High-performance liquid chromatography electrospray ionization tandem mass spectrometry for the detection and quantitation of pyrrolizidine alkaloid-derived DNA adducts in vitro and in viwo

Article abstract of DOI:10.1021/tx900402x

Pyrrolizidine alkaloid-containing plants are widespread in the world and are probably the most common poisonous plants affecting livestock, wildlife, and humans. Pyrrolizidine alkaloids require metabolism to exert their genotoxicity and tumorigenicity. We have determined that the metabolism of a series of tumorigenic pyrrolizidine alkaloids in Vitro or in ViVo generates a common set of (±)-6,7-dihydro-7-hydroxy-1-hydroxymethyl-5H-pyrrolizine (DHP)-derived DNA adducts that are responsible for tumor induction. The identification and quantitation of the DHP-derived DNA adducts formed in ViVo and in Vitro were accomplished previously by 32P-postlabeling/HPLC methodology. In this article, we report the development of a sensitive and specific liquid chromatography-electrospray ionization-tandem mass spectrometry (HPLC-ES-MS/MS) method to detect DHP-derived DNA adducts formed in Vitro and in ViVo. The method is used to quantify the levels of DHP-2′-deoxyguanosine (dG) and DHP-2′-deoxyadenosine (dA) adducts by multiple reaction monitoring (MRM) analysis in the presence of known quantities of isotopically labeled DHP-dG and DHP-dA internal standards. This HPLC-ES-MS/MS method is accurate and precise. When applied to liver samples from rats treated with the pyrrolizidine alkaloids riddelliine and monocrotaline, the method provided significant new information regarding the mechanism of DNA adduct formation.

Full text of DOI:10.1021/tx900402x

Chem. Res. Toxicol. 2010, 23, 637–652
637
High-Performance Liquid Chromatography Electrospray Ionization
Tandem Mass Spectrometry for the Detection and Quantitation of
Pyrrolizidine Alkaloid-Derived DNA Adducts in Vitro and in ViWo
Peter P. Fu,*,† Ming W. Chou,† Mona Churchwell,† Yuping Wang,† Yuewei Zhao,†
Qingsu Xia,† Gonc¸alo Gamboa da Costa,† M. Matilde Marques,‡ Frederick A. Beland,† and
Daniel R. Doerge†
National Center for Toxicological Research, Jefferson, Arkansas 72079, and Centro de Qu´ımica Estrutural,
Complexo I, Instituto Superior Te´cnico, UniVersidade Te´cnica de Lisboa, P-1049-001 Lisboa, Portugal
ReceiVed NoVember 2, 2009
Pyrrolizidine alkaloid-containing plants are widespread in the world and are probably the most common
poisonous plants affecting livestock, wildlife, and humans. Pyrrolizidine alkaloids require metabolism to
exert their genotoxicity and tumorigenicity. We have determined that the metabolism of a series of
tumorigenic pyrrolizidine alkaloids in Vitro or in ViVo generates a common set of (()-6,7-dihydro-7-
hydroxy-1-hydroxymethyl-5H-pyrrolizine (DHP)-derived DNA adducts that are responsible for tumor
induction. The identification and quantitation of the DHP-derived DNA adducts formed in ViVo and in
Vitro were accomplished previously by 32P-postlabeling/HPLC methodology. In this article, we report
the development of a sensitive and specific liquid chromatography-electrospray ionization-tandem mass
spectrometry (HPLC-ES-MS/MS) method to detect DHP-derived DNA adducts formed in Vitro and in
ViVo. The method is used to quantify the levels of DHP-2′-deoxyguanosine (dG) and DHP-2′-
deoxyadenosine (dA) adducts by multiple reaction monitoring (MRM) analysis in the presence of known
quantities of isotopically labeled DHP-dG and DHP-dA internal standards. This HPLC-ES-MS/MS method
is accurate and precise. When applied to liver samples from rats treated with the pyrrolizidine alkaloids
riddelliine and monocrotaline, the method provided significant new information regarding the mechanism
of DNA adduct formation.
Central Africa, the West Indies, China, Jamaica, Canada,
Europe, New Zealand, Australia, and the U.S., rendering it
highly possible that pyrrolizidine alkaloid-containing plants are
the most common poisonous plants affecting livestock, wildlife,
and humans (4, 5, 8, 9, 13-15).
Introduction
Pyrrolizidine alkaloids are heterocyclic compounds, most of
which are derived from esters of basic alcohols, known as the
necine bases. The hydrolyzed products of pyrrolizidine alkaloids
are a necine base and a necic acid. The structures and numbering
system of the several representative necine bases, platynecine,
retronecine, heliotridine, and otonecine that form the basis of a
variety of commonly studied pyrrolizidine alkaloids are shown
in Figure 1.
Pyrrolizidine alkaloids are phytochemicals that are constitu-
tively produced by plants as secondary metabolites for exerting
a defense mechanism against insect herbivores (1-3). As a
consequence, pyrrolizidine alkaloids are common constituents
of hundreds of plant species of different unrelated botanical
families distributed in many geographical regions in the world
(1, 4-11). Pyrrolizidine alkaloids are found in more than 12
higher plant families of the Angiosperms, among which three
families, Compositae (Asteraceae), Boraginaceae, and Legu-
minosae (Fabaceae), contain the most toxic pyrrolizidine
alkaloids. It has been reported that about 3% of the world’s
flowering plants contain toxic pyrrolizidine alkaloids (8, 11, 12).
More than 660 pyrrolizidine alkaloids and pyrrolizidine alkaloid
N-oxides have been identified in over 6,000 plants, and about
half of them exhibit hepatotoxic activity (4, 5, 13). Toxic
pyrrolizidine alkaloid-containing plants grow in South Africa,
Pyrrolizidine alkaloids are among the first naturally occurring
carcinogens identified in plants. In 1954, a pure pyrrolizidine
alkaloid, retrorsine, was found to induce liver tumors in
experimental animals (16). Subsequently, lasiocarpine and
riddelliine were found to induce liver tumors in National
Toxicology Program (NTP1) chronic tumorigenicity bioassays
(17, 18). Because of the concern about human exposure to
genotoxic and tumorigenic pyrrolizidine alkaloids, in 1989, the
International Programme on Chemical Safety (IPCS) determined
that pyrrolizidine alkaloids are a threat to human health and
safety (10). To date, all of the known tumorigenic pyrrolizidine
alkaloids are based upon a retronecine, heliotridine, or otonecine
structure (5) (Figure 2).
1 Abbreviations: DHR, dehydroretronecine or (-)-R-6,7-dihydro-7-
hydroxy-1-hydroxymethyl-5H-pyrrolizine; DHP, (()-6,7-dihydro-7-hydroxy-
1-hydroxymethyl-5H-pyrrolizine; PNK, cloned T4 polynucleotide kinase;
MN, micrococcal nuclease; SPD, spleen phosphodiesterase; dG, 2′-
deoxyguanosine; dA, 2′-deoxyadenosine; DHP-dG-1 and DHP-dG-2,
epimeric pairs of 7-(deoxyguanosin-N2-yl)dehydrosupinidine; DHP-dG-3,
9-(deoxyguanosin-N2-yl)dehydrosupinidine; DHP-dG-4, 5-(deoxyguanosin-
N2-yl)dehydrosupinidine; DHP-dA-1 and DHP-dA-2, epimeric pairs of
7-(deoxyadenosin-N6-yl)dehydrosupinidine; DHP-dA-3, x- (deoxyadenosin-
N6-yl)-dehydrosupinidine; DHP-dA-4, y-(deoxyadenosin-N6-yl)dehydrosu-
pinidine; HPLC-ES-MS/MS, high-performance liquid chromatography-
electrospray ionization-tandem mass spectrometry; NCTR, National Center
for Toxicological Research; NTP, National Toxicology Program; MRM,
multiple reaction monitoring; LOD, limit of detection.
* Corresponding author. Tel: +1 870-543-7207. Fax: +1 870-543-7136.
E-mail: peter.fu@fda.hhs.gov.
† National Center for Toxicological Research.
‡ Universidade Te´cnica de Lisboa.
10.1021/tx900402x  2010 American Chemical Society
Published on Web 01/15/2010

638 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
Figure 1. (A) Representative necine bases of pyrrolizidine alkaloids and (B) structures of (()-6,7-dihydro-7-hydroxy-1-hydroxymethyl-5H-pyrrolizine
(DHP); dehydroretronecine [(-)-(R)-6,7-dihydro-7-hydroxy-1-hydroxymethyl-5H-pyrrolizine (DHR)]; and dehydroheliotridine [(+)-(S)-6,7-dihydro-
7-hydroxy-1-hydroxymethyl-5H-pyrrolizine (DHH)].
Figure 2. Nomenclature and structures of representative tumorigenic retronecine-, heliotritridine-, and otonecine-type pyrrolizidine alkaloids.
Pyrrolizidine alkaloids themselves are not toxic and require
metabolic activation to exert their acute and chronic toxicity.
The mechanism of tumor formation induced by pyrrolizidine
alkaloids was not known until 2001, when it was determined
that riddelliine induces liver tumors (i.e., hemangiosarcoma in
male and female rats and male mice, and a lower incidence of
hepatocellular adenoma and carcinoma in male and female rats)
through a genotoxic mechanism mediated by (()-6,7-dihydro-
7-hydroxy-1-hydroxymethyl-5H-pyrrolizine (DHP)-derived DNA
adduct formation (19).
with the preferential induction of liver hemangiosarcomas by
riddelliine. Furthermore, metabolism of riddelliine by human
liver microsomes resulted in a metabolic pattern and DNA
adduct profile that were very similar to those formed in rat liver,
indicating that the results of in ViVo and in Vitro mechanistic
studies with experimental rodents are highly relevant to humans
(22). Since riddelliine induces liver tumors in rodents and the
DHP-derived DNA adducts are responsible for liver tumor
induction, these results suggest that riddelliine can be genotoxic
to humans via DHP-derived DNA adduct formation. Partly
because of these mechanistic findings, the NTP has classified
riddelliine as “reasonably anticipated to be a human carcinogen”
(23).
The levels of the DHP-derived DNA adducts correlated
closely with the tumorigenic potencies in rats fed different doses
of riddelliine (19-21). We subsequently determined that the
DHP-derived DNA adduct levels formed in ViVo in hepatic
endothelial cells, the cells of origin for the hemangiosarcomas,
were significantly greater than those in parenchymal cells (20).
These results indicated that the levels of riddelliine-induced
DNA adducts in specific populations of liver cells correlated
Subsequent mechanistic studies on the tumorigenic pyrrolizi-
dine alkaloids, retrorsine, monocrotaline, retronecine, clivorine,
lasiocarpine, and heliotrine, conducted under similar experi-
mental conditions, also generated a common set of DHP-derived
DNA adducts (24-29). While riddelliine, retrorsine, monocro-

