Stability of N-glycosidic bond of (5′ S)-8,5′-Cyclo-2′- deoxyguanosine

Article abstract of DOI:10.1021/tx300302a

8,5′-Cyclopurine deoxynucleosides are unique tandem lesions containing an additional covalent bond between the base and the sugar. These mutagenic and genotoxic lesions are repaired only by nucleotide excision repair. The N-glycosidic (or C1′-N9) bond of 2′-deoxyguanosine (dG) derivatives is usually susceptible to acid hydrolysis, but even after cleavage of this bond of the cyclopurine lesions, the base would remain attached to the sugar. Here, the stability of the N-glycosidic bond and the products formed by formic acid hydrolysis of (5′S)-8,5′-cyclo-2′-deoxyguanosine (S-cdG) were investigated. For comparison, the stability of the N-glycosidic bond of 8,5′-cyclo-2′,5′-dideoxyguanosine (ddcdG), 8-methyl-2′-deoxyguanosine (8-Me-dG), 7,8-dihydro-8-oxo-2′- deoxyguanosine (8-Oxo-dG), and dG was also studied. In various acid conditions, S-cdG and ddcdG exhibited similar stability to hydrolysis. Likewise, 8-Me-dG and dG showed comparable stability, but the half-lives of the cyclic dG lesions were at least 5-fold higher than those of dG or 8-Me-dG. NMR studies were carried out to investigate the products formed after the cleavage of the C1′-N9 bond. 2-Deoxyribose generated α and β anomers of deoxyribopyranose and deoxyribopyranose oligomers following acid treatment. S-cdG gave α- and β-deoxyribopyranose linked guanine as the major products, but α and β anomers of deoxyribofuranose linked guanine and other products were also detected. The N-glycosidic bond of 8-Oxo-dG was found exceptionally stable in acid. Computational studies determined that both the protonation of the N7 atom and the rate constant in the bond breaking step control the overall kinetics of hydrolysis, but both varied for the molecules studied indicating a delicate balance between the two steps. Nevertheless, the computational approach successfully predicted the trend observed experimentally. For 8-Oxo-dG, the low pK_a of O⁸ and N3 prevented appreciable protonation, making the free energy for N-glycosidic bond cleavage in the subsequent step very high.

Full text of DOI:10.1021/tx300302a

Article
pubs.acs.org/crt
Stability of N‑Glycosidic Bond of (5′S)‑8,5′-Cyclo-2′-deoxyguanosine
Rajat S. Das, Milinda Samaraweera, Martha Morton,† Jose A. Gascon, and Ashis K. Basu*
́
́
Department of Chemistry, University of Connecticut, Storrs, Connecticut 06269, United States
S
* Supporting Information
ABSTRACT: 8,5′-Cyclopurine deoxynucleosides are unique tandem
lesions containing an additional covalent bond between the base and the
sugar. These mutagenic and genotoxic lesions are repaired only by
nucleotide excision repair. The N-glycosidic (or C1′-N9) bond of 2′-
deoxyguanosine (dG) derivatives is usually susceptible to acid hydrolysis,
but even after cleavage of this bond of the cyclopurine lesions, the base
would remain attached to the sugar. Here, the stability of the N-glycosidic
bond and the products formed by formic acid hydrolysis of (5′S)-8,5′-cyclo-
2′-deoxyguanosine (S-cdG) were investigated. For comparison, the stability
of the N-glycosidic bond of 8,5′-cyclo-2′,5′-dideoxyguanosine (ddcdG), 8-
methyl-2′-deoxyguanosine (8-Me-dG), 7,8-dihydro-8-oxo-2′-deoxyguano-
sine (8-Oxo-dG), and dG was also studied. In various acid conditions, S-
cdG and ddcdG exhibited similar stability to hydrolysis. Likewise, 8-Me-dG
and dG showed comparable stability, but the half-lives of the cyclic dG lesions were at least 5-fold higher than those of dG or 8-
Me-dG. NMR studies were carried out to investigate the products formed after the cleavage of the C1′-N9 bond. 2-Deoxyribose
generated α and β anomers of deoxyribopyranose and deoxyribopyranose oligomers following acid treatment. S-cdG gave α- and
β-deoxyribopyranose linked guanine as the major products, but α and β anomers of deoxyribofuranose linked guanine and other
products were also detected. The N-glycosidic bond of 8-Oxo-dG was found exceptionally stable in acid. Computational studies
determined that both the protonation of the N7 atom and the rate constant in the bond breaking step control the overall kinetics
of hydrolysis, but both varied for the molecules studied indicating a delicate balance between the two steps. Nevertheless, the
computational approach successfully predicted the trend observed experimentally. For 8-Oxo-dG, the low pKa of O8 and N3
prevented appreciable protonation, making the free energy for N-glycosidic bond cleavage in the subsequent step very high.
cells.7 It prevents the binding of TATA binding protein and
INTRODUCTION
■
strongly reduces transcription in vivo.11 Both R-cdA and S-cdA
The tandem DNA lesions, 8,5′-cyclopurine 2′-deoxynucleo-
sides (cPDNs), formed by ionizing radiation and other
processes that generate reactive oxygen species, are unique in
block the elongation of a primer in vitro by mammalian DNA
polymerase δ and T7 DNA polymerase,6 whereas mammalian
DNA polymerase η can bypass R-cdA but not S-cdA.12
that they contain damage to both the purine base and the 2′-
Recently, (5′S)-8,5′-cyclo-2′-deoxyguanosine (S-cdG) was
shown to be significantly mutagenic in Escherichia coli inducing
primarily S-cdG→A mutations.13,14 The structures of oligo-
deoxynucleotide duplexes containing a site-specific S-cdG
lesion placed opposite dC, dT, or dA in the complementary
strand were obtained by a combination of solution NMR
spectroscopy and molecular dynamics calculations.15,16 These
studies showed that the S-cdG deoxyribose is in the O4′-exo
(west) pseudorotation in each pair. The Watson−Crick base
pairing was conserved in the S-cdG·dC pair, whereas the S-
cdG·dT pair adopts a wobble pairing. In contrast, no hydrogen
bonding was observed for the S-cdG·dA pair, which differs in
conformation from the dG·dA mismatch pair.
