Properties of the OH adducts of hydroxy-, methyl-, methoxy-, and amino-substituted pyrimidines: Their dehydration reactions and end-product analysis

Article abstract of DOI:10.1021/jp0117720

Reactions of hydroxyl radicals (OH) with 2-amino-4-methyl pyrimidine (AMP), 2-amino-4,6-dimethyl pyrimidine (ADMP), 2-amino-4-methoxy-6-methyl pyrimidine (AMMP), 2-amino-4-hydroxy-6-methyl pyrimidine (AHMP), 4,6-dihydroxy-2-methyl pyrimidine (DHMP), 2,4-d

Full text of DOI:10.1021/jp0117720

J. Phys. Chem. A 2002, 106, 2497-2504
2497
Properties of the OH Adducts of Hydroxy-, Methyl-, Methoxy-, and Amino-Substituted
Pyrimidines: Their Dehydration Reactions and End-Product Analysis
T. L. Luke,† T. A. Jacob,† H. Mohan,‡ H. Destaillats,§ V. M. Manoj,† P. Manoj,† J. P. Mittal,*,‡
M. R. Hoffmann,*,§ and C. T. Aravindakumar*,†
School of Chemical Sciences, Mahatma Gandhi UniVersity, Kottayam 686 560, India, Radiation Chemistry and
Chemical Dynamics DiVision, Bhabha Atomic Research Centre, Mumbai, 400 085, India, and
W.M. Keck Laboratories, California Institute of Technology, Pasadena, California 91125
ReceiVed: May 9, 2001; In Final Form: September 13, 2001
Reactions of hydroxyl radicals (•OH) with 2-amino-4-methyl pyrimidine (AMP), 2-amino-4,6-dimethyl
pyrimidine (ADMP), 2-amino-4-methoxy-6-methyl pyrimidine (AMMP), 2-amino-4-hydroxy-6-methyl py-
rimidine (AHMP), 4,6-dihydroxy-2-methyl pyrimidine (DHMP), 2,4-dimethyl-6-hydroxy pyrimidine (DMHP),
6-methyl uracil (MU), and 5,6-dimethyl uracil (DMU) have been studied by pulse radiolysis and steady-state
•
radiolysis techniques at different pH values. The second-order rate constants of the reaction of OH with
these systems are of the order of (2-9) × 109 dm3 mol-1 s-1 at near neutral pH. The difference in the
spectral features of the intermediates at near neutral pH and at higher pH (10.4) obtained with these pyrimidines
are attributed to the deprotonation of the OH adducts. The G(TMPD•+) obtained at pH ∼ 6, from the electron-
transfer reactions of the oxidizing intermediates with the reductant, N,N,N′,N′-tetramethyl-p-phenylenediamine
(TMPD), are in the range (0.2-0.9) × 10-7 mol J-1 which constituted about 3-16% oxidizing radicals.
These yields were highly enhanced at pH 10.5 in the case of AHMP, DHMP, DMU, and MU (G(TMPD•+)
) 3.8-5.5 = 66-95% oxidizing radical). On the basis of these results, it is proposed that a nonoxidizing
C(6)-ylC(5)OH radical adduct is initially formed at pH 6 which is responsible for the observed transient
spectra. The high yield of TMPD•+ at higher pH is explained in terms of a base-catalyzed conversion (via a
dehydration reaction) of the initially formed C(6)-ylC(5)OH adduct (nonoxidizing) to C(5)-ylC(6)OH adduct
which is oxidizing in nature. Among the selected pyrimidines, such a dehydration reaction was observed
only with those having a keto (or hydroxy) group at the C(4) position of the pyrimidine ring. Qualitative
analyses of the products resulting from the OH adducts of DHMP (at pH 4.5) and DMHP (at pH 6) were
carried out using HPLC-ES-MS and a variety of products have been identified. Glycolic and dimeric products
were observed as the major end-products. The product profiles of both DHMP and DMHP have shown that
the precursors of the products are mainly the C(6)-ylC(5)OH and the H adduct radicals. The identified products
are formed mainly by disproportionation and dimerization reactions of these radicals. The mechanistic aspects
are discussed.
Introduction
oxidizing and reducing OH adducts was reported in the case of
adenine.7 In purine nucleosides and nucleotides, 25% of the
products were reported to be formed by the primary attack of
•OH at the sugar moiety.3
Reactions of primary free radicals such as hydroxyl radical
(•OH), hydrogen atom (•H), and hydrated electron (e-aq),
generated by the radiolysis of water, with nucleic acid compo-
nents have been a subject of increased attention for the last many
decades using pulse and steady-state radiolysis techniques.1-4
The reactions of •OH with both purines and pyrimidines are of
In the case of pyrimidines, the main reaction is also the
addition at C(5)-C(6) double bond while the double bond
between N(3) and C(4) of cytosine is a potential additional
site.8-12 The major OH radical adducts are, therefore, C(5)-
ylC(6)-hydroxy and C(6)-ylC(5)-hydroxy radicals. The former
has oxidizing properties with respect to N,N,N′,N′-tetramethyl-
p-phenylenediamine (TMPD) and the latter has reducing proper-
ties with respect to tetranitromethane (TNM). The ratio of the
oxidizing to reducing radicals depends on the nature of the
substituents present in the C(5) and C(6) position of the
pyrimidine ring whereas such substituent effect is not predomi-
nant with pyrimidines substituted at N(1) position.9,10 The
deprotonation of the OH adduct normally occurs at N(1) when
there is a hydrogen at N(1) position. On the bases of the redox
properties of the radicals, it has been identified that the
distribution of C(6)-yl radical in uracil is 80% while that of
thymine, cytosine and 1-methyluracil are 56, 87, and 65%
•
particular interest because OH is the major damaging agent
that leads to DNA lesions due to ionizing radiation.3 •OH reacts
with both purines and pyrimidines at almost diffusion controlled
rate (g109 dm3 mol-1 s-1).5 It generally undergoes addition at
C(4), C(5), and C(8) positions of the purine ring. The resulting
OH adducts in the case of 2′-deoxyguanosine were reported to
have both oxidizing [G(4)OH] and reducing [G(5)OH and G(8)-
OH] properties with respect to some known oxidants and
reductants.6,7 A more or less similar distribution of both
* To whom correspondence should be addressed. Fax: +91-481-597731.
