Screening for DNA adducts by data-dependent constant neutral loss-triple stage mass spectrometry with a linear quadrupole ion trap mass spectrometer

Article abstract of DOI:10.1021/ac802096p

A two-dimensional linear quadrupole ion trap mass spectrometer (LIT/MS) was employed to simultaneously screen for DNA adducts of environmental, dietary, and endogenous genotoxicants, by data-dependent constant neutral loss scanning followed by triple-stag

Full text of DOI:10.1021/ac802096p

Anal. Chem. 2009, 81, 809–819
Screening for DNA Adducts by Data-Dependent
Constant Neutral Loss-Triple Stage Mass
Spectrometry with a Linear Quadrupole Ion Trap
Mass Spectrometer
Erin E. Bessette,† Angela K. Goodenough,†,‡ Sophie Langoue¨ t,§,| Isil Yasa,† Ivan D. Kozekov,
Simon D. Spivack,X and Robert J. Turesky*,†
Division of Environmental Health Sciences, Wadsworth Center, New York State Department of Health,
Albany, New York 12201, Bristol-Myers Squibb, P.O. Box 4000, Princeton, New Jersey 08543, INSERM U620,
Universite´ de Rennes I, 35043 Rennes, France, EA SeRAIC, IFR 140, 35043 Rennes, France, Department of
Chemistry, Center in Molecular Toxicology, and the Vanderbilt Institute for Chemical Biology, Vanderbilt University,
Nashville, Tennessee 37235, and Division of Pulmonary Medicine, Albert Einstein College of Medicine/Montefiore
Medical Center, Bronx, New York 10461
A two-dimensional linear quadrupole ion trap mass spectrom-
eter (LIT/MS) was employed to simultaneously screen for DNA
adducts of environmental, dietary, and endogenous genotoxi-
cants, by data-dependent constant neutral loss scanning fol-
lowed by triple-stage mass spectrometry (CNL-MS3). The loss
of the deoxyribose (dR) from the protonated DNA adducts
([M + H - 116]+) in the MS/MS scan mode triggered the
acquisition of MS3 product ion spectra of the aglycone
adducts [BH2]+. Five DNA adducts of the tobacco carcino-
gen 4-aminobiphenyl (4-ABP) were detected in human
hepatocytes treated with 4-ABP, and three DNA adducts
of the cooked-meat carcinogen 2-amino-3,8-dimethylimi-
dazo[4,5-f]quinoxaline (MeIQx) were identified in the livers
of rats exposed to MeIQx, by the CNL-MS3 scan mode.
Buccal cell DNA from tobacco smokers was screened for
DNA adducts of various classes of carcinogens in tobacco
smoke including 4-ABP, 2-amino-9H-pyrido[2,3-b]indole
(ArC), and benzo[a]pyrene (BaP); the cooked-meat car-
cinogens MeIQx, ArC, and 2-amino-1-methyl-6-phenylmi-
dazo[4,5-b]pyridine (PhIP); and the lipid peroxidation
products acrolein (AC) and trans-4-hydroxynonenal (HNE).
The CNL-MS3 scanning technique can be used to simulta-
neously screen for multiple DNA adducts derived from
different classes of carcinogens, at levels of adduct modi-
fication approaching 1 adduct per 108 unmodified DNA
bases, when 10 µg of DNA is employed for the assay.
for cross-species extrapolation of the biologically effective dose
and for human risk assessments of exposure to chemical
carcinogens.2,3 The 32P-postlabeling method has been a mainstay
in biomonitoring of DNA adducts, because of its sensitivity and
ability to detect many different classes of DNA adducts.4 HPLC
coupled with electrochemical or fluorescence detection has also
been used to measure certain types of oxidative DNA lesions5
or polycyclic aromatic hydrocarbon (PAH) adducts.6 Immuno-
histochemistry (IHC) is a third technique that has been
employed to monitor multiple classes of carcinogen-DNA
adducts.7 A major limitation in all of these analytical methods
is the lack of information that they provide on the structure of
the DNA adduct. Hence, the identity of the lesion remains
equivocal.
Sensitive mass spectrometric (MS) methods have been estab-
lished during the past 2 decades, to detect DNA adducts. These
MS techniques provide varying degrees of structural information
on the adduct. Accelerator mass spectrometry (AMS) is the most
sensitive MS-based technique,8 but it requires dosing the subject
with a radiolabeled substrate.9 Both tritiated and 14C-isotopically
labeled carcinogens have been used for adduct detection by
AMS.8 However, structural information is limited to the
detection of the radioactive isotope; then a chromatographic
technique and coelution of the adduct with a nonradioactive
reference compound are required to provide evidence of adduct
identity. Gas chromatography-negative ion chemical ionization/
mass spectrometry (GC-NICI/MS) has been used to success-
The covalent modification of DNA by chemical mutagens is
regarded as the initiating step in chemical carcinogenesis.1 The
measurement of DNA adducts is an important end point, both
(1) Miller, E. C. Cancer Res. 1978, 38, 1479–1496
(2) Swenberg, J. A.; La, D. K.; Scheller, N. A.; Wu, K. Y. Toxicol. Lett. 1995,
82-83, 751–756
(3) Poirier, M. C.; Beland, F. A. Chem. Res. Toxicol. 1992, 5, 749–755
(4) Beach, A. C.; Gupta, R. C. Carcinogenesis 1992, 13, 1053–1074
(5) Floyd, R. A.; Watson, J. J.; Wong, P. K.; Altmiller, D. H.; Rickard, R. C.
Free Radical Res. Commun. 1986, 1, 163–172
(6) Jankowiak, R.; Small, G. J. Anal. Chem. 1989, 61, 1023A–1024A
(7) Poirier, M. C.; Santella, R. M.; Weston, A. Carcinogenesis 2000, 21, 353–
359
(8) Turteltaub, K. W.; Dingley, K. H. Toxicol. Lett. 1998, 102-103, 435–439
(9) Farmer, P. B.; Brown, K.; Tompkins, E.; Emms, V. L.; Jones, D. J.; Singh,
R.; Phillips, D. H. Toxicol. Appl. Pharmacol. 2005, 207, 293–301
.
.
.
* Corresponding author. Robert J. Turesky, Ph.D. Division of Environmental
Health Sciences, Wadsworth Center, New York State Department of Health.
Phone: 518-474-4151. Fax: 518- 473-2095. E-mail: Rturesky@wadsworth.org.
† New York State Department of Health.
.
.
.
‡ Bristol-Myers Squibb.
§ INSERM U620, Universite´ de Rennes I.
.
| EA SeRAIC, IFR 140.
.
Vanderbilt University.
X Albert Einstein College of Medicine/Montefiore Medical Center.
.
