Reactions of 9-substituted guanines with bromomalondialdehyde in aqueous solution predominantly yield glyoxal-derived adducts.

Article abstract of DOI:10.1039/b405117c

Reactions of 9-ethylguanine, 2'-deoxyguanosine and guanosine with bromomalondialdehyde in aqueous buffers over a wide pH-range were studied. The main products were isolated and characterized by (1)H and (13)C NMR and mass spectroscopy. The final products formed under acidic and basic conditions were different, but they shared the common feature of being derived from glyoxal. Among the 1 : 1 adducts, 1,N(2)-(trans-1,2-dihydroxyethano)guanine adduct (6) predominated at pH < 6 and N(2)-carboxymethylguanine adduct (10a,b) at pH > 7. In addition to these, an N(2)-(4,5-dihydroxy-1,3-dioxolan-2-yl)methylene adduct (11a,b) and an N(2)-carboxymethyl-1,N(2)-(trans-1,2-dihydroxyethano)guanine adduct (12) were obtained at pH 10. The results of kinetic experiments suggest that bromomalondialdehyde is significantly decomposed to formic acid and glycolaldehyde under the conditions required to obtain guanine adducts. Glycolaldehyde is oxidized to glyoxal, which then modifies the guanine base more readily than bromomalondialdehyde. Besides the glyoxal-derived adducts, 1,N(2)-ethenoguanine (5a-c) and N(2),3-ethenoguanine adducts (4a-c) were formed as minor products, and a transient accumulation of two unstable intermediates, tentatively identified as 1,N(2)-(1,2,2,3-tetrahydroxypropano)(8) and 1,N(2)-(2-formyl-1,2,3-trihydroxypropano)(9) adducts, was observed.

Full text of DOI:10.1039/b405117c

View Article Online / Journal Homepage / Table of Contents for this issue
Reactions of 9-substituted guanines with bromomalondialdehyde
in aqueous solution predominantly yield glyoxal-derived adducts
Anne-Mari Ruohola, Niangoran Koissi, Sanna Andersson, Ilona Lepistö, Kari Neuvonen,*
Satu Mikkola* and Harri Lönnberg
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland
Received 5th April 2004, Accepted 6th May 2004
First published as an Advance Article on the web 16th June 2004
Reactions of 9-ethylguanine, 2Ј-deoxyguanosine and guanosine with bromomalondialdehyde in aqueous buffers
over a wide pH-range were studied. The main products were isolated and characterized by 1H and 13C NMR and
mass spectroscopy. The final products formed under acidic and basic conditions were different, but they shared the
common feature of being derived from glyoxal. Among the 1 : 1 adducts, 1,N 2-(trans-1,2-dihydroxyethano)guanine
adduct (6) predominated at pH < 6 and N 2-carboxymethylguanine adduct (10a,b) at pH > 7. In addition to these,
an N 2-(4,5-dihydroxy-1,3-dioxolan-2-yl)methylene adduct (11a,b) and an N 2-carboxymethyl-1,N 2-(trans-1,2-
dihydroxyethano)guanine adduct (12) were obtained at pH 10. The results of kinetic experiments suggest that
bromomalondialdehyde is significantly decomposed to formic acid and glycolaldehyde under the conditions
required to obtain guanine adducts. Glycolaldehyde is oxidized to glyoxal, which then modifies the guanine base
more readily than bromomalondialdehyde. Besides the glyoxal-derived adducts, 1,N 2-ethenoguanine (5a–c) and
N 2,3-ethenoguanine adducts (4a–c) were formed as minor products, and a transient accumulation of two unstable
intermediates, tentatively identified as 1,N 2-(1,2,2,3-tetrahydroxypropano) (8) and 1,N 2-(2-formyl-1,2,3-
trihydroxypropano) (9) adducts, was observed.
The adduct formation was studied over a wide pH-range, and
Introduction
the predominant stable products were isolated and character-
ized. The results showed that under any conditions employed,
neither the etheno nor formyletheno products were the pre-
dominant products, but those derived from glyoxal actually
prevailed. More detailed kinetic investigations were carried out
under neutral and slightly acidic conditions to study the
mechanism of the formation of these unexpected adducts and
to compare the results to those previously obtained with
9-substituted adenines.
Human beings are exposed every day to numerous alkylating
agents of environmental origin. Such compounds can also be
formed endogenously as a result of metabolism. Alkylation of
DNA bases is deleterious for living organisms.1 Reactions
of nucleophilic amine functions with electrophiles block the
normal hydrogen bonding within double-helical DNA, result-
ing in miscoding during the DNA synthesis and, hence, form-
ation of a mutation. Cyclic DNA adducts obtained by reactions
with bifunctional electrophiles, such as malondialdehyde,2–4
acrolein,5–8 crotonaldehyde,7–10 halogenated acetaldehydes,11–15
halogenated malondialdehydes,16–18 vinyl chloride and chloro-
oxirane,19–21 and mucohalo acids,22–25 have been extensively
studied over several decades. Such studies serve a dual purpose.
