Metabolism of (R)- and (S)-3-(phenylamino)propane-1,2-diol in C57BL/6- and A/J-strain mice. Identification of new metabolites with potential toxicological significance to the Toxic Oil Syndrome

Article abstract of DOI:10.1021/tx010001k

The Toxic Oil Syndrome was a massive food-borne intoxication that occurred in Spain in 1981. Epidemiological studies point to 3-(phenylamino)propane-1,2-diol (PAP) derivatives as the putative toxic agents. We report further identification of metabolites c

Full text of DOI:10.1021/tx010001k

Chem. Res. Toxicol. 2001, 14, 1097-1106
1097
Meta bolism of (R)- a n d
(S)-3-(P h en yla m in o)p r op a n e-1,2-d iol in C57BL/6- a n d
A/J -Str a in Mice. Id en tifica tion of New Meta bolites w ith
P oten tia l Toxicologica l Sign ifica n ce to th e Toxic Oil
Syn d r om e
J ordi Bujons,† Margarita G. Ladona,*,‡ Angel Messeguer,† Anna Morato´,† and
Coral Ampurdane´s‡
Department of Biological Organic Chemistry, Instituto de Investigaciones Qu´ımicas y Ambientales
de Barcelona, and Department of Medical Bioanalysis, Instituto de Investigaciones Biome´dicas de
Barcelona, Consejo Superior de Investigaciones Cient´ıficas, Barcelona, Spain
Received J anuary 5, 2001
The Toxic Oil Syndrome was a massive food-borne intoxication that occurred in Spain in
1981. Epidemiological studies point to 3-(phenylamino)propane-1,2-diol (PAP) derivatives as
the putative toxic agents. We report further identification of metabolites cleared in urine of
A/J and C57BL/6 mice in which (R)- and (S)-3-(phenylamino)propane-1,2-diol were administered
intraperitoneally. This investigation is an extension of previous studies carried out with the
racemic compound [Ladona, M. G., Bujons, J ., Messeguer, A., Ampurdane´s, C., Morato´, A.,
and Corbella, J . (1999) Chem. Res. Toxicol. 12, 1127-1137]. Both PAP enantiomers were
extensively metabolized, and several metabolites were eliminated in urine. The HPLC profiles
of the urine samples of both mouse strains treated with each enantiomer were qualitatively
similar, but differences were found in a relatively higher proportion of several detected
metabolites in mice treated with (R)-PAP compared with those treated with (S)-PAP. The main
urine metabolite continues to be 2-hydroxy-3-(phenylamino)propanoic acid (1), which confirms
our previous results obtained with rac-PAP. In addition to the detection of other metabolites
already reported in our previous paper, interesting evidence is provided on the presence of
4-aminophenol and paracetamol conjugates in the urine samples from both mouse strains.
The detection of these metabolites suggests the in vivo formation of quinoneimine PAP
derivatives. Indeed, some quinoneimine species (11 and 12), as well as other PAP metabolites
(13) that bear modifications in the alkyl chain, have been tentatively identified in mouse urine.
These metabolic findings might imply a potential toxicological significance for the Toxic Oil
Syndrome.
amino)propane-1,2-diol (PAP) (aromatic amines) (1, 7, 8),
with the latter, the PAP esters, being strongly associated
with TOS (8, 9). In particular, epidemiological studies
has revealed that the levels of the 1,2-dioleyl ester of
3-(phenylamino)propane-1,2-diol (OOPAP) correlate with
patient morbidity in the corresponding households; there-
fore, this compound is currently considered to be the
putative toxic compound responsible for the intoxication
(9, 10). OOPAP was generated by chemical processes
during oil refining at the ITH Co., and the epidemic was
considered to have emerged from a single source in that
refinery (11).
In tr od u ction
The Toxic Oil Syndrome (TOS)1 appeared in Spain in
1981 as a massive food-borne intoxication that affected
more than 20 000 people (1). TOS was attributed to the
ingestion of rapeseed oil, which had been adulterated
with aniline, illegally refined, and delivered for human
consumption (1-6). Two types of chemical species derived
from aniline were identified in oil batches, i.e., fatty acid
anilides (amides) and fatty acid esters of 3-(phenyl-
* To whom correspondence should be addressed at Cartagena 401,
AT-2, 08024 Barcelona, Spain. E-mail: mglqob@cid.csic.es.