10.1021/ac802096p CCC: $40.75 2009 American Chemical Society
Published on Web 12/16/2008
Analytical Chemistry, Vol. 81, No. 2, January 15, 2009 809

fully measure several classes of DNA adducts, following
chemical derivatization.10 Because only one ion is usually
monitored, structural information provided about the adduct
is limited. Atmospheric pressure ionization (API) techniques
are the most robust analytical MS methods used to detect
and DNA adduct measurements.23-26 The QIT/MS permits ion
storage and sequential ejection of ions, as a function of the mass,
through adjustment of the strength of the quadrupole field holding
the ions. This scan filter enables consecutive reaction monitoring
(CRM) or multistage scan events (MSn), and the rapid-scanning
capacity of the instrument permits the routine acquisition of
full product ion spectra, thereby enabling extensive mass
spectral characterization of the analytes. The recently devel-
oped two-dimensional LIT/MS has been advertised to have
superior ion storage volume and trapping efficiencies, as well
as reduced space-charge effects27 and increased sensitivity, in
comparison to some predecessor three-dimensional QIT/MS
instruments.28-30 We recently employed a LIT/MS to quantitate
DNA adducts of PhIP,25 an experimental animal carcinogen and
likely human carcinogen,31 that is formed in cooked meat.32 The
high ion dissociation efficiency of the LIT/MS resulted in the
recovery of >90% of the total ion counts of the PhIP adducts as it
transitioned from MSn to MSn+1 stage scan mode and enabled
us to characterize DNA adducts of PhIP formed in experimental
animals, with full scan spectra at MS3 and MS4 stage scan
modes.25
LC-MS has generally been applied to the analysis of one to
several DNA adducts of a specific carcinogenic agent.33,34 How-
ever, humans are continuously exposed to genotoxicants in the
environment and the diet, as well as endogenously as a conse-
quence of oxidative cellular stress. Thus, a robust MS screening
method that can simultaneously analyze multiple adducts derived
from multiple sources of exposures would represent a major
advance in biomonitoring of DNA adducts. There is one published
report that describes the use of TQ/MS/MS and the CNL scan
mode transition [M + H]+ f [M + H - 116]+, attributed to
the loss of dR, to monitor putative DNA adducts in lung tissue
of tobacco smokers.35 Numerous analytes were detected;
however, it was not known whether many of these analytes
were actually DNA adducts or simply other compounds that
lost 116 Da upon collision-induced dissociation (CID) in MS/
MS. A variety of data-dependent scan techniques employing
-12
nonvolatile and thermally labile DNA adducts.10
Liquid chromatography in combination with electrospray
ionization (ESI) tandem mass spectrometry (LC-ESI/MS/MS)
has been used to analyze numerous classes of carcinogen-DNA
adducts.10,12 LC-ESI/MS with triple quadrupole tandem mass
spectrometry (TQ/MS/MS) has been the primary instrumentation
for the quantification of DNA adducts. For trace-analysis studies,
the TQ/MS/MS system is operated in the selected reaction
monitoring (SRM) scan mode. The protonated adducts ([M +
H]+) are selectively transmitted by the first mass analyzer (Q1)
and are subjected to collision-induced dissociation (CID),
typically with argon gas, in the second quadrupole (Q2). These
collision conditions generally result in the loss of deoxyribose
[M + H - 116]+, to form the protonated base aglycone adducts
[BH2]+ as the principal product ions; these are then selectively
transmitted through the third quadrupole (Q3). The SRM scan
mode is very effective for precise quantification because of the
rapid duty cycle of TQ/MS/MS. Under higher collision-energy
conditions, the aglycone [BH2]+ ion undergoes extensive
fragmentation, and the product ion scan mode can be used to
obtain structural information about the adduct.11 However, the
slow scanning rate of the product ion scan mode generally
precludes the use of this mode for characterization of DNA
adducts in vivo;13 the scanning method lacks sensitivity,
because only a small fraction (<1%) of the total ions enter the
,15
MS.14 Thus, in trace-analysis measurements by TQ/MS/MS,
the analyst must rely solely on the characteristic retention time
(tR); often, only a single SRM transition ([M+ H - 116]+) is
used as a criterion for analyte identification.
Quadrupole ion trap mass spectrometry (QIT/MS)16 has been
used in a variety of bioanalytical applications, including proteom-
ics,17 oligonucleotide sequencing,18,19 analyses of dietary and
environmentalcontaminants,20characterizationofdrugmetabolites,21,22
(23) Li, L.; Chiarelli, M. P.; Branco, P. S.; Antunes, A. M.; Marques, M. M.;
Goncalves, L. L.; Beland, F. A. J. Am. Soc. Mass Spectrom. 2003, 14, 1488–
1492
.
(24) Turesky, R. J.; Goodenough, A. K.; Ni, W.; McNaughton, L.; LeMaster,
D. M.; Holland, R. D.; Wu, R. W.; Felton, J. S. Chem. Res. Toxicol. 2007,
(10) Koc, H.; Swenberg, J. A. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci.
20, 520–530
(25) Goodenough, A. K.; Schut, H. A.; Turesky, R. J. Chem. Res. Toxicol. 2007,
20, 263–276
(26) Liu, X.; Lovell, M. A.; Lynn, B. C. Chem. Res. Toxicol. 2006, 19, 710–718
(27) Schwartz, J. C.; Senko, M. W.; Syka, J. E. P. J. Am. Soc. Mass Spectrom.
2002, 13, 659–669
.
2002, 778, 323–343
(11) Turesky, R. J.; Vouros, P. J. Chromatogr., B: Anal. Technol. Biomed. Life
Sci. 2004, 802, 155–166
(12) Singh, R.; Farmer, P. B. Carcinogenesis 2006, 27, 178–196
(13) Paehler, A.; Richoz, J.; Soglia, J.; Vouros, P.; Turesky, R. J. Chem. Res.
Toxicol. 2002, 15, 551–561
(14) McLuckey, S. A.; Van Berkel, G. J.; Goeringer, D. E.; Glish, G. L. Anal.
Chem. 1994, 66, 689A–696A
(15) Gangl, E.; Utkin, I.; Gerber, N.; Vouros, P. J. Chromatogr., A 2002, 974,
91–101
(16) Paul, W. Agnew. Chem., Int. Ed. 1990, 29, 739–748
(17) Mikesh, L. M.; Ueberheide, B.; Chi, A.; Coon, J. J.; Syka, J. E.; Shabanowitz,
J.; Hunt, D. F. Biochim. Biophys. Acta 2006, 1764, 1811–1822
(18) Jackson, P. E.; Scholl, P. F.; Groopman, J. D. Mol. Med. Today 2000, 6,
271–276
(19) Chowdhury, G.; Guengerich, F. P. Angew. Chem., Int. Ed. 2008, 47, 381–
384
(20) Andreu, V.; Pico, Y. Curr. Anal. Chem. 2005, 1, 241–265
(21) Anari, M. R.; Sanchez, R. I.; Bakhtiar, R.; Franklin, R. B.; Baillie, T. A. Anal.
Chem. 2004, 76, 823–832
(22) Ma, L.; Wen, B.; Ruan, Q.; Zhu, M. Chem. Res. Toxicol. 2008, 21, 1477–
1483
.
.
.
.
.
.
.
(28) Thermo Electron. Product Support Bulletin 117, Thermo Electron Corpora-
tion: Waltham, MA, 2006.
(29) Bonilla, L. E.; Thakur, R.; Guzzetta, A.; Jaffe, J. D. Application Note 328,
2006.
.
.
(30) Riter, L. S.; Gooding, K. M.; Hodge, B. D.; Julian, R. K., Jr. Proteomics 2006,
.
6, 1735–1740.
(31) National Toxicology Program. Report on Carcinogenesis, 11th ed.; U.S.
Department of Health and Human Services, Public Health Service: Research
Triangle Park, NC, 2005.
.
.
(32) Sugimura, T.; Wakabayashi, K.; Nakagama, H.; Nagao, M. Cancer Sci.
2004, 95, 290–299
(33) Churchwell, M. I.; Beland, F. A.; Doerge, D. R. Chem. Res. Toxicol. 2002,
15, 1295–1301
(34) Lao, Y.; Villalta, P. W.; Sturla, S. J.; Wang, M.; Hecht, S. S. Chem. Res. Toxicol.
2006, 19, 674–682
(35) Kanaly, R. A.; Hanaoka, T.; Sugimura, H.; Toda, H.; Matsui, S.; Matsuda,
T. Antioxid. Redox Signaling 2006, 8, 993–1001
.
.
.
.
.
.
.
.
