Faculty of Pharmacy (B.B.H., M.E.A., G.-L.C., N.B.) and Department
of Zoology (E.H.), University of Manitoba, Winnipeg, Manitoba, Canada;
Department of Biology, University of Winnipeg, Winnipeg, Manitoba,
Canada (E.H.B.); Pathology Section, National Heart, Lung and Blood
Institute, National Institutes of Health, Bethesda, Maryland (V.J.F.,
Z.-X.Y); and Department of Pharmacology, University of Pittsburgh
School of Medicine and Cancer Institute, Pittsburgh, Pennsylvania
(J.C.Y.)
The bisdioxopiperazines, including dexrazoxane (ICRF-187), are
catalytic or noncleavable complex-forming inhibitors of DNA topoisomerase II that do not produce DNA strand breaks. In this study
we show that dexrazoxane inhibits the division of Chinese hamster ovary
(CHO) cells resulting in marked increases in cell size (up to 80 µm
in diameter), volume (up to 150-fold greater), and ploidy (as high as
32N). This last result indicates that the dexrazoxane-induced DNA
reduplication was restricted to once per cell cycle. Kinetic analysis
of the flow cytometry data indicated that the conversion between
successively higher ploidy levels was progressively slowed at longer
times of exposure to dexrazoxane. Both the protein and DNA content of
dexrazoxane-treated CHO cells increased linearly over time in the same
proportion. Light and electron microscopic studies of
dexrazoxane-treated cells showed ring-like multilobulated nuclei.
Immunohistochemical staining of dexrazoxane-treated cells showed that
F-actin and acetylated 
-tubulin were present in large, highly
organized networks. Immunohistochemical staining of the
dexrazoxane-treated CHO cells also showed that the topoisomerase II
colocalized with the DNA of the multilobulated nuclei. Staining of
-tubulin revealed that the dexrazoxane-treated cells contained
multiple centrosomes, indicating that dexrazoxane prevents cytokinesis
but not centrosome reduplication. It is concluded that dexrazoxane
inhibits CHO cytokinesis in cells by virtue of its ability to inhibit
topoisomerase II.
 |
Introduction |
Topoisomerase
II (EC 5.99.1.3) alters DNA topology by catalyzing the passing of an
intact DNA double helix through a transient double-stranded break made
in a second helix (Corbett and Osheroff, 1993
). Topoisomerase II has a
critical role in DNA metabolism that includes replication,
transcription, and recombination and is required for the separation of
chromosomes during mitosis.
The bisdioxopiperazine dexrazoxane (ICRF-187, Zinecard) and its analogs
(ICRF-159 (razoxane), ICRF-154, and ICRF-193) are strong catalytic
inhibitors of mammalian DNA topoisomerase II (Ishida et al., 1991;
Hasinoff et al., 1995) that inhibit this enzyme without inducing DNA
strand breaks. The bisdioxopiperazines have been proposed to act by
trapping the enzyme in the form of a closed ATP-modulated protein clamp
(Roca et al., 1994), thus preventing the formation or stabilization of
cleavable complexes. Dexrazoxane may promote an energy-dependent
inappropriate binding of topoisomerase II to DNA after the resealing
step. The inability of bisdioxopiperazines to induce DNA strand breaks
is in contrast to such cleavable complex-forming antitumor drugs as the
anthracyclines doxorubicin and daunorubicin, the epipodophyllotoxins
etoposide and teniposide, and amsacrine. These drugs are thought to be
cytotoxic by virtue of their ability to stabilize a cellular toxic
covalent topoisomerase II-DNA intermediate (the cleavable complex) and are called topoisomerase II poisons (Corbett and Osheroff, 1993). The
characterization and the sequencing of the topoisomerase II from
three dexrazoxane-resistant cell lines (Hasinoff et al., 1997; Yalowich
et al., 1998; Wessel et al., 1999) have identified functional mutations
in the amino terminal region, specifically at the clamp part of the
dimer interface of the enzyme (Yalowich et al., 1998) and at the ATP
binding site (Wessel et al., 1999), suggesting that dexrazoxane binds
in this interface region. A recent report has shown that overexpression
of human topoisomerase II in yeast results in sensitization to
dexrazoxane (Jensen et al., 2000). The pattern of cell killing by the
bisdioxopiperazines was consistent with a poisoning of topoisomerase II
by a novel mechanism involving the accumulation of closed clamp
conformations trapped on DNA that interfered with DNA transcription or metabolism.
Dexrazoxane is the (+)-(S)-enantiomer of racemic ICRF-159
(razoxane), which was originally developed as an antitumor agent. However, dexrazoxane now is clinically used to reduce
doxorubicin-induced cardiotoxicity (Hasinoff, 1998; Hasinoff et al.,
1998). Under physiological conditions dexrazoxane undergoes a slow
ring-opening hydrolysis to ADR-925 (Hasinoff, 1998; Hasinoff et al.,
1998), an analog of EDTA. Dexrazoxane likely exerts its
cardioprotective effects through its rings-opened hydrolysis product
ADR-925 by virtue of its ability to strongly chelate free iron, or to
quickly and efficiently remove iron from its complex with doxorubicin (Hasinoff, 1998; Hasinoff et al., 1998), thus reducing
doxorubicin-induced iron-based oxygen free radical damage. The cell
growth inhibitory properties of daunorubicin and etoposide are
antagonized by dexrazoxane (Sehested et al., 1993). Our previous study
also showed that dexrazoxane can antagonize doxorubicin- and
daunorubicin-mediated cytotoxicity (Hasinoff et al., 1996). Dexrazoxane
probably acts by preventing the formation of or the destabilizing of
the covalent cleavable complex as it does for etoposide (Hasinoff et
al., 1997). Because dexrazoxane is used clinically in combination with
doxorubicin in the treatment of breast cancer, it is important that the
effects of dexrazoxane on cell growth be characterized. In this study we report on the ability of dexrazoxane to inhibit the division of
Chinese hamster ovary (CHO) cells and induce an extensive endopolyploidy.
| |
Materials and Methods |
Drugs and Chemicals.
