Metabolism of N'-nitrosonornicotine enantiomers by cultured rat esophagus and in vivo in rats

Article abstract of DOI:10.1021/tx990171l

People who use tobacco products are exposed to considerable amounts of N'-nitrosonornicotine (NNN), a well-established esophageal carcinogen in rats. NNN is believed to play a significant role as a cause of esophageal and oral cavity cancer in smokers and

Full text of DOI:10.1021/tx990171l

192
Chem. Res. Toxicol. 2000, 13, 192-199
Meta bolism of N′-Nitr oson or n icotin e En a n tiom er s by
Cu ltu r ed Ra t Esop h a gu s a n d in Vivo in Ra ts
Edward J . McIntee and Stephen S. Hecht*
University of Minnesota Cancer Center, Minneapolis, Minnesota 55455
Received October 6, 1999
People who use tobacco products are exposed to considerable amounts of N′-nitrosonornicotine
(NNN), a well-established esophageal carcinogen in rats. NNN is believed to play a significant
role as a cause of esophageal and oral cavity cancer in smokers and snuff dippers. The
carcinogenicity of NNN is dependent on its metabolic activation. However, virtually all studies
carried out to date on NNN metabolism have used racemic material. In this study, we examined
the metabolism of [5-3H]-(S)-NNN and [5-3H]-(R)-NNN in cultured rat esophagus and in vivo
in rats. Cultured rat esophagus metabolized (S)-NNN (1 µM) predominantly to products of
2′-hydroxylation, 4-oxo-4-(3-pyridyl)butanoic acid (keto acid) and 4-hydroxy-1-(3-pyridyl)-1-
butanone (keto alcohol). In contrast, the major metabolite of (R)-NNN under these conditions
was 4-hydroxy-4-(3-pyridyl)butanoic acid (hydroxy acid), a product of NNN 5′-hydroxylation.
The 2′-hydroxylation:5′-hydroxylation metabolite ratio ranged from 6.22 to 8.06 at various time
intervals in the incubations with (S)-NNN, while the corresponding ratios were 1.12-1.33 in
the experiments with (R)-NNN. These differences were statistically significant (P < 0.001).
Since 2′-hydroxylation is believed to be the major metabolic activation pathway of NNN in the
rat esophagus, the results demonstrate that (S)-NNN is metabolically activated more
extensively than (R)-NNN in this tissue, and therefore may be more carcinogenic. Rats were
treated with 0.3 mg/kg of [5-3H]-(R)-NNN, [5-3H]-(S)-NNN, or racemic [5-3H]NNN by gavage,
and the urinary metabolites were analyzed. The major metabolites were hydroxy acid and
keto acid. As in the in vitro studies, products of 2′-hydroxylation predominated in the urine of
the rats treated with (S)-NNN while products of 5′-hydroxylation were more prevalent in the
rats treated with (R)-NNN. 2′-Hydroxylation:5′-hydroxylation metabolite ratios ranged from
1.66 to 2.04 in the urine at various times after treatment with (S)-NNN, while the ratios were
0.398-0.450 for the rats treated with (R)-NNN (P < 0.001). The results of this study provide
new insights into NNN metabolism in rats and suggest that the carcinogenicity of (S)-NNN,
the predominant enantiomer in tobacco products, may be greater than that of (R)-NNN or
racemic NNN.
an esophageal tumor incidence of 71% (5). When NNN
is swabbed in the oral cavity of rats together with the
related tobacco-specific nitrosamine 4-(methylnitrosamino)-
1-(3-pyridyl)-1-butanone (NNK), tumors of the oral cavity
develop (6). NNN causes respiratory tract tumors in mice
and hamsters and tumors of the nasal mucosa in mink
(4). These data, together with the exposure levels dis-
cussed above, strongly support the involvement of NNN
as a causative factor for esophageal cancer in smokers
and cancer of the oral cavity in people who use smokeless
tobacco products (3, 4).
In tr od u ction
The tobacco-specific nitrosamine N′-nitrosonornicotine
(NNN)1 is present in substantial quantities in unburned
tobacco as well as cigarette smoke. Levels of NNN in the
five leading brands of moist snuff sold in the United
States vary from 3.04 to 8.73 µg/g of dry weight of tobacco
(1). NNN levels in cigarette tobacco range from 45 to
121 454 ng/cigarette and in mainstream smoke from 5
to 1353 ng/cigarette based on studies of products from
12 different countries (2, 3). These extensive international
data clearly demonstrate that people who use tobacco
products are exposed to significant amounts of NNN.
Numerous studies show that NNN causes tumors of
the esophagus and nasal mucosa in rats (4). When
administered in drinking water, NNN produces mainly
esophageal tumors, while nasal tumors are more common
in rats treated with NNN by injection (4). Treatment of
rats with 5 ppm NNN in the drinking water results in
The metabolism of NNN is summarized in Scheme 1
(4). Pyridine N-oxidation produces N′-nitrosonornicotine
1-N-oxide (NNN-N-oxide), while denitrosation and oxida-
tion yield norcotinine; both pathways result in detoxifi-
cation. All positions of the pyrrolidine ring are hydrox-
ylated. The major hydroxylation reactions occur R to the
N-nitroso group, yielding 2′-hydroxy-NNN and 5′-hy-
droxy-NNN. Hydroxylation at the â-carbons occurs only
to a limited extent. 2′-Hydroxy-NNN undergoes sponta-
neous ring opening, producing a pyridyloxobutyl diazo-
hydroxide. This intermediate reacts with H2O producing
4-hydroxy-1-(3-pyridyl)-1-butanone (keto alcohol), which
in turn is oxidized to 4-oxo-4-(3-pyridyl)butanoic acid
1
Abbreviations: diol, 4-hydroxy-1-(3-pyridyl)-1-butanol; hydroxy
acid, 4-hydroxy-4-(3-pyridyl)butanoic acid; keto acid, 4-oxo-4-(3-py-
ridyl)butanoic acid; keto alcohol, 4-hydroxy-1-(3-pyridyl)-1-butanone;
lactol, 5-(3-pyridyl)-2-hydroxytetrahydrofuran; NNK, 4-(methylnitro-
samino)-1-(3-pyridyl)-1-butanone; NNN, N′-nitrosonornicotine; NNN-
N-oxide, N′-nitrosonornicotine 1-N-oxide.
