helpful for more accurate evaluation of their biological
properties. In this context, we tried to develop the new
method for the synthesis of phytoene in mammalian
cells. And, establishment of mammalian cells producing phytoene was succeeded by the introduction of crtB
gene, which encodes phytoene synthase. These cells
were proven to acquire the resistance against carcinogenesis, as well as oxidative stress. Thus, usefulness
of phytoene was proven.
2. Anti-carcinogenic activity of
Among the carotenoids, β-carotene has been expected
to be the most promising candidate as cancer preventive agent. Thus, mainly β-carotene has been tested
for cancer-preventive activity in interventional trials,
i.e., two Linxian trials (Linxian 1 and Linxian 2), the
Alpha-Tocopherol Beta-Carotene (ATBC) Cancer Prevention Study, the β-Carotene and Retinol Efﬁciency
Trial (CARET), the Physicians’ Health Study (PHS)
and the Skin Cancer Prevention Study (SCPS). In
addition to these studies, we have recently completed
the intervention trial with supplementation of the
mixture of natural carotenoids (lycopene, β-carotene,
α-carotene, and others) plus α-tocopherol, and the
analysis of the results is now going on (Jinnno K,
Nishino H, et al. 2002, unpublished information, see
the Section 2.8. Multicarotenoids).
In the Linxian 1 study, a protective effect of supplemental β-carotene, vitamin E, and selenium was reported
with regard to the incidence and mortality rates of gastric cancer when compared with untreated subjects.
In the Linxian 2 study, the relative risk for cancer
mortality was 0.97 in men and 0.92 in women (not
At the end of follow-up in ATBC cancer prevention
study, 894 cases of lung cancer were reported. The
numbers of lung cancer cases by intervention group
were 204 in α-tocopherol, 242 in β-carotene, 240 in
α-tocopherol, plus β-carotene, and 208 in placebo.
The group receiving β-carotene had a 16% higher incidence of lung cancer than those not given β-carotene.
The excess risk associated with β-carotene supplementation was concentrated mainly among people who
currently smoked more than 20 cigarettes per day and
who drank more than 11 g/day of ethanol.
In the CARET, the relative risk of lung cancer incidence was 1.3 in the groups treated with β-carotene
and retinal (p = 0.02), 1.4 in the current smoker
group treated with β-carotene and retinal, and 1.4 in
the asbestos-exposed group treated with β-carotene
In the PHS, no signiﬁcant modiﬁcation in risk was
In the SCPS, the relative risk for skin cancer was 1.4
in the smoker group treated with β-carotene.
In recent studies, we found that α-carotene induced the
G1-arrest in the process of cell cycle . Since various
agents which induce G1-arrest have been proven to
have cancer preventive activity, we evaluated anticarcinogenic activity of α-carotene. α-Carotene showed
higher activity than β-carotene to suppress the tumorigenesis in skin, lung, liver and colon [3,4].
In skin tumorigenesis experiment, two-stage mouse
skin carcinogenesis model was used. Seven-week-old
female ICR mice had their backs shaved with electric clipper. From 1 week after initiation by 100 µg
of 7,12-dimethylbenz[a]anthracene (DMBA), 1.0 µg
of 12-O-tetradecanoylphorbol-13-acetate (TPA) was
applied twice a week. α- or β-Carotene (200 nmol) was
applied with each TPA application. The higher potency
of α-carotene than β-carotene was observed. The percentage of tumor-bearing mice in the control group was
69%, whereas the percentages of tumor-bearing mice
in the groups treated with α- and β-carotene were 25%
and 31%, respectively. The average number of tumors
per mouse in the control group was 3.7, whereas the
α-carotene-treated group had 0.3 tumors per mouse
(p < 0.01, Student’s t-test). β-Carotene treatment also
decreased the average number of tumors per mouse
(2.9 tumors per mouse), but the difference from the
control group was not signiﬁcant.