E-mail: mgu@vsnl.com.
† School of Chemical Sciences.
‡ Radiation Chemistry and Chemical Dynamics Division.
§ W.M. Keck Laboratories.
10.1021/jp0117720 CCC: $22.00 © 2002 American Chemical Society
Published on Web 02/26/2002

2498 J. Phys. Chem. A, Vol. 106, No. 11, 2002
Luke et al.
TABLE 1: Second-Order Rate Constantsa and the
respectively.11 A third type of radical derived by a H-abstraction
from the methyl group of pyrimidine has also been reported.11
An interesting property of the reducing radicals (C(6)-ylC(5)-
OH) of pyrimidines is their base-catalyzed dehydration reaction
to yield oxidizing radicals (C(5)-ylC(6)OH) which can oxidize
TMPD to TMPD•+.9,10 To look at this important transformation
reaction in basic pHs more closely, a series of pyrimidines
substituted by methyl, hydroxy, amino, and methoxy groups at
different positions have been selected and investigated the effect
of these various substituents as well as their positions in the
pyrimidine ring on the dehydration reaction. TMPD has been
utilized to determine the ratio of oxidizing to non-oxidizing
radicals and to demonstrate the transformation reaction. The
selected pyrimidines include 2-amino-4-methyl pyrimidine
(AMP), 2-amino-4,6-dimethyl pyrimidine (ADMP), 2-amino-
4-methoxy-6-methyl pyrimidine (AMMP), 2-amino-4-hydroxy-
6-methyl pyrimidine (AHMP), 4,6-dihydroxy-2-methyl pyrim-
idine (DHMP), 2,4-dimethyl-6-hydroxy pyrimidine (DMHP),
6-methyl uracil (MU). and 5,6-dimethyl uracil (DMU). Both
the kinetics of the reaction and the spectral nature of the
intermediates were also investigated at different pH values.
One of the difficult areas in the field of free radical chemistry
of biomolecules is the end product analyses. This is mainly due
to low concentrations of the products (≈1 × 10-6 mol dm-3 or
less) and hence cannot be easily analyzed using the normal
analytical techniques. GC-MS analysis after rotary evaporation
followed by silylation of the dry products has made some
progress in this area.12 However, such studies are restricted only
to few DNA model systems. In the present study, we have used
a more sophisticated analytical technique, HPLC connected to
electrospray mass spectrometer (HPLC-ES-MS), and analyzed
the end products directly in aqueous medium resulting from
Absorption Maxima of the Intermediates Obtained for the
Reaction of OH Radicals with the Selected Pyrimidines
pH 6
k2/109
pH 10.4
b
compound
AMP
(dm3 mol-1 s-1
)
λ
max (nm)
λ
max (nm)
2.0
7.2
5.8
315
550
310
550
320
455
320
550
320
565
300
460
750
300
520
290
390
480
330
455
310
425
310
440
ADMP
AHMP
AMMP
DHMP
6.5
300
515
420
5.6,c (3.7)d
DMHP
DMU
MU
6.2
5.9
9.0
325
450
435
410
a These were determined from the average value of slopes of kobs vs
concentration plots for intermediates having two maxima (these two
values were nearly the same). b The deviation in the calculation of these
values was between 5 and 10%. c Measured at pH 4.5. d From reference
19.
radiolysis set up have been described elsewhere.15,16 The pH of
the selected pyrimidines was found to be around 6.0 when
dissolved in water and was maintained the same for all of the
experiments at near neutral pH. However, a pH of 4.5 was
maintained in the case of DHMP in order to keep it in its neutral
form as the pK values are 0.21, 6.35, and 12.9.17
N2O-saturated aqueous solutions of DHMP (1 × 10-3 mol
dm-3) and DMHP (1 × 10-3 mol dm-3) were irradiated in a
60Co-γ-source (maximum dose ∼2 kGy and the dose rate was
determined by ceric sulfate dosimetry18) and then injected into
a Hewlett-Packard 1100 HPLC system with UV detection and
coupled to a mass spectrometer through an electrospray interface
(ES-MS). The mass specta were recorded for both positive and
negative ions in the range 50-1000 m/z units at 3 s/scan and a
skimmer voltage of 3.5 kV. A 3 m, 100 × 4 mm Hypersil BDS-
C18 column (Hewlett-Packard) was used. The eluent was
Millipore-Milli-Q water and the flow rate was fixed at 0.5 mL/
min.
•
the reaction of OH with DHMP and DMHP. It will be shown
that HPLC-ES-MS is an ideal choice for the direct detection of
such radiolytic products and that this technique may have
advantages over GC-MS analysis in which the products have
to undergo lengthy and rigorous steps before being analyzed.
