Measurement of urinary excretion of 5-hydroxymethyluracil in human by GC/NICI/MS: Correlation with cigarette smoking, urinary TBARS and etheno DNA adduct

Article abstract of DOI:10.1016/j.toxlet.2004.11.009

5-Hydroxymethyluracil (5-HMU) is derived from radiation in addition to endogenous oxidative DNA damage and it is one of the most abundant DNA adducts. Human 5-HMU-DNA-glycosylase has been shown to repair this lesion. Whether urinary levels of 5-HMU is a valid biomarker for oxidative DNA damage in vivo has been investigated. However, controversial results on its relation to cigarette smoking were reported. To facilitate analysis of urinary 5-HMU in epidemiological studies, a highly sensitive and specific assay based on stable isotope dilution gas chromatography/negative ion chemical ionization/mass spectrometry was developed. The limit of detection for N¹,N ³-bis(pentafluorobenzyl)-HMU is 10 fg (20 amol) (S/N = 4) injected on column and the limit of quantification in urine was 0.7 nM of 5-HMU. Using as little as 10 μL of human urine samples, levels of urinary 5-HMU in 21 healthy volunteers were accurately quantified. No correlation was observed between urinary 5-HMU levels and cigarette smoking. However, there was a statistically significant association between urinary levels of 5-HMU and thiobarbituric acid-reactive substances (r = 0.71, p = 0.0003). In addition, urinary 5-HMU levels also correlated with urinary levels of 1,N⁶-ethenoadenine (r = 0.54, p = 0.01). These findings suggest that this assay should be valuable in assessing the role of urinary 5-HMU as a biomarker of oxidative DNA damage and repair.

Full text of DOI:10.1016/j.toxlet.2004.11.009

Toxicology Letters 155 (2005) 403–410
Measurement of urinary excretion of 5-hydroxymethyluracil
in human by GC/NICI/MS:
Correlation with cigarette smoking, urinary TBARS
and etheno DNA adduct
Hauh-Jyun Candy Chen∗, Chan-Fu Wu, Ju-Li Huang
Department of Chemistry and Biochemistry, National Chung Cheng University, 160 San-Hsing, Ming-Hsiung, Chia-Yi 62142, Taiwan
Received 9 September 2004; received in revised form 16 November 2004; accepted 18 November 2004
This article is dedicated to Professor Iwao Ojima for his 60th birthday
Abstract
5-Hydroxymethyluracil (5-HMU) is derived from radiation in addition to endogenous oxidative DNA damage and it is one of
the most abundant DNA adducts. Human 5-HMU-DNA-glycosylase has been shown to repair this lesion. Whether urinary levels
of 5-HMU is a valid biomarker for oxidative DNA damage in vivo has been investigated. However, controversial results on its
relation to cigarette smoking were reported. To facilitate analysis of urinary 5-HMU in epidemiological studies, a highly sensitive
and specific assay based on stable isotope dilution gas chromatography/negative ion chemical ionization/mass spectrometry was
developed. The limit of detection for N1,N3-bis(pentafluorobenzyl)-HMU is 10 fg (20 amol) (S/N = 4) injected on column and
the limit of quantification in urine was 0.7 nM of 5-HMU. Using as little as 10 ␮L of human urine samples, levels of urinary
5-HMU in 21 healthy volunteers were accurately quantified. No correlation was observed between urinary 5-HMU levels and
cigarette smoking. However, there was a statistically significant association between urinary levels of 5-HMU and thiobarbituric
acid-reactive substances (r = 0.71, p = 0.0003). In addition, urinary 5-HMU levels also correlated with urinary levels of 1,N6-
ethenoadenine (r = 0.54, p = 0.01). These findings suggest that this assay should be valuable in assessing the role of urinary
5-HMU as a biomarker of oxidative DNA damage and repair.
© 2004 Elsevier Ireland Ltd. All rights reserved.
