Cross-linked thymine-purine base tandem lesions: Synthesis, characterization, and measurement in γ-irradiated isolated DNA

Article abstract of DOI:10.1021/tx015594d

5-(Phenylthiomethyl)-2′-deoxyuridine has been recently shown to be a specific photolabile precursor of 5-(2′-deoxyuridilyl)methyl radical that is involved in the formation of tandem base lesion with vicinal guanine in oxygen-free aqueous solution. The thionucleoside was incorporated by either liquid or solid-phase phosphoramidite synthesis into dinucleoside monophosphates with a 2′-deoxyadenosine residue as the vicinal nucleoside located either at the 3′ or 5′-extremity. UV-C irradiation of the modified dinucleoside monophosphate under anaerobic conditions gives rise to cross-linked thymine^CH2-C8adenine tandem base lesions which were isolated and characterized by ¹H NMR and mass spectrometry analyses. The formation of the latter tandem lesions involved an intramolecular addition of the 5-(2′-deoxyuridilyl)methyl radical to the C8 of the adenine moiety. A sensitive and specific assay aimed at monitoring the formation of the four thymine^CH2-C8purine adducts, namely d(T∧G), d(G∧T), d(T∧A), d(A∧T), within DNA, was designed. This was based on a liquid chromatography analysis coupled to tandem mass spectrometry (HPLC-MS/MS) detection of the dinucleoside monophosphates which were quantitatively released from γ-irradiated DNA and oligodeoxyribonucleotides by enzymatic hydrolysis. The four lesions were detected in both single-stranded oligodeoxyribonucleotide and isolated DNA upon exposure to γ-radiation in oxygen-free aqueous solution. It was found that the tandem guanine-thymine lesions were produced more efficiently than the adenine-thymine cross-links. Moreover, a significant sequence effect was observed. Thus, the yield of formation of the tandem lesions is higher when the purine base is located at the 5′ position of the 5-(2′-deoxyuridilyl)methyl radical.

Full text of DOI:10.1021/tx015594d

598
Chem. Res. Toxicol. 2002, 15, 598-606
Cr oss-Lin k ed Th ym in e-P u r in e Ba se Ta n d em Lesion s:
Syn th esis, Ch a r a cter iza tion , a n d Mea su r em en t in
γ-Ir r a d ia ted Isola ted DNA
Sophie Bellon, J ean-Luc Ravanat, Didier Gasparutto, and J ean Cadet*
Laboratoire des Le´sions des Acides Nucle´iques, Service de Chimie Inorganique et Biologique,
UMR 5046, De´partement de Recherche Fondamentale sur la Matie`re Condense´e, CEA-Grenoble,
F-38054 Grenoble Cedex 9 (France)
Received December 20, 2001
5-(Phenylthiomethyl)-2′-deoxyuridine has been recently shown to be a specific photolabile
precursor of 5-(2′-deoxyuridilyl)methyl radical that is involved in the formation of tandem base
lesion with vicinal guanine in oxygen-free aqueous solution. The thionucleoside was incorporated
by either liquid or solid-phase phosphoramidite synthesis into dinucleoside monophosphates
with a 2′-deoxyadenosine residue as the vicinal nucleoside located either at the 3′ or
5′-extremity. UV-C irradiation of the modified dinucleoside monophosphate under anaerobic
conditions gives rise to cross-linked thymineCH2-C8adenine tandem base lesions which were
1
isolated and characterized by H NMR and mass spectrometry analyses. The formation of the
latter tandem lesions involved an intramolecular addition of the 5-(2′-deoxyuridilyl)methyl
radical to the C8 of the adenine moiety. A sensitive and specific assay aimed at monitoring
the formation of the four thymineCH2-C8purine adducts, namely d(T∧G), d(G∧T), d(T∧A),
d(A∧T), within DNA, was designed. This was based on a liquid chromatography analysis coupled
to tandem mass spectrometry (HPLC-MS/MS) detection of the dinucleoside monophosphates
which were quantitatively released from γ-irradiated DNA and oligodeoxyribonucleotides by
enzymatic hydrolysis. The four lesions were detected in both single-stranded oligodeoxyribo-
nucleotide and isolated DNA upon exposure to γ-radiation in oxygen-free aqueous solution. It
was found that the tandem guanine-thymine lesions were produced more efficiently than the
adenine-thymine cross-links. Moreover, a significant sequence effect was observed. Thus, the
yield of formation of the tandem lesions is higher when the purine base is located at the 5′
position of the 5-(2′-deoxyuridilyl)methyl radical.
reaction in oxygen-free aqueous solution of DNA (7). The
structure of the latter vicinal base lesion involves a
covalent bond between the C8 carbon atom of a guanine
residue and the carbon atom of the methyl group of an
adjacent thymine. The tandem lesion was initially de-
scribed in a GT containing tetramer oligonucleotide
following exposure to X-radiation in deaerated aqueous
solution (7). Interestingly, both tandem base lesions
result from a single free-radical initiating event which
leads to the formation of the double modifications.
The d(T∧G)1 and d(G∧T) cross-linked tandem base
lesions were isolated from UV photolysis under anaerobic
conditions of dinucleoside monophosphates that con-
tained a photolabile precursor of 5-(2′-deoxyuridilyl)-
methyl radical (9, 10). The latter bridged thymine-
guanine adducts have been characterized by NMR and
MS analyses. Recently, the reactivity of 5-(2′-deoxyuri-
dilyl)methyl radical, was investigated by Greenberg et
al (11, 12). The latter radical was generated by Norrish
Type I photocleavage of the 5-(1-(3-phenyl-2-oxopropyl))-
2′-deoxyuridine motif, which implies abstraction of the
hydrogen atom from the methyl group. It was concluded
that the generation of the radical may induce the damage
transfer from the nucleobase to an adjacent nucleotide
in DNA under hypoxic conditions.