Dexrazoxane (a gift from Pharmacia & Upjohn, Columbus, OH) was freshly prepared in media directly before use
to avoid hydrolysis (Hasinoff, 1998; Hasinoff et al., 1998). PBS
(Na+ 154; Cl
141;
K+ 4;
PO43 10 mM) was from Life
Technologies Inc. (Burlington, Canada). Other drugs and chemicals not
listed were obtained from Sigma-Aldrich (Oakville, Canada). Cellular
DNA and protein determinations were done on 1% (w/v) SDS-lysed
cells using a fluorescent Hoechst 33258 dye assay (Cesarone et al.,
1979) and a 96-well plate spectrophotometric Bradford assay (Bradford,
1976), with appropriate calf thymus DNA and BSA as standards, respectively.
Cells and Cell Culture.
CHO cells (type AA8, CRL-1859),
obtained from the American Type Culture Collection (Rockville, MD),
were, as indicated, grown either as adherent cultures on plastic or as
suspension cultures in spinner flasks in Dulbecco's -modified
Eagle's medium (Life Technologies Inc.) containing 20 mM HEPES
(Sigma), 100 units/ml penicillin G (Life Technologies Inc.), 100 µg/ml streptomycin (Life Technologies Inc.), 10% (v/v) fetal bovine
serum (Life Technologies Inc.) in an humidified atmosphere of 5%
CO2 and 95% (v/v) air at 37°C (pH 7.1). Our
highly dexrazoxane-resistant DZR cell line (1500-fold), derived from
the CHO parent line, has been described previously (Hasinoff et al.,
1997; Yalowich et al., 1998). The spectrophotometric multiwell plate
MTT assay, which measures the ability of the cells to reduce MTT, has
also been described (Hasinoff et al., 1995). The ability of the cells
to exclude trypan blue was used to assess cell viability. Cells were
counted on a model ZF Coulter counter (Coulter
Electronics, Hialeah, FL).
Cell Sizing Analysis.
Changes in cell size distribution that
occurred after dexrazoxane exposure were determined by morphometric
analysis from digitized microscope images of the cells obtained using
SigmaScan (Jandel Scientific, San Rafael, CA) to obtain the image
area A from which the diameter d was calculated
from d = 2
. The
cell counts were expressed in the form of a distribution as a function
of cell volume.
Flow Cytometry.
Changes in cell cycle progression and DNA
ploidy were determined with and without treatment with dexrazoxane.
Approximately 8 × 106 cells in PBS were
fixed as a single cell suspension in 70% (v/v) cold ethanol overnight
at 
20°C. Samples were subsequently washed with PBS, resuspended in
a 0.1% (v/v) Triton X-100 solution containing 0.02 mg/ml propidium
iodide and 0.2 mg/ml DNase-free RNase A, incubated for 15 min at
37°C, and stored on ice for a maximum of a couple of hours before
analysis. An EPICS V multiparameter flow cytometer (Coulter
Electronics) with an argon laser (excitation 488 nm, and emission 610 nm) was used in the analysis.
Microscopy.
Attached cells were viewed with differential
interference contrast and/or epifluorescence optics on a Zeiss Photo II
microscope, and images were captured with a Sony 3-chip color
charge-coupled device camera using the Northern Eclipse imaging system
(Empix Imaging, Mississauga, Canada). Immunofluorescence preparations were examined using the appropriate epifluorescence filter sets for
fluorescein, rhodamine, and Hoechst fluorophores. For F-actin staining,
cells grown on cover slips were fixed in 4 g/100 ml paraformaldehyde in
a microfilament stabilizing buffer (60 mM PIPES, 35 mM HEPES, 10 mM
EGTA, 2 mM MgCl2, pH 6.9). They were washed three
times (10 min each) in PBS containing 0.2% (v/v) Triton X-100.
Subsequently, they were stained with rhodamine-labeled phalloidin (5 µg/ml, from D. Wieland, University of Heidelberg) for 1 h,
washed in PBS, and then stained in Hoechst 33258 (5 µg/ml) in PBS for
30 min. After a washing in PBS they were mounted in an anti-fade
mounting medium (Molecular Probes, Eugene, OR) and imaged. For
visualizing microtubules, an antibody against acetylated -tubulin
(C3B9 from K. Gull, University of Manchester) was used at a 1:5
dilution. Cells were fixed as described above in a microtubule stabilizing buffer for 1 h, washed in PBS with 0.2% (v/v) Triton X-100 (3 times), and blocked in PBS-Triton X-100 with 0.5 g/100 ml BSA
for 30 min. Incubation with the primary antibody for 1 h was
followed by washings (five times, 10 min each) and a 1-h incubation
with a goat anti-mouse secondary antibody labeled with fluorescein
isothiocyanate (FITC; Sigma) and diluted 1:128. It was then washed and
mounted as described above. The staining of -tubulin to show
centrosomes was performed as described for -tubulin using a mouse
monoclonal antibody (TU-30 from P. Draber, Czech Republic). For
electron microscopy, cells grown in suspension in the presence or
absence of dexrazoxane were fixed by resuspending the cell pellet in
0.5 ml of 2.5 g/100 ml glutaraldehyde in PBS. They were then postfixed
with 1 g/100 ml osmic acid in 0.1 M phosphate buffer (pH 7.3),
dehydrated with graded ethanols and propylene oxide, and embedded in
epoxy resin. Ultrathin sections stained with uranyl acetate and lead
citrate were viewed on a 1200EX electron microscope (JEOL, London, UK).
