The fate of C5′ radicals of purine nucleosides under oxidative conditions

Article abstract of DOI:10.1021/ja800763j

The factors that influence the reactivity of C5′ radicals in purine moieties under aerobic conditions are unknown not only in DNA, but also in simple nucleosides. 5′,8-Cyclopurine lesions are the result of a rapid C5′ radical attack to the purine moieties before the reaction with oxygen. These well-known lesions among the DNA modifications were suppressed by the presence of molecular oxygen in solution. Here we elucidate the chemistry of three purine-substituted C5′ radicals (i.e., 2′-deoxyadenosin- 5′-yl, 2′-deoxyinosin-5′-yl, and 2′-deoxyguanosin- 5′-yl) under oxidative conditions using γ-radiolysis coupled with product studies. 2′-Deoxyadenosin-5′-yl and 2′-deoxyinosin- 5′-yl radicals were selectively generated by the reaction of hydrated electrons (e_aq^-) with 8-bromo-2′-deoxyadenosine and 8-bromo-2′-deoxyinosine followed by a rapid radical translocation from the C8 to the C5′ position. Trapping these two C5′ radicals with Fe(CN)₆^3- gave corresponding hydrated 5′-aldehydes in good yields that were isolated and fully characterized. When an oxygen concentration in the range of 13-266 μM (typical oxygenated tissues) is used, the hydrated 5′-aldehyde is accompanied by the 5′,8-cyclopurine nucleoside. The formation of 5′,8-cyclopurines is relevant in all experiments, and the yields increased with decreasing O₂ concentration. The reaction of HO^? radicals with 2′-deoxyadenosine and 2′-deoxyguanosine under normoxic conditions was also investigated. The minor path of C5′ radicals formation was found to be ca. 10% by quantifying the hydrated 5′-aldehyde in both experiments. Rate constants for the reactions of the 2′-deoxyadenosin-5′-yl with cysteine and glutathione in water were determined by pulse radiolysis to be (2.1 ± 0.5) × 10⁷ and (4.9 ± 0.6) × 10 ⁷ M^-1 s^-1 at 22°C, respectively.

Full text of DOI:10.1021/ja800763j

Published on Web 06/04/2008
The Fate of C5′ Radicals of Purine Nucleosides under
Oxidative Conditions
Fabien Boussicault, Panagiotis Kaloudis, Clara Caminal, Quinto G. Mulazzani, and
Chryssostomos Chatgilialoglu*
ISOF, Consiglio Nazionale delle Ricerche, Via P. Gobetti 101, 40129 Bologna, Italy
Received January 30, 2008; E-mail: chrys@isof.cnr.it
Abstract: The factors that influence the reactivity of C5′ radicals in purine moieties under aerobic conditions
are unknown not only in DNA, but also in simple nucleosides. 5′,8-Cyclopurine lesions are the result of a
rapid C5′ radical attack to the purine moieties before the reaction with oxygen. These well-known lesions
among the DNA modifications were suppressed by the presence of molecular oxygen in solution. Here we
elucidate the chemistry of three purine-substituted C5′ radicals (i.e., 2′-deoxyadenosin-5′-yl, 2′-deoxyinosin-
5′-yl, and 2′-deoxyguanosin-5′-yl) under oxidative conditions using γ-radiolysis coupled with product studies.
2′-Deoxyadenosin-5′-yl and 2′-deoxyinosin-5′-yl radicals were selectively generated by the reaction of
hydrated electrons (eaq-) with 8-bromo-2′-deoxyadenosine and 8-bromo-2′-deoxyinosine followed by a rapid
radical translocation from the C8 to the C5′ position. Trapping these two C5′ radicals with Fe(CN)63- gave
corresponding hydrated 5′-aldehydes in good yields that were isolated and fully characterized. When an
oxygen concentration in the range of 13-266 µM (typical oxygenated tissues) is used, the hydrated 5′-
aldehyde is accompanied by the 5′,8-cyclopurine nucleoside. The formation of 5′,8-cyclopurines is relevant
in all experiments, and the yields increased with decreasing O2 concentration. The reaction of HO• radicals
with 2′-deoxyadenosine and 2′-deoxyguanosine under normoxic conditions was also investigated. The minor
path of C5′ radicals formation was found to be ca. 10% by quantifying the hydrated 5′-aldehyde in both
experiments. Rate constants for the reactions of the 2′-deoxyadenosin-5′-yl with cysteine and glutathione
in water were determined by pulse radiolysis to be (2.1 ( 0.5) × 107 and (4.9 ( 0.6) × 107 M-1 s-1 at 22
°C, respectively.
Introduction
sites, 5′,8-cyclopurine nucleotides, strand breaks terminated with
a variety of sugar residues, and freely diffusible degradation
Hydroxyl radicals (HO•) are known to react with DNA either
by hydrogen abstraction from the 2-deoxyribose units or by
addition to the base moieties causing deleterious chemical
modifications.1 The order of reactivity of HO• radicals toward
the various hydrogen atoms of the 2-deoxyribose moiety is under
debate.2 The proposed order that also parallels the exposure to
solvent of the 2-deoxyribose hydrogen atoms (i.e., H5′ > H4′
> H3′ ≈ H2′ ≈ H1′)3,4 was recently challenged by a few
publications.5,6 Computational data in a simple nucleotide
indicated that the most probable sites of H-atom abstraction by
radical species are the H5′ and H4′ positions.7 Under biologically
relevant conditions (i.e., aerobic), the corresponding carbon-
centered radicals resulted in the formation of oxidized abasic
products.8–13 There is growing evidence that the oxidation of
2-deoxyribose in DNA plays a critical role in the genetic
toxicology of oxidative stress and inflammation.2
The chemistry of 2-deoxyribose oxidation in DNA and model
nucleosides or nucleotides has been the subject of periodic
reviews.1,2,8–10,12,13 Several mechanistic aspects such as the 5′-
oxidation pathways are not well understood. Under aerobic
conditions, the forming C5′ radical 1 reacts with oxygen to
generate the peroxyl radical 2, which partitions between two
reaction channels. One path yields the 5′-aldehyde terminus 3,
while the other gives the 5′-(2-phosphoryl-1,4-dioxobutane)
residue 4 with the concomitant loss of a base (Scheme 1).14–16
(1) (a) von Sonntag, C. The Chemical Basis of Radiation Biology; Taylor
& Francis: Philadelphia, PA, 1987. (b) von Sonntag, C. Free-Radical-
Induced DNA Damage and Its Repair: A Chemical PerspectiVe;
Springer-Verlag: Berlin, 2006.
(8) (a) Goldberg, I. H. Acc. Chem. Res. 1991, 24, 191–198. (b) Dedon,
P. C.; Goldberg, I. H. Chem. Res. Toxicol. 1992, 5, 312–332.
(9) Pratviel, G.; Bernadou, J.; Meunier, B. Angew. Chem., Int. Ed. Engl.
1995, 34, 746–769.
(2) Dedon, P. C. Chem. Res. Toxicol. 2008, 21, 206–219.
(3) Balasubramanian, B.; Pogozelski, W. K.; Tullius, T. D. Proc. Natl.
Acad. Sci. U.S.A. 1998, 95, 9738–9743.
(10) Pogozelski, W. K.; Tullius, T. D. Chem. ReV. 1998, 98, 1089–1107.
(11) Burrows, C. J.; Muller, J. G. Chem. ReV. 1998, 98, 1109–1147.
