Reactions of beta-carotene with cigarette smoke oxidants. Identification of carotenoid oxidation products and evaluation of the prooxidant/antioxidant effect.

Article abstract of DOI:10.1021/tx980263v

Recent intervention trials reported that smokers given dietary beta-carotene supplementation exhibited an increased risk of lung cancer and overall mortality. beta-Carotene has been hypothesized to promote lung carcinogenesis by acting as a prooxidant in the smoke-exposed lung. We have examined the interactions of cigarette smoke with beta-carotene in model systems. Both whole smoke and gas-phase smoke oxidized beta-carotene in toluene to several products, including carbonyl-containing polyene chain cleavage products and beta-carotene epoxides. A major product of the reaction was identified as 4-nitro-beta-carotene, which was formed by nitrogen oxides in smoke. Both cis and all-trans isomers of 4-nitro-beta-carotene were detected. The hypothesis that smoke-driven beta-carotene autoxidation exerts prooxidant effects was tested in a liposome system. Lipid peroxidation in dilinoleoylphosphatidylcholine liposomes exposed to gas-phase smoke was modestly inhibited by the incorporation of 0.1 mol % beta-carotene. Both the lipid soluble antioxidant alpha-tocopherol and the water soluble antioxidant ascorbate were oxidized more slowly by gas-phase smoke exposure in liposomes containing beta-carotene. These data indicate that beta-carotene exerts weak antioxidant effects against smoke-induced oxidative damage in vitro. It is unlikely that a prooxidant effect of beta-carotene occurs under biologically relevant conditions or is responsible for an increased incidence of lung cancer observed in smokers who consume beta-carotene supplements.

Full text of DOI:10.1021/tx980263v

Chem. Res. Toxicol. 1999, 12, 535-543
535
Rea ction s of â-Ca r oten e w ith Ciga r ette Sm ok e Oxid a n ts.
Id en tifica tion of Ca r oten oid Oxid a tion P r od u cts a n d
Eva lu a tion of th e P r ooxid a n t/An tioxid a n t Effect
Daniel L. Baker,†,‡ Ed S. Krol,† Neil J acobsen,§ and Daniel C. Liebler*,†
Department of Pharmacology and Toxicology, College of Pharmacy, and Department of Chemistry,
University of Arizona, Tucson, Arizona 85721-0207
Received December 14, 1998
Recent intervention trials reported that smokers given dietary â-carotene supplementation
exhibited an increased risk of lung cancer and overall mortality. â-Carotene has been
hypothesized to promote lung carcinogenesis by acting as a prooxidant in the smoke-exposed
lung. We have examined the interactions of cigarette smoke with â-carotene in model systems.
Both whole smoke and gas-phase smoke oxidized â-carotene in toluene to several products,
including carbonyl-containing polyene chain cleavage products and â-carotene epoxides. A major
product of the reaction was identified as 4-nitro-â-carotene, which was formed by nitrogen
oxides in smoke. Both cis and all-trans isomers of 4-nitro-â-carotene were detected. The
hypothesis that smoke-driven â-carotene autoxidation exerts prooxidant effects was tested in
a liposome system. Lipid peroxidation in dilinoleoylphosphatidylcholine liposomes exposed to
gas-phase smoke was modestly inhibited by the incorporation of 0.1 mol % â-carotene. Both
the lipid soluble antioxidant R-tocopherol and the water soluble antioxidant ascorbate were
oxidized more slowly by gas-phase smoke exposure in liposomes containing â-carotene. These
data indicate that â-carotene exerts weak antioxidant effects against smoke-induced oxidative
damage in vitro. It is unlikely that a prooxidant effect of â-carotene occurs under biologically
relevent conditions or is responsible for an increased incidence of lung cancer observed in
smokers who consume â-carotene supplements.
a steady-state concentration of radicals resulting from
nitrogen oxide-driven autoxidation of low-molecular weight
hydrocarbons present in the smoke stream (11).
In tr od u ction
Lung cancer is the leading cause of cancer death in
the United States for both men and women (1). Obser-
vational epidemiology has shown that increased dietary
intakes and elevated blood levels of â-carotene are
associated with a decreased risk of lung cancer (2-4).
However, recent intervention trials in heavy smokers
indicated increased lung cancer incidences and deaths
in those participants receiving â-carotene supplementa-
tion versus placebo (5, 6). These results raise questions
about the mechanism(s) by which supplemental â-caro-
tene enhanced lung cancer in these cohorts.
Despite considerable interest in â-carotene and lung
cancer in the past decade, little is known about the
chemistry and consequences of interactions between
cigarette smoke and â-carotene. Handelman et al. have
shown that â-carotene is depleted from human plasma
exposed to gas-phase cigarette smoke, but the resulting
products have not been examined (12). â-Carotene is
known to react with NO2 in hexane solution, although
oxidation products have not been characterized (13).
Everett and co-workers have shown by pulse radiolysis
that NO2 oxidizes â-carotene in solution via the â-caro-
tene radical cation (eq 1) (14, 15).
Cigarette smoke is a complex mixture of literally
thousands of compounds, many of which are known or
suspected human carcinogens (7, 8). Smoke may be
divided into two main fractions on the basis of separation
by a glass fiber filter (8). Tar is defined as that portion
of whole smoke trapped by the filter and contains
quinone/hydroquinone redox couples (8-10) that are
capable of generating superoxide anion catalytically (8).
Gas-phase smoke passes through the filter and contains
Car + NO2 f Car+• + NO2
(1)
-
Peroxyl radicals, which are abundant in smoke (8, 11),
have been shown previously to oxidize â-carotene to
epoxides and to aldehyde and ketone chain cleavage
products (16-19) (see Figure 1 for structures discussed
in the text).
â-Carotene is known to exert antioxidant effects in
model systems, apparently by acting as a chain-breaking
antioxidant. However, Burton and Ingold reported that
â-carotene antioxidant activity is diminished at high
oxygen tensions (20). This apparently reflects the ability
of oxygen to add reversibly to â-carotene-derived radicals,
thus generating peroxyl radicals that facilitate carotenoid
* To whom correspondence should be addressed: Department of
Pharmacology and Toxicology, College of Pharmacy, The University
of Arizona, P.O. Box 210207, Tucson, AZ 85721-0207. Telephone: (520)
626-4488. Fax: (520) 626-2466. E-mail: liebler@pharmacy.arizona.edu.
† Department of Pharmacology and Toxicology, College of Pharmacy.
‡ Present address: Department of Physiology and Biophysics,
University of Tennessee at Memphis, 894 Union Ave., Memphis, TN
38163.
§ Department of Chemistry.
10.1021/tx980263v CCC: $18.00 © 1999 American Chemical Society
Published on Web 05/19/1999

536 Chem. Res. Toxicol., Vol. 12, No. 6, 1999
Baker et al.
otene in toluene and identified 4-nitro-â-carotene as a
novel smoke-dependent oxidation product.
