A convenient preparative method for the 1,N2-cyclic adducts of guanine nucleosides and nucleotides with crotonaldehyde

Article abstract of DOI:10.1016/S0040-4020(02)01047-5

The treatment of guanine nucleosides and nucleotides with excess crotonaldehyde in pH 8.0 phosphate buffer containing an equimolar amount of L-arginine at 50°C for 2h resulted in the selective formation of the corresponding cyclic 1,N²-propano

Full text of DOI:10.1016/S0040-4020(02)01047-5

Tetrahedron 58 (2002) 8413–8416
A convenient preparative method for the 1,N2-cyclic adducts of
guanine nucleosides and nucleotides with crotonaldehyde
*
Magoichi Sako and Isamu Yaekura
Medicinal Informatics, Gifu Pharmaceutical University, 5-6-1, Mitahora-higashi, Gifu 502-8585, Japan
Received 10 August 2002; accepted 22 August 2002
Abstract—The treatment of guanine nucleosides and nucleotides with excess crotonaldehyde in pH 8.0 phosphate buffer containing an
equimolar amount of L-arginine at 508C for 2 h resulted in the selective formation of the corresponding cyclic 1,N 2-propano adducts as a
mixture of its diastereomers. q 2002 Elsevier Science Ltd. All rights reserved.
Crotonaldehyde (CA) is a mutagen and carcinogen1,2
commonly existing in the human environment as a
component of mobile source emissions, tobacco smoke,
and other thermal degradation products, and as a metabolite
of lipid peroxidations.3 Recently, it has been documented
that, in shuttle vector plasmids propagated in human cells,
CA induces a variety of mutations including dGuo–dThd
transversions, dGuo–dAdo transitions, and tandem base
substitutions,4 and also that a 1,N 2-cyclic adduct (cf. 1a) of
20-deoxyguanosine with CA has been detected in tissues of
laboratory animals in the range of three adducts/109
nucleotides after intake of CA via food followed by
enzymatic treatment with DNase I, phosphodiesterase I,
Figure 1.
and alkaline phosphatase.5 To study the molecular mech-
anisms explaining the mutagenic and carcinogenic effects of
CA, a variety of the cyclic 1,N 2-propano adducts (cf. 1) of
the guanine nucleosides and nucleotides were required.
However, the formation of the 1,N 2-cyclic adducts 1 in the
reactions of guanosine derivatives with CA are not effective
at 378C or required severe conditions (e.g. at 908C for
16 h),6 for example, the treatment of 20-deoxyguanosine
with excess CA in pH 7 phosphate buffer at 378C for 1 day
affords a small amount of 1a, together with the unstable but
isolatable N 2-(3-hydroxybutylidene) dGuo adduct as a
major product and trace amounts of other products such as
N7,8-cyclic adducts and the N 2-paradol dGuo adduct
(Fig. 1).6a
are very effective for the selective formation of the 1,N 2-
cyclic propano adducts (1).
A mixture of 20-deoxyguanosine and excess CA was heated
at 508C for 2 h in 0.1 M phosphate buffer (pH 8.0)
containing an equimolar amount of L-arginine. After
adjustment to pH 5, chromatographic separation of the
reaction mixture using a reversed-phase column allowed the
isolation of the cyclic 1,N 2-propane dGuo adduct (1a) in
94% yield as a mixture of its diastereomers (ca. 1:1),7 with
recovery of the starting arginine. Analogous results were
obtained using a half molar amount of L-arginine (see Fig.
2) or other amino acids such as L-lysine and L-cysteine in
place of L-arginine, though the efficiency for the formation
of 1a is dependent on the nature of the functional group of
the employed amino acids as shown in Table 1. On the other
hand, the use of an acidic buffer as the solvent (see Section
1) or the use of acidic amino acids in place of L-arginine (see
Table 1) caused a decrease in the formation of 1a but
recovery of most of the starting deoxyguanosine.
We now report that the employment of L-arginine as an
additive and a slightly basic buffer as the solvent in the
reaction of the guanine nucleosides and nucleotides with CA
Keywords: preparative method; cyclic 1,N 2-propano guanine adduct;
guanine nucleosides and nucleotides; crotonaldehyde.
