Cancer Prevention
Reversal of Hypermethylation and Reactivation of p16 INK4a, RARb,
and MGMT Genes by Genistein and Other Isoflavones from Soy
Ming Zhu Fang,1 Dapeng Chen,1 Yi Sun,1 Zhe Jin,1 Judith K. Christman,2 and Chung S. Yang1
Abstract
Purpose: We have previously shown the reactivation of some methylation-silenced genes in
cancer cells by ( )-epigallocatechin-3-gallate, the major polyphenol from green tea. To determine whether other polyphenolic compounds have similar activities, we studied the effects of
soy isoflavones on DNA methylation.
Experimental Design: Enzyme assay was used to determine the inhibitory effect of genistein
on DNA methyltransferase activity in nuclear extracts and purified recombinant enzyme.
Methylation-specific PCR and quantitative real-time PCR were employed to examine the DNA
methylation and gene expression status of retinoic acid receptor h (RARb), p16INK4a , and O 6methylguanine methyltransferase (MGMT) in KYSE 510 esophageal squamous cell carcinoma
cells treated with genistein alone or in combination with trichostatin, sulforaphane, or 2Vdeoxy-5-aza-cytidine (5-aza-dCyd).
Results: Genistein (2-20 Amol/L) reversed DNA hypermethylation and reactivated RARb,
p16INK4a , and MGMT in KYSE 510 cells. Genistein also inhibited cell growth at these concentrations. Reversal of DNA hypermethylation and reactivation of RARb by genistein were also
observed in KYSE 150 cells and prostate cancer LNCaP and PC3 cells. Genistein (20-50
Amol/L) dose-dependently inhibited DNA methyltransferase activity, showing substrate- and
methyl donor ^ dependent inhibition. Biochanin A and daidzein were less effective in inhibiting
DNA methyltransferase activity, in reactivating RARb, and in inhibiting cancer cell growth. In
combination with trichostatin, sulforaphane, or 5-aza-dCyd, genistein enhanced reactivation of
these genes and inhibition of cell growth.
Conclusions:These results indicate that genistein and related soy isoflavones reactivate methylation-silenced genes, partially through a direct inhibition of DNA methyltransferase, which
may contribute to the chemopreventive activity of dietary isoflavones.
Methylation of CpG islands in the promoter region is a key
epigenetic mechanism for the silencing of many genes
including tumor suppressor genes, as well as genes encoding
hormone receptors, DNA repair enzymes, mediators of the
apoptosis pathway, and detoxification enzymes (1, 2). Unlike
many other genomic alterations that occur during carcinogenesis, DNA methylation is potentially reversible via the use of
preventive or therapeutic agents (3, 4). 5-Aza-cytidine and 2Vdeoxy-5-aza-cytidine (5-aza-dCyd), nucleoside analogue inhibitors of DNA methytransferases, have been widely used in
Authors’ Affiliations: 1Susan Lehman Cullman Laboratory for Cancer Research,
Department of Chemical Biology, Ernest Mario School of Pharmacy, Rutgers, The
State University of New Jersey, Piscataway, New Jersey and 2Department of
Biochemistry and Molecular Biology and UNMC Eppley Cancer Center, University
of Nebraska Medical Center, Omaha, Nebraska
Received 2/24/05; revised 6/28/05; accepted 7/8/05.
Grant support: NIH grants CA105331, CA88961, and CA65871 (C.S. Yang).
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Chung S. Yang, Department of Chemical Biology, Ernest
Mario School of Pharmacy, Rutgers University, 164 Frelinghuysen Road,
Piscataway, NJ 08854-8020. Phone: 732-445-5360; Fax: 732-445-0687;
E-mail: csyang@ rci.rutgers.edu.
F 2005 American Association for Cancer Research.
doi:10.1158/1078-0432.CCR-05-0406
www.aacrjournals.org
attempts to reverse abnormal DNA hypermethylation in cancer
cells and restore ‘‘silenced’’ gene expression. However, clinical
utility of these nucleoside analogue DNA methyltransferase
inhibitors has been somewhat limited by myelosuppression
and other side effects (5). We have previously shown the
reactivation of methylation-silenced retinoic acid receptor
h (RARb), p16INK4a (p16), O 6-methylguanine methyltransferase (MGMT), and human mutL homologue 1 (hMLH1)
genes by ( )-epigallocatechin-3-gallate (EGCG), the major
polyphenol from green tea, and this activity has been
attributed to the inhibition of DNA methyltransferase by
EGCG (6). To determine whether other polyphenolic compounds have similar activities, we tested the ability of soy
isoflavones to inhibit DNA methylation.
Genistein (4V,5,7-trihydroxyisoflavone), the major isoflavone
present in soy bean, has been shown to prevent carcinogenesis
in animal models for tumor development at different organ
sites (7). Many mechanisms have been proposed for this
activity. Some are believed to be closely related to the
estrogenic and antiestrogenic activities of genistein (7).
Prepubertal exposure to soy or genistein reduced mammary
carcinogenesis in rats treated with carcinogens, possibly by
modulating the development of the mammary end buds
(8, 9). Various soy products containing genistein have been
found to inhibit the growth of transplanted human prostate
carcinoma, reduce the incidence of poorly differentiated
7033
Clin Cancer Res 2005;11(19) October 1, 2005
Cancer Prevention
prostate adenocarcinoma in a transgenic mouse model, and
inhibit 2-amino-1-methyl-6-phenylimidazo[4,5]pyridine induced rat prostate carcinogenesis (10 – 13). Carcinogenesis or
metastasis in stomach, colon, bladder, and lung is also
inhibited by genistein and related isoflavones (14 – 20).
The effects of genistein on various cancer cell lines have been
extensively studied. It has been reported that genistein can
inhibit cancer cell growth (21 – 23), induce apoptotic cell death
accompanied by cell cycle arrest at G2-M phase, and inhibit
angiogenesis (17, 24 – 28). The precise molecular mechanisms
responsible for these activities, however, are not clearly known.
One potential mechanism that has recently received considerable attention is that genistein may be involved in regulation of
gene activity by modulating epigenetic events such as DNA
methylation and/or histone acetylation directly or through an
estrogen receptor dependent process (29 – 31). This hypothesis
is supported by a report indicating that dietary genistein causes
epigenetic changes in mouse prostate (32). Genistein can also
up-regulate mRNA expression of the BRCA1 gene during
mammary tumorigenesis, which is frequently inactivated by
epigenetic events in breast cancer (8).