810 Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

QIT/MS have been used to monitor post-translational modifica-
B[a]PDE with dG as described.42 The adducts were purified
by SPE followed by HPLC purification13,39 and were isolated as
a mixture of unresolved isomers. The (6R/S)-3-(2′-Deoxyribos-
1′-yl)-5,6,7,8-tetrahydro-6-hydroxypyrimido[1,2-a]purine-
10(3H)one (6-hydroxy-PdG),43 the (8R/S)-3-(2′-deoxyribos-1′-yl)-
5,6,7,8-tetrahydro-8-hydroxypyrimido[1,2-a]purine-10(3H)one (8-
hydroxy-PdG),44 and the isomeric 1,N2-8-hydroxypropano-2′-
deoxyguanosine adducts of HNE (dG-HNE)45 were synthesized
as previously reported.
,22,37
tions of proteins,36 and to characterize drug metabolites.21
The neutral-loss scan mode, by the traditional definition, cannot
be performed on the LIT/MS. However, neutral loss ion maps
can be obtained retrospectively from the full set of MS/MS data,
once the acquisition is complete. We sought to determine whether
the rapid scanning speed and high sensitivity provided by the LIT/
MS can be exploited to screen for multiple DNA adducts of
multiple classes of genotoxicants by data-dependent CNL-MS3
,
PhIP-Modified Calf Thymus DNA. [3H]-PhIP-modified CT
DNA was provided by Dr. F. Kadlubar (UAMS, Little Rock,
AR).38 The extent of PhIP modification was estimated at 1
adduct per 106 unmodified DNA bases.
using the neutral loss of the dR moiety in the MS/MS scan
mode to trigger the acquisition of MS3 product ion spectra of
the aglycone ion [BH2+] adducts in a single analysis.
4-ABP- and BaP-Modified Calf Thymus DNA. [3H]-4-ABP-
CT DNA (62 adducts per 108 unmodified DNA bases)46 and
[3H]-BaP-modified CT DNA (111 adducts per 108 unmodified
DNA bases)47 were provided by Dr. F. Beland (NCTR,
Jefferson, AR).
EXPERIMENTAL SECTION
Caution. PhIP, MeIQx, AC, 4-ABP, BaP, HNE, AC, and their
derivatives are genotoxic and/or carcinogenic agents, and they
should only be handled in a well-ventilated fume hood with the
appropriate protective clothing.
DNA Adduct Formation of 4-ABP in Human Hepatocytes.
A human liver sample was obtained from a patient undergoing
liver resection for primary hepatoma via the Biological Resource
Center (CHRU Pontchaillou, Rennes, France). The research
protocol met French legal guidelines and the requirements of the
local institutional ethics committee. Hepatocytes were isolated by
a two-step collagenase perfusion procedure.48 The liver paren-
chymal cells were seeded, at a density of 106 viable cells/35
cm2dish, in 2 mL of Williams’ medium with supplements.48
Once the medium was renewed, 4-ABP (10 µM) in DMSO (0.1%
v/v) was added and the cells were incubated for 24 h. Cellular
DNA was isolated by a chloroform/phenol extraction method.4
Animal Treatment with MeIQx. Male Fischer-344 rats
(220-260 g of body wt, Iffa Credo, L’Abresle, France) were dosed
orally with MeIQx-HCl (10 mg/kg body wt) in 10 mM phosphate-
buffered saline (pH 7.4).13 Control animals received phosphate-
buffered saline. The animals were sacrificed 24 h after treatment
and liver DNA was isolated as previously described.13
Human Buccal Cells. Subjects (tobacco smokers >20 ciga-
rettes per day and on a noncontrolled diet), swished mouthwash
(Scope brand, 10 mL) vigorously for 1 min and then expelled it
into a 50 mL polypropylene Falcon tube. The biospecimens were
rendered anonymous and stored at -20 °C until processed. DNA
was isolated by phenol/chloroform extraction as previously
described.49 This study was approved by the Institutional Review
Boards at the Wadsworth Center and the Albert Einstein College
of Medicine.
Chemicals. MeIQx and PhIP were purchased from Toronto
Research Chemicals (Toronto, ON, Canada). 2-Nitro-9H-pyrido[4,5-
b]indole (NO2-ARC) was a kind gift from Dr. D. Miller, NCTR
(Jefferson, AR). (±)-r-7,t-8-Dihydroxy-t-9,10-epoxy-7,8,9,10-tet-
rahydrobenzo[a]pyrene ((±) (anti)B[a]PDE)) was purchased
from the NCI Chemical Carcinogen Reference Standards
Repository, Midwest Research Institute (Kansas City, MO). Calf
thymus (CT) DNA, deoxyguanosine (dG), DNase I (Type IV,
bovine pancreas), alkaline phosphatase (from E. coli), and
nuclease P1 (from Penicillium citrinum) were purchased from
Sigma (St. Louis, MO). [13C10]dG was purchased from Cam-
bridge Isotopes (Andover, MA). Phosphodiesterase I (from
Crotalus adamanteus venom) was purchased from GE Health-
care (Piscataway, NJ). All solvents used were high-purity B &
J Brand from Honeywell Burdick and Jackson (Muskegon, MI).
ACS reagent-grade formic acid (88%) was purchased from J.T.
Baker (Phillipsburg, NJ). Hypersep filter SpinTips C-18 (20 mg)
were acquired from Thermo Scientific (Palm Beach, FL).
Preparation of the DNA Adduct Standards. N-(Deoxygua-
nosin-8-yl)-PhIP (dG-C8-PhIP),25,38 N-(deoxyguanosin-8-yl)-MeIQx
(dG-C8-MeIQx), and 5-(deoxyguanosin-N2-yl)-MeIQx (dG-N2-
,39
MeIQx),13 and N-(deoxyguanosin-8-yl)-2-amino-9H-pyrido[2,3-
b]indole (dG-C8-ARC)40 were prepared by reaction of their
N-acetoxy HAA derivatives with dG or [13C10]dG (5 mg) in 100
mM K2HPO4 buffer (pH 8.0). N-(Deoxyguanosin-8-yl)-4-ami-
nobiphenyl (dG-C8-ABP) was prepared by reaction of N-hy-
droxy-4-ABP with pyruvonitrile, followed by reaction with dG
or [13C10]dG.41 The (±)-anti-B[a]PDE-derived dG adducts, 10-
(deoxyguanosin-N2-yl)-7,8,9-trihydroxy-7,8,9,10-tetrahydroben-
zo[a]pyrene (dG-N2-BaP) were prepared by reaction of (±)-anti-
(42) Ruan, Q.; Kim, H. Y.; Jiang, H.; Penning, T. M.; Harvey, R. G.; Blair, I. A.
Rapid Commun. Mass Spectrom. 2006, 20, 1369–1380
(43) Nechev, L. V.; Kozekov, I. D.; Brock, A. K.; Rizzo, C. J.; Harris, T. M. Chem.
Res. Toxicol. 2002, 15, 607–613
(44) Nechev, L. V.; Harris, C. M.; Harris, T. M. Chem. Res. Toxicol. 2000, 13,
421–429
(45) Wang, H.; Rizzo, C. J. Org. Lett. 2001, 3, 3603–3605
(46) Beland, F. A.; Doerge, D. R.; Churchwell, M. I.; Poirier, M. C.; Schoket,
B.; Marques, M. M. Chem. Res. Toxicol. 1999, 12, 68–77
.
.
(36) Witze, E. S.; Old, W. M.; Resing, K. A.; Ahn, N. G. Nat. Methods 2007, 4,
.
798–806
(37) Mutlib, A.; Lam, W.; Atherton, J.; Chen, H.; Galatsis, P.; Stolle, W. Rapid
Commun. Mass Spectrom. 2005, 19, 3482–3492
(38) Lin, D.; Kaderlik, K. R.; Turesky, R. J.; Miller, D. W.; Lay, J. O., Jr.; Kadlubar,
F. F. Chem. Res. Toxicol. 1992, 5, 691–697
(39) Turesky, R. J.; Rossi, S. C.; Welti, D. H.; Lay, J. O., Jr.; Kadlubar, F. F.