Exp er im en ta l P r oced u r es
Ch em ica ls. All-trans â-carotene was purchased from Fluka
Chemical Corp. (Ronkonkoma, NY). Research grade cigarettes
(1R3) were obtained from the University of Kentucky Tobacco
and Health Research Institute (Lexington, KY). Cambridge filter
pads were obtained from Performance Systematics (Caledonia,
MI). Linoleic acid was purchased from Nu Chek Prep (Elysian,
MN). 12-Hydroxylauric acid, soy lipoxidase (type I, EC 1.13.11.12),
L-ascorbic acid, and ferrozine were obtained from Sigma Chemi-
cal Co. (St. Louis, MO). Rhodium (5% on alumina), PtO2,
pentafluorobenzoyl chloride, and diisopropylethylamine were
purchased from Aldrich Chemical Co. (Milwaukee, WI). D-R-
Tocopherol was a gift from Henkel Fine Chemicals (LaGrange,
IL). BLPC1 and DLPC were obtained from Avanti Polar Lipids
(Alabaster, AL). All other chemicals were of the highest purity
available and were used without further purification.
In str u m en ta tion . Product analyses were performed with a
Hewlett-Packard model 1040M diode-array detector coupled to
a Hewlett-Packard 1050 four-channel gradient pump (Hewlett-
Packard, Palo Alto, CA). Hewlett-Packard Chemstation software
(version A.02.02) was utilized for system control and data
analysis. LC/MS was performed on a Finnigan TSQ-7000 triple-
quadrupole mass spectrometer with an atmospheric pressure
chemical ionization source (Finnigan MAT, San J ose, CA).
All NMR spectra were acquired on a Bruker AM-500 spec-
trometer operating at a 1H frequency of 500.13 MHz, using an
inverse broad-band 5 mm probe (22). Samples were dissolved
in 0.6 mL of benzene-d6 at low light levels, and spectra were
acquired at 30 °C with lights off and the spectrometer bore
capped. All two-dimensional (2D) spectra were acquired without
spinning in the TPPI (23) mode with 64 repetitions per t1 value
by collecting 2048 complex data points in t2 and 750 real data
points in t1. For inverse experiments, no 13C decoupling was used
during acquisition. NMR data were analyzed using the Felix95
software package (BIOSYM/Molecular Simulations, San Diego,
CA). A skewed, 45°-shifted sine-bell window was used for HMQC
(24, 25) spectra, and a sine-bell window function was used for
DQF-COSY (26) spectra. In both cases, zero-filling was employed
to give a final matrix of 2048 (F2) × 1024 (F1) real data points.
The 2D spectra included a 1H chemical shift range from 8 to 0
ppm and a 13C chemical shift range from 150 to 0 ppm. In
HMQC spectra of U-13C samples, the cross-peaks were visibly
F igu r e 1. Structures and numbering schemes for all-trans
â-carotene and the â-carotene oxidation products that are
discussed herein.
1
broadened in the F1 dimension by J CC coupling, but this did
not significantly affect peak overlap or assignments. The
1
residual C6H5D peak was used as a reference for H (7.158 ppm)
and 13C (128.33 ppm).
autoxidation (eqs 2-4).
Capillary gas chromatography was performed on a Hewlett-
Packard 5890 series II gas chromatograph equipped with a
Hewlett-Packard 7673 autosampler, a cool on-column injector,
and an electron capture detector (Hewlett-Packard, Avondale,
PA). Chemstation software 3365 series II was used for system
control and data processing. Samples were analyzed on a 0.53
mm × 20 m DB-5 capillary column (J &W Scientific, Folsom,
CA).
Car + R• f R-Car•
R-Car• + O2 f R-Car-OO•
(2)
(3)
R-Car-OO• + other molecules f oxidative damage
(4)
GC/MS was performed on a Fisons-VG MD-800 analyzer
coupled to a Carlo Erba 8000 series gas chromatograph fitted
This oxygen-driven â-carotene autoxidation has been
offered as the basis for possible prooxidant effects of
â-carotene. Recently, Mayne et al. proposed that smoke-
driven â-carotene autoxidation may enhance smoke-
induced oxidative damage in the lungs of smokers taking
supplemental â-carotene (21).
To test the hypothesis that â-carotene amplifies smoke-
induced oxidative damage, we have investigated the
oxidation of â-carotene by smoke and the effect of
â-carotene on smoke-induced lipid peroxidation and
antioxidant consumption in homogeneous solution and
liposomal model systems. We have characterized the
products of cigarette smoke-dependent oxidation of â-car-
with
a Fisons A200S autosampler and on-column injector
(Fisons Instruments, Beverly, MA). Analyses were carried out
in the electron ionization mode at 70 eV. Fisons-VG Lab-Base
system software was used for system control and data process-
ing. Samples were analyzed on a 0.25 mm × 30 m DB-5ms
capillary column (J &W Scientific).
1
Abbreviations: APCI, atmospheric-pressure chemical ionization;
BLPC, bovine liver phosphatidylcholine; DLPC, dilinoleoylphosphati-
dylcholine; DQF-COSY, double-quantum-filtered correlation spectros-
copy; GC/MS, gas chromatography/mass spectrometry; HFAME, hy-
droxy fatty acid methyl ester; HMQC, heteronuclear multiple-quantum
coherence; LC/MS, liquid chromatography/mass spectrometry.

Smoke Oxidation of â-Carotene
Chem. Res. Toxicol., Vol. 12, No. 6, 1999 537
Ciga r ette Sm ok e Oxid a tion of â-Ca r oten e in Solu tion .
â-Carotene (250 µM in toluene) was exposed to whole or gas-
phase 1R3 cigarette smoke, 200 mL every 15 min at 40 °C for
3 h using a smoking apparatus consisting of a 20 mL syringe,
a three-way stopcock, and an in-line filter. Gas-phase smoke
was that fraction of smoke that passed through a Cambridge
glass fiber filter. Unoxidized â-carotene and products resulting
from the oxidation of â-carotene in solution by whole or gas-
phase 1R3 cigarette smoke were separated on a 4.6 mm × 250
mm Allsphere ODS-2 5 µm HPLC column (Alltech Associates,
Deerfield, IL) with gradient elution at a rate of 1.0 mL/min (19).
Initially, the mobile phase consisted of 100% solvent A (85:15:
0.1 v/v/w acetonitrile/methanol/ammonium acetate). At 8 min,
solvent B (2-propanol) was introduced in a linear gradient to
30% by 12 min. From 25 to 30 min, the solvent was returned to
100% solvent A using a linear gradient. Compounds were
detected by monitoring the UV/vis absorbance.
F igu r e 2. Smoke oxidation of â-carotene in toluene solution.
Depletion of â-carotene (250 µM) by gas-phase (9) and whole
smoke (b) and generation of product R-IV by gas-phase (0) and
whole smoke (O) exposure.
Id en tifica tion of â-Ca r oten e Oxid a tion P r od u cts. â-Car-
otene (500 µM in toluene) was oxidized with gas-phase smoke
(200 mL every 5 min) at 40 °C for 1 h. Products were separated
by reverse-phase HPLC on a 10 mm × 300 mm Allsphere ODS-2
5 µm column using a mobile phase of acetonitrile/methanol/2-
propanol (59.5:10.5:30 v/v/v) at a rate of 4 mL/min with detection
at 450 nm. Fraction R-IV was collected for further analysis by
cyano HPLC. Cyano HPLC was carried out on a 10 mm × 300
mm Allsphere cyano column, and compounds were eluted with
hexane/ethyl acetate (99:1 v/v) at a rate of 4.0 mL/min with
detection at 450 nm (17). Individual products were isolated and
characterized by LC/MS and NMR.
2:1 v/v chloroform/methanol. The supernatants were evaporated
under N2, and the fatty acids were transesterified with 0.5 M
KOH in methanol (0.5 mL) at room temperature for 45 min.