Keywords: Cigarette smoking; 1,N6-Ethenoadenine; 5-Hydroxymethyluracil; Mass spectrometry; Oxidative DNA damage; TBARS; Urine
Abbreviations: PFB2-5-HMU; N1,N3-bis(pentafluorobenzyl)-5-HMU; GC/NICI/MS, gas chromatography negative ion chemical ionization
mass spectrometry; 5-HMU, 5-hydroxymethyluracil; LC/ESI/MS/MS, liquid chromatography/electrospray ionization tandem mass spectrome-
try; PFB, pentafluorobenzyl; SIM, selective ion monitoring; SPE, solid-phase extraction; TBARS, thobarbituric acid-reactive substances
∗
Corresponding author. Tel.: +886 524 281 76; fax: +886 527 210 40.
E-mail address: chehjc@ccunix.ccu.edu.tw (H.-J.C. Chen).
0378-4274/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.toxlet.2004.11.009

404
H.-J.C. Chen et al. / Toxicology Letters 155 (2005) 403–410
1. Introduction
tate the analysis of urinary 5-HMU in epidemiological
studies and to examine its correlation with the index
of lipid peroxidation and lipid peroxidation-induced
DNA adducts in the urine.
The production of reactive oxygen species within
the cells causes damage of biomolecules and it plays
an important role in carcinogenesis and aging. An
important type of oxidative DNA damage is hy-
droxylation of DNA bases, leading to formation of
8-oxo-7,8-dihydroguanine; 5-hydroxymethyluracil (5-
HMU); 5-hydroxyuracil; 8-oxo-7,8-dihydroadenine,
etc. Hydroxylated DNA adducts have been considered
as biomarkers of oxidative DNA damage (Shigenaga
and Ames, 1991). Cigarette smoking produces vari-
ous reactive oxygen species and results in elevation of
oxidative stress and lung cancer formation (Pryor and
Stone, 1993; Hecht, 1999).
2. Materials and methods
2.1. Subjects
The subjects were aged between 20 and 68 years.
The population consists of 9 smokers and 12 non-
smokers, including 3 female non-smokers. The range
of number of cigarettes smoked per day was between
5 and 25 and that of years of smoking was between 3
and 50 years.
Although the biological consequences of 5-HMU in
human are not fully understood, specific glycosylase
activities have been identified which remove 5-HMU
from DNA (Rusmintratip and Sowers, 2000; Baker et
al., 2002). Two possible sources account for formation
of 5-HMU: via hydroxylation of thymine, forming the
HmU:A mismatched base pair, or hydroxylation fol-
lowed by deamination of 5-methylcytosine, resulting
in HmU:G base pair formation. Although the occur-
rence of the mismatched HmU:A base pair greatly ex-
ceeds that of HmU:G, the latter is removed much more
efficiently than the former by human 5-HMU glycosy-
lase activities (Rusmintratip and Sowers, 2000; Baker
et al., 2002). Thus, the conversion of 5-methylcytosine
to 5-HMU is a potential pathway for the generation of
5-methylcytosine to T-transition mutations often found
in human tumors.
Increased 5-HMU formation in tissue DNA has
been connected with human diseases, and 5-HMU is
considered a biomarker of oxidative DNA damage
and of breast cancer (Djuric et al., 1991, 1996;
Frenkel et al., 1993). Bianchini and co-workers
reported that urinary excretion of 5-HMU is higher
than that of other oxidized nucleobases, including
8-oxo-7,8-dihydroguanine, and it is significantly
higher in smokers than in non-smokers (Bianchini
et al., 1998). However, Pourcelot and co-workers
showed that urinary excretion of HMU was similar
in the smokers and non-smokers groups (Pourcelot et
al., 1999). In this study, a highly sensitive and specific
assay based on stable isotope dilution gas chro-
matography/negative ion chemical ionization/mass
spectrometry (GC/NICI/MS) was developed to facili-
2.2. Materials
5-Hydroxymethyluracil, pentafluorobenzyl bro-
mide (PFB-Br), diisopropylethylamine, anhydrous
methanol, and anhydrous phosphorous pentoxide
were obtained from Aldrich Chemical Co. (Mil-
waukee, WI). [13C2,D2]5-Hydroxymethyluracil (4,5-
13C2, 98%; 5,5ꢀ-D2, 98%) was purchased from Cam-
bridge Isotope Laboratories, Andover, MA. Bond Elut
C18–OH and Si solid-phase extraction (SPE) columns
(500 mg, 3 mL) were from Varian (Harbor City, CA).