deoxyribose sugar moiety.1 These lesions exist as 5′R and 5′S
diastereomers and have been detected in vitro and in vivo in
DNA derived from many different cells and organisms.2−4
Although these lesions were discovered in the 1960s,5 they
received more attention in recent years prompted by two
reports in 2000 showing that in mammalian cells 8,5′-cyclo-
2′deoxyadenosine (cdA) diastereomers are repaired by
nucleotide excision repair (NER) and not by base excision
repair.6,7 However, a study in 2010 provided evidence for
involvement of the DNA repair enzyme NEIL1 in NER of R-
cdA and S-cdA in mice, although the mechanism remains
unclear.8 R-cdA is more efficiently repaired by NER than S-cdA,
indicating that the stereochemistry of these lesions are
important.6 The cPDNs have been claimed to play a role in
neurologic diseases in certain Xeroderma Pigmentosum
patients with defects in NER.9 It was also shown that S-cdA
accumulates in vivo in genomic DNA of csb−/− mice,
suggesting that this lesion may accumulate in Cockayne
syndrome patients.10 S-cdA has been reported to be a strong
block of gene expression in Chinese hamster ovary and human
For many DNA lesions, structural changes of the purine or
pyrimidine residue accompany destabilization of the N-
glycosidic bond. The N7 position of guanine is the most
nucleophilic site among the DNA bases, and 2′-deoxyguanosine
Received: July 3, 2012
Published: October 1, 2012
© 2012 American Chemical Society
2451
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
2996). A gradient elution program with 100% water to 15%
acetonitrile with a flow rate of 1 mL/min in 20 min was used.
Formic Acid Hydrolysis. The acid hydrolysis was monitored
either at regular intervals after treatment with a specific acid
concentration, or the reaction time was kept constant, but the acid
concentration was varied. For the time-course experiments, ∼3 nmol
of samples were incubated in 10% (v/v) formic acid at 37 °C for
various times, and an aliquot was neutralized with ice cold ammonia
solution. To determine the effect of increasing acid concentration on
N-glycosidic bond stability, ∼3 nmol of samples were incubated in
various concentrations of formic acid for 4 h, and the reaction mixtures
were neutralized with ice-cold ammonia. After drying, a known
amount (1 nmol) of adenosine was added as internal standard to each
sample, and the total volume was made up to 100 μL by adding water.
The samples were injected on to a C18 analytical column. The amount
of hydrolyzed product formed or the remaining nucleoside was
measured by integrating the area under peak at 254 nm.
NMR Analysis. For dG and 8-Me-dG, NMR kinetic experiments
were performed on a Bruker Avance III 400 MHz NMR spectrometer
in 10% DMSO-d6, 20% DCOOD and 70% D2O. For ddcdG, the data
including 13C data were obtained through direct detection methods.
The experiments to determine the acid decomposition products were
run on a Bruker Avance I 400 MHz NMR spectrometer. Since the
solubility of S-cdG in the above solvent mixture was poor, the spectra
were recorded on a Varian Inova 600 MHz NMR spectrometer with a
1H detect cold probe, and the 13C data were acquired using indirect
detection from the 13C HSQC and HMBC. For a typical hydrolysis
experiment, approximately 30 mM solution of the substrate was
prepared in 600 μL of 20% DCOOD, 70% D2O, and 10% D6-DMSO.
This solution was immediately introduced into the spectrometer for
kinetic studies, and the probe temperature was maintained at 37 °C.
Because of the very poor solubility of S-cdG, the sample was prepared
in a similar way, and the turbid solution was centrifuged at 14000 rpm
for 2 min. The clear supernatant was used for the kinetic studies. The
ratio of the products formed from 2-deoxyribose was determined from
the integrals of the two deoxyribopyranose and the oligomer peaks
from the C1′ protons between 4.5 and 5.5 ppm and from the C2′
protons between 1.4 and 2.5 ppm. Similarly, for ddcdG and S-cdG the
integrals of the C2′ protons between 1.5 and 2.7 ppm were used.
Computational Methods. Geometry optimization of reactant and
transition states were carried out with the quantum chemistry package
Gaussian 09,25 using density functional theory (DFT) with the
B3LYP26,27 functional and basis set 6-31++G(d,p). Solvent effects
were taken into account using the polarized continuum model
(PCM)28 with water as an implicit solvent. Transition state structures
were located using the synchronous transit-guided quasi-Newton
method.29 Harmonic vibrational frequencies were calculated for the
stationary points. This frequency analysis was also used to calculate
Gibbs free energies at 310.0 K, needed to determine activation free
energy barriers and rate constants.
(dG) is particularly susceptible to acid-catalyzed depurina-
tion.17 The pKa of N7-protonated dG in DNA has been
estimated to be between 2.5 and 3.0.18 In low pH, therefore,
SN1 type hydrolysis occurs via protonation at the N7 position
of dG followed by cleavage of the C1′-N9 bond. Theoretical
studies show that for dG at neutral pH, the hydrolysis consists
of two steps, but the Gibbs free energy barrier for the first step
is very high.19 By contrast, the N7-protonated dG follows a
different two-step mechanism with a much lower Gibbs free
energy barrier for the first step,19,20 classified as DN + AN
reaction (or DN*AN where “*” indicates that a short-lived
oxacarbenium ion is formed).21 Because of the C8−C5′
covalent bond in cPDNs, it is anticipated that the purine and
the deoxyribose will remain attached even after the cleavage of
the C1′-N9 bond. The hydrolysis of the N-glycosidic bond of S-
cdA has been shown to be ∼40-fold more stable than dA in
acid,22 but the chemical stability of the N-glycosidic bond of
either R- or S-cdG is not known. Herein, we have investigated
the stability of the N-glycosidic bond of S-cdG in comparison to
dG and several analogues of S-cdG, including a cyclic analogue
(ddcdG), an acyclic analogue (8-Me-dG), and 8-Oxo-dG
(Chart 1). The products formed following hydrolysis of the
C1′-N9 bond were also investigated.
Chart 1. Structures of the Compounds Used in This Study
for N-Glycosidic Bond Cleavage: S-cdG, ddcdG, 8-Me-dG, 8-
Oxo-dG, and dG
MATERIALS AND METHODS
Protonation of the N7 atom in the nucleoside (Nu) induces the
cleavage of the N-glycosidic bond followed by nucleophilic
substitution of the formed oxacarbenium ion by water, leading to
the corresponding products (Scheme 1).
■
S-cdG and ddcdG were synthesized and characterized as reported.15,23
8-Me-dG was prepared according to Kohda et al.24 8-Oxo-dG and dG
were purchased from Berry and associates (Dexter, MI). Bulk solvents
were purchased either from Sigma Aldrich (St. Louis, MO) or Fisher
Scientific (Clifton, NJ). Deuterated solvents were purchased from
Cambridge Isotope Laboratories (Andover, MA). Water used for this
study was obtained from a Barnstead Nanopure System (Thermo
Scientific, Dubuque, IA). All reagents used were of the highest purity
available. The silica gel used for flash chromatography (40−63 μm)
and gravitational column (63−200 μm) were purchased from Sorbent
Technology (Norcross, GA). NMR data were analyzed using
TOPSPIN software and referenced against a residual solvent peak.
