Formation of 2-chloroinosine from guanosine by treatment of HNO2 in the presence of NaCl

Article abstract of DOI:10.1016/S0968-0896(01)00173-0

We investigated the reaction of Guo with nitrous acid in the presence of NaCl. When 1 mM Guo was incubated with 100 mM NaNO₂ and 2 M NaCl in sodium acetate buffer at pH 3.2 and 37°C, 2-chloroinosine (2-Cl-Ino) was produced in addition to oxanosine (Oxo) and xanthosine (Xao). The yield of 2-Cl-Ino was 0.033 mM at an incubation time of 2 h. Under the same reaction conditions, GMP and dGuo gave rise to the corresponding 2-chloro derivatives with comparable yields. All the 2-chloro derivatives were fairly stable (t_1/2>360 h) at physiological pH and temperature. To elucidate the reaction mechanism of the chlorination, the diazoate derivative of Guo, a reaction intermediate of the Guo-HNO₂ system, was employed as a starting compound. When the diazoate was incubated with 2 M NaCl in a neutral solution, 2-Cl-Ino was produced in addition to Oxo and Xao. These results suggest that the 2-chloro derivatives can be produced from foodstuffs in the human stomach and may have potential importance as a carcinogen causing gastric cancer.

Full text of DOI:10.1016/S0968-0896(01)00173-0

Bioorganic & Medicinal Chemistry 9 (2001) 2937–2941
Formation of 2-Chloroinosine from Guanosine by Treatment of
HNO2 in the Presence of NaCl
Toshinori Suzuki,a Hiroshi Ide,b Masaki Yamada,a Takashi Moriia
and Keisuke Makinoa,*
aInstitute of Advanced Energy, Kyoto University, Gokasho, Uji 611-0011, Japan
bDepartment of Mathematical and Life Sciences, Graduate School of Science, Hiroshima University, Kagamiyama,
Higashi-Hiroshima 739-8526, Japan
Received 4 January 2001; accepted 12 May 2001
Abstract—We investigated the reaction of Guo with nitrous acid in the presence of NaCl. When 1mM Guo was incubated with 100
mM NaNO2 and 2 M NaCl in sodium acetate buffer at pH 3.2 and 37 ꢀC, 2-chloroinosine (2-Cl-Ino) was produced in addition to
oxanosine (Oxo) and xanthosine (Xao). The yield of 2-Cl-Ino was 0.033 mM at an incubation time of 2 h. Under the same reaction
conditions, GMP and dGuo gave rise to the corresponding 2-chloro derivatives with comparable yields. All the 2-chloro derivatives
were fairly stable (t1=2>360 h) at physiological pH and temperature. To elucidate the reaction mechanism of the chlorination, the
diazoate derivative of Guo, a reaction intermediate of the Guo–HNO2 system, was employed as a starting compound. When the
diazoate was incubated with 2 M NaCl in a neutral solution, 2-Cl-Ino was produced in addition to Oxo and Xao. These results
suggest that the 2-chloro derivatives can be produced from foodstuffs in the human stomach and may have potential importance as
a carcinogen causing gastric cancer. # 2001Elsevier Science Ltd. All rights reserved.
Introduction
gastric cancer.6 Neither is there any laboratory evidence
whatsoever to indicate that salt per se is a carcinogen.6
Despite the long-standing and widespread reduction in
the incidence and mortality rates, gastric cancer is still
one of the most common neoplastic causes of death
world-wide.1 Based on the observation that gastric can-
cer mortality and stroke mortality were strongly corre-
lated, a hypothesis was presented in 1969 that salt could
be involved in the aetiology of gastric cancer.2 There-
after, a large number of epidemiological and laboratory
studies relevant to the assessment of salt for carcino-
genic potential were reported. Although many case
control studies found a positive association between
high salt intake and the stomach cancer incidence, the
risk factor tends to show a great variability in different
population.3 Animal studies showed that salt had co-
carcinogenic and promoting effects when administered
with an N-nitrosoguanidine, a carcinogen.4,5 Recently,
Cohen and Roe have reviewed these findings and con-
cluded that there is no reliable epidemiological evidence
to indicate that dietary salt affects the incidence of
Another potential risk factor for gastric cancer is nitrite
(NOꢁ2 ). The reaction of NOꢁ2 with biomolecules has
been well investigated. NOꢁ2 is protonated and results in
nitrous acid (HNO2) in the stomach. HNO2 produces
mutagenic N-nitroso compounds from secondary
amines in foodstuff.7 HNO2 can also react directly with
components of nucleic acids such as Guo, Ado, and
Cyd, generating deamination products, that is xantho-
sine (Xao), inosine (Ino), and uridine (Urd), respec-
tively.8 Among these nucleosides, Guo reacts with
HNO2 more rapidly than Ado and Cyd9 and gives rise
to several products including oxanosine (Oxo) which is
the
second
major
product
and
potentially
mutagenic.10ꢁ12 In addition to direct ingestion of NO2ꢁ
as a content of foodstuffs or a contaminant of drinking
water, NOꢁ2 is generated in the saliva from ingested
NOꢁ3 by nitrate-reducing microorganisms.13,14 Since the
amount of NOꢁ2 generated from NOꢁ3 in vivo is 2-fold
greater than that ingested directly, dietary intake of
NOꢁ3 can be important for gastric cancer. Many epi-
demiological and laboratory studies have been reported
to assess carcinogenic potential of NOꢁ3 and NO2ꢁ.
