Teratogenesis, Carcinogenesis, and Mutagenesis Supplement 1:171–186 (2003)

Chromosomal Aberrations Induced by

5-Azacytidine Combined With VP-16
(Etoposide) in CHO-K1 and XRS-5
Cell Lines

A.P.A. Guimarães,1 F.L. Dias,2 R.S. Cardoso,1 S.N. Kronka,3 and

E.T. Sakamoto-Hojo1,2n
1
Departamento de Genética, Faculdade de Medicina de Ribeirão Preto,
Universidade de São Paulo, SP, Brasil
2
Departamento de Biologia, Faculdade de Filosofia Ciências e Letras de Ribeirão
Preto, Universidade de São Paulo, SP, Brasil
3
Departamento de Ciências Exatas, Faculdade de Ciências Agrárias e Medicina
Veterinária da UNESP-Campus de Jaboticabal, SP, Brasil

A cytogenetic study was carried out with 5-azacytidine (5-azaC) and etoposide
(VP-16) in CHO-K1 and XRS-5 (mutant cells deficient for double-strand break
rejoining) cell lines to verify the interaction effects of the drugs in terms of
induction of chromosomal aberrations. 5-azaC is incorporated into DNA causing
DNA hypomethylation, and VP-16 (inhibitor of topoisomerase II enzyme) is a
potent clastogenic agent. Cells in exponential growth were treated with 5-azaC for
1 h, following incubation for 7 h, and posttreatment with VP16 for the last 3 h. In
K1 cells, the combined treatments induced a significant reduction in the
aberrations induced in the X and ‘‘A’’ (autosome) chromosomes, which are the
main target for 5-azaC. However, in XRS-5 cells, the drug combination caused a
significant increase in the aberrations induced in those chromosomes, but with a
concomitant reduction in the randomly induced-aberrations. In addition, each
cell line presented characteristic cell cycle kinetics; while the combined treatment
induced an S-arrest in K1 cells, alterations in cell cycle progression were not
found for XRS-5, although each drug alone caused a G2-arrest. The different cell
responses presented by the cell lines may be explained on the basis of the evidence
that alterations in chromatin structure caused by 5-aza-C probably occur to a
different extent in K1 and XRS-5 cells, since the mutant cells present a typical

Contract grant sponsor: FAPESP; Contract grant sponsor: CAPES; Contract grant sponsor: CNPq.

n
Correspondence to: Elza T. Sakamoto-Hojo, Departamento de Biologia, Faculdade de Filosofia Ciências
e Letras de Ribeirão Preto, Universidade de São Paulo, Av. Bandeirantes 3900, 14040-901 Ribeirão Preto,
SP, Brasil. E-mail: etshojo@usp.br

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/tcm.10072

r 2003 Wiley-Liss, Inc.

172 Guimarães et al.
hyper-condensed chromosome structure (especially the X- and ‘‘A’’ chromo-
somes), but, alternatively, 5-aza-C could induce reactivation of DNA repair genes
in XRS-5 cells. Teratogenesis Carcinog. Mutagen. Suppl. 1:171–186, 2003.


c 2003 Wiley-Liss, Inc.

Key words: 5-azacytidine; etoposide; chromosomal aberrations; DNA methylation

INTRODUCTION
In chemotherapy, drug combinations are commonly prescribed for different
types of malignant neoplasias with the purpose to improve the chemotherapeutic
regimens. However, in most cases, the mechanism by which the cytotoxic effect of a
certain antitumoral drug can be modulated by another compound is poorly
understood. The anti-leukemic drug 5-Azacytidine (5-azaC) is a cytidine analogue
with a nitrogen replacing the carbon atom at 5-position of the pyrimidine ring [1].
When 5-azaC is incorporated into the DNA (instead of cytidine), the nitrogen atom
is not methylated, resulting in a hypomethylated state, which has been correlated
with gene activity and cellular differentiation [2, 3].
Besides its application in anticancer therapy, 5-azaC is widely used in many
fields of biological research, as a useful tool to study the role of DNA methylation at
different cellular processes, on the basis of its property to cause DNA
hypomethylation. There is considerable evidence that methylated sites contribute
to the silencing of gene activity in various biological contexts [4], and genes silenced
by hypermethylation can be reactivated by 5-aza-C [5,6].
DNA methylation normally occurs at cytosine residues within the CpG
dinucleotides, which are represented at lower frequency than that expected
throughout the eukaryotic genome [7, 8]. In contrast, regions of the genome defined
as CpG islands contain the predicted frequency of CpG dinucleotides that are
usually unmethylated, except for those located at the inactive X-chromosome and
parentally imprinted genes [9, 10]. Bender et al. [11] demonstrated that 5-azaC is able
to slow the growth of tumor cells during subsequent cell passages by reactivating
growth-regulatory genes (p16) silenced by de novo methylation.
It has been reported that an intensive regimen of etoposide (VP16) combined
with 5-azacytidine was efficient to treat patients with refractory or relapsed acute
nonlymphocytic leukemia (ANLL), and a complete remission rate of 45–60% was
achieved [12, 13]. Etoposide is a commercially available antineoplastic drug classified
as an inhibitor of DNA-topoisomerase II enzyme, which catalyses the breakage and
rejoining of both DNA strands during replication [14].
In general, inhibitors of topo II are widely used in anticancer chemotherapy,
and the cytotoxic and antitumoral effect of these inhibitors resides in their ability to
inhibit the topo II activity at an intermediate step following DNA scission, before
the re-ligation step [15]. This intermediate is termed cleavable complex and consists
of a topo II monomer that is bound covalently to the 50 termini of a DNA double-
strand break (DSB) [16], thus introducing high levels of transient protein-associated
DNA breaks in treated cells. These breaks become the target for recombination and
sister chromatid exchange, and also generate large insertions and deletions,
consequently producing chromosomal aberrations [17].
In an earlier study, a negative interaction effect between 5-azaC and inhibitors
of topo II enzyme (ellipticine and teniposide) was observed in CHO cells, particularly
 5-Azacytidine Combined With Etoposide 173
a significant reduction in the production of chromosomal aberrations [18]. The
effects of combined treatments with 5-azaC and the antitumoral compound cytosine
arabinoside (Ara-C) were also studied in CHO-K1 compared with XRS-5 cells; a
significant reduction (35–68%) in the frequencies of chromosome aberrations was
observed in both cell lines, but only for the aberrations that were randomly
distributed, i.e., excluding those located on the X-chromosome [19].
XRS-5 cells were described to be deficient for DNA double-strand break
rejoining and more sensitive to the induction of chromosome aberrations following
irradiation [20], or treatment with inhibitors of DNA topoisomerase II enzyme [21].
The aim of the present work was to study the interaction effects of 5-azaC (added to
cell cultures during the S-phase) with VP16 posttreatment at the G2-phase of the cell
cycle, concerning the induction of chromosome aberrations in Chinese hamster
ovary fibroblast cell lines: CHO-K1 (wild-type) and XRS-5 (mutant cells). By
comparing the cell response between both cell lines, it can be expected to obtain
additional information on the mechanism by which the DNA damage induced by
5-azaC would interact (during one cell cycle) with the subsequent effects induced by
etoposide, an inhibitor of topo II enzyme.

