Analysis of M1G-dR in DNA by aldehyde reactive probe labeling and liquid chromatography tandem mass spectrometry

Article abstract of DOI:10.1021/tx049853l

A novel method for the measurement of pyrimido[1,2-α]purin-10(3H)one (M₁G) has been developed. Previous methods for analysis of M ₁G have been confounded by the fact that this lesion exists in equilibrium between a ring-closed form and a ring-opened aldehyde form. Poor detection sensitivity of the aldehydic form can result from loss of the adduct during analysis by its reaction with amines or proteins. We utilized the aldehyde reactive probe (ARP) to produce a stable ARP-M₁G-deoxyribose (ARP-M₁G-dR) conjugate to minimize adduct loss. This conjugate has increased the hydrophobicity that enhances separation of ARP-M₁G-dR from unmodified DNA nucleosides by using solid phase extraction. In addition, measuring ARP-M₁G-dR by selective reaction monitoring (SRM) of the [ARP-M₁G-dR + H]⁺ (635) → [M₁G + H] ⁺ (188) transition increases the detection sensitivity by nearly an order of magnitude relative to the measurement of M₁G-dR by SRM. For accurate measurement, analytical standard (AS) DNA and internal standard (IS) DNA were used. High purity ¹⁵N-labeled DNA was isolated from Escherichia coli that had been grown in minimum salt medium containing ( ¹⁵NH₄)SO₄. The ¹⁵N-DNA and calf thymus DNA were treated with malondialdehyde to induce a high number of M ₁G adducts to prepare the IS and AS DNA, respectively. A consistent calibration curve was established from the analysis of 200 μg of blank DNA, 23 ng of IS DNA (400 fmol of ¹⁵N₅-M₁G-dR), and AS DNA containing 0-810 fmol of M₁G-dR. With the use of this novel IS DNA and selective labeling, this assay is a useful tool for monitoring oxidative stress-induced DNA damage from small amounts of DNA without the need of a specific antibody or laborious procedures. By this assay, two M ₁G adducts/10⁸ guanines can readily be detected. Furthermore, this approach should be applicable to the analysis of other aldehydic DNA adducts as well as the measurement of an array of DNA lesions.

Full text of DOI:10.1021/tx049853l

Chem. Res. Toxicol. 2005, 18, 51-60
51
Analysis of M G-dR in DNA by Aldehyde Reactive Probe
1
Labeling and Liquid Chromatography Tandem Mass
Spectrometry
Yo-Chan Jeong, Ramiah Sangaiah, Jun Nakamura, Brian F. Pachkowski,
Asoka Ranasinghe, Avram Gold, Louise M. Ball, and James A. Swenberg*
Department of Environmental Sciences and Engineering,
University of North Carolina at Chapel Hill, North Carolina 27599-7431
Received May 26, 2004
A novel method for the measurement of pyrimido[1,2-a]purin-10(3H)one (M
1
G) has been
developed. Previous methods for analysis of M G have been confounded by the fact that this
1
lesion exists in equilibrium between a ring-closed form and a ring-opened aldehyde form. Poor
detection sensitivity of the aldehydic form can result from loss of the adduct during analysis
by its reaction with amines or proteins. We utilized the aldehyde reactive probe (ARP) to produce
a stable ARP-M
has increased the hydrophobicity that enhances separation of ARP-M
DNA nucleosides by using solid phase extraction. In addition, measuring ARP-M
1
G-deoxyribose (ARP-M
1
G-dR) conjugate to minimize adduct loss. This conjugate
G-dR from unmodified
G-dR by
1
1
+
+
selective reaction monitoring (SRM) of the [ARP-M
transition increases the detection sensitivity by nearly an order of magnitude relative to the
measurement of M G-dR by SRM. For accurate measurement, analytical standard (AS) DNA
and internal standard (IS) DNA were used. High purity N-labeled DNA was isolated from
1 1
G-dR + H] (635) f [M G + H] (188)
1
15
1
5
Escherichia coli that had been grown in minimum salt medium containing ( NH
4
)
2
SO
N-DNA and calf thymus DNA were treated with malondialdehyde to induce a high number
of M G adducts to prepare the IS and AS DNA, respectively. A consistent calibration curve
was established from the analysis of 200 µg of blank DNA, 23 ng of IS DNA (400 fmol of
4
. The
1
5
1
1
5
5 1 1
N -M G-dR), and AS DNA containing 0-810 fmol of M G-dR. With the use of this novel IS
DNA and selective labeling, this assay is a useful tool for monitoring oxidative stress-induced
DNA damage from small amounts of DNA without the need of a specific antibody or laborious
8
procedures. By this assay, two M
1
G adducts/10 guanines can readily be detected. Furthermore,
this approach should be applicable to the analysis of other aldehydic DNA adducts as well as
the measurement of an array of DNA lesions.
Introduction
mutation (4, 5). Base propenal, a byproduct of deoxyribose
oxidation, has been identified as a key intermediate,
producing the tricyclic DNA adduct, pyrimido[1,2-a]-
The role of endogenous DNA damage in genetic dis-
eases including cancer is an important issue in current
research. Reactive oxygen species (ROS) include endog-
1
purin-10(3H)one (M G) (6). More frequently, ROS can
1
abstract a hydrogen from polyunsaturated fatty acids to
produce a carbon-centered fatty acid radical, which can
be further oxidized to form a lipid peroxy radical. Lipid
peroxy radicals abstract hydrogen from neighboring
phospholipids, producing additional lipid peroxy radicals.
As a result, there may be up to 2 orders of magnitude
amplification in the production of free radicals (7, 8).
Lipid hydroperoxides are degraded to multiple reactive
byproducts such as malondialdehyde (MDA), 4-hydroxy-
nonenal, crotonaldehyde, and acrolein, which can damage
genomic DNA (9). Lipid peroxidation-induced DNA ad-
enous agents capable of causing multiple types of DNA
damage through different pathways. Estimations of such
oxidative DNA damage range from 10000 to over 200000
events per cell (1, 2). ROS can directly react with DNA
bases or deoxyribose, producing modified bases, abasic
sites, and DNA strand breaks (3). While base oxidation
and abasic sites are the most frequent insults to DNA,
substantial evidence supports that secondary DNA dam-
age produced from byproducts of primary DNA damage
or lipid peroxidation may play an important role in
6
ducts include tricyclic purine adducts such as 1,N -
2
*
To whom correspondence should be addressed. Tel: 919-966-6139.
ethenodeoxyadenosine, N ,3-ethenodeoxyguanosine, and
Fax: 919-966-6123. E-mail: james_swenberg@unc.edu.
3
-(2′-deoxy-â-D-erythro-pentofuranosyl)pyrimido[1,2-a]pu-
1
1 1
M G, pyrimido[1,2-a]purin-10(3H)one; M G-dR, 3-(2′-deoxy-â-
D-erythro-pentofuranosyl)pyrimido[1,2-a]purin-10(3H)one; ARP, alde-
hyde reactive probe (N′-aminoxymethylcarbonylhydrazino D-biotin);
SPE, solid phase extraction; SRM, selective reaction monitoring; AS,
analytical standard; IS, internal standard; CTD, calf thymus DNA;
MDA, malondialdehyde; E. coli, Escherichia coli; ROS, reactive oxygen
species; TEMPO, 2,2,6,6-tetramethylpiperidine-1-oxyl; MSM, minimum
salt medium; COSY, correlated spectroscopy; HSQC, heteronuclear
single quantum coherence; HMBC, heteronuclear multiple bond cor-
relation; NOESY, nuclear overhauser effect spectroscopy.
rin-10(3H)one (M G-dR) (10). MDA, which has been
1
extensively used as a biomarker for lipid peroxidation,
is the most mutagenic byproduct. â-Hydroxyacrolein, a
tautomer of MDA, is highly reactive toward guanine, pro-
ducing M
of ROS and MDA during lipid peroxidation chain reac-
tions, a relatively high number of M G-dR lesions have
1
G (11). Because of the continuous production
1
1
0.1021/tx049853l CCC: $30.25 © 2005 American Chemical Society
Published on Web 12/09/2004

52
Chem. Res. Toxicol., Vol. 18, No. 1, 2005
Jeong et al.
been reported in animal and human tissues (12, 13). In
human pancreatic tissue, the level of M G in the DNA
exceeded that of 8-oxoguanine, the most commonly
measured form of oxidative DNA damage. Therefore, it
1040A diode array detector combined with an Agilent Technolo-
gies (Palo Alto, CA) 1100 Series quaternary pump and degasser.