Experimental Section
Commercially available AMP, ADMP, AHMP, AMMP,
DMHP, DHMP, DMU, MU, and TMPD (Aldrich) were used
without further purification. The radiolysis of water produces
three highly reactive species (e-aq, •H, and •OH) in addition to
the formation of inert or less reactive molecular products (H2,
H2O2, etc.) (reaction 1).
H2O Df eaq-, H•, H2, H2O2, H3O+
(1)
Results and Discussion
I. Pulse Radiolysis. The second-order rate constants of the
In N2O-saturated solutions, eaq- are quantitatively converted to
reaction of •OH with the selected pyrimidines were found to be
•OH (reaction 2), and therefore, the transient species available
diffusion controlled ((2-9) × 109 dm3mol-1 -1
s ) at near neutral
•
for reaction are OH (G(•OH) ) 5.3-6.8 × 10-7 mol J-1 13
)
pH and the values are tabulated in Table 1. These were
determined from the plot of the pseudo first-order formation of
intermediates (kobs) versus the concentration of the substrate.
The kobs was monitored at the respective absorption maxima of
the intermediates (see Table 1). Such plots gave straight line
graphs with good correlation coefficient (g0.99) with all the
compounds. The absorption build-up of the intermediate at 320
nm and a typical kobs versus concentration plot obtained with
AHMP (at pH 6) are shown in Figure 1. The high rate constants
obtained for these compounds are in good agreement with other
pyrimidines reported earlier.5
and •H (G(•H) ) 0.6 × 10-7 mol J-1) (where G value represents
the radiolytic yield).
N2O + eaq- f •OH + OH- + N2
(2)
N2O-saturated aqueous solutions typically contained 10-3 mol
dm-3 of the substrate were pulse irradiated using a linear
accelerator delivering 7 MeV electron pulses of 50 ns duration.
KSCN dosimetry was used to determine the dose per pulse using
Gꢀ(500 nm) ) 2.15 × 10-4 m2 J-1 in aerated solutions,14 and the
dose per pulse was normally kept at 15 Gy. Low dose per pulse
of 5 Gy was used for the study of electron-transfer reaction
between the OH adducts and TMPD. The details of the pulse
The transient absorption spectra showed two absorption
maxima at pH 6 with the selected pyrimidines except for DHMP
which has only a single peak at 420 nm. The transient absorption

OH Adducts of Substituted Pyrimidines
J. Phys. Chem. A, Vol. 106, No. 11, 2002 2499
Figure 1. (a) A build-up of the intermediate at 320 nm ([AHMP] ) 2 × 10-4 mol dm-3) at pH 6. (b) kobs vs concentration plot at 320 nm at pH
6. (c) Absorbance measured at 2 µs after the pulse versus pH at 700 nm. (d) decay trace of the intermediate at 300 nm at pH 10.4. (e) The TMPD•+
build-up at 565 nm ([AHMP] ) 2 × 10-3 mol dm-3 and [TMPD] ) 5 × 10-5 mol dm-3) at pH (i) 6.0 and (ii) 10.4, obtained in N2O-saturated
aqueous solutions of 2-amino-4-hydroxy-6-methyl pyrimidine (AHMP)
second-order kinetics. The spectrum at higher pH (∼10.4) also
has similar characteristics with the same maxima but with higher
absorbance values (Figure 1S, see Supporting Information). In
the case of AMMP, the spectrum has λmax at 300 and 515 nm
at pH 6. At higher pH (10.4) there is a slight change in the
maxima which are at 300 and 520 nm. This spectrum was found
to decay by second-order kinetics. The intermediate obtained
with AMP has absorption maxima around 315 and 550 nm at
pH 6. Identical spectra were obtained at higher pH except there
is a slight difference in the lower λmax, which is at 320 nm.
The transient spectra obtained with DHMP is different from
that obtained for the other pyrimidines. In this case the spectrum
is characterized by its single λmax at around 420 nm at pH 4.5.
The spectrum obtained at 2 µs after the pulse at higher pH (10.4)
showed three absorption maxima at 290, 390, and 480 nm
Figure 2. Transient absorption spectra obtained in N2O saturated
(Figure 3). These spectral features are in line with an earlier
aqueous solutions of 2-amino-4-hydroxy-6-methyl pyrimidine (AHMP)
report on the OH adduct of DHMP.19 The initial spectrum
(1 × 10-3 mol dm-3) at 2 µs after the pulse at pH 6 (O) and at 2 µs
undergoes a very fast first-order type decay (k ) 6 ×105 s-1
)
(2) and 40 µs (×) after the pulse at pH 10.4 (dose per pulse ) 15 Gy).
followed by a slow build up of absorption around 290-300
nm. The spectrum with DMHP showed two absorption maxima
at 325 and 450 nm at pH 6 and at pH 10.4 with nearly same
absorbance values (Figure 2S, see Supporting Information). The
transient spectra obtained with DMU and MU have two maxima
(295 and 410 nm for DMU and 285 and 410 nm for MU) at
pH 6. The transients at their lower abosorption maxima (at 295
and 285 nm, respectively, from DMU and MU) were found to
undergo a very fast decay. We have recently identified these
peaks as due to the deprotonated DMU and MU (at N(1)) which
result from the reaction of radiolytically formed OH- in N2O-
saturated solution (reaction 2) even at this short time scale and
the fast decay is only a (re)protonation.20 In the present case
too, a similar deprotonated DMU and MU (at N(1)) could be
formed and therefore, the peaks at 295 and 285 nm are very
unlikely to be due to the OH adducts. The time-resolved
spectrum obtained with AHMP at pH 6 has maxima around
320 and 455 nm (Figure 2). The initial spectrum was found to
undergo a second-order decay. However, when the pH was
raised to 10.5, the spectrum showed three maxima. An additional
peak appeared around 750 nm while the lower λmax is blue
shifted to 300 nm with a shoulder around 330 nm (Figure 2).