2.3. Verification of the structure of PFB2-5-HMU
To a 4 mL vial containing dried 5-HMU (1.0 mg,
7.8 ␮mol) was added a solution of diisopropy-
lethylamine (28 ␮L, 186 mmol) and PFB-Br (20 ␮L,
132 mmol) in 0.2 mL anhydrous methanol and the re-
action mixture was stirred at 42 ◦C for 1 h. The reaction
mixture was evaporated under vacuum and purified
by two Si SPE columns to afford 3.0 mg PFB2-5-
HMU (76% yield) as a white solid. 1H NMR (CDCl3)
δ 7.30 (C6–H, 1H), 5.28 (N3–CH2 of PFB, 2H),
5.20 (N1–CH2 of PFB, 2H), 4.40 (C5–CH2OH, 2H).
NICI/MS (assignment and relative abundance in paren-
theses) m/z 321 ([M–PFB]−, 100%), 305 ([M–PFB-
O]−, 25%).
2.4. Assay procedures
Urine samples collected over a 24 h period was
stored as 1.0 mL aliquots in 1.5 mL Eppendorff tubes

H.-J.C. Chen et al. / Toxicology Letters 155 (2005) 403–410
405
at the –84 ◦C freezer. The sample was centrifuged at
23,000 × g for 10 min at 4 ◦C. The creatinine contents
were analyzed by a picric acid method (Heinegard
and Tiderstrom, 1973). The supernatant (10 ␮L) of
pretreated urine sample was added [13C2,D2]5-HMU
(0.81 ng in 10 ␮L of water) and 0.1 M potassium phos-
phate buffer (pH 7.0, 100 ␮L) and the mixture was
loaded on a pre-conditioned C18–OH SPE column.
After the sample was loaded, the column was washed
with 1 mL of water and collected by elution with the
next 2 mL of water. The fraction containing 5-HMU
and [13C2,D2]5-HMU was dried over phosphorous
pentoxide, followed by addition of PFB-Br (20 ␮L)
and dii-sopropylethylamine (28 ␮L) in 0.1 mL anhy-
drous methanol under argon atmosphere and incu-
bation at 42 ◦C for 1 h. After the reaction mixture
was evaporated to dryness in vacuum, it was dis-
solved in dichloromethane (0.2 mL) and the pentaflu-
orobenzylated adduct was purified by a Si SPE col-
umn pre-conditioned with dichloromethane. The col-
umn was washed with 3 mL of 1% methanol in
dichloromethane (v/v) and 1 mL of 5% methanol in
dichloromethane (v/v) and PFB2-5-HMU was eluted
with 2 mL of 5% methanol in dichloromethane (v/v).
The eluant was evaporated, transferred to an in-
sert, evaporated to dryness, reconstituted in 40 ␮L
of acetone and a 1 ␮L aliquot was analyzed by GC/
NICI/MS.
to 300 ◦C at a gradient of 40 ◦C/min and remained at
300 ◦C for 5 min. PFB2-5-HMU and [13C2,D2]PFB2-
5-HMU for were detected and quantified at m/z 321
and 325, respectively, under the SIM mode for the re-
spective [M–181]− fragment ions. The quantification
of HMU was based on intrapolation of the ratio of the
peak area of PFB2-5-HMU versus [13C2,D2]PFB2-5-
HMU on the calibration curve.
2.6. Assay calibration
A fixed amount of [13C2,D2] 5-HMU (0.81 ng) in
10 ␮L of water was added as internal standard to each
sample containing various amounts of 5-HMU ranging
from 0, 1, 5, 10, 20, 50, 100, and 300 pg in 0.1 M potas-
sium phosphate buffer (pH 7.0, 100 ␮L) in a final vol-
ume of 120 ␮L. The samples were processed through
the same procedures for urine samples, i.e. adduct
enrichment by a C18–OH SPE column, pentafluo-
robenzylation, Si SPE purification, and GC/NICI/MS
analysis. The quantification of 5-HMU was based on
intrapolation of the ratio of the peak area of PFB2-
5-HMU versus that of [13C2,D2]PFB2-5-HMU to the
calibration curve obtained from at least duplicated ex-
periments.