The higher potency of α-carotene than β-carotene
in the suppression of tumor promotion was conﬁrmed by other two-stage carcinogenesis experiment,
i.e., 4-nitroquinoline 1-oxide (4NQO)-initiated and
glycerol-promoted ddY mouse lung carcinogenesis
model. 4NQO (10 mg/kg body weight) was given
by a single s.c. injection on the ﬁrst experimental
day. Glycerol (10% in drinking water) was given as
tumor promoter from experimental week 5 to week
30 continuously. α- or β-Carotene (at the concentration of 0.05%) or vehicle as a control was mixed as
an emulsion into drinking water during the promotion
stage. The average number of tumors per mouse in
the control group was 4.1, whereas the α-carotenetreated group had 1.3 tumors per mouse (p < 0.001).
β-Carotene treatment did not show any suppressive
effect on the average number of tumors per mouse, but
rather induced slight increase (4.9 tumors per mouse).
In liver carcinogenesis experiment, spontaneous
liver carcinogenesis model was used. Male C3H/He
mice, which have a high incidence of spontaneous
liver tumor development, were treated for 40 weeks
with α- and β-carotene (at the concentration of 0.05%,
mixed as an emulsion into drinking water) or vehicle
as a control. The mean number of hepatomas was
signiﬁcantly decreased by α-carotene treatment as
compared with that in the control group; the control
group developed 6.3 tumors per mouse, whereas the
α-carotene-treated group had 3.0 tumors per mouse
(p < 0.001). On the other hand, the β-carotene-treated
group did not show a signiﬁcant difference from the
control group, although a tendency toward a decrease
was observed (4.7 tumors per mouse).
A short-term experiment to evaluate the suppressive
effect of α-carotene on colon carcinogenesis, effect on
N-methylnitrosourea (MNU, three intrarectal administrations of 4 mg in week 1)-induced colonic aberrant
crypt foci formation (ACF) was examined in Sprague–
Dawley (SD) rats. α- or β-Carotene (6 mg, suspended in
0.2 ml of corn oil, intragastric gavage daily) or vehicle
as control were adminstered during weeks 2 and 5. The
mean number of colonic ACF in control group was 63,
whereas α- or β-carotene-treated group had 42 (significantly lower than that in the control group: p < 0.05)
and 56, respectively. Thus, the greater potency of
α-carotene than β-carotene was again observed.
Lutein is dihydroxy-form of α-carotene, and distributed among variety of vegetables, such as kale, spinach
and winter squash, and fruits, such as mango, papaya,
peaches, prunes and oranges.
Epidemiological study in Paciﬁc Islands indicated
that people with high intake of all three of β-carotene,
α-carotene and lutein had the lowest risk of lung
Thus, the effect of lutein on lung carcinogenesis
was examined. Lutein showed anti-tumor promoting
activity in a two-stage carcinogenesis experiment in
lung of ddY mice, initiated with 4NQO and promoted
with glycerol. Lutein, 0.2 mg in 0.2 ml of mixture of
olive oil and Tween 80 (49 : 1), was given by oral
intubation three times a week during tumor promotion
stage (25 weeks). Treatment with lutein showed a tendency of decrease of lung tumor formation; the control
group developed 3.1 tumors per mouse, whereas the
lutein-treated group had 2.2 tumors per mouse.
The anti-tumor promoting activity of lutein was
conﬁrmed by another two-stage carcinogenesis experiment, i.e., it showed antitumor promoting activity in
a two-stage carcinogenesis experiment in skin of ICR
mice, initiated with DMBA and promoted with TPA
and mezerein. At 1 week after the initiation by 100 µg
of DMBA, TPA (10 nmol) was applied once, and then
mezerein (3 nmol for 15 weeks, and 6 nmol for subsequent 15 weeks) twice a week. Lutein (1 µmol, molar
ratio to TPA = 100) was applied twice (45 min before
and 16 h after TPA application). At the experimental
week 30, average number of tumors per mouse in
the control group was 5.5, whereas the lutein-treated
group had 1.9 tumors per mouse (p < 0.05).
Lutein also inhibited the development of ACF in SD
rat colon induced by MNU (three intrarectal administrations of 4 mg in week 1). Lutein (0.24 mg, suspended
in 0.2 ml of corn oil, intragastric gavage daily) or vehicle as control were administrated during weeks 2–5.