The reaction mixtures were separated on C-18 Platinum EPS
column (Alltec, Deerfield, IL, 150 mm × 4.6 mm, 5 µm). Full
1
scan and tandem mass spectra were obtained using a Finnigan
is imperative to accurately measure M
the biological significance of M G in disease (14).
Attempts to analyze M G have been hampered by its
1
G to determine
DECA
(Woburn, MA) LCQ
ion trap mass spectrometer. Quantifica-
1
1
tion of ARP-M G-dR and its IS were performed using a Finnigan
1
Quantum triple-quadrupole mass spectrometer connected to a
Finnigan Surveyor Micro-LC. NMR data were obtained on a
Varian (Palo Alto, CA) INOVA 500 spectrometer at 500 MHz
multiple chemical forms that change during DNA analy-
sis and hydrolysis. When the adduct is present in double-
1
13
stranded DNA, base pairing of dC opposite M
1
G facili-
for H and C NMR at 125 MHz. NMR spectra were recorded
2
1
13
tates ring opening to N -oxopropenyl-dG (15). The
6
in DMSO-d at ambient temperature. H and C chemical shifts
2
and coupling constants were determined by spectral simulation
using MestRe-C 3.5.1 software (mestrec@usc.es).
aldehydic moiety in N -oxopropenyl-dG is reactive toward
amine compounds (15, 16), which can hinder accurate
measurements due to competitive reactions that produce
unstable conjugates (17). Furthermore, the ring-opened
form is frequently not recognized by the anti-M G
1
antibody that has been used in immunoaffinity purifica-
tion or immunoslot-blot measurement (2). In the present
1
study, a stable conjugate of M G was generated by
reaction with aldehyde reactive probe (ARP) (Scheme 1).
ARP has been successfully used to produce stable con-
jugates with ring-opened, aldehydic AP sites for mea-
surement by a slot blot method (18, 19). The strategy of
M
1
G, M
1 1 1
G-dR, and ARP-M G-dR Synthesis. M G and
M
1
G-dR were prepared by the reaction of guanine with tetra-
methoxypropane according to a published procedure (25) with
slight modification. Ten millimoles of tetraethoxypropane was
added dropwise to 50 mM guanine in 200 mL of 1 M hydrochloric
acid. After 30 min at 70 °C, the reaction mixture was cooled to
0
°C on ice. Excess guanine was precipitated after the addition
of K CO to adjust the reaction mixture to neutral pH. After
2
3
centrifugation at 3000 rpm for 20 min, the supernatant was
divided into two parts for purification of M G or synthesis of
1
1 1
M G-dR. To isolate M G, the reaction mixture was lyophilized,
and then, the remaining solids were extracted twice with
methanol followed by concentration of the extract to 1 mL using
this research for measuring M
ing stable ARP-M G-dR conjugates in DNA prior to
enzymatic hydrolysis of the DNA to the nucleoside level
1
G-dR consisted of produc-
1
a vacuum evaporator. M
by column chromatography on silica gel with methanol:chloro-
form (12:88) elution. M G-dR was prepared by incubating the
crude M G fraction with the same volume of aqueous solution
1
G was purified from the concentrate
and subsequent quantification of ARP-M
spectrometry.
1
G-dR by mass
1
1
With several major instrumental improvements during
the past decade, mass spectrometry has gained more
popularity in DNA adduct analysis due to its outstanding
specificity and sensitivity (20-23). Application of a stable
isotope-labeled DNA adduct as an internal standard (IS)
dramatically enhances the accuracy of adduct measure-
ment (24). However, the accuracy and precision of mass
spectrometry can be undermined if the use of IS bases
or nucleosides cannot account for artifacts that arise
during DNA hydrolysis, such as additional oxidation of
unmodified nucleotides; incomplete hydrolysis of DNA to
nucleosides; or interactions between DNA adducts and
containing trans-N-deoxyribosylase (400 µg), 5 mmol of deoxy-
cytidine, and 100 mmol of MES. The enzymatic glycosylation
of M
lyophilized and extracted with methanol for subsequent column
chromatography with silica gel. M G-dR standard was eluted
with methanol:dichloromethane (8:92). M G and M G-dR stan-
1
G proceeded for 2 h at 37 °C. The reaction mixture was
1
1
1
dards were characterized by electrospray ionization mass
1
spectrometry (ESI-MS) and H NMR (data not shown).
1
To test the reactivity of ARP toward M G-dR, equimolar (1
mM) M G-dR and ARP were incubated in 1% acetic acid solution
1
for 30 min at 70 °C. The reaction mixture was separated by
HPLC. Three fractions (conjugates 1, 2, and 3; Figure 1) were
collected and reanalyzed by HPLC. For the scaled-up synthesis,
other molecules. In addition, M
opened aldehyde form by base pairing with dC, which
does not happen to the free M G base or its nucleoside
under the same conditions. This behavior can limit the
use of M G base or M G-dR as a standard in analysis.
1
G is hydrolyzed to a ring-
9
1
mg of M G-dR was dissolved in 6 mL of 0.2% acetic acid
solution followed by the addition of 6 mL of 10 mM ARP solution.
The reaction mixture was incubated for 2 days at 4 °C. The
precipitate was collected by centrifugation and washed twice
with 6 mL of cold water. The precipitate was lyophilized to
1
1
1
1
Furthermore, measurement is affected by the competitive
reaction of the aldehyde moiety of the ring-opened form
with amines as well as differences in hydrolysis ef-
remove any remaining water (recovery, 13.6 mg). H NMR (500
MHz, DMSO-d
6
): 11.16, 11.05 (2 × bs, 1H total, NH9syn/anti),
ARP ARP
9.77 (bs, 0.5H, NH15syn/anti ), 9.74 (bs, 1H total, NH14syn/anti ),
ARP
9.73 (bs, 0.5 H, NH15syn/anti
), 9.61 (d, 0.5H, NH5syn), 9.49 (d,
1
ficiency. To overcome these pitfalls, M G analytical
0
J
.5H, NH5anti), 8.10, 8.06 (2 × s, 1H, H2syn/anti), 7.98 (d, 0.5 H,
standard (AS) and IS were prepared from high purity
genomic DNA and 15N-labeled DNA, respectively. This
7,8 ) 10.2 Hz, H8anti), 7.60 (dd, 0.5 H, J6,7 ) 14.1, J6,5 ) 11.1
Hz, H6syn), 7.52 (dd, 0.5 Hz, J6,7 ) 13.9, J6,5 ) 10.8 Hz, H6anti),
1
approach mimics M G-dR in sample DNA and is expected
to produce reliable results that account for artifacts that
can arise during sample analysis.