2.7. Measurement of thiobarbituric acid-reactive
substances (TBARS)
2.5. GC/NICI/MS analysis of PFB2-5-HMU
The level of TBARS in human urine was measured
as malondialdehyde (MDA) using a modified method
(Liu et al., 1998). Urine was centrifuged at 22,000 × g
for 30 min. A 280 ␮L aliquot of the supernatant was
mixed with 2 mL of 0.1N HCl, 0.3 mL of 10% phos-
photungstic acid, 20 ␮L of 2 mM butylated hydroxy-
toluene (BHT), and 1 mL of 0.7% thiobarbituric acid
(TBA). The mixture was heated for 60 min at 90 ◦C,
followed by addition of saturated sodium chloride so-
lution (0.5 mL) and extraction of MDA-TBA adduct
with 5 mL of 1-butanol twice. The fluorescence inten-
sity of the MDA-TBA complex in the n-butanol layer
was quickly measured in a fluorescence spectropho-
tometer using excitation wavelength of 525 nm and
emission wavelength of 550 nm. The experiment was
performed in triplicates and the urinary concentration
of TBARS was expressed as ␮M MDA. A calibra-
tion curve was constructed for each run using 1,1,3,3-
tetramethoxypropane as the standard.
The GC/NICI/MS analysis was performed using a
Hewlett Packard 6890 GC with 5973 mass selective
detector (MSD) with the negative ion chemical ioniza-
tion (NICI) source (Agilent Technologies, Palo Alto,
CA). The filament was operated at 120 eV with the ion
source at 150 ◦C. The analyses were carried out with
a cool-on-column inlet, a precolumn (Restek, 2.5 m,
0.53 mm, deactivated silica), and a Rtx-5MS capil-
lary column (Restek, Dellefonte, PA, 0.25 mm × 30 m,
0.25 ␮m film thickness) inserted into the ion source.
Methane (99.999% pure) was the reagent gas with a
flow rate of 2.0 mL/min and the pressure at the ion
gauge was 2.2 × 10−4 Torr. Helium was used as the
carrier gas (99.999% pure) at a flow rate of 1.2 mL/min.
The oven temperature of GC was held at 50 ◦C for the
first 2 min, raised to 230 ◦C at a gradient of 40 ◦C/min
andmaintainedat230 ◦Cfor26 minbeforebeingraised

406
H.-J.C. Chen et al. / Toxicology Letters 155 (2005) 403–410
2.8. Analysis of urinary 1,N6-ethenoadenine
Levels of urinary 1,N6-ethenoadenine were deter-
mined by the isotope dilution GC/NICI/MS method
published previously (Chen and Chiu, 2003).
2.9. Statistical analysis.
All results are reported as the means standard de-
viation (S.D.). Statistical analysis of two groups was
performed by Mann–Whitney U-test and p ≤ 0.05 was
considered significant. Spearman rank correlation was
used to calculate correlation coefficient. GraphPad In-
Stat version 3.00 for Windows 95, GraphPad Software,
San Diego California USA (www.graphpad.com) was
used for these analyses.
3. Results
Urinary levels of 5-HMU were analyzed by the
isotope dilution gas chromatography/negative ion
chemical ionization/mass spectrometry (GC/NICI/
MS) after pentafluorobenzylation as N1,N3-bis(penta-
fluorobenzyl)-HMU (PFB2-5-HMU), which showed
good response with GC/NICI/MS despite of its free
hydroxyl group. Using modified procedures, the limit
of detection (LOD) for standard PFB2-5-HMU in-
jected on column was improved as 20 amol (or 10 fg
with a S/N ratio of 4), compared to the previous
report (Kresbach et al., 1989). The entire assay in-
cluded adduct enrichment by a C18–OH solid-phase
extraction (SPE) column, followed by electrophore-
labeling and post-derivatization cleanup with a Si
SPE column before analysis by GC/NICI/MS. The
recovery of the entire assay was 58%. Injecting one
fortieths of the sample, the limit of quantification
(LOQ) was 1.0 pg (7.1 fmol), defined as the low-
est amount of the analyte which showed quantitative
linearity in the calibration curve (Fig. 1). The lin-
ear correlation coefficient (γ2) was 0.9995. The y-
intercept was the peak area ratio of PFB2-5-HMU ver-
sus [13C2,D2]PFB2-5-HMU in the control samples.