Chem. Res. Toxicol. 1992, 5, 479–490
(40) Frederiksen, H.; Frandsen, H.; Pfau, W. Carcinogenesis 2004, 25, 1525–
1533
(41) Jones, C. R.; Sabbioni, G. Chem. Res. Toxicol. 2003, 16, 1251–1263
.
.
.
.
(47) Beland, F. A.; Churchwell, M. I.; Von Tungeln, L. S.; Chen, S.; Fu, P. P.;
.
Culp, S. J.; Schoket, B.; Gyorffy, E.; Minarovits, J.; Poirier, M. C.; Bowman,
E. D.; Weston, A.; Doerge, D. R. Chem. Res. Toxicol. 2005, 18, 1306–1315
(48) Langoue¨t, S.; Paehler, A.; Welti, D. H.; Kerriguy, N.; Guillouzo, A.; Turesky,
R. J. Carcinogenesis 2002, 23, 115–122
(49) Lum, A.; Le Marchand, L. Cancer Epidemiol. Biomarkers Prev. 1998, 7,
719–724
.
.
.
.
.
.
Analytical Chemistry, Vol. 81, No. 2, January 15, 2009 811

Enzyme Digestion and Solid Phase Extraction (SPE) of
DNA Adducts. The enzymatic digestion conditions used for the
hydrolysis of DNA employed DNase I for 1.5 h, followed by
incubation with nuclease P1 for 3 h, and then alkaline phosphatase
and phosphodiesterase for 18 h. Thereafter, three volumes of cold
DNA adducts. The MS3 scan mode was triggered if the
aglycone adduct [M + H - 116]+ was within the top 10 most
abundant fragment ions in the MS/MS product ion spectrum.
RESULTS AND DISCUSSION
C
2H5OH (200 proof) were added to the hydrolysis mixture.25
The efficacy of the data-dependent CNL-MS3 scan mode to
detect DNA adducts was evaluated with DNA samples obtained
from rat liver and human hepatocytes exposed to a single
carcinogenic agent, and from buccal cell DNA samples col-
lected from individuals who are exposed to numerous carci-
SPE DNA Adduct Enrichment Procedure. The C2H5OH/
DNA digest solution was centrifuged at 15 000g for 2 min. The
supernatant containing DNA adducts was removed and dried
by vacuum centrifugation. Samples were purified by SPE, using
SpinTips. The DNA digest extracts were resuspended in 10%
CH3OH in 0.1% HCO2H (0.25 mL) and applied to a SpinTip,
which was placed into a vacuum manifold. The SpinTip was
then washed with 10% CH3OH in 0.1% HCO2H (2 × 0.25 mL),
to remove nonmodified 2′-deoxynucleosides. For studies de-
signed to screen for 6- and 8-hydroxy-PdG, the SpinTip was
washed with 0.1% HCO2H (2 × 0.25 mL). The desired adducts
were eluted with CH3OH containing 0.1% HCO2H (0.2 mL) into
silylated glass insert capillary LC vials (Microliter Analytical
Supplies, Suwanee, GA). Samples were evaporated to dryness
by vacuum centrifugation and reconstituted in 1:1 DMSO/H2O
(20 µL).
LC-MS Parameters. Chromatography was performed with
an Agilent 1100 series capillary LC system (Agilent Technologies,
Palo Alto, CA) equipped with an Aquasil C18 column (0.32 mm ×
250 mm) from Thermo Fisher (Bellafonte, PA). Samples (2 µL)
were injected, and analytes were separated with a gradient. The
solvent conditions were held at 100% A (solvent composition, 0.01%
HCO2H and 10% CH3CN) for 2 min, followed by a linear
gradient to 100% B (solvent composition, 95% CH3CN containing
0.01% HCO2H) over 30 min at a flow rate of 6 µL/min. The MS
instrumentation was an LTQ mass spectrometer (ThermoElec-
tron, San Jose, CA). The Xcalibur version 2.07 software was
used for data manipulations. Analyses were conducted in the
positive ionization mode and employed an Advance nanospray
source from Michrom Bioresource Inc. (Auburn, CA). Repre-
sentative optimized instrument tuning parameters were as
follows: capillary temperature 225 °C; source spray voltage 3.5
kV; source current 2.8 µA; no sheath gas, sweep gas, or
auxiliary gas was employed; capillary voltage 40 V; tube lens
voltage 110 V; and in-source fragmentation 10 V. The normal-
ized collision energies were set at 24 and 34, and the isolation
widths were set at 4.0 and 1.5 Da, respectively, for the MS2
and MS3 scan modes. The activation Q was set at 0.35, and
the activation time was 30 ms. Helium was used as the collision
damping gas in the ion trap and was set at a pressure of 1
mTorr. One µscan was used for data acquisition. The automatic
gain control (AGC) settings were full MS target 30 000 and
MSn target 10 000, and the maximum injection time was 50
ms.
,50
nogenic agents in tobacco smoke and the diet.32 We moni-
tored DNA adducts of aromatic amines, HAAs, PAHs, and
aldehydes that form in tobacco smoke (4-ABP, ARC, PhIP, BaP,
and AC);50-52 DNA adducts of HAAs formed in cooked meats
(PhIP, MeIQx, ARC)32 and R,ꢀ-unsaturated aldehydes (AC, HNE),
which are produced endogenously through lipid peroxidation as
a consequence of oxidative stress.53 The structures of the DNA
adducts investigated are shown in Figure 1 (The chemical
structures of the carcinogens are provided in Supporting Informa-
tion, Figure S-1).
We evaluated the sensitivity and specificity of the data-
dependent CNL-MS3 scan mode to detect adducts in global
scanning or with the use of a targeted adduct ion mass list.
The monitoring of DNA adducts in the global scan mode is an
intensity-dependent MS/MS experiment; MS/MS acquistion
is carried out on the most abundant ions detected in the full-
scan mode. With the use of a targeted mass-list, the MS/MS
acquisition is automatically triggered once a listed adduct ion
is detected. The loss of the neutral dR fragment moiety in the
MS/MS scan mode [M + H - 116]+ is used to activate the
acquisition of the MS3 product ion spectra of the aglycone ion
[BH2+].
4-ABP-DNA Adduct Formation in Human Hepatocytes.
In the first application, we have assayed DNA adduct formation
in cultured human hepatocytes exposed to 4-ABP (10 µM). This
aromatic amine arises in tobacco smoke;51 it is an experimental
animal carcinogen and a human urinary bladder carcinogen.54 The
major adduction product of 4-ABP occurs at the C8 atom of dG
and the N-hydroxylated amine group of 4-ABP, to produce dG-
C8-ABP.54 Two other isomeric dG adducts have been character-
ized: a hydrazo linked adduct, N-(deoxyguanosin-N2-yl)-4-amino-
biphenyl (dG-N2-N4-ABP) and an adduct formed at the N2 atom
of dG and the C-3 atom of 4-ABP, 3-(deoxyguanosin-N2-yl)-4-
aminobiphenyl (dG-N2-ABP).54,55 The dA adduct has been
identified as N-(deoxyadenosin-8-yl)-4-aminobiphenyl (dA-C8-
ABP)55 (Figure 1).
Adducts of 4-ABP formed in hepatocytes were monitored by
scanning of the protonated ions [M + H]+ from the mass range
of 400 to 500 Da, a range that encompasses the molecular
masses of all known adducts of 4-ABP. The mass chromato-
grams of putative adducts in the full-scan, MS/MS, CNL [M
+ H - 116]+, and the data-dependent CNL-MS3 scan modes
CNL-MS3 Data Dependent Scanning. A signal threshold
of 200 counts was required in full-scan MS [M + H]+ and MS/
MS [M + H - 116]+ scan modes to trigger the acquisition of
the MS3 scan. The instrument scanned from 300 to 600 Da in
MS full scan mode and from 100 to 600 Da in CRM scan modes.