For laser scanning confocal fluorescence microscopy, cells were grown
on coverslips in the presence or absence of dexrazoxane and were fixed
in 10% (v/v) formalin in PBS. After treatment with RNase A and
staining with propidium iodide, the cells were examined with a confocal
microscope (Leica TCS, Deerfield, IL) equipped with an argon-krypton
laser (excitation at 568 nm).
Induction of apoptosis or loss of cell viability after dexrazoxane
treatment were evaluated by epifluorescence microscopy using an XF-19
filter set (Omega Optical, Brattleboro, VT) after double staining with
ethidium bromide and acridine orange (Mercille and Massie, 1994). The
charged ethidium bromide is taken up only by necrotic cells possessing
a damaged membrane, and this results in orange fluorescence upon
binding of the dye to DNA. Neutral acridine orange is taken up by cells
that may or may not possess an intact membrane, which results in green
fluorescence upon binding to DNA. Positive apoptotic controls for
comparison were obtained by exposing attached CHO cells to 10 µM
camptothecin for times up to 48 h. The percentage of cells that
were apoptotic ranged from 40% at 24 h to 85% at 48 h.
Nonviable controls for comparison were obtained by exposing the cells
overnight to 70% (v/v) ethanol (Mercille and Massie, 1994).
Immunofluorescence staining of dexrazoxane-treated cells for
topoisomerase II was carried out essentially as described (Chaly and
Brown, 1996). CHO cells were seeded onto coverslips and allowed to
attach for 24 h, and they were then treated with 100 µM
dexrazoxane for 72 h and fixed. CHO cells not treated with
dexrazoxane were seeded onto coverslips 1 day before fixation. All
steps were carried out at room temperature and samples were washed
(2 × 5 min) in PBS between each step unless otherwise noted.
Cells were fixed in freshly prepared 3% (v/v) paraformaldehyde in PBS
for 5 min and permeabilized in 0.2% (v/v) Triton X-100 in PBS for 5 min. Samples were then incubated for 45 min in PBS-diluted (1:10)
rabbit polyclonal antibody to human topoisomerase II (Topogen,
Columbus, OH). Samples were subsequently washed in PBS as before,
except for a prolonged second wash (2 h). Samples were then incubated in the dark for 45 min in PBS-diluted (1:800) goat anti-rabbit IgG FITC
conjugate (Sigma). Finally, cells were counterstained with
4',6-diamidino-2-phenylindole (0.5 µg/ml in PBS for 10 min), washed,
mounted, and imaged on an epifluorescence microscope with the
appropriate filter pairs for 4',6-diamidino-2-phenylindole and FITC.
Data are cited as the mean ± S.E., unless otherwise indicated.
Nonlinear least-squares curve fitting and graphing was done using
SigmaPlot 5 (SPSS, Chicago, IL). The exact solutions to the first order
differential equations were obtained using the symbolic math program
Maple V (release 5.1, Waterloo Maple, Waterloo, Canada).
| |
Results |
Effect of Dexrazoxane on Cell Growth and Morphology of CHO
Cells.
The results in Fig. 1a
compare the effect of dexrazoxane on the CHO and the
dexrazoxane-resistant DZR cell lines as measured by an MTT assay (72-h
growth in the presence of a single dose). The CHO and DZR cell lines
had IC50 values of 5.2 ± 0.4 µM and 2200 ± 200 µM, respectively. Daily replacement of the medium
with fresh dexrazoxane resulted in an essentially unchanged CHO
IC50 of 5.0 ± 0.4 µM (data not shown).
Nearly identical results were obtained in Fig. 1b, where
dexrazoxane-mediated growth inhibition was determined by cell counting
(IC50 values of 3.5 ± 0.1 µM and 1600 ± 100 µM for CHO and DZR cells, respectively). At 20 to
1000 µM dexrazoxane, MTT absorbance from CHO cells leveled off at 0.2 absorbance units, which is 28% of the absorbance observed for untreated control cells (Fig. 1a). By comparison, over this dexrazoxane concentration range, CHO cell number was reduced to approximately 2%
of the number of cells in untreated control plates (Fig. 1b). Because
2.5 × 104 CHO cells were plated initially,
these results indicate that cell growth was completely inhibited at 20 µM dexrazoxane (Fig. 1b), yet these cells have an ability to reduce
MTT to tetrazolium blue out of proportion to their numbers as indicated
by the 490-nm absorbance at a level of 28% of that of untreated
controls (Fig. 1a). The results shown in Fig. 1c demonstrate that CHO
cell growth, as determined by cell number, is quickly and completely
halted upon the addition of 100 µM dexrazoxane, compared with
resistant DZR cells, which grew exponentially at nearly the same rate
as CHO cells in the absence of dexrazoxane. Under these conditions, where media was replaced daily with fresh media containing 100 µM
dexrazoxane, the dexrazoxane-treated CHO cells maintained a relatively
high viability (Fig. 1d) (82% at 120 h) despite the fact that
cell division was completely inhibited. Together these results indicate
that the dexrazoxane-treated CHO cells that have ceased dividing are
viable and have a greater ability to reduce MTT, compared with the same
number of untreated CHO cells. In addition, initial microscopic
examination revealed that dexrazoxane-treated CHO cells were increased
in size, thereby providing a potential explanation for the enhanced
mitochondrial reduction of MTT (Fig. 1a). Quantitative cell sizing
experiments were then carried out.