Mass spectral analyses were performed on a Qstar Elite mass
spectrometer (Applied Biosystems, Carlsbad, CA) using collision-
induced dissociation to detect the fragmentation pattern.
Scheme 1
The nucleosides 8-Me-dG, dG, ddcdG, and S-cdG involve a rapid
pre-equilibrium between the protonated and unprotonated states of
N7 atom before the slow rate determining step (k2), thus leading to an
overall rate constant (kobs) as follows:
k2
10−pKa
k
[H+] = 2 [H+]
Ka
kobs
=
HPLC Analysis. HPLC analyses were performed on a reversed
phase column (Synergi 4 μm Hydro-RP 80 Å LC Column 250 × 4.6
mm, Phenomenex, Torrance, CA) using a Waters 1525 HPLC system
(Waters, Milford, MA) equipped with a diode-array detector (model
(1)
where k2 is the rate constant for the cleavage of the N-glycosidic bond
(rate determining step), calculated according to the Eyring equation:
2452
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
†
kBT
h
k2 =
e−Δ G/RT
(2)
where Δ†G is the free energy of activation, kBT is the thermal energy
(T = 310.15 K), R is the gas constant, and h the Planck constant. The
pKa at the N7 position was calculated with the pKa module in the
quantum chemistry package Jaguar (Schrodinger, Inc.).30 The half-life
̈
was computed as t1/2 = 0.693/kobs. Equations 1 and 2 imply that, at
constant k2, the lower the pKa, the more stable is the bond (larger t1/2).
Similarly, at constant pKa, the lower k2, the more stable is the bond.
RESULTS
■
N-Glycosidic Bond Cleavage of S-cdG. To determine the
identity of the products formed via cleavage of the N-glycosidic
Figure 3. Kinetics of N-glycosidic bond cleavage of S-cdG in 10%
formic acid at 37 °C monitored by HPLC. The inset shows the
disappearance of the S-cdG peak (green diamonds) and appearance of
the hcdG peak (pink squares) with time.
Figure 1. Reverse-phase HPLC profile of S-cdG following incubation
with 10% formic acid at 37 °C for 7 h.
Figure 4. 1H NMR time course analysis (shown here for the first two
hours) (calibrated with the formic acid peak at 8.26 ppm) of S-cdG in
10% DMSO-d6, 20% DCOOD, and 70% D2O at 37 °C. The H-1′ peak
(moved downfield to 6.52 ppm) was integrated to measure the loss of
S-cdG. Starting from the bottom panel at zero time, spectra were taken
every 15 min.
products, the first one eluting at 10.8 min followed by a small
broad peak eluting at 11.7 min. We referred to the first 10.8
min peak as hydrolyzed cdG (hcdG). Mass spectral analysis of
hcdG showed a mass ion of 284.1 Da, representing the
protonated molecular ion (MH+) of a H2O (18 Da) addition
product of S-cdG (m/z 266), as well as the sodium adduct ion
(MNa+) at 306 Da (Figure 2). Mass fragmentation of hcdG was
analyzed by collision-induced dissociation (CID MS/MS),
which exhibited ions at m/z 266, 248, and 230 owing to three
consecutive water eliminations, plus 194 and 180 (Figure 2
inset). Since the mass of S-cdG is 266, the ion at m/z 180
originated from the base moiety plus the 5′-CHOH part of the
sugar. The ion at m/z 194 is likely to be a fragment ion
composed in addition to the base and 5′-CHOH, the 4′-CH
(plus an H). Taken together, these data suggest that hcdG is
the hydrolysis product of S-cdG. We considered it likely that
the cleavage of the C1′-N9 bond followed by the addition of a
water molecule generates hcdG, which is consistent with a
stepwise formation of the oxacarbenium ion intermediate that
undergoes nucleophilic attack by the water molecule in an SN1
mechanism (DN*AN) (Scheme 2). The relative amount of the
second broad peak at 11.7 min varied from one injection to
another and, upon reinjection, it gave >95% of the first peak,
hcdG, suggesting that it is one of the unstable intermediates.
When dG was incubated with acid under similar conditions,
it was 77% converted to guanine in 2 h (when only 18% S-cdG
Figure 2. Mass spectrometry analysis of hcdG. MS-MS of 284.1 peak
is shown in the inset.
Scheme 2. Postulated Mechanism of Acid-Catalyzed N-
Glycosidic Bond Cleavage of S-cdG
bond of S-cdG, we treated it with 10% formic acid at 37 °C and
analyzed the products by reverse-phase HPLC. Figure 1 shows
the HPLC chromatogram after 7 h of treatment, in which part
of S-cdG (the 13.7 min peak) was converted to two other
2453
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
Figure 6. HPLC monitoring of N-glycosidic bond cleavage of S-cdG
(red filled squares), ddcdG (green filled triangles), Me-dG (purple
filled circles), and dG (blue filled squares) after 4 h of treatment at 37
°C with increasing concentrations of formic acid.
of S-cdG, including the time course of generation of the hcdG
1
peak in the inset. The hydrolysis was also monitored by H
NMR (Figure 4), except that we used 20% DCOOD, 10%
DMSO-d6, and 70% D2O due to the poor solubility of S-cdG in
10% deuterated formic acid. The kinetics of hydrolysis was
followed by integrating the H-1′ peak area at 6.52 ppm since it
became smaller as more substrate was hydrolyzed. When
similar experiments were performed with dG, ddcdG, and 8-
Me-dG, it was apparent that the hydrolysis of S-cdG was similar
to that of ddcdG, whereas dG hydrolysis was comparable to 8-
Me-dG (Figure 5). As shown in Figure 5A, in which reverse-
phase HPLC monitoring was employed, and in Figure 5B,
which utilized NMR monitoring of the intensity of H-1′ proton
signal, the kinetics of hydrolysis of the four compounds
followed an analogous trend, despite the differences in acid
conditions. The half-lives (t1/2) of the four compounds show a
similar pattern (Figure 5C), also indicating that S-cdG
hydrolysis was more than 5 times slower than dG. We also
determined that 8-Oxo-dG was much more resistant to
hydrolysis than the four other nucleosides evaluated here, in
that >99% of 8-Oxo-dG was recovered after incubation in 10%
formic acid for 7 days at 37 °C (data not shown). In another set
of experiments, we varied the formic acid concentrations but
analyzed the products in each case after a 4 h time interval
(Figure 6). The concentration of formic acid needed to obtain
50% depurination of dG was 1.7% compared to 19% for N-
glycosidic bond cleavage of S-cdG, whereas the same for ddcdG
and 8-Me-dG were 18.9% and 0.9%, respectively. Again, the
acid concentration needed to cause 50% hydrolysis of S-cdG
was similar to that of ddcdG, whereas the same for 8-Me-dG
and dG was comparable. However, compared to dG or 8-Me-
dG a 10-fold acid higher concentration was needed to
hydrolyze 50% S-cdG or ddcdG. On the basis of these three
sets of experiments, we conclude that the N-glycosidic bond of
S-cdG is 5−10-times more stable to acid hydrolysis than that of
dG.