The initial spectrum undergoes a fast decay, (first-order type, k
∼ 106 s-1) with a slow absorption build-up around 300 nm
region. This is found to continue over 300 µs and then undergoes
a bimolecular decay. A typical trace obtained at 300 nm is also
shown in Figure 1. An absorbance versus pH plot measured at
2 µs after the pulse at 700 nm, shown in Figure 1, gave an
inflection point at pH 9. The time-resolved spectra obtained for
ADMP at pH 6 have absorption maxima at 310 and 550 nm.
The spectrum at 2.5 µs after the pulse was found to decay by

2500 J. Phys. Chem. A, Vol. 106, No. 11, 2002
Luke et al.
•OH generally undergoes addition reaction at C(5)-C(6)
double bond of the pyrimidines according to earlier reports.8-11
The resulting transient spectra are characterized by their
absorption maxima around 320 and 430 nm. On the bases of
the reactions of the OH adducts with TMPD, a comparatively
high yield of the reducing radicals (C(5)OH-C(6)yl radicals)
were reported to be formed with a number of uracil derivatives.10
In agreement with these reports we propose a similar OH
addition at the C(5) and C(6) positions of the selected pyrim-
idines at pH 6 as well as at pH 10.4 (reaction 3). The transient
spectra obtained with these compounds are also having two
absorption maxima in the same region except for AMP, ADMP,
and AMMP where the second λmax is in the 520-550 nm region.
This slight difference in the second λmax could be due to the
substituent effect.
Figure 3. Transient absorption spectra obtained in N2O saturated
aqueous solutions of DHMP (1 × 10-3 mol dm-3) at 1 µs after the
pulse at pH 6 (2) and at 2 µs (O) and 90 µs (() after the pulse at pH
10.5 (dose per pulse ) 15 Gy). Inset: (a) intermediate trace obtained
at 290 nm at pH 10.4. (b) The TMPD•+ build-up at 565 nm obtained
with DHMP (2 × 10-3 mol dm-3) in the presence of TMPD at pH
10.4 (5 × 10-5 mol dm-3) (dose per pulse ) 5 Gy).
TABLE 2: The G(TMPD•+) (×107/mol J-1) Values and the
Percentage of Oxidizing Radicals Obtained for the Selected
Pyrimidines at pH 6 and 10.4
The G(TMPD•+) values at near neutral pH, given in Table
2, show a clear account of the yield of oxidizing radicals. Such
low yield of oxidizing radicals is in agreement with the earlier
reports on the OH adducts of thymine and uracil derivatives.10,11
The oxidizing property of the C(5)-ylC(6)OH radical can be
understood as its mesomeric structure is either an oxygen-
centered radical (at C(4)-O) or a nitrogen centered (at N(3)).
Therefore, this gives an additional support for the assignment
of the spectra to the formation of C(6)-ylC(5)OH adduct in the
case of all the selected pyrimidines which could act as reducing
radical as was reported in the case of uracil and thymine.11
Furthermore, there can be a possibility of H-abstraction reaction
from the methyl group of these compounds leading to allyl type
radicals as reported in the case of thymine.11 However, these
radicals cannot be reducing unlike that of thymine. Therefore,
if formed, these radicals could also contribute to the TMPD•+
yield. But this is expected to be a minor process.
percentage of
G(TMPD•+) at pH
oxidizing radicalsa at pH
compound
6.0
10.4
6.0
10.4
AMP
0.72
0.91
0.37
0.72
0.44b
0.37
0.34
0.19
0.68
0.75
5.49
0.74
5.18
0.32
3.83
4.76
12.5
15.7
6.4
12.5
7.5
6.3
5.9
11.8
13.0
94.6
12.8
89.3
5.5
ADMP
AHMP
AMMP
DHMP
DMHP
DMU
66.0
82.0
MU
3.2
a Percentages are calculated based on a maximum yield of the
.
intermediates, G(intermediates) ≈ 5.7 × 10-7 mol J-1 b At pH 4.5.
transient spectra obtained with MU at 10.4 showed absorption
maxima at 310 and 440 nm (Figure 3S, see Supporting
Information) and that of DMU at 10.4 has maxima at 310 and
425 nm (Figure 4S, see Supporting Information). In both cases,
the spectra recorded at higher time scales (>40 µs) have only
a single λmax around 440 nm.
The transient absorption spectra at higher pHs (∼10.4) were
different from those at lower pHs either by their higher
absorption coefficient or by the shift in their λmax (including
additional peaks such as those in the cases of AHMP and
DHMP). Such differences in the case of uracil, thymine, and
cytosine were explained in terms of the formation of deproto-
nated OH adducts at higher pHs.8,18 In the present case too,
such an interpretation is logical. The absorbance versus pH plot
in the case of AHMP (Figure 1) gave a clear pK type curve
with inflection point around pH 9. AHMP has pKa values as
4.1 and 9.4. Therefore, it can be concluded that the species
existing at pH 10.4 is the deprotonated form of the OH adduct
of AHMP. On the bases of these observations and on the
previous reports,8,19 the spectral differences at lower and higher
pHs, are attributed to the protonation-deprotonation reaction
of the OH adducts. However, in the case of DMHP, there was
no observable difference between the absorption spectra at pH
6 and 10.5 and hence it must be assumed that the pKa value of
the OH adduct is higher than pH 10.5 and that the spectrum
existing at both the pH are the neutral OH adduct of DMHP.