The reaction was terminated by the addition of 2.0 mL of 2.0 M
phosphate buffer, and hydroxy fatty acid methyl esters were
extracted with 4.5 mL of hexane. The extracts were evaporated
under N2 and derivatized with 1.5% pentafluorobenzoyl chloride
and 0.75% diisopropylethylamine in toluene (325 µL) at 65 °C
for 60 min. Samples were analyzed by capillary gas chroma-
tography with electron capture detection; 1.0 µL injections were
made on-column at 100 °C. After 1 min, the oven temperature
was increased to 210 °C at a rate of 25 °C/min and finally to
270 °C at a rate of 4.5 °C/min for a total analysis time of 33
min. The injector temperature was maintained 3 °C above the
oven temperature, and the detector was maintained at 275 °C
throughout the run.
Atmospheric-pressure chemical ionization LC/MS was per-
formed on a 4.6 mm × 250 mm Allsphere cyano 5 µm column,
with a mobile phase of hexane/ethyl acetate (99:1 v/v) at a rate
of 1.0 mL/min. Mass spectra were obtained in both the positive
and negative modes. Double-quantum-filtered 1H COSY spectra
were obtained in benzene-d6 at 30 °C.
An a lysis of An tioxid a n ts in BLP C Lip osom es Oxid ized
w ith Ciga r ette Sm ok e. Stock solutions of bovine liver phos-
phatidylcholine (30 µmol) and appropriate amounts of â-carotene
or R-tocopherol were evaporated together under N2 and resus-
pended in 250 µL of ethanol (30). The ethanol mixture was
injected into phosphate-buffered saline that was being rapidly
stirred to generate small unilamellar vesicles. In some experi-
ments, 1 mmol of ascorbate was added to each 10 mL of
phosphate-buffered saline solution immediately prior to ethanol
injection.
â-Carotene was analyzed by reverse-phase HPLC using a
Spherisorb ODS-2 5 µm column (Alltech Associates) eluted with
methanol/hexane (85:15 v/v) at a rate of 1.5 mL/min. Detection
was by UV/vis absorbance at 450 nm. Samples (0.5 mL) were
extracted with 1.5 mL of ethyl acetate, supplemented with
R-tocopherol propionate as an internal standard, and dried
under N2. R-Tocopherol was assessed by GC/MS as described
previously (31). Ascorbate was assessed spectrophotometrically
through a modification of the method of Butts and Mulvihill
(32). Samples consisted of 200 µL of liposomal suspension, 450
µL of 2.2 mM ferrozine, 50 µL of 8.3 mM ferric ammonium
sulfate, 400 µL of water, and 900 µL of ethanol. Samples were
analyzed by UV/vis absorbance at 562 nm.
Oxid a tion of [13C]-â-Ca r oten e. â-Carotene (98% 13C) was
obtained from Martek Corp. (Columbia, MD) as a crude algal
hexane extract, which contained all-trans â-carotene as well as
several mono cis â-carotene isomers. Mono cis isomers of
â-carotene were converted to the all-trans form using the
method of Foote et al. (27). All-trans â-carotene was purified
by reverse-phase HPLC on a 10 mm × 300 mm Allsphere ODS-2
5 µm column with a mobile phase of methanol/hexane (85:15
v/v) at a rate of 4.5 mL/min. The structure and isotopic purity
of all-trans per-[13C]-â-carotene were analyzed by UV/vis ab-
sorbance, LC/MS, and NMR. Purified all-trans per-[13C]-â-
carotene (150-200 µM in toluene) was oxidized with 100 mL of
gas-phase smoke every 5 min at 40 °C for 1 h. Products were
anlayzed by reverse-phase and cyano HPLC as described above.
Individual 13C-labeled â-carotene oxidation products were col-
lected for NMR characterization by HMQC (28) in benzene-d6
at 30 °C.
An a lysis of HF AME in DLP C Lip osom es Oxid ized w ith
Ciga r ette Sm ok e. DLPC liposomes were prepared by ethanol
injection (29) with or without incorporation of 0.1 mol %
â-carotene. The liposomal suspensions were oxidized with 200
mL of whole or gas-phase cigarette smoke every 15 min for 3 h
at 40 °C. Samples (250 µL) were drawn and added to 750 µL of
water and 100 µL of BHT (10 mM in ethanol). Lipids were
extracted with 4.5 mL of chloroform/methanol (2:1 v/v). The
pooled organic fractions were evaporated under N2. Lipid
hydroperoxides were reduced to corresponding alcohols with
NaBH4 (ca. 1 mg) in 0.5 mL of methanol. The reaction was
terminated with the addition of 1.0 mL of water and the product
extracted with 4.5 mL of chloroform/methanol (2:1 v/v). The
organic layer was evaporated under N2 in screw-cap Reactivials
(Pierce, Rockford, IL). The extent of fatty acyl unsaturation was
reduced with 5% rhodium on alumina (ca. 1 mg) and H2 in 0.5
mL of methanol at room temperature for 30 min. Following
centrifugation, the methanol supernatant was supplemented
with 2.5 nmol of 12-hydroxymethyl laurate internal standard
and removed and the catalyst was washed twice with 1.0 mL of
Sta tistica l An a lysis. Data were tested for statistical sig-
nificance (p < 0.05) using ANOVA.
Resu lts
Cigar ette Sm oke Oxidation of â-Car oten e in Tolu -
en e. â-Carotene was consumed by cigarette smoke
exposure in toluene solution. Unfiltered whole smoke
depleted the â-carotene level by 27% over 3 h at 40 °C,
whereas gas-phase smoke depleted the â-carotene level
by 75% under the same conditions (Figure 2). Both gas-
phase and whole cigarette smoke treatment of â-carotene
in solution generated several oxidation products (Figures
3 and 4). The distribution of products appeared to be

538 Chem. Res. Toxicol., Vol. 12, No. 6, 1999
Baker et al.
F igu r e 3. Reverse-phase HPLC product profile from gas-phase
smoke oxidation of â-carotene (500 µM) in toluene solution.
F igu r e 5. Cyano HPLC separation of reverse-phase HPLC-
purified product R-IV generated from gas-phase smoke oxidation
of â-carotene (500 µM) in toluene solution.
Ta ble 2. UV/Vis a n d LC/MS Da ta for Com p ou n d s C-I-VII
P u r ified by Cya n o Ch r om a togr a p h y fr om P r od u ct R-IV
UV/vis λmax
positive APCI
negative APCI
peak
(nm)
(m/z, [M + H]+)
(m/z, [M - H]-)
C-I
C-II
338, 445,a 471
339, 445, 471
339, 445, 476
460
342, 456
342, 454
NDb
ND
ND
582
582
582
582
580
580
580
ND
ND
ND
ND
C-III
C-IV
C-V
C-VI
C-VII
346, 452
a
b
The most intense absorbance bands are underlined. Not
detected.
F igu r e 4. Reverse-phase HPLC product profile from whole
smoke oxidation of â-carotene (500 µM) in toluene solution.
where â-carotene solutions were exposed to NO2 in air
(30 ppm), product profiles were virtually identical to
those from smoke exposures (data not shown).