(12) Chatgilialoglu, C.; O’Neill, P. Exp. Gerontol. 2001, 36, 1459–1471.
(13) Greenberg, M. M. Org. Biomol. Chem. 2007, 5, 18–30.
(14) (a) Kawabata, H.; Takeshita, K.; Fujiwara, T.; Sugiyama, H.; Matsura,
T.; Saito, I. Tetrahedron Lett. 1989, 30, 4263. (b) Angeloff, A.; Dubey,
I.; Pratviel, G.; Bernadou, J.; Meunier, B. Chem. Res. Toxicol. 2001,
14, 1413–1420.
(4) Sy, D.; Savoye, C.; Begusova, M.; Michalik, V.; Charlier, M.;
Spotheim-Maurizot, M. Int. J. Radiat. Biol. 1997, 72, 147–155.
(5) Xue, L.; Greenberg, M. M. Angew. Chem., Int. Ed. Engl. 2007, 46,
561–564.
(6) Roginskaya, M.; Bernhard, W. A.; Marion, R. T.; Razskazovskiy, Y.
Radiat. Res. 2005, 163, 85–89.
(15) (a) Chen, B.; Bohnert, T.; Zhou, X.; Dedon, P. C. Chem. Res. Toxicol.
2004, 17, 1406–1413. (b) Montevecchi, P. C.; Manetto, A.; Navacchia,
M. L.; Chatgilialoglu, C. Tetrahedron 2004, 60, 4303–4308.
(7) Toure, P.; Villena, F.; Melikyan, G. G. Org. Lett. 2002, 4, 3989–
3992.
9
10.1021/ja800763j CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 8377–8385 8377

A R T I C L E S
Boussicault et al.
Scheme 1. Fate of C5′ Radical under Aerobic Conditions
Scheme 2. Chemical Studies of Hydrated Electron (eaq-) with
8-Bromo-2′-deoxyadenosine (6)
ously demonstrated by using pulse radiolysis techniques that the
reaction of hydrated electrons (eaq-) with 8-bromo-2′-deoxyad-
enosine (6) produces bromide anion and the corresponding C8
radical, which abstracts intramolecularly a hydrogen atom from
the C5′ position, affording selectively the 2′-deoxyadenosin-5′-yl
radical 7 (Scheme 2).24 Radical 7 undergoes cyclization with a
rate constant of kc ) 1.6 × 105 s-1 to give the heteroaromatic
radical 8. Furthermore, the reactivity of C5′ radical 7toward
K3Fe(CN)6 and O2 was examined in competition with the
cyclization process.24 Similar reactions with almost identical
rate constants have been observed with 8-bromo-2′-deoxyinosine
(10), which affords the 2′-deoxyinosin-5′-yl radical.25 In this
article, we extended the latter approach for site-specific genera-
tion of C5′ radical to include the reaction of 7 with biologically
relevant thiols and, more importantly, to study the fate of purine-
substituted C5′ radical under aerobic conditions or in the
presence of oxidants. Furthermore, we reinvestigated the reaction
of radiation-generated HO• radicals with 2′-deoxyadenosine and
2′-deoxyguanosine under aerobic conditions.
The 5′,8-cyclo-2′-deoxyadenosine and 5′,8-cyclo-2′-deox-
yguanosine moieties, also denominated as 5′,8-cyclopurines, are
the result of a rapid C5′ radical attack to the purine moiety to
give 5 before the reaction with oxygen (Scheme 1). These
lesions are observed among the DNA modifications.17,18 They
have also been identified in mammalian cellular DNA in vivo,
where their levels are enhanced by conditions of oxidative
stress.19–21 It was also reported that the formation of 5′,8-
cyclopurines was suppressed by the presence of molecular
oxygen in solution.22
Interestingly, these lesions do give neither strand breaks nor
abasic sites, and therefore they have not been considered in the
overall reactivity of HO• radicals toward the various hydrogen
atoms of the 2-deoxyribose moieties. The factors that influence
the reactivity of C5′ radicals in purine moieties under aerobic
conditions are unknown not only in DNA and related systems,
but also in simple nucleosides.
Results and Discussion
Radiolytic Production26,27 of Transients. Radiolysis of neutral
water leads to the species eaq-, HO•, and H• as shown in eq 1,
where the values in parentheses represent the chemical radiation
There are only a few methods available for site-specific
generation of C5′ radicals. We have recently demonstrated that
photolysis of 5′-tert-butyl ketone derivatives of thymine and
2′-deoxyguanosine leads to formation of thymidin-5′-yl and 2′-
deoxyguanosin-5′-yl radicals, respectively.23 We have also previ-
-
yields (G) expressed in µmol J-1. The reactions of eaq with
substrates were studied by irradiating solutions containing 0.25
M t-BuOH. Hydroxyl radicals were scavenged efficiently by
t-BuOH, (eq 2, k2 ) 6.0 × 108 M-1 s-1), whereas hydrogen
atoms reacted only slowly (eq 2, k2 ) 1.7 × 105 M-1 s-1). The
presence of N2O transforms efficiently eaq- into the O•- species
(eq 3, k3 ) 9.1 × 109 M-1 s-1). The HO• radical [pKa(HO•) )
11.9] is in equilibrium with its conjugated base O•- (eq 4/-4,
k4 ) 1.2 × 1010 M-1 s-1, k-4 ) 1 × 108 s-1).
(16) (a) Recent results on DNA adducts derived from electrophilic products
of 5′-oxidation of deoxyribose in DNA and on oligonucleotides
containing 5′-aldehyde at one terminus have been also reported. See:
Chen, B.; Vu, C. C.; Byrns, M. C.; Dedon, P. C.; Peterson, L. A.
Chem. Res. Toxicol. 2006, 19, 982–985. (b) Kodama, T.; Greenberg,
M. M. J. Org. Chem. 2005, 70, 8122–8129.
(17) Dizdaroglu, M.; Jaruga, P.; Rodriguez, H. Free Radical Biol. Med.
2001, 30, 774–784.
H2O ' eaq- (0.27), HO• (0.28), H• (0.062)
(1)
(18) Jaruga, P.; Birincioglu, M.; Rodriguez, H.; Dizdaroglu, M. Biochem-
istry 2002, 41, 3703–3711.
(19) Randerath, K.; Zhou, G. D.; Somers, R. L.; Robbins, J. H.; Brooks,
P. J. J. Biol. Chem. 2001, 276, 36051–36057.
(24) (a) Chatgilialoglu, C.; Guerra, M.; Mulazzani, Q. G. J. Am. Chem.
Soc. 2003, 125, 3839–3848. (b) Flyunt, R.; Bazzanini, R.; Chatgilia-
loglu, C.; Mulazzani, Q. G. J. Am. Chem. Soc. 2000, 122, 4225–4226.
(25) Russo, M.; Jimenez, L. B.; Mulazzani, Q. G.; D’Angelantonio, M.;
Guerra, M.; Miranda, M. A.; Chatgilialoglu, C. Chem.-Eur. J. 2006,
12, 7684–7693.
(20) D’Errico, M.; Parlanti, E.; Teson, M.; Bernardes de Jesus, B. M.;
Degan, P.; Calcagnile, A.; Jaruga, P.; Bjørås, M.; Crescenzi, M.;
Pedrini, A. M.; Egly, J.-M.; Zambruno, G.; Stefanini, M.; Dizdaroglu,
M.; Dogliotti, E. EMBO J. 2006, 25, 4305–4315.