ARP
7
(
.28 (d, 0.5 H, J7,8 ) 9.8 Hz, H8syn), 6.41 (bs, 1H, NH3 ), 6.35
ARP
bs, 1H, NH1 ), 6.33 (dd, 0.5 H, J7,6 ) 14.1, J7,8 ) 9.8 Hz,
H7syn), 6.24 (dd, 1 H, J1′,2′′ ) 14.1, J1′,2′ ) 7.3, H1′), 5.82 (dd, 0.5
H, J7,6 ) 13.9, J7,8 ) 10.2 Hz, H7anti), 5.31 (bs, 1 H, OH3′), 4.90
Materials and Methods
ARP
(
bs, 1 H, OH5′), 4.47, 4.43 (2 × s, 1 H total, CH
2
17syn/anti
),
4.39 (bs, 1 H, H3′), 4.30 (m, 1 H, H8 ), 4.11 (m, 1 H, H4ARP),
ARP
Materials. HPLC grade solvents were purchased from Fisher
Scientific (Raleigh, NC). ARP was purchased from Dojindo
Molecular Technologies (Gaithersburg, MD). trans-N-Deoxyri-
bosylase was a generous gift from F. Peter Guengerich (Vander-
3.83 (m, 1 H, H4′), 3.56-3.58 (m, 2 H, H5′,H5′′), 3.09 (m, 1 H,
ARP
ARP ARP
ARP
H5 ), 2.82 (dd, 1 H, Jgem ) 12.5, J7cis ,8 ) 2.8 Hz, H7trans ),
2.58 (m, 1 H, H2′′), 2.57 (bd, 1 H, Jgem ) 12.5 Hz, H7cisARP), 2.12
(bt, 2 H, JCH 11
1
5
ARP
ARP
ARP
bilt University, Nashville, TN). N-ammonium sulfate was
purchased from Spectra Stable Isotopes (Columbia, MD). Unless
stated otherwise, all other chemicals and enzymes were pur-
chased from Sigma-Aldrich Chemical Co. (St. Louis, MO) and
were used without further purification.
,
2
CH 12
) 6.5 Hz, CH
2
12 ), 1.62 (m, 1 H,
), 1.53 (m, 2 H, CH 11 ), 1.47 (m, 1 H,
), 1.33 (m, 2 H, CH 10 ) ppm. C (125 MHz,
2
2
ARP
ARP
CH
CH
2
9
9
diastereotopic
ARP
diastereotopic
2
ARP
13
2
ARP
ARP
ARP
6
DMSO-d ): 169.8 (C13 ), 167.0 (C16 ), 161.6 (C2 ), 152.1
(C8anti), 149.0 (C8syn), 148.6 (C3a), 137.3 (C2), 135.9 (C6syn), 134.8
Instrumentation. HPLC analysis to characterize the ARP-
(C6anti), 117.7 (C10a), 102.4 (C7anti), 98.7 (C7syn), 87.9 (C4′), 83.1
(C1′), 71.1 (C3′), 61.1 (C4ARP), 59.2 (C8 ), 55.5 (C5 ), 39.9
ARP
ARP
M
1
G-dR conjugate was performed using a Hewlett-Packard

Analysis of M
1
G-dR in DNA
Chem. Res. Toxicol., Vol. 18, No. 1, 2005 53
Figure 1. HPLC-UV analysis of the reaction mixture of M
1 1
G-dR and ARP. An equal molar ratio of ARP and M G-dR was incubated
in a weak acid solution. The resulting mixture was separated by the HPLC connected to a photodiode array detector.
(
2
C7ARP), 39.6 (C2′), 32.9 (C12ARP), 27.1 (C10ARP), 27.0 (C9ARP),
mg; KH
300 mg; Na-citrate‚2H
mg; FeCl ‚6H O, 25 mg; CuSO
0.127 mg; ZnCl , 0.0128 mg; and CoCl
into a final volume of 1 L]. E. coli (DH5R F′IQ, Invitrogen,
Carlsbad, CA) was inoculated by spatula in 1 mL of N-MEM
followed by culturing for 2 days at 37 °C in an environmental
shaker. One milliliter of fresh medium was mixed with 30 µL
of bacterial culture for subculturing. After two additional
subcultures, 100 µL of the bacterial culture was transferred to
2
PO
4
, 1.6 g; Na
2
HPO
O, 570 mg; CaCl
‚5H O, 0.24 mg; MnSO
‚H O, 0.494 mg dissolved
4
, 5.3 g; glycerol, 20 g; MgSO
4
‚7H
EDTA,
‚H O,
2
O,
ARP
4.3 (C11 ) ppm.
Blank DNA Preparation for M
2
2
, 0.566 mg; Na
2
1
G-dR Assay. High quality
3
2
4
2
4
2
genomic DNA was prepared from snap-frozen calf thymus
tissue (26). The frozen tissue sample (2 g) was homogenized in
2
2
2
1
5
5
0 mL of lysis buffer containing 10 mM HEPES (pH 7.4), 1 mM
EDTA, 1% SDS, and 2 mM 2,2,6,6-tetramethylpiperidine-1-oxyl
TEMPO). Proteinase K was added to the homogenate at a final
concentration of 8 units/mL and incubated in a cold room (4
C) overnight. Protein was precipitated by the addition of one-
(
°
third volume of protein precipitation solution (Promega, Madi-
son, WI) followed by centrifugation at 4000g for 20 min. The
nucleic acid was isolated from the supernatant by the addition
of an equal volume of 2-propanol. The precipitated nucleic acid
pellet was washed with 70% ethanol and reconstituted in 45
mL of 10 mM HEPES buffer (pH 7.4) with 0.2 mM TEMPO.
RNase A (36 KeU) and RNase T1 (12.5 kU) were added to the
nucleic acid solution, and the mixture was incubated for an hour
at 37 °C. DNA was precipitated by the sequential addition of
15
1
L of fresh N-MSM. After incubation for 3 days, the bacteria
were collected by centrifugation for 20 min at 6000g. Lysis buffer
40 mL) was added followed by homogenization for DNA
(
15
isolation. The N-labeled DNA was isolated from the bacteria
by the protocol used for CTD isolation. The recovery was 6 mg
based on UV absorbance measurement. To induce a high
15
number of M
1
G adducts in the DNA, N-labeled DNA or CTD
was treated with MDA for IS DNA or AS DNA, respectively.
MDA was prepared from tetramethoxypropane in 1 M HCl by
hydrolysis at 45 °C for 45 min. Three milligrams DNA was
incubated for 2 h in a cold room (4 °C) with 6 mL of an aqueous
solution containing 50 mM MDA. Unreacted MDA was removed
by two precipitations of the DNA with ethanol as described
above. Additionally, the DNA was washed with water several
times on a Centricon-100 microfilter (Millipore, Billerica, MA)
to remove any remaining MDA, salt, and small DNA fragments.
2
.25 mL of 4 M NaCl and 90 mL of cold ethanol. The DNA was
washed once with 90 mL of 70% ethanol solution. Following
removal of the ethanol solution by centrifugation, the precipi-
tated DNA was dried on ice for 15 min. The DNA pellet was
reconstituted in 30 mL of HPLC grade water. The DNA
concentration was determined using UV absorbance at 260 nm
in TE buffer (pH 8.0). The DNA yield was 46 mg.
To prepare M
thymus DNA (CTD) was reduced to 5,6-dihydro-M
5 mL of CTD solution (23 mg) was mixed with an equal volume
of an aqueous solution containing 10 mM hydroxylamine and
0 mM sodium cyanoborohydride. The reaction mixture was
1
G-dR blank DNA, endogenous M
1
G-dR in calf
To determine the M
was hydrolyzed in 150 µL of 10 mM HCl solution at 70 °C for
0 min. The depurinated DNA bases were separated by HPLC
on a Platinum EPS column (5 µm, 4.6 mm × 150 mm) and
monitored by UV. M G was monitored with fluorescence (Ext,
63 nm; Emt, 500 nm) by a LS-40 fluorescence detector (Applied
Biosystems, Foster City, CA), and the M G concentration was
1
G number in AS or IS DNA, 30 µg of DNA
1
G-dR. Briefly,
1
3
2
1
incubated for an hour at 45 °C with mild agitation. After the
reaction mixture cooled to room temperature, the DNA was
precipitated by adding NaCl and ethanol as described above.