10.1021/tx990171l CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/24/2000

Metabolism of N′-Nitrosonornicotine in Rats
Chem. Res. Toxicol., Vol. 13, No. 3, 2000 193
Sch em e 1. Meta bolism of NNN Ba sed on Stu d ies in La bor a tor y An im a lsa
a
See ref 4 for further details.
Ap p a r a tu s. HPLC analyses were carried out with a Waters
Associates (Milford, MA) system consisting of two model 510
pumps, a model 440 UV absorbance detector, an IN/US Systems
model 2 â-ram radioflow detector (IN/US Systems, Inc., Fair-
field, NJ ), and either a Rheodyne model 7725 manual injector
(Rheodyne Inc., Cotati, CA) or a WISP model 710 Waters
Intelligent Sample Processor. Mass spectra were obtained on a
Finnigan TSQ 7000 instrument (Finnigan MAT/Themoquest,
San J ose, CA). NMR spectra were obtained on a Varian Inova
300 MHz instrument (Varian, Inc., Palo Alto, CA). Optical
rotations were determined with a Perkin-Elmer 241 MC pola-
rimeter. Confirmation of stereochemistry was achieved via GC
and GC with nitrosamine selective detection (Thermedics Detec-
tion, Inc., Chelmsford, MA) using a 30 m × 0.25 mm i.d., 0.25
µm film thickness, Cyclosil-B chiral column (J & W Scientific,
Folsom, CA) as previously described (10).
(keto acid) or reduced to 4-hydroxy-1-(3-pyridyl)-1-bu-
tanol (diol). The pyridyloxobutyl diazohydroxide formed
by 2′-hydroxylation of NNN also alkylates DNA, yielding
adducts which release keto alcohol on hydrolysis. 5′-
Hydroxylation of NNN gives 5′-hydroxy-NNN which ring
opens to a diazohydroxide that ultimately produces
mainly 5-(3-pyridyl)-2-hydroxytetrahydrofuran (lactol)
and 4-hydroxy-4-(3-pyridyl)butanoic acid (hydroxy acid).
DNA adduct formation by this pathway has not been
demonstrated, but is likely based on studies of the related
nitrosamine N-nitrosopyrrolidine (7). Studies to date
favor the view that 2′-hydroxylation of NNN is the major
metabolic activation pathway (4). This pathway predomi-
nates in rat target tissues such as esophagus and nasal
mucosa, and adducts formed by 2′-hydroxylation have
been detected in these tissues (8, 9).
NNN has a chiral center at the 2′-position. Recently,
we demonstrated that the enantiomeric composition of
NNN in cigarette tobacco, moist snuff, and chewing
tobacco ranges from 63 to 95% (S)-NNN (10). Only one
previous study examined the metabolism of (S)-NNN in
rats (11). At a dose of 100 mg/kg, urinary metabolites of
(S)-NNN were similar to those of racemic NNN. In the
study presented here, we synthesized the NNN enanti-
omers labeled with tritium in the pyridine ring and
investigated their metabolism by cultured rat esophagus
and in vivo in rats.
The following HPLC systems were used: (A) a 4.6 mm × 250
mm Luna 5 µm C-18 column (Phenomenex, Torrance, CA), with
a guard column, eluted with a linear gradient from 0 to 30%
CH3OH in 20 mM NaH2PO4 (pH 5.0) for 60 min at a flow rate
of 1 mL/min; (B) the same column used for system A eluted with
a linear gradient from 0 to 8% CH3OH/H2O (95:5) in 20 mM
NaH2PO4 (pH 7.0) for 15 min, held for 15 min, then eluted with
a linear gradient from 8 to 23% CH3OH/H2O (95:5) in 20 mM
NaH2PO4 (pH 7.0) for 30 min, then held for 20 min, and then
eluted with a linear gradient from 23 to 35% CH3OH/H2O (95:
5) in 20 mM NaH2PO4 (pH 7.0) for 15 min; and (C) a 3.9 mm ×
300 mm Waters µBondapak C-18 column eluted with a linear
gradient from 0 to 25% CH3OH in 25 mM CH3CO2NH4 (pH 4.5)
for 60 min.
Ch em ica ls. NNN and its metabolites were synthesized by
procedures described in the literature: NNN (12), NNN-N-oxide
(13), keto alcohol (14), diol (15), lactol (16), and keto acid and
hydroxy acid (17). Norcotinine was purchased from Toronto
Research Chemicals Inc. (North York, ON). Enantiomers of
Exp er im en ta l P r oced u r es
Ca u tion : NNN is carcinogenic and mutagenic and therefore
should be handled with extreme care, using appropriate protec-
tive clothing and ventilation at all times.