MATERIALS AND METHODS

Cell Lines
CHO-K1 and XRS-5 were kindly provided by Pof. A.T. Natarajan (Leiden
University Medical Center, The Netherlands). The cells were maintained as
monolayers growing at 371C in 25-cm2 flasks containing HAM-F10 þ DEM (Sigma,
St. Louis, MO) culture medium supplemented with 15% fetal calf serum (Cultilab,
Brazil), antibiotics (0.01 mg/ml streptomycin and 0.005 mg/ml penicillin), and 2.38
mg/ml Hepes (Sigma). All the experiments were carried out with CHO-K1 and
XRS-5 cells at the 5–10th culture passage after thawing.

Chemicals
Etoposide (VP-16) was obtained from Bristol-Myers Squibb, and 5-aza-C from
Sigma. Since 5-aza-C is unstable in aqueous solution, it was stored at 201C and
prepared in phosphate-buffered saline (PBS) immediately before use. VP-16 was
prepared in deionized distilled water at the concentrations of 0.1, 10, and 100 mg/ml.

Cell Treatment With 5-aza-C and/or VP-16

Cells in exponential growth were seeded (1  106 cells per flask) and allowed to
grow for 20 h. 5-azaC was added to CHO-K1 cell cultures at the concentrations of 2,
4, and 8 mM, while for XRS-5 cells, the concentrations were 1, 2, and 4 mM. In both
cases, 5-azaC treatments (in a 60-min pulse) were performed in serum-free medium, 8
h before fixation, followed by 0.5 or 1.0 mg/ml VP-16, which was added to the
cultures of CHO-K1 cells at the G2-phase (last 3 h). For the mutant XRS-5 cells,
lower concentrations of 0.01, 0.02, and 0.05 mg/ml were chosen on the basis of the
results from preliminary experiments. VP-16 treatment was performed in a 30-min
pulse, followed by three washes in PBS solution and incubation in complete medium
at 371C. Before establishing such levels of concentrations, many preliminary
174 Guimarães et al.
experiments were carried out to evaluate the level of drug cytotoxicity, by
determining the mitotic indexes in addition to the frequencies of chromosomal
aberrations. The cultures were incubated with colcemid (1 mg/ml) during the last 90
and 120 min (before fixation) for XRS-5 and CHO-K1 cells, respectively. Additional
experiments were carried out using different fixation time points: 3, 4, and 5 h.

Chromosome Spread Preparations and Cytogenetic Analysis

Cells were harvested by standard cytogenetic methods (hypotonic shock with
1% sodium citrate, methanol/acetic acid (3:1) as fixative, and metaphase block with
colcemid. The air-dried chromosome preparations were stained with 2% Giemsa
diluted in phosphate buffer. Following the criteria established in the literature [22],
metaphases with an acceptable chromosome number (2172) were analyzed to
determine the frequencies of chromosomal aberrations. One hundred metaphases
were analyzed per culture to score the chromatid or isochromatid breaks, as well as
chromosome exchanges. The gap lesions were also included in the analysis. The
mitotic indexes were obtained by counting the number of mitotic cells in a total of
2,000 cells per treatment. In the combined treatments, the expected values for the
CA frequencies were calculated as the sum of the effects caused by each drug alone
minus the control value; these values were used to calculate the magnitude of
reduction.

Cell Cycle Analysis

Similar experiments (as described for cytogenetic analysis) were carried out with
both cell lines, and the cell cycle analysis was performed by staining the cells
with propidium iodide and applying the detergent-trypsin method described by
Vindelov et al. [23]. Ten thousand cells were analyzed using a FACSCAN flow
cytometer (Becton Dickinson, San Jose, CA), and cells were distributed along the
different phases of the cell cycle according to the DNA content, which was
performed on gated nuclei using the software program (CellFit) provided by Becton
Dickinson. The Z-Test for comparison of proportions was applied to compare the
distribution of cells at different phases of the cell cycle. Instead of 10,000, an
‘‘n’’ ¼ to 5,000 cells was considered to minimize the sensitivity of the test and accept
only the strictly significant differences. The probability of error considered
was r0.05.

Statistical Analysis
Multi-factorial analysis of variance with repeated measures was applied to the
data, followed by the Tukey test in order to verify the interaction effects of 5-azaC
treatment combined with VP16 concerning the induction of chromosomal
aberrations.