Due to the reported high urinary 5-HMU levels, 10 uL
of urine sample was used for each analysis and the
adduct levels were all above the LOQ. The concen-
tration quantification limit was thus determined as
0.7 nM.
Fig. 1. GC/NICI/MS analysis of PFB2-5-HMU in a non-smoker’s
urine. Urine (10 ␮L) was added [13C2,D2]5-HMU (0.81 ng), en-
riched with a C18–OH SPE column, derivatized with PFB-Br, puri-
fied by a Si SPE column, and analyzed by GC/NICI/MS under SIM
modeasdescribedinthe(Section2). Thepeakof22.97 minatm/z321
represents 12 pg (24 fmol) of PFB-5-HMU and the one of 22.88 min
at m/z 325 is [13C2,D2]PFB-5-HMU. The result corresponds to a
5-HMU concentration of 13.35 ng/mL.
Fig. 2 showed a representative GC/MS chromatog-
raphy of PFB2-5-HMU in a non-smoker’s urine sam-
ple; that of a smoker’s urine was similar (not shown).
To separate interference peaks, an isothermal GC
condition was used. [13C2,D2]PFB2-5-HMU eluted
at 22.88 min at m/z 325, while PFB2-5-HMU was
22.97 min at m/z 321. The small difference in reten-
tion times (ca. 0.11 min) was observed consistently
in the calibration curve and in all the urine analyses.
Quantification of urinary 5-HMU was based on intrap-
olation of the peak area ratio of PFB2-5-HMU versus
[13C2,D2]PFB2-5-HMU in the calibration curve.
Levels of urinary 5-HMU in the 21 healthy vol-
unteers were in the range between 4.1 ng/mL and
23 ng/mL, which were consistent with the previous re-
ports (Bianchini et al., 1998; Ravanat et al., 1999).
Acceptable variances (R.S.D. < 10%) were found for

H.-J.C. Chen et al. / Toxicology Letters 155 (2005) 403–410
407
Fig. 2. Calibration curve for the GC/NICI/MS analysis of 5-HMU. Samples containing various amounts (0–300 pg) of 5-HMU were added to
a fixed amount of [13C2,D2]5-HMU (0.81 ng) and subjected to the assay procedures described in the (Section 2) (r2 = 0.9995). The data are
combined from at least separate experiments in duplicates. The ratio of each analyte to the internal standard was calculated based on the peak
areas.
Table 1
Precision of the GC/MS assay for urinary HMU
Samples
Adduct levels (ng/mL) R.S.D. (%) (n = 5)
Day 1
Day 2
Day 3
Interday variation R.S.D. (%)
No. 1
No. 2
No. 3
18.3 0.3 (1.8)
8.5 0.9 (10)
5.7 0.1 (1.6)
20.7 0.1 (0.5)
7.3 0.1 (1.9)
6.8 0.3 (4.4)
20.3 0.4 (1.9)
7.1 0.3 (4.4)
6.2 0.2 (3.2)
5.4
8.0
7.4
Values in parenthesis are in percent.
all the samples. The averaged intraday and inter-
day precision of the assay was 3.3% (R.S.D., n = 5)
and 6.9% (R.S.D., n = 3), respectively (Table 1).
The concentrations of 5-HMU in smokers’ and
non-smokers’ urine were determined as 95 36 nM
and 80 40 nM, respectively. Despite normalization
of the adduct levels by creatinine concentration
and/or body weight, no statistically significant cor-
relation was found between urinary 5-HMU levels
and cigarette smoking using Mann–Whitney U-test
(Table 2). However, an extremely significant associ-
ation (p = 0.0003) was found between urinary levels
of 5-HMU and thiobarbituric acid-reactive substances
(TBARS) with a Spearman correlation coefficient (r)
of 0.71 (Fig. 3). Furthermore, urinary 5-HMU levels
were found to correlate with urinary levels of 1,N6-
ethenoadenine(r = 0.54, p = 0.01)(Fig. 4), butnot3,N4-
ethenocytosine (r = 0.34, p = 0.13).