DNA Adducts from Pyrrolizidine Alkaloids
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 639
HPLC grade. (-)-(R)-6,7-Dihydro-7-hydroxy-1-hydroxymethyl-5H-
pyrrolizine (DHR, dehydroretronecine) (Figure 1) was prepared by
barium hydroxide-catalyzed hydrolysis of monocrotaline followed
by dehydrogenation of the resulting retronecine with o-bromanil
(19). The 7R stereochemistry of DHR was established previously
by circular dichroism (CD) spectroscopic analysis (26).
Instrumentation. A Waters HPLC system (Milford, MA),
consisting of a Model 600 controller, a Model 996 photodiode array
detector, and a 600 pump, was used for the separation and
purification of the DHP-derived DNA adducts.
taline, and retronecine are retronecine-type pyrrolizidine alka-
loids, clivorine is an otonecine-type, and heliotrine and lasio-
carpine are heliotridine-type pyrrolizidine alkaloids (Figure 2).
The formation of the same DHP-derived DNA adducts from
the metabolism of all of these tumorigenic pyrrolizidine alkaloids
indicates that these DHP-derived DNA adducts are potential
biomarkers of pyrrolizidine alkaloid exposure and tumorigenicity
(5, 25, 30).
Pyrrolizidine alkaloids have long been found as contaminants
in many human food sources and may potentially cause
worldwide human health problems (1, 5, 6, 13, 31-39). A
particular concern is that tumorigenic pyrrolizidine alkaloids
have been found in Chinese herbal plants and herbal dietary
supplements sold in the U.S. (40), including comfrey, coltsfoot,
and borage. We recently reported that the DHP-derived DNA
adducts are formed in the livers of female F344 rats gavaged
orally with three dietary supplements (comfrey root extract,
comfrey compound oil, and coltsfoot root extract) or an extract
of a Chinese herbal plant, Flos farfara (Kuan Tong Hua) (40).
This indicates that pyrrolizidine alkaloid-containing herbal plants
and herbal dietary supplements may pose a human health risk.
DNA adducts are considered to be a biomarker for assessing
the risk of chemically induced carcinogenesis. We previously
developed a 32P-postlabeling/HPLC method that enabled us to
establish that the DHP-derived DNA adducts formed in ViVo
are potential biomarkers of pyrrolizidine alkaloid exposure and
tumorigenicity (41). With this method, we identified and
quantified the DHP-derived DNA adducts in rodents treated with
pure pyrrolizidine alkaloids (e.g., riddelliine, retrorsine, and
monocrotaline), Chinese herbal plants, and herbal dietary
supplements purchased in the U.S. and Canada. These DNA
adducts were also formed from the rat and mouse liver
microsomal metabolism of several different pyrrolizidine alka-
loids (22, 24-29, 40). We also determined that the same DHP-
derived DNA adducts were present in leukocytes of rats fed
riddelliine, suggesting that DHP-derived DNA adducts in blood
can serve as potential biomarkers of pyrrolizidine alkaloid
exposure (30).
1H-Nuclear magnetic resonance (NMR) experiments were carried
out at 301 K on a Bruker AM 500 spectrometer (Bruker Instruments,
Billerica, MA) operating at 500 MHz. Samples were dissolved in
DMSO-d6 or DMSO-d6 with a trace of D2O. NOE difference and
homonuclear decoupling experiments were conducted to assist in
assigning proton resonances.
Treatment of Rats with Pyrrolizidine Alkaloids. Procedures
involving the care and handling of rats were reviewed and approved
by the National Center for Toxicological Research (NCTR)
Laboratory Animal Care and Use Committee. Female F344 rats
were obtained from the NCTR breeding colony as weanlings and
maintained on a 12-h light-dark cycle. At 8-10 weeks of age,
the rats were orally gavaged for 3 consecutive days with riddelliine,
at daily doses of 0, 0.1, 1, 2, and 5 mg/kg body weight in 0.5 mL
of 75% DMSO, or with monocrotaline, at daily doses of 1 and 5
mg/kg in the same solvent and volume. Twenty-four hours after
the last dose, the rats were sacrificed by exposure to CO2 followed
by exsanguination. Livers were removed, rinsed with cold saline,
and stored at -70 °C before DNA was isolated for DNA adduct
analysis. DNA samples were extracted using RecoverEase DNA
Isolation Kits (Stratagene, Cedar Creek, TX) according to the
manufacturer’s instructions. The concentration of the DNA was
determined spectrophotometrically.
National Toxicology Program (NTP) Female F344 Rat
Liver Samples. Riddelliine-modified rat liver DNA was obtained
from female F344 rats treated by gavage with riddelliine in a study
conducted by the NTP (42). The rats had been treated under the
same experimental conditions used for the two year chronic
tumorigenicity study, with the exception that there were six animals
per dose group and that the animals were sacrificed after six months
of treatment. Briefly, female F344/N rats (7-8 weeks old) were
gavaged with 0, 0.01, 0.033, 0.1, 0.33, or 1.0 mg of riddelliine/kg
body weight in 0.1 M phosphate buffer five times a week and
continued until sacrifice at six months after the first dosing day.
After sacrifice, liver tissue was collected, stored at -78 °C, and
shipped by the NTP to NCTR for DNA adduct analysis. Rat liver
DNA was isolated according to the method of Beland et al. (43).
Synthesis of DHP-dG and DHP-dA Adducts. DHP-dG adducts
were prepared by an adaptation of the method of Robertson (44).
Briefly, to a solution of dG (53 mg, 0.2 mmol) in 10 mL of 25
mM potassium phosphate (pH 7.4) was added DHR (15 mg, 0.1
mmol) under argon over a period of 5 min. The reaction mixture
was heated at 37 °C for 15 h. After cooling, the reaction mixture
was filtered through a 0.22-µm Millipore filter to remove insoluble
materials. The solvent was removed under vacuum, and the resulting
residue was separated by HPLC using a Prodigy 5 µm ODS column
(Phenomenex, Torrance, CA, 4.6 mm × 250 mm), by eluting with
a methanol and water linear gradient (0-20 min, 20-30%
methanol; 30-50 min 30-50% methanol). The flow rate was 1.0
mL/min, and UV absorbance was monitored at 254 nm. DHP-dA
adducts were prepared similarly, with the exception that for the
HPLC separation, the UV wavelength was set at 266 nm.
The yield of DHP-dA adduct formation is very low. In order to
obtain sufficient amounts of pure DHP-dA adducts for structural
There are several limitations associated with 32P-postlabeling/
HPLC methods for DNA adduct analysis. These include (i) the
risk from exposure to high levels of radioactivity, (ii) DNA
adduct instability under the experimental conditions, (iii)
frequent problems with enzymatic reactions, and (iv) the lack
of structural information. In view of these limitations, it is
desirable to develop alternate methods for the detection and
quantitation of DHP-derived DNA adducts in ViVo and in Vitro,
in particular from human samples. In this article, we report the
successful development of an HPLC-ES-MS/MS methodology
for the identification and quantitation of DHP-derived DNA
adducts in ViVo and in Vitro.
Experimental Procedures
Caution: Riddelliine, monocrotaline, and DHP are carcinogenic
in laboratory animals. They should be handled with extreme care,
using proper personal protectiVe equipment and a well-Ventilated
hood.
Chemicals. Monocrotaline, o-bromanil, calf thymus DNA (so-
dium salt, type I), 2′-deoxyguanosine (dG), 2′-deoxyadenosine (dA),
glucose-6-phosphate, glucose-6-phosphate dehydrogenase, and nic-
otinamide adenine dinucleotide phosphate (NADP+) were purchased
from the Sigma Chemical Co. (St. Louis, MO). Riddelliine was
obtained from the National Toxicology Program (NTP). [15N5]dG
1
determination by H NMR spectroscopic analyses and for deter-
mination of their UV molar extinction coefficients, 8.0 g of
monocrotaline was used to prepare retronecine (2.8 g, 75% yield)
(19). Dehydrogenation of the retronecine (2.8 g) with o-bromoanil
produced DHR (2.14 g, 77% yield) (19), which when reacted with
dA (14 g) generated a mixture of four DHP-dA adducts. Upon
HPLC purification, pure DHP-dA-1 and DHP-dA-2 were obtained.
13
and [15N5, C10]dA were obtained from Spectra Stable Isotopes
(Columbia, MD). Alkaline phosphatase was acquired from Roche
Diagnostics Corporation (Indianapolis, IN). All solvents used were