In this preclinical study, we examined the effect of genistein
and its combination with other agents on the reactivation of
methylation-silenced RARb, p16, and MGMT genes, as well as
its effect on the activity of DNA methyltransferase and histone
deacetylase (HDAC) in the esophageal squamous cell carcinoma cell line KYSE 510. These genes were selected for study
because they have been shown to be progressively inactivated
by DNA hypermethylation in human esophageal squamous
carcinogenesis and to be reactivated in KYSE 510 cells by 5-azadCyd and EGCG (33 – 35). Our results show that genistein
inhibits DNA methyltransferase and reverses the methylation
status, with accompanying reexpression of RARb, p16, and
MGMT genes. These activities may contribute to the chemopreventive activity of genistein.
Materials and Methods
Cell lines and cell culture
The human esophageal squamous cell carcinoma cell lines KYSE
510 and KYSE 150 were a gift from Dr. Yutaka Shimada (Kyoto
University, Kyoto, Japan). The cells were maintained in 5% CO2
atmosphere at 37jC in RPMI 1640/Ham’s F-12 mixed (1:1) medium
containing 5% fetal bovine serum. To determine the dose-dependent
changes, KYSE 510 cells were treated with 2, 5, 10, or 20 Amol/L of
genistein (Sigma, St. Louis, MO) or 8.7 Amol/L of 5-aza-dCyd (Sigma)
administered in fresh culture medium every other day, when the
medium was changed, for 6 days. For the time-course study, cells were
treated with 5 Amol/L of genistein in fresh culture medium on days 0,
2, 4, and 5. Prostate cancer cell lines LNCaP and PC3 were obtained
from American Type Culture Collection (Manassas, VA) and were
grown in RPMI 1640 containing 10% fetal bovine serum. KYSE 150,
LNCaP, and PC3 cells were treated with 10 or 20 Amol/L of genistein
for 6 days as described above. For the combination study, KYSE 510
cells were treated with either 5 Amol/L of genistein or 2 Amol/L 5-azadCyd for 5 days and then cultured for 1 additional day in fresh
medium with 0.5 Amol/L trichostatin or 15 Amol/L sulforaphane; or
treated with 5 Amol/L of genistein or 2 Amol/L 5-aza-dCyd alone or
together for 5 days and 1 additional day in fresh medium.
Bisulfite modification and methylation-specific PCR
DNA was extracted from the cells using the DNeasy tissue kit
(Qiagen, Valencia, CA) following the procedure of the manufacturer
Clin Cancer Res 2005;11(19) October 1, 2005
and modified by a bisulfite reaction as described by Herman et al.
(36). Primers specific for the methylated and unmethylated DNA
were the same as we previously reported (6). To verify the specificity
of the PCR, we used normal human placental DNA (Sigma) as a
negative control and CpGenome universal methylated DNA (Chemicon, Temecula, CA) as a positive control for methylation-specific
PCR. Methylation-specific PCR was carried out using a nested twostage PCR approach (37) with similar PCR conditions as previously
described (6). In brief, stage I PCR was done on bisulfite-modified
DNA to amplify the CpG-rich promoter regions of the p16, RARb, or
MGMT gene with the primers that recognize the bisulfite-modified
template, but do not discriminate between methylated and unmethylated alleles. DNA fragments from the first PCR were used for
methylation-specific PCR with primers specific for methylated and
unmethylated DNA. Amplification was carried out using a 9700
Perkin-Elmer thermal cycler (Applied Biosystems, Foster City, CA).
Reverse transcription-PCR
Total RNA was isolated from cells using Tri reagent (Sigma). Reverse
transcription of RNA was done using the Advantage RT-for-PCR kit
(Clontech, Palo Alto, CA). The primers and conditions used in the PCR
for the genes analyzed were previously summarized (6). The PCR was
carried out in an approximate linear range (for example, 25 to 28 cycles
for p16, 30 to 35 cycles for RARb, and 30 to 35 cycles for MGMT). The
PCR products were separated on 2% agarose gel containing ethidium
bromide and then photographed. Negative controls for PCR were run
under the same conditions without RNA or reverse transcriptase.
Glyceraldehyde-3-phosphate dehydrogenase (G3PDH) mRNA was used
as an endogenous control.
Real-time PCR
Total RNA was extracted using ABI Prism 6100 Nucleic Acid
PrepStation. Analysis of relative gene expression for RARb, p16, and
MGMT was done using real-time quantitative PCR and the
comparative threshold cycle method (Ct). All reagents, kits, and
instruments used in RNA extraction, reverse transcription, and realtime PCR were purchased from Applied Biosystems. In the reverse
transcription step, 12.5 ng of RNA were reversely transcribed to cDNA
using the High Capacity cDNA Archive Kit following the instruction
manual. In the real-time PCR step, cDNA was amplified with Assayson-Demand Products containing two gene-specific primers for RARb,
p16, and MGMT, respectively, and one TaqMan MGB probe (6-FAM
dye – labeled) using the TaqMan Universal PCR Master Mix in ABI
Prism 7000 Sequencing Detector. Thermal cycling conditions included
2 minutes at 50jC, 10 minutes at 95jC, 40 cycles of 95jC for 15
seconds, and 60jC for 1 minute according to the TaqMan Universal
PCR Protocol. No-template control was included in each assay.
b-Actin was used as an endogenous control and vehicle control was
used as a calibrator. Each sample was run in triplicate. The
comparative Ct method was used to calculate the relative changes in
gene expression in the ABI Prism sequence detection system and
Microsoft Excel as previously described by Livak and Schmittgen (38)
and Ariani et al. (39). The relative changes of gene expression were
calculated using the following formula: Fold change in gene expresDCt (untreated control)]
sion, 2 – DDCt = 2 [DCt (genistein treated samples)
,
where DCt = Ct (detected gene)
Ct (h-actin) and Ct represents
threshold cycle number.
DNA methyltransferase assays
Assays with nuclear extracts as the source of DNA methyltransferase.