(21) These lesions represent a unique class of oxidative DNA lesions in
that they are repaired by the nucleotide excision repair pathway but
not by base excision repair or direct repair. For a review, see: Brooks,
P. J. Neuroscience 2007, 145, 1407–1417.
(26) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J. Phys.
Chem. Ref. Data 1988, 17, 513–886.
(27) Ross, A. B.; Mallard, W. G.; Helman, W. P.; Buxton, G. V.; Huie,
R. E.; Neta, P. NDRL-NIST Solution Kinetics Database, ver. 3;
National Institute of Standards and Technology: Gaithersburg, MD,
1998.
(22) Dizdaroglu, M. Biochem. J. 1986, 238, 247–254.
(23) Manetto, A.; Georganakis, D.; Gimisis, T.; Leondiadis, L.; Mayer,
P.; Carell, T.; Chatgilialoglu, C. J. Org. Chem. 2007, 72, 3659–3666.
9
8378 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008

Fate of C5′ Radicals of Purine Nucleosides
A R T I C L E S
•
HO•/H• + t-BuOH f (CH3)2C(OH)CH2 + H2O/H2 (2)
eaq- + N2O f O•- + N2
(3)
(4)
HO• + HO- a O•- + H2O
Reactivity of 2′-Deoxyadenosin-5′-yl Radical toward Thiols. The
reactivity of radical 7 toward cysteine (CySH) and glutathione
(GSH) was investigated by pulsing Ar-purged solutions of 1
mM 6 containing 0.25 M t-BuOH and different concentrations
of thiol (0-5 mM) at natural pH.28 The reactions were
monitored at 360 nm, where the radical 8 has a λmax with ꢀ )
9600 M-1 cm-1 (Scheme 2).24 The absorption at 360 nm was
found to decrease by increasing the concentration of thiol (see
Figure 1 for CySH and Supporting Information for GSH). This
decrease was almost identical to the one expected from the
-
partition of eaq between 6 (k ) 1.6 × 1010 M-1 s-1 24
)
and
CySH (k ) 8.5 × 109 M-1 s-1).27 On the other hand, the first-
order growth (kobs) increased linearly with increasing CySH
concentration; that is, kobs ) kc + kx[CySH] (insets a and b).
From the slopes of kobs versus [RSH], the bimolecular rate
constants were found to be (2.1 ( 0.5) × 107 and (4.9 ( 0.6)
× 107 M-1 s-1 for CySH and GSH, respectively.
Figure 1. Dependence of ∆Αbsorbance at 360 nm on [CySH] obtained
from the pulse radiolysis of Ar-purged, unbuffered solutions containing 1
mM 6 and 0.25 M t-BuOH; optical path ) 2.0 cm, dose ) 20 Gy. The
solid line represents an empirical fit to the data. Insets: (a) Time dependence
of ∆Αbsorbance at 360 nm in the absence (upper curve) and in the presence
(lower curve) of 5 mM CySH. (b) Dependence of kobs for buildup at 360
nm on [CySH].
Table 1 summarizes the results obtained for the trapping of
C5′ radical 7, together with the analogous reactions of the C1′
radical (2′-deoxyuridin-1′-yl).29 It is worth emphasizing that the
rate constants for the reaction of C5′ radical with thiols are 1
order of magnitude higher than that of the anologous reaction
of C1′ radical (presumably both enthalpic and entropic factors
play a role) and that GSH is two times faster than CySH for
both radicals. The oxygen trapping is quite fast and similar for
the two species, whereas the oxidation by Fe3+ is ca. 40 times
faster for C5′ radical.
Table 1. Rate Constants for Reaction of 2′-Deoxyadenosin-5′-yl
and 2′-Deoxyuridin-1′-yla
rate constant (M-1 s-1
)
trapping agent
2′-deoxyadenosin-5′-yl
2′-deoxyuridin-1′-ylb
CySH
GSH
O2
(2.1 ( 0.5) × 107
(4.9 ( 0.6) × 107
(1.8 ( 1) × 109c
(4.2 ( 0.4) × 109c
(2.9 ( 0.4) × 106
(4.4 ( 0.3) × 106
2 × 109
Fe3+d
ca. 1 × 108
a Reactions in neutral water at 22 ( 2 °C. b Data from ref.29 c Data
Oxidation of 2′-Deoxyadenosin-5′-yl and 2′-Deoxyinosin-
5′-yl Radicals. C5′ radicals such as 2′-deoxyadenosin-5′-yl (7)
and 2′-deoxyinosin-5′-yl were found to react with K3Fe(CN)6
with rate constants higher than 109 M-1 s-1 (cf. Table 1),
presumably affording the corresponding cation because the
electron transfer is thermodynamically quite favorable; that is,
E°[Fe(CN)63-/Fe(CN)64-] ) 0.36 V30a,b and E°[CH3CHO, H+/
CH3CH(•)OH] ) -1.25 V (cf. Scheme 2).30c We investigated
these reactions by detailed product studies. On the basis of the
above-mentioned kinetic data, it is expected that in the presence
of a few millimolar concentration of K3Fe(CN)6 the reaction
follows this path. Since eaq- reacts with K3Fe(CN)6 (k ) 3.1 ×
enosine or 8-bromo-2′-deoxyinosine (k ) 1.6 × 1010 M-1
s-1),24,25 the amount of K3Fe(CN)6 should be added preferably
in portions to keep a steady-state concentration e1 mM.31
Ar-purged aqueous phosphate-buffered solutions (pH 7.1) of
1.52 mM 6 (or 1.04 mM 10) containing 0.25 M t-BuOH and 1
mM K3Fe(CN)6 were initially γ-irradiated with a dose of 1.5
3-
from ref.24 d Fe(CN)6
with 2′-deoxyadenosin-5′-yl and FeCl3 with
2′-deoxyuridin-1′-yl.
kGy. Subsequently, amounts of K3Fe(CN)6 (1 mM) were added
at 1.5 kGy dose intervals to keep the concentration of oxidant
in the desired range. Both sets of reactions were monitored by
HPLC after passage of the reaction mixtures through an ion-
exchange resin to eliminate the iron salts. The HPLC analyses
are shown in Figure 2A,B for the two 8-bromopurine nucleo-
sides. In both sets of experiments, there is the consumption of
starting bromide (blue peaks) and the formation of a main
product (red peaks). After workup, the hydrated 5′-aldehydes 9
and 11 were isolated as compounds of quite good stablility32
and fully characterized (Scheme 3).
The insets of Figure 2A,B show the change in concentration
of the bromide (blue) and the hydrated 5′-aldehyde (red). On
the basis of the consumption of starting material, the yields of
hydrated 5′-aldehydes 9 and 10 are 85 and 87%, respectively.
Two minor products eluted before the red peaks in all chro-
matograms and were assigned to the corresponding free base
and (5′R)-5′,8-cyclopurine nucleoside (in order of appearance)
in comparison with the authentic samples (vide infra). Their
yields are estimated to be in the range of 6-8% each.33
Aldehydic protons usually exhibit a peak at 9-10 ppm at 1H
NMR, while the corresponding carbon atoms usually absorb
downfield from 160 ppm at 13C NMR. On the contrary,
compounds 9 and 11 exhibited a peak at ∼5.2 ppm at 1H NMR
109 M-1 s-1 27
only 5 times slower than 8-bromo-2′-deoxyad-
)
(28) The pH of the solutions decreases from 7.5 to 5.5 when the amount
of CySH (0-5 mM) is increased.