In tr od u ction
Ionizing radiation, high-intensity laser UV photolysis,
and UV-A photosensitization that generate reactive
oxygen species and/or mediate one-electron oxidation
processes are known to induce oxidative damage to
nucleic acids. This includes altered bases, strand breaks,
abasic sites, DNA-DNA, and DNA-protein cross-links
(1-3). During the last two decades, major efforts were
made to isolate and characterize modified nucleobases
within isolated DNA including tandem base lesions. Two
different classes of vicinal base damage were identified
by Box et al. (4-7). They are constituted by two adjacent
modified bases, such as in 8-oxo-7,8-dihydro-2′-deoxygua-
nosylyl-(3′f5′)-N-(2-deoxy-â-D-erythro-pentofuranosyl)-
formylamine (8-oxodG-dâF). The lesion 8-oxodG-dâF that
contains an oxidized guanine and an adjacent pyrimidine
base remnant was first identified in short oligodeoxyri-
bonucleotides upon exposure to X-rays in aerated aque-
ous solution. Then, the characterization and the mea-
surement of the latter modification within isolated
γ-irradiated DNA have been reported (4-6, 8). The
second type of double lesion involves a covalent bridge
between a thymine and a vicinal guanine (4, 6, 7). The
•
adduct has been found to be generated by OH radical
One of the two main objectives of the present work was
to extend our previous studies on the formation of d(G∧T)
* To whom the correspondence should be addressed. Phone: +33
4-38-78-49-87. Fax: +33 4-38-78-50-90. E-mail: cadet@drmfc.ceng.cea.fr.
10.1021/tx015594d CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/15/2002

Cross-Linked Thymine-Purine Tandem Lesions
Chem. Res. Toxicol., Vol. 15, No. 4, 2002 599
million), using the residual proton signal of D2O (δH ) 4.65) as
and d(T∧G) adducts to other vicinal tandem bridged
thymine-purine base in γ-irradiated isolated DNA. Thus,
two new tandem base lesions were isolated, namely
thymine-adenine adduct d(T∧A) and its reversed se-
quence d(A∧T). The latter tandem modifications were
generated using a similar approach than that previously
described for the thymine-guanine vicinal base lesions
(9, 10). The UV-C reactive precursor of 5-(2′-deoxyuridi-
lyl)methyl radical was incorporated into ODN located
either 3′ or 5′ to a 2′-deoxyadenosine nucleoside.
the external reference.
Ma ss Sp ectr om etr y Mea su r em en ts. All modified and
unmodified oligonucleotides were characterized by electrospray
ionization mass spectrometry measurement (ESI-MS) using a
LCQ ion-trap model mass spectrometer from Thermo-Finnigan
(San J ose, CA). Typically, for the analysis in the negative mode,
0.1 AU260nm of the dinucleoside monophosphates was dissolved
in a mixture of acetonitrile and water (50/50, v/v) that contained
1% of triethylamine. For the measurements performed in the
positive mode, the samples were dissolved in a water/methanol
mixture (50/50, v/v).
In addition, an assay aimed at measuring the four
tandem base lesions d(G∧T), d(T∧G), d(A∧T), and d(T∧A)
within isolated DNA was designed. The four lesions were
efficiently separated by high-performance liquid chro-
matography, and the detection of the damage was
achieved using the highly accurate electrospray ioniza-
tion tandem mass spectrometry (ESI-MS/MS) technique.
The four vicinal base lesions were produced in both ODNs
and isolated DNA upon exposure to γ-rays in oxygen-
free aqueous solution. Interestingly, it was found that
the tandem lesions were produced in a much higher yield
in the purine(5′f3′)thymine compounds than in the
reversed sequence within both oligonucleotides and calf
thymus DNA. Moreover, the two tandem base lesions
including the guanine residue are generated more ef-
ficiently than the respective tandem base damage bearing
the adenine moiety.
High -P er for m a n ce Liqu id Ch r om a togr a p h y. System A:
Reversed-phase HPLC (Hypersil C18 column, 5 µm, 250 × 4.6
mm) with a mixture of acetonitrile and 25 mM ammonium
formate buffer (AF, pH 6.2) as the eluents [100% AF (2 min),
linear gradient from 0 to 20% of acetonitrile (50 min)] at a flow
rate of 1 mL min-1. UV detection at 260 nm. System B:
Reversed-phase HPLC (Hamilton PRP3, polymeric phase col-
umn, 10 µm, 305 × 7.0 mm) with a mixture of acetonitrile and
10 mM TEAA buffer as the eluents [100% TEAA (5 min), linear
gradient from 0 to 8% of acetonitrile (10 min), then isocratic
TEAA-acetonitrile (92/8) v/v (10 min); after isocratic 100% TFA
(1%) (10 min) and finally a gradient from 0 to 10% acetonitrile
(40 min)] with a flow rate of 2.5 mL min-1. UV detection set at
260 nm. System C: Reversed-phase HPLC (Hypersil C18 column,
5 µm, 250 × 4.6 mm) with a mixture of acetonitrile and 25 mM
ammonium formate buffer (AF, pH 6.2) as the eluents [100%
AF (2 min), linear gradient from 0 to 20% of acetonitrile (30
min)] at a flow rate of 1 mL min-1, UV detection at 260 nm.
System D: Reversed-phase HPLC (Uptisphere ODB, 3 µm, 150
× 2 mm); elution with acetonitrile and TEAA (5 mM) [linear
gradient: from 0 to 40% of a 20% acetonitrile solution in TEAA
(10 min) and then from 40% to 60% (20 min) at a flow rate of
0.2 mL/min].
Exp er im en ta l Section
Gen er a l P r oced u r es a n d Ma ter ia ls. The silica gel (70-
200 µm) used for the low-pressure column chromatography was
purchased from SDS (Peypin, France). Thin-layer chromatog-
raphy was performed on DC Kieselgel Polygram SilG/UV254
(0.2 mm) plastic sheets from Macherey-Nagel (Dueren, Ger-
many). Deuterated solvents were purchased from Acros (Geel,
Belgium). Buffers for high performance liquid chromatography
(HPLC) were prepared using water purified with a Milli-Q
apparatus (Milford, MA).
Syn th etic P r oced u r es (Scheme 2). 3′-O-Acetyl-5-(phenylth-
iomethyl)-2′-deoxyuridine (2) was prepared as previously de-
scribed (10). Then, the modified nucleoside was coupled by
the phosphoramidite liquid-phase synthesis method to com-
mercial pacdAdo building block (1); subsequently, the deprotec-
tion and purification of the dinucleoside monophosphate 4
[5′-d(ATSPh)-3′] were achieved as reported below.