As shown in Table 2, the percentage of oxidizing radicals
calculated from the G(TMPD•+) values were almost constant
for AMP, ADMP, AMMP, and DMHP at lower and higher pH
values, but for AHMP, DHMP, DMU, and MU the calculated
percentage of oxidizing radicals were 95, 89, 66 and 82%,
respectively, at higher pH. The high difference in the G(T-
•
The reaction of OH was carried out at low doses in the
presence of TMPD which is an effective reductant and can be
oxidized to TMPD•+ by transferring one electron to an oxidizing
intermediate.9-11 The build-up of TMPD•+ has been monitored
with solutions contained 2 × 10-3 mol dm-3 pyrimidine and 5
× 10-5 mol dm-3 TMPD at 565 nm. The yields of the radical
cation, G(TMPD•+), were calculated for all these pyrimidines
at near neutral pH and at higher pH (∼10.4), and are sum-
marized in Table 2. The percentage of oxidizing radicals were
thus calculated for all the pyrimidines based on a maximum
G(TMPD•+) as 5.7 × 10-7 mol J-1 and are also tabulated in
Table 2. The G(TMPD•+) values are almost the same at pH 6
and pH 10.4 for AMP, ADMP, AMMP, and DMHP ((0.3-
0.9) × 10-7 mol J-1, Table 2). However, in the cases of AHMP,
DHMP, DMU, and MU, a much higher yield of TMPD•+ ((3.8-
5.5) × 10-7 mol J-1) is obtained at higher time scales (Table
2). A well-defined build-up of TMPD•+ was obtained at pH
10.5 against the feeble absorbance at pH 6 in all these cases
and the typical traces obtained in the case of AHMP and DHMP
are shown in Figures 1 and 3, respectively (also in Figures 3S
and 4S for MU and DMU, respectively: see Supporting
Information).

OH Adducts of Substituted Pyrimidines
J. Phys. Chem. A, Vol. 106, No. 11, 2002 2501
SCHEME 1
B (due to its mesomeric form B′) can be oxidizing in nature.
The redox properties of similar radicals such as from thymine
and 1-methyl uracil are already well documented.11
The fast first-order decay of the transients at higher pHs is,
therefore, proposed as the dehydration of the initially formed
reducing radicals in the cases of AHMP, DHMP, DMU, and
MU. It is also important to note that such dehydration of the
reducing radicals and the transformation reactions have been
observed only with AHMP, DHMP, DMU, and MU. The major
structural difference of these compounds with others is that these
contain either a keto or hydroxy group at the C(4) of the
pyrimidine ring. It is therefore obvious that the chemical
environment in the cases of other pyrimidines, such as ADMP,
AMP, AMMP, and DMHP, is not conducive for an OH-
elimination reaction.
MPD•+) value at lower and higher pHs gives an indication of
the change in the oxidizing property of the intermediate radicals.
A two component build-up of TMPD•+ absorption has been
reported with uracil at pH>9 where the initial component
depended on the concentration of TMPD.10 The second com-
ponent, on the other hand, was dependent only on the pH (and
not on the concentration) indicating a slow transformation (in
the pulse scale) of the initial non-oxidizing species into an
oxidizing species. Therefore, in the present case, we have
monitored the build-up of TMPD•+ at a higher time scale with
the pyrimidines at pH ∼ 10.4 and the G(TMPD•+) values were
determined (Table 2). The mechanism of the change in the
oxidizing properties at higher pH values, observed in the case
of AHMP, DHMP, DMU, and MU, is proposed on the basis of
the conversion of the initially formed (deprotonated) OH adduct
(C(6)-ylC(5)OH) to C(5)-ylC(6)OH. In the case of cytosine, this
occurs by the dehydration of C(5)OH adducts (reducing in
nature) which ultimately leads to the formation of a radical site
at C(5) or oxygen at C(2) position.9,10 Therefore, a similar
conversion of the initially formed C(6)-ylC(5)OH to an oxidizing
radical at basic pHs is proposed as shown in the case of AHMP
(Scheme 1).
A similar type of dehydration reaction of the C(6)-ylC(5)-
OH radical in basic medium is proposed for DHMP, DMU, and
MU. In the case of DHMP, EPR evidence is available for the
formation of an oxyradical at basic pHs19 and this assignment
is in good agreement with our observation of a high G(TMPD•+)
value at pH 10.5 (Table 2). The relatively low value of
G(TMPD•+) obtained with DMU could be understood on the
basis of possible formation of two kinds of allyl type radicals
(A and B) from a H-abstraction from the methyl groups. It could
be probable that the radical A is formed in high yield compared
to the radical B. The radical A could be reducing, but cannot
be transformed into an oxidizing radical. The distribution of
Some additional spectral features obtained with AHMP,
DHMP, DMU, and MU at high pHs further gave some
indication of the optical absorption spectra of the oxidizing
radicals. The absorption trace obtained with AHMP at 300 nm,
given in Figure 1, showed a slow build-up of absorption after
a very fast first-order decay (corresponding to the dehydration)
of the transient. As this is not due to the formation of a stable
product (it started a second-order decay after about 300 µs),
this is likely due to the absorption of the oxidizing radical which
is formed after the dehydration reaction. A similar absorption
build-up was observed in the case of DHMP as well (Figure
3). While the initial spectra had undergone a fast decay at the
maxima 290, 390, and 480 nm, the later spectrum (90 µs after
the pulse) appears to have a single prominent peak at 300 nm.