The DNA pellet was washed with 70% ethanol and reconstituted
in 45 mL of water. To remove any remaining reducing reagent,
DNA was precipitated once again using ethanol. Following
washing with 70% ethanol to remove NaCl, the DNA pellet was
dried briefly on ice and reconstituted in 15 mL of water followed
by storage at -80 °C in small aliquots (500 µL) until it was ready
to be used.
3
1
calculated based on a standard calibration curve. Total DNA
concentration was calculated based on Gua and Ade measured
by UV absorbance at 254 nm after separation by HPLC to give
1
the ratio of M G-dR to nucleotides in the standard DNA.
Bleomycin Treatment of DNA. Aqueous solutions contain-
ing different amounts of bleomycin with the same molar ratio
of ferrous ammonium sulfate were added to a CTD solution
(final concentration, 0.3 mg/mL) in 50 mM potassium phosphate
buffer (pH 7.4). The mixture was incubated for 2 h at 37 °C
with mild agitation. The incubation was terminated by transfer-
ring the sample to ice. DNA was precipitated with ethanol and
then washed extensively with water on Centricon-30 microfilters
following ethanol precipitation.
M
1
G-dR Assay Standard and IS DNA Preparation.
1
5
Stable isotope-labeled DNA containing
pared from Escherichia coli cultured in minimum salt medium
5 1
N -M G-dR was pre-
1
5
(
MSM) containing ( NH
)
4 2
4
SO as a single nitrogen source for
1
5
bacterial growth. N-MSM was prepared according to a pub-
15
4 2 4
lished procedure (27) with minor modifications [( NH ) SO , 700

54
Chem. Res. Toxicol., Vol. 18, No. 1, 2005
Jeong et al.
Analysis of M G-dR in DNA. Sample or blank DNA (200
1
µg) with a known amount of M
1
G-dR was mixed with IS DNA
-M G-dR. The DNA mix was
corresponding to 400 fmol of 5N
1
5
1
incubated with 2 mM ARP in 0.1% acetic acid solution for 30
min at 50 °C. Aliquots (10 µL) were subjected to mild acid
hydrolysis followed by HPLC-UV measurements of purine base
to determine DNA concentration in the sample mix. Unbound
ARP in the sample solution was removed by ethanol precipita-
tion of DNA followed by reconstitution in 450 µL of 20 mM
2
HEPES buffer (pH 7.8) containing 1 mM MgCl . The DNA was
hydrolyzed to nucleosides by DNase I treatment at a final
concentration of 100 U/mL for 20 min at 37 °C, followed by the
addition of a mixture containing alkaline phosphatase, phos-
phodiesterase I, and phosphodiesterase II to final concentrations
of 5, 0.06, and 0.1 U/mL, respectively. The DNA and enzyme
mixture was further incubated at 37 °C for 1 h with mild
agitation. After the removal of the enzymes by microfiltration,
the filtrate was placed on an solid phase extraction (SPE)
column (HLB-60, Waters, Milford, MA) that had been preacti-
vated with methanol and equilibrated with water. The SPE
column was washed once with 1 mL of water and twice with 3
1
mL of 20% methanol. ARP-M G-dR and the IS were eluted with
3
mL of 50% methanol in water. The solvent was removed by
vacuum evaporation. The samples were reconstituted in 50 µL
of 10% methanol in water and stored at -80 °C until analysis.
+
LC-ESI -MS/MS Analysis. ARP-M
1
G-dR and IS were sepa-
rated on a Platinum EPS C18 column (3 µm, 2 mm × 150 mm)
at a flow rate of 0.2 mL/min. After sample injection, the column
was washed with 5% acetonitrile in 8.3 mM acetic acid for 2
min. The acetonitrile was gradually increased to 50% over 10
min, kept at 50% for 3 min, followed by reequilibration to initial
conditions. The LC eluant was delivered to waste for the initial
Figure 2. ESI mass spectrum for ARP-M
1
G-dR conjugate 2.
(
A) Full scan ESI mass spectrum for conjugate 2. (B) Tandem
+
mass spectrum for parent molecular ion of ARP-M
1
G-dR + H
(
m/z ) 635).
1
from direct reaction of M G with ARP (data not shown).
7
min, and then, the flow was introduced to the mass spectrom-
eter for 2 min followed by diversion to waste until the end of
analysis. The analysis was performed by positive ion electro-
spray using selective reaction monitoring (SRM). Nitrogen gas
was used for the sheath and auxiliary gas. The needle spray
voltage was maintained at 4.5 kV. The ion skimmer was placed
at a 90° angle relative to the spray needle, which was main-
tained at a voltage of 70 V. The capillary temperature was 350
Peak 1 accounted for <5% of the total products, estimated
from peak integrals of the HPLC trace recorded at 254
nm. Peaks 2 and 3 gave similar full-scan ESI-MS spectra,
with ions at m/z 635, corresponding to pseudo molecular
+
ions [M + H] and m/z 657, corresponding to sodium
+
adducts [M + Na] (Figure 2A). Peak 2 was major; the
ratio of peaks 2/3 was 95:5 (Figure 1). MS/MS of the
pseudo molecular ions induced by low collision energy is
°
C. The first and third quadrupole mass analyzers were used
for unit mass resolution, and the second quadrupole was used
for ion fragmentation. Fragmentation was accomplished at a
collision energy of 48 V using argon gas at a pressure of 1.5
mTorr. Two ion transitions, m/z 635f m/z 188 and m/z
consistent with the glycosylated conjugate structure ARP-
+
M
1
G-dR, yielding daughter ions at m/z 188 [M
1
G + H] ,
+
+
332 [ARP + H] , and 519 [ARP-M
Five structures can be considered for ARP-M
1
G + H] (Figure 2B).
G-dR con-
6
40f m/z 193, were used to monitor ARP-M
1
G-dR and
1
1
5
ARP- N
5 1
-M G-dR.
jugates, with equilibration possible between all struc-
tures. In addition to a propeno ring-closed form, four ring-
opened isomers are possible as a result of cis/trans isom-
erism about the C6-C7 double bond and syn/anti con-
Results and Discussion
G-dR Conjugates. Analysis of the reaction
ARP-M
1
figurations of the aldoxime. In contrast to M
1
G-dR, the
between M G-dR and ARP by HPLC indicated three
product peaks in addition to unreacted M G-dR. UV
1
conjugates ARP-M G-dR are expected to be present in
1
1
predominantly ring-opened forms in aqueous solution.
Peaks 2 and 3 were collected and reanalyzed by HPLC
and LC-MS. Each peak yielded UV and LC-MS traces in
which peaks 2 and 3 appeared in the same ratio as in
the initial trace, suggesting equilibration between prod-
ucts that was slow on the chromatographic time scale.