(29) Chatgilialoglu, C.; Ferreri, C.; Bazzanini, R.; Guerra, M.; Choi, S.-
Y.; Emanuel, C. J.; Horner, J. H.; Newcomb, M. J. Am. Chem. Soc.
2000, 122, 9525–9533.
(30) (a) All redox potentials are vs NHE. (b) Hausler, K. E.; Lorenz, W. J.
In Standard Potentials in Aqueous Solution; Bard, A. J., Parsons, R.,
Jordan, J. Eds.; Marcel Dekker: New York, 1985; p 408. (c) Wardman,
P. J. Phys. Chem. Ref. Data 1989, 18, 1637–1755.
(31) K3Fe(CN)6 can also trap H• atoms (k ) 6.3 × 109 M-1 s-1 27
)
and
Me2C(OH)CH2• radicals (k ≈ 3 × 106 M-1 s-1). See: Candeias, L. P.;
Wolf, P.; O’Neill, P.; Steenken, S. J. Phys. Chem. 1992, 96, 10302–
10307.
(32) An aqueous solution of 9 or 11 (1 mM) left at room temperature for
24 h and additionally heated at 40 °C for 12 h did not result in any
decomposition.
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8379

A R T I C L E S
Boussicault et al.
Scheme 4. Proposed Mechanisms for the Fate of C5′ Radical and
Formation of Hydrated 5′-Aldehyde under Oxidative Conditions
(the pKa of HOO• Is 4.8)
ventional radical-polar crossover reaction,35 for the conversion
of 8-bromo derivatives 6 and 10 to hydrated 5′-aldehyde 9 and
11 in good yields. Scheme 4 shows the mechanism that involves
oxidation of C5′ radical 12 by Fe3+ to the carbocation 14 and
the subsequent addition of water to form the hydrated 5′-
aldehyde 16.
Figure 2. (A) HPLC runs of γ-irradiation of 6 (1.52 mM) in Ar-purged
phosphate-buffered aqueous solutions (pH ) 7.1) containing t-BuOH (0.25
M) and 1 mM K3Fe(CN)6 at a dose rate of 8.2 Gy min-1. Extra amounts
of K3Fe(CN)6 (1 mM) were added at 1.5 kGy dose intervals. The HPLC
peaks of 6 are highlighted in blue, while the peaks of the product 9 are
highlighted in red. The inset shows the concentration of 6 (blue) and 9
(red) as a function of the irradiation dose. (B) HPLC runs of γ-irradiation
of 10 (1.04 mM) in Ar-purged phosphate-buffered aqueous solution (pH )
7.1) containing t-BuOH (0.25 M) and 1 mM K3Fe(CN)6 at a dose rate of
8.7 Gy min-1. Extra amounts of K3Fe(CN)6 (1 mM) were added at 1.5
kGy dose intervals. The HPLC peaks of 10 are highlighted in blue, while
the peaks of the product 11 are highlighted in red. The inset shows the
concentration of 10 (blue) and 11 (red) as a function of the irradiation dose.
Effect of Oxygen Concentration on the Fate of 2′-Deoxy-
adenosin-5′-yl and 2′-Deoxyinosin-5′-yl Radicals. Our method
of selective generation of C5′ radicals in purine nucleosides is
also suitable for investigating these reactions under aerobic
conditions. Air-saturated aqueous phosphate-buffered solution
(pH 7.1) of 1.58 mM 6 containing 0.25 M t-BuOH was
γ-irradiated at a dose rate of 8.2 Gy min-1. During γ-radiolysis,
continuous bubbling of air through the sample ensured the
constant concentration of O2. The reaction was monitored by
HPLC at 1.5 kGy dose intervals up to 6 kGy as shown in Figure
3A. An analogous experiment was performed with 0.95 mM
10, where the reaction was monitored at 1 kGy dose intervals
up to 7 kGy. Figure 3B shows the HPLC runs at 0, 2, 5, and 7
kGy. The consumption of the starting material (blue peaks) was
accompanied by the formation of a main product (red peaks),
which is assigned to the corresponding hydrated aldehyde (9
or 11) in comparison with the authentic sample. The insets of
Figure 3A,B show the change in concentration of the starting
bromide and the main product. On the basis of the consumption
of starting material, the yields of hydrated aldehydes 9 and 11
are 42 and 76%, respectively.
Scheme 3. One-Pot Synthesis of Hydrated 5′-Aldehydes 9 and 11
from the Reaction of Corresponding 8-Bromopurine Nucleoside
3-
with Hydrated Electrons in the Presence of Fe(CN)6
The mechanism we envisaged for the formation of hydrated
5′-aldehydes under these conditions is also outlined in Scheme
4. Reaction of C5′ radical 12 with oxygen gives the peroxyl
radical 13. The rate constant for the oxygen trapping of
(33) Analogous experiments of bromide 6 replacing K3Fe(CN)6 with FeCl3,
CuCl, or CuCl2 were also carried out. In all cases, the yields of 9
were much poorer because of (i) the lower conversion of starting
material, the rate constants of eaq- with FeCl3, CuCl, or CuCl2 being
10-20 times faster than that of K3Fe(CN)6,27 and (ii) the formation
of similar amounts of 5′,8-cyclo-2′-deoxyadenosine, indicating a slower
quenching of C5′ radical by the oxidants FeCl3, CuCl, or CuCl2.
(34) (a) Liu, S.; Wnuk, S. F.; Yuan, C.; Robins, M. J.; Borchardt, R. T.
J. Med. Chem. 1993, 36, 883–887. (b) Wnuk, S. F.; Yuan, C.-S.;
Borchardt, R. T.; Balzarini, J.; De Clercq, E.; Robins, M. J. J. Med.
Chem. 1997, 40, 1608–1618.
and ∼90 ppm at 13C NMR, a clear indication that only the
hydrated form is present in aqueous solutions. In the ribo series,
adenosine 5′-aldehyde has been previously synthesized and
shown also by NMR to exist in the hydrated form in aqueous
solution.34 Therefore, the radiolytic conditions allowed to
develop a synthetically useful one-pot procedure by an uncon-
(35) Murphy, J. A. In Radicals in Organic Synthesis; Renaud, P., Sibi,
M. P. Eds.; Wiley-VCH: Weinheim, Germany, 2001; pp 298-315.
9
8380 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008

Fate of C5′ Radicals of Purine Nucleosides
A R T I C L E S
Scheme 5. Fate of C1′ Peroxyl Radicals40
constant of 1.4 × 104 s-1, followed by hydrolysis with formation
of the ribonolactone 19 (Scheme 5).40
In the above-mentioned experiments, a consistent percentage
-
•- 41
•-
of eaq and H• produces O2
,
whereas additional O2 is
produced in the reaction of C5′ radicals with oxygen. However,
•-
no evidence of further reaction between O2 with the starting
8-bromopurine nucleosides is found. To investigate deeply this
issue, O2-saturated phosphate-buffered solutions of lower con-
centrations of 8-bromopurine derivatives (0.14 mM) containing
0.25 M t-BuOH were γ-irradiated as usual. Under these
conditions, where O2 quenched nearly all the primary reducing
species forming O2•-, the consumption of the starting nucleoside
was not observed.37b
Two minor products eluted before the red peaks in all
chromatograms of Figure 3A,B and were assigned (in order of
appearance) to the corresponding free base and (5′R)-5′,8-
cyclopurine nucleoside in comparison with the authentic
samples.24,25 To understand better the relation of these two
minor products with the hydrated 5′-aldehyde, we performed
analogous experiments by varying the oxygen concentration.