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 1.
Effect of dexrazoxane on the growth of attached CHO
cells. a, MTT growth inhibition assay after growth for 72 h for
CHO ( ) and DZR ( ) cells treated with a single initial dose of
dexrazoxane. Error bars shown are S.D. b, cell counting growth
inhibition assay at 72 h growth with a single initial dose of
dexrazoxane for attached CHO () and DZR () cells. c, the relative
growth in cell number of suspension cultures of CHO ( ) and DZR ()
cells in the presence of 100 µM dexrazoxane (in which the media was
replaced daily with fresh media containing drug), and suspension
cultures of CHO () cells grown in the absence of dexrazoxane. For
clarity the relative cell number data for the DZR cells have been
normalized to an initial value of 2 to offset this data from the CHO
cell data. d, viability of dexrazoxane-treated CHO cells (measured by a
trypan blue dye exclusion assay) grown as described in c.
|
|
Cell sizing was performed on dexrazoxane-exposed CHO cells grown in
suspension culture (Fig. 2a) for various
times. Although untreated CHO cells displayed a narrow distribution of
sizes, the dexrazoxane-exposed cells displayed broadened peak maxima that shifted to higher cell volumes at longer times of exposure to
dexrazoxane. The mean CHO cell volumes increased linearly
(r2 = 0.99) with time after
dexrazoxane treatment (Fig. 2b). The DZR cells, however, did not
display any increase in mean cell volume upon dexrazoxane treatment
(Fig. 2b).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2.
Changes in cell size and volume of
dexrazoxane-treated cells. a, sizing of CHO cells grown in suspension
culture after 0 (), 12 (), 24 ( ), 48 ( ), and 120 h
( ) of exposure to dexrazoxane. The cells were suspended in fresh
medium containing 100 µM dexrazoxane at 0 h and 24 h.
Thereafter, the cells were resuspended daily in fresh medium without
dexrazoxane. Data obtained at 72 and 96 h are not plotted for
clarity. The data were normalized to the same total cell count at each
time. b, increase in mean cell volume with time for CHO () and DZR
( ) cells as calculated from the data in a.
|
|
To further characterize the effect of dexrazoxane on CHO cell growth,
the protein and DNA content of dexrazoxane-treated cells were also
determined. As shown in Fig. 3, a and b,
CHO cells grown in suspension culture that were exposed to 100 µM
dexrazoxane, at 0 and 24 h with media replacement, showed
approximately a 4-fold increase in both average cell protein (Fig. 3a)
and DNA (Fig. 3b) content by 120 h. By comparison, the DZR cells
showed no significant changes upon exposure to dexrazoxane
(r2 = 0.04). Both the mean protein and
DNA content per million cells also increased linearly with time
(r2 = 0.96 and 0.96, respectively).
Comparison of the slopes of the data obtained at different times showed
no significant increase in protein/DNA ratios (Fig. 3c) of CHO cells
exposed to dexrazoxane compared with controls (P > .2). These findings indicate that the dexrazoxane-induced increase in
the size of CHO cells does not result in unbalanced growth, at least as
far as protein and DNA content is concerned.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 3.
Changes in protein and DNA content and protein/DNA
ratios in CHO and DZR cells. Protein (a), DNA (b), and protein/DNA
ratios (c) in CHO () and DZR () cells grown in the presence of
100 µM dexrazoxane and control (untreated) CHO () cells. The cells
were grown for different times in suspension culture as indicated, and
the data shown are expressed as cell protein or DNA content per million
cells. The error bars are S.E.s from four replicate samples.
Dexrazoxane caused significant increases (P < .001) in
protein and DNA content compared with untreated control (0 h) cells at
all times as determined by an unpaired Student's t test.
, CHO + 0 µM dexrazoxane; , CHO + 100 µM dexrazoxane; ,
DZR + 100 µM dexrazoxane.
|
|
Cell cycle analysis using flow cytometry after staining with propidium
iodide was carried out on both CHO (Fig.
4) and DZR cells (data not shown) that
had been treated with 100 µM dexrazoxane for various periods of time.
The data obtained were expressed as the logarithm of the propidium
iodide fluorescence, rather than the more conventional linear
fluorescence measure, because cell populations of higher ploidy were
observed. Care was taken in these experiments to examine
microscopically the propidium iodide-stained cell suspensions for
aggregated cells that could result in falsely positive ploidy levels.
In all cases doublets were less than 1%, and triplets or larger were
not detectable. On a logarithmic DNA fluorescence scale there is an
equidistant separation between peaks that differ in ploidy by a
constant factor of 2. Such peaks correspond to cells with 2N, 4N, 8N,
16N, and 32N ploidy (where N is the haploid DNA content) and can be
easily distinguished in Fig. 4. Study of the back correlation of the cell size light scatter data (not shown) to DNA fluorescence showed that the high ploidy peaks were due to large cells. The results in Fig.