Nuclease P1 (Penicillium citrinium), bovine intestinal mucosa
phosphodiesterase (3′-exo), and calf thymus DNA were obtained
from Sigma (St. Louis, MO). Calf spleen phosphodiesterase (5′-
exo) and calf intestinal alkaline phosphatase were purchased
from Boehringer Mannheim (Mannheim, Germany).
Syn th esis of P r od u ct 3. Commercially available 5′-O-
dimethoxytrityl-N6-phenoxyacetyl-2′-deoxyadenosine-3′-O-(â-
cyanoethyl-N,N-diisopropyl)phosphoramidite (1) (1 g, 1.12 mmol)
was dissolved in 30 mL of dry acetonitrile and 3′-O-acetyl-5-
(phenylthiomethyl)-2′-deoxyuridine (2) (486 mg, 1.24 mmol)
dissolved in 30 mL dry acetonitrile and 427 mg of tetrazole (6
mmol) were added. The resulting solution was stirred at room
temperature for 20 min and then the reaction mixture was
partly reduced. Then, 40 mL of 0.1 M iodine solution in THF-
water-pyridine was added and the reaction mixture was stirred
for 45 min. Finally the reaction was quenched by addition of 10
mL of 1 M Na2S2O3 and 100 mL of chloroform. The organic layer
was washed with water and dried by addition of Na2SO4 prior
to be evaporated to dryness under vacuum. The resulting yellow
oil was dissolved in 1% TFA in methylene chloride solution and
stirred for 30 min. Then, the organic solution was evaporated
to dryness under vacuum. The residue thus obtained was
purified by chromatography on a silica gel column using a step
gradient of methanol in methylene chloride (from 0 to 5%).
Evaporation to dryness of the appropriate fractions yielded 3
(1.12 g, 53%). Rf (CHCl3/CH3OH 95/5): 0.32. ESI-MS (positive
mode): m/z 892.9 [M + H]+, 915.1 [M + Na]+.
The aqueous solutions of DNA and oligonucleotides were
deaerated by bubbling for 15 min with a nitrogen stream prior
to UV or gamma irradiation.
NMR Mea su r em en ts. 1H NMR spectra of 400 MHz were
recorded on a U400 (Varian) operating in the Fourier transform
mode. The chemical shifts are reported in ppm (parts per
1
Abbreviations: MMTrCl, 4-monomethoxytrityl chloride; TFA, tri-
fluoroacetic acid; DCM, dichloromethane; ESI-MS, electrospray ioniza-
tion mass spectrometry; ESI-MS/MS, electrospray ionization tandem
mass spectrometry; MRM, multiple reaction monitoring; ODN, oli-
godeoxyribonucleotide; AF, ammonium formate buffer; AU, absorbance
unit; TEAA, triethylammonium acetate; pac, phenoxyacetyl; TSPh
,
5-(phenylthiomethyl)-2′-deoxyuridine; 8-oxodGuo, 8-oxo-7,8-dihydro-
2′-deoxyguanosine; DHT, 5,6-dihydrothymine; d(T∧G), 2′-deoxy-8-[[1-
(2-deoxy-â-D-erythro-pentofuranosyl)-1,2,3,4-tetrahydro-2,4-dioxo-5-
pyrimidinyl]methyl]-3′-guanylic acid intramol. 3′,5′′′-ester; d(G∧T), 2′-
deoxy-8-[[1-(2-deoxy-â-D-erythro-pentofuranosyl)-1,2,3,4-tetrahydro-2,4-
dioxo-5-pyrimidinyl]methyl]-5′-guanylic acid intramol. 5′,3′′′-ester;
d(T∧A), 2′-deoxy-8-[[1-(2-deoxy-â-D-erythro-pentofuranosyl)-1,2,3,4-tet-
rahydro-2,4-dioxo-5-pyrimidinyl]methyl]-3′-adenylic acid intramol. 3′,5′′′-
ester; d(A∧T), 2′-deoxy-8-[[1-(2-deoxy-â-D-erythro-pentofuranosyl)-
1,2,3,4-tetrahydro-2,4-dioxo-5-pyrimidinyl]methyl]-5′-adenylic acid
intramol. 5′,3′′′-ester; 5′-d(ATSPh)-3′, 2′-deoxyadenylyl-(3′f5′)-5-(phe-
nylthiomethyl)-2′-deoxyuridine; 5′-d(GTSPh)-3′, 2′-deoxyguanylyl-(3′f5′)-
5-(phenylthiomethyl)-2′-deoxyuridine; 5′-d(AT)-3′, 2′-deoxyadenylyl-
(3′f5′)-thymidine; 5′-d(AHMU)-3′, 2′-deoxyadenylyl-(3′f5′)-5-(hydroxy-
methyl)-2′-deoxyuridine.
The dinucleoside monophosphate 4 was obtained after depro-
tection of 3 (480 mg, 0.54 mmol), by treatment with concentrated
aqueous ammonia (30%, 30 mL) at room temperature for 2 h.
After evaporation of the solvent to dryness under vacuum, the
crude 2-mer 4 [5′-d(ATSPh)-3′] was purified by reversed-phase
HPLC (system A) with a retention time of 42 min.
ESI-MS (positive mode): m/z 664.0 [M + H]+, 686.0 [M +
1
Na]+, 702.0 [M + K]+. H NMR in D2O (400 MHz): δ (in ppm)

600 Chem. Res. Toxicol., Vol. 15, No. 4, 2002
Bellon et al.
F igu r e 1. HPLC elution profile of the UV-C irradiated 4 on a C18 Hypersil analytical column (system A).
Sch em e 1. Str u ctu r e of th e Th ym in eCH2-C8P u r in e Ta n d em Ba se Lesion s Stu d ied in Th is Wor k
8.0 (s, 1H, H6T); 7.9 (s, 1H, H2A); 7.1 (m, 5H, H-arom PhS); 7.0
(s, 1H, H8A); 6.20 (t, 1H, H1′T); 5.92 (t, 1H, H1′A); 4.17 (m, 2H,
H3′T, H3′A); 4.08 (m, 1H, H4′T); 3.97 (m, 1H, H4′A); 3.86 (m, 2H,
room temperature for 4 h. Thereafter, solvents were removed
by evaporation under vacuum. Then, the crude 5′-DMTr-
oligomers were purified and deprotected on-line by reversed-
phase HPLC (system B) (14). The purity of the collected fractions
was controlled by HPLC analyses (system C) and ESI mass
spectrometry measurements.