† Department of Biological Organic Chemistry, Instituto de Inves-
tigaciones Qu´ımicas y Ambientales de Barcelona.
Extensive efforts to investigate the potential toxicity
of aniline derivatives in animal species were carried out
with unsuccessful results to reproduce the full spectrum
of symptoms observed in humans (12). However, few
studies reported partial toxicity in mice; e.g., PAP and
its mono-oleyl ester (OPAP) were toxic to some mouse
strains when administered intraperitoneally (13, 14), but
lacked toxicity when administered orally. In this respect,
it has also been shown that, when administered orally
to rats, OOPAP is hydrolyzed in the intestine to the
corresponding mono-oleyl ester and PAP (15). Although
‡ Department of Medical Bioanalysis, Instituto de Investigaciones
Biome´dicas de Barcelona.
1
Abbreviations: TOS, Toxic Oil Syndrome; EMS, Eosinophilia-
Myalgia Syndrome; OLA, oleanilide; PAP, 3-(phenylamino)propane-
1,2-diol; OPAP, monooleyl ester of PAP.; OOPAP dioleyl ester of PAP.;
MPAP, DPAP, mono and diacyl derivatives of PAP.; PAA, 3-(phenyl-
amino)-L-alanine; â-NF, â-naphthoflavone; TEAA, triethylammonium
acetate; TEAC, triethylammonium hydrogen carbonate; PBS, phos-
phate buffered saline (137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4,
1.8 mM KH2PO4, pH 7.4); BSTFA, O,N-bis(trimethylsilyl)trifluoro-
acetamide; EI-MS, electron impact mass spectrometry; HRP, horse-
radish peroxidase.
10.1021/tx010001k CCC: $20.00 © 2001 American Chemical Society
Published on Web 07/31/2001

1098 Chem. Res. Toxicol., Vol. 14, No. 8, 2001
Bujons et al.
Ch a r t 1. P AP a n d Rela ted Str u ctu r es In vestiga ted
a s P u ta tive Meta bolites of P AP in C57BL/6- a n d
A/J -Str a in Mice
that absorption through the intestinal cells into the
human body of enantiomerically enriched PAP, MPAP,
and DPAP can be expected. Consequently, enantiomeric
distribution and metabolism differences should be taken
into account in the toxicokinetics of these compounds and
their association with TOS symptoms.
With these findings and the evidence of intestinal
formation of PAP enantiomers, herein we extend our
previous work on in vivo disposition in mice of rac-PAP
by reporting the results of analogous studies carried out
with the pure (R)- and (S)-PAP enantiomers. This study
led us to the identification of new PAP metabolites in
mouse urine samples which may have toxicological
significance for TOS.
Ma ter ia ls a n d Meth od s
Ch em ica ls a n d Rea gen ts. Aniline (Aldrich, 99%) and
glycidol (Merck, 98%) were previously distilled and stored at
-20 °C. (R)- and (S)-glycidol, 4-aminophenol, N-(4-hydroxy-
phenyl)glycine, 4-acetamidophenol (paracetamol), and O,N-
bis(trimethylsilyl)trifluoroacetamide (BSTFA) were obtained
from Aldrich (Germany). [U-14C]Aniline hydrochloride (129 mCi/
mmol) was from Amersham (U.K.). Olive oil, â-glucuronidase
from Helix pomatia and Escherichia coli, and â-naphthoflavone
(â-NF) were obtained from Sigma (St. Louis, MO). Methanol,
acetonitrile, isopropyl alcohol, dichloromethane, and other com-
mon solvents were of HPLC quality and were purchased from
Merck (Darmstadt, Germany). Triethylammonium acetate
(TEAA, 1 M, pH 7.0) and triethylammonium hydrogen carbonate
(TEAC, 1 M, pH 8.4) buffers were obtained from Fluka (Switz-
erland). General laboratory reagents were obtained from local
sources and were of analytical grade. OptiPhase ‘HiSafe’ cocktail
was purchased from Fisons Chemicals (England).