On one hand, development of sensitive methods for detection
of mutations within DNA depends on an understanding of the
structure and stability of the adducts.26–31 On the other hand
adducted nucleosides are useful diagnostic tools. Some cyclic
DNA adducts, such as 1,N 6-ethenoadenine,12,32 3,N 4-etheno-
cytocine,32 N 2,3-ethenoguanine,33 and 1,N 2-propenoguanine,2–4
are fluoresescent and can be used as biomarkers.3,12,16,32,33
Reactions between nucleic acid bases and malondialde-
hydes,28 halogenated malondialdehydes18 and their congener,
triformylmethane,34–36 have previously been studied in our
laboratory. The reaction of bromomalondialdehyde (BMA)
with adenosine has been shown to yield, under slightly acidic
conditions, two products, viz. 1,N 6-ethenoadenosine and its
formyl derivative, suggested to be formed via a common acyclic
intermediate.18 As a continuation of the mechanistic work on
the reactions of halogenated malondialdehydes with nucleic
acid bases, the present work studies the reactions of BMA with
9-substituted guanines. It has been reported that 1,N 2-etheno
and 1,N 2-formyletheno25 products are formed in reactions with
3-chloro-4-dichloromethyl-5-hydroxy-2(5H )-furanone, similar
to the situation with adenine nucleosides. Formation of
N 2,3-etheno products is of particular interest, since such
adducts are potentially mutagenic37 and exceptionally
susceptible to depurination.38
Results
Kinetic measurements
To avoid the complications that depurination of guanine
nucleosides or their adducts could possibly result in, 9-ethyl-
guanine (1a) was used as a model compound in most of the
kinetic studies. At pH > 5, the reactions of 2Ј-deoxyguanosine
(1b) were also studied. The reactions were usually carried out at
60 ЊC in 10 mmol dmϪ3 solutions of BMA, the pH of which was
adjusted with 0.1 mol dmϪ3 buffer solution. The progress of the
reactions was followed by analyzing aliquots withdrawn at
appropriate intervals by RP-HPLC. The initial concentration
of the starting material was lower than 0.5 mmol dmϪ3 and,
hence, its disappearance obeyed pseudo first-order kinetics. The
rate constants obtained as a function of pH are shown in Fig. 1.
The data previously obtained with 9-methyladenine is included
for comparative purposes.
As seen from Fig. 1, the reaction of 9-ethylguanine with
BMA exhibits an inverse dependence of the rate on acidity at
pH < 5, a broad maximum around pH 6, and a rate-retardation
at pH > 7. The overall effect of pH on the reaction rate is,
however, rather modest; at pH < 5 the rate constant is decreased
by approximately one order of magnitude when the hydronium
ion concentration becomes thousand-fold. As discussed below
in more detail, several parallel reactions in all likelihood con-
tribute to the disappearance of the starting material and most
of them are preceded by complicated transformation reactions
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1943

View Article Online
Fig. 1 pH-Rate profiles for the reactions of various 9-substituted
guanines (< 0.5 mmol dmϪ3) with BMA (10 mmol dmϪ3) at 60 ЊC.
Notation: 9-ethylguanine (1a;᭺), 9-ethyl-O6-methylguanine (2; ∆),
9-ethyl-1-methylguanine (3; ꢀ). The dotted line refers to the reaction of
9-methyladenine with BMA.18
of BMA. Accordingly, it is not possible to give a simple
mechanistic explanation for the shape of the rate profile.
For comparative purposes, the reactions of 9-ethyl-O6-
methylguanine (2) and 9-ethyl-1-methylguanine (3), which may
be regarded as models of the two tautomeric forms of 1a, were
studied under the same conditions. As seen from Fig. 1, the pH
– rate profile of the reaction of 9-ethyl-1-methylguanine with
BMA closely resembles that of the corresponding reaction of
9-ethylguanine, but the rate constants are only one fifth of those
obtained with 1a. The reaction of 9-ethyl-O6-methylguanine,
mimicking the rare tautomer of 1a, is more than one order of
magnitude faster than that of 1a, and the shape of the pH–rate
profile resembles that of the reaction of 9-methyladenine,
exhibiting a maximum at pH 4. Accordingly, 9-ethylguanine
appears to react through its major tautomeric form. 2Ј-Deoxy-
guanosine (1b) reacts with BMA as fast as 9-ethylguanine at pH
> 5 (data not shown) and, hence, seems to be a relevant model
for the studies of the reactions of guanine nucleosides
under neutral and alkaline conditions. In more acidic solutions,
2Ј-deoxyguanosine undergoes depurination.
Fig.
9-ethylguanine (1a; < 0.5 mmol dmϪ3) with BMA (10 mmol dmϪ3
at pH 4.1 and 60 ЊC. Annotation: tR 4.3 min: 1,N 2-(1,2,2,4-
tetrahydroxypropano)-9-ethylguanine and
1,N 2-(2-formyl-1,2,3-
2
RP-HPLC chromatograms of the reaction mixtures of
)
trihydroxypropano)-9-ethylguanine (8 ϩ 9), tR 5–6 min 1,N 2-(1,2-
dihydroxyethano)-9-ethylguanines (6 and 7), tR 8 min N 2,3-etheno-
9-ethylguanine (4a) and tR 9 min 9-ethylguanine (1a). b: RP-HPLC
chromatograms of the reaction mixtures of 9-ethylguanine (1a; < 0.5
mmol dmϪ3) with BMA (10 mmol dmϪ3) at pH 7.4 and 60 ЊC.
Annotation: tR 4.3 min: 1,N 2-(1,2,2,4-tetrahydroxypropano)-9-ethyl-
guanine and 1,N 2-(2-formyl-1,2,3-trihydroxypropano)-9-ethylguanine
(8 ϩ 9), tR 5–6 min 1,N 2-(1,2-dihydroxyethano)-9-ethylguanines (6 and
7), tR 9 min 9-ethylguanine (1a) and tR 16 min 1,N 2-etheno-9-ethyl-
guanine (5a).
Product analyses
Fig. 2 shows the RP-HPLC chromatograms obtained for the
reaction of 9-ethylguanine with BMA at pH 4.1 and 7.4. As
seen, the previously reported etheno products22,25 are present
in the product mixture (cf. Table 1), but only as minor com-
ponents. The fluorescent N2,3-etheno product (4a) was
observed at a retention time clearly shorter (tR 8.5 min) than
that of its non-fluorescent 1,N 2-etheno isomer (5a) (tR 17 min).
Low pH favours the formation of the 2,N 3-adduct compared to
the 1,N 2-adduct and the other products. No clear evidence
for the formation of the corresponding formyletheno
products could be obtained, in contrast to the situation with
9-methyladenine.18
In addition to the signals of the etheno derivatives, three
UV-absorbing peaks appeared. The one with tR 4.2 min
predominated during the early stages of the reaction (Fig. 3).