The mean number of colonic ACF in control group
at week 5 was 69, whereas lutein-treated group had
40 (signiﬁcantly lower than that in the control group:
p < 0.05) .
Zeaxanthin is dihydroxy-form of β-carotene, and distributed in our daily foods, such as corn and various
vegetables. Since awareness of zeaxanthin as a beneﬁcial carotenoid is achieved recently, available data for
zeaxanthin are little.
Recently, some features of zeaxanthin were elucidated. For example, zeaxanthin suppressed TPAinduced expression of early antigen of Epstein–Barr
virus in Raji cells. TPA-enhanced 32 Pi-incorporation
into phospholipids of cultured cells was also inhibited
Anti-carcinogenic activity of zeaxanthin in vivo was
also examined. For example, it was found that spontaneous liver carcinogenesis in C3H/He male mice was
suppressed by the treatment with zeaxanthin (at the
concentration of 0.005%, mixed as an emulsion with
Anti-metastatic activity of zeaxanthin was also
Table 1. Effect of lycopene on tumorigenesis in mouse lung
Lycopene occurs in our diet, predominantly in tomatoes
and tomato products.
Recently, the exceptionally high singlet oxygen
quenching ability of lycopene was found [6,7].
Epidemiological study in elderly Americans indicated that high tomato intake was associated with
50% reduction of mortality from cancers at all sites
. And the case-control study in Italy showed potential protection of high consumption of lycopene from
tomatoes against cancers of digestive tract . Inverse
association between high intake of tomato products
and prostate cancer risk was also reported .
Studies on anti-carcinogenic activity of lycopene in
animal models were carried out in mammary gland,
liver, lung, skin and colon [4,11].
The study in mice with a high rate of spontaneous mammary tumors showed that intake of lycopene
delayed and reduced tumor growth.
Lycopene showed anti-tumor promoting activity in
a two-stage carcinogenesis experiment in lung of ddY
mice, initiated with 4NQO and promoted with glycerol. Lycopene, 0.2 mg in 0.2 ml of mixture of olive
oil and Tween 80 (49 : 1), was given by oral intubation three times a week during tumor promotion stage
(25 weeks). Treatment with lycopene resulted in the
signiﬁcant decrease of lung tumor formation; the control group developed 3.1 tumors per mouse, whereas
the lycopene-treated group had 1.4 tumors per mouse
(p < 0.05) (Table 1).
The anti-tumor promoting activity of lycopene was
conﬁrmed by another two-stage carcinogenesis experiment, i.e., it showed antitumor promoting activity in
a two-stage carcinogenesis experiment in skin of ICR
mice, initiated with DMBA and promoted with TPA.
Lycopene (160 nmol, molar ratio to TPA = 100) was
applied with each TPA application. At the experimental week 20, average number of tumors per mouse
in the control group was 8.5, whereas the lycopenetreated group had 2.1 tumors per mouse (p < 0.05).
Lycopene also inhibited the development of ACF
in SD rat colon induced by MNU (three intrarectal
administration of 4 mg in week 1). Lycopene (0.12 mg,
suspended in 0.2 ml of corn oil, intragastric gavage
daily) or vehicle as control were administered during weeks 2–5. The mean number of colonic ACF in
p < 0.05, b p < 0.05.
Male ddY mice were used. 4NQO (10 mg/kg body weight),
dissolved in a mixture of olive oil and cholesterol (20 : 1), was
given by a single s.c. injection on the ﬁrst experimental day.
Glycerol (10% in drinkig water) was given as tumor promoter
from experimental week 5 to week 30 continuously. Lycopene,
0.2 mg in 0.2 ml of mixture of olive oil and Tween 80 (49 : 1),
was given by oral intubation three times a week during tumor
promotion stage (25 weeks).
Male C3H/He mice at the age of 6 weeks were used. Lycopene,
0.005% in drinking water, was given during the whole period
of experiment (40 weeks).
control group at week 5 was 69, whereas lycopenetreated group had 34 (signiﬁcantly lower than that in
the control group: p < 0.05).
Spontaneous liver carcinogenesis in C3H/He male
mice was also suppressed. Treatment for 40 weeks with
lycopene (at the concentration of 0.005%, mixed as an
emulsion into drinking water) resulted in the signiﬁcant
decrease of liver tumor formation as shown in Table 1
(p < 0.005).