γ-Ir r a d ia tion of Isola ted DNA a n d ODN. Deaerated
aqueous solutions of either calf thymus DNA or ODNs (0.5 mg
mL-1) were exposed to the γ rays of a 60Co source immersed in
a pool. The dose rate was 18 Gy.min-1. The selected irradiation
times ranged from 0 to 16 min.
UV-C Ir r a d ia tion of Oligon u cleotid es. Oxygen-free aque-
ous solutions of oligonucleotides containing the thiophenyl
nucleoside (1 AU260nm mL-1 solution) were exposed to UV-C light
provided by 8 lamps (λmax ) 254 nm) from a Rayonet photore-
actor (The Southern New England Ultraviolet Company, Ham-
den, CT). The irradiation times ranged from 5 to 10 min.
P h otoch em ica l Syn th esis of d (A∧T). 1 AU260nm mL-1 of 4
in 10 mM sodium phosphate oxygen-free buffer (pH 7) that
contained 100 mM NaCl, was exposed to UV-C radiation. The
H
5′5′′T); 3.68 (m, 2H, H5′5′′A); 3.51 (m, 2H, CH2-T); 2.64 (m, 2H,
H2′T, H2′A); 2.06 (m, 1H, H2′′T); 1.70 (m, 1H, H2′′A).
Solid -P h a se Syn t h esis of Oligod eoxyr ib on u cleot id es.
The synthesis of oligodeoxyribonucleotides was performed at 1
µmol scale using the “Pac phosphoramidite” chemistry (13), with
retention of the 5′ terminal DMTr group (“trityl-on” mode). The
standard 1 µmole DNA cycle was used on a 392 DNA synthesizer
(Applied Biosystems Inc, Palo Alto, CA) with a few modifica-
tions. First, the duration of the condensation was increased 4
times for the modified nucleoside phosphoramidite (120 s
instead of 30 s for a normal nucleoside phosphoramidite).
Second, a 0.3 M solution of phenoxyacetic acid in anhydrous
tetrahydrofuran was used for the capping step.
Dep r otection a n d P u r ifica tion of Oligod eoxyr ibon u cle-
otid es. Upon completion of the synthesis, the alkali-labile
protecting groups of the oligodeoxyribonucleotides were removed
by treatment with concentrated aqueous ammonia (30%) at

Cross-Linked Thymine-Purine Tandem Lesions
Chem. Res. Toxicol., Vol. 15, No. 4, 2002 601
Sch em e 2. Syn th etic P a th w a y Used for th e
P r ep a r a tion of th e Din u cleosid e Mon op h osp h a te
5′-[ATSP h ]-3′ 4a
a
Reagents: (a) tetrazole, CH3CN, 20 min, RT; (b) I2, THF/H2O/
pyridine, 45 min, RT; (c) 1% TFA, CH2Cl2, 30 min, RT; (d) NH4OH
30%, 2H, RT.
F igu r e 2. ESI-MS and ESI-MS/MS spectra (in the negative
mode) of the tandem base damage d(A∧T).
main products were isolated by analytical HPLC (system A).
The d(A∧T) tandem base lesion was obtained in a 10% yield.
ESI-MS (Nega tive Mod e) of d (A∧T) Ta n d em Ba se Le-
sion . (Figure 3) m/z 552.2 [M - H]-, 574.3 [M - 2H + Na]-;
MS2 m/z 454.2 [M - H]-; MS3 m/z 258.3 [M - H]-.
1H NMR (400 MHz, D2O) of d (A∧T) Ta n d em Ba se Lesion .
δ (in ppm): 8.0 (s, 1H, H2A); 7.8 (m, 1H, H6T); 6.3 (q, 1H, H1′A);
6.2 (t, 1H, H1′T); 4.8 (m, 1H, H3′A); 4.3 (m, 2H, H3′T, H4′A); 4.0
(m, 5H, CH2T,H5′5′′T, H4′T); 3.7 (m, 2H, H5′5′′A); 3.1 (m, 1H, H2′A);
2.4 (q, 1H, H2′′A); 2.2 (q, 2H, H2′2′′T).
En zym a tic Digestion of Din u cleosid e Mon op h osp h a tes
by Nu clea se P 1 a n d Alk a lin e P h osp h a ta se. A total of 0.5
AU260nm of oligonucleotide in water (45 µL) was digested at 37
°C by incubation for 2 h with 5 units of nuclease P1 (1 unit/µL)
in a 30 mM NaOAc and 0.1 mM ZnSO4 aqueous solution (pH
5.5). Then, 5 µL of Tris 500 mM, 1 mM EDTA (pH 8.5), and 2
units of alkaline phosphatase were added. The resulting mixture
was subsequently incubated at 37 °C for 1 h and then diluted
in 50 µL of 25 mM AF (pH 6.2). After centrifugation, the mixture
was finally resolved by reversed-phase HPLC (system A). The
different products were identified by co-injection with synthetic
standards and electrospray ionization mass spectrometry analy-
sis in the negative mode.
ESI-MS (negative mode) of d(AT) Dinucleoside Monophos-
phate: m/z 554.3 [M - H]-.
1H NMR (400 MHz, D2O) of d (AT) Din u cleosid e Mon o-
p h osp h a te. δ (in ppm): 8.2 (s, 1H, H2A); 8.0 (s, 1H, H8A); 7.3
(m, 1H, H6T); 6.3 (t, 1H, H1′A); 6.0 (t, 1H, H1′T); 4.7 (m, 2H,
H3′A); 4.4 (q, 1H, H3′T); 4.2 (q, 1H, H4′A); 4.0 (m, 1H, H4′T); 3.9
(m, 2H, H5′5′′T); 3.7 (m, 2H, H5′5′′A); 2.7 (m, 2H, H2′2′′A); 1.5 (d,
3H, CH3T).