RESULTS
Clastogenic Effects of 5-aza-C
In the experiments performed with CHO-K1 cells, 2, 4, and 8 mM 5-azaC
treatments induced an increase in the frequencies of chromosomal aberrations as a
 5-Azacytidine Combined With Etoposide 175
function of drug concentration, and by the Tukey test analysis, the differences
observed in the CA frequencies were significant, Po0.05 and Po0.01, for the
treatments with 4 and 8 mM, respectively. Only 5-azaC 8 mM induced a significant
(Po0.05) reduction in the mitotic index (Table I). Similar experiments were carried
out with XRS-5 cells by treating them with 1, 2, and 4 mM of 5-aza-C, which induced
a significant (Po0.01) increase in the CA frequencies as compared with the control
levels, but the differences were not statistically significant (P40.05) among the
different treatments. The mitotic indexes suffered a significant (Po0.05) decrease for
the highest 5-azaC treatments (2 and 4 mM) (Table I).
In both cell lines, CHO-K1 and XRS-5, 5-azaC induced a high incidence of
breaks in the X-chromosome, and also in a specific autosome called ‘‘A’’
chromosome, varying from 28 to 76% and 23 to 60%, respectively, for each cell
line (Table II).

Clastogenic Effects of VP-16

Experiments with CHO-K1 cells showed that 0.5 and 1.0 mg/ml VP-16 induced a
statistically significant (Po0.05) increase in the CA frequencies, and compared with
the control value, the differences were much higher (Po0.01) (Table I). In XRS-5
cells, lower VP-16 concentrations were tested (0.01, 0.02, and 0.05 mg/ml), and they
induced a significant (Po0.01) increase in the CA frequencies in relation to the
control level. The analysis of the total number of abnormal cells also showed an
increase as a function of drug concentration (Table I). Although VP-16 decreased the
mitotic indexes, the differences with the control values were statistically significant
(Po0.05) only for K1 cells.

Interaction Effects of the Combined (5-Azac þ VP16) Treatments

Comparison between the cell responses of both cell lines to the combined
treatments indicated that each one presented a characteristic pattern. A lack of
interaction effect occurred in the frequency of total CAs (considering the aberrations
in the whole genome), except in the case of XRS-5 treated with the highest
concentration of two drugs in combination. In spite of this, in the wild-type K1 cells,
the combined treatments induced a significant (Po0.05) reduction in the aberrations
induced in the X and ‘‘A’’ chromosomes, ranging from 35 to 55% in relation to the
expected values (Fig. 1A), while in XRS-5 cells there was a significant (Po0.05)
increase in the aberrations induced in those chromosomes (18 to 37%),
independently of VP-16 concentration, except the treatment with 5-azaC 4 mM
(Table II; Fig. 1B); however, concomitantly, the mutant XRS-5 cells presented a
significant (Po0.05) reduction in the aberrations randomly-induced in the genome,
i.e., those scored in all chromosomes, excluding X and ‘‘A.’’
In K1 cells, the mitotic indexes for the combined treatments did not differ
among the different cultures, but the differences were significant (Po0.05) in
relation to the control value (Table I). A significant reduction (Po0.01) in the
mitotic indexes (compared with the control) was found for XRS-5 cells submitted to
the combined treatments with 2 and 4 mM 5-aza-C þ VP16 (Table II).
Considering the toxicity caused by each drug and the possibility of a delay in the
cell cycle, and also cell death, additional experiments were performed using the same
protocol of cell treatment, but different fixation times (3, 4, or 5 h) after drug
TABLE I. Chromosomal Aberrations Obtained in the Combined Treatments With 5-azaC and VP16 in CKO-K1 and XRS-5 Cell Lines, Considering the
Total Aberrations in the Genomew
Treatments CHO-K1 Treatments XRS-5

5-aza-C VP-16 MI(%)7SE Abnormal cells7SE CAs/100cells7SE 5-aza-C VP-16 (mg/ml) MI(%)7SE Abnormal cells7SE CAs/100 cells7 SE
(mM) (mg/ml) (mM)

F F 7.470.7 3.670.7 471.0 F F 7.070.2 1170.6 11.770.3

nn
2 F 6.470.9 7.670.9 10.770.7 1 F 5.070.2 23.772.6 3074.4
n n nn
4 F 4.870.2 2175.0 2572.1 2 F 3.770.3 33.372.7 36.772.0
n nn n nn
8 F 4.170.6 3872.5 44.371.9 4 F 3.370.4 41.773.2 49.374.3
n n nn
F 0.5 3.770.2 1771.0 18.771.4 F 0.01 4.670.5 22.373.8 23.374.6
n nn nn
F 1.0 2.470.1 4178.5 5576.7 F 0.02 3.870.2 3172.0 3973.22
n n nn
2 0.5 3.570.3 20.371.8 26.370.9 F 0.05 3.870.7 4973.2 67.778.8
n nn nn
2 1.0 2.170.4 4174.7 57.374.3 1 0.01 3.770.5 2972.1 3373.2
n nn nn
4 0.5 3.570.4 n27.672.4 32.372.7 1 0.02 4.870.2 41.570.5 48.573.5
n nn
4 1.0 2.170.5 n51.379.4 +63.777.4 1 0.05 4.370.3 53.774.7 70.373.7
n nn nn
8 0.5 2.070.7 n4176.1 48.378.2 2 0.01 4.270.7 4072.1 43.774.5
n nn nn
8 1.0 1.270.3 n57.6710.7 72.6715.3 2 0.02 4.270.9 41.373.8 47.372.5
n nn
F F F F F 2 0.05 3.070.8 5574.93 73.773.7
n nn
F F F F F 4 0.01 3.770.8 41.371.8 5074.5
n nn
F F F F F 4 0.02 3.370.7 4771.0 5672.5
n nn
F F F F F 4 0/05 2.470.7 5171.2 6471.5
w
Three hundred cells were analyzed in three independent experiments.
n
Po0.05, nnPo0.01.
Significant difference aanalyzed by the Tukey test.
MI=mitotic index; CAs=chromosomal aberrations; SE=standard error of the mean.
TABLE II. Distribution of Chromosomal Aberrations at the Preferential Sites (X+‘‘A’’) for 5-azaC and in the Rest of the Genome [Total-(X+‘‘A’’)] in
CHO-K1 CHO-K1 and XRS-5 Cells Submitted to the Combined Treatments With 5-azaC and VP-16
CHO-K1 XRS-5

