cyclin A, cyclin B and
p34cdc2, the blots were blocked for 1 h in Phosphate bu€ered
saline/0.3% Tween-20 (PBST) containing 5% Carnation
nonfat dry milk and incubated overnight at 48C in a
1 : 1000 dilution of antiserum in the same blocking solution.
Blocking was not performed for T, p53 and ERK blots. All
membranes were washed three times with PBST, incubated at
room temperature with 0.75 mCi/ml 125I-labeled protein A
(ICN, Irvine, CA) in PBST, and washed four times in PBST
before detection by indirect autoradiography using Kodak XOMAT ®lm and an intensifying screen. Bands were
quanti®ed by analysis of the probed membranes with a
Molecular Dynamics PhosphorImager and ImageQuant
software (Sunnyvale, CA).
Histone H1 kinase assay
Asynchronous cells were washed with warm PBS,
detached from the monolayer by incubation in PBS/
0.5 mM EDTA, lysed by resuspension in lysis bu€er
(40 mM HEPES pH 7.5, 250 mM NaCl, 15 mM MgCl2,
1% Triton X-100, 5 mM NaF, 5 mM b-glycerophosphate,
2 mM EGTA, 2 mM EDTA, 0.5 mM phenylmethanesulfonyl ¯uoride, 1 mM Na4VO3, 10 mg/ml each of aprotinin,
leupeptin and pepstatin) containing 18% sucrose, passed
ten times through a 25-gauge needle and kept on ice for
30 min. The extract was then clari®ed by centrifugation,
aliquoted, stored at 7808C, and assayed for protein
concentration as described above (Peterson, 1977). Cyclin
A, cyclin B and p34cdc2 were immunoprecipitated from
50 mg of extract using a 1 : 100 dilution of the appropriate
antiserum at 48C for 1 h, followed by a 1 h incubation at
48C in 20 mg/ml protein A-Sepharose beads. After
immunoprecipitation, the beads were washed twice in
0.5 ml lysis bu€er and three times in 0.5 ml washing bu€er
(20 mM HEPES pH 7.5, 15 mM EGTA, 20 mM MgCl2).
Histone H1 kinase activity was determined by incubation
of beads in 40 ml of washing bu€er containing 1 mM
dithiothreitol, 10 mg/ml A kinase inhibitory peptide
(Sigma), 0.5 mg/ml histone H1 (Gibco-BRL), and 50 mM
ATP (5500 c.p.m./pmol) for 15 min at 308C. The reaction
was stopped by adding 40 ml of 26 Laemmli sample
2391
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
2392
bu€er (Laemmli, 1970) and boiling for 10 min. Proteins
were separated on a 12% SDS-polyacrylamide minigel,
®xed in 25% (w/v) trichloroacetic acid, and stained with
Coomassie Brilliant Blue R-250. As described above for
immunoblots, a PhosphorImager was used to quantify 32P
counts incorporated into histone H1 bands and to
determine relative kinase activities.
Radiation-induced mitotic delay assay
Replicate 35 mm dishes were seeded with 46104 cells.
Two days later, appropriate dishes were exposed to 1.5 Gy
from a 60Co gamma source, and dishes were ®xed at
various time points by inversion over methanol vapour.
After staining for 10 min with 1 mg/ml Hoechst 33258, the
frequency of cells with mitotic ®gures was determined by
¯uorescence microscopy. 2000 cells were scored for each
time point.
Cell cycle synchronization, irradiation and ¯ow cytometry
Replicate 60 mm dishes were seeded with 26105 cells. Two
days later, medium supplemented with 2 mM thymidine
was added. Fourteen hours later, medium was replaced
with thymidine-free medium. After 9 h, medium was
replaced with fresh medium containing 2 mM thymidine.
Cells were released from the second thymidine block 12 h
later by changing to fresh medium. 5.8 h post-release, cells
were exposed to 1.5 Gy from a 60Co gamma source and
harvested for ¯ow cytometry as described below for the
mitotic spindle checkpoint assay. At least 16104 cells were
analysed on a Becton-Dickinson FACScan using BectonDickinson CellFIT anlaysis software.