4. Discussion
Compare to 32P-postlabelling technique or im-
munoaffinity chromatography-based assays for
analysis of DNA adducts, mass spectrometry pro-
vides chemical characterization of the analyte.

408
H.-J.C. Chen et al. / Toxicology Letters 155 (2005) 403–410
Table 2
Levels of 5-HMU in human urine samples and statistical data
Smokersa (n = 9)
Non-smokersa (n = 12)
p-valueb
5-HMU (ng/mL)
5-HMU (nM)
5-HMU/creatinine (nmol/mmol)
5-HMU/body weight/ creatinine (pmol/(kg/g)
12.2 4.7 (7.5 − 22.9)
95 36 (58−178)
10.3 5.5 (4.1 − 20.1)
80 40 (32 − 156)
8.9 2.0 (6.2 − 13.7)
1.2 0.1 (0.6 − 1.9)
0.35
0.35
0.32
0.98
10.9 4.7 (7.4 − 23.2)
1.3 0.6 (0.7 − 2.9)
a
Mean S.D. (range).
b
The p-values were obtained by comparing adduct levels among smokers vs. non-smokers using nonparametric Mann–Whitney U-test.
Incorporation of stable isotope of the analyte, which
has identical chemical property as the analyte,
provides detailed monitoring of recovery in each
step of the analytical procedure and thus ensures
accurate quantification of the analyte. Analysis
of oxidized bases by GC/MS often used electron
impact ionization MS after derivatization with
silylating agents. The silylation conditions require
high temperature, e g. 110–150 ◦C, which might
cause artifactual formation of oxidized DNA adducts
(Halliwell and Dizdaroglu, 1992; Ravanat et al.,
1995). In this assay, pentafluorobenzylation was
performed at 42 ◦C and the background levels in the
control experiments were very low. The GC/MS assay
with electrophore-labeled analyte using negative ion
chemical ionization also provides much higher sensi-
tivity than methods using electron impact ionization.
Furthermore, the use of disposable SPE columns
in this assay eliminated the possibility of carryover
from collecting the adduct-containing fractions by
HPLC and the tedious cleanup between sample runs
(Bianchini et al., 1998; Ravanat et al., 1999). Levels
of HMU measured in this study (6.2–23.2 nmol/mmol
creatinine or 32–178 nM) are in the same range,
3.2–18.7 nmol/mmol creatinine, as those reported by
Bianchini and co-workers (Bianchini et al., 1998) or
121 56 nM by Ravanat and co-workers (Ravanat
et al., 1999). The agreement in the range of adduct
levels in these three studies suggest that our new assay
is reliable and it can be useful in epidemiological
studies.
An alternative technique, namely the liquid chro-
matography/electrospray ionization tandem mass spec-
trometry (LC/ESI/MS/MS), has recently been used
for analysis of DNA adducts in tissue and biological
fluidic samples with high specificity. To the best of
Fig. 3. Correlation between the urinary 5-HMU and TBARS levels of 21 human urine samples.

H.-J.C. Chen et al. / Toxicology Letters 155 (2005) 403–410
409
Fig. 4. Correlation between the urinary 5-HMU and 1,N6-ethenoadenine levels of 21 human urine samples.
our knowledge, there has been no report on analysis
of 5-HMU by LC/ESI/MS/MS to date. On the other
hand, the deoxyribonucleoside 5-(hydroxymethyl)-2ꢀ-
deoxyuridine has been measured by LC/ESI/MS/MS in
cellular DNA (Frelon et al., 2000). We attempted to de-
velop a LC/ESI/MS/MS assay for urinary 5-HMU, but
the sensitivity of this base adduct was poor and severe
matrix effect suppressing the analyte signal was ob-
served (Chen et al. unpublished results). Although the
GC/MS assay described in this study is somewhat trou-
blesome, requiring adduct enrichment by SPE columns
and derivatization steps, it is highly sensitive and quan-
titative. Nonetheless, the SPE and derivatization steps
help in removing the interferences in the complex mix-
ture of urine.