Figure 5. Comparison of the kinetics of N-glycosidic bond cleavage of
S-cdG (red filled squares), ddcdG (green filled triangles), 8-Me-dG
(purple filled circles), and dG (blue filled squares) in (A) 10% formic
acid at 37 °C monitored by HPLC and (B) 20% DCOOD, 10%
DMSO-d6, and 70% D2O at 37 °C monitored by 1H NMR of the H-1′
peak. Each point represents at least three independent determinations.
Their half-lives calculated from panel A (dark blue) and B (orange)
and normalized to that of dG are shown in panel C. The half-lives
determined by the computational study (Table 2) are also shown here
in dark green.
was hydrolyzed), suggesting that S-cdG was significantly more
resistant to acid hydrolysis. In order to further investigate the
N-glycosidic bond stability of S-cdG, we compared the kinetics
of cleavage of this cyclopurine nucleoside with 8,5′-cyclo-2′,5′-
dideoxyguanosine (ddcdG), a cyclic analogue, and 8-methyl-2′-
deoxyguanosine (8-Me-dG), an acyclic analogue of cdG (Chart
1). We selected these two compounds because, like S-cdG, each
compound contains an sp3-hybridized carbon attached to
guanine-C8, but only ddcdG contains an additional ring like S-
cdG. We also analyzed the N-glycosidic bond stability of dG
and 8-Oxo-dG, the latter containing an oxygen substitution at
guanine-C8. In one set of experiments, we monitored the N-
glycosidic bond cleavage at 37 °C in 10% formic acid by
reverse-phase HPLC. Figure 3 shows the kinetics of hydrolysis
NMR Analysis of N-Glycosidic Bond Cleavage of S-
cdG. In order to understand the products formed by acid
hydrolysis of dG and dG derivatives, we followed the reaction
of 2-deoxyribose in deuterated formic acid. After 24 h in 20%
DCOOD, 10% DMSO-d6, and 70% D2O at 37 °C, both 13C-
1
and H NMR indicated that 2-deoxyribose no longer had any
detectable signal of the deoxyribofuranose form. As shown in
1
Figure 7 (panels a and b), H NMR indicated that four major
products were generated from 2-deoxyribose, which included α-
2454
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
Figure 7. (a) 1H NMR spectrum of 2-deoxyribose following acid hydrolysis in 20% deuterated formic acid. (b) COSY spectrum of the same, which
leads to four products: α- and β-deoxyribopyranose (red and blue line, respectively), and oligomerized α- and β-deoxyribopyranose (purple and
green line, respectively).
deoxyribopyranose (35%), β-deoxyribopyranose (30%), and
two deoxyribopyranose oligomers (15% each) (Scheme 3).
There also were other unidentified minor products. Additional
details of the spectral data are provided in Table S1
(Supporting Information). Formic acid hydrolysis of dG and
8-Me-dG under similar conditions showed a nearly identical
product profile, in addition to the heterocyclic base (data not
shown). It is noteworthy that when DCOOD was replaced with
DCl, extensive polymerization and degradation of the sugar
took place, making the product profile difficult to analyze (data
not shown).
2455
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
and ddcdG are locked in the anti orientation (Figure 9). The
pKa of the protonated S-cdG was found lower than the other
three nucleosides (Table 1). The computational prediction of
the overall rate constants for the N-glycosidic bond cleavage
(Table 1) allowed us to determine the factors for acid-catalyzed
bond stability across all the dG derivatives studied. The
experimentally observed order of stability and the ratio of half-
lives of the nucleosides, using dG as a reference, showed a very
good qualitative agreement with the calculated values (Table 2
and Figure 5C). According to eq 2, both the protonation of the
N7 atom and the rate constant in the bond breaking step
control the overall kinetics. Lower pKa implies a lower number
of protonated N7 molecules, thus diminishing the probability
for the N-glycosidic bond cleaving. However, a higher
activation barrier in this last process will also disfavor cleavage
of the bond. Considering all these contributions in Table 1, it is
clear that neither pKa nor k2 remain constant for the molecules
studied. Thus, there appears to be a delicate balance between
the two steps. Let us consider, for instance, the marked
difference between the pKa at the N7 position for S-cdG (2.7)
and ddcdG (3.7). This difference alone affects the overall rate
constant by an order of magnitude. Thus, in this case, it is the
pKa that is mostly the determining factor, although a higher k2
of S-cdG relative to ddcdG partly counterbalanced its low pKa.
The origin of the more acidic proton in S-cdG is due to the
unfavorable dipole−dipole interaction between the 5′-OH and
the N7-H atom. When dG and 8-Me-dG, with almost identical
stability, are compared, there appears to be a compensation of
effects of almost equal magnitude. The slightly lower pKa in dG
tends to make the reaction slower by a factor of 2, while the
actual water-mediated cleavage is faster by roughly a factor of
2.5. Finally, in the case of 8-Oxo-dG, stability is almost entirely
determined by its inability to get protonated. Indeed,
calculation of the pKa at the O8 position gives a negative
value (not reported in Table 1). In this case, protonation can
actually occur at the alternative N3 position. However, the
previously reported pKa value at N3 is also very small.37 Thus,
significant protonation cannot occur in weak acid conditions,
making the free energy for cleavage in the next step very high.
This result is consistent with 8-Oxo-dG having the maximum
stability found experimentally relative to that of all other
nucleosides studied in this work. Thus, we infer that the high
stability of 8-Oxo-dG is largely dominated by the inability of the
base to be protonated and not by the strength of the N-
glycosidic bond.
Scheme 3. Postulated Mechanism of Formic Acid-Catalyzed
Reaction of 2-Deoxyribose
The 13C chemical shift data of ddcdG in neutral and acidic
conditions are shown in Table S2 in the Supporting
Information. The upfield shift of −5.4 ppm at C5 and the
downfield shift of 4.9 ppm at C8 on acidification with 10%
deuterated formic acid suggest protonation of N7,31,32
consistent with the general mechanism of N-glycosidic bond
cleavage of dG derivatives as postulated in Scheme 2.