PR-HPLC-ESI/MS analysis revealed that this signal actually
resulted from co-elution of two compounds having MHϩ values
284 and 296, respectively. On prolonged treatment, these
products disappeared. Neither of them was sufficiently stable to
be preparatively isolated. The other two signals that appeared at
tR 5.0 and 5.9 min, both exhibited a MHϩ value of 238. The
concentration of the respective products continuously increased
with time, the dependence of peak area on time showing a
slight upward curvature during the early stages of the reaction
(Fig. 3). Hence these products clearly predominated during the
late stages of the reaction. They were isolated from the reaction
mixture obtained at pH 5.5 (60 ЊC) and were characterized by
1H and 13C NMR spectroscopy (Tables 2 and 3). The isolated
compound turned out to be predominantly 1,N 2-(trans-1,2-
dihydroxyethano)-9-ethylguanine (6). Shapiro et al.39 have
shown previously that this adduct may be obtained in a solid
form from glyoxal and guanine. The identity of 6 formed in the
reaction of 9-ethylguanine with BMA was verified by spiking
with a sample prepared in such a manner from glyoxal and
9-ethylguanine.
Since the product having MHϩ 238 appeared as two signals
in the original HPLC chromatogram, the occurrence of 6 as
1
various stereoisomers was studied by H NMR spectroscopy.
The cyclic trans adduct proved to be the thermodynamically
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1944

View Article Online
Table 1 1H NMR chemical shifts for the adducts of 9-substituted
296) adducts. Consistent with the suggested structures, 8 exhib-
ited ESI-MS signals at m/z 266 (MHϩ Ϫ H2O), 248 (MHϩ
guaninesa
Ϫ
2H2O) and 220 (MHϩ Ϫ H2O–HCOOH) and 9 at 180 (MHϩ of
1a). Their UV-absorption maxima were rather similar to those
of 9-ethylguanine. Accordingly, 8 exhibited, at pH 4.5, an
absorption maximum at 252 nm with a shoulder at 278 nm (cf.
1a 253 nm with shoulder 275 nm), and 9 at pH 7.4 a maximum
at 256 nm with a shoulder at 283 nm (cf. 1a 254 with shoulder
274 nm). While 6 was obtained by reacting 9-ethylguanine with
either BMA or glyoxal, products 8 and 9 could only be obtained
from BMA; the reaction of 9-ethylguanine with glyoxal gave no
such products. The plausible mechanisms for the formation of
6, 8 and 9 are discussed below.
The products obtained under alkaline conditions (pH > 8)
were different. Again products having MHϩ 238 and 296 were
formed, but they were not identical with 6 or 9, respectively. The
product having MHϩ 238 was isolated and assigned by 1H and
13C NMR spectroscopy as N 2-carboxymethyl-9-ethylguanine
(10a) (Tables 2 and 3). The corresponding adduct, N 2-carb-
oxymethylguanosine (10b), was obtained on treating guanosine
with BMA in a phosphate buffer at pH 10 and 70 ЊC for 72 h
(10b). Consistent with structures 10a and 10b, the coupling
between the N 2H and the two methylene protons is clearly
detected (J 3.9 Hz and J 4.1 Hz, respectively). The presence
of a carboxylic acid group may well explain the fact that the
retention time of 10a is very sensitive to the acidity of the
eluent.
4a
5a
11b
1-H
N 2-H
8-H
7.93(s)
7.82(d, J 1.8)
7.14(d, J 1.8)
7.89(s)
7.56(d, J 2.6)
7.37(d, J 2.6)
8.13(s)
a-H
b-H
c-H
d-H
7.60(d, J 2.6)
7.41(d, J 2.6)
7.43
7.90
b
c
d
Others
a As ppm from TMS in DMSO-d6. Coupling constants in Hz. b Ethyl
group resonances at 1.41(t, J 7.3) and 4.42(q, J 7.3). c Ethyl group
resonances at 1.37(t, J 7.3) and 4.08(q, J 7.3). d For the –HOH–CHOH–
moiety an AAЈXXЈ proton spectrum was observed: 3JHC–CH 6.7, 3JHO–CH
8.4, JHO–C–CH 1.1.41 Sugar proton resonances at 5.82(d, J 5.8), 4.48(q,
4
J 5.5), 4.12(q, J 4.4), 3.91(q, J 4.1), 3.64(m) and 3.55(m). Sugar hydroxy
resonances 5.39(d, J 6.1), 5.12(d, J 4.7) and 5.06(t, J 5.3).
The product having MHϩ 296 is in all likelihood N 2-(4,5-
dihydroxydioxolan-2-yl)methylene-9-ethylguanine (11a). The
respective guanosine adduct (11b) was isolated from the same
reaction mixture as 10a and characterized by 1H and 13C NMR
spectroscopy (Tables 1 and 3). The structure proposed for 11b is
consistent with the appearance of an AAЈXXЈ type proton
spectrum for the –HOHCHOH– moiety41 and the fact that the
UV spectrum is rather similar to that of 9-ethylguanine. In
addition to 10a and 11a, a 1 : 2 adduct, N 2-carboxymethyl-
1,N 2-(trans-1,2-dihydroxyethano)-9-ethylguanine (12) was
formed (Tables 1 and 3). The mechanisms for the formation of
10a, 11a and 12 are discussed below.
The products which the reaction of 2Ј-deoxyguanosine with
BMA yielded at pH > 5 were similar to those described above
for 9-ethylguanine (Fig. 4). As mentioned above, at pH < 5 the
predominant reaction was depurination, and significant
amounts of 1,N 2-ethenoguanine (5d) and N 2,3-ethenoguanine
Fig. 3 Time-dependent product distribution for the reaction of
9-ethylguanine (1a; <0.5 mmol dmϪ3) with BMA at pH 5.3 and 60 ЊC.