Elevated levels of lipid peroxidation could lead
to damage of DNA producing exocyclic etheno
DNA adducts, such as 1,N6-ethenoadenine and 3,N4-
ethenocytosine (Chung et al., 1996). Although our re-
sults showed that urinary 5-HMU excretion was not
associated with cigarette smoking, it correlated with
urinary TBARS, an index of lipid peroxidation, as
well as urinary 1,N6-ethenoadenine, a biomarker of
oxidative DNA damage derived from lipid peroxida-
tion (Chen and Chiu, 2003). On the other hand, Faure
and co-workers reported that urinary 5-HMU levels
were significantly increased by the treatment of adri-
amycin, an anticancer drug producing reactive oxy-
gen species, with elevated TBARS and lowered an-
tioxidant vitamins in plasma of cancer patients (Faure
et al., 1996). Recently, urinary 5-(hydroxymethyl)-2ꢀ-
deoxyuridine was found to be discriminating of oxida-
tive stress from tobacco smoke (Harman et al., 2003).
Although there are other possible sources of urinary
DNA adducts, it is accepted that excretion of adducted
DNA bases and the deoxyribonucleosides represent
the result of repairing DNA damage by base exci-
sion repair and nucleoside excision repair mechanisms,
respectively. These adducts have been used to esti-
mate DNA adduction in the body, as demonstrated for
8-oxo-7,8-dihydroguanine and 8-oxo-7,8-dihydro-2ꢀ-
deoxyguanosine (Loft and Poulsen, 1998; Gackowski
et al., 2003). Thus, the urinary 5-HMU assay presented
in this study suggests its potential applicability to stud-
ies of oxidative DNA damage and repair in humans.
Acknowledgment
This work was supported by grants from National
Science Council of Taiwan and from National Chung
Cheng University (to H.-J.C.C.).

410
H.-J.C. Chen et al. / Toxicology Letters 155 (2005) 403–410
Urinary excretion of three nucleic acid oxidation adducts and iso-
References
prostane F“2@a measured by liquid chromatography-mass spec-
trometry in smokers, ex-smokers, and non-smokers. Free Radic.
Biol. Med. 35, 1301–1309.
Baker, D., Liu, P., Burdzy, A., Sowers, L.C., 2002. Characterization
of the substrate specificity of a human 5-hydroxymethyluracil
glycosylase activity. Chem. Res. Toxicol. 15, 33–39.
Bianchini, F., Donato, F., Faure, H., Ravanat, J.L., Hall, J., Cadet, J.,
1998. Urinary excretion of 5-(hydroxymethyl)uracil in healthy
volunteers: effect of active and passive tobacco smoke. Int. J.
Cancer 77, 40–46.
Hecht, S.S., 1999. Tobacco smoke carcinogens and lung cancer. J.
Natl. Cancer Inst. 91, 1194–1210.
Heinegard, D., Tiderstrom, G., 1973. Determination of serum cre-
atinine by a direct colorimetric method. Clin. Chim. Acta 43,
305–310.
Kresbach, G.M., Itani, M., Saha, M., Rogers, E.J., Vouros, P.,
Giese, R.W., 1989. Ester and related derivatives of ring
N-pentafluorobenzylated 5-hydroxymethyluracil. Hydrolytic
stability, mass spectral properties, and trace detection
by gas chromatography-electron-capture detection, gas
chromatography-electron-capture negative ion mass spectrom-
etry, and moving-belt liquid chromatography-electron-capture
negative ion mass spectrometry. J. Chromatogr. A 476, 423–438.
Liu, C.S., Chen, H.W., Lii, C.K., Chen, S.C., Wei, Y.-H., 1998. Al-
terations of small-molecular-weight antioxidants in the blood of
smokers. Chem. Biol. Interact. 116, 143–154.
Chen, H.-J.C., Chiu, W.-L., 2003. Detection and quantification
of 1,N6-ethenoadenine in human urine by stable isotope dilu-
tion capillary gas chromatography/negative ion chemical ioniza-
tion/mass spectrometry. Chem. Res. Toxicol. 16, 1099–1106.