The absorption trace at 290 nm (in the inset of Figure 3) has
also shown a build-up similar to AHMP. This trace was found
to undergo a second-order decay after about 500 µs (data not
shown). Therefore, in this case, too, it is likely that the radical
which is formed after the dehydration reaction of C(6)-ylC(5)-
OH of DHMP, contribute to the absorption build-up around 300
nm. Such a difference in the time-resolved absorption spectra
in the case of both DMU and MU was also prominent. The
initial spectra were found to have two absorption maxima where
the later spectra (85 and 40 µs respectively after the pulse) have
only a single λmax (Figures 3S and 4S, see Supporting Informa-
tion) which undergo a bimolecular decay. These spectral features
also give some indication of the contribution of absorption from
the oxidizing radicals of DMU and MU.
II. Product Analysis. Qualitative HPLC-ES-MS analysis of
N2O-saturated solution of DHMP (pH∼4.5) and DMHP (pH∼6)
after γ-irradiation gave a large number of mass peaks with
varying m/z values.21 Among these mass peaks a number of
m/z values were selected for the most probable products
(protonated or deprotonated under mild ionization conditions).
The identified products along with their m/z values are sum-
marized in Tables 3 and 4. Figure 5S, see Supporting Informa-
tion, illustrates a typical mass spectra obtained from a given
chromatographic peak in the case of DMHP. The probable
mechanism of the formation of the identified products resulting
from the proposed radical intermediates from DHMP and
DMHP, is shown in Schemes 2 and 3, respectively. As discussed
in section I, C(5)OH adduct is the major radical intermediate
in the case of DHMP and DMHP at near neutral pH. Therefore,
the products are likely to result from the disproportionation and
dimerization reactions of these radicals.
these radicals, on the other hand, cannot be determined using
the common oxidants or reductants as both the C(6)-ylC(5)OH
and the allyl type radical A (due to its mesomeric form A′) can
be reducing in nature while C(5)-ylC(6)OH and the allyl radical
Although, the presence of C(6)OH adduct is evidenced by
the formation of TMPD•+ (see section I) no products which
has a likely precursor as C(6)OH were obtained from the product
profile at this pH (pH 4.5). This could be due to the very low

2502 J. Phys. Chem. A, Vol. 106, No. 11, 2002
TABLE 3: Products Identified from the Reaction of OH with DHMP Using HPLC-ES-MS Analysis at pH 4.5
Luke et al.
•
products
molecular formula
molecular weight
m/z
a
b
c
d
e
5,6-dihydro-5,6-dihydroxy-2- methyl-(3H)-pyrimidin-4-one
5,6-dihydroxy-2-methyl-(3H)-pyrimidin-4-one
5-hydroxy-2-methyl-(3H)- pyrimidin-4-one
bis(5-hydro-5,6-dihydroxy-2-methyl-(3H)-pyrimidin-4-one-6yl)
5-hydroxy-2-methyl-(3H)-pyrimidin-4-one-6-
(5-hydro-5,6-dihydroxy-2-methyl-(3H)-pyrimidin-4-one)
bis(5-hydroxy-2-methyl-(3H)-pyrimidin-4-one-6-yl)
5-hydroxy-2-methyl-(3H)- pyrimidin-4,6-dione
5,5-dihydro-6-hydro-6-hydroxy-2-
C5N2O3H8
C5N2O3H6
C5N2O2H6
C10N4O6H14
C10N4O5H12
144
142
126
286
268
+144,-143
-141, +143
-126
-286
-267.5,+268.5
f
g
h
C10N4O4H10
C5N2O3H6
C5N2O2H8
250
142
128
+250
-141,+143
+129
methyl-(3H)-pyrimidin-4-one
i
j
6-hydroxy-2-methyl-(3H)- pyrimidin-4-one
bis(5,5-dihydro-6-hydroxy-2-methyl-(3H)-
pyrimidin-4-one-6-yl)
C5N2O2H6
C10N4O4H14
126
254
-125,+127
-254
k
bis(2-methyl-(3H)-pyrimidin-4-one-6-yl)
C10N4O2H10
218
-217
•
TABLE 4: Products Identified from the Reactionof OH with DMHP Using HPLC-ES-MS Analysis at pH 6
products
molecular formula
molecular weight
m/z
l
5,6-dihydro-5,6-dihydroxy-2,4-dimethyl pyrimidine
5,6-dihydroxy-2,4-dimethyl pyrimidine
5-hydroxy-2,4-dimethyl pyrimidine
bis(5-hydro-5,6-dihydroxy-2,4-dimethyl pyrimidine-6-yl)
5-hydroxy-2,4-dimethyl-6-(5-hydro-5,6-dihydroxy-2,4-
dimethyl pyrimidine) pyrimidine
C6N2O2H10
C6N2O2H8
C6N2OH8
C12N4O4H18
C12N4O3H16
142
140
124
282
264
-142,+142
-139, +141
-124
m
n
o
p
+283
-264
q
r
s
t
u
bis(5-hydroxy 2,4-dimethyl pyrimidine-6-yl)
5-hydro-5-hydroxy-2,4-dimethyl pyrimidin-6-one
5,5-dihydro-6-hydro-6-hydroxy-2,4-dimethyl pyrimidine
6-hydroxy-2,4-dimethyl pyrimidine
bis(5,5-dihydro-6-hydroxy-2,4-
dimethyl pyrimidine-6-yl)
C12N4O2H14
C6N2O2H8
C6N2OH10
C6N2OH8
246
140
126
124
250
-245
-139, +141
-125, +126
-123, +125
+250
C12N4O2H18
v
bis(2,4-dimethyl pyrimidine-6-yl)
C12N4H14
214
-213, +215
SCHEME 2

OH Adducts of Substituted Pyrimidines
J. Phys. Chem. A, Vol. 106, No. 11, 2002 2503
yield of C(6)OH adduct as evidenced by the low yield of
TMPD•+ (see section I). On the other hand, the product mass
peaks clearly gave evidence for the formation of products
resulting from the H adducts of both DHMP and DMHP
(Schemes 2 and 3).