1
The H NMR spectra of S-cdG in DMSO-d6 is shown in
Figure S1 in the Supporting Information. For S-cdG in acid, we
could identify the guanine linked α- and β-deoxyribopyranose
by C1′ chemical shift and H1′ coupling constants (Figure 8 and
Table S3, Supporting Information). The guanine linked
deoxyribofuranoses were identified by C1′ and C2′ chemical
shifts. No oligomers were formed, presumably because of the
steric effect of the attached guanine, but there was an
unidentified product. We hypothesize this product to be a
septanose derivative, formed via the attack of N7 or N9 of
guanine to the aldehyde carbon of the ring-opened deoxyribose
(Scheme 4). A septanose derivative was also detected in the
reaction mixture of formic acid-hydrolyzed ddcdG, which
cannot form the deoxyribopyranoses (Figure S2, Supporting
Information). But definitive characterization of these com-
pounds must await further investigation. The hydrolysis of 81%
S-cdG generated the deoxyribopyranose derivatives as the
major products (46%) with a small fraction (12%) of
deoxyribofuranose derivatives and 23% other products, which
included the suspected guanine linked septanose (Figure 8 and
Scheme 4).
It is interesting to note that HPLC analyses of S-cdG
hydrolysis gave one major peak, in addition to the
unhydrolyzed starting material, as shown in Figure 1. Since
the molecular mass of the various products shown in Scheme 4
is the same, we suspect that the 10.8 min peak in Figure 1
included all these different hcdG isomers. Our attempts to
separate these molecules by HPLC using a variety of columns
and conditions were unsuccessful, which may suggest that these
isomeric deoxypyranoses, deoxyfuranoses, and other products
remain in equilibrium.
Insights from the Computational Study. Computational
studies are often a valuable complement to experimental
observations. Accordingly, using a theoretical approach, we
explored the effects of protonation of S-cdG in comparison to
dG and the dG derivatives investigated here. Our goal was to
compute the pKa values of the protonated nucleosides as well as
the activation energies and rate constants of the process
represented in Scheme 1 so that their half-lives can be
determined. Optimized geometries for reactants and their
transition states were in agreement with structures previously
reported.19,33,34 In particular, in agreement with the literature
data,35,36 the minimum energy structure of dG and 8-Oxo-dG
was in the anti conformation, whereas that of 8-Me-dG was
measured in the syn conformation (Figure 9). Evidently, S-cdG
Another relevant question is about the intrinsic effect of the
cyclic structure of S-cdG. To answer this question, it is
appropriate to select molecules that have a hydroxyl group at
the 5′ position (i.e., S-cdG and dG). We proceeded by breaking
the C5′-C8 bond in S-cdG and capping the two ends with
hydrogen atoms. We carried out this by only permitting
rotation of the CHOH group, while maintaining the C5′-C8
distance. This allowed us to estimate the effect of the cyclic
form alone. As a result of this bond breaking, the pKa changed
from 2.7 to 3.1, which is very close to the pKa value of the fully
relaxed acyclic compound, dG (Table 1). This implies that the
C5′-C8 bond is a necessary condition to make protonated N7
more acidic in S-cdG than in dG.
DISCUSSION
■
8,5′-Cyclopurine 2′-deoxynucleosides are biologically impor-
tant DNA damages. Learning the chemical features of these
molecules is critical for analyzing their biological effects. The
2456
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
Figure 8. (a) 1H NMR spectrum of S-cdG following acid hydrolysis in 20% deuterated formic acid. (b) COSY spectrum of the same, which leads to
four products: α- and β-deoxyribopyranose (red and blue line), α- and β-deoxyribofuranose (green line), suspected septanose (fuchsia line),
unhydrolyzed S-cdG (black line).
objective of this investigation on the chemical stability of the N-
glycosidic bond of S-cdG was to provide a better understanding
of this lesion. Using different acid conditions, we established
that S-cdG and ddcdG were much more stable to acid than dG
or 8-Me-dG. However, of the five compounds examined, 8-
Oxo-dG was the most resistant to acid hydrolysis.
Role of the Six-Membered Ring and the Sugar Pucker
of S-cdG. For nonenzymatic N-glycosidic bond cleavage,
conformation changes in the sugar ring to promote favorable
2457
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
a
Scheme 4. Formic Acid-Catalyzed Ring Opening and Rearrangement of S-cdG, According to the NMR Result
a
The postulated structure of a septanose derivative is shown. These and other unidentified products have been detected at ∼23%.
recent NMR investigations of S-cdG in DNA opposite three
different partner bases revealed that in each case S-cdG
deoxyribose shifts to the O4′-exo (“West”) pseudorotation,
which is likely maintained in the S-cdG nucleoside, as was
reported for the S-cA riboside.39 The NMR studies15,16 showed
that the six-membered ring C8-N9-C1′-O4′-C4′-C5′ adopts a
half-chair conformation with the O4′ and C4′ puckered. The
O4′-exo (“West”) is considered an unstable conformer, owing
to the axial orientation of the bulkier substituents. In this
conformation, the lone pair of O4′ of the sugar seems suitably
aligned with the C1′-N9 bond, which would allow the oxygen
to stabilize the developing positive charge on C1′ during the
bond breaking step. Therefore, a faster hydrolysis is expected,
although the magnitude of this effect may be small.40 In
contrast, increased stability of the N9-C1′ bond for S-cdG was
experimentally observed (Table 2). The computational study
indicated that the ring formation in S-cdG results in an
unfavorable dipole−dipole interaction between the 5′-OH and
the N7-H atom. This interaction gives rise to increased acidity
of the proton at N7 (Table 1), thereby enhancing the acid
stability of the N9-C1′ bond. It is tempting to attribute this
stability to the formation of the new six-membered bicyclic ring
system since both S-cdG and ddcdG exhibit comparable
stability to hydrolysis; however, the mechanism is more
complex because N7-H of ddcdG is less acidic than any
nucleoside examined in this study, except that a much lower
rate constant for step 2 (k2) in ddcdG hydrolysis relative to S-
cdG (and others) counterbalanced its high pKa.
Mechanistic Considerations and the Acid Hydrolysis
Products. The 13C chemical shift data in neutral and acidic
conditions were consistent with the general mechanism of N-
glycosidic bond cleavage of dG derivatives as postulated in
Scheme 2. It has been well-established for more than 40 years
that protonation gives rise to substantial acceleration to
depurination of dG,41 and theoretical studies show that the
transition state energy is lowered by ∼10 kcal mol−1.42 In the
DN*AN mechanism,21 the N7 protonated dG leads to the
generation of an oxacarbenium ion intermediate in the first
step,41,43,44 with a Gibbs free energy barrier of 19 kcal mol−1.19
In the second step, nucleophilic attack by the water molecule
on the oxacarbenium ion to yield the final products is
Figure 9. Computed minimum energy structures of dG (anti), 8-Me-
dG (syn), ddcdG (anti), and S-cdG (anti) with a reactive water
molecule.