Figure 3. (A) HPLC runs of γ-irradiation of 6 (1.58 mM) in air-saturated
phosphate-buffered aqueous solution (pH ) 7.1) containing t-BuOH (0.25
M) at a dose rate of 8.2 Gy min-1. The HPLC peaks of 6 are highlighted
in blue, while the peaks of the product 9 are highlighted in red. The inset
shows the concentration of 6 (blue) and 9 (red) as a function of the
irradiation dose. (B) HPLC runs of γ-irradiation of 10 (0.95 mM) in air-
saturated phosphate-buffered aqueous solution (pH ) 7.1) containing
t-BuOH (0.25 M) at a dose rate of 8.1 Gy min-1. The HPLC peaks of 10
are highlighted in blue, while the peaks of the product 11 are highlighted
in red. The inset shows the concentration of 10 (blue) and 11 (red) as a
function of the irradiation dose.
The solubility of molecular oxygen in H2O is 1.33 × 10-3
M
at 22 °C and 1 atm partial pressure.42 The air-saturated solution
corresponds to 2.66 × 10-4 M of O2, which is ca. 6 times higher
than typical well-oxygenated tissues; that is, [O2] = 4 × 10-5
M (the oxygen concentration is even lower in the nucleus).43
We performed additional experiments using 5 or 1% O2, which
corresponds to 6.65 × 10-5 and 1.33 × 10-5 M, respectively,
where the cyclization should be favored. Indeed, by decreasing
the O2 concentration the yields of hydrated 5′-aldehyde and
(5′R)-5′,8-cyclopurine nucleoside decreased and increased,
respectively. The yields of free base also increased.
2′-deoxyadenosin-5′-yl or 2′-deoxyinosin-5′-yl is 2 × 109 M-1
To quantify these changes, analysis of the data was performed
in terms of radiation chemical yield. The disappearance of the
starting material (mol kg-1) divided by the absorbed dose (1Gy
) 1 J kg-1) gives the radiation chemical yield (G). Figure 4
shows examples of the G of disappearance of starting bromide
6 and the G of formation of products vs dose for 2.66 × 10-4
and 1.33 × 10-5 M of oxygen, respectively.37b
s-1 24,25
The peroxyl radical 13 could decay either via a cyclic
.
transition state leading to HOO• radical and aldehyde 15 or via
heterolytic cleavage to generate carbocation 14 and superoxide
radical anion. The two pathways are formally identical.36,37
Although the analogous R-hydroxyethylperoxyl radical is
reported to eliminate HOO• with formation of acetaldehyde (k
) 50 s-1 at 20 °C),39 the heterolytic cleavage has an important
precedent in nucleosides. Indeed, the C1′ peroxyl radical 17
has been demonstrated to split to 18 and superoxide with a rate
In Tables 2 and 3, all the data are collected for the two series,
showing that the sum of the yields of the three products, on the
basis of the converted starting bromide, are ca. 70% for 6 and
ca. 90% for 10. Taking into account that G(eaq-) + G(H•) )
0.33 µmol J-1 (eq 1), the G(-6) ) 0.27 and G(-10) ) 0.23
µmol J-1 in air-saturated experiments correspond to 82 and 70%
(36) The pKa of HOO• is 4.8. See: Bielski, B. H. T.; Cabelli, D. E.; Arudi,
R. L. J. Phys. Chem. Ref. Data 1985, 14, 1041–1100.
-
of the reaction of eaq and H• with each derivative. These
(37) (a) Hydrogen peroxide was also found to be a major reaction product.
Air-saturated aqueous phosphate-buffered solution (pH 7.5) of 2.2 mM
6 containing 0.25 M t-BuOH was γ-irradiated at a dose rate of 7.7
Gy min-1. During γ-radiolysis, continuous bubbling of air through
the sample ensured the constant concentration of O2. Hydrogen
percentages reflect fairly well the competition existing between
the reduction of starting bromide (k ) 1.6 × 1010 M-1 s-1 24,25
)
and molecular oxygen (k ) 1.9 × 1010 M-1 s-1 26,27
)
by solvated
peroxide was determined iodometrically at 0, 1, and 2 kGy.38
A
G(H2O2) ≈ 0.18 µmol J-1 was found, considerably higher than the
G(H2O2) ) 0.07 produced from the water radiolysis. (b) See
Supporting Information for more details.
(40) (a) Chatgilialoglu, C.; Gimisis, T. Chem. Commun. 1998, 1249–1250.
(b) Emanuel, C. J.; Newcomb, M.; Ferreri, C.; Chatgilialoglu, C. J. Am.
Chem. Soc. 1999, 121, 2927–2928.
(38) Allen, A. O.; Hochanadel, C. J.; Ghormley, J. A.; Davis, T. W. J.
Phys. Chem. 1952, 56, 575–586.
(41) The rate constants for the reaction of eaq- and H• with O2 to produce
•-
O2 are 1.9 × 1010 and 1.2 × 1010 M-1 s-1, respectively.27
(39) (a) Bothe, E.; Behrens, G.; Schulte-Frohlinde, D. Z. Naturforsch., B:
Chem. Sci. 1977, 32, 886–889. (b) Bothe, E.; Schuchmann, M. N.;
Schulte-Frohlinde, D.; von Sonntag, C. Z. Naturforsch., B: Chem. Sci.
1983, 38, 212–219.
(42) Battino, R.; Rettich, T. R.; Tominaga, T. J. Phys. Chem. Ref. Data
1983, 12, 163–178.
(43) (a) Zander, R. Z. Naturforsch., C: Biosci. 1976, 31, 339–352. (b)
Zander, R. AdV. Exp. Med. Biol. 1976, 75, 463–467.
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8381

A R T I C L E S
Boussicault et al.
Scheme 6. Fate of 2′-Deoxyadenosin-5′-yl radical and the
Formation of (5′R)-5′,8-Cyclo-2′-deoxyadenosine (20) and
(5′S)-5′,8-Cyclo-2′-deoxyadenosine (21) under Aerobic Conditions
conditions and the presence of protecting groups.44,45 A (5′R)/
(5′S) ) 6:1 ratio was found in γ-irradiated aqueous solutions.24
In the present study, a (5′R)/(5′S) ≈ 6:1 ratio is also found in
all experiments.37b The cyclization of C5′ radical 7 occurs
through its pro-(5′R) and pro-(5′S) to give 8a and 8b radicals,
respectively, which are in the chair conformation (Scheme 6).
It was suggested that the 8a/8b ratio depends on the energies
of the two chair transition states, where pro-(5′R) and pro-(5′S)
conformations have the 5′OH in axial and equatorial arrange-
ment, respectively, and only the pro-(5′R) conformer can be
stabilized by hydrogen bonding, involving either the oxygen of
the sugar ring or the N3 of the base.24,45 Therefore, such
hydrogen bonding is most likely the origin of the observed
diastereoselective ratio. Oxidative rearomatization of 8a and 8b
completes the reaction mechanism, affording the diastereomers
20 and 21 (Scheme 6).46 Similar results were also reported for
the (5′R)- and (5′S)-isomers of 5′,8-cyclo-2′-deoxyinosine,25 and
in the present study the observed diastereomeric ratio of ca.
5:1 is in favor of (5′R)-isomer 22.37b Our findings clearly
indicate that the formation of 5′,8-cyclopurines in the range of
13-266 µM of oxygen is relevant and, in agreement with
Scheme 6, the yield increases with decreasing O2 concentration.