194 Chem. Res. Toxicol., Vol. 13, No. 3, 2000
McIntee and Hecht
[5-3H]NNN were synthesized from the appropriate 5-bromonor-
nicotine precursor by Moravek Biochemicals Inc. (Brea, CA) as
described previously (18).
supplemented with glutamine (2 mM), gentamycin (100 µg/mL),
insulin (1 milliunit/mL), and dexamethasone (0.4 µg/mL). To
each dish was added 60 µL of 0.9% saline containing either 100
µM [5-3H]-(R)-NNN (4.45 Ci/mmol) or [5-3H]-(S)-NNN (4.59 Ci/
mmol) for a final concentration of 1 µM. The dishes were placed
in a 37 °C controlled atmosphere incubator atop a rocking
(2′R,2′S)-N′-[(1R,2S,5R)-Men th oxycar bon yl]-5-br om on or -
n icotin e (3 a n d 4). To a solution of racemic 5-bromonornicotine
(2 g, 8.81 mmol) dissolved in 75 mL of ether and triethylamine
(1.84 mL, 13.2 mmol) at 0 °C under an N2 atmosphere was added
(1R,2S,5R)-menthyl chloroformate (2, Aldrich Chemical Co.,
Milwaukee, WI; 2.46 mL, 11.3 mmol) via neat syringe over the
course of 8 min with vigorous stirring. Precipitate formed as
the chloroformate was added. The mixture was stirred at 0 °C
for 10 min and at room temperature for 1.5 h. Following suction
filtration, the ethereal solution was washed with 10% NaOH (2
× 35 mL), dried with anhydrous Na2SO4, filtered, and concen-
trated to give a dark orange-yellow oil. Purification by silica
gel (70-230 mesh) flash column chromatography (1.25 in. × 16
in.) with elution by CHCl3 (500 mL) and then with CHCl3/CH3-
OH (98:2) afforded 2.2 g (61%) of the title compound as a viscous
oil: 1H NMR (CDCl3) δ 8.53 (1H, s, pyr-H2), 8.38 (1H, s, pyr-
H6), 7.64 (1H, s, pyr-H4), 4.9-4.8 (1H, m, CHO), 4.4-4.1 (1H,
m, H2′), 3.68 (2H, m, H5′), 2.37 (1H, m, OCHCH), 2.1-1.8 (4H,
m, H3′, H4′), 1.62-1.39 (2H, m, OCHCH2), 1.36-0.75 [15H, m
containing singlets of varying intensity, (CH3)2CH, CH3CHCH2-
CH2]; ESI MS m/z [M + H]+ 408.8 and 410.8.
Sem ip r ep a r a tive HP LC Resolu tion of N′-(Men th oxy-
ca r bon yl)-5-br om on or n icotin e. A semipreparative Whatman
Partisil 10 µm column was employed with a hexane/ethyl
acetate/CHCl3 (73:23:4) mixture as the eluent. All separations
were carried out at a flow rate of 5 mL/min and monitored at
280 nm. A typical sample size consisted of 12.5 mg of the
diastereomeric carbamate mixture 3 and 4 dissolved in 50 µL
of the eluent. Diastereomer 3 eluted at approximately 17 min
and was nearly baseline separated from 4. The collected
fractions were combined on the basis of purities determined by
reinjection on the HPLC system. The combined solutions of each
pure diastereomer obtained from 40 separations were concen-
trated to dryness in vacuo, giving 320 mg of the first diastere-
omer, (2′R)-N′-[(1R,2S,5R)-menthoxycarbonyl]-5-bromonornic-
otine (3) [[R]2D0 48.47° (c 1.77, CH2Cl2)], and 250 mg of (2′S)-N′-
[(1R,2S,5R)-menthoxycarbonyl]-5-bromonornicotine (4) [[R]D20
-117.56° (c 1.885, CH2Cl2)].
apparatus. The dishes were rocked at
a rate of 20 cpm,
alternately exposing the sections to the atmosphere (95% O2/
5% CO2) and to the culture medium. Aliquots of 450 µL of
medium were removed at 1, 2, 4, 6, 8, 10, and 23 h and placed
in 1.5 mL Eppendorf tubes. Tubes were then placed on dry ice
and stored at -20 °C until the contents were assayed.
An a lysis of Meta bolites. Medium samples were allowed to
thaw at room temperature for 40 min. The samples were then
vortexed, and a 0.2 mL aliquot was removed and filtered
through a 0.45 µm nylon Acrodisc filter. One hundred microliters
of each sample was then subjected to HPLC analysis using
system A. Metabolites were detected and quantified using
radioflow detection. Metabolites were identified by co-injection
with reference metabolites using HPLC system A. Selected
samples were reinjected on HPLC systems B and C to confirm
the identity of the metabolites. Metabolite concentrations were
expressed as picomoles per milliliter of medium.
Meta bolism in Vivo. Male F-344 rats (Charles River,
Kingston, NY) were maintained on NIH-07 pellet diet (Dyets,
Inc., Bethlehem, PA) and were housed individually in glass
metabolism cages. The rats (240-260 g) were divided into three
groups of three rats each. Each rat received a single intragastric
gavage dose (0.3 mg/kg of body weight) of either racemic [5-3H]-
NNN, [5-3H]-(R)-NNN, or [5-3H]-(S)-NNN in 0.5 mL of saline.
The amount of radioactivity and the specific activity of each
compound were as follows: racemic [5-3H]NNN, 55.2 µCi and
108.9 µCi/µmol; [5-3H]-(R)-NNN, 49.6 µCi and 97.9 µCi/µmol;
and [5-3H]-(S)-NNN, 52.2 µCi and 103 µCi/µmol. Urine was
collected after 6, 12, 18, 24, 36, and 48 h in vessels cooled to
0-4 °C. Urine samples were frozen on dry ice and stored at
-20 °C until they were assayed.
An a lysis of Ur in e. Urine samples were allowed to thaw at
room temperature for 40 min. The samples were vortexed, and
a 0.3 mL aliquot was removed and filtered through a 0.45 µm
nylon Acrodisc filter. One hundred microliters of each sample
was then subjected to HPLC analysis using system A. Metabo-
lites were detected and quantified using radioflow detection.