Ta ble 1. Ma jor P r od u cts Id en tified by Rever se-P h a se
HP LC An a lysis of Ga s-P h a se Sm ok e Oxid a tion s of
â-Ca r oten e in Tolu en e Solu tion
Peak R-IV was collected and subjected to further
fractionation by cyano HPLC which resolved peak R-IV
into two distinct groups of products (Figure 5 and Table
2). The first consisted of compounds C-I-III, which all
produced strong signals at m/z 580 in negative ion APCI
LC/MS analysis and weak signals at m/z 582 in the
positive ion mode. The second group contained com-
pounds C-IV-VI, and C-VII, which exhibited signals at
m/z 582 in positive ion APCI LC/MS, but weak signals
at m/z 580 in the negative ion mode. Thus, all of these
products appeared to result from substitution of one
â-carotene hydrogen with two oxygens and one nitrogen.
Products C-I-VII were treated with methylene blue
and light according to the method of Foote et al. (27) to
convert cis isomers to the thermodynamically favored all-
trans form. C-I and C-II both converted to C-III, as
assessed by retention time and UV/vis absorbance changes
(Table 3). Likewise, methylene blue/light treatment
converted compounds C-VI and C-VII to C-V (Table 3).
Thus, C-I and C-II were identified as cis isomers of C-III,
and C-VI and C-VII were identified as cis isomers of C-V.
Unfortunately, these compounds proved to be relatively
unstable, and generating sufficient quantities (1 mg) of
all seven for NMR analyses was not possible. To increase
sensitivity in heteronuclear experiments, per-[13C]-â-
carotene was used to generate products labeled at all
carbon positions with 13C. Analysis of the oxidation
products by HMQC permitted 13C shifts of all nonqua-
ternary carbons to be observed in samples containing ca.
100 µg of material. Because C-III was both the most
abundant and apparently the most stable of the com-
pounds, both HMQC (on per-13C-labeled material) and
retention UV/vis λmax positive APCI
peak
R-I
identity
time (min)
(nm)
(m/z, [M + H]+)
â-apo-14′-carotenal
7.2
9.8
398a
422
442
454
428, 452
446, 474
452, 476
311
351
377
582
553
553
537
R-II â-apo-12′-carotenal
R-III â-apo-10′-carotenal
R-IV 4-nitro-â-carotene
R-V 5,8-epoxy-â-carotene
R-VI 5,6-epoxy-â-carotene
R-VII â-carotene
10.6
16.9
18.3
19.8
25.2
a
The most intense absorbance bands are underlined.
similar with both treatments. However, gas-phase smoke
formed greater quantities of these products than did
whole smoke. Among the â-carotene oxidation products
identified by HPLC retention, UV/vis absorbance, and
LC/MS were â-carotene chain cleavage products and
â-carotene epoxides (Table 1). Spectral data and HPLC
retention times corresponded to previously published
data for these compounds (16, 17, 19). In addition, a novel
product, R-IV, eluted at 16.9 min in this system (Figures
3 and 4). R-IV eluted prior to the â-carotene epoxides,
but after the polar chain cleavage products, suggesting
a compound with intermediate polarity. The initial
characterization of R-IV by positive ion LC/MS showed
a signal at m/z 582, which corresponds to a [M + H]+ ion
for nitro-â-carotene, in which a â-carotene hydrogen is
replaced by a nitro group. CID of this ion resulted in a
prominent ion at m/z 536 corresponding to the loss of the
nitro group, m/z 46 (data not shown). The time course of
formation of this product fraction by whole and gas-phase
smoke exposures is shown in Figure 2. In experiments

Smoke Oxidation of â-Carotene
Chem. Res. Toxicol., Vol. 12, No. 6, 1999 539
4.58 ppm (1H) and 87.9 ppm (13C) with a one-bond J CH
value of 147 Hz, typical of a CH group with a single
electronegative substituent. In the DQF-COSY spectrum
of natural abundance C-III, this resonance (4.58 ppm) is
correlated to two peaks at 1.67 and 2.02 ppm. These two
resonances correlate in the HMQC spectrum of per-
[13C]C-III to a single 13C position (24.8 ppm), indicating
an X-CH-CH2 fragment. The resonance at 2.02 ppm is
also correlated in the DQF-COSY spectrum to a reso-
nance at 1.08 ppm which is part of another CH2 fragment.
The fragment X-CH-CH2-CH2 can only result from nitra-
tion of â-carotene at the 4-position, and this is the basis
of all further assignments of C-III (see Table 4). Con-
sistent with this conclusion is the fact that HMQC cross-
peaks are not found near the 2′, 3′, and 4′ cross-peaks
for C-III, while all other cross-peaks are located near and
have structures similar to their counterparts in the
unsubstituted half of the structure. The observation of
Ta ble 3. Meth ylen e Blu e/Ligh t Tr ea tm en t of Com p ou n d s
Isola ted by Cya n o Ch r om a togr a p h y fr om Com p ou n d
R-IV
untreated
methylene blue/light-treated
retention
peak time (min)
UV/vis λmax
retention
time (min)
UV/vis λmax
(nm)
(nm)
C-I
C-II
5.8
6.2
6.7
7.0
7.4
8.0
9.3
336, 442,a 468
338, 444, 470
450, 476
462
460, 486
452
6.7
6.7
6.7
7.0
7.3
7.2
7.3
450, 476
450, 476
450, 476
460, 486
460, 488
460, 488
458, 486
C-III
C-IV
C-V
C-VI
C-VII
454
.a
The most intense absorbance bands are underlined.
Ta ble 4. NMR Ch a r a cter iza tion of P r od u ct C-III a n d
Com p a r ison w ith All-Tr a n s â-Ca r oten e
1H (ppm)
13C (ppm)
positiona C-III âCb multc (Hz) C-III âCd (Hz) COSY
J HH
1J CH
1
five distinct H chemical shifts for the five protons in the
spin system (2a, 2b, 3a, 3b, and 4) is consistent with the
creation of a chiral center at C-4. These assignments were
confirmed by the DQF-COSY spectrum of natural abun-
dance C-III, which exhibits long-range correlations from
the methyl signals at positions 18, 19, and 20 to olefinic
resonances 7, 10, and 14, respectively. Similar correla-
tions were observed on the unsubsituted half of the
molecule (18′, 19′, and 20′ to 7′, 10′, and 14′, respectively).
Comparison of chemical shifts of C-III with those of per-
[13C]-â-carotene shows that the only positions with 13C
chemical shift differences of >2 ppm are 4, 2, 3, 18, and
7 (in order of decreasing differences), all within or directly
attached to the six-membered ring. 13C shift differences
of >1 ppm are also observed for carbons 17, 8, and 10,
the latter two indicating a change in the conjugation of
the unsaturated system. Likewise, the only positions with
1H chemical shift differences of >0.15 ppm are 4, 2b, 3a,
2a
2b
3a
3b
4
7
1.56 1.50
1.08
2.02 1.61
1.67
m
m
m
m
34.5 40.0
z
z
2b, 3b
2a, 3a
25.2 19.7 128 2b, 3b, 4
123 2a, 3a, 4
88.2 33.4 147 3a, 3b
124.3 126.9 153 8, 18
140.5 138.7 153
133.5 131.9 139 11, 19
4.58 1.98 br t
5
6.03 6.33
6.26 6.39
6.23 6.34
6.72 6.78 dd
6.50 6.47
6.35 6.32 ol
6.67 6.68
1.00 1.15
0.86 1.15
1.69 1.89
1.83 1.94
1.87 1.88
1.50 1.50
1.61 1.61
1.98 1.98 br t
6.33 6.33 ol
6.39 6.39
6.34 6.34 ol
6.79 6.78 dd
6.47 6.47
d
d
d
16
16
11
8
7
10
11
12
14
15
16
17
18
19
20
2′ab
3′ab
4′ab
7′
11, 15 125.0 125.5 147 10, 12
d
15
139.1 138.1 152 11
133.9 133.2 145 15, 20
130.7 130.7 144 14
28.8 29.2 123
27.4 29.2 123
19.2 22.0 125 7
m
s
s
s
s
s
m
m
12.6 12.9 124 10
12.8 12.9 122 14
40.0 40.0 121 3′ab
19.7 19.7 121 2′ab, 4′ab
33.3 33.4 122 3′ab, 18′
127.0 126.9 148 4′ab, 18′
138.6 138.7 152 7′
6
1
7, and 17 (in order of decreasing differences). The H and
13C chemical shifts on the unsubstituted half of the
molecule are essentially identical to those for all-trans
â-carotene.