Peak 2 was collected by semipreparative HPLC and char-
spectra of the HPLC peaks were qualitatively similar,
with λmax in the range of 310-320 nm characteristic of
aldoximes (Figure 1). This result is consistent with the
report of Schnetz-Boutaud et al. (17) who characterized
conjugates of hydroxylamine and methoxyamine with
M
1
G-dR. On the basis of HPLC retention times, the
products contained in the three peaks were significantly
more hydrophobic than M G-dR, supporting the forma-
tion of conjugates with ARP. By full-scan ESI-MS, peak
showed ions at m/z 519 and 541, corresponding in mass-
1
1
acterized by one-dimensional (1D) H NMR, H correlated
1
1
spectroscopy (COSY), H nuclear overhauser effect spec-
1
13
troscopy (NOESY), and H- C heteronuclear shift cor-
relation experiments since rigorous structural assign-
ment of the analyte standard is imperative. The presence
1
+
to-charge ratio to the pseudo molecular ion [M + H] of
a deglycosylated conjugate ARP-M G and its sodium
1
1
of two compounds was established by the 1D H NMR
adduct, respectively. MS/MS of the pseudo molecular ion
spectrum. A single set of sugar and biotin signals was
apparent. In addition, two sets of signals were resolved,
each present in a ratio of 1:2 relative to biotin and sugar
based on the signal integrations. Hyperfine shift patterns
and multiplicities of signals within these two sets were
is consistent with an aglycone, yielding significant daugh-
+
ter ions at m/z 188 and 332 corresponding to [M G + H]
1
and [ARP + H]+ (data not shown). Peak 1 was also
identical in UV and mass spectra to the adduct formed

Analysis of M
1
G-dR in DNA
Chem. Res. Toxicol., Vol. 18, No. 1, 2005 55
Figure 3. 1H COSY expansion showing the NH5-H6-H7-H8 connectivities of the syn (s) and anti (- - -) isomers of the ARP-M
conjugate.
1
G-dR
similar to those reported for H6, H7, and H8 of the
mixtures of anti/syn aldoximes derived from the reaction
of M G-dR with hydroxylamine and methoxyamine (17).
1
On the basis of this observation and the UV spectrum of
HPLC peak 2 characteristic of aldoxime derivatives of
M
1
G, the 1D NMR spectrum was tentatively attributed
to a 1:1 mixture of anti/syn ARP-M G-dR conjugates, in
1
which sugar and biotin signals of the isomers were not
resolved. The two-dimensional (2D) NMR experiments
confirmed the connectivities expected for the side chain
C6-C7-C8 of the ring-opened mixture of aldoximes and
connectivity of the sugar and biotin signals with the sets
1
of resonances assigned to the isomers. The H COSY
spectrum (Figure 3; Supporting Information, Figure S1)
shows connectivities between the proton resonance at
5
.82 ppm and signals at 7.52 and 7.98 ppm, consistent
with the assignments H7, 5.82; H6, 7.52; and H8 7.98
ppm. This pattern of shifts is identical to that assigned
to the anti aldoximes by Schnetz-Boutaud et al. (17). A
similar pattern of connectivities between the signal 6.33
ppm and signals at 7.60 and 7.28 ppm (Figure 3) allows
the assignment H7, 6.33; H6, 7.60; and H8 7.28 ppm for
the syn aldoxime. Further confirmation of the assign-
ments is provided by the heteronuclear multiple bond
correlation (HMBC) spectrum (Supporting Information,
1
1
Figure 4. Expansion of the H NOESY spectrum of ARP-M G-
dR in the region of 6.6-10.2 ppm showing interaction between
H6,H8 and NH5,H6 for anti and syn isomers. Connectivities
are syn (s) and anti (- - -).
S4). 13C signals of the carbons R to anti aldoximes are
uniformly observed to be shifted downfield with respect
to the R carbon signals of the corresponding syn isomers
(28). Thus, the C hyperfine shifts of the C7 signals
1
confirm the anti/syn assignments for the ARP-M G-dR
conjugates and support configurations assigned to the
oxime derivatives reported in (17) for which no rationale
was provided. Coupling constants J6,7 ) 13.9 and 14.1
Hz determined for the vinylic protons of anti and syn
isomers, respectively, are in a range that is not definitive
for a cis or trans configuration (29). However, strong
cross-peaks between proton pairs H6,H8 and NH5,H7 of
2
13
Figure S2), which shows J(C-H) coupling between H8
and C6 and C7 for both isomers. Additional connectivity
in the COSY spectrum was observed between a doublet
at 9.49 ppm and H6 of the anti isomer and a doublet at
9
.61 ppm and H6 of the syn isomer (Figure 3). The
absence of the low-field doublets in the heteronuclear
single quantum coherence (HSQC) spectrum (Figure S3)
allows the assignment of these signals to NH5 of the side
1
3
chain. The C resonances of C7anti and C7syn appear at
02.4 and 98.7 ppm, respectively, in the HSQC spectrum,
1
1
both anti and syn isomers observed in the H NOESY
confirmed by cross-peaks with H6anti/syn and H8anti/syn in
the HMBC experiment (Suppporting Information, Figure
spectrum (Figure 4) are possible only in the trans
configuration. Hence, the conjugates can be assigned syn,

56
Chem. Res. Toxicol., Vol. 18, No. 1, 2005
Scheme 1. Formation of M
Jeong et al.
1
G-dR and Its Conjugate to ARP
trans and anti, trans geometries. 3J(C-H) coupling
between H1′ and C2 and C3a establishes connectivity
between the deoxyribose and the M G moiety. The 2D
1
spectra do not establish connectivity between H2 or C2
and any of the signals that can definitively be assigned
to the anti or syn isomer. Consequently, it is not possible
to assign the H2 signals to a specific isomer. Singlets at
coupled to C16ARP, are assigned as NH15syn/antiARP. Two
additional resonances at 11.16 and 11.05 ppm appear in
a region characteristic of the amido signal of guanine and
are assigned as NH9syn/anti. Consistent with identifica-
ARP
ARP
tion as NH signals, NH14syn/anti
, NH15syn/anti
, and
NH9syn/anti are absent from the HSQC spectrum. Because
none of the resolved anti and syn ARP signals show
connectivity with atoms of the NH5-C6-C7-C8 sequence
4
1
2
.43 and 4.47 ppm are assigned to ARP methylene CH -
ARP
7
by comparison with the hyperfine shift of uncon-
1
of the ring-opened M G moiety, they cannot be identified
2
ARP
1
13
jugated ARP and J(C-H) coupling with carbonyl C16
with a specific syn or anti isomer. The H and C signals
of the minor component(s) in peak 2 were obscured by
the signals of the major conjugates, and no structural
information could be derived. Peak 3 was not collected
specifically, as the rechromatography experiment deter-
mined that the collection of peak 3 would give a mixture
of anti/syn isomers and the minor product in the same
ratio as peak 2.
in the HMBC spectrum. Absence of a cross-peak between
the two methylene proton signals in the COSY spectrum
confirms that they are independent singlets resolved for
the isomeric conjugates, rather than diastereotopic gemi-
nal methylene protons, and that an ARP moiety corre-
sponds to each anti and syn conjugate. Of the remaining
1
signals resolved for the anti and syn isomers, broad H
singlets at 9.73, 9.74, and 9.77 ppm are identified as the
Mass Spectral Analysis of ARP-M
gates. An LC-ESI-MS/MS method was utilized to mea-
sure ARP-M G-dR because of the high sensitivity and
specificity of this technique. Two transitions were com-
1
G-dR Conju-
ARP
ARP
2
carbazido protons NH14 ,NH15
by J(C,H) coupling
ARP
with the carbonyl carbons at C16
(167.0 ppm) and
1
ARP
ARP
C13
(169.8). The assignment of C16
is based on
2
ARP
+
cross-peaks with ARP methylene protons CH
17 , and
is identified by cross-peaks with ARP linker
pared by SRM, 635 f 188 (ARP-M
H ) and 635 f 332 (ARP-M G-dR + H f ARP + H ),
1
1
G-dR + H f M G +
1
ARP
+
+
+
C13
ARP
ARP
protons CH
2
12
and CH
2
11 . Overlapping reso-
in terms of sensitivity, signal consistency, and interfer-
ence from DNA matrix. Both transitions showed excellent
sensitivity for the detection of M G-dR, with the first
1
nances at 9.74 ppm, coupled to C13, are assigned as
ARP
NH14syn/anti
, while resonances at 9.73 and 9.77 ppm,

Analysis of M
1
G-dR in DNA
Chem. Res. Toxicol., Vol. 18, No. 1, 2005 57
providing a more consistent signal and less interference
from sample matrix. The lowest quantity of ARP-M G-
dR detectible was between 200 and 500 attomoles on
column. Consistent quantification for ARP-M G-dR was
achieved for as little as 2 fmol, which exhibited less than
0% interinjection variation between six injections. When
comparing the level of detection for M G-dR and M
with that of ARP-M G-dR, there was an improvement in
sensitivity by 10- and 100-fold, respectively. This superior
sensitivity makes it possible to detect one M
G-dR/108
nucleotides (n.t.) using as little as 100 µg of DNA.