TOS disease was considered species-specific for humans
(12), from animal studies it can be postulated that aniline
derivative toxicokinetics may differ between animals and
humans. Thus, to ascertain the potential toxicity of these
compounds and seek an animal model, investigation
efforts should focus on determining factors such as
metabolism, distribution, and elimination in comparison
between animals and humans.
Ca u tion : As PAP derivatives have been implicated in TOS,
special precautions should be taken when handling these sub-
stances to avoid potential risks (use gloves, masks, and ventilated
hood cabinets when handling solutions and powder).
Therefore, our research line focused on investigating
the metabolic profile of TOS patients in parallel with
determining PAP derivative metabolism and distribution
in mouse strains, and, further, applying the acquired
knowledge to obtain an animal model for in-depth
analysis of TOS pathogenic mechanisms. In this respect,
we recently reported that TOS survivors exhibit a specific
metabolic profile determined by molecular analyses on
genetic polymorphisms of enzyme metabolism (16). In
addition, we have studied the biotransformation and
clearance of racemic PAP (rac-PAP) in the C57BL/6 and
A/J mouse strains (17), a well-known murine model for
human acetylation polymorphism. Metabolic species of
PAP (Chart 1) administered intraperitoneally, in par-
ticular 2-hydroxy-3-(phenylamino)propanoic acid 1, were
identified for the first time as major metabolites in the
urine of these mouse strains. Nevertheless, the structure
of additional minor PAP metabolites present in those
urine samples remained to be elucidated. Further on this
line of research, we investigated the in vitro stereoselec-
tivity of the lipase-catalyzed hydrolysis of the mono and
dioleyl PAP esters as a model of the transformations that
PAP esters may undergo in the human intestinal tract.2
The results of these studies indicate that these esters can
be regio- and stereoselectively hydrolyzed and, therefore,
Syn th esis of P AP Der iva tives. Unless stated otherwise,
organic solutions obtained from the treatment of crude reaction
mixtures were dried over MgSO4. Purification of metabolites by
flash chromatography was performed using 35-70 µm silica gel
(SDS, France). Reactions were monitored either by GC, HPLC,
or TLC. GC analyses were performed on a Hewlett-Packard
5890 Series II gas chromatograph, provided with a FID detector
and a 15 m HP-5 capillary column. HPLC analyses were
performed on a Hewlett-Packard 1100 system provided with a
DAD detector and a HP ODS Hypersil column (5 µm, 125 × 4
mm), using mixtures of HPLC grade acetonitrile and 10 mM
TEAA (pH 6.8) buffer as eluents. Thin-layer chromatography
(TLC) analyses were performed on Merck Kielsegel 60 F254
plates (aluminum sheets, 0.2 mm thickness) using mixtures of
hexanes-EtOAc as eluents, and were developed by UV irradia-
tion at 254 nm. Radioactivity was determined with an LKB 1217
Rackbeta scintillation counter following the addition of 10 mL
of OptiPhase ‘HiSafe’ cocktail. The 1H (300 MHz) and 13C (75
MHz) NMR spectra were registered with a Varian Unity 300
spectrometer. Spectra were taken in neutralized CDCl3 unless
otherwise stated. Chemical shifts are given in ppm relative to
tetramethylsilane for 1H and deuteriochloroform for 13C as
internal standards. The GC/MS analyses and EI-MS spectra (70
eV) were obtained using a Fisons model MD 800 mass spec-
trometer coupled to a Fisons GC 8000 apparatus equipped with
a 25 m HP-5 capillary column. Elemental microanalyses were
carried out at the “Servei de Microana`lisi” of the I.I.Q.A.B. using
a 1108 Carlo Erba analyzer.
2
Ladona, M. G., Morato´, A., Messeguer, A., Ampurdane´s, C., and
[U-14C]-rac-PAP (5.7 mCi/mmol) was prepared from [U-14C]-
aniline hydrochloride as previously described (17). (R)-PAP and
(S)-PAP were synthesized by reaction of aniline with (S)- and
(R)-glycidol, respectively, according to our previous experience.2
Bujons, J . Stereoselective hydrolysis of 1,2-dioleyl-3-phenylamino-
propane-1,2-diol by human pancreatic lipase and bovine lipoprotein
lipase: toxicokinetic implications in Toxic Oil Syndrome. Manuscript
in preparation.