Table 1. Computed Values of pKa at the N7 Position,
Activation Energies, and Rate Constants for Step 2 in the
Reaction
molecule
pKa
k2
activation energy (kcal mol−1
)
dG
3.2
3.5
3.7
2.7
98.2 × 10−4
37.7 × 10−4
12.0 × 10−4
65.9 × 10−4
21.09
21.52
22.27
21.30
8-Me-dG
ddcdG
S-cdG
Table 2. Ratio of the Half-Life of a Nucleoside to That of dG
ratio
(experimentally, in
10% HCOOH
monitored by
HPLC)
ratio (experimentally, in
20% DCOOD, 10%
DMSO-d6 and 70% D2O
monitored by NMR)
ratio
molecule
dG
(theoretically)
1.00
1.15
1.00
0.70
1.00
1.01
8-Me-
dG
ddcdG
5.44
5.55
2.96
2.71
2.13
4.04
S-cdG
orbital interactions have been suggested to play a role.38
Significantly higher N-glycosidic bond stability of S-cdG and
ddcdG compared to dG and 8-Me-dG implies that the new six-
membered bicyclic ring structure provides enhanced stability,
which is consistent with an earlier investigation that reported S-
cdA to be ∼40-fold more stable to acid than dA.22 However,
the computational studies indicate that such generalizations
may not be valid and that the changes in each step of the
pathway should be examined carefully as it is altered by the
change in structure and conformation of the molecule. Two
2458
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
considered a nearly barrier-less process.19 For the cPDNs,
previous DFT calculations show a large negative enthalpy for
the N-glycosidic bond cleavage reactions of the protonated
form.45 Our calculations of the hydrolysis of this set of
nucleosides, based on DFT with the B3LYP26 functional and
basis set 6-31++G(d,p), accurately predicted the trend observed
experimentally. We also found exceptionally high stability of the
N-glycosidic bond of 8-Oxo-dG in acid, which was attributed to
its inability to protonate the base.
Present Address
†Department of Chemistry, University of Nebraska, Lincoln,
NE.
Funding
This work was supported by the National Institute of
Environmental Health Sciences, NIH (grant ES013324).
Upgrade of the 400 MHz NMR at the University of
Connecticut Chemistry Department was supported by NSF
through the NSF-CRIF program.
What Happens After the Cleavage of the C1′-N9 Bond?
For abasic sites, after the water addition the mixture of cyclic
hemiacetals is the predominant form, and the acyclic aldehyde
form was estimated to be less than 1%.46,47 However, to our
knowledge, there are no reports on the type of products formed
from ring-opened furanose sugars with a purine attached to it.
To examine the products formed after the cleavage of C1′-N9
bond, we first analyzed the products generated from 2-
deoxyribose under acid conditions used in the S-cdG
hydrolysis. From the sugar, only α and β anomers of
deoxyribopyranose and deoxyribopyranose oligomers were
detected. With formic acid hydrolyzed S-cdG, α- and β-
deoxyribopyranose linked guanine were the major products.
However, instead of the oligomeric products found in the
reaction mixture of 2-deoxyribose, S-cdG gave a fraction of the
two anomers of deoxyribofuranose linked guanine and other
cyclic products (Scheme 4). As noted by others,47 the ring-
opened aldehyde form was nearly undetectable. Oligomeric
products were not detected in ddcdG hydrolysis as well. We
believe that the steric effect of guanine at the 5′ position of the
sugar prevented oligomerization in S-cdG and ddcdG, but for S-
cdG, six-membered deoxyribopyranoses were the major
products.
Biological Implications. Base excision repair proteins can
accelerate the hydrolysis of the N-glycosidic bond in the range
107−1012-fold.40 Although a variety of mechanisms are known
to operate, a prevalent mechanism employs a water molecule to
attack the anomeric carbon (C1′) of the nucleotide. For the
cPDNs, no DNA glycosylase activity was detected in
mammalian cellular extracts,6,7 but NEIL1 was suspected to
be involved in the repair of cdA.8 In a more recent in vitro
work, it was found that seven repair proteins, including NEIL1,
which repair oxidative DNA damage, do not directly interact
with S-cdG or S-cdA.48
Notes
The authors declare no competing financial interest.
ABBREVIATIONS
■
cPDN, 8,5′-cyclopurine 2′-deoxynucleoside; S-cdG and R-cdG,
(5′S)- and (5′R)-8,5′-cyclo-2′-deoxyguanosine, respectively; S-
cdA and R-cdA, (5′S)- and (5′R)-8,5′-cyclo-2′-deoxyadenosine,
respectively; ddcdG, 8,5′-cyclo-2′,5′-dideoxyguanosine; 8-Oxo-
dG, 7,8-dihydro-8-oxo-2′-deoxyguanosine; 8-Me-dG, 8-methyl-
2′-deoxyguanosine; 2-deoxyribose, 2-deoxy-D-erythro-pentose;
NER, nucleotide excision repair; HSQC, heteronuclear single
quantum coherence; HMBC, heteronuclear multiple bond
correlation
REFERENCES
■
(1) Chatgilialoglu, C., Ferreri, C., and Terzidis, M. A. (2011) Purine
5′,8-cyclonucleoside lesions: chemistry and biology. Chem. Soc. Rev. 40,
1368−1382.
(2) Jaruga, P., and Dizdaroglu, M. (2008) 8,5′-Cyclopurine-2′-
deoxynucleosides in DNA: mechanisms of formation, measurement,
repair and biological effects. DNA Repair 7, 1413−1425.
(3) Jaruga, P., Xiao, Y., Nelson, B. C., and Dizdaroglu, M. (2009)
Measurement of (5′R)- and (5′S)-8,5′-cyclo-2′-deoxyadenosines in
DNA in vivo by liquid chromatography/isotope-dilution tandem mass
spectrometry. Biochem. Biophys. Res. Commun. 386, 656−660.
(4) Dizdaroglu, M., Jaruga, P., and Rodriguez, H. (2001)
Identification and quantification of 8,5′-cyclo-2′-deoxy-adenosine in
DNA by liquid chromatography/ mass spectrometry. Free Radical Biol.
Med. 30, 774−784.
(5) Keck, K. (1968) [Formation of cyclonucleotides during
irradiation of aqueous solutions of purine nucleotides]. Z. Naturforsch.,
B: Chem. Sci. 23, 1034−1043.
(6) Kuraoka, I., Bender, C., Romieu, A., Cadet, J., Wood, R. D., and
Lindahl, T. (2000) Removal of oxygen free-radical-induced 5′,8-purine
cyclodeoxynucleosides from DNA by the nucleotide excision-repair
pathway in human cells. Proc. Natl. Acad. Sci. U.S.A. 97, 3832−3837.