4 show that dexrazoxane treatment does not cause a cell cycle blockage
at G2/M, but rather that the CHO cells continue to cycle to high ploidy levels. Flow cytometry studies showed no
increase in ploidy level (data not shown) of the DZR cells similarly
treated with dexrazoxane. The percentages of CHO cells in each of 2N,
4N, 8N, 16N, and 32N ploidy levels are plotted as a function of time in
Fig. 5a. These percentages were
determined by measuring the total number of cells from one peak valley
to the next (Fig. 5a). Although this analysis does not take into account cells that were in S phase (approximately 9% of the total at
time 0), it should, nonetheless, give an approximate measure of the
dexrazoxane-treated CHO cells that were present in each ploidy level.
Given the complex nature of the cell cycle distribution curves at
longer times, a more sophisticated analysis taking into account what
was presumably up to four different S phases was not attempted. As
shown in Fig. 5a the 2N peak was observed to decrease smoothly with
time of exposure to dexrazoxane, whereas the 4N peak increased from its
initial value of 18%, leveled off, and then decreased. The 8N peak
went through a similar increase and then decreased, whereas the 16N and
the 32N peaks only increased. A preliminary analysis using nonlinear
least-squares fitting to a 2-parameter exponential decay equation
indicated that the decrease in the percentage of CHO cells in 2N with
time was closely approximated by an exponential decay process (fit
curve not shown). This result indicates that a first order process was
occurring in which the rate of loss of the 2N DNA was proportional to
the amount of 2N present at any time.
View larger version (34K):
[in this window]
[in a new window]
|
Fig. 4.
Cell cycle analysis of CHO cells grown for different
times in suspension culture in the presence of 100 µM dexrazoxane.
The cells were resuspended daily in fresh medium containing
dexrazoxane. After staining with propidium iodide, the cells were
analyzed by flow cytometry. A total of 10,000 cells were analyzed for
each time. Although the results shown were from one experiment, they
are representative of three separate experiments. The cell counts are
plotted on the vertical axis, and the integrated red fluorescence is
plotted on a logarithmic scale. The 2N, 4N, 8N, 16N, and 32N ploidy
levels, which differ by a factor of 2, are equally spaced on a
logarithmic scale and are indicated by the vertical arrows.
|
|
View larger version (27K):
[in this window]
[in a new window]
|
Fig. 5.
Ploidy levels in CHO cells treated with dexrazoxane.
a, percentage of CHO cells showing ploidy levels of 2N (), 4N (),
8N (), 16N ( ), and 32N () after being grown in suspension
culture in the presence of 100 µM dexrazoxane for different times.
The solid curves were calculated from a nonlinear least-squares fit of
the concatenated data to equations solved for the linear first order
differential equations of Scheme 1. The data were obtained from the
flow cytometry values shown in Fig. 4. b, relative amount of DNA
at different times of culture. Data were obtained as described in the
text. c, first order rate constants for the conversion of each ploidy
level to the next highest level.
|
|
Given that the 2N DNA decayed in an exponential process, it was decided
to attempt a fit to the kinetics of the data in Fig. 5a, assuming that
the interconversions to each of the other ploidy levels were also first
order processes as shown in Scheme 1:
|
(1)
|
The differential rate expressions for Scheme 1 are:
The rate constants are given by k2N,
k4N, k8N, and
k16N, and the percentage of each DNA
ploidy level is given by (2N), (4N), etc. The
exact solutions to the linear set of ordinary differential rate
expressions for the scheme above, with appropriate initial boundary
conditions, gave expressions yielding sums of exponential terms for
each of 2N, 4N, 8N, 16N,
and 32N DNA as a function of time. The whole concatenated
data set was simultaneously fit to all of these expressions by
nonlinear least-squares analysis to give the eight best-fit parameters
of Table 1. The curves in Fig. 5a were
calculated from the best-fit parameters and are seen to be a very good
fit to the data. The r2 for the fit
was 0.988 and the S.E. of the estimate was 2.2, both of which also
indicate the quality of the fit to the model of Scheme 1 was very good.
The rate constants k2N,
k4N, k8N, and k16N are plotted in Fig. 5c and show that
the interconversion from one ploidy level to the next highest level was
progressively slowed as the number of cell cycles increases. The rate
constant decreases by approximately one-half for interconversion from
one ploidy level to the next. All cell preparations were stained at the
same time under the same conditions, and the flow cytometry was carried
out with the same instrumental settings, thus, the total relative
amount of DNA per 10,000 dexrazoxane-treated CHO cells as a function of
time can be obtained from the percentage of DNA present in each ploidy
level from the data in Fig. 5a from the expression
DNArel = 2(2N) + 4(4N) + 8(8N) + 16(16N) + 32(32N). As shown in
Fig. 5b, DNArel increased linearly with time
(r2 = 0.98). This result is in
excellent agreement with the linearity displayed on the plot of DNA
content as a function of time (Fig. 3b), and which was measured by
directly determining cellular DNA content and thus, is additional
support for the model of Scheme 1.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Kinetic analysis of the interconversion of ploidy levels of
dexrazoxane-treated CHO cells
The best-fit rate constants and initial ploidy level parameters for the
conversions of the ploidy levels were obtained from nonlinear least
squares fitting of the cell cycle kinetic data of dexrazoxane-treated
CHO cells.
|
|
Light, Fluorescence, and Transmission Electron Microscopy of
Dexrazoxane-Treated Cells.
CHO cells that had been exposed to
dexrazoxane ceased to divide and continued to increase in size (Figs.
6-9). At 14 days, cells as large as 80 µm in diameter were observed, which corresponds to a 150-fold
increase in volume compared with untreated cells. The attached
dexrazoxane-treated CHO cells, as shown in the differential interference contrast micrographs (Fig. 6, c, d, and e), displayed a
variety of morphologies. DNA staining of the dexrazoxane-treated CHO
cells with the fluorescent Hoechst 33258 dye (Fig.