Cultured KYSE 510 cells were harvested, and nuclear extracts were
prepared with a nuclear extraction kit (Pierce, Rockford, IL). The
DNA methyltransferase assay was modified in our lab according to
the published methods (40, 41). In brief, the nuclear extracts (4.5 Ag
of protein) were incubated for 1.5 hours at 37jC with 20 nmol/L
(0.75 Ag) of poly(deoxyinosine-deoxycytosine) [poly(dI-dC)poly(dIdC)] and 10 Amol/L of S-adenosyl-L-[methyl-3H]methionine (2 ACi;
7034
www.aacrjournals.org
Reactivation of p16 INK4a , RARb, and MGMT by Genistein
and washed on glass fiber filters before analysis for 3H-methyl by liquid
scintillation counting in aqueous fluor. To measure the effect of genistein
on the rate of substrate methylation, sets of duplicate reactions were
initiated by adding poly(dI-dC)poly(dI-dC) in the presence of
increasing concentrations of EGCG or genistein dissolved in DMSO as
described above.
Cell growth inhibition and other studies
KYSE 510 cells were treated with 1, 2, 5, 10, and 20 Amol/L genistein,
biochenin A, or daidzein added in fresh culture medium every other day
for 6 days or with 5 Amol/L genistein for 1, 2, 4, or 6 days. Living cells
were counted after staining with 0.04% trypan blue. After treatment for 2
or 6 days, the cells were photographed under an inverted microscope for
morphologic analysis. In other experiments, KYSE 510 cells were treated
with 2, 5, 10, and 20 Amol/L genistein for 3 days; cell proliferation was
determined with the BrdUrd ELISA method according to the instruction
manual (Roche Diagnostics Co., Indianapolis, IN) and apoptosis was
determined with fluorescence 4V,6-diamidino-2-phenylindole staining or
caspase-3 immunocytochemistry staining. For colony formation efficiency assay, KYSE 510 cells were treated with 2, 5, 10, or 20 Amol/L
genistein for 2 days and then cultured for 10 additional days in fresh
medium, and then colony numbers were analyzed.
Fig. 1. Alterations of methylation status and mRNA expression levels of RARb,
p16, and MGMT genes after treatment of genistein. Methylation status and mRNA
expression level were determined with methylation-specific PCR and RT-PCR,
respectively. UMSB, unmethylation-specific band; MSB, methylation-specific band.
A , dose- and time-dependent alterations of methylation status and mRNA
expression levels. Left, KYSE 510 cells were treated with 2, 5, 10, or 20 Amol/L
of genistein or 8.7 Amol/L of 5-aza-dCyd (DAC) for 6 days. Right, the cells were
treated for1, 2, 4, and 6 days with 5 Amol/L of genistein. B, alterations of methylation
status and mRNA expression level of RARb gene in KYSE 150, LNCaP, and PC3
cells. Cells were treated with 10 or 20 Amol/L of genistein for 6 days.
Amersham, Piscataway, NJ) in a total volume of 40 AL of reaction
buffer (pH 7.4), containing 20 mmol/L Tris-HCl, 10% glycerol (v/v),
10 mmol/L EDTA, 0.2 mmol/L phenylmethylsulfonylfluoride, and 2
mmol/L 2-mercaptoethanol. Genistein was dissolved in DMSO, and
all incubations were adjusted to contain an equivalent concentration
of DMSO (10%), which had no significant effect on DNA
methyltransferase activity. Reactions were initiated by the addition
of nuclear extracts and stopped by mixing with 200 AL of a solution
containing 1% SDS, 3% 4-aminosalicylate, 5% butanol, 2.0 mmol/L
EDTA, 125 mmol/L NaCl, 0.25 mg/mL carrier salmon testes DNA
(Life Technologies, Inc., Gaithersburg, MD), and 1 mg/mL protease
K. DNA was extracted using the phenol/chloroform method, purified
with ethanol precipitation, and then washed thrice with 70%
ethanol. The radioactivity in the pellets was counted in a scintillation
counter. Each assay was done in duplicate. Background levels were
determined in incubations without the template poly(dI-dC)poly(dIdC). For kinetics, the results were analyzed using GraphPad Prism 4
(GraphPad Software, San Diego, CA).
Assays with purified recombinant DNA methyltransferase 1. Recombinant murine DNA methyltransferase 1 (DNMT1) with an NH2terminal hexahistidine tag added for purification was expressed in
Spodoptera frugiperda 9 cells and purified as previously described (42).
Standard reaction mixtures (25 AL) contained 60 nmol/L recombinant
DNMT1, 15 Amol/L S-adenosyl-L-[methyl-3H]methionine (specific
activity, 275 GBq/mmol), and 10 ng poly(dI-dC)poly(dI-dC) (average
chain length: 3,000) in M2 buffer [100 mmol/L Tris (pH 8), 10 mmol/L
EDTA, 10 mmol/L DTT, 200 Ag/mL bovine serum albumin] and were
incubated for 1 hour at 37jC. Reactions were terminated by the addition
of NaOH to a final concentration of 0.3 mol/L. Salmon sperm DNA (100
Ag/mL) was added as a carrier, and the solution brought to 860 mmol/L
with perchloric acid to precipitate DNA. Precipitated DNA was collected
www.aacrjournals.org
Fig. 2. Quantitative determination of mRNA expression levels of p16, RARb,
or MGMT genes by real-time PCR. KYSE 510 cells were treated with 2, 5,
10, or 20 Amol/L of genistein (G) or 8.7 Amol/L of 5-aza-dCyd for 6 days
and mRNA expression levels were determined using real-time PCR with
TaqMan MGB system. The results were analyzed using a comparative Ct
method. h-Actin was used as an endogenous control. Each sample was run
in triplicate. A , amplification plots of RARb and b-actin are shown on right
and left side, respectively. Y axis, DRn = Rn
baseline (Rn, the normalized
reporter). X axis, amplification cycle number. B, relative quantification for
mRNA expression levels of p16, RARb, or MGMT. Columns, mean (n = 3);
bars, SD. Genistein significantly induced reexpression of p16, RARb, and
MGMT in a concentration-dependent pattern (r = 0.82-0.87, P < 0.0001).
7035
Clin Cancer Res 2005;11(19) October 1, 2005
Cancer Prevention
Fig. 3. Induction of RARh mRNA expression by different isoflavonoids. KYSE 510
cells were treated with different concentrations of genistein, biochenin A,
or daidzein for 6 days, and then the RARh mRNA levels were determined with
RT-PCR. G3PDH was used as an internal control. Representative of three
independent experiments. The band intensity was determined with densitometry;
columns, mean (n = 3); bars, SD. Different letters indicate a significant difference
(P <0.05) based on one-wayANOVA and Tukey’s honest significant difference test.