Treatments Chromosomal aberrationsa Treatments Chromosomal aberrationsa

5-aza-C (mM) VP-16 (mg/ml) Total–(X+‘‘A’’)7SE (%) X+‘‘A’’7SE (%) 5-aza-C (mM) VP-16 (mg/ml) Total–(X+‘‘A’’)7SE (%) X+‘‘A’’7SE (%)

F F 471.7 (100) 0.0 (0.0) F F 11.470.6 (99.9) 0.370.6 (0.1)

2 F 7.770.8 (71.9) 3.070.5 (28.1) 1 F 23.076.5 (76.7) 7.070.5 (23.3)
4 F 11.771.6 (42.7) 14.37 2.3 (57.3) 2 F 15.172.1 (40.4) 21.671.7 ( 59.6)
8 F 10.773.2 (24.1) 33.675.3 (75.9) 4 F 31.378.4 (63.5) 18.071.4 (36.5)
F 0.5 18.772.5 (100) 0.0 ( 0.0) F 0.01 23.078.0 (99.9) 0.370.6 (0.1)
F 1.0 55.0711.5 (100) 0.0 (0.0) F 0.02 39.075.6 (100) 0.0 (0.0)
2 0.5 24.372.3 (92.4) 2.071.2 (7.6) F 0.05 66.4715.4 ( 99.6) 1.370.1 (0.4)
2 1.0 54.777.1 (95.4) 2.670.7 (4.6) 1 0.01 28.075.3 (85.0) 5.070.4 (15.0)
4 0.5 22.078.8 (71.2) 9.374.6 (28.8) 1 0.02 45.0728.0 (92.8) 3.570.3 (7.2)
4 1.0 57.4711.7 (90.1) 6.371.6 (9.9) 1 0.05 60.075.7 (85.3) 10.370.8 ( 14.7)
8 0.5 32.7710.2 (67.6) 15.675.1 (32.4) 2 0.01 28.474.0 (65.0) 15.371.2 (35.0)
8 1.0 49.3721.8 (70.0) 23.3 74.7 (30.0) 2 0.02 25.375.6 (53.6) 22.071.7 (46.4)
F F F F 2 0.05 52.375.5 (71.0) 21.471.6 (29.0)
F F F F 4 0.01 25.476.9 (50.7) 24.671.9 (49.3)
F F F F 4 0.02 34.772.7 (62.0) 21.371.6 (38.0)
F F F F 4 0.05 40.672.0 (63.5) 23.471.7 (36.5)
n
Percentage shown is in relation to the total number of CAs.
178 Guimarães et al.

Fig. 1. Frequencies of chromosomal aberrations in CHO-K1 (A) and XRS-5 (B) cells submitted to the
combined treatments with 5-azaC and VP16, taking into consideration those induced in the X and ‘‘A’’
chromosomes, and those which were randomly-induced in the total genome. The expected values for CA
frequencies were calculated as the sum of the effect caused by each drug alone minus the control value.
Three hundred cells were analyzed for each cell line. (*Statistically significant, Po0.05).
 5-Azacytidine Combined With Etoposide 179
removal (Table III). The results obtained for both cell lines indicated an increase in
the mitotic indexes for the cultures harvested at later times, but without variation in
the CA frequencies. K1 cells treated with VP-16 1.0 mg/ml presented similar CA
frequencies, and the mitotic indexes were recovered at later times. The interaction
effect was not significant for the combined treatments (5-aza-C þ VP-16), and the
magnitude of reduction in the CA frequencies did not suffer a significant difference
among the different harvesting times, and the results were similar to those obtained
in the previous experiments. Similarly, in XRS-5 cells, an increase in the mitotic
index was also observed, being higher at later times, but without variation in the CA
frequencies. Therefore, 5-aza-C þ VP-16 combined treatments induced an interac-
tion effect in XRS-5 cells (32.5 to 39.6% of reduction in CA frequencies), being
statistically significant (Po0.05), but independent of the time of cell harvesting
(Table III). These results indicated that the reduction in the CA frequencies is not a
consequence of drug cytotoxicity leading to a blockade in the cell cycle.

Cell Cycle Kinetics

Cell cycle analysis was performed by flow cytometry in order to study the
alterations in the cell cycle progression in response to the treatment with 5-aza-C and
VP-16. Each cell line presented characteristic cell cycle kinetics. While CHO-K1 cell
line without any treatment presented 58.9, 27.3, and 13.8% of cells at G0/G1, S, and
G2/M cell cycle phases, XRS-5 untreated cells showed 49.4, 19.6, and 31.0%,
respectively. Small variations on the cell cycle progression were observed in response
to the treatment with 5-aza-C or VP-16 alone, and a statistically significant increase
in the percentage of cells (B46%) undergoing the S-phase was observed in K1 cells
submitted to the combined treatments (5-aza-C þ VP-16) in comparison to the
control value (27.3%). Concomitantly, we observed a significant decrease (Po0.05)
in the incidence of K1 cells at G0/G1 and G2/M phases for the combined treatment
(8 mM 5-aza-C þ 1.0 VP-16) (Fig. 2). On the other hand, XRS-5 cells treated with
5-aza-C suffered an increase in the percentage of cells at G2/M transition, and this
increase was considerable for 4 mM 5-azaC, i.e., on the border of statistical
significance (P = 0.0512). While the treatment with 0.01 mM VP-16 alone induced a
significant increase (Po0.05) in the incidence of cells undergoing the G2/M phase,
the combined treatments (5-aza-C 4mM þ VP-16) did not induce significant
variations in the distribution of XRS-5 cells at different phases of the cell cycle
(Fig. 2).