640 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
On the basis of 1H NMR spectroscopic analysis in the presence of
a known quantity of tert-butanol as internal standard, 3.8 mg of
DHP-dA-1 (0.07% yield) and 3.6 mg of DHP-dA-2 (0.067% yield)
were obtained.
DHP-H6a), 3.82 (1H, m, dG-H4′), 3.84-3.90 (1H, m, DHP-H7b),
3.92-3.97 (1H, m, DHP-H7a), 4.26 (1H, t, J ) 14, DHP-H9b),
4.35 (1H, two sets of d, J ) 14.2, DHP-H9a), 4.35 (1H, m, dG-
H3′), 4.74 (1H, dd, J ) 6.4, 7.6, DHP-H5), 6.11 (1H, d, J ) 2.7,
DHP-H2), 6.17 (1H, m, dG-H1′), 6.68 (1H, two sets of d, J ) 2.7,
DHP-H3), 7.89 (1H, s, dG-H8).
1H NMR spectroscopic data for each of the isolated adducts
follow.
DHP-dA-1. 1H NMR (DMSO-d6) δ 2.23-2.31 (1H, m, dA-H2′′),
2.35-2.47 (1H, m, DHP-H6b), 2.68-2.78 (1H, m, dA-H2′), 2.85-2.93
(1H, m, DHP-H6a), 3.49-3.56 (1H, two sets of d, J ) 4.5, dA-H5′′),
3.60-3.66 (1H, two sets of d, J ) 4.5, dA-H5′), 3.85-3.93 (2H, m,
DHP-H5b, dA-H4′), 4.01-4.25 (3H, m, DHP-H5a, H9a, H9b),
4.38-4.45 (1H, m, dA-H3′), 4.48 (1H, bs, OH), 5.10-5.45 (2H, bs,
overlapped, OH), 5.78 (1H, bs, DHP-H7), 6.04-6.10 (1H, two sets
of d, J ) 2.5, DHP-H2), 6.35-6.38 (1H, dd, dA-H1′), 6.60-6.67 (1H,
two sets of d, J ) 2.5, DHP-H3), 8.07 (1H, bs, dA-N6H), 8.27 (1H,
bs, dA-H8), 8.35 (1H, s, dA-H2).
1
DHP-dG-1. H NMR (DMSO-d6) δ 2.21-2.25 (1H, m, dG-
H2′′), 2.34-2.38 (1H, m, DHP-H6b), 2.62-2.68 (1H, m, dG-H2′),
2.86-2.90 (1H, m, DHP-H6a), 3.49-3.54 (1H, m, dG-H5′′),
3.57-3.60 (1H, m, dG-H5′), 3.80-3.83 (1H, m, dG-H4′), 3.89-3.94
(1H, m, DHP-H5b), 3.99-4.05 (1H, m, DHP-H5a), 4.28 (1H, bd,
DHP-H9b), 4.32 (1H, bd, DHP-H9a), 4.38-4.40 (1H, m, dG-H3′),
4.51 (1H, bs, OH/NH), 4.87 (1H, bt, OH/NH), 5.25-5.28 (2H, m,
DHR-H7, OH/NH), 6.09 (1H, d, J ) 2.6, DHP-H2), 6.20 (1H, dd,
J ∼ 7, dG-H1′), 6.66 (1H, d, J ) 2.6, DHP-H3), 6.73 (1H, bs,
OH/NH), 7.93 (1H, s, dG-H8), 10.31 (1H, s, dG-NH1).
DHP-dA-1. 1H NMR (DMSO-d6/D2O) δ 2.25-2.31 (1H, m, dA-
H2′′), 2.33-2.47 (1H, m, DHP-H6b), 2.67-2.73 (1H, m, dA-H2′),
2.87-2.96 (1H, m, DHP-H6a), 3.48-3.56 (1H, two sets of d, J )
4.5, dA-H5′′), 3.59-3.62 (1H, two sets of d, J ) 4.5, dA-H5′),
3.84-3.92 (2H, m, DHP-H5b, dA-H4′), 4.04-4.23 (3H, m, DHP-
H5a, H9a, H9b), 4.37-4.43 (1H, m, dA-H3′), 5.72 (1H, bs, DHP-
H7), 6.04-6.06 (1H, two sets of d, J ) 2.5, DHP-H2), 6.34 (1H,
dd, dA-H1′), 6.59-6.65 (1H, two sets of d, J ) 2.5, DHP-H3)
8.25 (1H, s, dA-H8), 8.32 (1H, s, dA-H2).
DHP-dG-1. 1H NMR (DMSO-d6/D2O) δ 2.21-2.26 (1H, m, dG-
H2′′), 2.31-2.37 (1H, m, DHP-H6b), 2.61-2.67 (1H, m, dG-H2′),
2.83-2.90 (1H, m, DHP-H6a), 3.80-3.83 (1H, m, dG-H4′),
3.87-3.91 (1H, m, DHP-H5b), 3.99-4.03 (1H, m, DHP-H5a), 4.25
(1H, d, J ) 12, DHP-H9b), 4.29 (1H, d, J ) 12, DHP-H9a),
4.37-4.39 (1H, m, dG-H3′), 5.25-5.27 (1H, m, DHP-H7), 6.08
(1H, d, J ) 2.5, DHP-H2), 6.19 (1H, dd, J ∼ 7, dG-H1′), 6.65
(1H, d, J ) 2.5, DHP-H3), 7.91 (1H, s, dG-H8).
1
DHP-dG-2. H NMR (DMSO-d6) δ 2.21-2.26 (1H, m, dG-
DHP-dA-2. 1H NMR (DMSO-d6) δ 2.23-2.32 (1H, m, dA-H2′′),
2.34-2.45 (1H, m, DHP-H6b), 2.68-2.78 (1H, m, dA-H2′), 2.84-2.94
(1H, m, DHP-H6a), 3.47-3.56 (1H, m, dA-H5′′), 3.57-3.66 (1H, m,
dA-H5′), 3.81-3.94 (2H, m, DHP-H5b, dA-H4′), 4.04-4.25 (3H, m,
DHP-H5a, H9a, H9b), 4.41 (1H, bs, dA-H3′), 4.47 (1H, bs, OH), 5.25
(1H, bs, OH), 5.33 (1H, bs, OH), 5.78 (1H, bs, DHP-H7), 6.06 (1H,
d, J ) 2.0, DHP-H2), 6.35-6.38 (1H, dd, dA-H1′), 6.60-6.67 (1H,
two sets of d, J ) 2.0, DHP-H3), 8.06 (1H, bs, dA-N6H), 8.27
(1H, bs, dA-H8), 8.35 (1H, s, dA-H2).
H2′′), 2.34-2.39 (1H, m, DHP-H6b), 2.62-2.68 (1H, m, dG-H2′),
2.86-2.90 (1H, m, DHP-H6a), 3.49-3.54 (1H, m, dG-H5′′),
3.57-3.60 (1H, m, dG-H5′), 3.81-3.83 (1H, m, dG-H4′), 3.89-3.94
(1H, m, DHP-H5b), 4.00-4.05 (1H, m, DHP-H5a), 4.29-4.34 (1H,
m, DHP-H9b), 4.32-4.34 (1H, m, DHP-H9a), 4.38-4.40 (1H, m,
dG-H3′), 4.51 (1H, t, OH/NH), 4.87 (1H, t, OH/NH), 5.25-5.28
(2H, m, DHP-H7, OH/NH), 6.09 (1H, d, J ) 2.5, DHP-H2), 6.20
(1H, dd, J ∼ 7.1, dG-H1′), 6.66 (1H, d, J ) 2.5, DHP-H3), 6.73
(1H, bs, OH/NH), 7.92 (1H, s, dG-H8), 10.28 (1H, s, dG-NH1).
DHP-dG-2. 1H NMR (DMSO-d6/D2O) δ 2.21-2.23 (1H, m, dG-
H2′′), 2.34-2.36 (1H, m, DHP-H6b), 2.63-2.67 (1H, m, dG-H2′),
2.83-2.90 (1H, m, DHP-H6a), 3.80-3.83 (1H, m, dG-H4′),
3.87-3.92 (1H, m, DHP-H5b), 4.00-4.03 (1H, m, DHP-H5a), 4.26
(1H, d, J ) 12, DHP-H9b), 4.30 (1H, d, J ) 12, DHP-H9a),
4.36-4.38 (1H, m, dG-H3′), 5.25-5.27 (1H, m, DHP-H7), 6.08
(1H, d, J ) 2.5, DHP-H2), 6.19 (1H, dd, J ∼ 7, dG-H1′), 6.65
(1H, d, J ) 2.5, DHP-H3), 7.91 (1H, s, dG-H8).
DHP-dA-2. 1H NMR (DMSO-d6/D2O) δ 2.24-2.32 (1H, m, dA-
H2′′), 2.33-2.44 (1H, m, DHP-H6b), 2.62-2.74 (1H, m, dA-H2′),
2.85-2.96 (1H, m, DHP-H6a), 3.47-3.56 (1H, two sets of d, J )
4.0, dA-H5′′), 3.65 (1H, m, dA-H5′, overlapped with D2O),
3.84-3.92 (2H, m, DHP-H5b, dA-H4′), 4.02-4.23 (3H, m, DHP-
H5a, H9a, H9b), 4.37-4.43 (1H, m, dA-H3′), 5.74 (1H, bs, DHP-
H7), 6.03-6.09 (1H, two sets of d, J ) 2.0, DHP-H2), 6.32-6.35
(1H, dd, dA-H1′), 6.60-6.65 (1H, two sets of d, J ) 2.0, DHP-
H3), 8.25 (1H, s, dA-H8), 8.32 (1H, s, dA-H2).
DHP-dG-3. 1H NMR (DMSO-d6) δ 2.19-2.26 (2H, m, dG-H2′′,
DHP- H6b), 2.54-2.68 (2H, m, dG-H2′, DHP-H6a), 3.48-3.53
(1H, m, dG-H5′′), 3.57-3.59 (1H, m, dG-H5′), 3.81-3.85 (2H,
m, dG-H4′, DHP-H5b), 3.99-4.03 (1H, m, DHP-H5a), 4.23-4.28
(1H, m, DHP-H9b), 4.31-4.37 (1H, m, DHP-H9a), 4.35-4.38 (1H,
m, dG-H3′), 4.81 (1H, m, OH/NH), 5.06-5.08 (1H, m, DHP-H7),
5.14-5.15 (1H, m, OH/NH), 5.26-5.27 (1H, m, OH/NH), 6.09
(1H, d, J ) 2.6, DHP-H2), 6.18 (1H, m, dG-H1′), 6.34 (1H, m,
OH/NH), 6.61 (1H, d, J ) 2.5, DHP-H3), 7.91 (1H, s, dG-H8),
10.40 (1H, s, dG-NH1).
Synthesis of Isotopically Labeled DHP-dG and DHP-dA
13
Adducts. DHP-[15N5]dG and DHP-[15N5, C10]dA adducts were
prepared as described above by reacting DHR (3.0 mg, 0.02 mmol)
with [15N5]dG (5 mg, 0.02 mmol) and [15N5,13C10]dA (5 mg, 0.02
mmol), respectively.
Chemical Reaction of DHR with Calf Thymus DNA. To
prepare DNA standards modified with DHR, 5 mg of calf thymus
DNA in 2 mL of 5 mM Tris-HCl buffer, pH 7.1, containing 0.1
mM EDTA (Tris-EDTA) was reacted with 0, 0.1, 0.5, 1, 5, and 50
µg of DHR in 100 µL of methanol at 37 °C for 3 h. After the
incubation, the reaction mixture was extracted twice with 5 mL of
a chloroform/isoamyl alcohol mixture (24/1, v/v). The DNA in the
aqueous phase was precipitated by adding 250 µL of 3 M sodium
acetate followed by an equal volume of cold 2-propanol and washed
with 70% ethanol. After the DNA was dissolved in Tris-EDTA,
the DNA concentration was determined spectrophotometrically. The
DNA solution was stored at -78 °C prior to hydrolysis and LC-
ES-MS/MS analysis.
The deoxynucleoside adducts were quantified by HPLC-ES-MS/
MS after standard enzymatic hydrolysis of the DNA to deoxy-
nucleosides (43). The hydrolysis mixtures from 100 µg aliquots of
the DHR-modified DNA were analyzed directly, without any adduct
enrichment procedure.
HPLC-ES-MS/MS Analysis of DHP-Derived DNA Adducts
Formed in Vitro and in ViWo. Quantification of DHP-DNA
Standards. Purified samples of the DHP-dG-1 and DHP-dG2
adduct standards were quantified spectrophotometrically on the basis
of the UV molar extinction coefficients 1.70 × 104 and 1.72 × 104
DHP-dG-3. 1H NMR (DMSO-d6/D2O) δ 2.18-2.25 (2H, m, dG-
H2′′, DHP- H6b), 2.55-2.66 (2H, m, dG-H2′, DHP-H6a), 3.78-3.83
(2H, m, dG-H4′, DHP-H5b), 3.98-4.03 (1H, m, DHP-H5a), 4.24
(1H, two sets of d, J ) 14, DHP-H9b), 4.32 (1H, two sets of d,
DHP-H9a), 4.33-4.38 (1H, m, dG-H3′), 5.03-5.08 (1H, m, DHP-
H7), 6.07 (1H, d, J ) 2.7, DHP-H2), 6.19 (1H, m, dG-H1′), 6.60
(1H, d, J ) 2.5, DHP-H3), 7.89 (1H, s, dG-H8).
1
DHP-dG-4. H NMR (DMSO-d6) δ 2.18-2.24 (1H, m, dG-
H2′′), 2.34-2.37 (1H, m, DHP-H6b), 2.56-2.65 (2H, m, dG-H2′,
DHP-H6a), 3.49-3.53 (1H, m, dG-H5′′), 3.54-3.57 (1H, m, dG-
H5′), 3.82 (1H, m, dG-H4′), 3.86-3.92 (1H, m, DHP-H7b),
3.95-4.00 (1H, m, DHP-H7a), 4.25-4.32 (1H, m, DHP-H9b),
4.33-4.36 (1H, m, DHP-H9a), 4.33-4.37 (1H, m, dG-H3′), 4.76
(1H, dd, DHP-H5), 4.85 (1H, bs, OH/NH), 5.25 (1H, d, OH/NH),
6.12 (1H, d, DHP-H2), 6.18 (1H, m, dG-H1′), 6.43 (1H, bs, OH/
NH), 6.70 (1H, two sets of d, DHP-H3), 7.90 (1H, s, dG-H8), 8.56
(1H, s, OH/NH), 10.38 (1H, bs, dG-NH1).
DHP-dG-4. 1H NMR (DMSO-d6/D2O) δ 2.19-2.24 (1H, m, dG-
H2′′), 2.34-2.39 (1H, m, DHP-H6b), 2.56-2.65 (2H, m, dG-H2′,