β-Cryptoxanthin seems to be a promising carotenoid,
since it showed the strongest inhibitory activity in
the in vitro screening test, i.e., β-cryptoxanthin suppressed TPA-induced expression of early antigen of
Epstein–Barr virus in Raji cells at the highest potency
among carotenoids tested . TPA-enhanced 32 Piincorporation into phospholipids of cultured cells was
also inhibited by β-crypto-xanthin. β-Cryptoxanthin is
distributed in our daily foodstuff, such as oranges, and
is one of the major carotenoids detectable in human
blood. Thus, it seems worthy to investigate more precisely. In this context, we further examined the anticarcinogenic activity in vivo.
β-Cryptoxanthin showed anti-tumor promoting
activity in a two-stage carcinogenesis experiment in
skin of ICR mice, initiated with DMBA and promoted
with TPA. β-Cryptoxanthin (160 nmol, molar ratio
to TPA = 1 : 100) was applied 1 h before each TPA
application. At week 20 of promotion, the percentage
of tumor-bearing mice in the control group was 64%,
whereas the percentages of tumor-bearing mice in the
group treated with β-cryptoxanthin were 29%. The
average number of tumors per mouse in the control
group was 2.7, whereas the β-crypto-xanthin-treated
group had 1.6 tumors per mouse (p < 0.05).
Effect of β-cryptoxanthin on colon carcinogenesis was also carried out. Four groups of F344 rats
(n = 25 each) received an intrarectal dose of 2 mg
MNU, 3 times a week for 5 weeks, and were fed the
diet supplemented with or without β-cryptoxanthin
(0.0025%). The colon cancer incidence at week 30 was
signiﬁcantly lower in the β-cryptoxanthin diet group
(68%) than in the control group (96%). The tumor multiplicity was also lower in the β-crypto-xanthin-treated
group (1.4 tumors per rat) than in the control group
(1.7 tumors per rat), but statistically not signiﬁcant.
2.7. Other carotenoids
In addition to carotenoids mentioned above, fucoxanthin, astaxanthin, capsanthin, crocetin and phytoene seem to be a promising carotenoid, since these
carotenoids showed the strong inhibitory activity in
screening test. Fucoxanthin is distributed in our daily
foodstuff, such as see weeds, and is one of the major
carotenoids which distributed in marine organisms.
Thus, it seems worthy to investigate more precisely.
It is now clear that various natural carotenoids are valuable to apply for cancer prevention. These carotenoids
may be suitable in combinational use, as well as
single use. In fact, we have recently found that multicarotenoids (i.e., mixture of various carotenoids, such
as β-carotene, α-carotene, lutein, lycopene and so on)
showed potent anti-carcinogenic activity.
For example, administration of a prototype multicarotenoids preparation (Table 2) resulted in the suppression of lung tumor promotion, as shown in Table 3.
Furthermore, we have recently proven that administration of natural multicarotenoids (mixture of
lycopene, β-carotene, α-carotene, and other natural
carotenoids) with α-tocopherol resulted in the significant suppression of liver tumor development in liver
Table 2. Composition of multicarotenoids
Table 3. Effect of oral administration of multicarotenoids on
the promotion of lung tumor formation by glycerol in 4NQOinitiated mice
p < 0.05, b p < 0.01.
Multicarotenoids (2 mg in 0.2 ml of oil, i.e., 3 times per week)
was given during the promoting period.
cirrhosis patients (Jinno K, Nishino H et al. patent
pending: #2002–022958, 2002.1.31., unpublished
data). Thus, multicarotenoids seem to be promising
for clinical use.
By the way, it is important that the effectiveness or
efﬁciency of multicarotenoids was different between
individuals, i.e., responders and non-responders were
founded in clinical trial. These difference may be
explained by single nucleotide polymorphisms (SNPs).