7, a, b, and c) showed a progressive
enlargement and multisegmentation of the nucleus. To determine whether
the internal organization of the dexrazoxane-treated CHO cells was
maintained, staining for F-actin (Fig. 7, d and e) and the microtubules
(Fig. 7f) was carried out. As shown in Fig. 7e, both the bundles of
F-actin filaments and the microtubules, although numerous, were well
defined and well organized, indicating that the growth of these cell
structures in the absence of cell division was not disrupted by
exposure to dexrazoxane.
View larger version (95K):
[in this window]
[in a new window]
|
Fig. 6.
Phase contrast photomicrographs of live CHO cells in
suspension after no exposure (a) or 144 h (b) of exposure to
dexrazoxane. The dexrazoxane-exposed attached CHO cells were treated
with fresh 100 µM dexrazoxane 24 h after attachment. At 48 h the culture medium was replaced with medium containing fresh 100 µM
dexrazoxane. The medium was replaced again at 96 h with medium
containing no dexrazoxane. After trypsinization the cells were
suspended in PBS containing trypan blue. Differential interference
contrast photomicrographs of live attached CHO cells after no exposure
(c), 72 h (d), or 144 h (e) exposure to dexrazoxane. The
treatment protocol was as described in b. The scales are the same in
all panels.
|
|
View larger version (66K):
[in this window]
[in a new window]
|
Fig. 7.
Epifluorescent photomicrographs of fixed attached
DNA-stained (with Hoechst 33258) CHO cells after no exposure (a),
72 h (b), or 144 h (c) exposure to 100 µM dexrazoxane. The
cells were treated as described in the legend to Fig. 5b. The scale is
the same in panels a, b, and c. Epifluorescent photomicrographs of
fixed attached CHO cells stained for F-actin after no exposure (d) and
96 h (e) of exposure to 100 µM dexrazoxane. Panel f shows a
fixed attached CHO cell stained for -tubulin that had been exposed
to 100 µM dexrazoxane for 96 h. The scales are the same in
panels d, e, and f.
|
|
To evaluate the effect of dexrazoxane on centrosome duplication,
attached CHO cells were stained for -tubulin with a mouse monoclonal
primary antibody and an FITC-conjugated secondary antibody (Fig.
8, a, b, and c). As shown in Fig. 8b,
dexrazoxane-treated CHO cells displayed multiple centrosomes. Double
staining for both DNA and centrosomes (Fig. 8c) showed that the
multilobulated nucleus formed a cap- or ring-like structure over the
multiple centrosomes. Confocal microscopy, followed by tridimensional
reconstruction using images obtained by Z axis scans of
dexrazoxane-treated CHO cells (Fig. 8, e and f) stained with propidium
iodide confirmed that multisegmented nuclei formed ring-like
structures.
View larger version (50K):
[in this window]
[in a new window]
|
Fig. 8.
Epifluorescent photomicrographs of fixed attached CHO
cells. Centrosome-stained (using anti--tubulin immunofluorescence)
CHO cells were imaged after no exposure (a) and 96 h (b and c) of
exposure to 100 µM dexrazoxane. The epifluorescent image in c is a
composite image showing a cell that has been stained for both DNA and
its centrosomes and imaged separately with filter sets for Hoechst
33258 and fluorescein, respectively. The centrosomes are located in a
pocket surrounded by DNA. The scales are the same in panels a, b, and
c. In the dexrazoxane-treated cells, multiple centrosomes can be seen
in the center of the cell. The cells were treated as described in the
legend to Fig. 6. The confocal epifluorescent photomicrographs in d, e,
and f show an attached fixed CHO cell stained for DNA with propidium
iodide, after no exposure (d) and after 96 h (e and f) of exposure
to dexrazoxane with the top view (e) and the side view (f) of the same
cell. The scales are the same in panels d, e, and f. Topoisomerase
II-stained attached CHO cells were imaged after they were either
untreated (g) or treated with 100 µM dexrazoxane for 72 h (h).
The same cells as in g and h, respectively, are shown counter-stained
for DNA in i and j.
|
|
Double staining for both DNA and topoisomerase II of CHO cells that
were untreated or were treated with 100 µM dexrazoxane for 72 h
(Fig. 8, g-i) showed that the DNA and topoisomerase II were
colocalized both in the untreated cells and generally quite evenly over
the multilobulated nuclei of the dexrazoxane-treated cells. However,
comparison of the images of Fig. 8, h and j, shows that the
dexrazoxane-treated cells show several regions of topoisomerase II
hyperfluorescence. Although this was not an uncommon feature of a large
number of cells that were examined, there are potential problems in
staining the much larger nuclei of the dexrazoxane-treated cells that
could give rise to this effect.
To further characterize the effect of dexrazoxane on CHO cell growth,
transmission electron microscopic studies were made of control and
dexrazoxane-treated cells (Fig. 9, a and
b). Multisegmented nuclei were clearly demonstrated in the
dexrazoxane-treated cells. The nuclear membranes in areas of
segmentation were closely apposed to each other, with little or no
intervening nuclear chromatin. Other cellular organelles, including the
mitochondria, were well defined and organized, again suggesting that
the inhibition of cell division did not disrupt orderly cell growth.
View larger version (130K):
[in this window]
[in a new window]
|
Fig. 9.