A window of 116 ± 0.5 Da was used to monitor the loss of the
dR moiety. The top four or seven ions were monitored in MS/
MS. For some studies, a mass list was used to monitor specific
(50) Hecht, S. S. Nat. Rev. Cancer 2003, 3, 733–744
(51) Patrianakos, C.; Hoffmann, D. J. Assoc. Off. Anal. Chem. 1979, 3, 150–154
.
.
.
(52) Matsumoto, T.; Yoshida, D.; Tomita, H. Cancer Lett. 1981, 12, 105–110
(53) Chung, F. L.; Nath, R. G.; Nagao, M.; Nishikawa, A.; Zhou, G. D.; Randerath,
K. Mutat. Res. 1999, 424, 71–81
(54) Beland, F. A.; Kadlubar, F. F. Environ. Health Perspect. 1985, 62, 19–30
.
.
.
(55) Swaminathan, S.; Hatcher, J. F. Chem. Biol. Interact. 2002, 139, 199–213
812 Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

Figure 1. Structures of DNA adducts under investigation.
Figure 2. Global CNL-MS3 data-dependent scanning of 4-ABP adducts in human hepatocytes, either (A) untreated or (B) 4-ABP-treated (10
µM). The upper chromatograms in each panel displays the full scan mode from 400 to 600 Da. The second chromatogram down is a collection
of all MS/MS spectral chromatograms. The third chromatogram displays all of the peaks that are filtered by the CNL [M + H]+ f [M + H -
116]+. The bottom chromatogram displays all of the peaks that were detected by the data-dependent CNL-MS3 scan mode. The respective tR
and area of response of the ion counts are shown.
are presented in Figure 2. The chromatograms in the full-scan
and MS/MS scan modes are complex. Processing of the MS/
MS data set with the CNL [M + H - 116]+ and data-dependent
CNL-MS3 filters considerably reduced background signals and
the number of detected peaks, resulting in greatly simplified
mass chromatograms. Three presumed DNA adducts, which
are minor constituents of the DNA digest, are detected in
treated hepatocytes but not in the control hepatocytes, when
the CNL [M + H - 116]+ and CNL-MS3 scan mode filters are
employed to process the data.
The masses of two compounds were observed at m/z 435 [M
+ H]+, a molecular mass consistent with an adduct formed from
the reaction of dG with HONH-ABP. The third adduct displayed
an ion at m/z 419 [M + H]+, a molecular mass consistent with
a reaction product formed between dA and HONH-ABP. The
major adduct was identified as dG-C8-ABP (tR ) 21.4 min), on
the basis of its MS3 product ion spectrum and its coelution
with the isotopically labeled [13C10]dG-C8-ABP (added prior to
the DNA digest at a level of 5 adducts per 108 DNA bases;
data not shown). On the basis of the response of the signal of
Analytical Chemistry, Vol. 81, No. 2, January 15, 2009 813

C8-PhIP25 or dG-C8-MeIQx (vide infra). The second-generation
product ions of guanyl-4-ABP at m/z 277.2 [BH2 - 42]+ and
m/z 249.2 [BH2 - 70]+ underwent fragmentation at MS4, to
produce the ion at m/z 195.2 as a prominent product ion
(Supporting Information, Figure S-2).
Both dG-N2-N4-ABP and dG-N2-ABP are plausible structures
of the isomeric dG adduct (tR ) 19.2 min). The MS3 product
ion spectrum of the aglycone [BH2]+ at m/z 319.3 displays a
prominent fragment ion at m/z 302.2, due to loss of NH3 and
several less abundant product ions (Figure 3B). Both guanyl-
N2-N4-ABP and guanyl-N2-ABP adducts can be expected to
undergo cleavage of the guanyl moiety to produce these
fragment ions. The structure of the adduct was tentatively
assigned as the dG-N2-N4-ABP isomer, on the basis of the MS4
product ion spectrum of the second-generation product ions
of the charged species at m/z 210.2 [BH2 - 109]+ (C4H3N3O)
(Supporting Information, Figure S-3 and Scheme S-1). The
fragment ion at m/z 141.1 [BH2 - 109 - 69]+ (C2H3N3) can be
formed by cleavage of the aniline ring at the C4 atom of 4-ABP
in the guanyl-N2-N4-ABP adduct, but this product ion seems
unlikely to arise from guanyl-N2-ABP at the MS4 scan stage.
The unambiguous characterization and definitive identification
of this adduct structure, by MS, will require a synthetic adduct
with stable 13C or 15N isotopes at defined positions of the
molecule.
Figure 3. CNL-MS3 product ion spectra of (A) dG-C8-ABP, (B)
proposed dG-N2-N4-ABP adduct, and (C) dA-C8-ABP.
the dG-C8-ABP relative to [13C10]dG-C8-ABP, the amount of dG-
C8-ABP formed was estimated to be at about 9.7 adducts per
106 DNA bases. Under the assumption that the ABP adducts
display similar ionization and dissociation efficiencies, when
analyzed by LIT/MS, the levels of the isomeric dG-ABP adduct
(tR ) 19.2 min) and the dA-ABP adduct (tR ) 23.9 min) were
∼10-fold lower levels than those of the dG-C8-ABP adduct.
Eight MS3 scans were acquired across the peak of dG-C8-ABP,
and four MS3 scans were acquired across the peaks of the
isomeric dG-ABP adduct and the dA-ABP adduct with CNL-
MS3, using global scanning.
The proposed structures of the adducts and the assignments
of the fragment ions and neutral losses in the MS3 and MS4 stage
spectra of 4-ABP dA and dG adducts (Figure 3) are tentative.
There are several possible isomeric adduct structures that could
undergo fragmentation to produce these product ions. An unam-
biguous elucidation of the structures of the adducts would require
exact mass measurements of the fragment ions, in conjunction
with NMR spectroscopy. The assigned structure of dG-C8-ABP
is supported by the MS3 product ion spectra of the aglycone
guanyl-C8-ABP (m/z 319.2) [BH2]+ (Figure 3A). Many of the
product ions were previously characterized by TSQ/MS/MS and
are attributed to fragmentation of the guanyl moiety.56 The product
ion formed at m/z 195.2 [BH2 - 124]+ (loss of C4H4N4O), occurs
through cleavage of the N7-C8 and C4-N9 bonds of the
The MS3 product ion spectrum of the dA-C8-ABP adduct
supports the assigned structure that has a linkage between the
C8 atom of dA and the 4-NH2 group of 4-ABP (Figure 3C). The
MS3 product ion spectrum aglycone [BH2]+ adduct at m/z 303.2
displays several fragments of the adenine moiety. The product
ion at m/z 195.1 [BH2 - 108]+ (C4H4N4) is proposed to occur
via cleavage of the N7-C8 and N9-C4 bonds to produce the
protonated 4-ABP containing a CN moiety. The second-
generation product ions of the adenyl-4-ABP adduct at m/z
276.2 [BH2 - 27]+ (HCN) underwent fragmentation at the MS4
stage to produce the 195.1 ion as the base peak in the product
ion spectra (Supporting Information, Figure S-4).