*Corresponding author. Tel.: +81-774-38-3517; fax: +81-774-38-
3524; e-mail: kmak@iae.kyoto-u.ac.jp
0968-0896/01/$ - see front matter # 2001Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(01)00173-0

2938
T. Suzuki et al. / Bioorg. Med. Chem. 9 (2001) 2937–2941
Case–control studies showed both positive and negative
association between NOꢁ3 or NO2ꢁ and gastric cancer.15
In certain animal studies, negative association were
reported.16,17 Therefore, although many efforts have
been made to determine whether or not high intake of
NOꢁ3 or NOꢁ2 is a cause of gastric cancer, it remains
inconclusive.15,18ꢁ20
(Oxo) (tR=10.5 min, lmax=245 and 287 nm), xantho-
sine (Xao) (tR=7.4 min, lmax=248 and 277 nm), and
unreacted Guo in the RP-HPLC analysis. Formation of
the product with tR=11.6 min was not observed in a
similar reaction without NaCl. Figure 1inset shows the
UV spectrum of the product. The product was isolated
by preparative RP-HPLC. Negative ion atmospheric
pressure chemical ionization (APCI-) LC/MS of this
product showed signals with m/z=303 and 301with
relative intensities of 1:3. When authentic 2-chlor-
oadenosine (2-Cl-Ado) was incubated with 100 mM
NaNO2 in 3 M acetate buffer (pH 3.7) at 37 ꢀC for 24 h,
a major peak with the same retention time (tR=11.6
min) and UV spectrum (lmax=256 nm) as the unknown
product appeared in the RP-HPLC chromatogram.
APCI-LC/MS analysis of the product formed from
2-Cl-Ado also showed signals with m/z=303 and 301(1:3).
It has been reported that 2-chloro-20-deoxyadenosine (2-
Cl-dAdo) is deaminated by HNO2 resulting in 2-chloro-20-
deoxyinosine (2-Cl-dIno).22 Combining these data, we
have concluded that the product is 2-chloroinosine (2-
Cl-Ino) (Fig. 2). Figure 3 shows the time courses of the
concentration changes in 2-Cl-Ino, Oxo, Xao, and Guo
under the above reaction conditions. The yields of
reaction products at the incubation time of 2 h were
3.3% 2-Cl-Ino, 13.4% Oxo, 68.9% Xao, and 9.7%
unreacted Guo. The yield of 2-Cl-Ino was found to be
proportional to the NaCl concentration: at 100 mM
(plasma concentration) and 500 mM, the yields were
0.15 and 0.76%, respectively.
Recently, an ecological study which examined the indi-
vidual importance of salt and NOꢁ3 intake was repor-
ted.21 The analysis showed that the importance of NO3ꢁ
in the stomach cancer mortality increased markedly
with a high sodium level. However, there is no study on
the molecular basis for the possible cooperative effect of
HNO2 and NaCl on carcinogenesis.
We report herein the formation of 2-chloroinosine (2-
Cl-Ino) from Guo with HNO2 in the presence of NaCl,
and discuss the reaction mechanism of the chlorination
based on the reactivity of a diazoate derivative of Guo.