Thus, SNPs is now analyzing in responders and
3. Production of phytoene in mammalian cells
3.1. Establishment of phytoene producing
mammalian cells, and analysis of
Phytoene, which is detectable in human blood, was
proven to suppress tumorigenesis in skin. And it was
suggested that antioxidative activity of phytoene may
play an important role in its action mechanism. In
order to conﬁrm the mechanism, more precise study
should be carried out. However, phytoene becomes
to be unstable when it is puriﬁed, and thus, is very
difﬁcult to examine the biological activities. Therefore, stable production of phytoene in target cells,
which may be helpful for the evaluation of their biological properties, was tried. As phytoene synthase
encoding gene, crtB, has already been cloned from
Erwinia uredovora , we used it for the expression
of the enzyme in animal cells. Mammalian expression
plasmids, pCAcrtB, to transfer the crtB gene to mammalian cells, were constructed as follows. First, the
sequence around the initiation codon of the crtB gene
on the plasmid pCRT-B was modiﬁed by PCR using
the primers to replace the original bacterial initiation
codon TTG with CTCGAGCCACCATG, which is a
composite of the typical mammalian initiation codon
ATG preceded by the Kozak consensus sequence and
a XhoI recognition site. The XhoI linker which harbors a cohesive end for the EcoRI site was ligated
to the EcoRI site at the 3 -end of the crtB gene, and
the 969-base pair (bp) XhoI fragment was cloned into
the XhoI site of the expression vector pCAGGS. The
resulting plasmid pCAcrtB drives the crtB gene by
the CAG promoter (modiﬁed chicken b-actin promoter coupled with cytomegalovirus immediate early
enhancer). In the pCAGGS vecter, a rabbit β-globin
polyadenylation signal is provided just downstream of
the XhoI cloning site.
Plasmids were transfected either by electroporation or lipofection. For the gene transfer to NIH3T3
cells, which were cultured in Dulbecco’s modiﬁed
minimum essential medium (DMEM) supplemented
with 4 mM l-glutamin, 80 U/ml penicillin, 80 µg/ml
streptomycin and 10% calf serum (CS), the parameter for electroporation using a Gene Pulser (BioRad)
was set at 1500 V/25mF with a DNA concentration of
12.5–62.5 µg/ml. Lipofection was carried out using
LIPOFECTAMINE (GIBCO BRL) according to the
protocol supplied by the manufacturer. pCAcrtB or
pCAGGS was cotransfected with the plasmid pKOneo,
which harbors a neomycin resistance encoding gene
(kindly provided by Dr. Douglas Hanahan, University
of California, San Francisco).
For Northern blot analysis, 20 µg of total RNA was
loaded onto a 1.2% formaldehyde agarose gel, electrophoresed and transferred to a nitrocellulose ﬁlter
(Nitroplus). The 969-bp XhoI fragment of the crtB gene
as mentioned above was labeled with [32 P]dCTP by the
random primer labeling method and used as a probe to
hybridize the target RNA on the ﬁlter.
NIH3T3 cells transfected with pCAcrtB showed
the expression of a 1.5 kb mRNA from the crtB
gene as a major transcript. That transcript was not
present in the cells transfected with the vector alone
Table 4. Expression of phytoene synthase mRNA and production of phytoene in mammalian cells by introduction of the
crtB gene from Erwinia uredovora
Phytoene synthase mRNA was detected by Northern blot
analysis, and phytoene in cell extracts was measured by HPLC.
For analysis of phytoene by HPLC, the lipid fraction
including phytoene was extracted from cells (107 –108 ).
The sample was subjected to HPLC (column: 3.9 by
300 mm, Nova-pakHR, 6µ C18, Waters) at a ﬂow rate
of 1 ml/min. To detect phytoene, UV absorbance of
the eluate at 286 nm was measured by a UV detector
Phytoene was detected as a major peak in HPLC
proﬁle in NIH3T3 cells transfected with pCAcrtB, but
not in control cells (Table 4). Phytoene was identiﬁed
by UV- and ﬁeld desorption mass-spectra.
Since lipid peroxidation is considered to play a
critical role in tumorigenesis, and it was suggested
that antioxidative activity of phytoene may play an
important role in its mechanism of anticarcinogenic
action, the level of phospholipid peroxidation induced
by oxidative stress in cells transfected with pCAcrtB
or vector alone was compared.