*
Corresponding author. Tel.: þ81-58-237-3931; fax: þ81-58-237-5979;
The structure of 1a was confirmed by comparison with the
e-mail: sako@gifu-pu.ac.jp
0040–4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(02)01047-5

8414
M. Sako, I. Yaekura / Tetrahedron 58 (2002) 8413–8416
Figure 2. Concentration-dependency on L-arginine in the formation of the
cyclic propano adduct 1a.
Scheme 1.
Table 1. Reactions of 20-deoxyguanosine with crotonaldehyde in the
presence of amino acids
1,N 2-cyclic adduct 1a (%)
principle, applicable to the regioselective modifications of
guanine residues in oligonucleotides.
Amino acids
After 1 h
After 2 h
L-Serine
34
31
7
36
25
48
92
43
43
9
49
30
59
99
L-Tyrosine
L-Aspartic acid
L-Histidine
L-Cysteine
L-Lysine
1. Experimental
1
All melting points are uncorrected. The H and 13C NMR
spectra were obtained at 400 and 75 MHz, respectively,
using deuterium oxide unless otherwise noted as the solvent.
Mass spectra were determined at an ionizing voltage of
70 eV. For the thin-layer chromatographic (TLC) analyses,
Merck precoated TLC plates [Merck No. 5715; silica gel 60-
F254; eluted with chloroform–methanol–acetic acid
(40:8:1)] and Shimadzu Dual-wavelength Frying-spot
Scanner CS-9000 (detected by 254 nm) were used. Column
chromatographic separation was performed with Sep-Pak
[Waters, Vac 35 cm3 (10 g)] C18 cartridges. The product
ratios of the diastereomeric products were determined by
HPLC analyses carried out using a LiChroCART column
(Cica-MERCK; 150-4.6) with isocratic elution by 5%
CH3CN in 10 mM sodium phosphate buffer (pH 7.0) for
5 min and then a gradient from 5 to 25% over the course of
20 min (flow rates: 1.0 mL/min; detected by 254 nm). Unless
otherwise noted, the materials obtained from commercial
suppliers were used without further purification.
L-Arginine
1H NMR spectral data previously reported.6b,8 For example,
its 1H NMR spectrum showed a broad doublet signal
assignable to the Me group at d 1.16 (3H, J¼6 Hz; dC
20.6) ppm, and characteristic signals for the methylene
group at d 1.43 and 2.05 (each 1H, each br t and br d,
J¼each 12 Hz; dC 34.7) ppm and the aminal methine group
at d 6.03 and 6.07 (each 1/2H, each br s; dC 72.0) ppm. In a
similar manner, the cyclic 1,N 2-propano Guo adduct (1b),
cyclic 1,N 2-propano 50-dGMP adduct (1c), cyclic 1,N 2-
propano 30-dGMP adduct (1d), and cyclic 1,N 2-propano
50-GMP adduct (1e) were obtained in almost quantitative
yields. The structural proof of the products 1b–e is based on
their spectral data (see Section 1).
The present method is based on the Michael addition of the
exocyclic amino group in the guanine ring to the a,b-
unsaturated aldehyde moiety in CA and subsequent
intramolecular cyclization leading to the corresponding
cyclic aminals 1. These steps could be significantly
accelerated by the addition of L-arginine to the medium
via the transient formation of a Shiff base (A) of L-arginine
with CA as shown in Scheme 1.9 The transient formation of
A during the reaction was evidenced by the 1H NMR
experiments carried out in a deuterated buffer (see Section
1). The employment of the basic buffer as a solvent is proper
for the nucleophilic additions of the guanine moiety to the
CvC and CvN bonds in the Schiff base A.
1.1. Preparation of cyclic 1,N 2-propano adducts of
guanine nucleosides and nucleotides (1a–e). General
procedure
To a suspension or solution of the appropriate guanosine
derivative (Sigma, .98% purity) (0.1 mmol) and L-arginine
(Aldrich, .98% purity) (17.4 mg, 0.1 mmol) in 0.1 M
phosphate buffer (pH 8.0) (2.0 mL) was added crotonalde-
hyde (CA, Kishida Chemical Co., .99% purity) (50 mL,
0.6 mmol) and then the mixture was heated at 508C for 2 h.