Notation: 1,N 2-(trans-1,2-dihydroxyethano)-9-ethylguanine (6, ᮀ),
1,N 2-(1,2,2,4-tetrahydroxypropano)-9-ethylguanine
and
1,N 2-(2-
formyl-1,2,3-trihydroxypropano)-9-ethylguanine (8 ϩ 9; ᭺) and N 2,3-
etheno-9-ethylguanine (4a; ᭿).
stable form, but in addition the cis-form (7) could be observed
immediately after dissolution of a solid product prepared
from 1a and glyoxal in DMSO-d6. Accordingly, the product
having MHϩ 238 may well appear as two separate signals in
HPLC chromatograms. Consistent with the structure of 6, the
coupling constants between the carbon bonded hydrogens are
small compared to the couplings to adjacent hydroxy protons,
as reported earlier for a cyclic glyoxal-2Ј-deoxyguanosine
adduct.40
As mentioned above, the products having MHϩ 284 and 296
could not be isolated. Accordingly, it is not clear whether they
lie on the pathway to the final product having MHϩ 238. In
other words, their degradation to non-chromophoric products
cannot be strictly excluded. We tentatively suggest that these
compounds are 1,N 2-(1,2,2,3-tetrahydroxypropano) (8, MHϩ
284) and 1,N 2-(2-formyl-1,2,3-trihydroxypropano) (9, MHϩ
Fig. 4 RP-HPLC chromatogram of the reaction mixture of 2Ј-
deoxyguanosine (1b; <0.5 mM) with BMA (10 mM) at pH 7.4
and 60Њ. Annotation: tR 5 min: 1,N 2-(1,2,2,4-tetrahydroxypropano)-
2Ј-deoxyguanosine and 1,N 2-(2-formyl-1,2,3-trihydroxypropano)-
2Ј-deoxyguanosine (8 ϩ 9), tR 6–7 min 1,N 2-(1,2-dihydroxyethano)-2Ј-
deoxyguanosines (6 and 7), tR 11 min 2Ј-deoxyguanosine (1b) and tR 19
min 1,N 2-etheno-2Ј-deoxyguanosine (5b).
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1945

View Article Online
Table 2 1H NMR chemical shifts for the adducts of 9-substituted guaninesa
6
10a
10b
12
1-H
8.41(br s)
6.61(t, J 3.9)
7.65(s)
8.20(br s)
6.69(t, J 4.1)
7.88(s)
N 2-H
8-H
8.75(d, J 0.7)
7.74(s)
7.74(s)
4.94(s)
5.38(s)
6.67(br s)
7.29(br s)
a-H
b-H
c-H
d-H
e-H
Others
4.85(dd, J 7.1, 0.7)
5.46(d, J 6.4)
6.48(d, J 7.1)
7.18(d, J 6.4)
3.49(t, J 3.9)
3.67(d, J 4.1)
e
3.80; 4,06
b
c
d
f
a As ppm from TMS in DMSO-d6. Coupling constants in Hz. b For the minor cis-isomer 7: N 2-H 8.42, a-H 5.07, b-H 5.68, c-H 6.77 and d-H 7.01.
Ethyl group resonances at 1.32(t, J 7.3) and 3.98(q, J 7.3). c Ethyl group resonances at 1.33(t, J 7.3) and 3.96(q, J 7.3). d Sugar proton resonances as
5.69(d, J 5.8), 4.52(t, J 5.3), 4.12(t, J 4.4), 3.88(q, J 4.1), 3.62(dd, J 11.7, 4.3) and 3.52(dd, J 11.7, 4.6). Sugar hydroxy resonances at 5.35(d, J 6.4),
5.12(d, J 4.4) and 4.95(t, J 5.3). e An AB proton spectrum was observed: 2JH–C–H 17.0. f Ethyl group resonances at 1.33(t, J 7.3) and 3.99(q, J 7.3).
Table 3 13C NMR chemical shifts for the adducts of 9-substituted
Firstly, BMA was decomposed to a significant extent during
guaninesa
the time period that the reactions of 9-ethylguanine were
followed. The disappearance of BMA can actually be followed
by RP-HPLC by using an acidic sodium formate buffer as an
eluent. An experiment in the absence of any nucleosidic
substrate shows that the half-life of the disappearance of BMA
under slightly acidic conditions at 60 ЊC is five days, the rate of
disappearance being rather independent of pH. Unfortunately
the products formed could not be observed by UV-detection.
1
6
10b
12
When the decomposition of BMA was followed by H NMR
spectroscopy, formate ion was the only product that could be
definitely assigned.
C-2
C-4
C-5
C-6
C-8
C-a
C-b
C-e
154.6
150.8
117.4
155.0
137.2
83.6
152.2
151.1
116.8
157.1
136.2
45.2
153.2
150.8
117.4
155.1
137.3
88.2
Secondly, indirect evidence for the decomposition of BMA
was obtained by a series of experiments where a reaction
solution (10 mmol dmϪ3 BMA in 0.1 mol dmϪ3 acetic acid
buffer pH 5.5) was incubated at 60 ЊC from one to seven days
before the reaction was started by adding the 9-ethylguanine
substrate. When 9-ethylguanine was added at the same time as
BMA, the high molecular weight products 8 and 9 were clearly
observed. In contrast, when the reaction was started after seven
days incubation of BMA in the buffer solution, only traces of
these products were formed. Furthermore, the latter reaction
was nearly 20 times as fast as that started immediately after the
preparation of the reaction solution. This increased reactivity
may well be accounted for by a significant amount of glyoxal
being formed in the reaction solution. In fact, the reaction
rate was comparable to that observed for the reaction of
9-ethylguanine with 2 mmol dmϪ3 glyoxal at pH 5.5; a value of
(5.9 0.2) × 10Ϫ5 sϪ1 was obtained for the pseudo first-order
rate constant. Under these conditions, the reaction with glyoxal
yielded only the dihydroxyethano adducts (6, 7).
84.0
171.6
82.4
47.9
C-f
Others
171.9
b
c
d
a As ppm from TMS in DMSO-d6. b Ethyl group resonances at 15.4 and
37.9. c Sugar carbon resonances at 87.1, 85.3, 73.4, 70.5, 61.7. d Ethyl
group resonances at 15.3 and 37.9.
(4d) were formed. The sigmoid product distribution curves
(data not shown) clearly showed that it is the guanine base
released upon depurination that reacted with BMA, yielding
the etheno products. Accordingly, unsubstituted guanine
appears to form etheno products considerably faster than
2Ј-deoxyguanosine.