En zym a tic Digestion of Din u cleosid e Mon op h osp h a tes
by Ca lf Sp leen P h osp h od iester a se (5′-exo). A total of 0.2
AU260nm of oligonucleotide in 20 µL of 0.02 M ammonium citrate

602 Chem. Res. Toxicol., Vol. 15, No. 4, 2002
Bellon et al.
F igu r e 3. Collision-induced dissociation mass of the pseudo-molecular ion of 5′-d(A∧T)-3′ and 5′-d(T∧A)-3′ (a), 5′-d(G∧T)-3′ and
5′-d(T∧G)-3′ (b) obtained in the negative electrospray ionization mode.
(pH 5) was digested at 37 °C by incubation with 10-3 units of
calf spleen phosphodiesterase for 1 h and the resulting mixture
was directly submitted to reversed-phase HPLC analysis (sys-
tem A). The different products were collected and analyzed by
electrospray ionization mass spectrometry in the negative mode.
mode was performed at -3300 V using a turbospray ionization
source (SCIEX, Thornill, Canada) with the turbospray gas
heated at 500 °C. The multiple reaction monitoring technique
(MRM) using the four transitions 552f454, 552f258, 568f470,
and 568f274 allowed a sensitive detection of d(A∧T), d(T∧A),
and d(G∧T), d(T∧G). The collision energy 45 eV was applied to
the pseudo-molecular ion. Accurate quantification was obtained
by external calibration.
En zym a tic Digestion of Din u cleosid e Mon op h osp h a tes
by Bovin e In testin a l Mu cosa P h osp h od iester a se (3′-exo).
Similarly, enzymatic digestion of oligonucleotides was performed
using 10-5 units of bovine intestinal mucosa phosphodiesterase
in 0.02 M ammonium citrate buffer (pH 9).
Resu lts a n d Discu ssion
En zym a tic Digestion of Isola ted DNA. Fractions of DNA
samples (100 µL, 50-100 µg) were incubated for 2 h at 37 °C
with 5 µL of nuclease P1 (1 unit/µL), 1 µL of calf spleen
phosphodiesterase (0.004 unit), and 10 µL of enzymatic buffer
(200 mM succinic acid, 100 mM CaCl2, 0.2 M ammonium citrate,
pH 5). Then, 10 µL of alkaline phosphatase buffer (500 mM
TRIS, 1 mM EDTA, pH 8.5) was added together with 5 units of
alkaline phosphatase and 0.003 unit of bovine intestinal mucosa
phosphodiesterase and the incubation was resumed for 2 h. The
enzymatic reaction was quenched by addition of 10 µL of HCl
0.1 M. Then, chloroform (100 µL) was added, and the resulting
solution was centrifuged. The aqueous layer was collected, and
its content was analyzed by HPLC-MS/MS.
P r ep a r a t ion
a n d
Ch a r a ct er iza t ion
of
Th ym in eCH2-C8Ad en in e Ta n d em Ba se Lesion s. The
thiophenyl derivative of 3′-monoacetylthymidine 2 was
prepared according to the multiple-step procedure re-
ported by Romieu et al. (10). This consisted in the
5-hydroxymethylation of 2′-deoxyuridine followed by a
specific monoacetylation and the substitution of the
acetyl group by thiophenol (15, 16). The latter compound
was condensed by liquid-phase synthesis with dA-cya-
noethyl phosphoramidite 1. After deprotection in an
ammoniac solution, [5′-d(ATSPh)-3′] was obtained and
purified by HPLC. The dinucleoside monophosphate 4
was characterized by electrospray ionization mass spec-
HP LC-ESI-MS/MS Detection of th e Mu ltip le Ba se Le-
sion s. The HPLC apparatus consisted of a 7100 Hitachi-Merck
pump (Merck, Darmstadt, Germany) connected to a SIL-9A
automatic injector (Shimadzu, Tokyo, J apan) and equipped with
an octadecylsilyl silica gel column Uptisphere 3 µm ODB (150
× 2 mm) (system D). Detection of the tandem lesions was
performed using a API 3000 tandem mass spectrometer (Perkin-
Elmer, Toronto, Canada). Electrospray ionization in the negative
1
trometry in the negative mode and H NMR analysis. As
described previously for [5′-d(GTSPh)-3′], the 5-(2′-deoxy-
uridilyl)methyl radical was generated into DNA frag-
ments by UV-C irradiation (λmax ) 254 nm); the isolation
and the characterization of the covalent bridged tandem

Cross-Linked Thymine-Purine Tandem Lesions
Chem. Res. Toxicol., Vol. 15, No. 4, 2002 603
Ta ble 1. Sequ en ces a n d Rela tive Molecu la r Ma ss (Da ) of
th e ODNs Syn th esized a n d Used in Th is Stu d y
base damage generated under anaerobic conditions have
been previously performed (10).
UV irradiation of 2-mer [5′-d(ATSPh)-3′] in oxygen-free
aqueous solution gave rise to several photoproducts
including simple and multiple base lesions as observed
by HPLC separation with UV detection (Figure 1). The
major products were isolated and characterized. Thus,
the predominant peak eluted at 25 min was found to be
[5′-d(AT)-3′], which is likely to be generated by atom
hydrogen trapping. The dinucleoside monophosphate
eluted at 24 min contains 5-(hydroxymethyl)uracil as a
modified thymine base [5′-d(AHMU)-3′]. The slower eluting
compound is the starting dinucleoside monophosphate.
The fastest eluting photoproduct in the chromatogram
is the cross-linked adenine-thymine that is searched. The
other compounds in the elution profile have not been yet
identified.
mass (Da)
sequences
length
calcd
found
5′-d(GTSPhG)-3′
3
3
15
15
1008.8
976.8
1008.5
977.3
5′-d(ATSPhA)-3′
5′-d(ATT CCG GTG GAA GCC)-3′
5′-d(CTT CCG ATA GAA GGA)-3′
4593.0
4601.0
4592.6
4600.8
Enzymatic processing by either a mixture of nuclease
P1 and alkaline phosphatase, or bovine intestinal mucosa
phosphodiesterase followed by calf spleen phosphodi-
esterase (3′-and 5′-exonucleases respectively), was inves-
tigated. HPLC and mass spectrometry analyses of the
reaction mixtures showed that no hydrolysis of the
phosphodiester bond of the dinucleoside monophosphate
occurred under the two enzymatic digestion conditions.