Oxidative stress was imposed by culturing the cells
in a Fe3+ /adenosine 5 -diphosphate (ADP) containing
medium (374 mM iron (III) chloride, 10 mM ADP
dissolved in DMEM) for 4 h. The cells were then
washed three times with a Ca2+ and Mg2+ -free phosphate buffered saline (PBS(−)), harvested by scraping,
washed once with PBS(−), suspended in 1 ml of
PBS(−) and freeze-thawed once. The lipid fraction
was extracted from the disrupted cells twice with
6 ml of chloroform/methanol (2 : 1). The chloroform
layer was collected, dried with sodium sulfate. The
sample was evaporated, and its residue was dissolved
in a small volume of HPLC solvent (2-propanol :
n-hexane : methanol : H2 O = 7 : 5 : 1 : 1) and then subjected to chemiluminescence-HPLC (CL-HPLC). The
lipid was separated with the column (Finepack SIL
NH2-5250 mm × 4.6 mm i.d. JASCO) by eluting with
the HPLC solvent (see above) at a ﬂow rate of 1 ml/min
at 35◦ C. Post column chemiluminescent reaction was
carried out in the mixture of 10 µg/ml cytochrome c and
2 µg/ml luminol in a borate buffer (pH 10.0) at a ﬂow
rate of 1.1 ml/min. To detect lipids, UV absorbance of
the eluate at 210 nm was measured by a UV-8011 detector (TOSOH), and chemiluminescence was detected
with a CLD-110 detector (Tohoku Electric Ind.).
The phospholipid hydroperoxidation level in the
cells transfected with pCAcrtB and conﬁrmed to produce phytoene by HPLC, was lower than that in the
cells transfected with vector alone (Table 5). Thus,
anti-oxidative activity of phytoene in animal cells was
It is of interest to test the effect of the endogenous
synthesis of phytoene on the malignant transformation
process which is newly triggered in noncancerous cells.
Thus, the study was carried out on the NIH3T3 cells
producing phytoene for its possible resistance against
oncogenic insult imposed by transfection of the activated H-ras oncogene. Plasmids with activated H-ras
gene was transfected to NIH3T3 cells with or without
phytoene production, and the rate of transformed focus
formation in 100 mm diameter dishes was compared.
As the results, it was proven that the rate of transformed
focus formation induced by the transfection of activated
H-ras oncogene was lower in the phytoene producing
cells than in control cells (Table 6).
This type of experimental method may be possible
to apply for the evaluation of anti-carcinogenic and/or
anti-oxidative activity of other phytochemicals, since
the cloning of genes for synthesis of various kinds of
substances in vegetables and fruits, has already been
accomplished. It is particularly useful to evaluate biological activity of unstable phytochemicals, such as
phytoene and other unstable carotenoids.
Table 5. Reduction of oxidative stress induced lipid
hydroperoxidation levels in cells producing phytoene
PEOOH/PC + PE
PCOOH: phosphatidylcholine hydroperoxide; PEOOH:
phosphatidylethanolamine hydroperoxide; PC: phosphatidylcholine; PE: phosphatidylethanolamine.
Table 6. Suppression of transformed focus formation
induced by activated H-ras gene in cells producing
Number of transformed foci
Valuable chemopreventive substances, including
carotenoids, may be produced in wide variety of foods
by means of bio-technology; this kind of new concept
may be named as ‘bio-chemoprevention’. As a prototype experiment, phytoene synthesis in animal cells is
demonstrated as described above. Since phytoene produced in animal cells was proven to prevent oxidative
damage of cellular lipids, it may become a valuable
factor in animal foods to reduce the formation of oxidized oils, which may be carcinogenic and hazardous
for health, as well as to keep freshness, resulting in
the maintenance of safety and good quality of foods.
Furtheremore, phytoene-containing foods are valuable to prevent cancer, since phytoene is known as an
Recently, we have succeeded to produce phytoene
in mice (Satomi Y, Nishino H et al. 2002, unpublished
data), and now analyze their properties. In the next step,
we are planning to produce phytoene in pigs and cows.