There is another piece of indirect evidence that suggests
glyoxal to be formed in the reaction solutions. The dihydroxy-
ethano adducts (6,7) are unstable in aqueous solutions, being
decomposed to glyoxal and 9-ethylguanine. This happened
when 6 isolated from the reaction solution was dissolved in a
freshly prepared solution of 10 mmol dmϪ3 BMA in an acetic
acid buffer, pH 5.2. However, if the BMA solution was kept
at 60 ЊC for nine days before 6 was added, only 20% of the
adduct decomposed, suggesting that an equilibrium between
the adduct and free 9-ethylguanine and glyoxal had been
settled. Addition of a small amount of glyoxal resulted
Routes to the glyoxal-derived adducts: evidence for intermediary
formation of glyoxal from BMA
A priori, two alternative pathways may be suggested for the
formation of the glyoxal-derived adducts (6,7,10–12). Either, the
starting material reacts with BMA and the adducts are formed
upon decomposition of an intermediate initially formed, or
alternatively, BMA is decomposed to glyoxal under the
experimental conditions. As indicated below, the results of the
kinetic experiments on the formation of the dihydroxyethano
adducts (6,7) are rather consistent with the latter pathway:
BMA is decomposed to glyoxal, and 9-substituted guanines
react with glyoxal considerably more readily than with BMA.
in total recovery of
9-ethylguanine.
6 and complete disappearance of
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1946

View Article Online
Scheme 1
Discussion
Mechanisms for the formation of the reaction products
As mentioned above, disappearance of BMA was observed to
be accompanied by formation of formate ion. As shown in
Scheme 1, the other expected product, hence, is bromoacetalde-
hyde or its hydrolysis product, glycolaldehyde, while oxidation
is required to obtain glyoxal. It has been shown previously that
glycolaldehyde, when reacted with 2Ј-deoxyguanosine, gives
diastereomeric glyoxal adducts similar to 6 under aerobic but
not under unaerobic conditions.42 Evidently glyoxal formed
by oxidation of glycoladehyde eventually reacts with 9-ethyl-
guanine, giving the dihydroxyethano adduct 6. The latter
reaction is well known.39,43
The carboxymethyl derivative 10a obtained under alkaline
conditions may be formed from the dihydroxyethano adduct
(6), as indicated in Scheme 2. This kind of a ring opening
reaction followed by a hydride ion migration has previously
been reported for a glyoxal-arginine adduct,44 and a similar
carboxymethyl derivative has also been obtained by treating
guanosine with glucose under oxidative conditions.45 Com-
pound 12 may be formed by the reaction of glyoxal with 10a.
Product 11a may simply be obtained by a nucleophilic attack of
N 2 of the guanine base on a glyoxal dimer (Scheme 3), the
formation of which is a well documented reaction.46,47
Scheme 3
products appear to be obtained by dimerization of BMA. The
effect of the BMA concentration on the product distribution
is consistent with such a mechanism. Although the overall
disappearance of 1a was approximately first-order in the
concentration of BMA, high BMA concentrations clearly
favoured the formation of 8 and 9 over 6. At a low BMA
concentration (2 mmol dmϪ3), these products were not observed
at all, whereas 6 was formed to a significant extent. Malon-
dialdehyde is known to undergo dimerization by an aldol
condensation mechanism.48 If 2-hydroxymalonaldehyde,
obtained by hydrolysis of BMA, behaves similarly, formation
of 8 and 9 could be expected to take place as shown in
Scheme 4.
Implications to reactions of BMA with adenosine
The results of the present work apparently argue against those
reported previously for the reaction of BMA with adenosine,
according to which the adducts are formed by a reaction
between the adenine base and intact BMA. This apparent dif-
ference, however, results from the different inherent reactivity
of these two bases towards BMA. The amidine group of the
adenine base appears to favour halogen substituted carbonyl
compounds over 1,2- or 1,3-dicarbonyl compounds. 9-Methyl-
adenine, for example, reacts readily with halogenenated
malondialdehydes giving etheno products (t < 1d in 10 mmol
½
dmϪ3 aq. BMA at pH 5 and 60 ЊC18), whereas no reaction with
glyoxal can be observed at pH 5.4 and 60 ЊC in one week (the
present work). Since the half-life for the decomposition of
BMA under such conditions is five days, it is clear that 9-substi-
tuted adenines react with intact BMA, and not with glyoxal
that possibly is present at low levels. Consistent with this, a
marked rate retardation was observed when the BMA solution
was incubated at 60 ЊC for 5 days before the adenosine substrate
was added, but even in this case the etheno adducts were the
only products observed.
In striking contrast to 9-substituted adenines, 9-substituted
guanines react preferably with 1,2-dicarbonyl compounds: the
half-lives of the reaction of 9-ethylguanine with 2 mmol dmϪ3
glyoxal and 10 mmol dmϪ3 chloroacetaldehyde at pH 5.4
and 60 ЊC are 2 h and 2 d, respectively. Accordingly, even
minor conversion of BMA to glyoxal is sufficient to make
glyoxal-derived adducts as the main products. In this respect,
1,3-dicarbonyl and α,β-unsaturated carbonyl compounds also
Scheme 2
As discussed above, adducts 8 and 9 were obtained on using
fresh solutions of BMA, while increasing of the glyoxal content
by ageing favoured the formation of 6 or 10a, depending on
pH. It should also be noted that neither 8 nor 9 were formed
on treating 9-ethylguanine with glyoxal. Accordingly, these
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1947

View Article Online
Scheme 4
seem to favour 9-substituted guanines over adenines: malon-
experiments were acquired using double quantum filtered
dialdehyde,3 triformylmethane,35,36 acrolein,5 and cronaldehyde9
all react faster with guanosine than with adenosine. α-Halo-
acetaldehydes, in turn, react less readily with guanosine than
with adenosine.11,13,20 Accordingly, it is expected that besides
glyoxal-derived adducts, those obtained by the reactions
through the 1,3-carbonyl groups (8, 9) predominate during the
early stages of the reaction of 9-ethylguanine with BMA.
Consistent with the preceding discussion, the tautomeric
form of the guanine base is crucial in dictating the reactivity.
9-Ethyl-O6-methylguanine (2), which exists in a tautomeric
form similar to that of 9-substituted adenine, reacts like
adenosine: the rate of reaction with BMA is the same, and the
reaction reaches its maximal rate under the same conditions.