After being acidified to pH 5 with 1N HCl, the resulting
solution was subjected to the reversed-phased column
eluting with 0, 10, 20, 30, 40, and then 50% methanol–
water (each 50 mL). The UV-positive fractions were
collected, evaporated to dryness, and triturated with diethyl
ether to obtain the desired products (1) as a powder in a pure
state. The 1H NMR spectrum of the first Ninhydrin-positive
fractions eluting with water showed the recovery of a large
It should be noted that no detectable formation of any stable
products0was observed in the reactions of other nucleosides
such as 2 -deoxyadenosine, 20-deoxycytidine, and thymidine
with CA in the presence of L-arginine under the same
conditions, suggesting that the present method is, in

M. Sako, I. Yaekura / Tetrahedron 58 (2002) 8413–8416
8415
amount of the starting arginine [d 1.68 (2H, m), 1.91 (2H, br
t, J¼6 Hz), 3.26 (2H, t, J¼6 Hz), 3.77 (1H, t, J¼5 Hz)].
1638 cm21; UV (MeOH): 275 (sh), 261 nm; Mass (FABþ):
418 [MþH]þ; HR-FABMS: 418.1136 (mmu: þ0.8; calcd
for C14H20N5O8P); 1H NMR d: 1.21 (3H, d, J¼6 Hz), 1.54
and 2.12 (each 1H, br t and br d, J¼12 Hz), 2.52 and 2.70
(each 1H, each m), 3.65 (1H, m), 3.67 (2H, m), 4.13 (1H, br
s), 4.78 (1H, br), 6.15 (1H, br t, J¼5 Hz), 6.18 (1H, br s),
7.82 (1H, s); 13C NMR d: 22.3, 36.4, 40.3, 43.0, 64.1, 73.7,
77.5, 86.3, 88.8, 117.8, 140.2, 152.6, 153.5, 159.9.
1.1.1. 8-Hydroxy-3-(4-hydroxy-5-hydroxymethyl-tetra-
hydrofuran-2-yl)-6-methyl-5,6,7,8-tetrahydro-3H-
1,3,4,5,8a-pentaaza-cyclopenta[b]naphthalen-9-one
(cyclic 1,N 2-propano dGuo adduct) (1a).6b,8 From
fractions eluted with 20–30% methanol-containing water;
31.7 mg (94% yield; as a 23:27 mixture of its diastereomers
isolatable on HPLC; retention time: 14.3 and 15.1 min); mp
155–1608C (from acetone); IR (KBr): 3394, 1686,
1.1.5. Phosphoric acid mono-[3,4-dihydroxy-5-(8-
hydroxy-6-methyl-9-oxo-5,7,8,9-tetrahydro-6H-
1,3,4,5,8a-pentaaza-cyclopenta[b]naphthalen-3-yl)-
tetrahydrofuran-2-ylmethyl] ester (cyclic 1,N 2-propano
50-GMP adduct) (1e). From fractions eluted with 10–20%
methanol-containing water; 41.4 mg (91% yield; containing
a small amount of its Na salt; as a 27:73 mixture of its
diastereomers isolatable on HPLC; retention time: 6.4 and
9.6 min); IR (KBr): 3412, 1686, 1637 cm21; UV (MeOH):
275 (sh), 262 nm; Mass (FABþ): 434 [MþH]þ; HR-
FABMS: 434.1071 (mmu: 20.6; calcd for C14H20N5O9P);
1H NMR d: 1.21 (3H, d, J¼6 Hz), 1.55 and 2.13 (each 1H,
br t and br d, each J¼12 Hz), 3.70 (2H, m), 3.96 (1H, m),
4.19 (1H, m), 4.34 (1H, br s), 4.58 (1H, m), 5.77 (1H, d,
J¼6 Hz), 6.19 (1H, br s), 7.92 (1H, s).