Metabolite concentrations were expressed as picomolar and were
adjusted for urine volume. Identification of metabolites was
initially determined by comparison of retention times of injected
reference metabolites using HPLC system A. Selected samples
were reinjected on HPLC systems B and C to confirm the
identity of the metabolites.
Sta tistica l An a lysis. Statistical comparison of 2′-hydroxy-
lated and 5′-hydroxylated metabolites was carried out using
SigmaStat (version 2.0). The ratio of 2′-hydroxylated to 5′-
hydroxylated metabolites was determined for each sample that
was analyzed. The groups were then compared using an
unpaired Student’s t test.
(2′R)-5-Br om on or n icotin e [(R)-1]. A solution of 300 mg of
3 in 3 mL of 10% HCl was heated in a 102 mm Teflon-lined
pressure tube at 110 °C for 60 h. The reaction vessel was cooled
in dry ice and opened without evidence of excess pressure having
developed. The reaction mixture was then washed with ether
(3 × 5 mL). The aqueous layer was basified with 10 N NaOH
and extracted with CHCl3 (4 × 5 mL). The CHCl3 extracts were
combined, dried (Na2SO4), filtered, and condensed. This gave
100 mg (60%) of (R)-1 as a golden oil: [R]2D0 52.85° (c 0.685,
1
CH2Cl2); H NMR (CDCl3) δ 8.49 (1H, s, pyr-H2), 8.44 (1H, s,
pyr-H6), 7.86 (1H, s, pyr-H4), 4.14 (1H, m, H2′), 3.17-3.11 (1H,
m, H5′), 3.09-2.98 (1H, m, H5′), 2.24-2.14 (2H, m, H3′ with
broad s for NH), 1.96-1.80 (2H, m, H4′), 1.77-1.57 (1H, m, H3′);
GC tR ) 51.57 min; ESI MS m/z [M + H]+ 226.6 and 228.6.
(2′S)-5-Br om on or n icotin e [(S)-1]. A solution of 230 mg of
4 in 3 mL of 10% HCl was heated in a 102 mm Teflon-lined
pressure tube at 110 °C for 60 h. The workup described above
produced 100 mg (79%) of (S)-1 as a golden oil: [R]2D0 -46.42°
Resu lts
The synthesis of [5-3H]-(S)- and -(R)-NNN is sum-
marized in Scheme 2. The method is based on that
described by Seeman et al. for resolution of nornicotine
(19). Derivatization of racemic 5-bromonornicotine (1)
with (1R,2S,5R)-menthyl chloroformate (2) gave the
diastereomeric carbamates 3 and 4, which were sepa-
rated by normal phase HPLC and hydrolyzed individu-
ally to (S)- and (R)-5-bromonornicotine. These were
converted to [5-3H]-(S)- and [5-3H]-(R)-NNN by catalytic
tritium exchange and nitrosation as previously described
(18). The products were analyzed by chiral column GC
with nitrosamine selective detection (10). The tritium-
labeled NNN enantiomers each coeluted with the corre-
1
(c 0.53, CH2Cl2); H NMR (CDCl3) δ 8.50 (1H, s, pyr-H2), 8.46
(1H, pyr-H6), 7.87 (1H, s, pyr-H4), 4.14 (1H, m, H2′), 3.18-3.13
(1H, m, H5′), 3.10-2.99 (1H, m, H5′), 2.24-2.14 (2H, m, H3′
with broad s for NH), 1.95-1.80 (2H, m, H4′), 1.75-1.55 (1H,
m, H3′); GC tR ) 52.28 min; ESI MS m/z [M + H]+ 226.6 and
228.6.
Meta bolism by Cu ltu r ed Ra t Esop h a gu s. Five male F-344
rats (240-260 g) were sacrificed by CO2 asphyxiation and the
esophagi removed. Each esophagus was split longitudinally into
two equal sections, each measuring ∼40 mm × 6 mm. The
fragments were placed with the adventitia side down in separate
60 mm × 15 mm scored culture dishes (Fisher Scientific, Itasca,
IL). Tissues were covered with 6 mL of Williams medium

Metabolism of N′-Nitrosonornicotine in Rats
Chem. Res. Toxicol., Vol. 13, No. 3, 2000 195
Sch em e 2. Syn th esis of [5-3H]-(R)-NNN a n d [5-3H]-(S)-NNN
Ta ble 1. 2′-Hyd r oxyla tion :5′-Hyd r oxyla tion Ra tios in th e Meta bolism of (S)-NNN a n d (R)-NNN by Cu ltu r ed Ra t
Esop h a gu sa
2′-hydroxylation:5′-hydroxylation ratio
1 h
2 h
4 h
6 h
8 h
10 h
23 h
(S)-NNN
(R)-NNN
Pb
8.06 ( 1.17
1.33 ( 0.134
<0.001
7.11 ( 0.764
1.15 ( 0.123
<0.001
6.89 ( 0.987
7.00 ( 1.03
6.79 ( 0.984
1.15 ( 0.0409
<0.001
6.56 ( 0.926
6.22 ( 0.614
1.16 ( 0.0644 1.18 ( 0.0935
1.13 ( 0.0704 1.12 ( 0.0783
<0.001
2.67 ( 1.15
0.008
2.63 ( 1.08
<0.001
2.62 ( 1.14
0.008
2.22 ( 0.794
(S)- + (R)-NNNc
3.43 ( 1.79
2.74 ( 1.38
2.58 ( 1.02
a
[5-3H]-(S)-NNN or [5-3H]-(R)-NNN was added to cultures of rat esophagus, and metabolites were analyzed at various times. 2′-
Hydroxylation is the sum of keto acid, keto alcohol, and diol, while 5′-hydroxylation is the sum of hydroxy acid and lactol. Keto acid and
b
hydroxy acid are not interconverted in cultured rat esophagus (20). Values are means ( SD. Comparisons of ratios from (S)-NNN to
(R)-NNN. c Calculated values.