Etheno adducts are formed as predominant products. 9-Ethyl-
1-methylguanine (3), which mimics the stable tautomer of
9-ethylguanine (1a), exhibits a pH-rate profile similar to that
obtained with 1a. The 1-methyl group, however, prevents the
reactions at this site, and hence the reaction of 3 with BMA is
nearly one order of magnitude slower than that of 9-ethylguan-
ine. Consistent with previous observations on the reactions
of 9-substituted guanines with malondialdehyde2–4,49 or tri-
formylmethane,35 the formation of cyclic N 2,3-adducts is a less
favourable process than that leading to 1,N 2-adducts.
COSY in phase-sensitive mode with spectral widths appropri-
ately optimized from the 1D spectra, and processed with
zero-filling (×2, ×4) and exponential weighting (1 Hz) applied
in both dimensions prior to Fourier transformation.
1D Carbon spectra were acquired with single-pulse excit-
ation, 45Њ flip angle, pulse recycle time of 3.5 sec, and with
spectral widths of 34 kHz consisting of 64 k data points (digital
resolution 0.52 Hz per pt), zero-filled to 128 k and with 1 Hz
exponential weighting applied prior to Fourier transformation.
2D Heteronuclear one-bond correlation experiments were
acquired using carbon detected CH-shift correlation (with
partial homonuclear decoupling in the f1 dimension) with
spectral widths appropriately optimised from the 1D spectra
and processed with zero-filling (×2, ×4), a 2Π/3-shifted sinebell
function, and exponential weighting applied in both dimen-
1
sions prior to Fourier transformation. A JHC coupling of 145
Hz was used.
Materials
2Ј-Deoxyguanosine, guanosine, O6-methylguanine, and 9-ethyl-
guanine were products of Sigma and 1-methylguanine a
product of Fluka. They all were used as received. HPLC grade
solvents were used for HPLC analyses and purifications. All the
electrolytes and buffer constituents were of reagent grade.
BMA was prepared by the method of Trofimenko.50
Experimental
9-Ethyl-O6-methylguanine (2) and 9-ethyl-1-methylguanine (3)
General
The HPLC analyses were carried out on a Perkin-Elmer
Integral 4000 HPLC equipped with a diode array detector and
the HPLC-ESI/MS analyses on a Perkin Elmer Sciex API 365
LC/MS/MS triple quadrupole mass spectrometer or on a
Fissions ZabSpectraTOF instrument.
Compounds 2 and 3 were synthesized by a procedure modified
from that of Kohda et al.51 The reaction of O6-methylguanine
(0.30 mmol) with ethyl iodide (0.75 mmol) in the presence of
excess of K2CO3 (10 cm3 DMF; 50 ЊC; 1 d) gave 9-ethyl-O6-
methylguanine (2). Similarly, 1-methylguanine (1.5 mmol) when
reacted with ethyl iodide (2.3 mmol) in the presence of excess
of K2CO3 (46 cm3 DMF; 50 ЊC; 20 h) yielded 9-ethyl-1-methyl-
guanine (3). The products were purified by semipreparative
RP-HPLC on a LiChrospher®100 RP-18 (250 × 10 mm, 5µm)
column and characterized by 1H NMR and mass spectroscopy.
The UV spectrum of both compounds was identical to that
of the corresponding 9-methylated compound.51 9-Ethyl-O6-
The NMR spectra were acquired on a JEOL Alpha 500
NMR spectrometer equipped with a 5 mm normal configur-
ation tunable probe or a 5 mm inverse z-axis field-gradient
probe operating at 500.16 for 1H, and 125.78 MHz for 13C. The
deuterium of the solvent was used as lock signal. The spectra
1
were recorded at 30 ЊC in DMSO-d6. The H and 13C spectra
were referenced internally to tetramethylsilane. The 1D proton
spectra were acquired with single-pulse excitation, 45Њ flip
angle, pulse recycle time of 9 sec and with spectral widths of 8
kHz consisting of 65 k data points (digital resolution 0.11 Hz
per pt), zero-filled to 128 k prior to Fourier transformation.
1 Hz of exponential weighting was usually applied prior to
Fourier transformation. 2D Homonuclear H,H-correlation
1
methylguanine: H NMR (DMSO-d6) δ 7.86 (s, 1H, 8-H), 6.36
(s, 2H, NH2), 4.02 (q, 2H, J 7.3 Hz, –CH2–), 3.94 (3H, –OCH3),
1.33 (t, 3H, J 7.3 Hz, –CH3). ESI-MS: m/z 194 (MHϩ). 9-Ethyl-
1-methylguanine: 1H NMR (DMSO-d6) δ 7.70 (s, 1H, 8-H), 6.98
(s, 2H, NH2), 3.96 (q, 2H, J 7.3 Hz, –CH2–), 2.50 (3H, –OCH3),
1.31 (t, 3H, J 7.3 Hz, –CH3). ESI-MS: m/z = 194 (MHϩ).
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1948

View Article Online
1,N 2-(1,2-Dihydroxyethano)-9-ethylguanine (6)
Kinetic measurements
Compound 6 was prepared by a procedure described for the
synthesis of 1,N 2-(1,2-dihydroxyethano)guanine.39 To 5 cm3 of
water, 9-ethylguanine (0.05 mmol) and glyoxal (0.5 mmol as
30 wt% solution in water) were added. The suspension was
stirred for 42 h at 60 ЊC. 20 mm3 of acetic acid was added after
18 h. The suspension was concentrated and kept at 7 ЊC for one
day and filtered. The product was washed with water and dried
in a vacuum. The 1H and 13C NMR spectroscopic data is given
in Tables 1 and 2. ESI-MS: m/z = 260 (MNaϩ), 238 (MHϩ),
220 (MHϩ Ϫ H2O), 192 (MHϩ Ϫ HCOOH). UVmax 251 nm, sh.
280 nm.