Because the mass of the ARP-M G-dR conjugate is nearly
twice that of ARP or M G-dR, there was a significant
Scheme 2. M
1
G-dR Assay Protocol from Sample
DNA
1
1
1
1
1
G
1
1
1
1
reduction in background noise from low molecular weight
contaminants. Lower noise levels in the chromatogram
should result in better resolution during the mass
spectral analysis of biological samples, as well as better
detection sensitivity.
In addition to improving detection sensitivity, conjuga-
1
tion with ARP increased the efficiency of ARP-M G-dR
enrichment from DNA samples. Labeling with ARP
significantly enhanced the recovery of the conjugate when
using SPE. No significant loss of the conjugate was
observed when using up to 20% methanol mobile phase,
which elutes almost all unmodified nucleosides as well
as unreacted ARP. In addition, the biotin moiety of ARP
allows for additional sample cleanup approaches such as
affinity purification with commercially available avidin
or an anti-biotin antibody. However, with the present
ing other DNA nucleotides. The blank DNA was washed
extensively to remove hydroxylamine and the reducing
reagent, which could hinder the reaction of ARP with
1
approach of measuring M G-dR after ARP labeling, no
additional cleanup beyond SPE was necessary. Because
of the extensive cleanup with SPE and the large molec-
ular mass of the conjugate, no significant interference
M
1
G-dR. UV spectral analysis for the blank DNA or
HPLC-UV analysis after mild acid hydrolysis showed
no difference between CTD and M G-dR assay blank
DNA. ARP-M G-dR in the CTD was analyzed with the
standard M G-dR assay protocol as described in Scheme
2. No detectable ARP-M G-dR signal was observed in
blank DNA as compared to the CTD, which contained
1
from the DNA sample matrix was observed for ARP-M
dR measurement.
1
G-
1
1
Assay Standard, Blank, and IS DNA Synthesis. To
improve the accuracy of the assay, ARP was applied to
1
8
DNA in order to label M
than labeling the free M G-dR nucleoside after enzymatic
hydrolysis of DNA. The advantage in DNA labeling is
that the M G-dR in the DNA has an unstable aldehydic
1
G-dR within the DNA, rather
1
approximately two M G-dR/10 n.t. (Figure 5). There was
no significant difference in the IS response between CTD
and blank DNA sample. These results confirmed that
1
1
most of the M G-dR in the CTD was selectively reduced
1
moiety that can react with proteins or chemicals to
decrease the level of DNA adducts. Additionally, labeling
DNA with ARP is preferable over labeling free nucleo-
sides because unbound ARP is conveniently removed by
washing DNA with ethanol and water. To reflect all of
the artifacts arising during ARP labeling and DNA
without any effect on the other nucleotides. It is critical
to use a valid blank DNA with an AS for DNA adduct
analysis since the concentration of DNA can be an
important variable affecting the efficiency of ARP label-
ing, sample DNA recovery, and hydrolysis by enzymes.
It is intuitive that proper compensation for DNA quantity
with a valid blank DNA should result in more accurate
measurements of biological DNA samples. The AS DNA
1
hydrolysis, valid M G-dR blank, AS, and IS DNA were
introduced. M G-dR assay blank DNA was prepared from
1
high quality DNA isolated from calf thymus tissue.
Because of continuous production of ROS resulting in
lipid peroxidation and DNA damage, there is endogenous
for M
treating it with MDA at physiological pH. To measure
G-dR in DNA, the DNA was hydrolyzed in 10 mM HCl
solution to depurinate M G, guanine, and adenine. Depu-
rinated M G and purine bases were separated by HPLC
1
G-dR assay was also prepared from CTD by
M
1
M
1
G-dR in most DNA isolated from tissue, including calf
thymus tissue. Endogenous M G-dR in the CTD was
reduced to 5,6-dihydro-M G-dR, which cannot react with
ARP. Goda and Marnett (30) demonstrated a reduction
of M G-dR using NaBH . However, CTD treated with
NaBH differed considerably from intact CTD in its UV
spectrum and water solubility probably due to indiscrimi-
nate reduction. To selectively reduce M G-dR without
affecting other nucleotides in DNA, milder reduction with
NaCNBH and hydroxylamine was carried out. While the
1
1
1
and quantified with a fluorescence detector and UV
detector, respectively. Fluorescence detection provided
1
accurate measurement for M
greater sensitivity than UV detection since interference
from other DNA bases was insignificant. M G AS DNA
contained 5 pmol M G-dR/µg DNA, which is equivalent
to 162 ng of DNA for 810 fmol of M G-dR standard, the
1
G with at least 10 times
1
4
4
1
1
1
1
highest level standard. The quantity of DNA from
standard DNA should not affect the analysis because it
is less than 0.05% of the sample DNA.
3
role of hydroxylamine is at the present time unclear, the
reduction does not proceed in its absence. However, the
inclusion of hydroxylamine in the reaction mixture
Synthesis of an IS for the assay posed a challenge since
1
resulted in selective reduction of M G-dR without affect-
stable isotope-labeled DNA cannot readily be prepared

58
Chem. Res. Toxicol., Vol. 18, No. 1, 2005
Jeong et al.
Figure 6. (A) Calibration curve for the M
1
G-dR assay and (B)
1
M G formation from bleomycin treatment. (A) From 0 to 810
fmol of AS DNA (three samples for each level) was mixed with
00 µg of blank DNA and 400 fmol of IS DNA followed by
processing with M G-dR assay protocol. The response was
expressed as the area ratio for ARP-M G-dR and its IS. (B) CTD
2
1
1
was treated with different amounts of bleomycin under physi-
ological conditions. Each point represents the mean ( SD from
four samples for each group.
15N IS. Production can be completed from a simple 2 or
3 day bacterial culture with subsequent DNA isolation
without any chemical or enzymatic synthesis. Such N-
1
Figure 5. Comparison of the M G-dR level between (A) 2 pmol
of IS DNA with 200 µg of blank DNA and (B) 200 µg of CTD
with 400 fmol of IS DNA. Samples were processed according to
15
1
the protocol described in the Materials and Methods. ARP-M G-
labeled DNA should be suitable for many other DNA
adduct assays, because the IS DNA will reflect any
artifacts produced during analysis, as commonly occurs
in sample DNA. In addition, the stable isotope-labeled
DNA is useful to investigate artifacts arising during the
DNA isolation/hydrolysis process. Artifactual formation
of DNA adduct during DNA isolation can be investigated
by comparisons of adduct levels between intact labeled
DNA and labeled DNA that has been processed by DNA
isolation from tissue samples. The differences in the
oxidative stress marker between intact labeled DNA and
labeled DNA after DNA isolation will reflect artifactual
formation of the DNA adduct.
1
5
5 1
dR and ARP- N -M G-dR were analyzed with SRM to investi-
gate the presence of AS in the blank DNA.
by chemical synthesis or molecular biological techniques.