Transmission electron micrographs of CHO cells after
no exposure (a) and after 96 h of exposure (b) to 100 µM
dexrazoxane. The cells were grown in suspension and were resuspended
daily in fresh medium containing 100 µM dexrazoxane. The untreated
control cells (a) show normal morphology, with oval or kidney-shaped
single nuclei. The CHO cells treated with dexrazoxane (b) were markedly
increased in size compared with control cells. They have well developed
rough endoplasmic reticulum and nuclear segmentations. It should be
noted that, due to the plane of sectioning, some areas of segmentation
appear as isolated nuclei. The scale bar is 5 µm and is the same for
both a and b.
|
|
Because previous reports had indicated that dexrazoxane caused
apoptosis in human leukemic CEM cells (Khelifa and Beck, 1999; Morgan
et al., 2000) and K562 cells (Synold et al., 1998), several indicators
of apoptosis were looked for in dexrazoxane-treated cells. Apoptosis
did not occur, as indicated by the lack of the following criteria: 1)
apoptotic features in the cells stained with ethidium iodide and
acridine orange (data not shown); 2) subdiploid peaks (Fig. 4) or any
increase in the number of particles seen by light scattering in flow
cytometry experiments (data not shown); 3) electron micrograph (Fig. 9)
findings of densely aggregated chromatin in the nuclei or cytoplasmic
condensation (Fig. 9) and disrupted mitochondria.
| |
Discussion |
Some of the cell growth inhibitory properties of dexrazoxane and
the other bisdioxopiperazines have been characterized previously (Hellmann and Field, 1970; Edgar and Creighton, 1981; Traganos et al.,
1981), because these compounds were originally developed as antitumor
drugs. After 14 days of continuous treatment with dexrazoxane, we
observed some CHO cells in suspension that were as large as 80 µm in
diameter, which corresponds to a 150-fold increase in volume. The
results of Fig. 3c indicate that both DNA and protein were continuing
to be synthesized in near normal proportions in dexrazoxane-treated CHO
cells and thus, unbalanced growth did not occur. The fact that the
-tubulin and F-actin networks were highly developed and well
organized suggests that cellular processes were not affected by the
inhibition of cytokinesis. The fact that the topoisomerase II and
DNA were nearly evenly colocalized on all of the nuclear lobes of the
dexrazoxane-treated cells (Fig. 8, i and j) indicated that the newly
synthesized DNA of these lobes had topoisomerase II associated with
it. The CHO-derived dexrazoxane-resistant DZR cell line, which is
1500-fold resistant to dexrazoxane (Hasinoff et al., 1997) [due to a
mutated topoisomerase II with a threonine to isoleucine mutation at
amino acid 48 (Yalowich et al., 1998)], showed no change in volume,
protein content, DNA content, or ploidy upon treatment with
dexrazoxane. These results indicate that dexrazoxane was inducing all
of the effects on CHO cells seen in this study through its ability to
inhibit topoisomerase II. The 100 µM concentration of dexrazoxane
used in this study is in the pharmacologically relevant concentration
range. Dexrazoxane clinically given at a dose of 600 mg/m2 yields a peak plasma concentration of 340 µM with an elimination t1/2 of 4.2 ± 2.9 h (Hochster et al., 1992).
For reasons that are unknown, mature neutrophils normally display
multisegmented nuclei (Campbell et al., 1995). Treatment with
anticancer drugs and other agents such as retinoic acid can also
produce multisegmented nuclei (Olins et al., 1998). Some of the
morphological features (large size and large multisegmented nuclei) and
high DNA content (high ploidy) of dexrazoxane-treated CHO cells are
also similar to those that naturally occur in megakaryocytes (Levine et
al., 1982; Nagata et al., 1997) in a process called endomitosis or
endopolyploidy (Nagata et al., 1997; Baatout et al., 1998). The
endogenously and exogenously induced mechanisms by which
endoreduplication can occur have been recently reviewed (Grafi, 1998).
Thrombopoietin-induced polyploidization of bone marrow megakaryocytes
also results in formation of multiple centrosomes (Nagata et al., 1997)
as seen in the dexrazoxane-treated CHO cells (Fig. 8, b and c). Phorbol
ester-treated erythroleukemic K562 cells, which can differentiate
toward a megakaryocytic lineage, were shown to have down-regulated
their topoisomerase II mRNA by a post-transcriptional mechanism
(Loflin et al., 1996). It is tempting to speculate that topoisomerase
II down-regulation or inhibition may be a determinant for megakaryocyte differentiation.
The high levels of polyploidization seen (Fig. 4) indicate that at
least three, and possibly four, cycles of DNA reduplication may have
occurred without the cells undergoing division. The distinct ploidy
levels seen indicate that the DNA reduplication had to have been
completed before the next round occurred. These results indicate that
suppression of topoisomerase II catalytic DNA strand passing activity
does not result in cell cycle blockage. Similarly, it was shown that
ICRF-193 does not prevent initiation of SV40 DNA replication (Permana
et al., 1994). This suggests that, in CHO cells at least, there is no
topoisomerase II activity-based cell cycle checkpoint. Likewise, it was
concluded from microscopic studies on ICRF-193-treated HeLa cells that
topoisomerase II function is not necessary for exit from
G2 (Downes et al., 1994) and that there is a
catenation-sensitive checkpoint mechanism for exiting from
G2 that is distinct from the
G2-damage checkpoint. The effect of ICRF-193 on
the dynamic behavior of HeLa cells double stained for DNA and - and
-tubulin was studied by fluorescence microscopy (Haraguchi et al.,
1997). ICRF-193 was shown to block chromosome segregation but not exit
from metaphase, and, thus, it was concluded that there is a metaphase
checkpoint that is monitored by functional spindle assembly but not by
completion of chromosome segregation. Chromosome decondensation and
spindle disassembly proceeded in the presence of ICRF-193 but with
delayed timing (Haraguchi et al., 1997).