Metabolism of (R)- and (S)-PAP in Mice
Chem. Res. Toxicol., Vol. 14, No. 8, 2001 1099
138.9 (C-1′), 116.5 (C-3′, C-5′), 116.3 (C-2′, C-6′), 109.7 (H3CCCH3),
73.7 (CHO), 67.1 (CH2O), 49.2 (CH2N), 26.9 (CH3), 25.3 (CH3);
EI-MS m/z 223(M+), 208, 122 (base peak); EI-MS (TMS deriva-
tive) m/z 295 (M+), 194 (base peak).
Sch em e 1. Syn th etic P a th w a y for th e Obten tion of
Qu in on eim in e 10
Oxid a tion of Dioxola n e 7. The oxidation of dioxolane 7 was
carried out following the procedure described by Dahlin and
Nelson for the case of N-acetyl-p-benzoquinoneimine (18). Thus,
7 (10 mg, 45 µmol) and Ag2O (7 mg, 30 µmol) were stirred in
dry chloroform (5 mL) at 20 °C until no further evolution of the
substrate was observed (4 h, HPLC monitoring). The crude
reaction mixture was filtered and the filtrate evaporated to
dryness to give a residue which revealed the presence of imine
9 as the major compound, 9: 1H NMR δ 7.86 (d, 1 H, J ) 5.1,
HCdN), 7.05 (d, 2 H, J ) 9.0, H-3′, H-5′), 6.81 (d, 2 H, J ) 9.0,
H-2′, H-6′), 4.76 (m, 1 H, H-CO), 4.29 (dd, 1 H, J 1 ) 8.6, J 2
)
6.6, CHH-O), 4.05 (dd, 1 H, J 1 ) 8.6, J 2 ) 6.3, CHH-O), 1.50
(s, 3 H, CH3), 1.44 (s, 3 H, CH3); 13C NMR δ 161.2 (C-3), 115.8
(C-3′, C-5′), 122.2 (C-2′, C-6′), 142.9 (C-1′), 155.1 (C-4′), 110.5
(C(CH3)2), 77.3 (C-2), 67.4 (C-1), 26.5 (CH3), 25.4 (CH3); EI-MS
m/z 221 (M+), 208, 120 (base peak); EI-MS (TMS derivative) m/z
293 (M+), 235, 192 (base peak).
To follow the evolution of the reaction by 1H NMR, the
reaction was repeated by placing in an NMR tube 18 mg (81
µmol) of dioxolane 7 dissolved in 0.7 mL of anhydrous CDCl3
and adding 22 mg of Ag2O (81 µmol). The initial formation of a
product was observed which was transformed into imine 9 in
the course of several hours. The spectroscopic data of this
intermediate were compatible with quinoneimine 8: 1H NMR
δ 7.32 (dd, 1 H, J 1 ) 10.3, J 2 ) 2.6),3 7.14 (dd, 1 H, J 1 ) 10.3,
J 2 ) 2.6),3 6.58 (m, 1 H, J 1 ) 10.3, J 2 ) 1.2),3 6.57 (m, 1 H, J 1
) 10.3, J 2 ) 1.2),3 4.56 (m, 1 H, H-CO), 4.20 (dd, 1 H, J 1 ) 8.4,
J 2 ) 6.5, CHH-O), 4.00 (dd, 1 H, J 1 ) 15.0, J 2 ) 5.9, CHH-N),
3.91 (dd, 1 H, J 1 ) 15.0, J 2 ) 6.2, CHH-N), 3.87 (dd, 1 H, J 1
)
8.4, J 2 ) 6.0, CHH-O), 1.43 (s, 3 H, CH3), 1.40 (s, 3 H, CH3);
13C NMR δ 142.8 (CArH), 133.0 (CArH), 131.7 (CArH), 125.2
(CArH), 109.7 (C(CH3)2), 75.5 (C-2), 67.6 (C-1), 56.2 (C-3), 26.8
(CH3), 25.3 (CH3). An aliquot of the crude mixture containing
this intermediate was filtered, treated with BSTFA, and ana-
lyzed by GC/MS. The main peak detected was that of the
O-silylated derivative of imine 9, described above. In addition,
a peak consistent with the structure of quinoneimine 8 was
detected [EI-MS m/z 221 (M+), 163, 120 (base peak)], but only
as a minor component probably due to the decomposition of the
product.