Results
Guo (1mM) was incubated with 100 mM NaNO and
2
2 M NaCl in 3 M sodium acetate buffer at pH 3.2 and
37 ꢀC. Figure 1shows the reversed phase (RP)-HPLC
chromatogram of the reaction mixture at the incubation
time of 2 h. An unknown product eluted [retention time
(tR)=11.6 min, lmax=256 nm] in addition to oxanosine
GMP was also treated with HNO2 and NaCl under
similar reaction conditions and the reaction was mon-
itored by RP-HPLC. GMP produced 2-chloroinosine
50-monophosphate
l
sphate (OMP) (tR=9.1min,
and xanthosine 50-monophosphate (XMP) (tR=6.3 min,
(2-Cl-IMP)
(tR=10.3
min,
max=256 nm) in addition to oxanosine 50-monopho-
lmax=245 and 287 nm)
lmax=249 and 276 nm). Identification of each product
was achieved by alkaline phosphatase digestion, which
dephosphorylated 2-Cl-IMP, OMP, and XMP to 2-Cl-
Ino, Oxo, and Xao, respectively. The yields of reaction
products at the incubation time of 2 h were 3.8% 2-Cl-
IMP, 15.5% OMP, 86.1% XMP, and 2.4% unreacted
GMP. Their yields were essentially comparable to those
for Guo.
Figure 1. RP-HPLC chromatogram for the reaction mixture of Guo
treated by HNO2 in the presence of NaCl. Guo (1mM) was incubated
with 100 mM NaNO2 and 2 M NaCl in 3 M sodium acetate buffer at
pH 3.2 and 37 ꢀC for 2 h. An HPLC system equipped with an ODS
column (4.6ꢂ150 mm) was used. The eluent was 100 mM triethy-
lammonium acetate (TEAA) buffer (pH 7.0) containing CH3CN and
the flow rate was 1mL/min. The CH 3CN concentration was increased
from 0 to 20% for 20 min in a linear gradient mode. Inset is an on-line
detected UV spectrum of 2-Cl-Ino.
A similar reaction was performed with dGuo. dGuo
produced 2-Cl-dIno (tR=12.9 min, lmax=256 nm) in
addition to 20-deoxyoxanosine (dOxo) (tR=12.1 min,
l
max=246 and 287 nm), 20-deoxyxanthosine (dXao)
Figure 2. Reaction products of Guo treated by HNO2 in the presence of NaCl. The numbers in the parentheses denote percentage yields of products
at the reaction time of 2 h. R=ribose.

T. Suzuki et al. / Bioorg. Med. Chem. 9 (2001) 2937–2941
2939
(tR=8.4 min, lmax=248 and 277 nm), and xanthine
(Xan) (tR=6.3 min, lmax=267 nm). Identification of 2-
Cl-dIno was achieved by negative ion APCI-LC/MS
analysis, which showed peaks of m/z=287 and 285 with
relative intensities of 1:3. RP-HPLC analysis also
revealed that HNO2-treated 2-Cl-dAdo gave a peak
with the same tR and UV spectrum of the product
formed from dGuo. The yields of reaction products at
the incubation time of 2 h were 3.4% 2-Cl-dIno, 12.6%
dOxo, 43.2% dXao and 30.0% Xan with 10.4%
unreacted dGuo.
of the Guo-diazoate (20 mL) was directly dropped into
2.5 M NaCl solution containing 500 mM phosphate
buffer of (pH 7.4, 80 mL). The final NaCl concentration
of the reaction mixture is 2 M. The reaction mixture was
incubated at 37 ꢀC for 1h and analyzed by RP-HPLC.
As shown in Figure 4, 2-Cl-Ino was produced in
addition to Oxo and Xao in the Guo-diazoate–NaCl
system. The relative yield of the products was 2-Cl-
Ino–Oxo–Xao=1:3.5:13.9. The ratio was similar to
that for the Guo–HNO2–NaCl system (2-Cl-Ino–Oxo–
Xao=1:4.0:20.7) (Fig. 5).
To examine the stability of the 2-chloro derivatives,
purified 2-Cl-Ino, 2-Cl-IMP, and 2-Cl-dIno (all 25 mM)
were incubated in 100 mM sodium phosphate buffer
under physiological conditions (pH 7.4, 37 ꢀC) up to 92 h
and analyzed by RP-HPLC. Although 2-Cl-IMP
underwent dephosphorylation slowly resulting in 2-Cl-
Ino exclusively (t1=2=360 h), 2-Cl-Ino and 2-Cl-dIno
remained intact after 92 h (data not shown).