We had previously found that the incorporation of
[methyl-3H]thymidine into the
DNA of CHO cells exposed for 24 h to 100 µM dexrazoxane was, on
a per cell basis, 51% of that observed in untreated control cells
(Hasinoff et al., 1999). It has also been shown that incubation with
razoxane for 22 h has relatively little effect (22-30%
inhibition) on RNA synthesis (Creighton and Birnie, 1970).
Razoxane induces an increase in the size of BHK-21S (Stephens and
Creighton, 1974) and RPMI 8402 human leukemic cells (Ishida et al.,
1991), similar to what we observed for dexrazoxane-treated CHO cells
(Figs. 6 and 7). Razoxane has also been previously shown (Edgar and
Creighton, 1981; Traganos et al., 1981) to induce 8N ploidy in BHK 21S,
murine leukemia L1210, and Friend leukemia cells. Dexrazoxane treatment
of CEM cells results in enlarged cells with 8N ploidy and induction of
apoptosis (Khelifa and Beck, 1999; Morgan et al., 2000). Although these
ploidy levels were lower than what we observed (Fig. 4), it is possible
that cells with ploidies exceeding these values were not detected in
these studies. This could have been due to the shorter times that the cells were exposed to razoxane or the fact that the fluorescence from
the flow cytometer was recorded on a limited linear scale, which tends
to flatten and obscure peaks of high ploidy. Treatment of promonocytic
U937 cells with the dexrazoxane analog ICRF-193 was shown to cause an
unquantified increase in cell mass and to allow an accumulation of
G2 DNA (Perez et al., 1997). The unbalanced protein/DNA ratios observed in that study may have been a consequence of the differentiation induced by ICRF-193 in U937 cells. Microtubule inhibitors such as colchicine, colcemid, and nocodazole are also known
to induce polyploidization and produce large cells due to a lack of a
functional spindle, either through reduplication of DNA in the absence
of an intervening mitosis or a premature exit from mitosis (Baatout et
al., 1998).
The rate constants k2N,
k4N, etc. for successive conversions to
higher ploidy levels was decreased (Fig. 5c) by similar factors of
0.67, 0.60, and 0.40, respectively, which yields a mean factor of
0.56 ± 0.08. This value is one-half within the S.E. Because each
successively higher ploidy level contains twice as much DNA, it is
tempting to speculate that the cycling time to the next ploidy level is
directly proportional to the amount of DNA that has to be reduplicated.
This result may be due to a lengthened S phase in which DNA
reduplication is occurring. The dependence of the cellular content of
DNA on the duration of the cell cycle may also be related to a
lengthening in another phase (possibly a delay from metaphase to
anaphase or a delay due to a blocked telophase). The bisdioxopiperazine
ICRF-193 does not delay progression of synchronized CHO cells, either
from M to G1 and S phase, or from early S to M
phase, although it did in HeLa cells (Ishida et al., 1994; Iwai et al.,
1997). The difference between the two cells types was ascribed to the
"relaxed" mitotic control of CHO cells, which are known to continue
cell cycle progression (with serial doublings to 16N DNA) without cell
division in the presence of antimicrotubule agents (Kung et al., 1990).
The finding of multiple centrosomes after 96 h of exposure to
dexrazoxane indicated that multiple rounds of centrosome duplication occurred in the absence of cytokinesis. The centrosome duplication cycle in Drosophila embryos can, however, operate even when
DNA replication is blocked by aphidicolin (Raff and Glover, 1989). We
found no evidence for dexrazoxane-induced apoptosis in CHO cells. The
ability of dexrazoxane to induce apoptosis would seem to depend on the
particular cell type, because dexrazoxane induced apoptosis in CEM
(Khelifa and Beck, 1999; Morgan et al., 2000) and K562 (Synold et al.,
1998) cells. The reason for this difference is not known.
In conclusion, this study has shown that the topoisomerase II inhibitor
dexrazoxane inhibits CHO cell division but allows cells to increase in
size, protein, and DNA content without detectable disruption in
cellular organization. The DNA was reduplicated to yield distinct
ploidy levels as high as 32N, indicating that each round of DNA
reduplication was completed before the next was started. Because
dexrazoxane inhibits topoisomerase II, preventing DNA decatenation, our
results also indicate that there is no decatenation-sensitive checkpoint in CHO cells. The ability of dexrazoxane to inhibit cytokinesis of cells may be responsible for the reported antimetastatic activity of razoxane (Zwilling et al., 1981) due to generation of
larger less mobile cancer cells. A reinvestigation of the
antimetastatic and antitumor effects of the bisdioxopiperazines may be
warranted if these cytostatic mechanisms can be further elucidated.
Additionally, the fact that dexrazoxane was growth inhibitory through
the induction of endoreduplication rather than through apoptosis
suggests that dexrazoxane may also be useful in treating tumors that
have defects in their apoptosis-transduction machinery.
The assistance of Dr. Edward Rector in the flow cytometry is
gratefully acknowledged.
CHO, Chinese hamster ovary;
FITC, fluorescein
isothiocyanate;
MTT, 3-[4,5-dimethylthiazol-2-yl]-2,5-tetrazolium
bromide;
PIPES, piperazine-N,N'-bis(2-ethanesulfonic acid).
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