An im a ls. A/J and C57BL/6 mice (9-10 weeks old, three
animals of each strain) used in the experiments were obtained
from Charles River (France) and Criffa (Spain), respectively.
Animals were kept in metabolic cages (Techniplast, Italy) for 3
days of acclimatization and induced with â-NF prior to admin-
istration of PAP. Protocols for induction treatment with â-NF
and PAP administration to the animals were identical to those
previously described (17). Identification of phenolic species and
PAP acid metabolites in our previous study prompted us to
induce mice with â-NF, since we expected that the enzymes
induced by the Ah-receptor enzyme battery would contribute
to the formation of such metabolic species. In this respect, we
observed that in C57BL/6 mice the urine clearance of metabo-
lites increased to levels similar to those observed in A/J mice
when pretreated with â-NF. Thus, in the present investigation
we used the same conditions, i.e., to treat mice with â-NF 80
mg/kg in olive oil distributed in three intraperitoneal injections.
A single total dose of (R)- or (S)-PAP (250 mg/kg, average animal
weight in each group: 22-24 g) spiked with 4 µCi of [U-14C]-
rac-PAP was given intraperitoneally. Urine and feces were
collected over a 24 h period and the animals euthanized. Organs
(lung, liver, kidney, heart, muscle, and spleen) were collected
in cold PBS (pH 7.4) and stored at -80 °C until analysis.
14C Ra d ioa ctivity Mea su r em en ts in Or ga n ic Tissu e a n d
Bod y F lu id s. Urine samples (20 µL) were mixed with 10 mL
4-(4′-Ben zyloxyp h en yl)a m in om et h yl-2,2-d im et h yl-1,3-
d ioxa cyclop en ta n e (6). This dioxolane was prepared from
benzyloxy derivative 5 (Scheme 1) (17). Briefly, a mixture of 5
(0.24 g, 0.9 mmol), 2,2-dimethoxypropane (5 mL), and a catalytic
amount of Amberlyst 15 was stirred for 2 h at 80 °C (GC
monitoring). The crude reaction mixture was cooled, filtered,
and evaporated. The residue was redissolved in tert-butyl methyl
ether, washed with NaHCO3, and dried. Final purification by
column chromatography on silica gel (4:1 hexane/EtOAc) af-
forded 0.20 g (73% yield) of compound 6 as a yellow oil. 1H NMR
δ 7.35 (5 H′′), 6.84 (d, 2 H, J ) 9.0, H-3′, H-5′), 6.53 (d, 2 H, J
) 9.0, H-2′, H-6′), 4.97 (2 H, CH2Ph), 4.32 (m, 1 H, H-CO), 4.06
(dd, 1 H, J 1 ) 8.2, J 2 ) 6.4, CHH-O), 3.74 (dd, 1 H, J 1 ) 8.2,
J 2 ) 6.4, CHH-O), 3.23 (dd, 1 H, J 1 ) 12.4, J 2 ) 4.4, CHH-N),
3.12 (dd, 1 H, J 1 ) 12.4, J 2 ) 6.6, CHH-N), 1.44 (s, 3 H, CH3),
1.36 (s, 3 H, CH3); 13C NMR δ 151.5 (C-4′), 142.4 (C-1′),137.5
(C-1′′), 128.4 (C-3′′, C-5′′), 127.7 (C-4′′), 127.4 (C-2′′, C-6′′), 116.1
(C-3′, C-5′), 114.2 (C-2′, C-6′), 109.3 (H3CCCH3), 74.5 (CHO),
70.7 (CH2Ph), 67.2 (CH2O), 47.5 (CH2N), 26.8 (CH3), 25.3 (CH3);
EI-MS m/z 313(M+), 91 (base peak). Elemental analysis for
C19H23NO3: C, 72.82; H, 7.40; N, 4.47. Found: C, 72.83; H, 7.40;
N, 4.37.