8′
d
16
10′
11′
12′
14′
15′
131.8 131.9 149 11′, 19′
11, 15 125.7 125.5 143 10′, 12′
d
15
138.0 138.1 148 11′
133.1 133.2 144 15′, 20′
130.7 130.7 144 14′
29.2 29.2 120
22.0 22.0 122 4′ab, 7′
12.8 12.9 123 10′
12.8 12.9 122 14′
DQF-COSY experiments performed on natural abun-
dence C-I and C-II confirmed that these compounds were
cis isomers of 4-nitro-â-carotene, as was suggested previ-
ously. Each peak appeared to be a mixture of mono cis
isomers where the cis bond was on both the same and
opposite halves of the structure with respect to the nitro
adduct. C-III constituted up to 30% and C-I-III up to
60% of the cyano HPLC-purified material resulting from
gas-phase cigarette smoke oxidation of â-carotene in
toluene solution (200 mL every 5 min at 40 °C for 1 h).
Lip id Oxid a tion in DLP C Lip osom es Exp osed to
Ciga r ette Sm ok e. To test the hypothesis that smoke-
driven â-carotene autoxidation exerts prooxidant effects,
we studied the effects of â-carotene on smoke-induced
lipid peroxidation of phospholipids in a liposome system.
Gas-phase smoke exposure resulted in a significant
increase in the extent of peroxidation of DLPC liposomes
as measured by the formation of HFAME (Figure 6). This
result was consistent for both the 9- and 13-OH isomers
of linoleic and total HFAME. Interestingly, the extent of
formation of the 9-OH appeared to exceed that of the 13-
OH in experiments with both gas-phase and whole
smoke. Incorporation of 0.1 mol % â-carotene in DLPC
liposomes slightly attenuated lipid peroxidation, but the
reduction in the extent of HFAME formation was not
statistically significant.
6.32 6.32 ol
6.68 6.68
m
s
s
s
s
16′/17′ 1.15 1.15
18′
19′
20′
1.81 1.81
1.94 19.4
1.88 1.88
a
b
Refer to Figure 5. per-[13C]-â-Carotene shifts from HMQC
included for comparison. c Multiplicity from one-dimensional 1H
spectrum of unlabled material: s, singlet; d, doublet; t, triplet;
br, broad; ol, overlapped; and m, multiplet. J CH
d
.
DQF-COSY (on natural abundance material) were used
to assess this product. â-Carotene exhibits C-2 symmetry
through the 15,15′ double bond, so both H and 13C NMR
1
resonances are identical for corresponding positions on
either side of the structure. However, once nitration
occurs, this symmetry is lost and the spectra become very
complex, particularly if substitution creates a chiral
center. HMQC data for per-[13C]C-III are shown in Table
4.
Nonquaternary positions on the nonadducted half of
the molecule in product per-[13C]C-III (positions 2′-20′)
show 1H and 13C chemical shifts which are nearly
identical to those in all-trans â-carotene. For consistency,
the HMQC spectrum of all-trans per-[13C]-â-carotene was
used for comparison. The most notable change in the
HMQC spectrum of per-[13C]C-III is the cross-peak at

540 Chem. Res. Toxicol., Vol. 12, No. 6, 1999
Baker et al.
F igu r e 6. Gas-phase smoke-dependent lipid peroxidation,
determined as hydroxy fatty acid methyl esters (HFAME), in
DLPC liposomes containing 0.0 mol % â-carotene (9) or 0.1 mol
% â-carotene (0). Groups with the same letters are significantly
different (p < 0.05).
F igu r e 9. Gas-phase smoke depletion of R-tocopherol from
BLPC liposomes containing 0.0 mol % â-carotene (9), 0.01 mol
% â-carotene (b), 0.1 mol % â-carotene (2), or 0.4 mol %
â-carotene ([).
F igu r e 7. Whole smoke-dependent lipid peroxidation, deter-
mined as hydroxy fatty acid methyl esters (HFAME), in DLPC
liposomes containing 0.0 mol % â-carotene (9) or 0.1 mol %
â-carotene (0). Groups with the same letters are significantly
different (p < 0.05).
F igu r e 10. Gas-phase smoke depletion of ascorbate from BLPC
liposomes containing 0.0 mol % â-carotene (9), 0.01 mol %
â-carotene (b), 0.1 mol % â-carotene (2), or 0.4 mol % â-carotene
([).
decreased â-carotene levels by 88% within 3 h. In BLPC
liposomes supplemented with R-tocopherol and ascorbate,
little or no depletion of either antioxidant was observed
upon whole smoke exposure (data not shown), whereas
significant depletion occurred with gas-phase exposures
(see below).
The oxidation of R-tocopherol and ascorbate was ex-
amined in BLPC liposomes exposed to gas-phase smoke
in the presence of varying concentrations of â-carotene.
The carotenoid produced a modest R-tocopherol sparing
effect that was statistically significant at 0.01 and 0.4
mol % but not at 0.1 mol % â-carotene (Figure 9).
Likewise, 100 mM ascorbate was modestly spared from
oxidation when BLPC liposomes were supplemented with
0.01 mol % â-carotene, but not 0.1 or 0.4 mol % â-carotene
(Figure 10). These data collectively indicate that â-car-
otene did not accelerate the smoke-induced oxidation of
R-tocopherol and ascorbate, but produced instead a
marginal sparing of the two antioxidants.
F igu r e 8. Depletion of â-carotene (0.1 mol %) by gas-phase
(9) and whole smoke (b) in BLPC liposomes.
Whole smoke exposure also significantly increased the
extent of lipid peroxidation of DLPC liposomes over that
in control liposomes (Figure 7). Incorporation of 0.1 mol
% â-carotene into the liposomal membrane resulted in a
small but statistically significant decrease in the extent
of formation of HFAME. These results indicate that
â-carotene did not increase the extent of HFAME forma-
tion in DLPC liposomes oxidized with either gas-phase
or whole smoke.
Discu ssion
â-Carotene autoxidation driven by smoke-borne oxi-
dants has been hypothesized to amplify smoke-induced
oxidative damage and contribute to increased lung cancer
incidence in â-carotene-supplemented smokers. Two ques-
tions are at the center of this hypothesis. First, what is
the chemistry of smoke-driven â-carotene oxidation?