DISCUSSION
Although many authors emphasize the importance of 5-aza-C as a hypomethy-
lating agent, there are few reports in the literature describing its activity in the
chromosome structure, and cytotoxicity. In vitro studies demonstrated that this
compound has efficient cytotoxic and clastogenic activity during the S-phase [18, 24],
probably due to DNA hypomethylation as a consequence of inhibition of DNA
methyltransferase activity [25].
In the present study, cells in exponential growth were treated with 5-aza-C 8 h
before fixation, when the majority of cells were undergoing the S-phase of the cell
cycle, and VP-16 was added at G2-phase (3 h before fixation), as a post-treatment. In
TABLE III. Chromosomal Aberrations and Mitotic Indexes Observed for Different Fixation Times in CHO-K1 and XRS-5 Cells Treated With 5-azaC at
S-Phase and Posttreated With VP-16 at G2-Phase of the Cell Cyclen
CHO-K1 cells XRS-5 cells

Treatments Treatments

5-aza-C VP-16 Fixation Time MI Total Reduction 5-aza-C VP-16 Fixation Time MI Total Reduction
(mM) (mg/ml) (h) (%) (%) (%) (mM) (mg/ml) (h) (%) (%) (%)

F F 3 6.6 0 F F F 3 6.2 10 F
F F 4 8.0 2 F F F 4 7.8 4 F
F F 5 8.6 1 F F F 5 7.5 8 F
F 1.0 3 2.9 55 F F 0.05 3 3.7 59 F
F 1.0 4 4.0 56 F F 0.05 4 6.2 55 F
F 1.0 5 5.1 53 F F 0.05 5 6.7 50 F
8.0 F 3 3.3 40 F 4 F 3 2.4 47 F
8.0 F 4 4.2 37 F 4 F 4 4.1 39 F
8.0 F 5 4.9 32 F 4 F 5 6.7 35 F
8.0 1.0 3 1.1 69 27.4 4 0.05 3 3.2 58 39.6
8.0 1.0 4 3.5 67 26.4 4 0.05 4 2.8 57 36.7
8.0 1.0 5 4.1 65 22.7 4 0.05 5 4.9 52 32.5
n
The magnitude of reduction in the frequencies of chromosomal aberrations was calculated using the expected values (sum of the effects caused by each drug alone minus
the control value)
MI= mitotic index;
CAs = chromosomal aberrations.
 5-Azacytidine Combined With Etoposide 181

Fig. 2. FACS profiles showing the percentage of K1 and XRS-5 cells at different phases of the cell cycle in
cultures treated with 5-azaC and VP-16. Ten thousand cells/treatment were scored by flow cytometry. In
both cell lines, 5-azaC treatments (in a 60-min pulse) were performed in serum-free medium, 8 h before
fixation, following by VP-16, which was added to the cultures at the G2-phase (3 h before the harvesting
time).