4-(4′-Hyd r oxyp h en yl)a m in om eth yl-2,2-d im eth yl-1,3-d i-
oxa cyclop en ta n e (7). A mixture of dioxolane 6 (0.10 g, 0.32
mmol) and 10% Pd/C (0.030 g, 0.032 mmol) in EtOH (7 mL)
was stirred under a hydrogen atmosphere at 20 °C. When the
reaction was completed (2 h, TLC monitoring), the crude
reaction mixture was filtered and evaporated to give the desired
product 7 (0.06 g, 85% yield). This compound was used without
further purification. 1H NMR δ 6.72 (d, 2 H, J ) 9.0, H-3′, H-5′),
6.67 (d, 2 H, J ) 9.0, H-2′, H-6′), 4.65 (2 H, NH, OH), 4.36 (m,
1 H, H-CO), 4.07 (dd, 1 H, J 1 ) 8.4, J 2 ) 6.3, CHH-O), 3.74
(dd, 1 H, J 1 ) 8.4, J 2 ) 6.0, CHH-O), 3.25 (dd, 1 H, J 1 ) 12.5,
J 2 ) 4.2, CHH-N), 3.14 (dd, 1 H, J 1 ) 12.5, J 2 ) 7.2, CHH-N),
1.44 (s, 3 H, CH3), 1.35 (s, 3 H, CH3); 13C NMR δ 150.7 (C-4′),
3
The chemical shifts and coupling constants for these absorptions
were confirmed through simulation using a LAOCOON based program
included in the software of the Varian Unity 300 instrument.

1100 Chem. Res. Toxicol., Vol. 14, No. 8, 2001
Bujons et al.
of OptiPhase ‘HiSafe’ scintillation cocktail and counted. The
radioactivity remaining in feces and tissues was counted in a
similar manner after total organ homogenization in PBS buffer
(pH 7.4). An aliquot of tissue homogenate (0.5-1 mL) was
counted, and total 14C disintegrations per minute in organs were
calculated. For equipment decontamination purposes, remaining
radioactivity, if any, in the cage cone-collectors, which would
indicate a radioactivity loss during the collecting period (urine
and feces), was measured.
An a lytica l Deter m in a tion s in Biologica l F lu id s. (1)
High -P er for m an ce Liqu id Ch r om atogr aph y (HP LC). HPLC
analyses were conducted on a Millenium Waters system equipped
with a 717 Autosampler and a 996 Photodiode Array Detector,
coupled on-line to a Berthold LB 507 radioactivity monitor
system. UV spectra were collected with a resolution of 3.6 nm.
C-18 Kromasil 100, 5 µm particle size, 250 × 4.6 mm columns
(Teknokroma, Spain) were used with 10 mM TEAA, pH 6.8
(solvent A), and acetonitrile (solvent B) as mobile phases.
Gradient elutions consisted of an initial isocratic step (0-5 min)
at 0.9 mL/min 95% A, followed by a linear ramp to reach 1.1
mL/min 60% A (5-20 min), and a final isocratic elution at 1.1
mL/min 60% A (20-30 min), similar to as previously described
(17).
Urine samples were used as reported (17). Briefly, 50 µL of
the urine samples was mixed with methanol (1:9 v/v) and
centrifuged. Supernatants were injected directly into the HPLC
system. For the separation of polar metabolites from PAP, urine
samples were buffered with 100 mM triethylammonium hydro-
gen carbonate (TEAC, pH 8.4) and extracted with a mixture of
dichloromethane-isopropyl alcohol (9:1) containing 1% am-
monia. The organic extracts were evaporated and redissolved
in TEAA (10 mM, pH 6.8)-acetonitrile (80:20) prior to the HPLC
analysis, while the aqueous phases were analyzed directly.
Chemical and enzymatic procedures were used for conjugate
cleavage and metabolite derivatization. The chemical procedure
consisted of treatment of the direct-urine sample or aqueous-
extract samples with 1 mL of 1 M HCl in methanol (HCl/MeOH)
for 15 min at 90 °C followed by evaporation of the solvents with
a stream of nitrogen. Direct-urine hydrolysates were further
extracted at pH 8.4 (TEAC) as described above. Samples treated
in this manner were subjected to either HPLC or GC/MS
analysis. Enzymatic hydrolysis consisted of sample treatment
with 2000 enzyme units/mL of sample-incubate of â-glu-
curonidase from E. coli or H. pomatia (which also exhibits
sulfatase activity) in the appropriate buffers for 16 h at 37 °C.