1571 cm21
; UV (MeOH): 275 (sh), 261 nm; Mass
(FABþ): 338 [MþH]þ; HR-FABMS: 338.1476 (mmu:
1
þ1.1; calcd for C14H20N5O5); H NMR d: 1.16 (3H, d,
J¼6 Hz), 1.43 and 2.05 (each 1H, br t and br d, J¼12 Hz),
2.34 and 2.55 (each 1H, each m), 3.63 (3H, m), 3.92 (1H,
m), 4.44 (1H, m), 5.97 (1H, m), 6.03 and 6.07 (each 1/2H,
each br s), 7.72 (1H, s); 13C NMR d: 20.6, 34.7, 39.4, 41.3,
62.5, 72.0 (2), 84.2, 87.8, 115.8, 137.9, 150.7, 151.7, 157.8.
1.1.2. 3-(3,4-Dihydroxy-5-hydroxymethyl-tetrahydro-
furan-2-yl)-8-hydroxy-6-methyl-5,6,7,8-tetrahydro-3H-
1,3,4,5,8a-pentaaza-cyclopenta[b]naphthalen-9-one
(cyclic 1,N 2-propano Guo adduct) (1b). From fractions
eluted with 10–20% methanol-containing water; 31.8 mg
(90% yield; as a 3:2 mixture of its diastereomers isolatable
on HPLC; retention time: 12.4 and 14.2 min); mp 177–
1.2. The formation of the cyclic 1,N 2-propano adduct 1a
in the reaction of 20-deoxyguanosine with CA
1798C (from acetone); IR (KBr): 3422, 1687, 1570 cm21
;
UV (MeOH): 275 (sh), 261 nm; Mass (FABþ): 354
pH-Dependency. CA (25 mL, 0.30 mmol) was added to a
mixed solution of 0.1 M phosphate buffer (pH 5.5, 6.5, 7.5,
or 8.5) (1.0 mL) containing 20-deoxyguanosine (14.3 mg,
0.05 mmol) and L-arginine (8.7 mg, 0.05 mmol) with
stirring. After continuous stirring at 508C for 1 h, the
product distributions in these reactions were estimated by
TLC densitometry. The yields of the desired product 1a
were as follows: 44% (pH 5.5), 58% (pH 6.5), 94% (pH 7.5),
and 95% (pH 8.5), along with recovery of the starting
deoxyguanosine.
[MþH]þ; HR-FABMS: 354.1406 (mmu: 20.8; calcd for
1
C14H20N5O6); H NMR d: 1.12 (3H, d, J¼6 Hz), 1.41 and
2.02 (each 1H, br t and br d, J¼12 Hz), 3.54 (1H, m), 3.68
(2H, m) 4.03 (1H, m), 4.24 (1H, m), 4.49 and 4.55 (each 3/5
and 2/5H, each t, J¼5 Hz), 5.65 (1H, m), 6.03 (1H, br s),
7.72 and 7.73 (each 2/5 and 3/5H, each s); 13C NMR d: 20.7,
34.7, 41.4, 62.3, 71.3, 72.2, 74.4, 85.9, 88.2, 116.1, 138.3,
150.9, 151.8, 158.0.
1.1.3. Phosphoric acid mono-[3-hydroxy-5-(8-hydroxy-6-
methyl-9-oxo-5,7,8,9-tetrahydro-6H-1,3,4,5,8a-pen-
taaza-cyclopenta[b]naphthalen-3-yl)-tet0rahydrofuran-2-
ylmethyl] ester (cyclic 1,N 2-propano 5 -dGMP adduct)
(1c). From fractions eluted with 10–20% methanol-
containing water; 40.4 mg (92% yield; as a 31:19 dia-
stereomeric mixture of its Na salt isolatable on HPLC;
retention time: 12.9 and 14.0 min); IR (KBr): 3420, 1686,
1637 cm21; UV (MeOH): 275 (sh), 262 nm; Mass (FABþ):
418 [MþH]þ; HR-FABMS: 418.1123 (mmu: 20.5; calcd
for C14H20N5O8P); 1H NMR d: 1.22 (3H, d, J¼6 Hz), 1.55
and 2.13 (each 1H, br t and br d, J¼12 Hz), 2.36 and 2.66
(each 1H, each m), 3.66 (1H, m), 3.90 (2H, m), 4.09 (1H, br
s), 4.61 (1H, br s), 6.17 (1H, br d, J¼7 Hz), 6.19 (1H, br s),
7.91 (1H, s); 13C NMR d: 21.0, 34.8, 39.5, 41.3, 65.5, 72.1,
72.4, 84.1, 86.8, 111.6, 137.9, 151.3, 152.1, 161.0.