To produce stable isotope-labeled DNA, E. coli was
utilized because it can produce all essential amino acids
and DNA molecules from a single nitrogen source. This
technique has been used to produce 15N- or C-labeled
proteins and RNA for chemical structural analysis (27,
13
3
1, 32). However, this methodology had not been applied
to the preparation of labeled dsDNA for use as an IS DNA
in mass spectral analysis. To ensure the production of
pure N-labeled E. coli, the bacteria were subcultured
1
5
15
three times using fresh N-containing medium. One liter
of the bacterial culture yielded about 6 mg of high grade
Assay Validation. The assay validation was ac-
complished by establishing assay sensitivity and accuracy
using ARP-M
1
G-dR standard and IS, AS, and blank DNA.
AS and blank DNA were analyzed with IS DNA to test
the assay sensitivity. As shown in Figure 6A, a linear
1
5
N-labeled DNA, which is sufficient to process up to
5000 samples. The labeled DNA was treated with MDA
1
1
to induce a high level of M G-dR (3.5 pmol/µg DNA). To
confirm the purity of the IS DNA, IS DNA containing 2
1
response was observed for AS DNA containing M G-dR
1
5
pmol of
lyzed for LC-MS/MS analysis. No detectable signal for
unlabeled ARP-M G-dR AS was observed in the chro-
matogram as shown in Figure 6A. On the basis of 98%
N
5
-M
1
G-dR was reacted with ARP and hydro-
in the range of 0-810 fmol in the presence of 400 fmol of
IS DNA and 200 µg of blank DNA. The slope constant
for the regression line (0.00255 ( 0.00004, n ) 4) was
very close to the theoretical value for 400 fmol of IS DNA
per standard sample. Because of the low level of noise
from the mass spectral analysis, the y-intercept of the
calibration curve was close to 0. The variation for the
lowest standard, 3 fmol, was 11%, which is acceptable
1
1
5
15
14
4 4
N purity in NH SO , the presence of N-purines
15
should account for less than 10 ppb guanine in the N-
labeled DNA. The purity of labeled DNA is far better than
that of chemically synthesized IS bases or nucleosides.
Besides, preparation of 15N IS DNA from labeled E. coli
is a time and cost efficient way to produce high quality
1
for the limit of detection. The 10 fmol M G-dR standard
showed only 4% variation between three samples. Varia-

Analysis of M
1
G-dR in DNA
Chem. Res. Toxicol., Vol. 18, No. 1, 2005 59
tion between four different days for 10 fmol standards
was less than 5%. The recovery ratio (detected M G-dR-
ARP/M G-dR in AS DNA) for 10 fmol standards was 69.3
6.5% from four samples. The concentration of M G-dR
in commercially available DNA was investigated using
CTD purchased from Sigma-Aldrich Co. The result was
References
1
(
(
1) Helbock, H. J., Beckman, K. B., Shigenaga, M. K., Walter, P. B.,
Woodall, A. A., Yeo, H. C., and Ames, B. N. (1998) DNA oxidation
matters: The HPLC-electrochemical detection assay of 8-oxo-
deoxyguanosine and 8-oxo-guanine. Proc. Natl. Acad. Sci. U.S.A.
1
(
1
9
5, 288-293.
2) Nakamura, J., and Swenberg, J. A. (1999) Endogenous apurinic/
apyrimidinic sites in genomic DNA of mammalian tissues. Cancer
Res. 59, 2522-2526.
(3) Bartsch, H., and Nair, J. (2000) Ultrasensitive and specific
detection methods for exocylic DNA adducts: Markers for lipid
peroxidation and oxidative stress. Toxicology 153, 105-114.
1
compared to previous reports showing different M G-dR
levels ranging from nondetectable (detection limit one
8
8
M
1
G-dR/10 n.t.) to 15 M
1
G-dR/10 n.t. We found 2.5 (
8
0
.9 M
1
G-dR/10 n.t. from four samples, a value within
the reported range.
The new assay was applied to assess the effects of
bleomycin on M G formation in the DNA. Dedon et al.
6) demonstrated the indirect formation of M G from base
(
(
(
4) Morinello, E. J., Ham, A. J., Ranasinghe, A., Nakamura, J., Upton,
P. B., and Swenberg, J. A. (2002) Molecular dosimetry and repair
2
of N ,3-ethenoguanine in rats exposed to vinyl chloride. Cancer
1
Res. 62, 5189-5195.
(
1
5) Ham, A. J., Ranasinghe, A., Koc, H., and Swenberg, J. A. (2000)
propenal produced from deoxyribose damage by a bleo-
mycin radical. Because base propenal is more reactive
than MDA toward guanine, 20 µM bleomycin was re-
4-Hydroxy-2-nonenal and ethyl linoleate form N ,3-ethenogua-
2
nine under peroxidizing conditions. Chem. Res. Toxicol. 13, 1243-
250.
6) Dedon, P. C., Plastaras, J. P., Rouzer, C. A., and Marnett, L. J.
1
ported to give an 11-fold increase of M
The same treatment resulted in a 13-fold increase in M
in the DNA as compared to a control DNA sample
1
G in the DNA.
(
1998) Indirect mutagenesis by oxidative DNA damage: Forma-
1
G
tion of the pyrimidopurinone adduct of deoxyguanosine by base
propenal. Proc. Natl. Acad. Sci. U.S.A. 95, 11113-11116.
8
(7) Maas, R. L., Ingram, C. D., Porter, A. T., Oates, J. A., Taber, D.
F., and Brash, A. R. (1985) Investigation of the chemical conver-
sion of hydroperoxyeicosatetraenoate to leukotriene epoxide using
stereospecifically labeled arachidonic acid. Comparison with the
enzymatic reaction. J. Biol. Chem. 260, 4217-4228.
8) Pryor, W. A., and Porter, N. A. (1990) Suggested mechanisms for
the production of 4-hydroxy-2-nonenal from the autoxidation of
polyunsaturated fatty acids. Free Radical Biol. Med. 8, 541-543.
9) Schauenstein, E., and Esterbauer, H. (1978) Formation and
properties of reactive aldehydes. Ciba Found. Symp. 67, 225-
containing 3.0 ( 1.1 M
dependent induction of M
range of 5-100 µM bleomycin in the reaction mixture
Figure 6B). The adduct levels as a result of high
1
G-dR/10 n.t. A concentration-
1
G-dR was observed in the
(
(
bleomycin exposure were similar to those of Ottender et
al. (33). However, we were not able to compare results
for control or low levels of bleomycin exposure due to the
limited sensitivity of the immunoslot blot method used
by Ottender et al.
(
2
44.
(
10) Marnett, L. J., and Burcham, P. C. (1993) Endogenous DNA
adducts: Potential and paradox. Chem. Res. Toxicol. 6, 771-785.
Conclusions
(11) Seto, H., Akiyama, K., Okuda, T., Hashimoto, T., Takesue, T.,
and Ikemura, T. (1981) Structure of a new modified nucleoside
formed by guanosine- malonaldehyde reaction. Chem. Lett. 707-
The new methodology using ARP to selectively label
G-dR significantly improves the measurement of this
DNA adduct. This approach effectively stabilizes M G-
7
08.
M
1
(
12) Rouzer, C. A., Chaudhary, A. K., Nokubo, M., Ferguson, D. M.,
Reddy, G. R., Blair, I. A., and Marnett, L. J. (1997) Analysis of
the malondialdehyde-2′-deoxyguanosine adduct pyrimidopurinone
in human leukocyte DNA by gas chromatography electron capture
negative chemical ionization mass spectrometry. Chem. Res.
Toxicol. 10, 181-188.