The effect of genistein on HDAC activity was carried out according to
the instruction manual of the HDAC Assay Kit (Fluorometric Detection,
Upstate Biotechnology, Lake Placid, NY). In brief, the reaction mixture
contained nuclear extract from KYSE 510 cells (4.0 Ag protein), substrate
(CH3-CO-NH-CH2-(CH2)3-Lys-X2-, 48 Amol/L), and genistein (5, 10,
20, 50, or 100 Amol/L) in 40 AL total volume. Trichostatin (1 Amol/L)
was used as a positive control. After incubation for 30 minutes at 25jC,
activator solution was added to release a fluorophore from the
deacetylated substrate. The product was determined by a fluorescence
plate reader (excitation, 360 nm; emission, 465 nm). Each assay was
done in triplicate.
Statistical analyses
Statistical significance between treatment and control groups was
evaluated using the Student’s t test. One-way ANOVA and Tukey’s
honest significant difference test were used for comparing the effects of
different treatments. Pearson’s correlation coefficient (r) with P value
was also determined to examine the association between concentration
and efficacy. All statistical analyses were done using GraphPad Instat.
genes began to appear after 2 days, whereas that from
methylated p16 was only observed after 6 days (Fig. 1A). The
mRNA expression of RARb, p16, and MGMT roughly correlated
with the appearance of the unmethylated PCR products. Under
these conditions, the loss of methylation was not apparent.
Similar results were obtained in a repeated experiment.
To determine whether the reactivation of methylationsilenced genes by genistein is a general phenomenon and occurs
in other cell lines, we studied the effects of genistein treatment on
the methylation status and mRNA levels of RARb in three other
human cancer cell lines (Fig. 1B). The appearance of an
unmethylation-specific band and mRNA band of RARb in KYSE
150 was observed after treating the cells with 10 or 20 Amol/L
genistein for 6 days. In the LNCaP and PC3 cells, unmethylationspecific band and mRNA band of RARb only appeared after
treating the cells with 20 Amol/L genistein for 6 days.
Methylation-specific bands in all three cell lines were slightly
decreased after treatment with 20 Amol/L of genistein. The
results show that the reactivation of methylation-silenced genes
by genistein does occur in different cell lines as a general
phenomenon, although the effective concentrations are somewhat different in the different cell lines.
To confirm the effect of genistein on the reactivation of these
genes, real-time quantitative PCR with TaqMan MGB detector was
employed to determine the mRNA expression levels of RARb,
p16, and MGMT genes in the KYSE 510 cells after treatment with
different concentrations of genistein. The results are summarized
in Fig. 2. The results show that the relative amount of mRNA
expression from these three genes was increased by treatment
with 5 Amol/L of genistein and increased further with higher
concentrations of genistein (10 or 20 Amol/L). However,
genistein was not as effective as 5-aza-dCyd (8.7 Amol/L). This
result was generally consistent with those from general reverse
transcription-PCR (RT-PCR) shown in Fig. 1A.
To compare the effects of different isoflavones from soy on
the reactivation of RARb gene, KYSE 510 cells were treated with
5, 10, or 20 Amol/L of genistein, biochanin A, or daidzein for 6
days and mRNA levels of RARb were determined by RT-PCR
(Fig. 3). All three compounds caused the reexpression of RARh
mRNA, but biochanin A and daidzein were significantly less
effective than genistein (P < 0.05).
Results
Inhibition of DNA methyltransferase and histone
deacetylase activities
Reversal of hypermethylation and reactivation of
RARb, p16, and MGMT by genistein
KYSE 510 cells, with hypermethylated RARb, p16, and
MGMT genes, showed only the methylation-specific bands of
these genes in methylation-specific PCR, with the loss of the
respective mRNA expression, as we previously reported (6, 34).
After treating the cells with 2, 5, 10, or 20 Amol/L genistein for
6 days, the unmethylation-specific bands of these three genes
were detectable at 2 Amol/L, with higher intensity at 5 Amol/L
and even higher intensity at 10 and 20 Amol/L (Fig. 1A). The
expression of mRNA from all three genes increased approximately proportional to the appearance of unmethylated DNA,
whereas PCR products from methylated DNA decreased as
genistein concentration increased. The reversal of hypermethylation and reexpression of these three genes by genistein were
similar to that produced by the classic DNA methyltransferase
inhibitor 5-aza-dCyd. On treating the cells with 5 Amol/L
genistein, PCR products from unmethylated RARb and MGMT
Clin Cancer Res 2005;11(19) October 1, 2005
Genistein exhibited a dose-dependent inhibition of DNA
methyltransferase activity with nuclear extracts from KYSE 510
cells as the enzyme source and poly(dI-dC)poly(dI-dC) as the
substrate. The IC50 of genistein was f67 Amol/L. The structural
analogues of genistein, biochanin A, and daidzein had weaker
inhibitory activities than genistein (Fig. 4A). Genistein also
showed a dose-dependent inhibitory effect on recombinant
DNMT1 activity with an IC50 of 30 Amol/L (Fig. 4B). In kinetic
studies with varying concentrations of poly(dI-dC)poly(dI-dC),
genistein decreased V max and increased K m, showing a mixed
inhibition with a K i c of 189.3 Amol/L for competitive and a K i n
of 80.2 Amol/L for noncompetitive actions (Fig. 4C). With
different concentrations of S-adenosyl-L-[methyl-3H]methionine, genistein decreased V max without a significant change of
K m, suggesting a noncompetitive inhibition with a K i n of 140.5
Amol/L (Fig. 4D). The same conclusion was reached when this
experiment was repeated thrice.
7036
www.aacrjournals.org
Reactivation of p16 INK4a , RARb, and MGMT by Genistein
HDAC activity was assayed using an artificial fluorescence
substrate, and the activity was completely inhibited by 1 Amol/L
trichostatin. Significant inhibition was observed at 5 Amol/L
genistein (13.2%), but the extent of inhibition was low even
at 100 Amol/L genistein (33%; Fig. 5).