Concentration-dependency on L-arginine. CA (25 mL,
0.30 mmol) was added to the mixed solutions of 0.1 M
phosphate buffer (pH 8.0) (1.0 mL) containing 20-deoxy-
guanosine (14.3 mg, 0.05 mmol) and L-arginine [none,
2.2 mg (0.012 mmol), 4.4 mg (0.025 mmol), or 8.7 mg
(0.05 mmol)] with stirring. After continuous stirring at
508C for 1 h, the product distributions in these reactions
were estimated by TLC densitometry. The yields of the
desired product 1a were illustrated in Fig. 2.
Concentration-dependency on CA. CA [6.2 mL
(0.075 mmol), 8.3 mL (0.1 mmol), 16.6 mL (0.2 mmol), or
24.9 mL (0.30 mmol)] was added to the mixed solutions of
0.1 M phosphate buffer (pH 8.0) (1.0 mL) containing 20-
deoxyguanosine (14.3 mg, 0.05 mmol) and L-arginine
(8.7 mg, 0.05 mmol) with stirring. After continuous stirring
at 508C for 1 h, the product distributions in these reactions
were estimated by TLC densitometry. The yields of the
desired product 1a were as follows: 70% (1.5 equiv.), 75%
(2.0 equiv.), 93% (4.0 equiv.), and 95% (6.0 equiv.), along
with recovery of the starting deoxyguanosine.
1.1.4. Phosphoric acid mono-[2-hydroxymethyl-5-(8-
hydroxy-6-methyl-9-oxo-5,7,8,9-tetrahydro-6H-
1,3,4,5,8a-pentaaza-cyclopenta[b]naphthalen-3-yl)-
tetrahydrofuran-3-yl] ester (cyclic 1,N 2-propano 30-
dGMP adduct) (1d). From fractions eluted with 20–30%
methanol-containing water; 40.8 mg (93% yield as a 27:23
diastereomeric mixture of its Na salt isolatable on HPLC;
retention time: 10.1 and 12.5 min); IR (KBr): 3398, 1686,
Other amino acids as an additive. CA (10 mL, 0.12 mmol)
was added to a mixed solution of 0.1 M phosphate buffer

8416
M. Sako, I. Yaekura / Tetrahedron 58 (2002) 8413–8416
(pH 8.0) (0.5 mL) containing 20-deoxyguanosine (5.7 mg,
0.02 mmol) and an appropriate amino acid (L-serine,
L-tyrosine, L-aspartic acid, L-histidine, L-L-lysine, or
L-cysteine (0.02 mmol) with stirring. After continuous
stirring at 508C for 1 or 2 h, the product distributions in
these reactions were estimated by TLC densitometry. The
yields of the desired product 1a were shown in Table 1.
References
1. (a) Eder, E.; Budiawan, B. Cancer Epidemiol., Biomarkers
Prev. 2001, 10, 883–888. (b) Eder, E.; Schuler, D.; Budiawan,
B. IARC Sci. Publ. 1999, 150, 219–232. (c) McGregor, D. B.;
Heseltine, E.; Moller, H. IARC Monographs on the Evaluation
of Carcinogenic Risks to Humans, Vol. 63; International
Agency for Research on Cancer: Lyon, France, 1995; pp
373–391.
1.3. Reactions of 20-deoxyadenosine, adenosine, 20-
deoxycytidine, cytidine, thymidine or uridine with CA in
the presence of L-arginine
2. Chung, F. L.; Tanaka, T.; Hechts, S. S. Cancer Res. 1986, 46,
1285–1289.