DNA Adducts from Pyrrolizidine Alkaloids
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 641
Scheme 1. Synthetic Preparation of DHP-dG Adductsa
a The absolute configuration assignments at the C7 positions of the necine base of DHP-dG-1 and DHP-dG-2 are arbitrary.
M-1 cm-1 at 256 nm (pH 7.0) reported by Robertson (44). DHP-
dA-1 and DHP-dA-2 adduct standards were also quantified spec-
trophotometrically using the following experimentally determined
UV molar extinction coefficients: 2.47 ( 0.26 × 104 and 2.48 (
0.14 x104 M-1 cm-1, respectively, at 271 nm in water and 2.21 (
0.23 × 104 and 2.18 ( 0.12 × 104 M-1 cm-1, respectively, in
methanol at 270 nm. The concentrations of the isotopically labeled
adduct standards for DHP-dG-1, DHP-dG-2, DHP-dA-1, and DHP-
dA-2 were quantified by LC-UV (256 nm) by comparison to the
unlabeled standards. Isotopically labeled standards were not avail-
able for DHP-dG-3, DHP-dG-4, DHP-dA-3, and DHP-dA-4;
isotopically labeled DHP-dG-2 was used as an internal standard
for DHP-dG-3 and DHP-dG-4, and isotopically labeled DHP-dA-2
was used as an internal standard for DHP-dA-3 and DHP-dA-4.
Response factors (i.e., unlabeled standard response/internal standard
response) were determined by adding known amounts of labeled
and unlabeled standards to DNA hydrolysates.
DNA Hydrolysis. DNA samples (typically ∼100 µg) obtained
from a reaction of DHR with calf thymus DNA, rats treated with
riddelliine or monocrotaline, and rats from the NTP study were
enzymatically hydrolyzed to nucleosides with micrococcal nuclease,
spleen phosphodiesterase, and nuclease P1 as previously described
(43). After adding approximately 50 fmol of each internal standard,
the samples were injected on the HPLC for ES-MS/MS analyses.
Liquid Chromatography. The liquid handling system consisted
of an Alliance 2795 separations module (Waters, Milford, MA), a
Dionex GP40 quaternary gradient pump (Dionex, Sunnyvale, CA),
and an automated switching valve (TPMV, Rheodyne, Cotati, CA).
The switching valve allowed for online sample cleanup. The valve
was used to divert the trap column effluent to either waste or to
the analytical column. The Alliance 2795 system was used for
sample injection, sample cleanup, and regeneration of the trap
column. The Dionex pump was used to back-flush the trap column
to the analytical column during analysis and to keep a constant
flow of mobile phase going into the mass spectrometer during the
sample loading and preparation periods.
column was equilibrated with the starting mobile phase. The total
run time for sample preparation and analysis was 34 min.
Mass Spectrometry. A Quattro Ultima quadrupole mass spec-
trometer (Waters, Milford, MA.), equipped with an ES interface,
was used with a source block of 100 °C and a desolvation
temperature of 400 °C. Nitrogen was the desolvation (750 L/h),
cone gas (100 L/h), and nebulizing gas. Argon was the collision
gas, at a collision cell pressure of 2.2 × 10-3 mBar. Positive ions
were acquired in the multiple reaction monitoring (MRM) mode
(dwell time of 0.2 s and interchannel delay of 0.03 s). The DHP-
dG adducts were monitored at the [(M + H)+] (m/z 403) to m/z
269 transition and the dG internal standards at the (M + H)+ (m/z
408) to m/z 274 transition. The DHP-dA adducts were monitored
at the [(M + H)+] (m/z 387) to m/z 253 transition and the dA
internal standards at the (M + H)+ (m/z 402) to m/z 263 transition.
The cone voltage was 25 V, and the collision energy was 14 eV
for all of the transitions. Samples were quantified by comparing
the areas of the unlabeled chromatogram peaks to those of the
corresponding labeled internal standard chromatogram peaks.
Results
Synthesis of DHP-dG Adducts. DHP-dG adducts were
synthesized by the reaction of DHR with dG (Scheme 1). The
resulting reaction products were purified by HPLC, first with a
Whatman ODS-3 column, to remove excess dG (data not
shown), and then with a Prodigy 5 µm ODS column (Figure
3A). The materials contained in the chromatographic peaks
eluted at 18.2 (DHP-dG-1) and 19.5 (DHP-dG-2) min showing
baseline separation and exhibited UV-visible absorption spectra
(Figure 4A) that were identical to those reported by Robertson
(44). The whole scan mass spectra of these compounds revealed
that both had protonated molecules (M + H)+ at m/z 403 (data
not shown). These products were also characterized by ES/MS
positive product ion spectra, which showed transition ions at
m/z 269 (DHP-guanine-OH), m/z 152 (guanine), and m/z 136
(DHP) (Figure 5).
Each sample was loaded onto a reverse phase trap column [Luna
C18(2), 2 mm × 30 mm, 3 µm, Phenomenex, Torrance, CA] with
95% water and 5% acetonitrile at a flow rate of 0.2 mL/min for
3.5 min. After switching the divert valve, the concentrated sample
zone was back-flushed from the trap column onto the analytical
column [Luna C18(2), 2 mm × 150 mm, 3 µm, Phenomenex] with
a water/acetonitrile gradient at 0.2 mL/min, and the sample
components were eluted into the mass spectrometer. The gradient
consisted of 10% acetonitrile for 2 min followed by a linear gradient
up to 20% acetonitrile over 20 min, and then reset to the initial
conditions. After 3 min, the valve was switched back to the load
position, and the trap column was cleaned of waste with 70% of
acetonitrile for 10 min at a flow rate of 0.5 mL/min. Then the trap
The CD spectra of DHP-dG-1 and DHP-dG-2 were similar
to those reported by Robertson (44). The CD Cotton effects of
DHP-dG-1 were approximate mirror images of those of DHP-
dG-2 (Figure 6A), indicating that these adducts were a pair of
epimers differing in the configuration of the stereocenter of the
1
DHR chromophore. The H NMR spectra of DHP-dG-1 and
DHP-dG-2 showed characteristic downfield shifts (ca. 0.27 ppm)
of H7 compared to H7 in DHR itself (Table 1), as reported by
Robertson (44). By contrast, the location of the resonances for
H2, H3, H5a, H5b, H9a, and H9b showed smaller downfield