Discussion
In the present study, it was shown that 2-Cl-Ino was
produced by the incubation of Guo with NaNO2 and
NaCl under mild acidic conditions. Similarly, 2-Cl-IMP
and 2-Cl-dIno were produced from GMP and dGuo,
respectively. The materials required for the reactions
(nucleic acid components, NaNO2, and NaCl) can be
supplied by ingested foods. In addition, the saliva sup-
plies additional NOꢁ2 at levels ranging from several to
several hundreds mg/L13,14 and the secreted gastric acid
solution contains about 160 mM of hydrochloric acid.24
These suggest that 2-chloro derivatives can be formed in
the human stomach. Since 2-chloro derivatives were
fairly stable under physiological conditions, it would
remain intact in the intestines and be absorbed. If the
absorbed 2-chloro derivatives are converted to the
deoxyribonucleoside triphosphate by salvage and/or de
novo pathways of purine nucleotide biosynthesis and
are incorporated into DNA by DNA polymerase, it
may have an important role in mutagenic or lethal
events in cells. Inosine 50-monophosphate dehydro-
genase (IMPDH), which catalyzes the oxidation of IMP
to XMP, plays an important role in the purine meta-
bolic pathways. Recently, it has been shown that human
type II IMPDH catalyzes dehalogenation of 2-Cl-IMP
resulting in XMP and Clꢁ.25 Since the turnover rate of
2-Cl-IMP was one-third of that observed for IMP, 2-Cl-
IMP was a reasonably good substrate of IMPDH. The
IMPDH was not inhibited irreversibly by 2-Cl-IMP.
The report on IMPDH implies that the 2-chloro
To elucidate the reaction mechanism for the formation
of 2-Cl-Ino in the Guo–HNO2–NaCl system, a diazoate
derivative of Guo (Guo-diazoate) was used. The RP-
HPLC fraction containing the Guo-diazoate was pre-
pared as described previously.23 The RP-HPLC fraction
Figure 4. RP-HPLC chromatogram for the reaction mixture of the
diazoate derivative of Guo incubated in the presence of NaCl. The
RP-HPLC fraction containing the diazoate (20 mL) was directly drop-
ped into 2.5 M NaCl solution (80 mL) containing 500 mM phosphate
buffer (pH 7.4) and incubated at 37 ꢀC for 1h. RP-HPLC conditions
were the same as those described in Figure 1.
Figure 3. The time courses of formation of 2-Cl-Ino (open triangle),
Oxo (closed triangle), Xao (closed square) and conversion of Guo
(closed circle) when Guo (1mM) was incubated with 100 mM NaNO
2
and 2 M NaCl in 3 M sodium acetate buffer at pH 3.2 and 37 ꢀC. The
concentration was determined by RP-HPLC analysis.

2940
T. Suzuki et al. / Bioorg. Med. Chem. 9 (2001) 2937–2941
Figure 5. Reaction products formed from the diazoate derivative of Guo (Guo-diazoate) in the presence of NaCl. The numbers in the parentheses
denote relative yields of products.
Experimental
derivatives may exist in human bodies and human cells
have a protection system to the 2-chloro derivatives.
Materials
The reaction of protonated NOꢁ2 (i.e., HNO2) is com-
plicated.26,27 Protonated NOꢁ2 generates reactive nitro-
gen species such as N2O3 and NO+, both of which
nitrosate amines. A nitrosated primary amine is con-
verted to hydroxylated compounds (R–OH) via several
intermediates such as diazohydroxide (R–N=N–OH),
diazoate (R–N=N–Oꢁ), diazonium (R–NꢃN+), and
alkyl cation (R+). The reaction between chloride (Clꢁ)
and the intermediates from Guo can result in 2-Cl-Ino.
In addition to N2O3 and NO+, nitrosyl chloride
(NOCl), another reactive nitrogen species, could be
produced when NOꢁ2 is protonated in the presence of
Clꢁ. NOCl nitrosates amines and chlorinates aromatic
compounds. Moreover, it is reported that in con-
centrated hydrochloric acid, NOꢁ2 facilitates the forma-
tion of Cl2, which can also chlorinate aromatic
compounds. The reaction between these reactive chlor-
inating species and Guo can lead to the formation of
2-Cl-Ino. Recently we have isolated an intermediate
formed in the reaction of Guo with HNO2 by RP-
HPLC.23 The intermediate was identified as a diazoate
derivative of Guo with a half-life of 5.6 min at pH 7.0
and 20 ꢀC. This intermediate turned out to be very use-
ful to study the reaction mechanism of the Guo–HNO2
system since various reaction conditions could be
employed. In the present study, when the diazoate deri-
vative was incubated with NaCl in neutral solution,
2-Cl-Ino was produced. This result clearly shows that
2-Cl-Ino is formed by the reaction between chloride
anion (Clꢁ) and the cation formed from the diazonium
derivative or its precursor (diazonium derivative).