Premature mitosis assay
25 cm2 ¯asks were seeded with 26105 cells. The next day,
the medium was replaced with 4 ml of medium containing
2.5 mM hydroxyurea (Sigma). Seven hours later, 1 ml of
medium, containing 2.5 mM hydroxyurea with or without
25 mM (56) ca€eine and 0.5 mg/ml (56) nocodazole, was
added to each ¯ask. After an additional 8 h, cells were
trypsinized, swollen for 10 min in 75 m M KCl, ®xed by
dropwise addition of 75% methanol/25% acetic acid, and
dropped onto slides. The cell spreads were stained for
10 min with 1 mg/ml Hoechst 33258 and examined by
¯uorescence microscopy. Premature mitotic cells were
identi®ed by their characteristic `pulverized' chromatin
morphology and by breakdown of the nuclear envelope
(Schlegel and Pardee, 1986). 3000 ± 5000 cells were scored
for each treatment group.
Mitotic spindle checkpoint assay
25 cm2 ¯asks were seeded with 26105 cells. The next day, the
medium was replaced with medium containing 0 or 50 ng/ml
nocodazole. At the appropriate times, cells were prepared for
¯ow cytometry by harvesting in trypsin and freezing in 100 ml
of 250 mM sucrose/40 mM sodium citrate/5% DMSO per
26105 cells. Before analysis, 26105 cells from each sample
were washed in PBS/1% BSA, and incubated in 150 ml of
PBS/0.5% Triton X-100/0.5 mM EDTA/100 mg/ml propidium iodide for 20 min, before adding 40 ml of 2 M HCl for
30 min and then adding 150 ml of 1 M Tris base. Five
thousand cells per sample were analysed on an Ortho
Cyto¯uorograf system 2151, using the Cytomation Cicero
program to compile ¯uorescence histograms.
Acknowledgements
We thank Amy Imrich, Dr Lester Kobzik and Glenn
Belinsky for technical assistance. We thank Drs A Dusty
Miller and Denise Galloway for the gift of pLXSN plasmid
and PA317 cells, Dr Helen Piwnica-Worms for the gift of
Cdc25C antibody and Dr Melanie Cobb for the gift of
ERK692 antiserum.
References
Agarwal ML, Agarwal A, Taylor WR and Stark GR. (1995).
Proc. Natl. Acad. Sci. USA, 92, 8493 ± 8497.
Bikel I, Montano X, Agha ME, Brown M, McCormack M,
Boltax J and Livingston DM. (1987). Cell, 48, 321 ± 330.
Boulton TG and Cobb MH. (1991). Cell Regulation, 2, 357 ±
371.
Brown M, Garvik B, Hartwell L, Kadyk L, Seeley T and
Weinert T. (1991). Cold Spring Harbor Symposia on
Quantitative Biology, 61, 359 ± 365.
Buckley MF, Sweeney KJ, Hamiltion JA, Sini RL, Manning
DL, Nicholson RI, deFazio A, Watts CK, Musgrove EA
and Sutherland RL. (1993). Oncogene, 8, 2127 ± 2133.
Busse PM, Bose SK, Jones RW and Tolmach LJ. (1977).
Radiation Res., 71, 666 ± 677.
Busse PM, Bose SK, Jones RW and Tolmach LJ. (1978).
Radiation Res., 76, 292 ± 307.
Chang TH and Schlegel R. (1996). J. Cell. Biochem., 60,
161 ± 172.
Choi YW, Lee IC and Ross SR. (1988). Mol. Cell. Biol., 8,
3382 ± 3390.
Cross SM, Sanchez CA, Morgan CA, Schimke MK, Ramel
S, Idzerda RL, Raskind WH and Reid BJ. (1995). Science,
267, 1353 ± 1356.
DeCaprio JA, Ludlow JW, Figge J, Shew JY, Huang CM,
Lee WH, Marsilio E, Paucha E and Livingston DM.
(1988). Cell, 54, 275 ± 283.
Donahue LM and Stein GH. (1988). Oncogene, 3, 221 ± 224.
Enoch T and Nurse P. (1990). Cell, 60, 665 ± 673.
Fan S, Smith ML, Rivet DJ, Duba D, Zhan Q, Kohn KW,
Fornace AJ Jr and O'Connor PM. (1995). Cancer Res., 55,
1649 ± 1654.
Fanning E and Knippers R. (1992). Annu. Rev. Biochem., 61,
55 ± 85.
Fukasawa K, Choi T, Kuriyama R, Rulong S and Vande
Woude GF. (1996). Science, 271, 1744 ± 1747.
Gautier J, Solomon MJ, Booher RN, Bazan JF and
Kirschner MW. (1991). Cell, 67, 197 ± 211.
Gould KL and Nurse P. (1989). Nature, 342, 39 ± 45.
Harlow E, Crawford LV, Pim DC and Williamson NM.
(1981). J. Virol., 39, 861 ± 869.