Enhanced reactivation of RARb, p16, and MGMT
and growth inhibition by combination of genistein
with trichostatin, sulforaphane, or 2V-deoxy-5-azacytidine
Treatment of KYSE 510 cells with trichostatin (0.5 Amol/L)
alone for 1 day or with 5-aza-dCyd (2 Amol/L) alone for
5 days reactivated the RARb gene, but had little or no effect
on p16 and MGMT (Fig. 6A). However, treatment of the cells
with 5 Amol/L genistein for 5 days and then with 0.5 Amol/L
trichostatin for 1 additional day significantly enhanced the
reactivation of these three genes. In combination with
2 Amol/L 5-aza-dCyd, genistein also significantly enhanced
the reexpression of RARb and p16, but not MGMT. The
reactivation caused by the combination of genistein with
trichostatin or 5-aza-dCyd was higher than the sum of the
effect of individual agents, suggesting a synergistic action. The
combination of 2 Amol/L 5-aza-dCyd and 0.5 Amol/L
trichostatin seemed to produce an additive effect on RARb
and a synergistic effect on p16 and MGMT. Under these
experimental conditions, trichostatin (0.5 Amol/L) was more
effective than genistein (5 Amol/L) in inhibiting cell growth,
and the combination of these two agents caused a more
pronounced inhibition (48%, P < 0.05). Similarly, 15 Amol/L
sulforaphane alone can reactivate the RARb gene, but cannot
induce reexpression of p16 and MGMT genes; however,
treatment of the cells with 5 Amol/L genistein for 5 days and
then with 15 Amol/L sulforaphane for 1 additional day
significantly enhanced the reactivation of these three genes,
especially p16 and MGMT, whereas the combination of
genistein and sulforaphane significantly enhanced the inhibitory effect on cell growth (P < 0.05) compared with genistein
or sulforaphane alone (Fig. 6B).
Effects on inhibition of cell growth and other
factors
After treatment with genistein, biochanin A, or daidzein for
6 days, a dose-dependent inhibition of cell growth was
observed (Fig. 7A). At concentrations as low as 2 Amol/L,
Fig. 4. Inhibition of 5-cytosine DNA methyltransferase activity by genistein. A, dose-dependent inhibition of DNA methyltransferase activity of nuclear extracts from
KYSE 510 cells by genistein, biochanin A, and daidzein. The reaction mixture contained nuclear extracts (4.5 Ag protein), poly(dI-dC)poly(dI-dC) (20 nmol/L), and
S-adenosyl-L-[methyl-3H]methionine (10 Amol/L, 2.0 ACi) in 40 AL incubation mixture containing 10% glycerol and 2 mmol/L 2-mercaptoethanol. The incubation time was
1.5 hours. B, dose-dependent inhibition of recombinant DNMT1activity. Standard reaction mixtures (25 AL) contained 60 nmol/L recombinant DNMT1, 15 Amol/L
S-adenosyl-L-[methyl-3H]methionine (specific activity, 275 GBq/mmol), and 10 ng poly(dI-dC)poly(dI-dC) (average chain length: 3,000) in M2 buffer [100 mmol/L Tris
(pH 8), 10 mmol/L EDTA, 10 mmol/L DTT, 200 Ag/mL bovine serum albumin] and were incubated for 1hour at 37jC. C, kinetic study with varying concentrations of poly
(dI-dC)poly(dI-dC). The reaction mixtures contained 20 Amol/L of S-adenosyl-L-[methyl-3H]methionine and different concentrations of poly(dI-dC)poly(dI-dC). Kin ,
noncompetitive inhibition constant; Kic , competitive inhibition constant. D, kinetic study with varying concentrations of S-adenosyl-L-[methyl-3H]methionine. The reaction
mixtures contained 120 nmol/L poly(dI-dC)poly(dI-dC) and different concentrations of S-adenosyl-L-[methyl-3H]methionine. Points, mean of three sets of the same
experiments; bars, SD. *, P < 0.05; **, P < 0.01, statistically significant difference from control according to Student’s t test. Genistein, biochanin A, and daidzein significantly
inhibited DNA methyltransferase activity of nuclear extracts and recombinant DNA methyltransferase in a concentration-dependent pattern (r = 0.95 to 0.98, P < 0.0001).
www.aacrjournals.org
7037
Clin Cancer Res 2005;11(19) October 1, 2005
Cancer Prevention
Fig. 5. Inhibition of HDAC activity by genistein. The reaction mixture contained
nuclear extract from KYSE 510 cells (4.0 Ag protein), substrate (48 Amol/L), and
genistein (5,10, 20, 50, or100 Amol/L) in 40 AL total volume.Trichostatin (1 Amol/L)
was used as a positive control. Incubation time is 30 minutes at 25jC. The product
was determined by a fluorescence plate reader (excitation, 360 nm; emission, 465
nm). Columns, mean (n = 3); bars, SD. *, P < 0.05; **, P < 0.01, statistically significant
difference from control according to Student’s t test. Genistein inhibited HDAC
activity in a concentration-dependent pattern (r = 0.89, P < 0.0001).
genistein inhibited cell growth by f30%, and at 20 Amol/L, by
more than 90%. The inhibitory effects of biochanin A and
daidzein were weaker, with biochanin A the weakest. In a timedependent study with 5 Amol/L genistein, slight growth
inhibition was seen after 4 days and significant inhibition
was observed after 6 days (Fig. 7A). Signs of toxicity and
growth inhibition were observed in cells treated for 2 or 6 days
with 10 or 20 Amol/L genistein; the greatest effect was observed
in cells treated with 20 Amol/L genistein for 6 days (Fig. 7B). In
cells treated with different concentrations of genistein for
3 days, cell proliferation was significantly inhibited after
treatment with 10 or 20 Amol/L genistein (Fig. 7C), but
induction of apoptosis was not observed even after treatment
with 20 Amol/L genistein (data not shown). In colony
formation assays of KYSE 510 cells, treatment with genistein,
especially at 10 or 20 Amol/L, for 2 days, significantly
decreased the number of colonies formed (Fig. 7D).
RT-PCR results with total RNA of KYSE 510 cells treated with
1, 2, 5, 10, and 20 Amol/L genistein for 6 days indicated that
genistein treatment did not affect the mRNA expression levels of
DNMT1, DNMT3a, DNMT3b, and methyl-CpG binding domain 2. The protein level of 5-methylcytosine DNA glycosylase,
an enzyme that may be involved in the removal of methylated
DNA (43), also did not change after treatment with 10 or
20 Amol/L of genistein for 24 hours (data not shown).
Discussion
The present study clearly shows that genistein reverses DNA
hypermethylation and reactivates the methylation-silenced
genes RARb, p16, and MGMT. This activity is similar to our
previous observations with EGCG. Genistein is also an
inhibitor of DNA methyltransferase activity in nuclear extracts
from KYSE 510 cells, with 14% inhibition at 20 Amol/L (IC50,
f67 Amol/L), and this activity is weaker than that of EGCG
Fig. 6. Effects of combination of genistein and trichostatin, sulforaphane, or 5-aza-dCyd on the reactivation of p16, RARb, and MGMT genes and cell growth. A, KYSE 510
cells were treated with or without either 5 Amol/L genistein or 2 Amol/L 5-aza-dCyd alone or together for 5 days and cultured for 1additional day in fresh medium with or
without 0.5 Amol/L trichostatin (TSA). B, KYSE 510 cells were treated with or without 5 Amol/L genistein for 5 days and cultured for 1additional day in fresh medium with or
without 15 Amol/L sulforaphane (SFN). The mRNA expression levels of p16, RARb, and MGMT genes were determined by RT-PCR and the band intensity was quantified
using densitometry and normalized to each control (mean F SD, n = 2). Representative of two independent experiments. Cell growth was analyzed using trypan blue
exclusion assay. Different letters indicate a significant difference (P < 0.05) based on one-wayANOVA and Tukey’s honest significant difference test (mean F SD, n = 3).