Figure 4. Irradiation of 8-bromo-2′-deoxyadenosine 6 in phosphate-
buffered aqueous solutions (pH ) 7.1) containing t-BuOH (0.25 M) under
aerobic conditions. Upper: 1.58 mM of 6 and [O2] ) 2.66 × 10-4 M. Lower:
1.77 mM of 6 and [O2] ) 1.33 × 10-5 M. The chemical irradiation yields
G(-6) (9), G(9) (b), G(adenine) (1) and G(20) (2), as a function of the
irradiation dose. The line extrapolation to a zero dose leads to the G values
given in Table 2.
Table 2. Radiation Chemical Yield (G) Extrapolated to Zero Dose
from the Irradiation of 8-Bromo-2′-deoxyadenosine 6 at Various
Oxygen Concentrationsa
6, mM
O2, M
G(-6)
G(9)
G(adenine)
G(20)
1.77
2.22
1.58
1.33 × 10-5
6.65 × 10-5
2.66 × 10-4
0.31
0.31
0.27
0.10
0.12
0.15
0.04
0.04
0.02
0.07
0.05
0.01
a Units of G in µmol J-1
.
Table 3. Radiation Chemical Yield (G) Extrapolated to Zero Dose
from the Irradiation of 8-Bromo-2′-deoxyadenosine 10 at Various
Oxygen Concentrationsa
10, mM
O2, M
G(-10)
G(11)
G(Hy)b
G(22)
1.17
0.94
0.95
1.33 × 10-5
6.65 × 10-5
2.66 × 10-4
0.29
0.28
0.23
0.10
0.12
0.16
0.07
0.05
0.03
0.09
0.07
0.04
a Units of G in µmol J-1 b Hy ) Hypoxanthine.
.
electrons, taking into consideration the initial concentration of
nucleoside (1.58 and 0.95 mM for 6 and 10, respectively) and
O2 (2.66 × 10-4 M). By decreasing the [O2], the G of
disappearance increased accordingly.
The 5′,8-cyclo-2′-deoxyadenosine is formed in two diaster-
eomeric forms depending on the configuration at C5′ position;
that is, (5′R)- and (5′S)-isomers (20 and 21, respectively).24,44
The (5′R)/(5′S) ratio varies substantially with experimental
The formation of free base (adenine or hypoxanthine)
accounts also substantially on the reaction products. Interest-
ingly, the yield decreased by increasing O2 concentration. The
base release mechanism probably involves several minor paths
of the sugar radicals C1′, C3′, and C4′.1 For example, the C1′
radical path would follow the mechanism reported in Scheme
5. We suggest that the primary alkyl radicals derived from the
reaction of HO•/H• with t-BuOH (eq 2) are able to abstract
hydrogen unselectively from the sugar moiety. By increasing
the oxygen concentration, the primary alkyl radical is trans-
(44) Jimenez, L. B.; Encinas, S.; Miranda, M. A.; Navacchia, M. L.;
Chatgilialoglu, C. Photochem. Photobiol. Sci. 2004, 3, 1042–1046.
9
8382 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008

Fate of C5′ Radicals of Purine Nucleosides
A R T I C L E S
formed to peroxyl radical whose reactivity toward hydrogen
abstraction is much more bland.47
Reaction of Hydroxyl Radicals with 2′-Deoxyadenosine
and 2′-Deoxyguanosine in the Presence of Oxygen. The above-
described approach of selectively generated C5′ radicals by
reaction of eaq- with 8-bromopurine derivatives does not operate
with 8-bromo-2′-deoxyguanosine because the electron adduct
undergoes a fast protonation at C8 to afford the one-electron-
oxidized 2′-deoxyguanosine.48,49 However, the two diastereo-
meric forms (5′S) and (5′R) of 5′,8-cyclo-2′-deoxyguanosine
were identified in γ-irradiation of N2O-saturated aqueous
solution of 2′-deoxyguanosine in a (5′R)/(5′S) ratio of 8.3:1.22,50
It is well-known that the reaction of HO• radicals with
2′-deoxyguanosine (23) occurs mainly at the base moiety by
addition (ca. 85%), and to a minor extent at the deoxyribose
unit by hydrogen abstraction (ca. 15%).51 Our recent results
have shown that the hydrogen abstraction from the C5′ position
is an important component of the sugar moiety.50 Therefore, it
is reasonable to expect substantial amounts of hydrated 5′-
aldehydes 24 in the reaction of HO• radicals with 23 under oxic
conditions (Scheme 7).52
Figure 5. HPLC run of γ-irradiation of 23 (1.44 mM) in N2O (80%)/O2
(20%)-purged phosphate-buffered aqueous solution (pH ) 7.1) with a dose
of 1.0 kGy (dose rate ) 8.0 Gy min-1). The peaks 1, 2, and 3 are the
starting materials (23), hydrated 5′-aldehyde (24), and guanine, respectively.
Inset: Radiation chemical yield G(-23) (9) and G(24) (b) as a function of
the irradiation dose.
Scheme 7. Reaction of Hydroxyl Radicals with 2′-Deoxyguanosine
in the Presence of Oxygen
An aqueous phosphate-buffered solution (pH 7.1) of 1.44 mM
23 was saturated with N2O(80%)/O2(20%) before irradiation.
During γ-radiolysis (dose rate ) 8.0 Gy min-1), continuous
bubbling of the gas mixture in the sample ensured the constant
concentration of O2 throughout the reaction time. The reaction
was monitored by HPLC at 0.5 kGy dose intervals up to 2 kGy
(Figure 5). The consumption of 23 (32% after 2 kGy) was
accompanied by the formation of two main products, identified
as guanine and the hydrated 5′-aldehyde 24. In an analogous
experiment after 6 kGy, the irradiated mixture was concentrated
to ca. 2 mL and passed through a column packed with RP-18
silica eluted with 0-2% acetonitrile/water. The hydrated 5′-
aldehyde 24 was isolated and fully characterized.52,53
The inset of Figure 5 shows the G of disappearance of 2′-
deoxyguanosine (23) and the G of formation of hydrated 5′-
aldehyde 24 vs dose. By extrapolating the radiation chemical
yields to zero dose, the values G(-23) ) 0.22 and G(24) )
0.04 µmol J-1 were found. Two important facts can be deduced:
(i) Considering G(HO•) ) 0.53 vs G(-23) ) 0.22, it can be
suggested that ca. 60% of the attack of HO• radicals on 23 is
not productive, affording 23 back. Interestingly, ca. 65% of HO•
radicals attack the guanine moiety to afford the deprotonated
one-electron-oxidized 2′-deoxyguanosine (25) that does not react
with oxygen.51 We suggest that under our experimental condition
the intermediate 25 is converted back to the 2′-deoxyguanosine.54,55
(ii) The ratio G(HO•)/G(24) approximately reflects the
percentage (ca. 8%) of HO• radicals attacking the H5′ of the
2′-deoxyguanosine.56
-
Under our normoxic conditions, 93% of the eaq was
converted to give further HO• (cf. eqs 3 and 4) for a total G(HO•)
-
) 0.53 µmol J-1, whereas the remaining eaq and H• atoms
were mainly converted to superoxide radical anion.41 The HO•
radicals are partitioned between guanine and sugar moieties of
23 with an overall rate constant of ∼5 × 109 M-1 s-1 27
.
Furthermore, the oxygen concentration would be high enough
to trap most of the alkyl radicals derived from the sugar moiety
on nucleoside, in analogy also with the reactions described in
the previous section.
(45) Navacchia, M. L.; Chatgilialoglu, C.; Montevecchi, P. C. J. Org. Chem.