sponding unlabeled standards, which have been synthe-
sized previously (10). The GC retention times of (S)- and
(R)-NNN were 41.7 and 44.9 min, respectively.
in rat esophagus cultured with (S)-NNN. In contrast,
hydroxy acid, a product of 5′-hydroxylation, was the
major metabolite of (R)-NNN. Figure 1C shows the
average of the metabolites of (S)- and (R)-NNN; these
are remarkably similar to that seen previously in the
metabolism of racemic NNN under the same conditions
(8). Using the data in panels A and B of Figure 1, we
calculate that, in a racemic mixture, 71% of the 2′-
hydroxylation products would be formed from (S)-NNN
and 69% of the 5′-hydroxylation products from (R)-NNN.
We added [5-3H]-(S)- or -(R)-NNN to cultures of rat
esophagi. HPLC analysis of metabolites showed peaks
corresponding in retention time to hydroxy acid, keto
acid, keto alcohol, diol, lactol, and norcotinine (Scheme
1). With the exception of lactol and norcotinine, the
identities of these metabolites in rat esophagus cultures
have been previously confirmed (8, 20). Lactol has been
established as a metabolite of NNN in rat esophageal
microsomes (21). In the study presented here, we con-
firmed the identities of lactol and norcotinine by HPLC
analysis in systems B and C. We also detected a small
peak corresponding in retention time to NNN-N-oxide,
but did not investigate it further.
Keto acid, keto alcohol, and diol are all products of 2′-
hydroxylation, while hydroxy acid and lactol are products
of 5′-hydroxylation. We calculated the 2′-hydroxylation:
5′-hydroxylation ratio for each enantiomer. These data,
which are summarized in Table 1, demonstrate that the
2′-hydroxylation:5′-hydroxylation ratio was significantly
greater in the metabolism of (S)-NNN than in that of (R)-
NNN at each time point. The 2′-hydroxylation:5′-hy-
droxylation ratios were also calculated at each time point
for the average of the metabolites produced from (S)-
We observed substantial differences between the me-
tabolism of (S)-NNN (Figure 1A) and (R)-NNN (Figure
1B). Keto acid and keto alcohol, products of 2′-hydroxy-
lation, were the major metabolites at every time point

196 Chem. Res. Toxicol., Vol. 13, No. 3, 2000
McIntee and Hecht
F igu r e 1. Formation of metabolites in rat esophagus cultured
with (A) [5-3H]-(S)-NNN and (B) [5-3H]-(R)-NNN. Panel C is the
calculated average of the data depicted in panels A and B: (0)
keto acid, (]) keto alcohol, ([) hydroxy acid, (O) norcotinine,
(4) diol, and (b) lactol.
F igu r e 3. Excretion of NNN metabolites in the urine of rats
treated with (A) [5-3H]-(S)-NNN, (B) [5-3H]-(R)-NNN, and (C)
racemic [5-3H]NNN: (0) keto acid, ([) hydroxy acid, (9) NNN-
N-oxide, (]) NNN, (4) diol, (O) norcotinine, and (b) lactol.
The major metabolites of NNN at this dose were
hydroxy acid and keto acid. Smaller amounts of NNN-
N-oxide, norcotinine, diol, lactol, and unchanged NNN
were also observed. Diol and lactol had not been reported
previously as urinary metabolites of NNN. Therefore, we
confirmed their identities by analysis in HPLC systems
B and C, in which they coeluted with standards. Time-
dependent excretion of urinary metabolites of (S)-NNN,
(R)-NNN, or racemic NNN is summarized in Figure 3A-
C. In the rats treated with (S)-NNN, keto acid was the
most prevalent urinary metabolite, followed by hydroxy
acid. However, in the rats treated with (R)-NNN, hydroxy
acid was the most prevalent metabolite, followed by keto
acid. The amounts of hydroxy acid and keto acid present
in the urine of the rats treated with racemic NNN were
consistent with these results. Total urinary excretion of
each metabolite is summarized in Figure 4.
2′-Hydroxylation:5′-hydroxylation ratios at each time
point are presented in Table 2. These results demonstrate
that (S)-NNN undergoes significantly more 2′-hydroxy-
lation than (R)-NNN. In the rats treated with racemic
NNN, approximately 66% of the 2′-hydroxylation prod-
ucts are formed from (S)-NNN while 74% of the 5′-
hydroxylation metabolites are produced from (R)-NNN.
F igu r e 2. Excretion of radioactivity in the urine of rats treated
with ([) [5-3H]-(R)-NNN, (0) [5-3H]-(S)-NNN, and (2) racemic
[5-3H]NNN.
NNN and (R)-NNN (Table 1). The ratios ranged from 2.22
to 3.43 which is consistent with previous results (8, 20).
For the in vivo studies, we chose a dose of 0.3 mg/kg
of body weight, administered by gavage. Rats treated
chronically with 5 ppm NNN in their drinking water, or
approximately 0.3 mg/kg of body weight per day, develop
tumors of the esophagus (5). Therefore, this dose is
similar to the daily dose in the carcinogenicity study. As
shown in Figure 2, more than 90% of the radioactivity
was excreted in the urine, consistent with previous
studies (22, 23). There was no difference in excretion
among the individual enantiomers and racemic mixture.