both cell lines, the increase in the CA frequencies caused by 5-azaC alone was
proportional to drug concentration, and 5-azaC induced high frequencies of
chromosome breaks at preferential sites, in the long arm of the X-chromosome and
also in one of the autosomes (called ‘‘A’’ chromosome), which showed a
characteristic uncoiled and elongated shape, as previously demonstrated at similar
conditions [18]. This effect probably occurs due to the presence of heterochromatic
regions, which are the main target for 5-azaC, since there is evidence indicating that
incorporation of 5-azaC into DNA during its replication and the consequent induced
182 Guimarães et al.
DNA hypomethylation may affect the chromosome condensation pattern, mainly in
the heterochromatic regions, which presents the highest extent of methylation
[26,27].
Experiments carried out with VP-16 tested alone at G2-phase demonstrated its
clastogenic activity in both cell lines, and the induction of aberrations occurred
proportionally to drug concentration. VP-16 does not bind directly to DNA, but it
interferes with topoisomerase II enzyme activity by stabilizing the cleavable complex
[28, 29]. The topo II enzyme participates in cell proliferation [30] and its maximum
activity occurs during the G2-phase of the cell cycle [31].
Comparison between the cell responses of both cell lines to the combined
treatments indicated that each one presented a characteristic pattern. A lack of
interaction effect occurred in the frequency of total CAs (considering the aberrations
in the whole genome), except in the case of XRS-5 treated with the highest
concentration of two drugs in combination. In spite of this, in the wild-type K1 cells
the combined treatments induced a significant reduction in the aberrations induced
in the X and ‘‘A’’ chromosomes, while in XRS-5 cells there was a significant increase
in the aberrations induced in those chromosomes; however, in contrast, the mutant
XRS-5 cells presented a significant reduction in the aberrations induced randomly in
the genome, i.e., those scored in all chromosomes, excluding X and ‘‘A.’’
It has been reported that DNA methylation can alter the distribution of the
binding sites for topo II enzyme [32]. This information gives support to suggest that
in cells submitted to the combined treatment at high drug concentrations (4 mM aza-
C þ 0.05 mg/ml VP-16), 5-azaC incorporation into DNA could change the cleavage
sites for the topo II inhibitor, protecting the DNA for the induction of damage by
VP-16, leading to the reduction in the aberrations induced by the combined
treatments in K1 and XRS-5 cells. However, in spite of the fact that 5-azaC can
induce chromatin decondensation, this effect could occur at less extension in XRS-5
cells, since their chromosomes are normally super-condensed; while in K1 cells 5aza-
C treatment may facilitate the binding of DNA repair enzymes, thus reducing the
induction of chromosome damage, as suggested by Slijepcevic and Natarajan [33], in
the mutant cells the hyper-condensation status of the chromosomes, especially in the
heterochromatic X and ‘‘A’‘ chromosomes, may make access to the elements of
DNA repair machinery difficult, leading to increased frequencies of aberrations in
those chromosomes, as observed only in XRS-5 cells. This is supported by the
evidence that the sensitivity of XRS-5 cells to chemicals has been attributed to their
super-condensed chromosomes, with consequent difficulty accessing DNA repair
enzymes [34].
Furthermore, alterations in the methylation pattern caused by 5-aza-C can lead
to changes in chromatin structure, and the extent by which this drug incorporates
into DNA in CHO-K1 and XRS-5 cells can be different, considering the typical
chromosome morphology of the mutant cells [35]. This aspect can influence the
binding of transcription factors, making gene transcription and repair activity in
XRS-5 cells difficult. Therefore, the differences in the cell response observed between
the cell lines can be a consequence of alterations in chromatin structure and in the
DNA-nuclear matrix, which are characteristics of XRS-5 cells [36, 37], but they also
depend on the chemical structure of the compounds. In a previous work, 5-azaC
administered to CHO cells during the S-phase of the cell cycle caused a significant
reduction in the frequency of chromosomal aberrations induced by the DNA
 5-Azacytidine Combined With Etoposide 183
topoisomerase II inhibitors, teniposide (VM26) and ellipticine (EPC), during the G2
phase of the cell cycle [18]. In another study, Sakamoto-Hojo et al. [19] demonstrated
that the combined treatment of 5-azaC with cytosine arabinoside (Ara-C) in K1 and
XRS-5 cells induced a significant reduction in the frequencies of aberrations that
were randomly distributed in the genome, but the same effect was not observed for
those located on the X-chromosome. In contrast, in this study VP-16 showed a
synergistic effect in the induced-aberrations at the preferential sites for 5-azaC, i.e.,
those located at X and ‘‘A’’ chromosomes, and this can be attributed to the different
interaction of the topo II inhibitor with the DNA in those super-condensed
chromosomes, and also to the reduced accessibility to DNA repair enzymes, as
already mentioned.
However, other authors have demonstrated that 5-azaC potentiates the
induction of sister chromatid exchange by mitomycin C in human lymphocytes
[38] and CHO cells [39], as well as cytotoxicity [40], indicating that the cell response
to chemicals in combination with 5-azaC should be influenced by the kind of drug,
the cell type, as well as the parameter analysed.
Another aspect to be considered in order to interpret the present results is
related to the evidence that the control of gene transcription is associated with the
methylation status, and that some ‘‘silent genes’’ (like the genes on the inactive
X-chromosome) are hypermethylated, but with the possibility of being reactivated
by demethylation induced by 5-aza-C treatment [6, 41]. Jeggo and Holliday [21]
described the isolation and the partial characterization of 6 XRS strains sensitive to
X-rays and deficient to DSB repair, showing that XRS cells reverted to the wild type
after 5-aza-C treatment, becoming more resistant to X-rays. In addition, other
reports provided evidence supporting that XRS-5 cells harbor an intact but non-
expressed KU80 allele, silenced by hypermethylation, which can be reactivated by
5-azaC [42]. Similarly, it can be suggested that in XRS-5 cells, the protective effect
of 5-aza-C (4 mM) combined with VP-16 (0.05 mg/ml) post-treatment at G2-phase
may be alternatively explained by the activation of DNA repair gene(s) in
consequence of DNA demethylation, leading to reduced frequencies of randomly
induced chromosomal aberrations; on the other hand, the wild-type CHO-K1 cells
may present a normal unmethylated copy of this gene, which is active in the DNA
repair process. However, considering the high level of induced-aberrations at
preferential sites (X and ‘‘A’’ chromosomes), which were the target for 5-azaC, it can
be suggested that the synergistic effect of this compound with VP-16 may be a
consequence of the characteristic chromatin structure of XRS-5 cells, which can
influence its interaction with DNA, or making difficult the access to repair enzymes,
as mentioned before.
Analysis of cell cycle progression showed that each cell line presented
characteristic cell cycle kinetics; without any treatment, K1 cells presented 58.9,
27.3, and 13.8% of cells at G0/G1, S and G2/M phases, while XRS-5 cells showed
49.4, 19.6, and 31.0%, respectively. Variations on the cell cycle progression were
observed in response to drug treatments, and a statistically significant increase in the
percentage of cells (B46%) undergoing the S-phase (S-phase arrest) was observed in
K1 cells treated with 5-aza-C þ VP-16, in comparison with the control value
(27.3%).
XRS-5 cells treated with each drug alone showed a G2-arrest, which is in
accordance with the results of Poot et al. [43], who demonstrated that 5-aza-C leads
184 Guimarães et al.
to an extended G2 phase, and can inhibit the cell division at this phase. However, the
combined treatment (5-aza-C 4 mM þ VP-16) did not change the cell cycle kinetics,
indicating that as a whole, the induced-chromosome damage was not sufficient to
cause a G2-arrest, being compatible with the reduction in the total aberrations for
the whole genome observed for the mutant cells. These results were reinforced by the
additional experiments, in which the cell cultures were harvested at different times
(3, 4, or 5 h). XRS-5 cells submitted to the combined treatment (5-aza-C þ VP-16)
showed a significant interaction effect (32.5 to 39.6% reduction in the total CA
frequencies), independently of the harvesting time. In both cell lines, an increase in
the mitotic indexes was observed, being higher at later times, showing a tendency to
recover the cell cycle progression. These results indicated that the interaction effect
of the combined treatments was not consequence of the high drug cytotoxicity, or
cell selection.