(7) Brooks, P. J., Wise, D. S., Berry, D. A., Kosmoski, J. V., Smerdon,
M. J., Somers, R. L., Mackie, H., Spoonde, A. Y., Ackerman, E. J.,
Coleman, K., Tarone, R. E., and Robbins, J. H. (2000) The oxidative
DNA lesion 8,5′-(S)-cyclo-2′-deoxyadenosine is repaired by the
nucleotide excision repair pathway and blocks gene expression in
mammalian cells. J. Biol. Chem. 275, 22355−22362.
(8) Jaruga, P., Xiao, Y., Vartanian, V., Lloyd, R. S., and Dizdaroglu, M.
(2010) Evidence for the involvement of DNA repair enzyme NEIL1 in
nucleotide excision repair of (5′R)- and (5′S)-8,5′-cyclo-2′-deoxy-
adenosines. Biochemistry 49, 1053−1055.
Increased N-glycosidic bond stability of S-cdG compared to
dG excludes the possibility of spontaneous depurination of this
lesion from DNA. Since base excision repair proteins cannot
repair this damage, NER remains the only mechanism for its
removal from DNA. It is therefore conceivable that in NER
compromised cells, this highly mutagenic lesion may persist for
a long time to exert its deleterious effects.
ASSOCIATED CONTENT
■
(9) Brooks, P. J. (2008) The 8,5′-cyclopurine-2′-deoxynucleosides:
candidate neurodegenerative DNA lesions in xeroderma pigmentosum,
and unique probes of transcription and nucleotide excision repair.
DNA Repair 7, 1168−1179.
S
* Supporting Information
Additional NMR data. This material is available free of charge
via the Internet at http://pubs.acs.org.
(10) Kirkali, G., de Souza-Pinto, N. C., Jaruga, P., Bohr, V. A., and
Dizdaroglu, M. (2009) Accumulation of (5′S)-8,5′-cyclo-2′-deoxy-
adenosine in organs of Cockayne syndrome complementation group B
gene knockout mice. DNA Repair 8, 274−278.
(11) Marietta, C., Gulam, H., and Brooks, P. J. (2002) A single 8,5′-
cyclo-2′-deoxyadenosine lesion in a TATA box prevents binding of the
AUTHOR INFORMATION
■
Corresponding Author
*Tel: 860-486-3965. Fax: 860-486-2981. E-mail: ashis.basu@
uconn.edu.
2459
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
TATA binding protein and strongly reduces transcription in vivo.
DNA Repair 1, 967−975.
(12) Kuraoka, I., Robins, P., Masutani, C., Hanaoka, F., Gasparutto,
D., Cadet, J., Wood, R. D., and Lindahl, T. (2001) Oxygen free radical
damage to DNA. Translesion synthesis by human DNA polymerase
eta and resistance to exonuclease action at cyclopurine deoxynucleo-
side residues. J. Biol. Chem. 276, 49283−49288.
(13) Jasti, V. P., Das, R. S., Hilton, B. A., Weerasooriya, S., Zou, Y.,
and Basu, A. K. (2011) 5′S)-8,5′-cyclo-2′-deoxyguanosine is a strong
block to replication, a potent pol V-dependent mutagenic lesion, and is
inefficiently repaired in Escherichia coli. Biochemistry 50, 3862−3865.
(14) Yuan, B., Wang, J., Cao, H., Sun, R., and Wang, Y. (2011) High-
throughput analysis of the mutagenic and cytotoxic properties of DNA
lesions by next-generation sequencing. Nucleic Acids Res. 39, 5945−
5954.
(15) Huang, H., Das, R. S., Basu, A. K., and Stone, M. P. (2011)
Structure of (5′S)-8,5′-cyclo-2′-deoxyguanosine in DNA. J. Am. Chem.
Soc. 133, 20357−20368.
(16) Huang, H., Das, R. S., Basu, A. K., and Stone, M. P. (2012)
Structures of (5′S)-8,5′-cyclo-2′-deoxyguanosine mismatched with dA
or dT. Chem. Res. Toxicol. 25, 478−490.
(17) Pullman, A., and Pullman, B. (1981) Molecular electrostatic
potential of the nucleic acids. Q. Rev. Biophys. 14, 289−380.
(18) Suzuki, T., Ohsumi, S., and Makino, K. (1994) Mechanistic
studies on depurination and apurinic site chain breakage in
oligodeoxyribonucleotides. Nucleic Acids Res. 22, 4997−5003.
(19) Rios-Font, R., Rodriguez-Santiago, L., Bertran, J., and Sodupe,
M. (2007) Influence of N7 protonation on the mechanism of the N-
glycosidic bond hydrolysis in 2 ′-deoxyguanosine. A theoretical study.
J. Phys. Chem. B 111, 6071−6077.
(20) Przybylski, J. L., and Wetmore, S. D. (2010) Modeling the
dissociative hydrolysis of the natural DNA nucleosides. J. Phys. Chem.
B 114, 1104−1113.
(21) Guthrie, R. D., and Jencks, W. P. (1989) IUPAC
recommendations for the representation of reaction-mechanisms.
Acc. Chem. Res. 22, 343−349.
(22) Theruvathu, J. A., Jaruga, P., Dizdaroglu, M., and Brooks, P. J.
(2007) The oxidatively induced DNA lesions 8,5′-cyclo-2′-deoxy-
adenosine and 8-hydroxy-2′-deoxyadenosine are strongly resistant to
acid-induced hydrolysis of the glycosidic bond. Mech. Ageing Dev. 128,
494−502.
(23) Romieu, A., Gasparutto, D., and Cadet, J. (1999) Synthesis and
characterization of oligonucleotides containing 5′,8-cyclopurine 2′-
deoxyribonucleosides: (5′R)-5′,8-cyclo-2′-deoxyadenosine, (5′S)-5′,8-
cyclo-2′-deoxyguanosine, and (5′R)-5′,8-cyclo-2′-deoxyguanosine.
Chem. Res. Toxicol. 12, 412−421.
(27) Lee, C. T., Yang, W. T., and Parr, R. G. (1988) Development of
the Colle-Salvetti correlation-energy formula into a functional of the
electron-density. Phys. Rev. B 37, 785−789.
(28) Cossi, M., Barone, V., Cammi, R., and Tomasi, J. (1996) Ab
initio study of solvated molecules: A new implementation of the
polarizable continuum model. Chem. Phys. Lett. 255, 327−335.
(29) Peng, C. Y., and Schlegel, H. B. (1993) Combining synchronous
transit and quasi-Newton methods to find transition-states. Israel J.
Chem 33, 449−454.
(30) Jaguar (2011) Schrodinger LLC, New York.
̈
(31) Schindler, M. (1988) Magnetic-properties in terms of localized
quantities 0.9. The DNA bases and the protonation of adenine. J. Am.