642 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
Figure 3. Reversed-phased HPLC profiles of (A) DHP-dG adducts and (B) DHP-dA adducts formed from reactions of DHR with 2′-deoxyguanosine
(dG) and 2′-deoxyadenosine (dA), respectively. For HPLC conditions, see Experimental Procedures.
ppm) compared to the corresponding protons in DHR, which
indicated proximity to the substitution site. These data were
consistent with adduct formation through substitution at H7 of
the DHP moiety. Each of the deoxyribose protons was clearly
evident in both DHP-dG-1 and DHP-dG2. In addition, reso-
nances could be assigned to H8 and N1H of dG, which excluded
substitution at C8, N1, or O6 of the dG moiety. On the basis of
the UV-visible absorption, mass, CD, and NMR spectra, the
structures of DHP-dG-1 and DHP-dG-2 were characterized as
a pair of epimeric DHP-dG adducts in which reaction had
occurred between C7 of DHP and N2 of dG (i.e., (()-7-
(deoxyguanosin-N2-yl)dehydrosupinidine) (Scheme 1).
The materials eluting at 25.5 (DHP-dG-3) and 25.8 (DHP-
dG-4) min (Figure 3A) had UV-visible absorption spectra
(Figure 4A) identical to those of DHP-dG-1 and DHP-dG-2.
Whole scan mass spectra revealed that the adducts had proto-
nated molecules (M + H)+ at m/z 403 (data not shown). Similar
to the mass spectrometric pattern of DHP-dG-1 and DHP-dG-
2, product ion mode mass spectrometric analysis indicated that
both DHP-dG-3 and DHP-dG-4 had transition ions at m/z 269
(DHP-guanine-OH), m/z 152 (guanine), and m/z 136 (DHP)
(Figure 5).
1
In contrast to DHP-dG-1 and DHP-dG-2, the H NMR
spectrum of DHP-dG-3 had a resonance for H7 of the DHP
moiety that was nearly identical to H7 of DHR (Table 1 and
Figure 7); likewise, H2, H3, and both pairs of geminal H5
and H6 protons were almost unaffected by adduct formation.
In addition, both H9a and H9b appeared as a set of mutually
coupled doublets (Table 1), which suggested significant
prochirality. As with DHP-dG-1 and DHP-dG-2, resonances
could be assigned to each of the deoxyribose protons, H8,
and N1H of dG. These NMR data indicated that reaction had
occurred at C9 of DHR through the exocyclic nitrogen of
dG, an interpretation supported by a fragment at m/z 164 in
the product ion mass spectrum, which is consistent with a
protonated guanine bearing a methylene group at the N2
position. Thus, on the basis of the UV-visible, mass, and
Figure 4. UV-visible absorption spectra of (A) 1, DHP-dG-1; 2, DHP-
dG-2; 3, DHP-dG-3; and 4, DHP-dG-4; and (B) 1, DHP-dA-1; 2, DHP-
dA-2; 3, DHP-dA-3; and 4, DHP-dA-4. Spectra were obtained in 25%
methanol in water.
shifts from the same resonances in DHR. Moreover, both
geminal H6 protons were substantially deshielded (ca. 0.18-0.26

DNA Adducts from Pyrrolizidine Alkaloids
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 643
Figure 5. Electrospray positive product ion mass spectra of the four DHP-dG adduct standards (A ) DHP-dG-1, B ) DHP-dG-2, C ) DHP-dG-3,
and D ) DHP-dG-4). The product ions of the (M + H)+ m/z 403 were acquired with a cone voltage of 25 V and a collision energy of 30 eV.
The 1H NMR spectrum of DHP-dG-4 had a single C-5 proton,
which was shifted downfield by approximately 0.8-1 ppm
relative to H5 in DHR and the other DHP-dG adducts (Table 1
and Figure 7). In contrast to the other adducts, there were two
H7 resonances in DHP-dG-4, and these were shifted upfield by
approximately 1-1.3 ppm compared to that of the H7 resonance
in DHR and the other DHP-dG adducts. As with the other DHP-
dG adducts, resonances were clearly evident for each of the
deoxyribose protons, H8, and N1H of dG. Thus, on the basis
of its UV-visible, mass, and 1H NMR spectroscopic data, DHP-
dG-4 was characterized as 5-(deoxyguanosin-N2-yl)dehydrosu-
pinidine.
The UV molar extinction coefficients of DHP-dG-1 and DHP-
dG-2 were previously reported by Robertson (44). On the basis
of the similarity of the UV spectra (Figure 4A) and assuming
that the molar extinction coefficients of DHP-dG-3 and DHP-
dG-4 are identical to those of DHP-dG-1 and DHP-dG-2, the
yields of these four adducts in the crude reaction mixture of
DHR with dG were estimated from the corresponding HPLC
peak areas as follows: DHP-dG-1, 0.89%; DHP-dG-2, 0.69%;
DHP-dG-3, 0.15%; and DHP-dG-4, 0.06%. Because of the
unstable nature of pyrrolic compounds, these adducts, particu-
larly DHP-dG-3 and DHP-dG-4, decomposed during HPLC
purification.
Synthesis of DHP-dA Adducts. DHP-dA adducts were
synthesized by the reaction of DHR with dA (Scheme 2) and
purified by HPLC, first with a Whatman ODS-3 column (not
shown) and then with a Prodigy 5 µm ODS column (Figure
3B). Whole scan mass spectrometric analysis revealed that the
materials contained in the chromatographic peaks eluted at 28.5
(DHP-dA-1), 29.5 (DHP-dA-2), 33.4 (DHP-dA-3), and 33.9
(DHP-dA-4) min had protonated molecules (M + H)+ at m/z
Figure 6. CD spectra of (A) DHP-dG-1 and DHP-dG-2 in methanol,
387 (data not shown); electrospray positive product ion mass
spectrometric analysis indicated transition ions at m/z 253 (DHP-
adenine-OH) and m/z 136 (DHP) (Figure 8), which suggested
that they were DHP-dA adducts.
and (B) DHP-dA-1 and DHP-dA-2 in water.
1H NMR spectroscopic data, DHP-dG-3 was characterized
as 9-(deoxyguanosin-N2-yl)dehydrosupinidine (Figure 7).

644 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
DHP-dA-1 and DHP-dA-2 had identical UV-visible absorp-
1
tion spectra (Figure 4B) and highly similar H NMR spectra
(Figure 9). As shown in Figure 6B, the CD Cotton effects of
DHP-dA-1 were an approximate mirror image of those of DHP-
dA-2, indicating that these adducts were a pair of epimers
differing in the configuration of one stereocenter in the DHR
1
chromophore. Similar to DHP-dG-1 and DHP-dG-2, the H
NMR spectra of DHP-dA-1 and DHP-dA-2 showed character-
istic downfield shifts (ca. 0.73 ppm) of H7 compared to that of
H7 in DHR (Figure 9 and Table 1). Resonances could be
assigned to H2, H3, H5a, H5b, H6a, H6b, H9a, and H9b. With
the exception of H9a and H9b, which were shielded by
approximately 0.2 ppm, these resonances were very similar to
those observed for DHP-dG-1 and DHP-dG-2. These data are
consistent with substitution through C7 of the DHR moiety, with
positioning of the methylene C9 protons in the shielding zone
of the adenine ring. Each of the deoxyribose protons was clearly
evident for the dA moiety. In addition, resonances could be
assigned to H2 and H8, and the N6H resonance appeared as a
single proton. On the basis of the UV-visible absorption, mass,
CD, and NMR spectra, the structures of DHP-dA-1 and DHP-
dA-2 were characterized as a pair of epimeric DHP-dA adducts
in which the reaction had occurred between C7 of DHP and N6
of dA [i.e., (()7-(deoxyadenosin-N6-yl)dehydrosupinidine]
(Scheme 2).
1
There was an insufficient amount of material for H NMR
spectroscopic characterization of DHP-dA-3 and DHP-dA-4. On
the basis of their mass and UV-visible spectroscopic charac-
teristics, DHP-dA-3 and DHP-dA-4 were tentatively assigned
as dA-dehydrosupinidine adducts. With the assumption that
DHP-dA-3 and DHP-dA-4 have molar extinction coefficients
similar to those of DHP-dA-1 and DHP-dA-2, upon HPLC
analysis, the crude yields of these adducts were estimated as
follows: DHP-dA-1, 0.1%; DHP-dA-2, 0.1%; DHP-dA-3,
0.01%; and DHP-dA-4, 0.01%.
Synthesis of Isotopically Labeled DHP-[15N5]dG and
DHP-[15N5, 13C10]-dA. DHP-[15N5]dG and DHP-[15N5, 13C10]dA
adducts were similarly prepared by reacting DHR with [15N5]dG
13
and [15N5, C10]dA, respectively. DHP-[15N5]dG-1, DHP-
[15N5]dG-2, DHP-[15N5]dG-3, and DHP-[15N5]dG-4 showed a
(M + H)+ at m/z 408 and a transition ion at m/z 274 (DHR-
13
[15N5]guanine-OH) (data not shown). DHP-[15N5, C10]dA-1,
DHP-[15N5, 13C10]dA-2, DHP-[15N5, 13C10]dA-3, and DHP-[15N5,
13
C ]dA-4 had a (M + H)+ at m/z 402 and a transition ion at
10
m/z 263 (data not shown).
Standard Characterization and Calibration Curves. HPLC-
ES-MS/MS calibration curves were generated for synthetically
prepared DHP-dG-1, DHP-dG-2, DHP-dA-1, and DHP-dA-2
versus their respective [15N5] and [15N5, 13C10] internal standards.
Response factors were generated from these curves. The curves
for the unlabeled compounds went from 10 to 500 fmol on-
13
column versus 50 fmol of the [15N5] and [15N5, C10] internal
standards. The curves were linear with r2 > 0.999.
[15N5] adduct standards did not exist for DHP-dG-3, DHP-
dG-4, DHP-dA-3, and DHP-dA-4. To generate response factors
for these adducts, 100 µg aliquots of hydrolyzed DNA were
fortified with the following mixtures: 50 fmol DHP-dG-3 and
55 fmol DHP-[15N5]dG-2; 25 fmol DHP-dG-4 and 55 fmol
DHP-[15N5]dG-2; 50 fmol DHP-dA-3 and 82 fmol DHP-[15N5,
13
C ]dA-2; and 134 fmol DHP-dA-4 and 82 fmol DHP-[15N5,
10
10
13
C ]dA-2. The final response factors obtained were as follows:
DHP-dG-1 vs DHP-[15N5]dG-1 ) 1.18; DHP-dG-2 vs DHP-
[15N5]dG-2 ) 1.20; DHP-dG-3 vs DHP-[15N5]dG-2 ) 2.10;
DHP-dG-4 vs DHP-[15N5]dG-2 ) 1.50; DHP-dA-1 vs DHP-