Heald R, McLoughlin M and McKeon F. (1993). Cell, 74,
463 ± 474.
Jat PS and Sharp PA. (1989). Mol. Cell. Biol., 9, 1672 ± 1681.
Johnson RT and Rao PN. (1970). Nature, 226, 717 ± 722.
Kaufmann WK, Levedakou EN, Grady HL, Paules RS and
Stein GH. (1995). Cancer Res., 55, 7 ± 11.
Keyomarsi K and Pardee AB. (1993). Proc. Natl. Acad. Sci.
USA, 90, 1112 ± 1116.
Koprowski H, Ponten JA, Jensen F, Ravdin RG, Moorhead
P and Saksela E. (1962). J. Cell. Comp. Physiol., 59, 281 ±
292.
Kung AL, Sherwood SW and Schimke RT. (1990). Proc.
Natl. Acad. Sci. USA, 87, 9553 ± 9557.
Laemmli UK. (1970). Nature, 227, 680 ± 685.
Lan J, Fogleman D and Lehman JM. (1989). Cytometry,
10, 205 ± 213.
Disregulation of mitosis by SV40 T antigen
T Hung-Tse Chang et al
Lane DP and Crawford LV. (1979). Nature, 278, 261 ± 263.
Lau CC and Pardee AB. (1982). Proc. Natl. Acad. Sci. USA,
79, 2942 ± 2946.
Lee MS, Ogg S, Xu M, Parker LL, Donoghue DJ, Maller JL
and Piwnica-Worms H. (1992). Mol. Biol. Cell., 3, 73 ± 84.
Levine DS, Sanchez CA, Rabinovitch PS and Reid BJ.
(1991). Proc. Natl. Acad. Sci. USA, 88, 6427 ± 6431.
Liebmann J, Cook JA, Lipschultz C, Teague D, Fisher J and
Mitchell JB. (1994). Cancer Chemotherapy & Pharmacology, 33, 331 ± 339.
Lin JY and Simmons DT. (1991a). J. Virol., 65, 2066 ± 2072.
Lin JY and Simmons DT. (1991b). J. Virol., 65, 6447 ± 6453.
Linzer DI and Levine AJ. (1979). Cell, 17, 43 ± 52.
Lopes NM, Adams EG, Pitts TW and Bhuyan BK. (1993).
Cancer Chemotherapy & Pharmacology, 32, 235 ± 242.
Lundgren K, Walworth N, Booher R, Dembski M,
Kirschner M and Beach D. (1991). Cell, 64, 1111 ± 1122.
Meikrantz W and Schlegel R. (1995). J. Cell. Biochem., 58,
160 ± 174.
Mignotte B, Larcher JC, Zheng DQ, Esnault C, Coulaud D
and Feunteun J. (1990). Oncogene, 5, 1529 ± 1533.
Miller AD and Rosman GJ. (1989). Biotechniques, 7, 980 ±
982, 984 ± 986, 989 ± 990.
Moreno S and Nurse P. (1994). Nature, 360, 236 ± 242.
Murray A. (1995). Cell, 81, 149 ± 152.
Nichols WW, Murphy DG, Cristofalo VJ, Toji LH, Greene
AE and Dwight SA. (1977). Science, 196, 60 ± 63.
Nigg EA. (1993). Curr. Opinion Cell Biol., 5, 187 ± 193.
Norbury C and Nurse P. (1992). Annu. Rev. Biochem., 61,
441 ± 470.
Oren M, Maltzman W and Levine AJ. (1981). Mol. Cell.
Biol., 1, 101 ± 110.
Oshima J, Steinmann KE, Campisi J and Schlegel R. (1993).
Oncogene, 8, 2987 ± 2993.
Paules RS, Levedakou EN, Wilson SJ, Innes CL, Rhodes N,
Tlsty TD, Galloway DA, Donehower LA, Tainsky MA
and Kaufmann WK. (1995). Cancer Res., 55, 1763 ± 1773.
Peterson GL. (1977). Anal. Biochem., 83, 346 ± 356.
Pines J and Hunter T. (1989). Cell, 58, 833 ± 846.
Pines J and Hunter T. (1990). Nature, 346, 760 ± 763.
Powell SN, DeFrank JS, Connell P, Eogan M, Pre€er F,
Dombkowski D, Tang W and Friend S. (1995). Cancer
Res., 55, 1643 ± 1648.
Quartin RS, Cole CN, Pipas JM and Levine AJ. (1994). J.
Virol., 68, 1334 ± 1341.