Discu ssion
The results of this study demonstrate clear stereose-
lectivity in rat esophageal metabolism of NNN enanti-

Metabolism of N′-Nitrosonornicotine in Rats
Chem. Res. Toxicol., Vol. 13, No. 3, 2000 197
Ta ble 2. 2′-Hyd r oxyla tion :5′-Hyd r oxyla tion Ra tios in th e Meta bolism of (S)-NNN, (R)-NNN, a n d Ra cem ic NNN in Vivo in
Ra tsa
2′-hydroxylation:5′-hydroxylation ratiob
6 h
12 h
18 h
24 h
36 h
48 h
(S)-NNN
(R)-NNN
racemic NNN
1.88 ( 0.043
0.398 ( 0.032
1.01 ( 0.057
1.66 ( 0.108
0.431 ( 0.047
0.880 ( 0.047
1.74 ( 0.237
0.450 ( 0.046
0.884 ( 0.020
1.81 ( 0.025
0.449 ( 0.019
0.750 ( 0.087
1.78 ( 0.081
0.438 ( 0.072
0.766 ( 0.040
2.04 ( 0.175
0.405 ( 0.042
0.786 ( 0.038
a
[5-3H]-(S)-NNN, [5-3H]-(R)-NNN, or racemic [5-3H]NNN (0.3 mg/kg) was administered to male F-344 rats by gavage, and urine was
b
collected at various times. Values are means ( SD. All differences among the three groups were significant at each time interval (P <
0.001) except (S)-NNN vs racemic NNN, 18 h (P ) 0.003); (R)-NNN vs racemic NNN, 24 h (P ) 0.004); and (R)-NNN vs racemic NNN,
36 h (P ) 0.002).
Sch em e 3. Som e P ossible P a th w a ys of NNN Meta bolism to Nor cotin in e
F igu r e 5. Structures of (S)-NNN, (R)-NNN, and NAB. Arrows
indicate the predominant positions of R-hydroxylation in cul-
tured rat esophagus.
esophagus and other target tissues of NNN compared to
liver, a nontarget site (8, 20, 21, 24, 25). Moreover,
comparative studies of NNN and the weak esophageal
carcinogen N′-nitrosoanabasine (NAB, Figure 5) demon-
strate that 2′-hydroxylation is strongly favored in the
metabolism of NNN by rat esophagus compared to that
F igu r e 4. Urinary metabolites in rats treated with [5-3H]-(R)-
of NAB (11). Collectively, these results suggest that (S)-
NNN (stippled bars), [5-3H]-(S)-NNN (white bars), and racemic
[5-3H]NNN (crosshatched bars).
NNN will be more carcinogenic to the rat esophagus than
(R)-NNN and that the carcinogenicity of NNN, based on
studies carried out to date exclusively with racemic
material, may have been underestimated. Considering
that (S)-NNN is the predominant enantiomer present in
tobacco, the results of a carcinogenicity study of the NNN
enantiomers could have considerable implications in risk
assessment. We are planning a comparative bioassay of
the NNN enantiomers and racemic NNN in rats.
Our results and those previously published indicate
that there is considerable stereoselectivity in rat esoph-
ageal metabolism of NNN and NAB (Figure 5). 5′-
Hydroxylation and 6′-hydroxylation are the predominant
pathways of metabolism of (R)-NNN and racemic NAB,
respectively, in cultured rat esophagus, while 2′-hydroxy-
lation is the predominant pathway of (S)-NNN metabo-
lism (11). Several studies provide strong evidence for the
presence of a high-affinity cytochrome P450 or P450s in
the rat esophagus which metabolize NNN, N-nitrosoben-
omers. (S)-NNN preferentially undergoes 2′-hydroxyla-
tion in this tissue, while (R)-NNN is preferentially
hydroxylated at the 5′-position. Previous studies carried
out with 1 µM racemic NNN have shown that 2′-
hydroxylation is favored compared to 5′-hydroxylation in
cultured rat esophagus, as well as in rat esophageal
microsomes. 2′-Hydroxylation:5′-hydroxylation ratios of
3 were reported in these studies (8, 20). The results
presented here indicate that 71% of the 2′-hydroxylation
products were formed from (S)-NNN while 69% of the
5′-hydroxylation products were formed from (R)-NNN. 2′-
Hydroxylation is believed to be the major pathway of
metabolic activation of NNN leading to esophageal tu-
morigenesis. Evidence in support of this hypothesis
includes the detection of DNA adducts resulting from 2′-
hydroxylation in rat esophagus cultured with NNN (8)
and higher 2′-hydroxylation:5′-hydroxylation ratios in

198 Chem. Res. Toxicol., Vol. 13, No. 3, 2000
McIntee and Hecht
zylmethylamine, and N-nitrosomethylamylamine (21,
26-28). These P450s have not been fully characterized,
but it is known that P450 2A3 is involved to a limited
extent (26, 27). The differences in metabolism shown here
between the NNN enantiomers, and among the NNN
enantiomers and NAB, should prove to be useful in
characterizing the enzymes involved in rat esophageal
metabolic activation of nitrosamines.
of metabolism of (S)-NNN by cultured rat esophagus and
in rats treated by gavage, while 5′-hydroxylation is the
favored pathway of (R)-NNN metabolism under these
conditions. These results provide new insights into mech-
anisms of metabolic activation of NNN and suggest that
the risk associated with exposure to NNN via tobacco
products may have been underestimated.
The results of the in vivo studies with racemic NNN
are consistent with those of a previous investigation (23).