Moreover, to confirm that the 3′ or 5′-adjacent nucleoside
of the tandem lesion was hydrolyzed, the 3-mer ODNs
5′-[ATSPhA]-3′ and 5′-[GTSPhG]-3′, prepared by the solid-
phase phosphoramidite method (Table 1), were exposed
to UV-C radiation under anaerobic conditions. In a
subsequent step, the irradiated mixture was analyzed by
HPLC-MS/MS before and after an enzymatic treatment
by the phosphodiesterase (3′-and 5′-exonucleases). Before
hydrolysis of the irradiated mixture of 5′-[ATSPhA]-3′, a
pseudo-molecular ion [M - H]- was found at m/z 866
using the HPLC-MS method. The latter ion was not
observed after the action of the 3′- and 5′-exonucleases.
In fact, a pseudo-molecular ion [M - H]- at m/z 552
corresponding to the dinucleoside monophosphates was
detected. Then, the release of the lesions was quantita-
tively assessed by the action of both the 3′-and 5′-
exonucleases either on ODN or DNA.
HP LC-ESI-MS/MS Assa y. Our objective was to de-
velop a quantitative method for the measurement of the
tandem base lesions d(A∧T), d(T∧A), d(G∧T), and d(T∧G),
within isolated DNA upon exposure to gamma radiation
in oxygen-free aqueous solution. In that respect, HPLC
separation combined with electrospray ionization tandem
mass spectrometry operating in the multiple reaction
monitoring (MRM) mode constitutes a specific and sensi-
tive detection method (4, 19-22). The full mass spectra
of both lesions d(T∧A) and d(A∧T) in the negative mode
exhibited a pseudo-molecular ion [M - H]- at m/z 552.
Two main fragments were observed in the MS/MS
spectra at m/z 454 and 258, respectively (Figure 3a). The
transitions 552f454 and 552f258 were monitored to
detect the tandem lesions in the MRM mode. The full
scan spectra of both d(T∧G) and d(G∧T) cross-linked
lesions recorded in the negative mode exhibited a pseudo-
molecular ion [M - H]- at m/z 568. Two main fragments
were observed in the MS/MS spectra of d(G∧T) similar
to d(T∧G) at m/z 470 and 274 (Figure 3b). Therefore, the
two major transitions 568f470 and 568f270 were
selected to detect such lesions. In addition, the HPLC
conditions used allow an efficient separation of the
sequence isomer on the ODB column (Figure 4). It may
be noted that the purine-thymine lesions are eluted
slower than the reversed sequence whereas the cross-
linked guanine-thymine adducts are eluted more rapidly
than the adenine-thymine tandem base lesions.
The targeted tandem damage, with the methyl carbon
atom of thymine linked to the C-8 carbon of adenine, [5′-
d(A∧T)-3′] eluted at 20 min, was isolated and character-
ized. The structure of this vicinal DNA lesion was
confirmed using tandem mass spectrometry (ESI-MS/MS,
Figure 2): the pseudo-molecular ion at m/z 552.2 ob-
tained in the negative mode is in agreement with the
calculated mass of 553.4. Collision-induced dissociation
of the pseudo-molecular ion provided additional struc-
tural information. The main fragment at m/z 454 results
from the loss of the sugar residue at the 5′-end of the
dinucleoside monophosphate. Further fragmentation of
the latter ion (m/z 454) gives rise to a fragment at m/z
258 corresponding to the thymine residue covalently
linked to adenine through the methylene bridge.
1
The H NMR spectra of [5′-d(A∧T)-3′] and [5′-d(AT)-
3′] dinucleoside monophosphates, that were collected
from HPLC runs, provided further support to the struc-
ture of the photoproducts. As striking features, both the
singlet resonance of the thymine methyl group and the
1
adenine H-8 signal are lacking in the H NMR spectrum
of [5′-d(A∧T)-3′]. Moreover, the signal corresponding to
the methylene bridge (CH2) in [5′-d(A∧T)-3′] resonates
as a multiplet pattern (δ ) 4.0 ppm). Further confirma-
tion of the assignment was inferred from 1H NMR
TOCSY experiments (see Supporting Information).
The dinucleoside monophosphate [5′-d(TSPhA)-3′] of
opposite sequence was synthesized using solid-phase
phosphoramidite method and a commercially available
adenine support. The tandem base lesion [5′-d(T∧A)-3′]
was isolated after UV-C irradiation under anaerobic
conditions. As described above, the fragmentation ob-
served by ESI-MS/MS confirmed the formation of a
methylene bridge between the C8 of the adenine base on
one hand and the methyl group of the thymine base on
the other hand.
The two new tandem base lesions d(A∧T) and d(T∧A)
were used as standards for the design and optimization
of an analytical method aimed at measuring their forma-
tion upon γ-irradiation of isolated DNA in oxygen-free
aqueous solution.
E n zym a t ic Hyd r olysis of DNA F r a gm en t s Th a t
Con ta in th e Ta n d em Lesion s. The ability for the
exonuclease to cleave the phosphodiester bond of the four
tandem base lesions, d(T∧A), d(A∧T), d(T∧G), d(G∧T)
(Scheme 1), obtained by UV-C irradiation of the corre-
sponding photoreactive dinucleoside monophosphate pre-
cursor was checked under different enzymatic hydrolysis
conditions (17, 18).
Mea su r em en t of t h e Ta n d em Lesion s w it h in
Isola ted DNA Exp osed to γ-Ra d ia tion . A deaerated
aqueous solution of calf thymus DNA was exposed to
doses of γ-radiation ranging from 0 to 280 Gy. The

604 Chem. Res. Toxicol., Vol. 15, No. 4, 2002
Bellon et al.
F igu r e 4. HPLC-MS/MS chromatograms of the four tandem base lesions (system D).
F igu r e 5. Radiation-induced formation of the four tandem base
lesions measured by HPLC-MS/MS (MRM mode) in calf-thymus
DNA upon exposure in oxygen-free aqueous solution to various
doses of γ-radiation.