CONCLUSIONS
The present results demonstrate that treatment of CHO-K1 and XRS-5 cells
with 5-aza-C can influence the induction of chromosome damage by the topo II
inhibitor VP-16, but the cell lines presented different cell response. In the wild-type
K1 cells, the combined treatments induced a significant reduction only in the
aberrations induced in the X and ‘‘A’’ chromosomes (which are the main target for
5-azaC), while in XRS-5 cells, the drug combination caused a significant increase in
the aberrations induced in those chromosomes, and a concomitant reduction in the
aberrations randomly induced in the genome.
Alterations in the methylation pattern caused by 5-aza-C can lead to changes in
chromatin structure, and the extent by which this drug incorporates into DNA in
CHO-K1 and XRS-5 cells can be different, considering the typical chromosome
morphology of the mutant cells. The data in the literature give the support to suggest
that in cells submitted to the combined treatment with high drug concentrations
(4 mM aza-C þ 0.05 mg/ml VP-16), 5-azaC incorporation into DNA could change the
cleavable sites for the topo II inhibitor, protecting the DNA for the induction of
damage by VP-16, leading to the reduction in the aberrations induced by the
combined treatments in K1 and XRS-5 cells. However, in spite of the fact that
5-azaC can induce chromatin decondensation, this effect could occur to a less extent
in XRS-5 cells, since their chromosomes are normally super-condensed. While in
K1 cells 5aza-C treatment may facilitate the binding of DNA repair enzymes,
thus reducing the induction of chromosome damage, in the mutant cells the
hyper-condensation status of the chromosomes, especially the X and ‘‘A’’
chromosomes, may make access to the elements of DNA repair machinery difficult,
leading to increased frequencies of aberrations in those chromosomes, as observed
only in XRS-5 cells. Alternatively, on the basis of data in the literature, another
possibility to explain the reduction in the CA frequencies is that 5-aza-C induces
reactivation of certain genes involved in the repair process.

ACKNOWLEDGMENTS
We are grateful to Luis A. Costa Júnior and Sueli A. Neves for technical assistance.
 5-Azacytidine Combined With Etoposide 185
REFERENCES