Radna RL, Caton Y, Jha KK, Kaplan P, Li G, Traganos F
and Ozer HL. (1989). Mol. Cell. Biol., 9, 3093 ± 3096.
Ray FA, Peabody DS, Cooper JL, Cram LS and Kraemer
PM. (1990). J. Cell. Biochem., 42, 13 ± 21.
Ray FA, Meyne J and Kraemer PM. (1992). Mutation Res.,
284, 265 ± 273.
Ray FA and Kraemer PM. (1993). Carcinogenesis, 14, 1511 ±
1516.
Rice RH, Steinmann KE, deGra€enried LA, Qin Q, Taylor
N and Schlegel R. (1993). Mol. Biol. Cell, 4, 185 ± 194.
Roberts JR, Allison DC, Donehower RC and Rowinsky EK.
(1990). Cancer Res., 50, 710 ± 716.
Rowley R, Hudson J and Young PG. (1992). Nature, 356,
353 ± 355.
Russell KJ, Wiens LW, Demers GW, Galloway DA, Plon SE
and Groudine M. (1995). Cancer Res., 55, 1639 ± 1642.
Russell P and Nurse P. (1986). Cell, 45, 145 ± 153.
Russell P and Nurse P. (1987). Cell, 49, 559 ± 567.
Schimke RT, Kung AL, Rush DF and Sherwood SW. (1991).
Cold Spring Harbor Symposia on Quantitative Biology, 56,
417 ± 425.
Schlegel R and Pardee AB. (1986). Science, 232, 1264 ± 1266.
Schlegel R, Belinsky GS and Harris MO. (1990). Cell Growth
Di€er., 1, 171 ± 178.
Sherr CJ. (1993). Cell, 73, 1059 ± 1065.
Solomon E, Borrow J and Goddard AD. (1991). Science,
254, 1153 ± 1160.
Stein GH. (1985). J. Cell. Physiol., 125, 36 ± 44.
Steinmann KE, Belinsky GS, Lee D and Schlegel R. (1991).
Proc. Natl. Acad. Sci. USA, 88, 6843 ± 6847.
Steinmann KE, Pei XF, Stoppler H, Schlegel R and Schlegel
R. (1994). Oncogene, 9, 387 ± 394.
Stewart N and Bacchetti S. (1991). Virology, 180, 49 ± 57.
Stewart N, Hicks GG, Paraskevas F and Mowat M. (1995).
Oncogene, 10, 109 ± 115.
Strausfeld U, Labbe JC, Fesquet D, Cavadore JC, Picard A,
Sadhu K, Russell P and DoreÂe M. (1991). Nature, 351,
242 ± 245.
Tack LC, Cartwright CA, Wright JH, Eckhart W, Peden
KW, Srinivasan A and Pipas JM. (1989). J. Virol., 63,
3362 ± 3367.
Tam SW and Schlegel R. (1992). Cell Growth Di€er., 3, 811 ±
817.
Tam SW, Belinsky GS and Schlegel R. (1995). J. Cell.
Biochem., 59, 339 ± 349.
Walworth NC and Bernards R. (1996). Science, 271, 353 ±
356.
Weinert TA and Hartwell LH. (1988). Science, 241, 317 ±
322.
Weinert TA and Hartwell LH. (1990). Mol. Cell. Biol., 10,
6554 ± 6564.
Weinert TA, Kiser GL and Hartwell LH. (1994). Genes Dev.,
8, 652 ± 665.
White AE, Livanos EM and Tlsty TD. (1994). Genes Dev., 8,
666 ± 677.
Wolman SR, Hirschhorn K and Todaro GJ. (1964).
Cytogenetics, 3, 45 ± 61.
Wright WE, Pereira-Smith OM and Shay JW. (1989). Mol.
Cell. Biol., 9, 3088 ± 3092.
Yamashita K, Yasuda H, Pines J, Yasumoto K, Nishitani H,
Ohtsubo M, Hunter T, Sugimura T and Nishimoto T.
(1990). EMBO J., 9, 4331 ± 4338.
Zambetti GP and Levine AJ. (1993). FASEB J., 7, 855 ± 865.
Zerler B, Roberts RJ, Mathews MB and Moran E. (1987).
Mol. Cell. Biol., 7, 821 ± 829.
Zhu JY, Abate M, Rice PW and Cole CN. (1991). J. Virol.,
65, 6872 ± 6880.
Zhu J, Rice PW, Gorsch L, Abate M and Cole CN. (1992). J.
Virol., 66, 2780 ± 2791.
2393