2006, 71, 4445–4452.
(46) Chatgilialoglu, C. Chem.-Eur. J. 2008, 14, 2310–2320.
(47) We cannot exclude minor paths of radical translocation in the
intermediate 7 to positions other than the C5′ of the sugar radical.
Theoretical calculation at the B3LYP/6-31G* level suggested that
radical translocations to the C5′ and C3′ positions have activation
energies of 3.2 and 11. 1 kcal mol-1, respectively. See: Chatgilialoglu,
C.; Duca, M.; Ferreri, C.; Guerra, M.; Ioele, M.; Mulazzani, Q. G.;
Strittmatter, H.; Giese, B. Chem.-Eur. J. 2004, 10, 1249–1255.
(48) Chatgilialoglu, C.; Caminal, C.; Guerra, M.; Mulazzani, Q. G. Angew.
Chem., Int. Ed. 2005, 44, 6030–6032.
On the basis of the above-described findings, it is reasonable
to expect substantial amounts of hydrated 5′-aldehydes 9 in the
reaction of HO• radicals with 2′-deoxyadenosine (26) under
(49) Chatgilialoglu, C.; Caminal, C.; Altieri, A.; Vougioukalakis, G. C.;
Mulazzani, Q. G.; Gimisis, T.; Guerra, M. J. Am. Chem. Soc. 2006,
128, 13796–13805.
(50) Chatgilialoglu, C.; Bazzanini, R.; Jimenez, L. B.; Miranda, M. A.
Chem. Res. Toxicol. 2007, 20, 1820–1824.
(51) Candeias, L. P.; Steenken, S. Chem.-Eur. J. 2000, 6, 475–484.
(52) For an analogous experiment, see: Langfinger, D.; von Sonntag, C. Z.
Naturforsch., C: Biosci. 1985, 40, 446–448.
(54) Candeias and Steenken also found that in the reaction of HO• radicals
with 9-methylguanine the depletion of the starting purine is incom-
plete.51
(55) A similar phenomenon has been also observed in the radiolytic
degradation of 2′-deoxyguanosine in aerated aqueous solution. See:
Douki, T.; Spinelli, S.; Ravanat, J.-L.; Cadet, J. J. Chem. Soc., Perkin
Trans. 1999, 2, 1875–1880.
(53) The formation of 5′-carboxaldehyde has been reported in the heavy
ion-mediated decomposition products of 2′-deoxyguanosine. See:
Gromova, M.; Nardin, R.; Cadet, J. J. Chem. Soc., Perkin Trans. 1998,
2, 1365–1374.
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8383

A R T I C L E S
Boussicault et al.
normoxic conditions. Evidence exists that some H-abstraction
occurs from the deoxyribose unit and in particular from C5′
position.57 Aqueous phosphate-buffered solution (pH 7.1) of
1.17 mM 26 was saturated with N2O(80%)/O2(20%) before
irradiation. During γ-radiolysis (dose rate ) 8.0 Gy min-1),
continuous bubbling of the gas mixture in the sample ensured
the constant concentration of O2 during the experiment. The
reaction was monitored by HPLC at 0.5 kGy dose intervals up
to 2 kGy.37b The consumption of the starting material (48%
after 2 kGy) was accompanied by the formation of two main
products which elute at shorter retention times and are assigned
to adenine and the hydrated 5′-aldehyde 9. In terms of radiation
chemical yields, the values G(-26) ) 0.37 and G(9) ) 0.06
µmol J-1 were found. Following the above-mentioned analysis,
we suggest that under our experimental condition ca. 30% of
the attack of HO• radicals on 26 affords adenine-derived radicals
that are converted back to starting nucleoside.58 On the other
hand, on the basis of the ratio G(HO•)/G(9), we suggest that
ca. 11% of HO• radical attack occurs at the H5′ of the
2′-deoxyadenosine.56
cyclopurine lesions in different environments. Their identifica-
tion in mammalian cellular DNA in vivo, where levels can be
enhanced by conditions of oxidative stress, deserves further
attention.19–21
Experimental Section
Materials. 8-Bromo-2′-deoxyadenosine, 2′-deoxyinosine, 2′-
deoxyguanosine, and 2′-deoxyadenosine were purchased from Berry
& Associates and used without any further purification. All other
chemicals and solvents were purchased from Sigma-Aldrich and
used as received. Solutions were freshly prepared using water
purified with a Millipore (Milli-Q) system; their pH was buffered
with potassium phosphates (Merck). 8-Bromo-2′-deoxyinosine was
synthesized by bromination of 2′-deoxyinosine.25 The solutions of
nucleosides were freshly prepared immediately before each
experiment.
Pulse Radiolysis. Pulse radiolysis with optical absorption
detection was performed as previously reported.24
Continuous Radiolysis. The experiments were performed at
room temperature (22 ( 2 °C) on 5-mL samples using 60Co-
Gammacells, with a dose rate of ca. 8 Gy min-1. The absorbed
radiation dose was determined with the Fricke chemical dosimeter,
by taking G(Fe3+) ) 1.61 µmol J-1 60
Reaction mixtures were
.
Conclusions
analyzed with a Zorbax SB-C18 column (4.6 × 150 mm) and eluted
in triethylammonium acetate buffer (20 mM, pH 7) with a 0-20%
acetonitrile linear gradient over 35 min with a flow rate of 1.0 mL/
min (detection at 254 nm).
The site-specific generation of C5′ radicals under various
oxygen concentrations has been achieved by the reaction of
hydrated electrons (eaq-) with 8-bromo-2′-deoxyadenosine or
8-bromo-2′-deoxyinosine. In the range of 13-266 µM of oxygen
(typical oxygenated tissues), the C5′ radicals are partitioned
between two reaction channels (i.e., the reaction with the O2
and an unimolecular rearrangement) providing the hydrated 5′-
aldehyde and the 5′,8-cyclopurine nucleoside, respectively. The
formation of 5′,8-cyclopurines is relevant in all experiments,
and the yield increased by decreasing O2 concentration. Under
aerobic conditions (266 µM of O2), the 2′-deoxyadenosin-5′-yl
radical affords a 15:1 ratio in favor of the hydrated 5′-aldehyde.
In the reaction of HO• radicals with 2′-deoxyadenosine or 2′-
deoxyguanosine, under analogous conditions, the 5′,8-cyclopu-
rines could not be quantified because the formation of C5′
radicals account only by ca. 10%. Without doubt, the most
relevant site of H-atom abstraction by HO• radical from
2-deoxyribose moiety of these two nucleosides results to be the
H5′ position,56 in agreement with the computational data.7 The
rate constant for the reaction of C5′ radical with GSH, the so-
called “repair reaction”, is found to be 5 × 107 M-1 s-1. Since
the intracellular level of GSH in mammalian cells is in the
0.5-10 mM range,59 the repair reaction, the trapping by O2,
and the cyclization process should be in competition. However,
it is not appropriate to extend the present findings to the behavior
of the double-stranded DNA. Local conformations due to the
supramolecular organization should influence considerably
(either accelerating or reducing) the cyclization rate, and
therefore it is necessary to provide measurements of 5′,8-
2′-Deoxyadenosine Hydrated 5′-Carboxaldehyde. Method A.