As the dose of NNN was decreased from 300 mg/kg to 3
mg/kg in our earlier study, there was a decrease in 5′-
hydroxylation and a corresponding increase in 2′-hy-
droxylation such that the 2′-hydroxylation:5′-hydroxyla-
tion ratios were 0.24 at 300 mg/kg and 0.84 at 3 mg/kg
(23). In the study presented here, with an NNN dose of
0.3 mg/kg, the 2′-hydroxylation:5′-hydroxylation ratio was
0.94. These results are also consistent with those ob-
tained in a study of NNN metabolism in rat liver
microsomes (21). A 2′-hydroxylation:5′-hydroxylation ra-
tio of 0.23 was observed at an NNN concentration of 200
µM, compared to 0.71 at 1 µM, demonstrating higher 2′-
hydroxylation:5′-hydroxylation ratios at lower substrate
concentrations (21). The liver contains a mixture of NNN
R-hydroxylating enzymes, probably P450s (21). Among
these, there is apparently a high-affinity enzyme which
favors 2′-hydroxylation over 5′-hydroxylation at low NNN
concentrations (21). This may be the same enzyme that
is present in the rat esophagus (21). In support of this
concept, the in vivo studies of the NNN enantiomers
demonstrated preferential 2′-hydroxylation of (S)-NNN
and 5′-hydroxylation of (R)-NNN, as seen in the esopha-
gus, although the differences were not as great. This is
probably due to the presence of multiple NNN R-hy-
droxylating enzymes in the liver, some of which have a
stronger proclivity for 5′-hydroxylation of (R)-NNN than
for 2′-hydroxylation of (S)-NNN.
Ack n ow led gm en t. This study was supported by
Grant CA-81301 from the National Cancer Institute. We
thank Nadia Felicia and Sharon Murphy for their help
with the esophageal organ culture experiments and
Patrick Kenney for assistance with the in vivo studies.
Refer en ces
(1) Hoffmann, D., Djordjevic, M. V., Fan, J ., Zang, E., Glynn, T., and
Connolly, G. N. (1995) Five leading U.S. commercial brands of
moist snuff in 1994: Assessment of carcinogenic N-nitrosamines.
J . Natl. Cancer Inst. 87, 1862-1869.
(2) Spiegelhalder, B., and Bartsch, H. (1996) Tobacco-specific nitro-
samines. Eur. J . Cancer Prev. 5, 33-38.
(3) Hoffmann, D., Brunnemann, K. D., Prokopczyk, B., and Djord-
jevic, M. V. (1994) Tobacco-specific N-nitrosamines and areca-
derived N-nitrosamines: chemistry, biochemistry, carcinogenicity,
and relevance to humans. J . Toxicol. Environ. Health 41, 1-52.
(4) Hecht, S. S. (1998) Biochemistry, biology, and carcinogenicity of
tobacco-specific N-nitrosamines. Chem. Res. Toxicol. 11, 559-603.
(5) Stoner, G. D., Adams, C., Kresty, L. A., Hecht, S. S., Murphy, S.
E., and Morse, M. A. (1998) Inhibition of N′-nitrosonornicotine-
induced esophageal tumorigenesis by 3-phenylpropyl isothiocy-
anate. Carcinogenesis 19, 2139-2143.
(6) Hecht, S. S., Rivenson, A., Braley, J ., DiBello, J ., Adams, J . D.,
and Hoffmann, D. (1986) Induction of oral cavity tumors in F344
rats by tobacco-specific nitrosamines and snuff. Cancer Res. 46,
4162-4166.
(7) Wang, M., Upadhyaya, P., Dinh, T. T., Bonilla, L. E., and Hecht,
S. S. (1998) Lactols in hydrolysates of DNA reacted with R-ac-
etoxy-N-nitrosopyrrolidine and crotonaldehyde. Chem. Res. Toxi-
col. 11, 1567-1573.
(8) Murphy, S. E., Heiblum, R., and Trushin, N. (1990) Comparative
metabolism of N′-nitrosonornicotine and 4-(methylnitrosamino)-
1-(3-pyridyl)-1-butanone by cultured rat oral tissue and esopha-
gus. Cancer Res. 50, 4685-4691.
Our results may provide some insights concerning the
origins of norcotinine in NNN metabolism. Possible
mechanisms of norcotinine formation are summarized in
Scheme 3. In mechanism a, NNN would undergo direct
denitrosation to nornicotine (6), followed by oxidation to
norcotinine. In mechanism b, NNN would lose HNO
yielding isomyosmine (7), which would be oxidized to
norcotinine. In mechanism c, NNN would be oxidized to
5′-radical 8. Radical 8 would give both isomyosmine (7)
and 5′-hydroxy-NNN, both of which could ultimately yield
norcotinine. A related radical has been proposed as an
intermediate in the R-hydroxylation and denitrosation of
N-nitrosodimethylamine (29). The results do not appear
to support mechanism c since radical 8 would be the
likely precursor to both norcotinine and hydroxy acid.
Therefore, the ratio of these two metabolites would be
expected to be constant in the metabolism of (S)-NNN
and (R)-NNN, but this was not observed, either in the
esophagus or in vivo. These results suggest that mech-
anism c is not important in norcotinine formation from
NNN in the rat, unless there are different mechanisms
for the two NNN enantiomers. The results would favor
the view that norcotinine is formed by route a or b,
although it should be noted that administration of
nornicotine (6) to rats did not result in detectable
amounts of norcotinine in urine (23).
(9) Trushin, N., Rivenson, A., and Hecht, S. S. (1994) Evidence
supporting the role of DNA pyridyloxobutylation in rat nasal
carcinogenesis by tobacco specific nitrosamines. Cancer Res. 54,
1205-1211.
(10) Carmella, S. G., McIntee, E. J ., Chen, M., and Hecht, S. S. (2000)
Enantiomeric composition of N′-nitrosonornicotine and N′-ni-
trosoanatabine in tobacco. Carcinogenesis (in press).