F igu r e 6. Relative formation of the tandem base lesions
d(Purine∧T) and d(T∧Purine) in the 15-mer oligonucleotides
containing either the GTG sequence or the ATA sequence at a
central position upon exposure to gamma rays (dose: 288 Gy).
The measurement was achieved by HPLC-MS/MS in MRM
mode.
measurement of the vicinal lesions including thymine on
one hand and each of the two purine bases on the other
hand was performed as described above. It was found
that the tandem lesions are formed approximately lin-
early with the radiation doses (Figure 5). In addition, the
HPLC-ESI-MS/MS analysis clearly showed that d(G∧T)
and d(T∧G) tandem base lesions are produced preferen-
tially. A likely mechanism for the radiation-induced
formation of the cross-linked thymine-purine base tan-
dem lesions is accounted for by the initial formation of
the photolabile thiophenyl precursor of the latter thymine
radical. Interestingly, a strong sequence effect was
observed for the formation of both thymine-adenine and
thymine-guanine cross-linked lesions. Thus, the tandem
damage is generated in a higher yield when the purine
base is located on the 5′-end. This sequence effect may
be explained within DNA by the distances between the
5-(2′-deoxyuridilyl)methyl radical and the carbon 8 of
the purine base. The two concerned carbons are sepa-
rated by 6.30 Å when the purine base (either guanine or
adenine) is located at the 3′-extremity (23). This dis-
•
5-(2′-deoxyuridilyl)methyl radical by OH-mediated hy-
drogen abstraction from the thymine methyl group. Then,
intramolecular addition to the C8 of the vicinal purine
base takes place as shown from model studies involving

Cross-Linked Thymine-Purine Tandem Lesions
Chem. Res. Toxicol., Vol. 15, No. 4, 2002 605
features of the latter damage as it was done for the
vicinal oxidative lesions 8-oxodG-dâF or 8-oxodG-DHT
(24-26). Emphasis will be placed on the evaluation of
the mutational potential of the cross-linked tandem lesion
and the ability for repair enzymes to excise such a
damage.
Ack n ow led gm en t. The contributions of Colette Leb-
run (LRI/SCIB/CEA-Grenoble) to the electrospray mass
spectrometry measurements of synthetic compounds and
Pierre-Alain Bayle (RMN/SCIB/CEA-Grenoble) to the
NMR analysis are gratefully acknowledged. We thank
Dr. Thierry Douki for helpful discussions on the HPLC-
MS/MS experiments. Financial support from the Comite´
de Radioprotection (Electricite´ de France) and the French
Research Ministry (PhD fellowship to S.B.) is acknowl-
edged.
Su p p or tin g In for m a tion Ava ila ble: ESI-MS and ESI-
MS/MS spectra (in the negative mode) of 5′-d(AT)-3′, H NMR
1D, and 2D TOCSY spectra of 5′-d(A∧T)-3′ tandem base lesion
and 5′-d(AT)-3′, radiation-induced formation of the cross-linked
thymine-purine base tandem lesions. This material is available
free of charge via the Internet at http://pubs.acs.org.
F igu r e 7. Relative formation of d(G∧T), d(T∧G) and d(A∧T),
d(T∧A) in the 3-mer oligonucleotides GTSPhG and ATSPhA,
respectively, upon UV-C irradiated in oxygen-free aqueous
solution (measured by HPLC-MS/MS in MRM mode).
1
tance is reduced to 3.58 Å when the purine base is
situated at the 5′-extremity. Such a sequence effect was
also observed in single-stranded 15-mer ODNs containing
the GTG or ATA sequences at a central position (Table
1). There is a 4-fold preferential formation of the 5′-
(purine-thymine)-3′ compared to the reversed sequence
(Figure 6).
Refer en ces
(1) von Sonntag, C. (1987) The Chemical Basis of Radiation Biology,
Taylor & Francis, New York.
(2) von Sonntag, C., and Schuchmann, H.-P. (1990) Radical-mediated
DNA damage in the presence of oxygen. Methods Enzymol. 186,
511-520.
(3) Steenken, S. (1989) Purine bases, nucleosides, and nucleotides:
aqueous solution redox chemistry and transformation reactions
of their radical cations and e- and OH adducts. Chem. Rev. 89,
503-520.
(4) Box, H. C., Patrzyc, H. B., Dawidzik, J . B., Wallace, J . C., Freund,
H. G., Iijima, H., and Budzinski, E. E. (2000) Double base lesions
in DNA X-irradiated in the presence or absence of oxygen. Radiat.
Res. 153, 442-446.
In addition, the 3-mer 5′-d[ATSPhA]-3′ and 5′-d[GTSPhG]-
3′, were submitted to UV-C irradiation in deaerated
solution and the enzymatic hydrolysis was applied. The
HPLC-ESI-MS/MS analysis confirmed that the formation
of the vicinal lesion with the purine located on 5′-position
was produced in a 95% yield (Figure 7), compared to the
reversed sequence.
(5) Maccubbin, A. E., Iijima, H., Ersing, N., Dawidzik, J . B., Patrzyc,
H. B., Wallace, J . C., Budzinski, E. E., Freund, H. G., and Box,
H. C. (2000) Double-base lesions are produced in DNA by free
radicals. Arch. Biochem. Biophys. 375, 119-123.
Con clu sion a n d P er sp ectives
(6) Box, H. C., Budzinski, E. E., Dawidzik, J . B., Wallace, J . C., and
Iijima, H. (1998) Tandem lesions and other products in X-
irradiated DNA oligomers. Radiat. Res. 149, 433-439.
(7) Box, H. C., Budzinski, E. E., Dawidzik, J . D., Wallace, J . C.,
Evans, M. S., and Gobey, J . S. (1996) Radiation-induced formation
of a crosslink between base moieties of deoxyguanosine and
thymidine in deoxygenated solutions of d(CpGpTpA). Radiat. Res.
145, 641-643.
(8) Bourdat, A.-G., Douki, T., Frelon, S., Gasparutto, D., and Cadet,
J . (2000) Tandem base lesions are generated by hydroxyl radical
within isolated DNA in aerated aqueous solution. J . Am. Chem.
Soc. 122, 4549-4556.