An Ar-purged aqueous phosphate-buffered solution (10 mM, pH
) 7.1) of 8-bromo-2′-deoxyadenosine (50 mg, 0.15 mmol, 1.5 mM)
containing 0.25 M t-BuOH and 1 mM K3Fe(CN)6 was initially
γ-irradiated with a dose of 1.5 kGy. Subsequently, amounts of
K3Fe(CN)6 (1 mM) were added at 1.5 kGy dose intervals to keep
the oxidant concentration in the desired range. The solution was
irradiated with a total dose of 6 kGy. Under these conditions, there
is ca. 40% conversion of the starting material and an 85% yield of
the aldehyde among the products. After passage of reaction crude
through an ion-exchange resin (Amberlite IRA-400) to eliminate
the iron salts, the volume of the mixture was reduced to ∼2 mL by
lyophilization and chromatographed on RP-18 silica using 0-4%
acetonitrile/water as the eluent. The fractions containing the
hydrated 5′-aldehyde were lyophilized to afford 13.6 mg of the
pure product as a white powder. Method B. An aqueous phosphate-
buffered solution (10 mM, pH ) 7.1) of 8-bromo-2′-deoxyadenosine
(50 mg, 0.15 mmol, 2 mM) was γ-irradiated under air bubbling
with a dose of 10 kGy. Under these conditions, there is a 70%
conversion of the starting material and a 50% yield of the aldehyde
among the products. The volume of the reaction crude was reduced
to ∼5 mL by lyophilization and chromatographed on RP-18 silica
using 0-4% acetonitrile/water as the eluent. The fractions contain-
ing the hydrated 5′-aldehyde were lyophilized to afford 18.7 mg
of the pure product as a white powder. 1NMR (400 MHz, D2O) δ
2.51 (m, 1H, H-2′′), 2.78 (m, 1H, H-2′′), 4.08 (dd, 1H, JH3′ ) 1.9,
JH5′ ) 3.7 Hz, H-4′), 4.67 (m, 1H, H-3′), 5.16 (d, 1H, JH4′ ) 3.7
Hz, H-5′), 6.41 (dd, 1H, JH2′ ) 6.1, JH2′′ ) 8.2 Hz, H-1′), 8.06 (s,
1H, H-2), 8.21 ppm (s, 1H, H-8). 13C NMR (100 MHz, D2O) δ
38.9 (C-2′), 71.5 (C-3′), 85.4 (C-1′), 89.2 (C-4′), 89.4 (C-5′), 118.8
(C-5), 140.5 (C-8), 148.1 (C-4), 152.3 (C-2), 155.4 (C-6). (+)-
ESI-MS: 267.8 [100 (M + H)+], (+)-ESI-MS/MS: 249.8, 135.9
UV (H2O, nm): 213.5, 257.0 (max), 228.5 (min). The assignment
of the NMR resonances was carried out by COSY and HSQC
experiments.
(56) It is possible to estimate roughly the total percentage of HO• radical
attacks on the deoxyribose unit by assuming that the free base can
result from attacks other than the one to the H5′ position.1,2 The values
of G(guanine) ) 0.03 and G(adenine) ) 0.06 µmol J-1 were found
in the experiments of HO• radical with 23 and 26, respectively. We
calculated that ca. 14 and 22% of HO• radicals attack the deoxyribose
unit of 2′-deoxyguanosine and 2′-deoxyadenosine, respectively, which
is in excellent agreement with the estimated values obtained from the
reducing properties of sugar-derived radicals by pulse radiolysis.51,58
(57) Mariaggi, N.; Cadet, J.; Teoule, R. Tetrahedron 1976, 32, 2385–2387.
(58) Vieira, A. J. S. C.; Steenken, S. J. Am. Chem. Soc. 1990, 112, 6986–
6994.
2′-Deoxyinosine Hydrated 5′-Carboxaldehyde. The above-
described Methods A and B were used for 8-bromo-2′-deoxyinosine
(50 mg). Method A. The hydrated 5′-aldehyde was isolated in 60%
(59) Meister, A.; Anderson, M. E. Annu. ReV. Biochem. 1983, 52, 711–
760.
(60) Spinks, J. W. T.; Woods, R. J. An Introduction to Radiation Chemistry,
3rd ed.; Wiley: New York, 1990; p 100.
9
8384 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008

Fate of C5′ Radicals of Purine Nucleosides
A R T I C L E S
yield (22.8 mg, 0.084 mmol). Method B. The hydrated 5′-aldehyde
was isolated in 25% yield (10 mg, 0.037 mmol). 1NMR (400 MHz,
D2O) δ 2.57 (m, 1H, H-2′), 2.87 (m, 1H, H-2′′), 4.08 (m, 1H, H-4′),
4.72 (m, 1H, H-3′), 5.17 (d, 1H, JH4′ ) 3.7 Hz, H-5′), 6.51 (dd,
1H, JH2′ ) 6.7, JH2′′ ) 7.4 Hz, H-1′), 8.20 (s, 1H, H-2), 8.26 ppm
(s, 1H, H-8). 13C NMR (100 MHz, D2O) δ 39.1 (C-2′), 71.5 (C-
3′), 85.5 (C-1′), 89.2 (C-4′), 89.4 (C-5′), 124.2 (C-5), 140.0 (C-8),
148.2 (C-2), 148.7 (C-4), 157.4 (C-6). m/z (ESI) 268.7 [M + H]+,
m/z (ESI-MS/MS): 136.9 UV (H2O, nm): 248.7 (max), 217.9 (min).
The assignment of the NMR resonances was carried out by COSY
and HSQC experiments.
2′-Deoxyguanosine Hydrated 5′-Carboxaldehyde. An aqueous
phosphate-buffered solution (10 mM, pH ) 7) of 2′-deoxguanosine
(40 mg, 2 mM) was γ-irradiated with a dose of 5 kGy under
continuous bubbling of a mixture of N2O (80%) and O2 (20%).
The reaction crude was concentrated to ∼2 mL and passed through
a column packed with RP-18 silica eluted with water. The samples
containing the hydrated 5′-aldehyde were lyophilized to afford 2
mg of the pure product as a white powder. 1NMR (400 MHz, D2O)
δ 2.47 (m, 1H, H-2′), 2.82 (m, 1H, H-2′′), 4.04 (m, 1H, H-4′),
4.42 (m, 1H, H-3′), 5.16 (d, 1H, JH4′ ) 4.0 Hz, H-5′), 6.35 (dd,
1H, JH2′ ) 6.1, JH2′′ ) 8.6 Hz, H-1′), 7.94 ppm (s, 1H, H-8). 13C
NMR (100 MHz, D2O) δ 38.6 (C-2′), 71.3 (C-3′), 85.6 (C-1′), 89.0
(C-4′), 89.5 (C-5′), 117.2 (C-5), 139.1 (C-8), 151.4 (C-4), 153.8
(C-2), 158.4 (C-6). m/z (ESI) 283.7 [M + H]+, 265.8 [M + H -
H2O]+ m/z (ESI-MS/MS): 151.8 UV (H2O, nm): 252.3, 273.8
(max), 220.2 (min). The assignment of the NMR resonances was
carried out by COSY and HSQC experiments.
Acknowledgment. This work was supported in part by the
European Community’s Marie Curie Research Training Network
under Contract MRTN-CT-2003-505086 [CLUSTOXDNA]. The
support and sponsorship concerned by COST Action CM0603 on
“Free Radicals in Chemical Biology (CHEMBIORADICAL)” are
kindly acknowledged. We thank M. Lavalle, A. Monti, and A.
Martelli for assistance with the pulse radiolysis experiments.
Supporting Information Available: Product studies and pulse
radiolysis experiments. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA800763J
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8385