Second, do radical intermediates in â-carotene oxidation
facilitate the oxidation of other biomolecules? Here we
have characterized the principal products of smoke-
driven â-carotene oxidation, including 4-nitro-â-carotene,
and we have demonstrated that, despite facile smoke-
An tioxid a n t Dep letion in BLP C Lip osom es Ex-
p osed to Ciga r ette Sm ok e. We next examined the
possibility that smoke-driven â-carotene autoxidation
accelerates the depletion of other antioxidants by smoke.
The depletion of â-carotene (0.1 mol %) in BLPC lipo-
somes was measured over time in both whole and gas-
phase smoke exposures. Three hours of exposure to whole
smoke (600 mL total) did not significantly deplete the
â-carotene level (Figure 8), whereas gas-phase smoke
(600 mL total) under the same conditions consistently

Smoke Oxidation of â-Carotene
Chem. Res. Toxicol., Vol. 12, No. 6, 1999 541
Sch em e 1. P r op osed Mech a n ism for th e
Con su m p tion of â-Ca r oten e a n d Gen er a tion of
â-Ca r oten e Oxid a tion P r od u cts by Ciga r ette
Sm ok e Oxid a n ts
dependent oxidation, â-carotene does not enhance the
peroxidation of membrane lipids or the depletion of the
levels of other antioxidants in a liposomal model. The
data strongly suggest that, although â-carotene is readily
oxidized by smoke, prooxidant effects are unlikely to
account for the apparent enhancement of lung cancer in
smokers taking this supplement.
â-Carotene is readily oxidized in toluene solution by
cigarette smoke. Several products result from the oxida-
tion of â-carotene in solution, the most prominent of
which is the novel product 4-nitro-â-carotene. The chain
cleavage products â-apo-10′-, 12′-, and 14′-carotenals and
the â-carotene epoxides 5,6-epoxy-â-carotene and 5,8-
epoxy-â-carotene also were identified on the basis of
retention times in reverse-phase HPLC, UV/vis spectra,
and molecular ions detected by APCI LC/MS (Table 1).
These compounds have been previously identified as
â-carotene oxidation products in several systems (17-
19, 33). All have been reported previously to arise from
peroxyl radical oxidation of â-carotene, and their appear-
ance in this system is consistent with the presence of
peroxyl radicals in smoke (8, 11). The reaction probably
generated chain cleavage products other than those
reported in Table 1 (e.g., ketone products), although they
evidently were not formed in high yield.
Gas-phase cigarette smoke is more effective at deplet-
ing the â-carotene level in solution (Figure 2) and
generating â-carotene oxidation products than is unfil-
tered whole smoke (Figures 3 and 4). Similar effects have
been noted previously in studies of the effect of smoke
and smoke fractions. It has been shown that the tar
fraction of smoke is a complex mixture of quinone/
hydroquine redox couples (8-10). This fraction of smoke
is reducing in nature and therefore can attenuate some
of the oxidizing characteristics of gas-phase smoke.
Smoke contains NO2 (8, 11) and peroxynitrite, both of
which carry out diverse oxidations. â-Carotene oxidation
by NO2/air produced the same products seen with ciga-
rette smoke exposures (data not shown). Taken together,
these considerations suggest that some combination of
volatile peroxyl radicals, nitrogen oxides, and peroxy-
nitrite are the principal smoke-borne oxidants of â-car-
otene.
The reverse-phase HPLC product fraction R-IV con-
tained several nitrogen oxide-modified â-carotene prod-
ucts (Figure 5). Compound C-III was identified as all-
trans 4-nitro-â-carotene which accounted for up to 30%
of these products under some oxidation conditions.
Products C-I-III and C-IV-VII comprised two distinct
compound classes. C-I-III exhibited UV/vis absorbance
maxima near 450 nm with fine structure similar to that
of â-carotene, which suggested that the â-carotene poly-
ene was intact. Fractions C-I and -II also exhibited an
absorbance maximum at 340 nm characteristic of cis
carotenoids. In contrast, compounds C-IV-VI and C-VII
displayed absorbance maxima above 450 nm and lacked
fine structure. C-V-VII also exhibited cis peaks at 340
nm.
LC/MS analysis also suggested differences between
these two distinct families of products. Negative ion APCI
LC/MS detected C-I-III at m/z 580. However, signals at
m/z 580 for the remaining four compounds were weak or
absent. Likewise, positive APCI LC/MS yielded strong
signals at m/z 582 for C-IV-VII, but signals at m/z 582
were weak or absent for C-I-III. Treatment of individual
cyano HPLC-purified compounds with methylene blue
and light (27) led to rapid conversion of cis isomers to
the corresponding all-trans forms. Accordingly, products
C-I and -II were converted to C-III, whereas C-VI and
-VII were converted to C-V.
Thus, the two product groups resolved by cyano HPLC
include (1) all-trans 4-nitro-â-carotene (C-III) and its two
cis isomers, C-I and -II, and (2) an apparently all-trans
product C-V and its cis isomers, C-VI and -VII, all of
which contain one nitrogen and two oxygens. Although
the latter group exhibits apparent molecular masses
identical to that of 4-nitro-â-carotene, they evidently are
different, perhaps in the arrangement of the nitrogen
oxide substituent. These products could be nitroso ad-
ducts of â-carotene epoxides, although this possibility
seems unlikely because epoxidation would necessarily
consume one polyene double bond and decrease the UV/
vis absorbance maximum well below the observed 452-
462 nm. These latter products may be nitrosooxy-
substituted â-carotene oxidation products formed either
by NO2/N2O4 or by peroxynitrite (34, 35). The nitrosooxy
substituent would be more easily protonated in APCI LC/
MS to generate the observed strong [M + H]+ ion and
correspondingly less able to stabilize an [M - H]- ion.
Products C-IV-VII are formed in smaller amounts and
appear less stable than C-I-III, and further work will
be required to unambiguously establish their identities.
A mechanistic scheme for the smoke-mediated oxida-
tion of â-carotene is proposed in Scheme 1. â-Carotene
can react with either NO2 or peroxyl radicals to generate
a â-carotene radical (15, 36, 37), likely through the
formation of the cation radical (eq 1) (14, 15). Deproton-
ation and subsequent reaction with a second NO2 can
lead to formation of a nitro or nitrosooxy product,
whereas other reactions of â-carotene radical intermedi-
ates with peroxyl radicals and oxygen form epoxides,
chain cleavage products, and â-carotene nonradical ad-
ducts as described previously (17, 18, 36).
The reactivity of â-carotene at the 4 position has been
noted previously. El-Tinay and Chichester showed that
oxidation of â-carotene in solution by N-bromosuccin-
imide generated 4-oxo-â-carotene (38). Woodall et al. (39)
reported 4-methoxy- and 4-ethoxy-â-carotene as products
of peroxyl radical oxidations of â-carotene in solutions

542 Chem. Res. Toxicol., Vol. 12, No. 6, 1999
Baker et al.
containing methanol and ethanol, respectively. Likewise,
Samokyszyn and Marnett showed that hydroperoxide-
dependent cooxidation of 13-cis-retinoic acid by prosta-
glandin H synthase generated 4-hydroxy-13-cis-retinoic
acid (40). Carbon 4 is one of four positions in â-carotene
allylic to the polyene and is the most electron rich, as it
resides at the terminus of the polyene structure (38).