4-ABP-DNA adduct formation in human hepatocytes was
further investigated through the use of the CNL-MS3 scan mode
with a targeted mass list of adducts of dG-ABP at m/z 435.2
and dA-ABP at m/z 419.2. The CNL-MS3 dependent scan mode
with the use of a targeted mass list provided coverage of 4-ABP
adducts that was superior to the coverage of the global
scanning mode: the number of MS3 product ion spectra
acquired across the peaks doubled for dG-C8-ABP and tripled
for dG-N2-ABP and dA-C8-ABP. Two additional isomeric dA-
ABP adducts were detected (tR ) 17.6 and 22.6 min) (Support-
ing Information, Figure S-5). The MS3 product ion spectra of
the aglycones [BH2]+ are presented in Supporting Information
,
Figure S-6). The structures of these adducts remain to be
determined.
guanine base. This mechanism of fragmentation is shared by
,23,57
many dG-C8-AA and dG-C8-HAA adducts.11
However, the
MeIQx DNA Adduct Formation in Rat Liver. In the second
application, we explored the efficacy of the data-dependent CNL-
MS3 scan mode in detecting DNA adducts in the livers of rats
treated with MeIQx. This HAA, which is formed in cooked beef,
is a potent bacterial mutagen and an experimental animal
carcinogen.32 Two isomeric dG adducts N-(deoxyguanosin-8-
yl)-MeIQx (dG-C8-MeIQx) and 5-(deoxyguanosin-N2-yl)-MeIQx
ion is produced at relatively low abundance in the MS3 product
ion spectrum of dG-C8-ABP, as compared to the corresponding
fragment ion observed in the MS3 product ion spectrum of dG-
(56) Ricicki, E. M.; Soglia, J. R.; Teitel, C.; Kane, R.; Kadlubar, F.; Vouros, P.
Chem. Res. Toxicol. 2005, 18, 692–699
(57) Chiarelli, M. P.; Lay, J. O., Jr. Mass Spectrom. Rev. 1992, 11, 447–493
.
.
814 Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

Figure 4. Targeted data-dependent scanning of DNA adducts in liver of (A) untreated and (B) MeIQx (10 mg/kg) treated rats. The upper
chromatogram displays the full scan mode from 400 to 600 Da. The second chromatogram down is a collection of all MS/MS chromatograms.
The third chromatogram displays all the peaks at m/z 479.2 (dG-MeIQx) that are filtered by the CNL [M + H]+ f [M + H - 116]+. The fourth
chromatogram displays all the peaks at m/z 479.2 (dG-MeIQx) that are detected by the data-dependent CNL-MS3. The fifth chromatogram
displays all the peaks at m/z 463.2 (dA-MeIQx) that are filtered by the CNL [M + H]+ f [M + H - 116]+. The bottom chromatogram displays
all the peaks at m/z 463.2 (dA-MeIQx) that are detected by the data-dependent CNL-MS3. The respective tR and area of response of the ion
counts are shown.
(dG-N2-MeIQx) have been reported to form by reaction of
N-acetoxy-MeIQx with dG or DNA (Figure 1).39 These adducts
also have been identified in rats treated with MeIQx.13 Global and
targeted mass [M + H]+ list scanning modes were employed
to screen for putative adducts of MeIQx formed with dC (m/z
439.2), dT (m/z 454.2), and dA (m/z 463.2), as well as with dG
(m/z 479.2). The dG-C8-MeIQx adduct (tR ) 20.1 min),
previously estimated in these liver samples at 1.8 ± 0.5 adducts
per 107 DNA bases by TQ/MS/MS,13 was detected by global
scanning with CNL-MS3 (four MS3 scans acquired across the
peak; data not shown). The dG-N2-MeIQx adduct, previously
estimated at 1.4 ± 0.3 adducts per 107 DNA bases in these DNA
samples,13 was not found here, under global scanning condi-
tions. However, when the targeted mass list was employed for
screening of adducts, dG-C8-MeIQx (12 MS3 scans acquired
across the peak), as well as dG-N2-MeIQx (tR ) 15.8 min) (6
MS3 scans acquired across the peak) and a previously unre-
ported dA adduct of MeIQx (tR ) 17.6 min) (7 MS3 scans
acquired across the peak) were easily identified. The mass
chromatograms for adducts in full-scan, MS/MS, CNL ([M +
H - 116]+), and data-dependent CNL-MS3 scan modes of the
targeted mass list are presented in Figure 4. The high
background signals precluded the detection of these adducts in
the full-scan and MS/MS-scan modes, but all three adducts are
observed when the CNL ([M + H - 116]+ and CNL-MS3 filters
are employed for data analysis. The response of synthetic dG-
N2-MeIQx was 10-fold lower than the response of dG-C8-MeIQx
under these CID conditions, a fact which explains the relatively
weaker signal of dG-N2-MeIQx in the DNA samples. We
surmise that the dA-MeIQx adduct is present in rat liver DNA
at a level of several adducts per 108 unmodified DNA bases.
The MS3 product ion spectra acquired by the CNL-MS3 scan
mode and the proposed mechanisms of fragmentation of the
isomeric dG-C8-MeIQx and dG-N2-MeIQx adducts, and of the
dA adduct, are presented in Figure 5. The product ion spectra
of dG-MeIQx adducts, acquired by TQ/MS/MS, were previously
described.11,13 Many of the same product ions are observed with
the LIT/MS. The MS3 product ion spectra of dA-MeIQx
aglycone [BH2]+ at m/z 347.3 (Figure 5C) provided rich
structural information about the structure of the adduct. Bond
formation is proposed to occur between the N6 atom of adenine
and the C-5 atom of the MeIQx heteronucleus to form
5-(deoxyadenosin-N6-yl)-MeIQx (dA-N6-MeIQx) via a carbenium
ion intermediate of N-acetoxy-MeIQx.39 The MS3 product ion
reveals extensive fragmentation of the adenine molecule.
Characteristic fragment ions arising through cleavage of
Analytical Chemistry, Vol. 81, No. 2, January 15, 2009 815

up to 258 ng per cigarette,61 while the amounts of 4-ABP, PhIP,
,62
and BaP range between 2 and 15 ng/cigarette.51 Although
MeIQx has not been detected in tobacco smoke, it, along with
PhIP, is formed in well-done cooked meats at the low parts-per-
billion levels.32 The expression of mRNA and catalytic activities
of P450 enzymes, which bioactivate 4-ABP, BaP, and HAAs, have
been detected in buccal cells,63-65 leading us to believe that these
carcinogens can form DNA adducts in buccal cells.
Buccal cell DNA samples were spiked with 6-hydroxy-PdG,
8-hydroxy-PdG, dG-HNE, dG-C8-MeIQx, and dG-C8-ARC as mon-
omeric adducts. dG-C8-PhIP, dG-C8-ABP, and dG-N2-BaP were
added as carcinogen-modified CT-DNA. All of the adducts were
added at a level of 5 adducts per 108 DNA bases into a final
quantity of 100 µg of buccal cell DNA. Spiking of buccal cell
DNA with these adducts enabled us to confirm the efficiency
of enzyme hydrolysis of carcinogen-modified CT-DNA and the
recovery of DNA adducts and to assess the limit of detection
(LOD) of the method.
The reconstructed ion mass chromatograms of DNA adducts
in buccal cell samples mined by the data-dependent CNL-MS4 scan
filter are presented in Figure 6, panels A (spiked samples) and
B (unspiked samples). All of the adducts spiked into the buccal
DNA were detected, and high-quality MS3 product ion spectra
were acquired (Figure 7). In the unspiked DNA samples, only
the 6- and 8-hydroxy-PdG adducts were unequivocally identified.
The number of MS3 scans across the peaks in the spiked DNA
samples ranged from 6 scans for dG-N2-BaP to more than 30
scans for the 6- and 8-hydroxy-PdG adducts. The response of
the signal for the 6- and 8-hydroxy-PdG adducts (tR between
17 and 18 min) was 10-50-fold greater than the response of
any other adduct (Figure 6A); the response also exceeded the
signal of the pure standards spiked into the DNA by 15-fold.