Chem. Soc. 110, 6623−6630.
(32) Seckarova, P., Marek, R., Malinakova, K., Kolehmainen, E.,
Hockova, D., Hocek, M., and Sklenar, V. (2004) Direct determination
of tautomerism in purine derivatives by low-temperature NMR
spectroscopy. Tetrahedron Lett. 45, 6259−6263.
(33) Zheng, Y., Xue, Y., and Yan, S. G. (2008) The effects of
oxidation and protonation on the N-glycosidic bond stability of 8-oxo-
2 ′-deoxyguanosine: DFT study. J. Mol. Struct.-Theochem. 860, 52−57.
(34) Greisch, J. F., Leyh, B., and De Pauw, E. (2009) Collision
induced dissociation of deuterium enriched protonated 2′-deoxygua-
nosine. Eur. Phys. J. D 51, 89−96.
(35) Lipscomb, L. A., Peek, M. E., Morningstar, M. L., Verghis, S. M.,
Miller, E. M., Rich, A., Essigmann, J. M., and Williams, L. D. (1995) X-
ray structure of a DNA decamer containing 7,8-dihydro-8-oxoguanine.
Proc. Natl. Acad. Sci. U.S.A. 92, 719−723.
(36) Xu, Y., Ikeda, R., and Sugiyama, H. (2003) 8-Methylguanosine:
a powerful Z-DNA stabilizer. J. Am. Chem. Soc. 125, 13519−13524.
(37) Jang, Y. H., Goddard, W. A., 3rd, Noyes, K. T., Sowers, L. C.,
Hwang, S., and Chung, D. S. (2002) First principles calculations of the
tautomers and pK(a) values of 8-oxoguanine: implications for
mutagenicity and repair. Chem. Res. Toxicol. 15, 1023−1035.
(38) Gorenstein, D. G. (1987) Stereoelectronic effects in
biomolecules. Chem. Rev. 87, 1047−1077.
(39) Haromy, T. P., Raleigh, J., and Sundaralingam, M. (1980)
Enzyme-bound conformations of nucleotide substrates. X-ray structure
and absolute configuration of 8,5′-cycloadenosine monohydrate.
Biochemistry 19, 1718−1722.
(40) Stivers, J. T., and Jiang, Y. L. (2003) A mechanistic perspective
on the chemistry of DNA repair glycosylases. Chem. Rev. 103, 2729−
2759.
(41) Zoltewicz, J. A., Clark, D. F., Sharpless, T. W., and Grahe, G.
(1970) Kinetics and mechanism of the acid-catalyzed hydrolysis of
some purine nucleosides. J. Am. Chem. Soc. 92, 1741−1749.
(42) Baik, M. H., Friesner, R. A., and Lippard, S. J. (2002)
Theoretical study on the stability of N-glycosyl bonds: why does N7-
platination not promote depurination? J. Am. Chem. Soc. 124, 4495−
4503.
(43) Hevesi, L., Wolfson-Davidson, E., Nagy, J. B., Nagy, O. B., and
Bruylants, A. (1972) Contribution to the mechanism of the acid-
catalyzed hydrolysis of purine nucleosides. J. Am. Chem. Soc. 94, 4715−
4720.
(44) Kampf, G., Kapinos, L. E., Griesser, R., Lippert, B., and Sigel, H.
(2002) Comparison of the acid-base properties of purine derivatives in
aqueous solution. Determination of intrinsic proton affinities of
various basic sites. J. Chem. Soc. Perkin Trans. 2, 1320−1327.
(45) Karwowski, B. T. (2010) The difference in stability between 5 ′
R and 5 ′ S diastereomers of 5 ′,8-cyclopurine-2 ′-deoxynucleosides.
DFT study in gaseous and aqueous phase. Cent. Eur. J. Chem. 8, 134−
141.
(46) Wilde, J. A., Bolton, P. H., Mazumder, A., Manoharan, M., and
Gerlt, J. A. (1989) Characterization of the equilibrating forms of the
aldehydic abasic site in duplex DNA by 17O-NMR. J. Am. Chem. Soc.
111, 1894−1896.
(47) Pierce, J., Serianni, A. S., and Barker, R. (1985) Anomerization
of furanose sugars and sugar phosphates. J. Am. Chem. Soc. 107, 2448−
2456.
(24) Kohda, K., Tsunomoto, H., Minoura, Y., Tanabe, K., and
Shibutani, S. (1996) Synthesis, miscoding specificity, and thermody-
namic stability of oligodeoxynucleotide containing 8-methyl-2′-
deoxyguanosine. Chem. Res. Toxicol. 9, 1278−1284.
(25) Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E.,
Robb, M. A., Cheeseman, J. R., Scalmani, G., Barone, V., Mennucci, B.,
Petersson, G. A., Nakatsuji, H., Caricato, M., Li, X., Hratchian, H. P.,
Izmaylov, A. F., Bloino, J., Zheng, G., Sonnenberg, J. L., Hada, M.,
Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima,
T., Honda, Y., Kitao, O., Nakai, H., Vreven, T., Montgomery, J. A., Jr.,
Peralta, J. E., Ogliaro, F., Bearpark, M., Heyd, J. J., Brothers, E., Kudin,
K. N., Staroverov, V. N., Kobayashi, R., Normand, J., Raghavachari, K.,
Rendell, A., Burant, J. C., Iyengar, S. S., Tomasi, J., Cossi, M., Rega, N.,
Millam, J. M., Klene, M., Knox, J. E., Cross, J. B., Bakken, V., Adamo,
C., Jaramillo, J., Gomperts, R., Stratmann, R. E., Yazyev, O., Austin, A.
J., Cammi, R., Pomelli, C., Ochterski, J. W., Martin, R. L., Morokuma,
K., Zakrzewski, V. G., Voth, G. A., Salvador, P., Dannenberg, J. J.,
̈
Dapprich, S., Daniels, A. D., Farkas, O., Foresman, J. B., Ortiz, J. V.,
Cioslowski, J., and Fox, D. J. (2009) Gaussian 09, Gaussian Inc.,
Wallingford, CT.
(26) Becke, A. D. (1993) Density-functional thermochemistry 0.3.
The role of exact exchange. J. Chem. Phys. 98, 5648−5652.
2460
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461

Chemical Research in Toxicology
Article
(48) Pande, P., Das, R. S., Sheppard, C., Kow, Y. W., and Basu, A. K.
(2012) Repair efficiency of (5′S)-8,5′-cyclo-2′-deoxyguanosine and
(5′S)-8,5′-cyclo-2′-deoxyadenosine depends on the complementary
base. DNA Repair, in press.
2461
dx.doi.org/10.1021/tx300302a | Chem. Res. Toxicol. 2012, 25, 2451−2461