Clin Cancer Res 2005;11(19) October 1, 2005
7038
www.aacrjournals.org
Reactivation of p16 INK4a , RARb, and MGMT by Genistein
Fig. 7. Effects of genistein on cell growth,
morphologic change, cell proliferation, and
colony formation efficiency assay. A , KYSE
510 cells were treated with different
concentrations of genistein, biochenin A,
or daidzein for 6 days, or with 5 Amol/L
genistein for different time periods, and then
their effects on cell growth were determined
by trypan blue exclusion assay. Genistein,
biochanin A, and daidzein significantly
inhibited cell growth with regard to
concentration (r = 0.94 to 0.99,
P < 0.0001) and time (r = 0.65,
P = 0.0017). B, after treatment for 2 days
(a-c) or 6 days (d-f), the morphology of the
cells treated with 0 Amol/L (a and d), 10
Amol/L (b and e), or 20 Amol/L (c and f)
genistein was photographed. Magnification,
100. C, KYSE 510 cells were treated with
different concentrations of genistein for
3 days, and then cell proliferation was
determined with the BrdUrd ELISA method.
Genistein significantly inhibited cell
proliferation in a concentration-dependent
pattern (r = 0.85, P < 0.0001). D, KYSE
510 cells were treated with different
concentrations of genistein for 2 days
and cultured for 10 additional days in fresh
medium, and then colony numbers were
analyzed. Genistein significantly
inhibited colony formation efficiency in a
concentration-dependent pattern
(r = 0.96, P < 0.0001). *, P < 0.05;
**, P < 0.01, statistically significant
difference from control according to
Student’s t test.
(IC50, f20 Amol/L). Kinetic studies indicate that genistein
inhibits DNA methyltransferase activity in a substrate- and
methyl donor – dependent manner, which differs from our
previous results with EGCG as a competitive inhibitor of DNA
methyltransferase. The inhibition activity of genistein was also
observed using recombinant DNMT1 as the enzyme source,
showing an IC50 of 30 Amol/L. In comparison with genistein,
biochanin A and daidzein are weaker inhibitors of DNA
methyltransferase, and they are also less effective in the
reactivation of RARb gene, suggesting a correlation between
the inhibition of DNA methyltransferase activity and reactivation of methylation-silenced genes by these dietary isoflavonoids. However, the fact that the effective concentrations of
genistein in the inhibition of DNA methyltransferase are higher
www.aacrjournals.org
than those for the reactivation of methylation-silenced genes
suggests that other factors are involved in the action of
genistein. Our results, however, suggest genistein did not exert
its effect by modulating the expression levels of DNMT1, 3a,
and 3b, methyl-CpG binding domain 2, and 5-methylcytosine
DNA glycosylase.
For the DNA hypermethylation-related silencing of RARb,
p16, and MGMT genes, histone deacetylation is also known to
be involved (44, 45). In addition to inhibition of DNA
methyltransferase, the presently observed weak inhibitory
effect of genistein on HDAC activity (13-17% inhibition at
5-20 Amol/L) may also contribute to the reactivation of these
genes. The weak inhibitory effects of genistein on DNA
methyltransferase and HDAC may have an intrinsic synergistic
7039
Clin Cancer Res 2005;11(19) October 1, 2005
Cancer Prevention
effect on the reactivation of methylation-silenced genes.
Genistein may also be expected to have a synergistic or an
additive effect with other HDAC or DNA methyltransferase
inhibitors. Indeed, our results (Fig. 6) indicate that genistein,
at 5 Amol/L, enhanced the activity of lower concentrations of
5-aza-dCyd (2 Amol/L) or trichostatin (0.5 Amol/L) in the
reactivation of RARb, p16, and MGMT genes. In addition,
combination of genistein and sulforaphane, a HDAC inhibitor
found in broccoli (46), also enhanced the reactivation of these
genes, especially p16 and MGMT genes. The possible synergistic
actions in the combination of these agents warrant detailed
investigation.
In theory, the demethylation and reactivation of p16 and
RARb by genistein should result in cell growth inhibition.
Genistein (5-20 Amol/L) significantly inhibited cell proliferation without induction of apoptosis and also affected long-term
cell survival after only 2 days of treatment, and this could be
related to the reexpression of p16. More studies, however, are
needed to determine the relationship between the reactivation
of these genes and the growth inhibition by genistein.
Oral administration of soy products containing genistein
and other isoflavones has been shown to inhibit tumorigenesis in different organs, and different mechanisms have been
proposed (7). In our present study, the effective concentration of genistein for DNA demethylation, transcriptional
reactivation of these three genes, and cell growth inhibition
was 2 to 10 Amol/L. These concentrations are lower than
those used in many other studies (26, 47, 48). A concentration
of 10 Amol/L is higher than the plasma level of genistein in
women consuming soy products (0.74-6.0 Amol/L; http://
www.aminoup.co.jp; refs. 49, 50), but lower micromolar
concentrations of genistein are still achievable. Because
genistein is a weak inhibitor of DNA methyltransferase,
genomic global hypomethylation is not expected to occur
due to dietary intake of soy isoflavones. Even if hypomethylation occurred, it would not be to the extent as to induce
genomic instability as has been reported with some DNA
methyltransferase inhibitors (51, 52).