1
dR in the DNA, enhances detection sensitivity, and
facilitates the separation of the ARP conjugate from other
nucleosides. Simple sample cleanup with SPE reduces
the time and labor for sample analysis and also obviates
the need for a specific antibody for immunoaffinity
purification. This ARP labeling method should be ap-
plicable not only to other ARP reactive aldehydic DNA
adducts but also simultaneously to multiple DNA adducts
with minor modification of instrumentation. Use of an
AS DNA and IS DNA with a valid blank DNA signifi-
cantly improves the assay accuracy and consistency. 15N-
labeled DNA from E. coli is a convenient and cost-
effective source of high quality 15N IS DNA. This DNA
has the potential for use in investigating the contribution
of artifacts in the measurement of DNA biomarkers for
oxidative stress. Overall, with the use of a valid standard,
this new assay should be a useful tool for measuring DNA
biomarkers of oxidative damage caused by various en-
dogenous and exogenous ROS inducing chemicals.
(
13) Chaudhary, A. K., Nokubo, M., Reddy, G. R., Yeola, S. N., Morrow,
J. D., Blair, I. A., and Marnett, L. J. (1994) Detection of
endogenous malondialdehyde-deoxyguanosine adducts in human
liver. Science 265, 1580-1582.
(
14) Kadlubar, F. F., Anderson, K. E., Haussermann, S., Lang, N. P.,
Barone, G. W., Thompson, P. A., MacLeod, S. L., Chou, M. W.,
Mikhailova, M., Plastaras, J., Marnett, L. J., Nair, J., Velic, I.,
and Bartsch, H. (1998) Comparison of DNA adduct levels associ-
ated with oxidative stress in human pancreas. Mutat. Res. 405,
1
25-133.
(
15) Mao, H., Schnetz-Boutaud, N. C., Weisenseel, J. P., Marnett, L.
J., and Stone, M. P. (1999) Duplex DNA catalyzes the chemical
rearrangement of a malondialdehyde deoxyguanosine adduct.
Proc. Natl. Acad. Sci. U.S.A. 96, 6615-6620.
(
16) Niedernhofer, L. J., Riley, M., Schnetz-Boutaud, N., Sanduwaran,
G., Chaudhary, A. K., Reddy, G. R., and Marnett, L. J. (1997)
Temperature-dependent formation of a conjugate between tris-
(hydroxymethyl)aminomethane buffer and the malondialdehyde-
DNA adduct pyrimidopurinone. Chem. Res. Toxicol. 10, 556-561.
17) Schnetz-Boutaud, N., Daniels, J. S., Hashim, M. F., Scholl, P.,
Burrus, T., and Marnett, L. J. (2000) Pyrimido[1,2-R]purin-10-
(
(
(
3H)one: A reactive electrophile in the genome. Chem. Res.
Acknowledgment. This research was supported in
part by the NIEHS Superfund Basic Research Program
P42-ES05948 and NIH Grants R42-ES11746, P30-
ES10126, and P30-CA16086.
Toxicol. 13, 967-970.
18) Nakamura, J., Walker, V. E., Upton, P. B., Chiang, S. Y., Kow,
Y. W., and Swenberg, J. A. (1998) Highly sensitive apurinic/
apyrimidinic site assay can detect spontaneous and chemically
induced depurination under physiological conditions. Cancer Res.
5
8, 222-225.
Supporting Information Available: 1H NMR of syn- and
(
19) Nakamura, J., Purvis, E. R., and Swenberg, J. A. (2003) Micro-
molar concentrations of hydrogen peroxide induce oxidative DNA
lesions more efficiently than millimolar concentrations in mam-
malian cells. Nucleic Acids Res. 31, 1790-1795.
anti-ARP-M
ARP-M
dR conjugates, and HMBC spectrum of ARP-M
1
G-dR conjugate mixture, full COSY spectrum of
1
G-dR conjugate mixture, HSQC spectrum of ARP-M
1
G-
G-dR. This
1
material is available free of charge via the Internet at http://
pubs.acs.org.
(20) Ham, A. J., Ranasinghe, A., Morinello, E. J., Nakamura, J., Upton,
P. B., Johnson, F., and Swenberg, J. A. (1999) Immunoaffinity/

60
Chem. Res. Toxicol., Vol. 18, No. 1, 2005
Jeong et al.
gas chromatography/high-resolution mass spectrometry method
(27) Batey, R. T., Battiste, J. L., and Williamson, J. R. (1995)
Preparation of isotopically enriched RNAs for heteronuclear NMR.
Methods Enzymol. 261, 300-322.
2
for the detection of N ,3-ethenoguanine. Chem. Res. Toxicol. 12,
1240-1246.
(
(
21) Koc, H., and Swenberg, J. A. (2002) Applications of mass
spectrometry for quantitation of DNA adducts. J. Chromatogr.
B: Biomed. Sci. Appl. 778, 323-343.
(
(
(
28) Hawkes, G. E., Herwig, K., and Robertson, L. W. (1974) Nuclear
13
magnetic resonance spectroscopy. Use of C spectra to establish
configurations of oximes. J. Org. Chem. 39, 1017-1028.
22) Yen, T. Y., Holt, S., Sangaiah, R., Gold, A., and Swenberg, J. A.
29) Pavia, D. L., Lampman, G. M., and Kriz, G. S. (2001) Introduction
to Spectroscopy. A Guide for Students of Organic Chemistry, 3
ed., Harcourt College Publishers, New York.
30) Goda, Y., and Marnett, L. J. (1991) High-performance liquid
chromatography with electrochemical detection for determination
of the major malondialdehyde-guanine adduct. Chem. Res. Toxi-
col. 4, 520-524.
6
(
1998) Quantitation of 1,N -ethenoadenine in rat urine by im-
munoaffinity extraction combined with liquid chromatography/
electrospray ionization mass spectrometry. Chem. Res. Toxicol.
1
1, 810-815.
(
23) Yen, T. Y., Christova-Gueoguieva, N. I., Scheller, N., Holt, S.,
Swenberg, J. A., and Charles, M. J. (1996) Quantitative analysis
2
of the DNA adduct N ,3-ethenoguanine using liquid chromatog-
(
31) Nikonowicz, E. P., and Pardi, A. (1992) Three-dimensional
raphy/electrospray ionization mass spectrometry. J. Mass Spec-
trom. 31, 1271-1276.
heteronuclear NMR studies of RNA. Nature 355, 184-186.
(
(
(
24) Fedtke, N., Boucheron, J. A., Turner, M. J., Jr., and Swenberg,
J. A. (1990) Vinyl chloride-induced DNA adducts. I: Quantitative
(32) Batey, R. T., Inada, M., Kujawinski, E., Puglisi, J. D., and
Williamson, J. R. (1992) Preparation of isotopically labeled
ribonucleotides for multidimensional NMR spectroscopy of RNA.
Nucleic Acids Res. 20, 4515-4523.
(33) Otteneder, M., Plastaras, J. P., and Marnett, L. J. (2002) Reaction
of malondialdehyde-DNA adducts with hydrazines-development
of a facile assay for quantification of malondialdehyde equivalents
in DNA. Chem. Res. Toxicol. 15, 312-318.
2
determination of N ,3-ethenoguanine based on electrophore label-
ing. Carcinogenesis 11, 1279-1285.
25) Schnetz-Boutaud, N. C., Mao, H., Stone, M. P., and Marnett, L.
J. (2000) Synthesis of oligonucleotides containing the alkali-labile
pyrimidopurinone adduct, M(1)G. Chem. Res. Toxicol. 13, 90-
95.
26) Lin, P. H., Nakamura, J., Yamaguchi, S., Asakura, S., and
Swenberg, J. A. (2003) Aldehydic DNA lesions induced by catechol
estrogens in calf thymus DNA. Carcinogenesis 24, 1133-1141.
TX049853L