(11) Hecht, S. S., and Young, R. (1982) Regiospecificity in the
metabolism of the homologous cyclic nitrosamines, N′-nitroso-
nornicotine and N′-nitrosoanabasine. Carcinogenesis 3, 1195-
1199.
(12) Hu, M. W., Bondinell, W. E., and Hoffmann, D. (1974) Synthesis
of carbon 14-labelled myosmine, nornicotine, and N′-nitrosonor-
nicotine. J . Labelled Compd. 10, 79-88.
(13) Hecht, S. S., Chen, C. B., and Hoffmann, D. (1980) Metabolic
R-hydroxylation and N-oxidation of N′-nitrosonornicotine. J . Med.
Chem. 23, 1175-1178.
(14) Hecht, S. S., Chen, C. B., Dong, M., Ornaf, R. M., Hoffmann, D.,
and Tso, T. C. (1977) Studies on non-volatile nitrosamines in
tobacco. Beitr. Tabakforsch. 9, 1-6.
(15) Hecht, S. S., Young, R., and Chen, C. B. (1980) Metabolism in
the F344 rat of 4-(N-methyl-N-nitrosamino)-1-(3-pyridyl)-1-bu-
tanone, a tobacco specific carcinogen. Cancer Res. 40, 4144-4150.
(16) Loozen, H. J . J ., Godefroi, E. F., and Besters, J . S. (1975) A novel
and efficient route to 5-arylated γ-lactones. J . Org. Chem. 40,
892-894.
(17) McKennis, H., Schwartz, S. L., Turnbull, L. B., Tamaki, E., and
Bowman, E. R. (1964) The metabolic formation of γ-(3-pyridyl)-
γ-hydroxybutyric acid and its possible intermediary role in the
mammalian metabolism of nicotine. J . Biol. Chem. 239, 3981-
3989.
(18) Wiley, J . C., Chien, D. H. T., Nungesser, N. A., Lin, D., and Hecht,
S. S. (1988) Synthesis of 4-(methylnitrosamino)-1-(3-pyridyl)-1-
butanone, 4-(carbethoxynitrosamino)-1-(3-pyridyl)-1-butanone, and
In summary, the results of this study clearly demon-
strate distinct differences in the metabolism of (S)- and
(R)-NNN. 2′-Hydroxylation is the predominant pathway

Metabolism of N′-Nitrosonornicotine in Rats
Chem. Res. Toxicol., Vol. 13, No. 3, 2000 199
N′-nitrosonornicotine labelled with tritium in the pyridine ring.
J . Labelled Compd. Radiopharm. 25, 707-716.
(19) Seeman, J . I., Chavdarian, C. G., and Secor, H. V. (1985) Synthesis
of the enantiomers of nornicotine. J . Org. Chem. 50, 5419-5421.
(20) Hecht, S. S., Reiss, B., Lin, D., and Williams, G. M. (1982)
Metabolism of N′-nitrosonornicotine by cultured rat esophagus.
Carcinogenesis 3, 453-456.
(21) Murphy, S. E., and Spina, D. A. (1994) Evidence for a high affinity
enzyme in rat esophageal microsomes which R-hydroxylates N′-
nitrosonornicotine. Carcinogenesis 15, 2709-2713.
(22) Chen, C. B., Hecht, S. S., and Hoffmann, D. (1978) Metabolic
R-hydroxylation of the tobacco specific carcinogen, N′-nitrosonor-
nicotine. Cancer Res. 38, 3639-3645.
(23) Hecht, S. S., Lin, D., and Chen, C. B. (1981) Comprehensive
analysis of urinary metabolites of N′-nitrosonornicotine. Carcino-
genesis 2, 833-838.
(24) Castonguay, A., Rivenson, A., Trushin, N., Reinhardt, J ., Spatho-
poulos, S., Weiss, C. J ., Reiss, B., and Hecht, S. S. (1984) Effects
of chronic ethanol consumption on the metabolism and carcino-
genicity of N′-nitrosonornicotine in F344 rats. Cancer Res. 44,
2285-2290.
(26) Gopalakrishnan, R., Morse, M. A., Lu, J ., Weghorst, C. M.,
Sabourin, C. L. K., Stoner, G. D., and Murphy, S. E. (1999)
Expression of cytochrome P450 2A3 in rat esophagus: relevance
to N-nitrosobenzylmethylamine. Carcinogenesis 20, 885-891.
(27) von Weymarn, L. B., Felicia, N. D., Ding, X., and Murphy, S. E.
(1999) N-Nitrosobenzylmethylamine R-hydroxylation and cou-
marin 7-hydroxylation: Catalysis by rat esophageal microsomes
and cytochrome P450 2A3 and 2A6 enzymes. Chem. Res. Toxicol.
12, 1254-1261.
(28) Chen, S. C., Wang, X., Xu, G., Zhou, L., Vennerstrom, J . L.,
Gonzalez, F., Gelboin, H. V., and Mirvish, S. S. (1999) Depenty-
lation of [3H-pentyl]methyl-n-amylnitrosamine by rat esophageal
and liver microsomes and by rat and human cytochrome P450
isoforms. Cancer Res. 59, 91-98.
(29) Wade, D., Yang, C. S., Metral, C. J ., Roman, J . M., Hrabie, J . A.,
Riggs, C. W., Anjo, T., Keefer, L. K., and Mico, B. A. (1987)
Deuterium isotope effect on denitrosation and demethylation of
N-nitrosodimethylamine by rat liver microsomes. Cancer Res. 47,
3373-3377.
(25) Brittebo, E. B., Castonguay, A., Furuya, K., and Hecht, S. S.
(1983) Metabolism of tobacco specific nitrosamines by cultured
rat nasal mucosa. Cancer Res. 43, 4343-4348.
TX990171L