(9) Bellon, S., Gasparutto, D., Romieu, A., and Cadet, J . (2001)
5-(Phenylthiomethyl)-2′-deoxyuridine as an efficient photoreactive
precursor to generate single and multiple lesions within DNA
fragments. Nucleosides, Nucleotides Nucleic Acids 20, 967-971.
(10) Romieu, A., Bellon, S., Gasparutto, D., and Cadet, J . (2000)
Synthesis and UV photolysis of oligonucleotides that contain
5-(phenylthiomethyl)-2′-deoxyuridine: a specific photolabile pre-
cursor of 5-(2′-deoxyuridilyl)methyl radical. Org. Lett. 2, 1085-
1088.
(11) Anderson, A. S., Hwang, J .-T., and Greenberg, M. M. (2000)
Independent generation and reactivity of 2′-deoxy-5-methyleneu-
ridin-5-yl, a significant reactive intermediate produced from
thymidine as a result of oxidative stress. J . Org. Chem. 65, 4648-
4654.
The two cross-linked lesions that involved the forma-
tion of a covalent bond between the methyl group of
thymine and the C8 of adenine were generated following
the methodology previously described for the formation
of thymine-guanine adducts. After isolation, the latter
tandem base lesions were characterized by electrospray
mass spectrometry and 1H NMR measurements. The
synthetic modified dinucleoside monophosphates d(A∧T)
and d(T∧A) together with the two other previously
described d(G∧T) and d(T∧G) lesions have been used as
standards for their measurement in ODNs and isolated
DNA following γ-irradiation under anaerobic conditions.
This was achieved by high-performance liquid chroma-
tography combined with electrospray ionization tandem
mass spectrometry in the multiple reaction monitoring
mode, subsequently to DNA enzymatic digestion. The
HPLC-MS/MS assay is an accurate and powerful method
to separate and measure the four radiation-induced
thymine-purine d(T∧A), d(A∧T), d(T∧G) and d(G∧T)
cross-linked base damage within single-stranded oligode-
oxyribonucleotides and isolated DNA in oxygen-free
aqueous solutions after their enzymatic release as di-
nucleoside monophosphate. Furthermore, an important
sequence effect was observed for the formation of both
lesion families. Finally, the tandem base lesions appear
to be important °OH-mediated lesions in isolated DNA.
Future work is aimed at delineating the biochemical
(12) Carter, N. K., Taverner, T., Schiesser, C. H., and Greenberg, M.
M. (2000) Chemical evidence for thiyl radical addition to the C6-
position of a pyrimidine nucleoside and its possible relevance to
DNA damage amplification. J . Org. Chem. 65, 8375-8378.
(13) Schulhof, J . C., Molko, D., and Te´oule, R. (1987) The final
deprotection step in oligonucleotide synthesis is reduced to a mild
and rapid ammonia treatment by using labile base-protecting
groups. Nucleic Acids Res. 15, 397-416.

606 Chem. Res. Toxicol., Vol. 15, No. 4, 2002
Bellon et al.
(14) Romieu, A., Gasparutto, D., Molko, D., and Cadet, J . (1997) A
convenient synthesis of 5-hydroxy-2′-deoxycytidine phosphora-
midite and its incorporation into oligonucleotides. Tetrahedron
Lett. 38, 7531-7534.
(15) Suzuki, Y., Matsuda, A., and Ueda, T. (1987) Synthesis of 6,5′-
cyclo-2′,5′-dideoxypyrimidine nucleosides. Chem. Pharm. Bull. 35,
1085-1092.
induced thymine photoproducts. J . Photochem. Photobiol. B: Biol.
54, 145-154.
(21) Douki, T., Court, M., Sauvaigo, S., Odin, F., and Cadet, J . (2000)
Formation of the main UV-induced thymine dimeric lesions
within isolated and cellular DNA as measured by high perfor-
mance liquid chromatography-tandem mass spectrometry. J . Biol.
Chem. 275, 11678-11685.
(16) Sowers, L. C., and Beardsley, P. G. (1993) Synthesis of oligo-
nucleotides containing 5-(hydroxymethyl)-2′-deoxyuridine at de-
fined sites. J . Org. Chem. 58, 1664-1665.
(17) 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′-deoxygua-
nosine. Chem. Res. Toxicol. 12, 412-421.
(18) Gasparutto, D., Ravanat, J .-L., Ge´rot, O., and Cadet, J . (1998)
Characterization and chemical stability of photooxidized oligo-
nucleotides that contain 2,2-diamino-4-[(2-deoxy-â-D-erythro-
pentofuranosyl)amino]-5(2H)-oxazolone. J . Am. Chem. Soc. 120,
10283-10286.
(19) Frelon, S., Douki, T., Ravanat, J .-L., Pouget, J .-P., Tornabene,
C., and Cadet, J . (2000) High-performance liquid chromatography-
tandem mass spectrometry measurement of radiation-induced
base damage to isolated and cellular DNA. Chem. Res. Toxicol.
13, 1002-1010.
(22) Ravanat, J .-L., Remaud, G., and Cadet, J . (2000) Measurement
of the main photooxidation products of 2′-deoxyguanosine using
chromatographic methods coupled to mass spectrometry. Arch.
Biochem. Biophys. 374, 118-127.
(23) Pullman, B., and Saran, A. (1976) Quantum-mechanical studies
on the conformation of nucleic acids and their constituents. Prog.
Nucleic Acid Res. Mol. Biol. 18, 215-322.
(24) Chaudhry, M. A., and Weinfeld, M. (1997) Reactivity of human
apurinic/apyrimidinic endonuclease and Escherichia coli exonu-
clease III with bistranded abasic sites in DNA. J . Biol. Chem.
272, 15650-15655.
(25) Bourdat, A.-G., Gasparutto, D., and Cadet, J . (1999) Synthesis
and enzymatic processing of oligonucleotides containing tandem
base damage. Nucleic Acids Res. 27, 1015-1024.
(26) David-Cordonnier, M.-H., Laval, J ., and O’Neill, P. (2001) Rec-
ognition and kinetics for excision of a base lesion within clustered
DNA damage by the Escherichia coli proteins Fpg and Nth.
Biochemistry 40, 5738-5746.
(20) Douki, T., Court, M., and Cadet, J . (2000) Electrospray-mass
spectrometry characterization and measurement of far-UV-
TX015594D