In this study, we showed that genistein and related soy
isoflavones can inhibit DNA methyltransferase activity,
reverse DNA hypermethylation, and reactivate methylationsilenced genes. Genistein in combination with other DNA
methyltransferase or HDAC inhibitors can enhance the
reactivation of methylation-silenced genes. It remains to be
determined whether dietary administration of genistein or
soy products could cause the reactivation of methylationsilenced genes in vivo. An attractive approach that needs to
be explored in vivo is the combination of genistein with
sulforaphane, and low concentrations of these two dietary
factors may produce a synergistic effect in the prevention of
hypermethylation-induced inactivation of tumor suppressor
genes. The combination of genistein with low doses of
trichostatin or 5-aza-dCyd for the prevention or reversal of
hypermethylation also warrants further investigations in vitro
and in vivo.
Acknowledgments
We thank Drs. Jungil Hong, Shengmin Sang, and Charles desBordes for helpful
discussions.
References
1. Herman JG, Baylin SB. Gene silencing in cancer in
association with promoter hypermethylation. N Engl J
Med 2003;349:2042 ^ 54.
2. Egger G, Liang G, Aparicio A, Jones PA. Epigenetics
in human disease and prospects for epigenetic therapy. Nature 2004;429:457 ^ 63.
3. Kopelovich L, Crowell JA, Fay JR. The epigenome as
a target for cancer chemoprevention. J Natl Cancer
Inst 2003;95:1747 ^ 57.
4. Karpf AR, Jones DA. Reactivating the expression of
methylation silenced genes in human cancer. Oncogene 2003;21:5496 ^ 503.
5. Christman JK. 5-Azacytidine and 5-aza-2V-deoxycytidine as inhibitors of DNA methylation: mechanistic
studies and their implications for cancer therapy.
Oncogene 2002;21:5483 ^ 95.
6. Fang MZ, Wang Y, Ai N, et al. Tea polyphenol
( )-epigallocatechin-3-gallate inhibits DNA methyltransferase and reactivates methylation-silenced
genes in cancer cell lines. Cancer Res 2003;63:
7563 ^ 70.
7. Dixon RA, Ferreira D. Genistein. Phytochemistry
2002;60:205 ^ 11.
8. Cabanes A, Wang M, Olivo S, et al. Prepubertal estradiol and genistein exposures up-regulate BRCA1
mRNA and reduce mammary tumorigenesis. Carcinogenesis 2004;25:741 ^ 8.
9. Fritz WA, Coward L, Wang J, Lamartiniere CA. Dietary genistein: perinatal mammary cancer prevention,
bioavailability and toxicity testing in the rat. Carcinogenesis 1998;19:2151 ^ 8.
10. Landstrom M, Zhang JX, Hallmans G, et al. Inhibitory effects of soy and rye diets on the development of
Dunning R3327 prostate adenocarcinoma in rats.
Prostate 1998;36:151 ^ 61.
11. Zhou JR, Gugger ET,TanakaT, Guo Y, Blackburn GL,
Clinton SK. Soybean phytochemicals inhibit the
growth of transplantable human prostate carcinoma
and tumor angiogenesis in mice. J Nutr 1999;129:
1628 ^ 35.
12. Mentor-Marcel R, Lamartiniere CA, Eltoum IE,
Greenberg NM, Elgavish A. Genistein in the diet reduce the incidence of poorly differentiated prostatic
adenocarcinoma in transgenic mice (TRAMP). Cancer
Res 2001;61:6777 ^ 82.
13. Hikosaka A, Asamoto M, Hokaiwado N, et al. Inhibitory effects of soy isoflavones on rat prostate
carcinogenesis induced by 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP). Carcinogenesis
2004;25:381 ^ 7.
14. Li D, Yee JA, McGuire MH, Murphy PA, Yan L.
Soybean isoflavones reduce experimental metastasis
in mice. J Nutr 1999;129:1075 ^ 8.
15. Guo JY, Li X, Browning JD, et al. Dietary
soy isoflavones and estrone protect ovariectomized ERaKO and wild-type mice from carcinogeninduced colon cancer. J Nutr 2004;134:179^82.
16. Arai N, Strom A, RafterJJ, Gustafsson JA. Estrogen
receptor h mRNA in colon cancer cells: growth effects
of estrogen and genistein. Biochem Biophys Res
Commun 2000;270:425 ^ 31.
17. Wietrzyk J, Boratynski J, Grynkiewicz G, Ryczynski
A, Radzikowski C, Opolski A. Antiangiogenic and
antitumour effects in vivo of genistein applied alone
or combined with cyclophosphamide. Anticancer Res
2001;21:3893 ^ 6.
18. Wietrzyk J, Opolski A, Madej J, Radzikowski C. The
antitumor effect of postoperative treatment with genistein alone or combined with cyclophosphamide in
mice bearing transplantable tumors. Acta Pol Pharm
2000;57 Suppl:5 ^ 8.
19. Tatsuta M, Iishi H, Baba M, Yano H, Uehara H,
Nakaizumi A. Attenuation by genistein of sodiumchloride-enhanced gastric carcinogenesis induced by
Clin Cancer Res 2005;11(19) October 1, 2005
7040
N-methyl-NV-nitro-N-nitrosoguanidine in Wistar rats.
Int J Cancer 1999;80:396 ^ 9.
20. Zhou JR, Mukherjee P, Gugger ET, Tanaka T, Blackburn GL, Clinton SK. Inhibition of murine bladder tumorigenesis by soy isoflavones via alterations in the
cell cycle, apoptosis, and angiogenesis. Cancer Res
1998;58:5231 ^ 8.
21. Dean NM, Kanemitsu M, Boynton AL. Effects of
the tyrosine-kinase inhibitor genistein on DNA synthesis and phospholipid-derived second messenger
generation in mouse 10T1/2 fibroblasts and rat liver
T51B cells. Biochem Biophys Res Commun 1989;
165:795 ^ 801.
22. Markovits J, Linassier C, Fosse P, et al. Inhibitory
effects of the tyrosine kinase inhibitor genistein on
mammalian DNA topoisomerase II. Cancer Res 1989;
49:5111 ^ 7.
23. Akiyama T, Ishida J, Nakagawa S, et al. Genistein,
a specific inhibitor of tyrosine-specific protein kinases.
J Biol Chem 1987;262:5592 ^ 5.
24. Li Y, Upadhyay S, Bhuiyan M, Sarkar FH. Induction
of apoptosis in breast cancer cells MDA-MB-231 by
genistein. Oncogene 1999;18:3166 ^ 72.
25. Lian F, Li Y, Bhuiyan M, Sarkar FH. p53-independent apoptosis induced by genistein in lung cancer
cells. Nutr Cancer 1999;33:125 ^ 31.