Con clu sion s
Cigarette smoke induces rapid â-carotene autoxidation
with the concomitant formation of 4-nitro-â-carotene, a
novel product derived from smoke-borne reactive nitrogen
species. This product may serve as a useful marker for
the interaction of â-carotene with smoke in biological
systems. Although smoke induces rapid â-carotene aut-
oxidation, a â-carotene “prooxidant effect” does not
amplify smoke-induced oxidation of liposomal membrane
lipids or the other antioxidants R-tocopherol or ascorbic
acid. It is unlikely that prooxidant effects of â-carotene
account for the increased incidence of lung cancer in
smokers taking â-carotene supplements.
This work is the first to describe 4-nitro-â-carotene as
a characteristic marker product for the interactions of
â-carotene with cigarette smoke. This product also is
formed in NO2-mediated oxidations of â-carotene and is,
to our knowledge, the first reported oxidation product of
â-carotene resulting specifically from exposure to a
reactive nitrogen species. Other recent reports have
shown that γ-tocopherol is an effective scavenger of
reactive nitrogen species (41-43). These interactions
generate 5-nitro-γ-tocopherol, a marker product for in-
teractions of γ-tocopherol with peroxynitrite (42), NO2
(41, 43), and smoke (D. L. Baker and D. C. Liebler,
unpublished observation). Although â-carotene may also
function as a trap for reactive nitrogen species, as well
as reactive oxygen species in vivo, â-carotene autoxida-
tion may limit its antioxidant effectiveness. Indeed, the
key question underlying this investigation is whether
â-carotene autoxidation drives oxidation of other biomol-
ecules.
Ack n ow led gm en t. We thank Thomas D. McClure,
Ph.D., for his expert assistance in the charaterization of
4-nitro-â-carotene by LC/MS and J eanne Burr and Arti
Arora, Ph.D., for their efforts in the generation and
isolation of smoke oxidation products. This work was
supported in part by NIH Grants CA56875, ES07091,
and ES06694.
Refer en ces
(1) Shopland, D. R. (1995) Tobacco use and its contribution to early
cancer mortality with special emphasis on cigarette smoking.
Environ. Health Perspect. 103 (Suppl. 8), 131-141.
(2) Peto, R., Doll, R., Buckley, J . D., and Sporn, M. B. (1981) Can
dietary â-carotene materially reduce human cancer rates? Nature
290, 201-208.
Our experiments permitted a clear test of the hypoth-
esis that â-carotene functions as a prooxidant when
exposed to cigarette smoke under ambient air. Our data
indicate that â-carotene acts as a weak antioxidant in
smoke-exposed liposomes. Exposures to both gas-phase
(Figure 6) and whole smoke (Figure 7) significantly
increased HFAME levels in DLPC liposomes. â-Carotene
incorporation into the membranes at 0.1 mol % modestly
decreased HFAME levels over those of control liposomes
without â-carotene. â-Carotene levels of 0.1 mol % are
similar to those achieved by dietary supplementation in
animals (44). In all cases, liposomes with â-carotene
exhibited lower levels of HFAME than did liposomes
without â-carotene. Thus, â-carotene did not increase the
level of smoke-induced lipid peroxidation. The oxidation
of other antioxidants in the presence and absence of
â-carotene was also examined. Both R-tocopherol in
BLPC liposomes and ascorbate in the liposome buffer
were depleted by smoke exposures. The presence of
â-carotene at 0.01, 0.1, and 0.4 mol % either modestly
spared or had no effect on either R-tocopherol (Figure 9)
or ascorbate (Figure 10) levels upon exposure to gas-
phase smoke. These results indicate that â-carotene
autoxidation, which occurs under these conditions (Figure
8), does not accelerate the depletion of other antioxidants
by smoke.
(3) Mayne, S. T. (1996) â-Carotene, carotenoids, and disease preven-
tion in humans. FASEB J . 10, 690-701.
(4) Ziegler, R. G. (1991) Vegetables, fruits, and carotenoids and the
risk of cancer. Am. J . Clin. Nutr. 53, 251S-259S.
(5) The R-Tocopherol BCPSG (1994) The effect of vitamin E and
â-carotene on the incidence of lung cancer and other cancers in
male smokers. N. Engl. J . Med. 330, 1029-1035.
(6) Omenn, G. S., Goodman, G. E., Thornquist, M. D., Balmes, J .,
Cullen, M. R., Glass, A., Keogh, J . P., Meyskens, F. L., Valanis,
B., Williams, J . H., Barnhart, S., and Hammar, S. (1996) Effects
of a combination of â-carotene and vitamin A on lung cancer and
cardiovascular disease. N. Engl. J . Med. 334, 1150-1155 (see
comments).
(7) Hoffmann, D., Djordjevic, M. V., and Hoffmann, I. (1997) The
changing cigarette. Prev. Med. 26, 427-434.
(8) Church, D. F., and Pryor, W. A. (1985) Free-radical chemistry of
cigarette smoke and its toxicological implications. Environ. Health
Perspect. 64, 111-126.
(9) Pryor, W. A., Prier, D. G., and Church, D. F. (1983) An electron
spin resonance study of mainstream and sidestream cigarette
smoke: the nature of the free radicals in gas-phase smoke and
in cigarette tar. Environ. Health Perspect. 47, 345-355.
(10) Pryor, W. A., Hales, B. J ., Premovic, P. I., and Church, D. F. (1983)
The radicals in cigarette tar: their nature and suggested physi-
ological implications. Science 220, 425-427.
(11) Pryor, W. A., Tamura, M., and Church, D. F. (1984) ESR spin-
trapping study of the radicals produced in NOx/olefin reactions:
A mechanism for the production of the apparently long-lived
radicals in gas-phase cigarette smoke. J . Am. Chem. Soc. 106,
5073-5079.
(12) Handelman, G. J ., Packer, L., and Cross, C. E. (1996) Destruction
of tocopherols, carotenoids, and retinol in human plasma by
cigarette smoke. Am. J . Clin. Nutr. 63, 559-565.
(13) Kikugawa, K., Hiramoto, K., Tomiyama, S., and Asano, Y. (1997)
â-Carotene effectively scavenges toxic nitrogen oxides: nitrogen
dioxide and peroxynitrous acid. FEBS Lett. 404, 175-178.
(14) Everett, S. A., Dennis, M. F., Patel, K. B., Maddix, S., Kundu, S.
C., and Willson, R. L. (1996) Scavenging of nitrogen dioxide, thiyl,
and sulfonyl free radicals by the nutritional antioxidant â-caro-
tene. J . Biol. Chem. 271, 3988-3994.
(15) Everett, S. A., Kundu, S. C., Maddix, S., and Willson, R. L. (1995)
Mechanisms of free-radical scavenging by the nutritional anti-
oxidant â-carotene. Biochem. Soc. Trans. 23, 230S-230S.
(16) Kennedy, T. A., and Liebler, D. C. (1992) Peroxyl radical scaveng-
ing by â-carotene in lipid bilayers. Effect of oxygen partial
pressure. J . Biol. Chem. 267, 4658-4663.
While this paper was in review, Wang et al. reported
that ferrets supplemented with â-carotene exhibited
squamous metaplasia in the lung, which was enhanced
by smoke exposure (45). Levels of â-carotene oxidation
products were elevated in ferret lung postnuclear frac-
tions from animals supplemented with â-carotene and
exposed to smoke in corresponding in vitro experiments
(45). Thus, although it is unlikely that a direct prooxidant
effect of â-carotene occurs in the smoke-exposed lung,
â-carotene oxidation products formed by smoke may
themselves exert procarcinogenic effects. This intriguing
hypothesis merits further investigation.