Indeed, high levels of 6- and 8-hydroxy-PdG adducts were detected
in the unspiked buccal DNA of tobacco smokers (Figure 6, panel
B).
The 6- and 8-hydroxy-PdG adducts exist as mixtures of
epimers,66 which are not completely resolved under these
chromatographic conditions. The 6-hydroxy-PdG adduct elutes
first and is the minor component of the mixture. The isomers are
distinguished on the basis of their MS3 product ion spectra
(Figure 7).60 The fragment ion at m/z 164.1 [BH2 - 44]+
(C2H4O), due to loss of ethenol, is only seen in the MS3 product
ion spectrum of 8-hydroxy-PdG.60 The MS3 product ion spectra
of 6- and 8-hydroxy-PdG acquired from the unspiked samples
were in excellent agreement (data not shown). The 6- and
8-hydroxy-PdG adducts were estimated at levels of above 5
adducts per 107 unmodified DNA bases in buccal cell DNA.
These levels are similar to those levels previously measured
Figure 5. CNL-MS3 product ion spectra of (A) dG-N2-MeIQx, (B)
dG-C8-MeIQx, and (C) the proposed dA-C8-MeIQx.
adenine at the C5-C6 and N1-C2 bonds or at the C5-C6 and
C6-N1 produce protonated MeIQx containing a cyanamide
moiety at m/z 254.3 [BH2 - 93]+ (C4H3N3) or a cyano moiety
at m/z 239.2 [BH2 - 108]+ (C4H4N4); both moieties are attached
to the C-5 atom of the MeIQx heteronucleus. Recently, a dA-
N6 adduct of the structurally related HAA, 2-amino-3-meth-
ylimidazo[4,5-f]quinoline (IQ), was identified in CT-DNA, by
high-resolution QIT/MS.58 Many of the pathways of fragmenta-
tion of the dA-N6-MeIQx and dA-N6-IQ adducts are remarkably
similar.
DNA Adduct Formation in Buccal Cell DNA of Tobacco
Smokers. In the third example, we have examined the efficacy
of the data-dependent CNL-MS3 scan mode, to screen for
multiple classes of DNA adducts in buccal cell DNA from
tobacco smokers. The oral and nasal cavities are the first
portals of entry for genotoxicants present in diet and tobacco
smoke. There are more than 60 known carcinogens present
in cigarette mainstream smoke,50 and 32P-postlabeling studies
have shown that a plethora of adducts are present in DNA of
the oral mucosa of smokers and nonsmokers alike.59 We
conducted data-dependent CNL-MS3 monitoring of DNA ad-
ducts derived from several ubiquitous genotoxicants that occur
in tobacco smoke or grilled meats or that form endogenously,
as a byproduct of cellular oxidative stress. AC and HNE, highly
reactive R,ꢀ-unsaturated aldehydes, are formed endogenously
through lipid peroxidation.53 AC also arises in cigarette smoke
at relatively high levels, ranging from 18 to 98 µg per
cigarette.60 Other tobacco genotoxicants studied here are
present at much lower levels in tobacco. ARC occurs at levels
(61) Yoshida, D.; Matsumoto, T. Cancer Lett. 1980, 10, 141–149
(62) Manabe, S.; Tohyama, K.; Wada, O.; Aramaki, T. Carcinogenesis 1991, 12,
1945–1947
.
.
(63) Vondracek, M.; Xi, Z.; Larsson, P.; Baker, V.; Mace, K.; Pfeifer, A.; Tjalve,
H.; Donato, M. T.; Gomez-Lechon, M. J.; Grafstrom, R. C. Carcinogenesis
2001, 22, 481–488
(64) Spivack, S. D.; Hurteau, G. J.; Jain, R.; Kumar, S. V.; Aldous, K. M.; Gierthy,
J. F.; Kaminsky, L. S. Cancer Res. 2004, 64, 6805–6813
(65) Liu, Y.; Sundqvist, K.; Belinsky, S. A.; Castonguay, A.; Tjalve, H.; Grafstrom,
R. C. Carcinogenesis 1993, 14, 2383–2388
.
(58) Jamin, E. L.; Arquier, D.; Canlet, C.; Rathahao, E.; Tulliez, J.; Debrauwer,
.
L. J. Am. Soc. Mass Spectrom. 2007, 18, 2107–2118
(59) Foiles, P. G.; Miglietta, L. M.; Quart, A. M.; Quart, E.; Kabat, G. C.; Hecht,
S. S. Carcinogenesis 1989, 10, 1429–1434
(60) Zhang, S.; Villalta, P. W.; Wang, M.; Hecht, S. S. Chem. Res. Toxicol. 2007,
20, 565–571
.
.
.
(66) Stone, M. P.; Cho, Y. J.; Huang, H.; Kim, H. Y.; Kozekov, I. D.; Kozekova,
A.; Wang, H.; Minko, I. G.; Lloyd, R. S.; Harris, T. M.; Rizzo, C. J. Acc.
Chem. Res. 2008, 41, 793–804.
.
816 Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

Figure 6. Targeted data-dependent scanning in CNL-MS3 of DNA adducts in (A) spiked buccal cells and (B) unspiked buccal cells. The
chromatograms show the CNL-MS3 traces for 6-hydroxy-PdG and 8-hydroxy-PdG (m/z 324.2), dG-HNE (m/z 424.2), dG-C8-ABP (m/z 435.2),
dG-C8-ARC (m/z 449.2), dG-C8-MeIQx (m/z 479.2), dG-C8-PhIP (m/z 490.2), and dG-N2-BaP (m/z 570.2); the area of response of the ion
counts are shown.
Figure 7. CNL-MS3 product ion spectra of DNA adducts recovered from spiked buccal cells and the aglycone adduct structures with proposed
mechanisms of fragmentation at the MS3 scan stage. The product ions of guanyl-N2-BaP (m/z 454.2) that arise at m/z 257.2 and 285.2 are
proposed to occur by elimination of H2O, followed by the expulsion of CO from the tetrahydro-BaP-7,8,9-triol cation species (m/z 303.1) in the
ion trap.
in gingival tissue of tobacco smokers by 32P-postlabeling.67
Relatively high levels of 6- and 8-hydroxy-PdG also have been
There are four diastereomers of dG-HNE,66 three of which are
resolved under the current chromatographic conditions. The
aglycone of dG-HNE [BH2]+ at m/z 308.2 undergoes dehydration
,60
identified in human lung35 and brain68 by LC-MS.
Analytical Chemistry, Vol. 81, No. 2, January 15, 2009 817

at the MS3 scan stage, to produce the ion at m/z 290.2 [BH2-
18]+ as the base peak. A second product ion at m/z 152.1 [BH2-
156]+ (C9H16O2) is attributed to cleavage of the HNE moiety.
The occurrence of dG-HNE in unspiked buccal cell DNA is
suggested by the two CNL-MS3 scans acquired on the peak at
tR 25.6 min (Figures 6B); however, the signal is weak and
identification of the adduct is equivocal.
selective and few spurious peaks were detected in any of our
DNA samples under these scanning parameters.
At high levels of DNA modifcation (>2 adducts per 107 bases),
the response of adducts in MS and MS/MS scan modes are
well above the background signals in the global scanning mode,
and the efficiency of the data-dependent CNL-MS3 scan mode
is high: almost every MS/MS scan that produced the neutral
fragment loss of 116 Da triggered the acquistion of the MS3
product ion scan. However, at lower levels of adduct modifica-
tion, when the background signal is relatively high, adducts
can be overlooked in the global scan mode. The use of a
targeted mass list, instead of global scanning, provides superior
coverage of the adducts and results in a 2- to 3-fold increase in
the number of MS3 scans acquired across the peaks. Moreover,
the postacquisition mining of the MS/MS data with [M + H
- 116]+ and CNL-MS3 scan filters allowed us to identify other
DNA adducts present at low abundance for both 4-ABP and
MeIQx for which we did not possess standards. Two previously
unreported dA adducts of 4-ABP, a novel dA adduct of MeIQx,
and the isomeric dG-N2-MeIQx adduct were identified with
CNL-MS3, when a mass list was employed for screening.