DNA Adducts from Pyrrolizidine Alkaloids
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 645
1
Figure 7. H NMR spectra of DHP-dG-3 and DHP-dG-4 measured in DMSO-d6.
Scheme 2. Synthetic Preparation of DHP-dA Adductsa
a The absolute configuration assignments at the C7 positions of the necine base of DHP-dA-1 and DHP-dA-2 are arbitrary.
[15N5, 13C10]dA-1 ) 0.93; DHP-dA-2 vs DHP-[15N5, 13C10]dA-
The levels of DHP-dG and DHP-dA adducts formed from
13
2 ) 0.91; DHP-dA-3 vs DHP-[15N5, C10]dA-2 ) 6.50; and
each of the reactions are tabulated in Table 2. These results
indicate that (i) DHP-dG and DHP-dA adducts were formed in
a dose-responsive manner; (ii) DHP-dA-3 and DHP-dA-4 were
the most abundant, followed by all of the DHP-dG adducts,
which were formed to a similar extent, and DHP-dA-1 and DHP-
dA-2 were formed to the least extent; and (iii) DHP-dA and
DHP-dG adducts could be detected even at the lowest treatment
group (0.1 µg of DHR reacted with 5 mg of calf thymus DNA).
13
DHP-dA-4 vs DHP-[15N5, C10]dA-2 ) 4.2.
HPLC-ES-MS/MS Analysis of DHP-dG and DHP-dA
Adducts Formed from the Reaction of Calf Thymus DNA
and DHR. DHP-DNA adducts were prepared by reactions of 5
mg of calf thymus DNA in 2 mL of 5 mM Tris-HCl buffer, pH
7.1, containing 0.1 mM EDTA with 0, 0.1, 0.5, 1, 5, and 50 µg
of DHR in 100 µL of methanol at 37 °C for 3 h. After enzymatic
hydrolysis of the resulting reaction mixtures, the quantity of
DHP-dG and DHP-dA adducts in each reaction was determined
by HPLC-ES-MS/MS through multiple reaction monitoring
(MRM) combined with the use of DHP-[15N5]dG and DHP-
HPLC-ES-MS/MS Analysis of DHR-dG and DHR-dA
Adducts Formed in ViWo. Female rats were dosed daily with
riddelliine (0.1, 1.0, 2.0, and 5.0 mg/kg body weight) or
monocrotaline (1.0 and 5.0 mg/kg body weight) for three
consecutive days. One day after the last treatment, DNA was
isolated from the livers for the analysis of DHP-dG and DHP-
dA adducts by HPLC-ES-MS/MS. Representative MRM
chromatograms of liver DNA from a rat dosed with 5 mg of
monocrotaline/kg body weight/day and a rat dosed with 5
mg of riddelline/kg body weight/day are shown in Figure
13
[15N5, C10]dA adducts as internal standards. Representative
multiple reaction monitoring (MRM) measurements for DHP-
dG and DHP-dA adducts formed by reacting 5 mg of calf
thymus DNA with 5 µg of DHR are shown in Figure 10, which
clearly indicates the presence of all four DHP-dG and DHP-
dA adducts.

646 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
Figure 8. Electrospray positive product ion mass spectra of the four DHP-dA adduct standards (A ) DHP-dA-1, B ) DHP-dA-2, C ) DHP-dA-3,
and D ) DHP-dA-4). The product ions of the MH+ m/z 387 were acquired with a cone voltage of 25 V and a collision energy of 30 eV.
1
Figure 9. H NMR spectra of DHP-dA-1 and DHP-dA-2 measured in DMSO-d6.
11. The data, which are summarized in Table 3, indicate that
for both riddelliine and monocrotaline (i) DHP-dG-3 and
DHP-dG-4 were formed in a dose-responsive manner and to
a greater extent than DHP-dA-3 and DHP-dA-4; and that (ii)
with the exception of DHP-dA-2, which was found in liver
DNA from rats treated with 5 mg of riddelliine/kg body
weight/day, DHP-dG-1, DHP-dG-2, DHP-dA-1, and DHP-
dA-2 were not detected in any treatment groups.
HPLC-ES-MS/MS Analysis of DHP-dG and DHP-dA
Adducts Formed in Rat Liver Samples from an NTP
Study. In a previous study, we used the 32P-postlabeling
methodology to determine the levels of DHP-derived DNA

DNA Adducts from Pyrrolizidine Alkaloids
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 647
Figure 10. Representative MRM chromatograms (relative signal intensity vs time) for DHP-dG and DHP-dA adducts formed during the in Vitro
reaction of 5 mg of calf thymus DNA with 5 µg of DHR. Internal standards (IS) were added in amounts of 50, 55, 50, and 80 fmol for DHP-dG-1,
DHP-dG-2, DHP-dA-1, and DHP-dA-2, respectively.
Table 2. Levels of DHP-dG and DHP-dA Adducts Formed from Reactions of 5 mg of Calf Thymus DNA in 2 mL of 5
mMTris-HCl Buffer, pH 7.1, Containing 0.1 mM EDTA and 0, 0.1, 0.5, 1, 5, or 50 µg of DHR in 100 µL of Methanol for 3 h
sample
levels of DHP-dG and DHP-dA/108 nucleotides
DHR DNAa
DHP-dG-1
DHP-dG-2
DHP-dG-3
DHP-dG-4
DHP-dA-1
DHP-dA-2
DHP-dA-3
DHP-dA-4
50 µg
5 µg
213.5
24.6
15.3
<LOD
<LOD
<LOD
1.5
164.6
15.5
10.4
<LOD
<LOD
<LOD
1.5
143.0
23.0
19.4
5.4
1.3
<LOD
1.2
180.6
30.1
23.1
5.8
<LOD
<LOD
1.5
20.1
2.3
1.9
<LOD
<LOD
0.3
21.9
3.5
3.0
0.3
<LOD
0.1
266.0
42.7
27.3
4.4
1.2
<LOD
0.1
366.0
60.7
37.8
6.3
1.6
<LOD
0.1
1 µg
0.5 µg
0.1 µg
0 µg
LODb
0.1
0.1
a The reaction consisted of the amount of DHR indicated in the presence of 5 mg of DNA. b LOD ) limit of detection based upon the analysis of
100 µg of DNA by HPLC-ES-MS/MS.
adducts in liver samples from an NTP study in which female
F344 rats were treated with riddelliine for six months (41). In
the current study, we assessed the same samples by HPLC-ES-
MS/MS. Representative chromatograms are shown in Figure
12, and the data are summarized in Table 4. As with the DNA
samples from the rats treated for three days with either riddelliine
or monocrotaline, DHP-dG-3- and DHP-dG-4 were formed in
a dose-responsive manner and to a greater extent than DHP-
dA-3 and DHP-dA-4. Likewise, DHP-dG-1, DHP-dG-2, DHP-
dA-1, and DHP-dA-2 were not detected in any treatment groups.
To develop the method, the eight DHP-dG and DHP-dA adducts,
namely, DHP-dG-1, DHP-dG-2, DHP-dG-3, DHP-dG-4, DHP-
dA-1, DHP-dA-2, DHP-dA-3, and DHP-dA-4, as well as their
13
[15N5]dG and [15N5, C10]dA adducts, were prepared syntheti-
cally. Upon determination of the optimal HPLC and mass
spectrometric conditions, we are able to detect these eight DHP-
derived dG and dA adducts in Vitro and/or in ViVo and quantify
the formation of these adducts in the multiple reaction monitor-
ing (MRM) mode. The DNA adducts assessed included those
formed from the reaction of DHR with calf thymus DNA in
Vitro, in liver DNA from female rats treated with different doses
of riddelliine or monocrotaline for three consecutive days, and
in liver DNA from female rats treated with riddelliine for six
months.
Discussion
Pyrrolizidine alkaloids are a class of genotoxic and tumori-
genic naturally occurring phytochemicals (3-6, 8, 26). In this
article, we report the development of a sensitive HPLC-ES-
MS/MS method for the detection of DHP-derived DNA adducts
and the utilization of this method for the detection and
quantification of these adducts from in Vitro and in ViVo samples.
In this developed HPLC-ES-MS/MS method, the sensitivity
for different DHP-DNA adducts varies in the range of 0.1-1.5
adducts in 108 nucleotides for the analysis of 100 µg samples.
This level of sensitivity is potentially adequate for measuring

648 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
Figure 11. Representative MRM chromatograms (relative signal intensity vs time) for rat liver DNA samples. The right panels show a 100 µg
aliquot of DNA from a rat dosed daily with 5 mg of riddelline/kg body weight for three days, and the left panels show a 100 µg sample of DNA
from a rat dosed daily with 5 mg of monocrotaline/kg body weight for three days. Internal standards (IS) were added in amounts of 50, 55, 50, and
80 fmol for DHP-dG-1, DHP-dG-2, DHP-dA-1, and DHP-dA-2, respectively.
Table 3. Levels of DHP-dG and DHP-dA Adducts Formed in the Livers of Female Rats Dosed with Riddelliine or
Monocrotaline for Three Consecutive Days
levels of DHP-dG and DHP-dA/108 nucleotidesa
treatment
(mg/kg body weight/day)
DHP-dG-1
DHP-dG-2
DHP-dG-3
DHP-dG-4
DHP-dA-1
DHP-dA-2
DHP-dA-3
DHP-dA-4
riddelliine
5
2
1
0.1
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
7.5
5.0
2.0
<LOD
10.2
4.9
2.6
<LOD
<LOD
<LOD
<LOD
0.5
1.0
0.5
0.3
0.1
1.4
0.6
0.3
0.1
<LOD
<LOD
<LOD
<LOD
monocrotaline
5
1
<LOD
<LOD
<LOD
<LOD
5.2
<LOD
6.9
<LOD
<LOD
<LOD
<LOD
<LOD
0.6
0.1
0.9
0.1
control
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
a Aliquots of DNA (100 µg) were assayed by HPLC-ES-MS/MS. The LOD for each of the adducts is given in Table 2.
adduct levels in exposed humans, provided that sufficient DNA
samples can be collected (e.g., lymphocyte DNA is present in
whole blood at 6 µg/mL), that the ingested doses are high
enough (e.g., detectable in rat liver from 3 doses of 0.1 mg/kg
bw riddelliine), and that samples are collected soon enough after
ingestion so that adduct loss through DNA repair does not occur.
and DHP-dA-4 (Table 2). In contrast, rats dosed with either
riddelliine or monocrotaline formed DHP-dG-3 and DHP-dG-4
as the major adducts, accompanied by small amounts of DHP-
dA-3 and DHP-dA-4 (Tables 3 and 4).
On the basis of the results of our previous mechanistic studies
conducted by 32P-postlabeling/HPLC, we had proposed a general
mechanism for DHP-derived DNA adduct formation from the
metabolism of all three types of tumorigenic pyrrolizidine
alkaloids (3, 5, 22, 24-26). Since the HPLC-ES-MS/MS method
provides further insight regarding adduct formation, we have
revised the proposed general metabolic activation mechanism,
which is presented in Scheme 3. In this Scheme, the retronecine-,
heliotridine-, and otonecine-type pyrrolizidine alkaloids are
metabolized to reactive pyrrolic esters (3, 5, 22, 24-26).
Hydrolysis and the subsequent loss of a hydroxyl group of the
There were marked differences in regioselectivity depending
upon the system being investigated. Reactions of DHR with
dG produced a pair of epimeric DHP-dG-1 and DHP-dG-2 as
the predominant adducts, with DHP-dG-3 and DHP-dG-4 being
formed in a much lower yield (Figure 3A). Likewise, DHR
reacted with dA in a regioselective manner to produce pre-
dominantly DHP-dA-1 and DHP-dA-2 (Figure 3B). When DHR
was reacted with calf thymus DNA, all four DHP-dG adducts
were formed to a similar but to a lower extent than DHP-dA-3