26. Li Y, Sarkar FH. Inhibition of nuclear factor nB
activation in PC3 cells by genistein is mediated via
Akt signaling pathway. Clin Cancer Res 2002;8:
2369 ^ 77.
27. Shao ZM, Wu J, Shen ZZ, Barsky SH. Genistein
inhibits both constitutive and EGF-stimulated invasion
in ER-negative human breast carcinoma cell lines.
Anticancer Res 1998;18:1435 ^ 9.
28. Shao ZM, Wu J, Shen ZZ, Barsky SH. Genistein
exerts multiple suppressive effects on human breast
carcinoma cells. Cancer Res 1998;58:4851 ^ 7.
www.aacrjournals.org
Reactivation of p16 INK4a , RARb, and MGMT by Genistein
29. Smirnoff P, Liel Y, Gnainsky J, Shany S, Schwartz B.
The protective effect of estrogen against chemically
induced murine colon carcinogenesis is associated
with decreased CpG island methylation and increased
mRNA and protein expression of the colonic vitamin D
receptor. Oncol Res 1999;11:255 ^ 64.
30. Demirpence E, Semlali A, OlivaJ, et al. An estrogenresponsive element-targeted histone deacetylase enzyme has an antiestrogen activity that differs from that
of hydroxytamoxifen. Cancer Res 2002;62:6519 ^ 28.
31. Badia E, Duchesne MJ, Semlali A, et al. Long-term
hydroxytamoxifen treatment of an MCF-7-derived
breast cancer cell line irreversibly inhibits the expression of estrogenic genes through chromatin remodeling. Cancer Res 2000;60:4130 ^ 8.
32. Day JK, Bauer AM, DesBordes C, et al. Genistein
alters methylation patterns in mice. J Nutr 2002;132:
2419 ^ 23S.
33. Xing EP, NieY, Wang LD, et al. Aberrant methylation
of p16INK4a and deletion of p15INK4b are frequent
events in human esophageal cancer in Linxian, China.
Carcinogenesis 1999;20:77 ^ 84.
34. Wang Y, Fang MZ, Liao J, et al. Hypermethylationassociated inactivation of retinoic acid receptor h in
human esophageal squamous cell carcinoma. Clin
Cancer Res 2003;9:5257 ^ 63.
35. Fang MZ, Jin Z, Wang Y, et al. Promoter hypermethylation and inactivation of O (6)-methylguanineDNA methyltransferase in esophageal squamous cell
carcinomas and its reactivation in cell lines. Int J Oncol
2005;26:615 ^ 22.
36. Herman JG, Graff JR, Myohanen S, Nelkin BD,
Baylin SB. Methylation-specific PCR: a novel PCR
www.aacrjournals.org
assay for methylation status of CpG islands. Proc Natl
Acad Sci U S A 1996;93:9821 ^ 6.
37. Palmisano WA, Divine KK, Saccomanno G, et al.
Predicting lung cancer by detecting aberrant promoter methylation in sputum. Cancer Res 2000;60:
5954 ^ 8.
38. Livak KJ, Schmittgen TD. Analysis of relative gene
expression data using real-time quantitative PCR and
the 2( DDC(T)) Method. Methods 2001;25:402 ^ 8.
39. Ariani F, Mari F, Pescucci C, et al. Real-time quantitative PCR as a routine method for screening large
rearrangements in Rett syndrome: report of one case
of MECP2 deletion and one case of MECP2 duplication. Hum Mutat 2004;24:172 ^ 7.
40. Belinsky SA, Nikula KJ, Baylin SB, Issa JP. Increased
cytosine DNA-methyltransferase activity is target-cellspecific and an early event in lung cancer. Proc Natl
Acad Sci U S A 1996;93:4045 ^ 50.
41. Adams RL, Rinaldi A, Seivwright C. Microassay for
DNA methyltransferase. J Biochem Biophys Methods
1991;22:19 ^ 22.
42. Brank AS, Van Bemmel DM, Christman JK. Optimization of baculovirus-mediated expression and
purification of hexahistidine-tagged murine DNA
(cytosine-C5)-methyltransferase-1 in Spodoptera
frugiperda 9 cells. Protein Expr Purif 2002;25:
31 ^ 40.
43. Zhu B, Benjamin D, Zheng Y, et al. Overexpression
of 5-methylcytosine DNA glycosylase in human embryonic kidney cells EcR293 demethylates the promoter of a hormone-regulated reporter gene. Proc
Natl Acad Sci U S A 2001;98:5031 ^ 6.
44. Nguyen CT, Gonzales FA, Jones PA. Altered chro-
7041
matin structure associated with methylation-induced
gene silencing in cancer cells: correlation of accessibility, methylation, MeCP2 binding and acetylation.
Nucleic Acids Res 2001;29:4598 ^ 606.
45. Nakagawachi T, Soejima H, Urano T, et al. Silencing
effect of CpG island hypermethylation and histone
modifications on O 6 -methylguanine-DNA methyltransferase (MGMT) gene expression in human cancer. Oncogene 2003;22:8835 ^ 44.
46. Myzak MC, Karplus PA, Chung FL, et al. A novel
mechanism of chemoprotection by sulforaphane: inhibition of histone deacetylase. Cancer Res 2004;64:
5767 ^ 74.
47. Hsieh CY, Santell RC, Haslam SZ, et al. Estrogenic
effects of genistein on the growth of estrogen
receptor-positive human breast cancer (MCF-7) cells
in vitro and in vivo. Cancer Res 1998;58:3833 ^ 8.
48. Suzuki K, Koike H, Matsui H, et al. Genistein, a soy
isoflavone, induces glutathione peroxidase in the human prostate cancer cell lines LNCaP and PC-3. Int J
Cancer 2002;99:846 ^ 52.
49. Adlercreutz CH, Goldin BR, Gorbach SL, et al. Soybean phytoestrogen intake and cancer risk. J Nutr
1995;125:757 ^ 70S.
50. Xu X, Harris KS, Wang HJ, et al. Bioavailability of
soybean isoflavones depends upon gut microflora in
women. J Nutr 1995;125:2307 ^ 15.
51. Eden A, Gaudet F, Waghmare A, et al. Chromosomal
instability and tumors promoted by DNA hypomethylation. Science 2003;300:455.
52. Gaudet F, Hodgson JG, Eden A, et al. Induction of
tumors in mice by genomic hypomethylation. Science
2003;300:489 ^ 92.
Clin Cancer Res 2005;11(19) October 1, 2005