Guo was obtained from Kohjin (Tokyo, Japan), and
dGuo from Wako Pure Chemicals (Osaka, Japan).
GMP, 2-chloroadenosine (2-Cl-Ado), 2-chloro-20-deoxy-
adenosine (2-Cl-dAdo), and xanthosine (Xao) were
purchased from Sigma (St. Louis, MO, USA). Authen-
tic oxanosine (Oxo), 20-deoxyoxanosine (dOxo), and
20-deoxyxanthosine (dXao) were prepared from Guo
and dGuo by HNO2 treatment, respectively, as reported
previously.10,23 All other chemicals of reagent grade
were purchased from Wako Pure Chemicals or Nacalai
Tesque (Osaka, Japan), and used without further puri-
fication. Water was purified with a Millipore Milli-QII
deionizer.
RP-HPLC conditions
The HPLC system consisted of Tosoh DP-8020 pumps
and a PX-8020 system controller. On-line UV spectra
were obtained with a Tosoh PD-8020 UV–vis photo-
diode-array detector. For reversed phase (RP-) HPLC,
a YMC-Pack ODS-A octadecylsilane column (4.6ꢂ150
mm, particle size 5 mm) was used. The eluent was 100
mM triethylammonium acetate (TEAA) buffer (pH 7.0)
containing CH3CN. The CH3CN concentration was
increased from 0 to 20% for 20 min in a linear gradient
mode. HPLC analysis was performed at room temperature
with a flow rate of 1mL/min.
Spectrometric measurements
Negative-ion atmospheric pressure chemical ionization
(APCI-) LC mass spectra were obtained by a Hitachi
M-2000 MS system. The sample was directly injected
into the MS system by an HPLC pump without separa-
tion columns. The following LC/MS conditions were
used: eluent, 100% CH3CN (isocratic); flow rate, 1mL/
min; vaporization temperature, 300 ꢀC; desolvation
temperature, 340 ꢀC; drift voltage, ꢁ195 V.
Conclusion
It has been found that 2-Cl-Ino, 2-Cl-IMP and 2-Cl-
dIno are formed from Guo, GMP and dGuo, respec-
tively, by HNO2 treatment in the presence of NaCl. The
intermediates involved in the reaction have been eluci-
dated as the diazoate derivatives which are generated
from the starting compounds by HNO2. These results
suggest that the 2-chloro derivatives can be produced
in the human stomach and may play unique role in
carcinogenesis.
Quantitative procedures
The concentrations of products were evaluated from
integrated peak areas of the HPLC chromatogram and
the molar extinction coefficients at 260 nm. The e260
values of 1.18ꢂ104 Mꢁ1 cmꢁ1 for 2-Cl-Ino, 2-Cl-IMP,
and 2-Cl-dIno, 5.1ꢂ103 Mꢁ1 cmꢁ1 for Oxo, OMP, and
dOxo, 7.8ꢂ103 Mꢁ1 cmꢁ1 for Xao, XMP, and dXao,

T. Suzuki et al. / Bioorg. Med. Chem. 9 (2001) 2937–2941
2941
8.5ꢂ103 Mꢁ1 cmꢁ1 for Xan, and 1.15ꢂ104 Mꢁ1 cmꢁ1
for Guo, GMP, and dGuo were used.28ꢁ30 In the quan-
titative analysis of the reaction products, the initial RP-
HPLC peak area was used as a standard.
9. Shapiro, R.; Pohl, S. H. Biochemistry 1968, 7, 448.
10. Suzuki, T.; Yamaoka, R.; Nishi, M.; Ide, H.; Makino, K.
J. Am. Chem. Soc. 1996, 118, 2515.
11. Suzuki, T.; Matsumura, Y.; Ide, H.; Kanaori, K.; Tajima,
K.; Makino, K. Biochemistry 1997, 36, 8013.
12. Suzuki, T.; Yoshida, M.; Yamada, M.; Ide, H.; Kobaya-
shi, M.; Kanaori, K.; Tajima, K.; Makino, K. Biochemistry
1998, 37, 11592.