DNA Adducts from Pyrrolizidine Alkaloids
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 649
Figure 12. Representative MRM chromatograms (relative signal intensity vs time) for liver DNA samples from an NTP study in which female
F344 rats were administered riddelliine. The left panel is a 100 µg aliquot of liver DNA from a control rat, and the right panel shows a 100 µg
sample of DNA from a rat dosed at 1 mg of riddelliine/kg body weight five days/week for six months. Internal standards (IS) were added in
amounts of 50, 55, 50, and 80 fmol for DHP-dG-1, DHP-dG-2, DHP-dA-1, and DHP-dA-2, respectively. The differences in retention times between
the left and right panels are due to the use of two different HPLC columns.
Table 4. Levels of DHP-dG and DHP-dA Adducts Formed in the Livers of NTP Female Rats Dosed with Riddelliine for Six
Months
levels of DHP-dG and DHP-dA/108 nucleotidesa
riddelliine (mg/kg b.w./day)
DHP-dG-1
DHP-dG-2
DHP-dG-3
DHP-dG-4
DHP-dA-1
DHP-dA-2
DHP-dA-3
DHP-dA-4
1
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
5.9 ( 2.4
2.3 ( 1.2
<LOD
8.0 ( 2.4
2.7 ( 1.2
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
0.5 ( 0.1
0.3 ( 0.1
0.1 ( 0.0
<LOD
0.8 ( 0.2
0.3 ( 0.1
0.1 ( 0.1
<LOD
0.33
0.1
0.1
0.033
0.01
0
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
<LOD
a Aliquots of DNA (100 µg) were assayed by HPLC-ES-MS/MS. The LOD for each of the adducts is given in Table 2. The data are presented as the
mean ( s.d., n ) 5.
pyrrolic ester lead to the formation of carbonium ions I and II.
Reaction of carbonium ion I with DNA produces C7-substituted
DHP-DNA adducts (i.e., DHP-dG-1, DHP-dG-2, DHP-dA-1,
and DHP-dA-2). Alternatively, carbonium ion I can undergo a
1,3-hydride shift to form carbonium ion III, which is stabilized
through resonance with iminium ion IV. The reaction of
carbonium ion III with DNA forms C5-substituted DHP-DNA
adducts (i.e., DHP-dG-4), whereas the reaction of carbonium
ion II with DNA leads to C9-substituted DHP-DNA adducts
(i.e., DHP-dG-3).
The pyrrolic ester (shown in Scheme 3) can also undergo
two consecutive hydrolysis steps to form DHP, which upon loss
of a hydroxyl group forms carbonium ions V and VI. Reaction
of these carbonium ions with DNA directly or after conversion
into carbonium ion VII through a 1,3-hydride shift generates
both DHP-C7 DNA adducts and DHP-C5 DNA adducts. The
DHP-C7 DNA adducts formed from carbonium ion I should
be kinetically controlled adducts, while the DHP-C5 DNA
adducts, mediated by the 1,3-hydride shift of carbonium ion I,
should be thermodynamically controlled adducts.
A comparison of the ratio of adducts obtained from in Vitro
reactions with DHR indicates that steric hindrance plays a
critical role in the regioselectivity of the reaction. When
reactions are conducted between DHR and the nucleosides dG
and dA, adducts arising from the reaction at C7 of DHP are
formed to a greater extent than adducts arising from substitution
at C5 or C9 of DHP. When DHR is reacted with DNA, the
ratio between C7-substituted adducts remains the same, as is
observed in the reactions with deoxynucleosides (DHP-dG-1/
2:DHP-dA-1/2; ∼10:1), while the proportion of adducts from
reaction at C5 and/or C9 (i.e., DHP-dG-3/4 and DHP-dA-3/4)
increases substantially (Table 2). This suggests that with the
deoxynucleosides, reactions occur primarily with carbonium ion
V, while with DNA, because of steric constraints, the equilib-
rium is shifted in favor of carbonium ions VI and VII (Scheme
3). The situation in ViVo appears to be more complex. The only

650 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
Scheme 3. Proposed General Mechanism Leading to DHP-Derived DNA Adduct Formation from the Metabolism of the Three
Types of Carcinogenic Pyrrolizidine Alkaloidsa
a Riddelliine, lasiocarpine, and clivorine are used as examples for the retronecine-type, heliotridine-type, and otonecine-type pyrrolizidine alkaloids,
respectively.
adducts detected in animals treated with riddelliine or mono-
crotaline are those obtained from the reaction at C5 and/or C9
of DHP (i.e., DHP-dG-3, DHP-dG-4, DHP-dA-3, and DHP-
dA-4; Tables 3 and 4). The lack of adducts arising from
substitution at C7 of DHP indicates that adduct formation in
ViVo does not occur with free DHR but rather must result from
pathways involving carbonium ions I (which rearranges to
carbonium ion III) and II, which leads to DHP-dG-3, DHP-
dG-4, DHP-dA-3, and DHP-dA-4. The other possibility to
explain the lack of DHP-dG-1, DHP-dG-2, DHP-dA-1, and
DHP-dA-2 adduct formation in ViVo is that these adducts are
indeed formed but are readily repaired.
It is worthy to note that pyrrolizidine alkaloids are capable
of binding at the C7 and the C9 positions of the necine base to
two sites in DNA or protein to form DNA or protein cross-
linking. The antimitotic, toxic, and carcinogenic activities of
pyrrolizidine alkaloids are thought to be caused, at least in part,
by these cross-links (45-50). At present, it is not known whether
or not DNA-interstrand cross-linking is involved, at least in part,
in the observed DNA adduct formation profile in ViVo and in
Vitro. It warrants further investigation.
On the basis of our comprehensive studies on the mech-
anisms of pyrrolizidine alkaloid carcinogenesis, we previously
proposed that DHP-derived DNA adducts are potential