Smoke Oxidation of â-Carotene
Chem. Res. Toxicol., Vol. 12, No. 6, 1999 543
(17) Kennedy, T. A., and Liebler, D. C. (1991) Peroxyl radical oxidation
of â-carotene: formation of â-carotene epoxides. Chem. Res.
Toxicol. 4, 290-295.
(33) McClure, T. D., and Liebler, D. C. (1995) A rapid method for
profiling the products of antioxidant reactions by negative ion
chemical ionization mass spectrometry. Chem. Res. Toxicol. 8,
128-135.
(34) Pryor, W. A., Castle, L., and Church, D. F. (1985) Nitrosation of
organic hydroperoxides by nitrogen dioxide/dinitrogentetroxide.
J . Am. Chem. Soc. 107, 211-217.
(35) Douki, T., Cadet, J ., and Ames, B. N. (1996) An adduct between
peroxynitrite and 2′-deoxyguanosine: 4,5-dihydro-5-hydroxy-4-
(nitrosooxy)-2′-deoxyguanosine. Chem. Res. Toxicol. 9, 3-7.
(36) Liebler, D. C., and McClure, T. D. (1996) Antioxidant reactions
of â-carotene: Identification of carotenoid-radical adducts. Chem.
Res. Toxicol. 9, 8-11.
(18) Mordi, R. C., Walton, J . C., Burton, G. W., Hughes, L., Ingold, K.
U., Lindsay, D. A., and Moffatt, D. J . (1993) Oxidative degradation
of â-carotene and â-apo-8′-carotenal. Tetrahedron 49, 911-928.
(19) Handelman, G. J ., van Kuijk, F. J ., Chatterjee, A., and Krinsky,
N. I. (1991) Characterization of products formed during the
autoxidation of â-carotene. Free Radical Biol. Med. 10, 427-437.
(20) Burton, G. W., and Ingold, K. U. (1984) â-Carotene: An unusual
type of lipid antioxidant. Science 224, 569-573.
(21) Mayne, S. T., Handelman, G. J ., and Beecher, G. R. (1996)
â-Carotene and lung cancer promotionin heavy smokerssa plaus-
able relationship? J . Natl. Cancer Inst. 88, 1513-1515.
(22) Cavanagh, J ., Fairbrother, W. J ., Palmer, A. G. I., and Skelton,
N. J . (1996) Protein NMR Spectroscopy, Principles and Practice,
Academic Press, San Diego, CA.
(23) Marion, D., and Wuthrich, K. (1983) Application of phase-sensitive
two-dimensional correlated spectroscopy (COSY) for measure-
ments of 1H-1H spin-spin coupling constants in proteins. Bio-
chem. Biophys. Res. Commun. 113, 967-974.
(24) Muller, L. (1979) Sensitivity enhanced detection of weak nuclei
using heteronuclear multiple quantum coherence. J . Am. Chem.
Soc. 101, 4481-4484.
(25) Bax, A., Griffey, R. H., and Hawkins, B. L. (1983) Correlation of
proton and nitrogen-15 chemical shifts by multiple quantum
NMR. J . Magn. Reson. 55, 301-315.
(26) Piantini, U., Sørenson, O. W., and Ernst, R. R. (1982) Multiple
quantum filters for elucidating NMR coupling networks. J . Am.
Chem. Soc. 104, 6800-6801.
(27) Foote, C. S., Chang, Y. C., and Denny, R. W. (1970) Chemistry of
singlet oxygen. XI. Cis-trans isomerization of carotenoids by
singlet oxygen and a probable quenching mechanism. J . Am.
Chem. Soc. 92, 5218-5219.
(28) Rance, M., Sorensen, O. W., Bodenhausen, G., Wagner, G., Ernst,
R. R., and Wuthrich, K. (1983) Improved spectral resolution in
COSY 1H NMR spectra of proteins via double quantum filtering.
Biochem. Biophys. Res. Commun. 117, 479-485.
(37) Grant, J . L., Kramer, V. J ., Ding, R., and Kispert, L. D. (1988)
Carotenoid cation radicals: electrochemical, optical, and EPR
study. J . Am. Chem. Soc. 110, 2151-2157.
(38) el-Tinay, A. H., and Chichester, C. O. (1970) Oxidation of
â-carotene. Site of initial attack. J . Org. Chem. 35, 2290-2293.
(39) Woodall, A. A., Lee, S. W.-M., Weesie, R. J ., J ackson, M. J ., and
Britton, G. (1997) Oxidation of carotenoids by free radicals:
relationship between structure and reactivity. Biochim. Biophys.
Acta 1336, 33-42.
(40) Samokyszyn, V. M., and Marnett, L. J . (1987) Hydroperoxide-
dependent cooxidation of 13-cis-retinoic acid by prostaglandin H
synthase. J . Biol. Chem. 262, 14119-14129.
(41) Cooney, R. V., Harwood, P. J ., Franke, A. A., Narala, K.,
Sundstro¨m, A. K., Berggren, P. O., and Mordan, L. J . (1995)
Products of γ-tocopherol reactions with NO2 and their formation
in rat insulinoma (RINm5F) cells. Free Radical Biol. Med. 19,
259-269.
(42) Hoglen, N. C., Waller, S. C., Sipes, I. G., and Liebler, D. C. (1997)
Reactions of peroxynitrite with γ-tocopherol. Chem. Res. Toxicol.
10, 401-407.
(43) Christen, S., Woodall, A. A., Shigenaga, M. K., Southwell-Keely,
P. T., Duncan, M. W., and Ames, B. N. (1997) γ-Tocopherol traps
mutagenic electrophiles such as NOx and compliments R-toco-
pherol: physiological implications. Proc. Natl. Acad. Sci. U.S.A.
94, 3217-3222.
(44) Liebler, D. C., Stratton, S. P., and Kaysen, K. L. (1997) Antioxi-
dant actions of â-carotene in liposomal and microsomal mem-
branes. Role of carotenoid-membrane incorporation and interac-
tion with R-tocopherol. Arch. Biochem. Biophys. 338, 244-250.
(45) Wang, X., Liu, C., Bronson, R. T., Smith, D. E., Krinsky, N. I.,
and Russell, R. M. (1999) Retinoid signaling and activator
protein-1 expression in ferrets given â-carotene supplements and
exposed to tobacco smoke. J . Natl. Cancer Inst. 91, 60-66.
(29) Kremer, J . M. H., van de Esker, M. W. J ., Pathmamanoharan,
C., and Wiersema, P. H. (1977) Vesicles of variable diameter
prepared by a modified injection method. Biochemistry 16, 3932-
3935.
(30) Batzri, S., and Korn, E. D. (1969) Single bilayer liposomes
prepared without sonication. Biochim. Biophys. Acta 298, 1015.
(31) Liebler, D. C., Burr, J . A., Philips, L., and Ham, A. L. (1996) Gas
chromatography-mass spectrometry analysis of vitamin E and
its oxidation products. Anal. Biochem. 236, 27-34.
(32) Butts, W. C., and Mulvihill, H. J . (1975) Centrifugal analyzer
determination of ascorbate in serum or urin with Fe3+/ferrozine.
Clin. Chem. 21, 1493-1497.
TX980263V