13. Tannenbaum, S. R.; Sinskey, A. J.; Weisman, M.; Bishop,
W. J. Natl. Cancer Inst. 1974, 53, 79.
14. Tannenbaum, S. R.; Weisman, M.; Fett, D. Food Cosmet.
Toxicol. 1976, 14, 549.
Enzymatic digestion of nucleotides
The solution of 2-Cl-IMP, OMP, and XMP was incu-
bated in 0.5 mL of 100 mM Tris–HCl buffer (pH 7.0)
ꢀ
containing alkaline phosphatase (1unit) at 37 C for 4 h.
Aliquots of the solution were analyzed by RP-HPLC.
The chromatograms of the digested 2-Cl-IMP, OMP,
and XMP showed the peaks of the corresponding
nucleosides (2-Cl-Ino, Oxo, and Xao, respectively).
15. Forman, D. Cancer Surv. 1989, 8, 443.
16. Maekawa, A.; Ogiu, T.; Onodera, H.; Furuta, K.; Mat-
suoka, C.; Ohno, Y.; Odashima, S. Food Chem. Toxicol. 1982,
20, 25.
17. Inai, K.; Aoki, Y.; Tokuoka, S. Gann 1979, 70, 203.
18. Forman, D.; Al-Dabbagh, S.; Doll, R. Nature 1985, 313,
620.
Acknowledgements
19. Mirvish, S. S. Nature 1985, 315, 461.
This work was partly supported by Grants-in-Aid for
Scientific Research from the Ministry of Education,
Science and Culture, Japan [to K.M. (11101001,
10151219, and 10878092)], and by a Grand from
‘Research for the Future’ Program of Japan Society for
the Promotion of Science (JSPS-RFTF97I00301).
20. Tannenbaum, S. R.; Correa, P. Nature 1985, 317, 675.
21. Joossens, J. V.; Hill, M. J.; Elliott, P.; Stamler, R.; Stam-
ler, J.; Lesaffre, E.; Dyer, A.; Nichols, R.; Kesteloot, H. Int. J.
Epidemiol. 1996, 25, 494.
22. Venner, H. Chem. Ber. 1960, 93, 140.
23. Suzuki, T.; Ide, H.; Yamada, M.; Endo, N.; Kanaori, K.;
Tajima, K.; Morii, T.; Makino, K. Nucleic Acids Res. 2000,
28, 544.
24. Guyton, A. C. Textbook of Medical Physiology; W. B.
Saunders: Philadelphia, 1991; pp 709–725.
References and Notes
25. Antonino, L. C.; Wu, J. C. Biochemistry 1994, 33, 1753.
26. Zollinger, H. Diazo Chemistry I—Aromatic and Hetero-
aromatic Compounds; VCH: Weinheim, 1994; pp 39–64.
27. Williams, D. L. H. Nitrosation; Cambridge University
Press: Cambridge, 1988; pp 10–16.
28. Kato, K.; Yagisawa, N.; Shimada, N.; Hamada, M.;
Takita, T.; Maeda, K.; Umezawa, H. J. Antibiot. 1984, 37,
941.
29. Sambrook, J.; Fritsch, E. F.; Maniatis, T. Molecular
Cloning: A Laboratory Manual; Cold Spring Harbor Labora-
tory: Plainview, NY, 1989.
30. Moschel, R. C.; Keefer, L. K. Tetrahedron Lett. 1989, 30,
1467.
1. Parkin, D. M.; Pisani, P.; Ferlay, J. Int. J. Cancer 1993, 54,
594.
2. Joossens, J. V.; Geboer, J. Nutr. Cancer 1981, 2, 250.
3. Kono, S.; Hirohata, T. Cancer Causes Control 1996, 7, 41 .
4. Takahashi, M.; Kokubo, T.; Furukawa, F.; Kurokawa, Y.;
Tatematsu, M.; Hayashi, Y. Gann 1983, 74, 28.
5. Takahashi, M.; Kokubo, T.; Furukawa, F.; Kurokawa, Y.;
Hayashi, Y. Gann 1984, 75, 494.
6. Cohen, A. J.; Roe, F. J. C. Food Chem. Toxicol. 1997, 35, 271.
7. Correa, P. Cancer Res. 1988, 48, 3554.
8. Schuster, H.; Schramm, G. Z. Naturforsch. 1958, 13b, 697.