DNA Adducts from Pyrrolizidine Alkaloids
Chem. Res. Toxicol., Vol. 23, No. 3, 2010 651
(16) Schoental, R., Head, M. A., and Peacock, P. R. (1954) Senecio
alkaloids; primary liver tumours in rats as a result of treatment with
(1) a mixture of alkaloids from S. jacobaea Lin.; (2) retrorsine; (3)
isatidine. Br. J. Cancer 8, 458–465.
biomarkers of pyrrolizidine alkaloid exposure and tumori-
genicity (3, 5, 8, 12, 20-22, 24-29, 51). With the new
information obtained in the present study (Table 3, Table 4,
and Scheme 3), the emphasis is focused on DHP-dG-3 and
DHP-dG-4, and possibly DHP-dA-3 and DHP-dA-4. Rid-
delliine and monocrotaline, the two pyrrolizidine alkaloids
studied by this HPLC-ES-MS/MS method, are retronecine-
type pyrrolizidine alkaloids. The DNA adduct formation
profiles from the metabolism of heliotridine-type and otone-
cine-type pyrrolizidine alkaloids have not yet been examined.
In order to validate the generality of the proposed metabolic
activation pathways shown in Scheme 3, mechanistic studies
with these two types of pyrrolizidine alkaloids are needed.
(17) National Toxicology Program. (1978) TR-39 Bioassay of Lasiocarpine
for Possible Carcinogenicity (CAS No. 303-34-4), NIH Publication No.
78-839, U.S. Department of Health and Human Services, Public Health
Service, National Institutes of Health, Research Triangle Park, NC.
(18) National Toxicology Program. (2003) TR-508 Toxicology and Carcino-
genesis Studies of Riddelliine (CAS No. 23246-96-0) in F344/N Rats and
B6C3F1 Mice (GaVage Studies), NIH Publication No. 03-4442, U.S.
Department of Health and Human Services, Public Health Service,
National Institutes of Health, Research Triangle Park, NC.
(19) Yang, Y. C., Yan, J., Doerge, D. R., Chan, P. C., Fu, P. P., and Chou,
M. W. (2001) Metabolic activation of the tumorigenic pyrrolizidine
alkaloid, riddelliine, leading to DNA adduct formation in vivo. Chem.
Res. Toxicol. 14, 101–109.
(20) Chou, M. W., Yan, J., Nichols, J., Xia, Q., Beland, F. A., Chan, P.-
C., and Fu, P. P. (2003) Correlation of DNA adduct formation and
riddelliine-induced liver tumorigenesis in F344 rats and B6C3F1 mice.
Cancer Lett. 193, 119–125.
(21) Chou, M. W., Jian, Y., Williams, L. D., Xia, Q., Churchwell, M.,
Doerge, D. R., and Fu, P. P. (2003) Identification of DNA adducts
derived from riddelliine, a carcinogenic pyrrolizidine alkaloid, in vitro
and in vivo. Chem. Res. Toxicol. 16, 1130–1137.
(22) Xia, Q., Chou, M. W., Kadlubar, F. F., Chan, P.-C., and Fu, P. P.
(2003) Human liver microsomal metabolism and DNA adduct forma-
tion of the tumorigenic pyrrolizidine alkaloid, riddelliine. Chem. Res.
Toxicol. 16, 66–73.
Acknowledgment. We thank Dr. Po-Chuen Chan of NTP
for supplying riddelliine and the NTP rat livers. This research
was supported in part by appointments (Y.W. and Y.Z.) to the
Postgraduate Research Program at the NCTR administered by
the Oak Ridge Institute for Science and Education through an
interagency agreement between the U.S. Department of Energy
and the FDA. This article is not an official U.S. Food and Drug
Administration (FDA) guidance or policy statement. No official
support or endorsement by the U.S. FDA is intended or should
be inferred.
(23) Federal Register. (2008) 73, 23463.
(24) Xia, Q., Chou, M. W., Lin, G., and Fu, P. P. (2004) Metabolic
formation of DHP-derived DNA adducts from a representative
otonecine type pyrrolizidine alkaloid clivorine and the extract of
Ligularia hodgsonnii hook. Chem. Res. Toxicol. 17, 702–708.
(25) Xia, Q., Chou, M. W., Edgar, J. A., Doerge, D. R., and Fu, P. P.
(2006) Formation of DHP-derived DNA adducts from metabolic
activation of the prototype heliotridine-type pyrrolizidine alkaloid,
lasiocarpine. Cancer Lett. 231, 138–145.
(26) Xia, Q., Yan, J., Chou, M. W., and Fu, P. P. (2008) Formation of
DHP-derived DNA adducts from metabolic activation of the prototype
heliotridine-type pyrrolizidine alkaloid, heliotrine. Toxicol. Lett. 178,
77–82.
(27) Wang, Y.-P., Fu, P., and Chou, M. (2005) Metabolic activation of the
tumorigenic pyrrolizidine alkaloid, retrorsine, leading to DNA adduct
formation in vivo. Int. J. EnViron. Res. Pub. Health 2, 74–79.
(28) Wang, Y.-P., Yan, J., Beger, R. D., Fu, P. P., and Chou, M. W. (2005)
Metabolic activation of the tumorigenic pyrrolizidine alkaloid, mono-
crotaline, leading to DNA adduct formation in vivo. Cancer Lett. 226,
27–35.
(29) Yan, J., Xia, Q., Chou, M. W., and Fu, P. P. (2008) Metabolic
activation of retronecine and retronecine N-oxide: formation of DHP-
derived DNA adducts. Toxicol. Ind. Health 24, 181–188.
(30) Yan, J., Nichols, J., Yang, Y.-C., Fu, P., and Chou, M. (2002) Detection
of riddelliine-derived DNA adducts in blood of rats fed riddelliine.
Int. J. Mol. Sci. 3, 1019–1026.
References
(1) Roeder, E. (1995) Medicinal plants in Europe containing pyrrolizidine
alkaloids. Pharmazie 50, 83–98.
(2) Roeder, E. (2000) Medicinal plants in China containing pyrrolizidine
alkaloids. Pharmazie 55, 711–726.
(3) Fu, P. P., Xia, Q., Chou, M. W., and Lin, G. (2007) Detection,
hepatotoxicity, and tumorigenicity of pyrrolizidine alkaloids in Chinese
herbal plants and herbal dietary supplements. J. Food Drug Anal. 15,
400–415.
(4) Mattocks, A. R. (1986) Chemistry and Toxicology of Pyrrolizidine
Alkaloids, Academic Press, London, UK.
(5) Fu, P. P., Xia, Q., Lin, G., and Chou, M. W. (2004) Pyrrolizidine
alkaloids: genotoxicity, metabolism enzymes, metabolic activation, and
mechanisms. Drug Metab. ReV. 36, 1–55.
(6) International Agency for Research in Cancer (IARC). (1976) Pyr-
rolizidine Alkaloids, in IARC Monographs on the EValuation of
Carcinogenic Risk of Chemicals to Man: Some Naturally Occurring
Substances, International Agency for Research in Cancer, Lyon,
France.
(7) Roeder, E., and Wiedenfeld, H. (2009) Pyrrolizidine alkaloids in
medicinal plants of Mongolia, Nepal and Tibet. Pharmazie 64, 699–
716.
(31) Mattocks, A. R. (1968) Toxicity of pyrrolizidine alkaloids. Nature
(8) Fu, P. P., Yang, Y.-C., Xia, Q., Chou, M. W., Cui, Y. Y., and Lin, G.
(2002) Pyrrolizidine alkaloids: tumorigenic components in Chinese
herbal medicines and dietary supplements. J. Food Drug Anal. 10,
198–211.
(9) Bull, L. B., Culvenor, C. C. J., and Dick, A. J. (1968) The
Pyrrolizidine Alkaloids: Their Chemistry, Pathogenicity and Other
Biological Properties, North-Holland Publishing, Amsterdam, The
Netherlands.
217, 723–728.
(32) Chauvin, P., Dillon, J. C., and Moren, A. (1994) An outbreak of
Heliotrope food poisoning, Tadjikistan, November 1992-March 1993,
Sante 4, 263-268.
(33) Prakash, A. S., Pereira, T. N., Reilly, P. E., and Seawright, A. A.
(1999) Pyrrolizidine alkaloids in human diet. Mutat. Res. 443, 53–67.
(34) Byron, J. (1998) Pyrrolizidine alkaloids in eggs; new alkaloid found
in potatoes. Food Chem. News 14, 6–7.
(10) International Programme on Chemical Safety (IPCS). (1989) Pyr-
rolizidine Alkaloids Health and Safety Guide, Health and Safety
Criteria Guide 26, WHO, Geneva, Switzerland.
(35) Culvenor, C. C. J., Edgar, J. A., Smith, J. A., and Hirono, I. (1976)
The occurrence of senecionine in Tussilago farfara. Aust. J. Chem.
29, 229–230.
(11) Smith, L. W., and Culvenor, C. C. (1981) Plant sources of hepatotoxic
(36) Winship, K. A. (1991) Toxicity of comfrey. AdVerse Drug React.
pyrrolizidine alkaloids. J. Nat. Prod. 44, 129–152.
Toxicol. ReV. 10, 47–59.
(12) Fu, P., Chou, M., Xia, Q., Yang, Y.-C., Yan, J., Doerge, D., and Chan,
P. (2001) Genotoxic pyrrolizidine alkaloids and pyrrolizidine alkaloid
N-oxides: mechanisms leading to DNA adduct formation and tum-
origenicity. J. EnViron. Sci. Health, Part C 19, 353–386.
(13) Stegelmeier, B. L., Edgar, J. A., Colegate, S. M., Gardner, D. R.,
Schoch, T. K., Coulombe, R. A., and Molyneux, R. J. (1999)
Pyrrolizidine alkaloid plants, metabolism and toxicity. J. Nat. Toxins
8, 95–116.
(14) Huxtable, R. J. (1980) Herbal teas and toxins: novel aspects of
pyrrolizidine poisoning in the United States. Perspect. Biol. Med. 24,
1–14.
(15) Steenkamp, V., Stewart, M. J., and Zuckerman, M. (2000) Clinical
and analytical aspects of pyrrolizidine poisoning caused by South
African traditional medicines. Ther. Drug Monit. 22, 302–306.
(37) Huxtable, R. J. (1989) Human Health Implications of Pyrrolizidine
Alkaloids and the Herbs Containing them, in Toxicants of Plant Origin,
Vol. 1 Alkaloids, pp 41-86 (Cheeke, P. R., Ed.) CRC Press, Boca
Raton, FL.
(38) Edgar, J. A., Roeder, E., and Molyneux, R. J. (2002) Honey from
plants containing pyrrolizidine alkaloids: a potential threat to health.
J. Agric. Food Chem. 50, 2719–2730.
(39) Betz, J. M., Eppley, R. M., Taylor, W. C., and Andrzejewski, D. (1994)
Determination of pyrrolizidine alkaloids in commercial comfrey
products (Symphytum sp.). J. Pharm. Sci. 83, 649–653.
(40) Chou, M. W., and Fu, P. P. (2006) Formation of DHP-derived DNA
adducts in vivo from dietary supplements and Chinese herbal plant
extracts containing carcinogenic pyrrolizidine alkaloids. Toxicol. Ind.
Health 22, 321–327.

652 Chem. Res. Toxicol., Vol. 23, No. 3, 2010
Fu et al.
(41) Yang, Y.-C., Yan, J., Churchwell, M., Beger, R., Chan, P.-C.,
Doerge, D. R., Fu, P. P., and Chou, M. W. (2001) Development of
(46) Kim, H. Y., Stermitz, F. R., Molyneux, R. J., Wilson, D. W., Taylor,
D., and Coulombe, R. A. (1993) Structural influences on pyrrolizidine
alkaloid-induced cytopathology. Toxicol. Appl. Pharmacol. 122, 61–
69.
(47) Kim, H. Y., Stermitz, F. R., and Coulombe, R. A. (1995) Pyrrolizidine
alkaloid-induced DNA-protein cross-links. Carcinogenesis 16, 2691–
2697.
(48) Coulombe, R. A., Jr., Drew, G. L., and Stermitz, F. R. (1999)
Pyrrolizidine alkaloids crosslink DNA with actin. Toxicol. Appl.
Pharmacol. 154, 198–202.
(49) Kim, H. Y., Stermitz, F. R., Li, J. K., and Coulombe, R. A., Jr. (1999)
Comparative DNA cross-linking by activated pyrrolizidine alkaloids.
Food Chem. Toxicol. 37, 619–625.
(50) Rieben, W. K., Jr., and Coulombe, R. A., Jr. (2004) DNA Cross-
linking by dehydromonocrotaline lacks apparent base sequence prefer-
ence. Toxicol. Sci. 82, 1–25.
(51) Fu, P., Xia, Q., Lin, G., and Chou, M. (2002) Genotoxic pyrrolizidine
alkaloids: mechanisms leading to DNA adduct formation and tum-
origenicity. Int. J. Mol. Sci. 3, 948–964.
a
32P-postlabeling/HPLC method for detection of dehydroretrone-
cine-derived DNA adducts in vivo and in vitro. Chem. Res. Toxicol.
14, 91–100.
(42) Chan, P. C., Haseman, J. K., Prejean, J. D., and Nyska, A. (2003)
Toxicity and carcinogenicity of riddelliine in rats and mice. Toxicol.
Lett. 144, 295–311.
(43) 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., and Doerge, D. R. (2005) High-
performance liquid chromatography electrospray ionization tandem
mass spectrometry for the detection and quantitation of benzo[a]py-
rene-DNA adducts. Chem. Res. Toxicol. 18, 1306–1315.
(44) Robertson, K. A. (1982) Alkylation of N2 in deoxyguanosine by
dehydroretronecine, a carcinogenic metabolite of the pyrrolizidine
alkaloid monocrotaline. Cancer Res. 42, 8–14.
(45) Hincks, J. R., Kim, H. Y., Segall, H. J., Molyneux, R. J., Stermitz,
F. R., and Coulombe, R. A. (1991) DNA cross-linking in mammalian
cells by pyrrolizidine alkaloids: structure-activity relationships. Toxicol.
Appl. Pharmacol. 111, 90–98.
TX900402X