Reactions were carried out in stoppered tubes, which were
immersed in a water bath, the temperature of which was kept at
60.0 0.1 ЊC. Aliquots were withdrawn at appropriate intervals,
immediately cooled down on an ice bath and stored in
the freezer until analyzed. The aliquots were analyzed by
RP-HPLC on a Hypersil ODS column (250 × 4 or 4,6 mm,
5
µm) using UV-detection at wavelengths 260 nm for
2Ј-deoxyguanosine and 9-ethylguanine, 279 nm for 9-ethyl-O6-
methylguanine and 255 nm for 9-ethyl-1-methylguanine.
Eluents were mixtures of MeCN and 0.05 mol dmϪ3 acetic acid
buffer (pH 4.3, I = 0.1 M with NH4Cl).
The same compound was isolated from a reaction mixture of
9-ethylguanine and BMA obtained as follows. BMA (20 mmol)
in 0.2 mol dmϪ3 acetic acid buffer (5 cm3; pH 5.5) was heated in
a sealed tube for one week at 60 ЊC. 9-Ethylguanine (6 mmol)
was added, and the suspension was kept at 60 ЊC for one
week. The product was purified by semipreparative HPLC on a
LiChrospher®100 RP-18 (5µm) column. In order to prevent
decomposition of the product, the elute was kept on an ice bath
or in the freezer and evaporations were made in a lyophilizator.
Calculation of the rate constants
Pseudo first order rate constants for the disappearance of the
starting material were calculated by using the integrated first
order rate law. The calculation was based on the decrease of the
signal area as a function of time.
References
Isolation of 9-ethylguanine adducts 10a and 12
BMA (1.51 mmol; 230 mg) in 0.5 mol dmϪ3 phosphate buffer
(pH 7.3, 50 cm3) was heated at 60 ЊC for one week. A mixture of
BMA (230 mg) and 9-ethylguanine (1.50 mmol; 300 mg) was
then added and the incubation was continued for an additional
5 d. The solution was concentrated to 10 cm3, filtered and
eluted through a LiChroprep RP-18 column ( 37 × 440 mm,
40–63 µm) with 0.1 mol dmϪ3 aqueous ammonium acetate
containing 3% MeCN (v/v). Two main fractions were collected,
concentrated to 5 cm3 and subjected to desalting on a Hyper-
prep HS-18 column (10 × 250 mm, 8 µm) using aqueous
acetonitrile as eluent.
1 P. M. Dansette, R. Snyder, M. Delaforge, G. G. Gibson, H. Greim,
D. J. Jollow, T. J. Monks and I. G. Sipes, Advances in Experimental
Medicine and Biology, vol. 500, Kluwer Academic/Plenum
Publishers, 2001.
2 R. C. Moschel and N. J. Leonard, J. Org. Chem., 1976, 41, 294–300.
3 H. Seto, T. Okuda, T. Takesue and T. Ikemura, Bull. Chem. Soc. Jpn,
1983, 56, 1799–1802.
4 A. K. Basu, S. M. OЈHara, P. Valladier, K. Stone, O. Mols and
L. J. Marnett, Chem. Res. Toxicol., 1988, 1, 53–59.
5 E. Eder and C. Hoffman, Chem. Res. Toxicol., 1993, 6, 486–496.
6 G. R. Reddy and L. J. Marnett, Chem. Res. Toxicol., 1996, 9, 12–15.
7 F.-L. Chung, R. Young and S. S. Hecht, Cancer Res., 1984, 44,
272–282.
8 D. W. Boerth, E. Eder, S. Hussain and C. Hoffman, Chem. Res.
Toxicol., 1998, 11, 284–294.
9 E. Eder and C. Hoffman, Chem. Res. Toxicol., 1992, 5, 802–808.
10 F.-L. Chung and S. S. Hecht, Cancer Res., 1983, 43, 1230–1235.
11 N. K. Kotchetkov, V. N. Shibaev and A. A. Kost, Tetrahedron Lett.,
1971, 1993–1997.
12 J. A. Secrist, III, J. R. Barrio, N. J. Leonard and G. Weber,
Biochemistry, 1972, 11, 3499–3506.
13 K. Kayasukada-Mikado, T. Hashimoto, T. Negishi, K. Negishi and
H. Hayatsu, Chem. Pharm. Bull., 1980, 28, 932–939.
14 J. T. Kusmierek and B. Singer, Chem. Res. Toxicol., 1992, 5, 634–638.
15 F. P. Guengerich and M. Persmark, Chem. Res. Toxicol., 1994, 7,
205–208.
Isolation of guanosine adducts
Guanosine (0.57 g, 2 mmol) and BMA (0.61 g, 4 mmol) were
stirred in a phosphate buffer (20 mL; pH 10.0; 70 ЊC, 72 h). The
solution was cooled to room temperature and evaporated to
dryness under reduced pressure. The crude mixture was dissol-
ved in 10 cm3 of water, the insoluble materials were filtered off,
and the filtrate was applied onto an RP-18 column (40 µm). The
products were eluted with a mixture of MeOH and a phosphate
buffer increasing the MeOH content in a stepwise manner: 0%
(20 min), 5% (1 h), 11% (1 h), 22% (30 min), 25% (3 h), and 30%
(1 h). The main products were collected as follows: 10b and
11b as a single peak with 11% MeOH, and 12 with 30% MeOH.
The fractions were concentrated under a reduced pressure
and desalted by semipreparative RP-chromatography using
7% aqueous acetonitrile as an eluent. Compounds 10b and
11b were separated during the desalting. The desalted fractions
were evaporated to dryness and analyzed by NMR
spectroscopy.
16 V. Nair, R. J. Offerman and G. A. Turner, J. Org. Chem., 1984, 49,
4021–4025.
17 L. Kronberg, S. Karlsson and R. Sjöholm, Chem. Res. Toxicol.,
1993, 6, 495–499.
18 S. Mikkola, N. Koissi, K. Ketomäki, S. Rauvala, K. Neuvonen and
H. Lönnberg, Eur. J. Org. Chem., 2000, 2315–2323.
19 F. P. Guengerich, Chem. Res. Toxicol., 1992, 5, 2–5.
20 M. K. Dosanjh, A. Chenna, E. Kim, H. Fraenkel-Konrat,
L. Samson and B. Singer, Proc. Natl. Acad, Sci. USA, 1994, 91,
1024–1028.