Hydrolysates were further extracted at pH 8.4 (TEAC) and
processed as described above.
different intermediates. Accordingly, the synthetic path-
way shown in Scheme 1 was attempted. Protection of
aminodiol 5 as acetonide followed by hydrogenolysis
afforded phenol 7. Treatment of 7 with silver oxide in
chloroform (18) and close control of the reaction by HPLC
and NMR revealed the initial formation of a major
product which presented only a UV maximum at 271 nm,
consistent with the oxidation of the aromatic ring. The
NMR data of this compound are consistent with those
expected for quinoneimine 8, in particular, the four
multiplets centered at 7.32, 7.14, 6.58, and 6.57 ppm in
the 1H NMR spectrum3 which were assigned, respec-
tively, to the four nonequivalent protons of the quinone-
imine moiety. This intermediate in deuteriochloroform
was slowly transformed into a new product which was
identified as imine 9, the compound resulting from the
prototropic rearrangement of quinoneimine 8. Especially,
the presence of a doublet at 7.86 ppm assigned to the
imine HCdN proton, an AB system (7.05 and 6.81 ppm)
for the aromatic protons, and a multiplet at 4.76 ppm
attributed to the HC-O- proton were in agreement with
the imine structure. Moreover, the GC/MS analysis of
silylated aliquots of this crude reaction mixture permitted
the detection of a main product with the molecular ion
at m/z 293 and the base peak at m/z 192. These peaks
appeared 2 units below the corresponding peaks observed
for the O-silylated derivative of phenol 7, and are
consistent with the imine 9 structure. Unfortunately, the
occurrence of quinoneimine 8 could not be definitively
confirmed by GC/MS owing to the decomposition of the
product during the derivatization with BSTFA and/or the
chromatographic process. In this respect, we systemati-
cally detected the presence of the phenol precursor 7 on
GC/MS analysis of samples containing quinoneimine 8
as the major component (1H NMR monitoring). It is worth
noting that for the quinoneimine derived from paraceta-
mol (18, 26), as well as other quinones (27), the detection
of products derived from their reduction due to traces of
residual water in the ion source of the mass spectrometer
has been reported.
Further analysis of the evolution of the above sample
by NMR showed a progressive decrease in the absorp-
tions attributed to imine 9, which was concomitant to the
appearance of a new compound identified as 4-amino-
phenol, based on its chromatographic and spectroscopic
data (not shown). In addition, a doublet centered at 9.73
ppm (J ) 1.8 Hz) indicative of the formation of an
aldehyde species was also observed. All these results can
be explained by the transformations postulated in Scheme
2. Thus, oxidation of aminophenol 7 would lead to the
generation of quinoneimine 8. This unstable intermediate
would rearrange to give imine 9, a compound that was
identified in the crude reaction mixture. Finally, hydroly-
sis of the imine would give rise to 4-aminophenol and
the corresponding aldehyde species. A similar process had
been proposed for the evolution of the quinoneimine
derived from N-(4-hydroxyphenyl)glycine (28), and a
more detailed study of this class of rearrangement for
N-alkyl monoquinoneimines has also been reported (29).
As anticipated, when the acetal deprotection of an
aliquot from the above crude reaction mixture, taken
(2) Ga s Ch r om a togr a p h y/Ma ss Sp ectr om etr y (GC/MS).
Samples for GC/MS analysis were lyophilized, redissolved in
acetonitrile, and treated with BSTFA, as described by Adachi
et al. (19), prior to injection into the GC/MS system described
above. Samples treated with HCl/MeOH were also redissolved
in acetonitrile after evaporation of solvents and injected either
directly or treated with BSTFA before injection into the GC/
MS system.