The data-dependent CNL-MS3 scan mode used with a
targeted mass list was about 5-fold less sensitive than the MS/
MS3 scan mode. However, the CNL-MS3 filter provided simpler
chromatograms with a higher signal-to-noise ratio than does
the MS/MS3 scan mode that can often result in the detection
of numerous peaks. The use of reconstructed chromatograms
of selected ions of MS/MS3 data improves the signal-to-noise
for DNA adducts, but the selection of ions requires prior
knowledge about the identity and structure of the adduct and
There are four isomeric dG-N2-BaP adducts,42 which are
partially resolved. The aglycone of dG-N2-BaP [BH2]+ (m/z
454), undergoes fragmentation at the MS3 scan stage, to
produce the fragment ion at m/z 303.1 [BH2 - 151]+
(C5H5N5O), due to loss of guanine, and ions at m/z 285.2 [BH2
- 169]+ (C5H7N5O2) and m/z 257.2 [BH2 - 197]+ (C6H7N5O3),
attributed to loss of guanine and H2O, followed by the expulsion
,47
of CO.42 CNL-MS3 spectra were also successfully acquired
on dG-C8-ABP, dG-C8-MeIQx, dG-C8-PhIP, and dG-C8-ARC in
spiked buccal cell DNA. The CNL-MS3 product ion spectra of
dG-C8-PhIP and dG-C8-ARC are shown in Figure 7. The product
ion spectra of all these adducts display all of the fragment ions
and at the same relative abundances that are observed for the
synthetic standards.
Putative DNA adducts of 4-ABP and BaP have been detected
in oral mucosa of smokers and nonsmokers, when assayed by
immunohistochemistry (IHC), at levels above 1 adduct per 107
,70
unmodified DNA bases.69 In this present pilot study of six
smoking subjects, we have yet to unequivocally identify these
lesions by LIT/MS (Figure 6, panel B), even though our LOD
(<5 adducts per 108 DNA bases) is lower than the LOD reported
by IHC. We suspect that the IHC technique is detecting a
variety of lesions in addition to or other than dG-C8-ABP or
dG-N2-BaP adducts.69,70 Since LC-ESI/MS is a concentration-
dependent detector,71 the use of liquid chromatography columns
with smaller inner diameters and lower flow rates should further
lower the LOD of adducts detected by the CNL-MS3 scan mode.
Scanning Parameters that Influence CNL-MS3 Data Ac-
quisition and the LOD. An ultralow ion abundance of 200
counts was chosen as the threshold in signal to trigger MS/MS
and CNL-MS3 scan events and provided the maximum number
of MS3 scans across the peaks. The CNL-MS3 scan event was
initiated only if the aglycone adduct [M + H - 116]+ was within
the top 10 most abundant fragment ions in the MS/MS product
ion spectrum. Restriction of the [M + H]+ f [M + H - 116]+
transition to the top two or three most abundant ions in the
MS/MS spectrum served to increase the number of scans
acquired across the peaks for some DNA adducts present at
levels above 2 adduct per 107 bases, but this level of stringency
can also diminish the number of data-dependent CNL-MS3
scans or can even result in the omission of adducts that are
present at low abundance. Remarkably, the CNL-MS3 scan
mode, used with a MS/MS tolerance of 0.5 Da, was relatively
,72
the mechanisms of fragmentation.25 The number of scans
acquired across the peaks in the CNL-MS3 scan mode with a
targeted mass list also is greater than the number of scans
acquired by MS/MS3, when six or more adducts are monitored
in the same segment of the MS/MS3 scan mode, and provides
superior MS3 product ion spectral data. The mass list can be
expanded to at least 12 adducts without effective diminution
of the number of CNL-MS3 scans acquired across the peaks.
CONCLUSIONS
The rapid duty cycle and high sensitivity of the LIT/MS
instrument permits simultaneous screening for an array of DNA
adducts having diverse structures, using the data-dependent CNL-
MS3 scan mode. The MS3 spectra of adducts can be acquired
at levels of DNA modificaton below 5 adducts per 108 DNA
bases, with 10 µg of DNA assayed on the column. The ability
to screen for multiple adducts of different classes of carcinogens
and to acquire MS3 product ion spectra for proof of adduct
identity in a single chromatographic run represent major
,12
advances in MS-based DNA adduct screening techniques,.11
The data-dependent CNL-MS3 scan mode is a powerful postac-
quisition data-mining technique for discovery and structural
elucidation of DNA adducts, a technique which may be used
in preclinical drug safety assessment.73 With further refine-
(67) Nath, R. G.; Ocando, J. E.; Guttenplan, J. B.; Chung, F. L. Cancer Res. 1998,
58, 581–584
(68) Liu, X.; Lovell, M. A.; Lynn, B. C. Anal. Chem. 2005, 77, 5982–5989
(69) Zhang, Y. J.; Hsu, T. M.; Santella, R. M. Cancer Epidemiol. Biomarkers Prev.
1995, 4, 133–138
.
.
.
(70) Romano, G.; Mancini, R.; Fedele, P.; Curigliano, G.; Flamini, G.; Giovagnoli,
(72) Grollman, A. P.; Shibutani, S.; Moriya, M.; Miller, F.; Wu, L.; Moll, U.;
Suzuki, N.; Fernandes, A.; Rosenquist, T.; Medverec, Z.; Jakovina, K.; Brdar,
B.; Slade, N.; Turesky, R. J.; Goodenough, A. K.; Rieger, R.; Vukelic, M.;
M. R.; Malara, N.; Boninsegna, A.; Vecchione, A.; Santella, R. M.; Cittadini,
A. Anticancer Res. 1997, 17, 2827–2830
(71) Gangl, E. T.; Annan, M. M.; Spooner, N.; Vouros, P. Anal. Chem. 2001,
73, 5635–5644
.
Jelakovic, B. Proc. Natl. Acad. Sci. U.S.A 2007, 104, 12129–12134
(73) Baillie, T. A. Chem. Res. Toxicol. 2008, 21, 129–137
.
.
.
818 Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

ments, this scanning technique can have important applications
in human biomonitoring of genotoxicants.
of Health, the French Cancer Societies, or Bristol-Myers Squibb.
The advice of Dr. Bill Mahn (Thermo Fisher) on the data-
dependent CNL-MS3 analyses is greatly appreciated. We thank
Dr. Paul Vouros, Northeastern University, Dr. Carmelo Rizzo,
Vanderbilt University, and Dr. Dieter Drexler, Bristol-Myers
Squibb, for their comments on this manuscript.
ACKNOWLEDGMENT
The project was supported by Grant R21ES-014438 (R.J.T.,
E.E.B., I.Y.) and Program Project Grant PO1ES-05355 (I.D.K.,
R.J.T.) from the National Institute of Environmental Health
Sciences, Grant R01CA-122320 from the National Cancer Institute
(R.J.T.), and grants from the Ligue Contre le Cancer (the Cote
d’armor, Morbihan and Maine et Loire committees) and the
Re´gion Bretagne, France (S.L.). The content is solely the
responsibility of the authors and does not necessarily represent
the official views of the National Cancer Institute, the National
Institute of Environmental Health Sciences, the National Institutes
SUPPORTING INFORMATION AVAILABLE
Additional information as noted in text. This material is
available free of charge via the Internet at http://pubs.acs.org.
Received for review October 3, 2008. Accepted November
18, 2008.
AC802096P
Analytical Chemistry, Vol. 81, No. 2, January 15, 2009 819