21 F. P. Guengerich, M. Persmark and W. G. Humpreys, Chem. Res.
Toxicol., 1993, 6, 635–648.
22 L. Kronberg, R. Sjöholm and S. Karlsson, Chem. Res. Toxicol.,
1992, 5, 852–855.
23 D. Asplund, L. Kronberg, R. Sjöholm and T. Munter, Chem. Res.
Toxicol., 1995, 8, 841–846.
24 L. Kronberg, D. Asplund, J. Mäki and R. Sjöholm, Chem. Res.
Toxicol., 1996, 9, 1257–1263.
25 T. Munter, F. Le Curieux, L. Kronberg and R. Sjöholm, Chem. Res.
Toxicol., 1999, 12, 42–52.
26 P. Yi, X. Sun, D. R. Doerge and P. P. Fu, Chem. Res. Toxicol., 1998,
11, 1032–1041.
The 1H and 13C NMR spectroscopic data is given in Tables 1
and 2. Compounds 10b: UVmax (H2O) 260 nm, UVmin (H2O) 226
nm. FAB-MS m/z (relative intensity): 683.2 [(2M ϩ H)ϩ, 45%],
364.2 (MNaϩ, 40%), 342.3 (MHϩ, 100%), 210.2 (MHϩ
Ϫ
C5H8O4, 40%). Compound 11b: UVmax (H2O) 260 nm, UVmin
(H2O) 222 nm. FAB-MS m/z (relative intensity): 400.0 (MHϩ,
100%), 268.2 (MHϩ Ϫ C5H8O4, 25%).
Reaction solutions
The pH of the reaction solutions was adjusted with formic acid
(pH 3–4), acetic acid (pH 4–5.5), MES (pH 5.5–6.5), HEPES
(pH 6.5–8) and CHES (pH 8.5–10) buffers. Below pH 3,
aqueous hydrogen chloride was employed for the same purpose.
The pH of the reaction solutions was checked with a pH-meter
at 60 ЊC. Ionic strength was adjusted with NaNO3 or NaCl.
27 M. Müller, F. J. Belas, I. A. Blair and F. P. Guengerich, Chem. Res.
Toxicol., 1997, 10, 242–247.
28 K. Hakala, S. Auriola, A. Koivisto and H. Lönnberg, J. Pharm.
Biomed. Anal., 1999, 21, 1053–1061.
29 M. Otteneder, J. P. Plastaras and L. J. Marnett, Chem. Res. Toxicol.,
2002, 15, 312–318.
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1949

View Article Online
30 I. Terashima, N. Suzuki and S. Shibutani, Chem. Res. Toxicol., 2002,
15, 305–311.
31 O. J. Schmitz, C. C. T. Wörth, D. Stach and M. Wiessler, Angew.
Chem., Int. Ed., 2002, 41, 445–448.
32 J. R. Barrio, J. A. Secrist, III and N. J. Leonard, Biochem. Biophys.
Res. Commun., 1972, 46, 597–604.
33 P. D. Sattsangi, N. J. Leonard and C. R. Frihart, J. Org. Chem., 1977,
42, 3292–3296.
H. Bartsch, International Agency for Research on Cancer, Lyon,
1999, pp. 155–168.
41 R. Laatikainen, M. Niemitz, S.-P. Korhonen and T. Hassinen,
PERCH NMR Software, version 2000, Kuopio University NMR
Research Group, Department of Chemistry, University of Kuopio,
P.O. BOX 1627, FIN-70211, Kuopio, Finland.
42 M. Awada and P. C. Dedon, Chem. Res. Toxicol., 2001, 14,
1247–1253.
34 K. Neuvonen, C. Zewi and H. Lönnberg, Acta Chem. Scand., 1996,
50, 1137–1142.
35 N. Koissi, K. Neuvonen, T. Munter, L. Kronberg and H. Lönnberg,
Nucleosides, Nucleotides, Nucleic Acids, 2001, 20, 1761–1774.
36 K. Neuvonen, N. Koissi and H. Lönnberg, J. Chem. Soc., Perkin
Trans. 2, 2002, 173–177.
37 B. Singer, S. J. Spengler, F. Chavez and J. T. Kusmierek,
Carcinogenesis, 1987, 8, 745–747.
38 C. Glemarec, Y. Besidsky, J. Chattopadhyaya, J. Kusmierek,
M. Lahti, M. Oivanen and H. Lönnberg, Tetrahedron, 1991, 47,
6689–6704.
43 R. Shapiro and J. Hachmann, Biochemistry, 1966, 5, 2799–2807.
44 M. A. Glomb and G. Lang, J. Agric. Food. Chem., 2001, 49,
1493–1501.
45 W. Seidel and M. Pischetsrieder, Bioorg. Med. Chem. Lett., 1998, 8,
2017–2022.
46 E. P. Whipple, J. Am. Chem. Soc., 1970, 92, 7183–7186.
47 C. Rabiller, New J. Chem., 1987, 11, 419–424.
48 B. T. Golding, N. Patel and W. P. Watson, J. Chem. Soc., Perkin
Trans. 2, 1989, 668–669.
49 L. J. Marnett, A. K. Basu, S. M. OЈHara, P. E. Weller,
A. F. M. M. Rahman and J. P. Oliver, J. Am. Chem. Soc., 1986, 108,
1348–1350.
50 S. Trofimenko, J. Org. Chem., 1963, 28, 3243–3245.
51 K. Kohda, K. Baba and Y. Kawazoe, Tetrahedron Lett., 1987, 28,
6285–6288.
39 R. Shapiro, B. I. Cohen, S.-J. Shiuey and H. Maurer, Biochemistry,
1969, 8, 238–245.
40 R. N. Loeppky, W. Cui, P. Goelzer, M. Park and Q. Ye, in Exocyclic
DNA adducts in mutagenesis and carcinogenesis, eds. B. Singer, and
O r g . B i o m o l . C h e m . , 2 0 0 4 , 2, 1 9 4 3 – 1 9 5 0
1950