Various natural carotenoids were proven to have anticarcinogenic activity, and seem to be useful for cancer
control in human. Some of them may also be applicable
for bio-chemoprevention project.
5. Key unanswered questions
Action mechanism of natural carotenoids should be
Assessment of usefulness of these natural
carotenoids in human cancer control project should be
carried out more extensively.
Taylor made-cancer prevention, including counterplan for non-responders to carotenoids, should
be developed. For this purpose, SNPs analysis and
comparison of SNPs between responders and nonresponders should be carried out systematically.
In any case, establishment of the most effective and
safe combination recipes of natural carotenoids for
each individual person is the most important problem.
This work was supported in part by grants from the
Program for Promotion of Basic Research Activities
for Innovative Biosciences (ProBRAIN), the Ministry
of Agriculture, Forestry, and Fisheries, the Ministry
of Health and Welfare, the Ministry of Education, Science and Culture, and Institute of Free Radical Control
(IFRC), Japan. The study was carried out in collaboration with research groups of Kyoto Prefectural University of Medicine, Akita University College of Allied
Medical Science, Kyoto Pharmaceutical University,
Food Research Institute, Fruit Tree Research Station,
National Cancer Center Research Institute, Shikoku
Cancer Center, Lion Co., Dainippon Ink & Chemicals, Inc., Kagome Co., Kirin Brewery Co., Koyo
Mercantile Co., Japan, Dr. Frederick Khachik, Department of Chemistry and Biochemistry, University of
Maryland, USA, and Dr. Zohar Nir, LycoRed Natural
Products Industries, Ltd., Israel. Authors are grateful
to Dr. Takashi Sugimura, Emeritus President, National
Cancer Center, Japan, for his kind encouragement
during this study.
1. Peto R, Doll R, Buckley JD, Sporn MB: Nature 290:
2. Murakoshi M, Takayasu J, Kimura O, Kohmura E,
Nishino H, Iwashima A, Okuzumi J, Sakai T, Sugimoto T,
Imanishi J, Iwasaki R: J Natl Cancer Inst 81: 1649–1652,
3. Murakoshi M, Nishino H, Satomi Y, Takayasu J,
Hasegawa T, Tokuda H, Iwashima A, Okuzumi J, Okabe H,
Kitano H, Iwasaki R: Cancer Res 52: 6583–6587, 1992
4. Narisawa T, Fukaura Y, Hasebe M, Ito M, Aizawa R,
Murakoshi M, Uemura S, Khachik F, Nishino H: Cancer
Lett 107: 137–142, 1996
5. Le Marchand L, Hankin JH, Kolonel LN, Beecher GR,
Wilkens LR, Zhao LP: Cancer Epidemiol Biomarkers Prev
2: 183–187, 1993
6. Stahl W, Sies H: Ann NY Acad Sci 691: 10–19, 1993
7. Ukai N, Lu Y, Etoh H, Yagi A, Ina K, Oshima S, Ojima F,
Sakamoto H, Ishiguro Y: Biosci Biotech Biochem 58:
8. Colditz GA, Branch LG, Lipnick RJ: Am J Clin Nutr 41:
9. Franceschi S, Bidoli E, La Veccia C, Talamini R,
D’Avanzo B, Negri E: Int J Cancer 59: 181–184, 1994
10. Giovannucci E, Ascherio A, Rimm EB, Stampfer MJ,
Colditz GA, Willett WC: J Natl Cancer Inst 87: 1767–1776,
11. Nagasawa K, Mitamura T, Sakamoto S, Yamamoto K: Anticancer Res 15: 1173–1178, 1995
12. Tsushima M, Maoka T, Katsuyama M, Kozuka M,
Matsuno T, Tokuda H, Nishino H, Iwashima A: Biol Pharm
Bull 18: 227–233, 1995
13. Misawa N, Nakagawa M, Kobayashi K, Yamano S, Izawa Y,
Nakamura K, Harashima K: J Bacteriol 172: 6704–6712,
Address for offprints: Hoyoku Nishino, Department of Biochemistry,
Kyoto Prefectural University of Medicine, Kawaramachi-Hirokoji,
Kamigyoku, Kyoto 602-8566, Japan; e-mail: firstname.lastname@example.org