Chung, F.-L., Chen, H.-J.C., Nath, R., 1996. Lipid peroxidation as
a potential endogenous source for the information or exocyclic
propano and etheno adducts. Carcionogenesis 17, 2105–2111.
Djuric, Z., Heilbrun, L.K., Reading, B.A., Boomer, A., Valeriote,
F.A., Martino, S., 1991. Effects of a low-fat diet on levels of
oxidative damage to DNA in human peripheral nucleated blood
cells. J. Natl. Cancer Inst. 83, 766–769.
Loft, S., Poulsen, H.E., 1998. Estimation of oxidative DNA damage
in man from urinary excretion of repair products. Acta Biochim.
Pol. 45, 133–144.
Pourcelot, S., Faure, H., Firoozi, F., Ducros, V., Tripier, M., Hee,
J., Cadet, J., Favier, A., 1999. Urinary 8-oxo-7, 8-dihydro-2ꢀ-
deoxyguanosine and 5-(hydroxymethyl) uracil in smokers. Free
Radic. Res. 30, 173–180.
Pryor, W.A., Stone, K., 1993. Oxidants in cigarette smoke. Radicals,
hydrogen peroxide, peroxynitrate, and peroxynitrite. Ann. N Y
Acad. Sci. 686, 12–27.
Ravanat, J.L., Guicherd, P., Tuce, Z., Cadet, J., 1999. Simultaneous
determination of five oxidative DNA lesions in human urine.
Chem. Res. Toxicol. 12, 802–808.
Djuric, Z., Heilbrun, L.K., Simon, M.S., Smith, D., Luongo, D.A.,
LoRusso, P.M., Martino, S., 1996. Levels of 5-hydroxymethyl-
2ꢀ-deoxyuridine in DNA from blood as a marker of breast cancer.
Cancer 77, 691–696.
Faure, H., Coudray, C., Mousseau, M., Ducros, V., Douki, T., Bian-
chini, F., Cadet, J., Favier, A., 1996. 5-Hydroxymethyluracil
excretion, plasma TBARS and plasma antioxidant vitamins in
adriamycin-treated patients. Free Radic. Biol. Med. 20, 979–983.
Frelon, S., Douki, T., Ravanat, J.-L., Pouget, J.-P., Tornabene,
C., 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.
Ravanat, J.-L., Turesky, R.J., Gremaud, E., Trudel, L.J., Stadler,
R.H., 1995. Determination of 8-oxoguanine in DNA by gas
chromatography-mass spectrometry and HPLC-electrochemical
detection: overestimation of the background level of the oxidized
base by the gas chromatography-mass spectrometry assay. Chem.
Res. Toxicol. 8, 1039–1045.
Rusmintratip, V., Sowers, L.C., 2000. An unexpectedly high excision
capacity for mispaired 5-hydroxymethyluracil in human cell ex-
tracts. Proc. Natl. Acad. Sci. USA 97, 14183–14187.
Shigenaga, M.K., Ames, B.N., 1991. Assays for 8-hydroxy-2ꢀ-
deoxyguanosine: a biomarker of in vivo oxidative DNA damage.
Free Radic. Biol. Med. 10, 211–216.
Frenkel, K., Karkoszka, J., Kim, E., Taioli, E., 1993. Recognition
of oxidized DNA bases by sera of patients with inflammatory
diseases. Free Radic. Biol. Med. 14, 483–494.
Gackowski, D., Speina, E., Zielinska, M., Kowalewski, J., Rozalski,
R., Siomek, A., Paciorek, T., Tudek, B., Olinski, R., 2003. Prod-
ucts of oxidative DNA damage and repair as possible biomarkers
of susceptibility to lung cancer. Cancer Res. 63, 4899–4902.
Halliwell, B., Dizdaroglu, M., 1992. The measurement of oxidative
damage to DNA by HPLC and GC/MS techniques. Free Radic.
Res. Commun. 16, 75–87.
Harman, S.M., Liang, L., Tsitouras, P.D., Gucciardo, F., Heward,
C.B., Reaven, P.D., Ping, W., Ahmed, A., Cutler, R.G., 2003.