3. (a) Satoh, K.; Hayakari, M.; Ookawa, K.; Satou, M.; Aizawa, S.;
Tanaka, M.; Hatayama, I.; Tsuchida, S.; Uchida, K. Mutat. Res.
2001, 483, 65–72. (b) Doerge, D. R.; Yi, P.; Churchwell, M. I.;
Preece, S. W.; Langridge, J.; Fu, P. P. Rapid Commun. Mass
Spectrom. 1998, 12, 1665–1672. (c) Chung, F. L.; Chen, H. J.
C.; Nath, R. G. Carcinogenesis 1996, 17, 2105–2111.
4. Kawanishi, M.; Matsuda, T.; Sasaki, G.; Yagi, T.; Matsui, S.;
Takebe, H. Carcinogenesis 1998, 19, 69–72.
To a mixed solution of an appropriate nucleoside
(0.02 mmol) in 0.1 M phosphate buffer (pH 8.0) (0.5 mL)
containing L-arginine (3.5 mg, 0.02 mmol), CA (10 mL,
0.12 mmol) was added and the mixture was heated at 508C
for 2 h with stirring. The TLC densitometric analyses of the
reaction mixtures showed no change in the starting nucleo-
sides. Similar results were obtained even after a prolonged
reaction time (e.g. for 1 day).
5. (a) Hecht, S. S.; McIntee, E. J.; Wang, M. Toxicology 2001,
166, 31–36. (b) Budiawan, B.; Schuler, D.; Eder, E. Arch.
Toxicol. 2000, 74, 404–414. (c) Nath, R.-G.; Ocando, J. E.;
Chung, F.-L. Cancer Res. 1996, 56, 452–456. (d) Foiles, P. G.;
Akerkar, S. A.; Miglietta, L. M.; Chung, F.-L. Carcinogenesis
1990, 11, 2059–2061.
1.4. 1H NMR experiments on the reaction of 20-deoxy-
guanosine with CA in the absence or presence of
L-arginine
To a solution of 20-deoxyguanosine (7.2 mg, 0.025 mmol)
or a mixed solution of 20-deoxyguanosine (7.2 mg,
0.025 mmol) and L-arginine (8.7 mg, 0.05 mmol) in 0.1 M
deuterated phosphate buffer (pH 8.0) (1.5 mL) was added
CA (12.5 mL, 0.15 mmol) and heated at 378C in a NMR
tube. These reactions were followed by 1H NMR spec-
trometry. In the experiments using L-arginine as an
additives, the transient formation of an intermediate having
characteristic peaks at d 1.99 (3H, d, J¼7 Hz), 6.50 (1H, m),
7.23 (1H, m), and 7.91 (1H, d, J¼8 Hz), assignable to the
Schiff base of L-arginine with CA, and the smooth formation
of the cyclic adduct 1a during the reaction were observed.
On the other hand, in the absence of L-arginine, the recovery
of most of the starting materials, 20-deoxyguanosine and CA
was observed even after 2 days.
6. (a) Wang, M.; McIntee, E. J.; Cheng, G.; Shi, Y.; Villalta, P. W.;
Hecht, S. S. Chem. Res. Toxicol. 2001, 14, 423–430. (b) Chung,
F.-L.; Hecht, S. S. Cancer Res. 1983, 43, 1230–1235.
7. The retention times (14.3 and 15.1 min) of the cyclic adducts 1a
on HPLC were consistent with those of the minor products6b
obtained in the absence of L-arginine. Therefore, the methyl and
hydroxyl groups on the cyclic 1,N 2-propane ring in 1a should
be in the trans-configuration, as reported by Chung et al.,6b
from the steric reasons in the intramolecular cyclization process
leading to the cyclic aminal.
8. Wang, M.; McIntee, E. J.; Cheng, G.; Shi, Y.; Villalta, P. W.;
Hecht, S. S. Chem. Res. Toxicol. 2000, 13, 1149–1157.
9. Cordova, A.; Notz, W.; Barbas, III., C. F. J. Org. Chem. 2002,
67, 301–303.