dehydration reaction of a. Another important reaction that
normally occurs with the OH adducts of pyrimidine is the
dimerization reaction.12,22-27 Reaction 9 leads to a dimer of the
type d and the subsequent water elimination reaction leads to
the products e and f as shown in reaction 10 and 11. The
formation of e and f through water elimination reactions is
supported by the report by K. M. Idriss Ali27 where the OH
adduct of uracil undergoes a similar water elimination reaction.
A probable product with the structure g is also proposed,
resulting from the dehydration of a trihydroxy derivative, as
shown in reactions 12 and 13. This product must have a m/z
value of 142 and the mass peaks showed m/z values as -141
and +143. However, the product b (from reaction 7) has also
similar m/z value and hence the existence of g could not be
A part of the identified products, summarized in Table 3 in
the case of DHMP, are proposed to be mainly resulted from
the disproportionation and dimerization reactions of radicals I
and II (as shown in Scheme 2). From the quantitative analysis
of the products in the case of 1,3-dimethyluracil (1,3-DMU)
using GC-MS it is known that the major products resulting from
C(5)OH adducts are the glycol and the dimer.12 A detailed
description of the formation of glycolic products from the
reaction of OH radicals in the radiolysis as well as in the fenton
reaction of 1,3-DMU using GC-MS has been reported recently.22
The HPLC-ES-MS data in the present case shows a similar
trend. The immediate product after the disproportionation
(reaction 7) of I is the 5,6-dihydro-5,6-dihydroxy-2-methyl-(3H)-
pyrimidin-4-one (a) and 5,6-dihydroxy-2-methyl-(3H)-pyrimi-
din-4-one (b). The pyrimidine glycols are known to be acid-
labile,12 and therefore, can undergo a water elimination as shown
in reaction 8. The product c, therefore, is a result of a
confirmed. An alternative mechanism for the formation of the
product g could be the reaction of radiolytically formed H2O2
SCHEME 3

2504 J. Phys. Chem. A, Vol. 106, No. 11, 2002
Luke et al.
which acts as an oxidizing agent as shown in reaction 17 and
forms an unstable trihydroxy-substituted derivative. The im-
mediate water elimination reaction might lead to a product like
g.
and 10.4 (Figures 1S-4S), and a typical ES-MS from DMHP
(Figure 5S). This material is available free of charge via the
Internet at http://pubs.acs.org.
Other identified products include m/z values corresponding
to h-k which can be explained only through an initial attack
of H• with DHMP. It is also formed in the reaction mixture
with a G value of 0.6 × 10-7 mol J-1 under the reaction
conditions that we used. The most likely mechanism for the
formation of the product h is the disproportionation of the H
adduct radical, II. The attack of H• at the C(5)-C(6) double
bond of pyrimidines, a reaction similar to •OH, is well
understood.28 In a recent study, it is demonstrated that C(6)-
ylC(5)-H radical is the major H adduct (II) in the case of
DHMP.29 It may be further noted that the product i is the starting
compound, DHMP, but is considered as the product partner of
h from the disproportionation reaction. The dimerization of the
radical II may lead to the product j. Although a single water
elimination is possible from j, m/z value corresponds to only k
was observed in the mass peak. On the other hand, our
assignment of the product k can be rationalized on the basis of
the report on the water elimination from a dimeric hydroxy
methyl radical resulted from the OH attack on methyl uracil
where such water-eliminated products were identified using
chromatographic techniques.27 It must be further noted that
though a substantial details of OH reaction products is available
in the literature, practically little information is available on the
end products from the reaction of H• with pyrimidines. The
product profile obtained with DMHP shows a similar reaction
route as in the case of DHMP. The identified products and the
mechanism of their formation from the C(5)OH adduct and from
the H adducts are presented in Table 4 and Scheme 3,
respectively.
In conclusion, a detailed investigation of the important pH-
dependent transformation of a nonoxidizing to oxidizing OH
adduct radicals from hydroxy, methyl-, methoxy-, and amino-
substituted pyrimidines is presented. Among the selected
pyrimidines the transformation of the nonoxidizing to oxidizing
radical reaction was observed only with pyrimidines having a
keto or hydroxy group at the C(4) position. A variety of new
stable products have been identified using HPLC-ES-MS
analysis and a detailed degradation pathway is proposed. The
product analysis from DHMP and DMHP gave indications that
these products mainly arise from the dispropotionation and
dimerization of the initially formed C(5)OH adduct as well as
the H adducts. To our knowledge, the identification of the
products resulted from the H adducts in N2O-saturated aqueous
solutions, is the first report of this kind.