1. Von Hoff D, Slavik M, Muggia F. 5-azacytine: a new anticancer drug with effectiveness in acute
myelogenous leukemia. Annu Int Med 1986;85: 237-245.
2. Jones PA. Altering gene expression with 5-azacytidine. Cell 1985;40:485–486.
3. Glover AB, Leyland-Jones B. Biochemistry of Azacytidine: a review. Cancer Treat Rep 1987;71:
959–964
4. Hollyday R, Monk M, Pugh JE, editors. DNA methylation and gene regulation. London: The Royal
Society; 1990.
5. Pinto A, Zagonel V. 5-aza-20 - deoxycytidine (Decitabine) and 5-aza-cytidine in the treatment of acute
myeloid leukemias and myelo dysplastic syndromes: past, present and future trends. Leukemia
(Baltimore) 1993;1:51–60.
6. Jones PA, Taylor SM. Cellular differentiation, cytidine analogs and DNA methylation. Cell
1995;20:85–93.
7. Bird AP. CpG- rich islands and the function of DNA methylation. Nature (Lond) 1986;321:
209–213.
8. Gardiner-Garden M, Frommer M. CpG island in vertebrate genomes. J Mol Biol 1987;196:261–282.
9. Li E, Berad C, Jaenish R. DNA Methylation in genomic imprinting. Nature (Lond) 1993;386:382–385.
10. Ferguson-Smith A, Sasaky LH, Cattanach BM, Surani MA, Parental Froelich-Ammon SJ,
Osheroff N. Topoisomease poisons: Harnessing the dark side of enzyme mechanisms. J Biol Chem
1995;270:21429–21432.
11. Bender MC; Pao MM, Jones PA. Inhibition of DNA methylation by 5-aza-20 deoxycytidine supresses
the growth of human tumor cell line. Cancer Res 1998;58:95–101.
12. Kalwinsky DK, Dahl GV, Mirro J, Jackson, CW, Look AT. Induction failures in childhood acute
nonlymphocytic leukemia: etoposide/5-azacytidine for cases refractory to daunorubicin/cytarabine.
Med Pediatr Oncol 1986;14:245–250.
13. Hakami N, Look AT, Steuber PC, Krischer J, Castleberry R, Harris R, Ravindranath Y, Vietti, TJ.
Combined etoposide and 5-azacitidine in children and adolescents with refractory or relapsed acute
nonlymphocytic leukemia: a Pediatric Oncology Group Study. J Clin Oncol 1987;5:1022–1025.
14. Smith PJ. DNA topoisomerase dysfunction: a new goal for antitumor chemotherapy. Bioassays
1990;12:167–172.
15. Liu LF, Chen AY. DNA topoisomerases: essential enzymes and lethal targets. Ann Rev Pharmacol
Toxicol 1994;34:191–218.
16. Boulton S, Kyle S, Durkacz BW. Mechanisms of enhancement of cytotoxicity in etoposide and
ionizing radiation-treated cells by the protein kinase inhibitors. Eur J Cancer 2000;36:535–541.
17. Hande KR. Etoposide: four decades of development of a topoisomerase II inhibitor. Euro J Cancer
1998;34:1514–1521.
18. Takahashi-Hyodo AS, Sakamoto-Hojo ET, Takahashi CS: Interaction effects of 5-azacytidine with
topoisomerase II inhibitor on CHO cells, as detected by cytogenetic analysis. Mutat Res 1999;
431:13–23.
19. Sakamoto-Hojo ET, Bı́cego-Nahas KC, Takahashi-Hyodo SA, Tavares DC, Dias FL: Chromosome
damage induced by 5-azacytidine under the influence of caffeine or cytosine arabinoside in CHO-K1
(wild-type) and XRS-5 (mutant) cell lines. Cancer Res Ther Cont 2001;11:125–134.
20. Kemp LM, Sedgwick SG, Jeggo PA. X-ray sensitive mutants of Chinese hamster ovary cells defective
in double-strand break rejoining. Mutat Res 1984;132:189–196.
21. Jeggo PA, Hollyday R: Azacytidine-induced reactivation of a DNA repair gene in Chinese Hamster
ovary cells. Mol Cell Biol 1986;6:2944–2949.
22. Holliday R. A diferent kind of Inherance. The methylation of DNA maybe a major ‘‘epigenetic’’
mechanisms by wich gene-activity patterns-as opposed to genes per se- are passed from one generation
of cells to another during development. Sci Am 1989;260:40–48.
23. Vindelov LL, Christensen IJ, Nissen NI. A detergent-trypsin method for the preparation of nuclei for
flow cytometric DNA analysis. Cytometry 1983;3:323–327.
24. Li LH, Olin EJ, Fraser TJ, Bhuyan BK. Phase specificity of 5-azacytidine against mammalian cells in
tissue culture. Cancer Res 1970;30:2770–2775.
25. Santi DV, Norment A, Garret CE. Covalent bond formation between a DNA-cytosine
methyltransferase and DNA containing 5-azacytosine. Proc Natl Acad Sci USA 1984;81:6993–6997.
186 Guimarães et al.
26. Jones PA, Taylor SM: Cellular differentiation, cytidine analogs and DNA methylation. Cell
1980;20:85–93.
27. Galloway SM, Armstrong MJ, Reuben C, Colman S, Brown B, Cannon C, Bloom AD, Nakamura F,
Ahmed M, Duck S, Rimpo J, Margolin BH, Resnick MA, Anderson B, Zeiger E. Chromosome
aberrations and sister chromatid exchange in chinese hamster ovary cells: evaluation of 108 chemicals.
Environ Mol Mutag 1987;10:1–175.
28. Chen GL, Liu Y, Rawe TC, Halligan BD, Tewey KM, Liu LF. Nonintercalative antitumor drugs
interfere with the breakage-reunion reaction of mammalian DNA topoisomerase II. J Biol Chem
1984;259:13560–13566.
29. Wang JC. Recents studies of CNA topoisomerasis. Biochim Biophys Acta 1987;909:1–9.
30. Sullivan DM, Glisson BS, Hodges PK, Smallewood-Kentro S, Ross WE. Proliferation dependence of
topoisomerase II mediated drug action. Biochemistry 1986;25:2248–2256.
31. Chow K, Ross WE. Topoisomerase-specific drug sensitivity in relation to cell cycle progression. Mol
Cell Biol 1987;7:3119–3123.
32. Leteurtre F, Kohlhagen G, Mark RF, Tanizawa LW, Kohn LW, Pommier I. Effects of DNA
methylation on topoisomerase I and II cleavage activities. J Biol Chem 1994;269:7893–7900.
33. Slijepcevic P, Natarajan AT. Fragile site at Xq21 in Chinese Hamster and its implications for the in
vitro chromosomal aberrations test. Mutagenesis 1995;10:353–355.
34. Holliday R, HO T. Gene silencing and endogenous DNA methylation in mammalian cells. Mutat Res
1998;400:361–368.
35. Weiss A, Keshet I, Razin A, Cedar H. DNA methylation in vitro: involvement of RNA. Cell
1996;86:708–718.
36. Yasui LS, Ling-Indeck L, Johnson-Wint B, Fink TJ, Molsen D. Changes in the Nuclear Structure in
the Radiation-Sensitive CHO Mutant Cell, xrs-5. Radiation Res 1991;127:269–277.
37. Schwartz JL, Brinkman WJ, Kasten L, Miller DW, Moan EI, Murphy YT, Stella D, Sedita BA.
Altered metaphase chromosome structure in xrs-5 cells is not related to its radiation sensitivity or
defective DNA break rejoining. Mutat Res 1995;328:119–126.
38. Lavia P, Ferraro M, Micheli A, Olivieri G. Effect of 5-azacytidine (5-azaC) on the induction of the
chromatid aberrations (CA) and sister-chromatid exchanges (SCE). Mutat Res 1985;149:463–467.
39. Shipley J, Sakai K, Tantravahi U, Fendrock B, Latt AS. Correspondence between effects of 5-
azacytidine on SCE formation, cell cycling and DNA methylation in Chinese hamster cells. Mutat Res
1985;150:333–345.
40. Smith KS, Folz BA., Adams EG, Bhuyan BK. Synergistic and additive combinations of several
antitumor drugs and other agents with the potent alkylating agent adozelesin. Cancer Chemother
Pharmacol 1995;35:471–482.
41. Razin A, Cedar H. DNA methylation in eukaryotic cells. Int Rev Cytol 1984;92:159–185.
42. Singleton BK, Priestley A, Steingrimsdottir H, Gell D, Blunt T, Jackson SP, Lehmann AR, Jeggo PA.
Molecular and biochemical characterization of xrs mutants defective in Ku80. Mol Cell Biol
1997;17:1264–1273.
43. Poot M, Koehler J, Rabinovitch OS, Hoehn H, Priest JH. Cell kinetic disturbances induced by
treatment of human diploid fibroblasts with 5-azacytidine indicate a major role for DNA methylation
in the regulation of the chromosome cycle. Human Genet 1990;84:258–262.