References and Notes
(1) Bertinchamps, A. J., Ed. Effect of Ionising Radiation on DNA;
Springer: Berlin, 1978; p 153.
(2) von Sonntag, C. Radiat. Phys. Chem. 1987, 30, 313.
(3) von Sonntag, C. Chemical Basis of Radiation Biology; Taylor and
Francis: London, 1987.
(4) Burrows, C. J.; Muller, J. G. Chem. ReV. 1998, 98, 1109.
(5) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. P. J.
Phys. Chem. Ref. Data 1988, 17 (2), 513.
(6) Simic, M. G.; Grossman, L.; Upton, A. D., Ed. In Mechanisms of
DNA Damage and Repair; Plenum Press: New York, 1986; p 39.
(7) Steenken, S. Chem. ReV. 1989, 89, 503.
(8) Hayon, E.; Simic, M. J. Am. Chem. Soc 1973, 95, 1029.
(9) Hazra, D. K.; Steenken, S. J. Am. Chem. Soc. 1983, 105, 4380.
(10) Fujita, S.; Steenken, S. J. Am. Chem. Soc. 1981, 103, 2540.
(11) Jovanivic, S. V.; Simic, M. G. J. Am. Chem. Soc. 1986, 108, 5968.
(12) Sheikhly, M. A.; von Sonntag, C. Z. Naturforsch. 1983, 38b, 1622.
(13) Schuler, R. H.; Haartzell, A. L.; Behar, B. J. Phys. Chem. 1981,
85, 192.
(14) Buxton, G. V.; Stiart. C. R., J. Chem. Soc., Faraday Trans. 1995,
91, 279-281. Fielden, E. M. In The Study of Fast Processes and Transient
Species by Electron Pulse Radiolysis; Baxendale, J. H., Busi, F., Eds.; D,
Reidel: Boston, 1982; p 49.
(15) Guha, S. N.; Moorthy, P. N.; Kishore, K.; Naik, D. B.; Rao, K. N.
Proc. Indian Acad. Sci. (Chem. Sci.) 1987, 99, 261.
(16) Panajkar, M. S.; Moorthy, P. N.; Shirke, N. D. BARC Rep. 1988,
1410.
(17) Hurst D. T. An Introduction to the Chemistry and Biochemistry of
Pyrimidines, Purines and Pteridines; J. Wiley and Sons: Chichester, 1980;
p 18.
(18) Mathews, R. W. Int. J. Appl. Radiat. Isot. 1982, 33, 1159.
(19) Novais, H. M.; Steenken, S. J. Phys. Chem. 1987, 91, 426-433.
(20) It must be noted that the deprotonation by radiolytically produced
OH- would take place only with those pyrimidines having a hydrogen at
the N(1) position. This means that among the selected pyrimidines, only
DMU and MU show this phenomenon. However, the absorbance due to
the deprotonated DMU and MU at pH 10.4 will not appear as these
compounds are already in the deprotonated form (pK’s are 9.8 and 9.5 for
DMU and MU, respectively) and therefore any ground-state absorption will
be taken care by the optical detection setup. The phenomenon of deproto-
nation and their fast (re)protonation of pyrimidines with hydrogen substituted
at N(1) is well documented in Aravindakumar, C. T.; Jacob, T. A.; Mohan,
H.; Mahal, H. S.; Mukherjee, T.; Mittal, J. P. Chem. Phys. Lett. 1998, 287,
645 and Greenstock, C. L.; Shragge, P. C.; Hunt, J. W. J. Phys. Chem.
1973, 77, 1624.
(21) The electrospray method does not provide a good estimation of
the relative yields for the various products observed, based on the intensity
of the mass signals detected. The sensitivity towards ionization in the spray
may be intrinsically very different for each analyte, and is also strongly
affected by the matrix. Therefore, no attempt is made to estimate the relative
yield of the products.
(22) Theruvathu, J. A.; Aravindakumar, C. T.; Flyunt, R.; von Sonntag,
J.; von Sonntag, C. J. Am. Chem. Soc. 2001, 123, 9007.
(23) Dizdaroglu, M.; Simic, M. G. Int. J. Radiat. Biol. 1984, 46, 241.
(24) Dizdaroglu, M.; Simic, M. G. Radiat. Res. 1984, 100, 41.
(25) Infante, G. A.; Jirathana, P.; Fendler, E. J.; Jendler, J. H. J. Chem.
Soc., Faraday Trans. 1 1974, 70 (7), 1162.
Acknowledgment. T. A. J. and V. M. M. are thankful to
the Board of Research in Nuclear Sciences (BRNS), Mumbai
and the Nuclear Science Centre, New Delhi, respectively, for a
fellowship. The financial support for this work is from the
BRNS.
(26) Infante, G. A.; Jirathana, P.; Fendler, E. J.; Jendler, J. H. J. Chem.
Soc., Faraday Trans. 1 1974, 70 (7), 1171.
(27) Idriss Ali, K. M. J. Radiat. Res. 1979, 20, 84.
(28) Deeble, D. J.; Das, S.; von Sonntag, C. J. Phys. Chem. 1985, 89,
5784.
(29) Luke, T. L.; Mohan, H.; Jacob, T.A.; Manoj, V. M.; Manoj, P.;
Mittal, J. P.; Destaillats, H.; Hoffmann, M. R.;Aravindakumar, C. T. J. Phys.
Org. Chem., in press.
Supporting Information Available: The absorption spectra
of the OH adducts of ADMP, DMHP, MU, and DMU at pH 6