Resu lts
Syn th esis of P AP Der iva tives. Synthesis of quinone-
imine 10 was initially attempted by oxidation of the
corresponding aminophenol precursor 2 using chemical
[oxidation with Fremy’s salt (20, 21)] and enzymatic
methods [HRP-catalyzed oxidation (22)]. However, efforts
carried out for isolating the desired product were unsuc-
cessful. An explanation of this failure could be the
potential instability and high polarity expected for 10
which would account for the erratic results observed in
the HPLC monitoring of the reaction course.4 These
antecedents advised us to design a synthesis where both
hydroxyl groups were protected until the last step, thus
facilitating the manipulation and monitoring of the
4
It is noteworthy that the literature describes mainly examples of
N-acylated quinoneimines, in particular the acetyl derivative (18), or
few cases of quinoneimines where this moiety is forming part of a more
complex condensed ring structure (23, 24). The isolation of such
electrophilic species is difficult because of their instability (25).

Metabolism of (R)- and (S)-PAP in Mice
Chem. Res. Toxicol., Vol. 14, No. 8, 2001 1101
Sch em e 2. P r op osed P a th w a y for th e F or m a tion
of 4-Am in op h en ol fr om P h en olic P AP Der iva tives
cone-collector walls and steel grid owing to urine dried
on the way down to the collecting tubes. This partial
radioactivity loss was reported in our previous study (17).
This last observation did not prevent us from proceeding
with analysis seeking the qualitative metabolic profile
in the urine samples since remaining radioactivity was
less than 2% of the administered dose in organ-tissues
and feces. Thus, the urine samples accumulated the most
substantial radioactivity clearance.
An a lytica l Deter m in a tion a n d Id en tifica tion of
Meta bolites in Ur in e Sa m p les. Figure 1 shows the
HPLC profiles of methanol-diluted urine aliquots ob-
tained from â-NF-induced, (S)- and (R)-PAP-treated mice.
Similarly to what was previously found in the experi-
ments with rac-PAP (17), the 14C-radioactivity traces
(Figure 1, panels A and B) of both mouse strains treated
with either enantiomer of PAP show a major peak at 16.2
min which, on average, accounts for approximately 62 (
2% of the total integrated area. The second major peak
in the radioactivity tracings of the C57BL/6 mice urine
is found close to 4 min. It accounts for about 20% of the
total integral, and it should contain the most polar PAP
metabolites. Similar results were obtained for the A/J
mice treated with (S)-PAP while those treated with (R)-
PAP showed the same peak with a lower intensity. Other
radioactive peaks were more clearly observed in this last
tracing in the 5-10 and 16-20 min regions, which were
less intense or absent in the other chromatograms.
Ta ble 1. 14C-P AP Clea r a n ce in Ur in e a n d F eces in a 24 h
P er iod
(S)-PAP-treated micea (R)-PAP-treated micea
A/J
C57BL/6
A/J
C57BL/6
14C-PAP doseb 3.1 × 107
3.1 × 107 2.5 × 107
2.4 × 107
urine (mL)
% of dose
feces (g)
2.00
59
2.00
52
1.23
54
1.80
63
3.58
0.80
3.60
0.45
3.10
0.54
4.06
1.80
% of dose
a
Three mice of each strain (total ) 6) were placed together in
a metabolic cage, constituting two groups treated either with (R)-
or with (S)-PAP (250 mg/kg average animal weight).
[
14C]-rac-
b
PAP total dose (dpm) administered together with the PAP enan-
tiomer to each animal group. Urine and feces samples were
collected over a 24 h period as detailed under Materials and
Methods.
The differential metabolism of (S)- and (R)-PAP be-
comes more evident in the UV-monitored chromatograms
(Figure 1, panels C and D). The peak at 16.2 min (peak
g) is again the major single peak on all traces. A minor
peak with a retention time close to 20 min (peak k ) is
also present in the four chromatograms. These two peaks
show the same UV spectrum (Figure 2) and can be
assigned to 2-hydroxy-3-(phenylamino)propanoic acid (1)
and PAP, respectively, based on the identity of their
retention times and spectra with those of synthetic
standards (17). The second major peak with a retention
time close to 4 min (peak a ) shows a nonhomogeneous
UV spectrum (not shown) which suggests that it corre-
sponds to the sum of several unresolved polar compo-
nents in the urine samples, some of them being highly
polar PAP metabolites. The spectra of peaks b-e and
h -i are also shown in Figure 2, while those for peaks f
and j are not included since they exhibit characteristics
of overlapping signals, similar to peak a . Taking the
spectrum of PAP (peak k ) as reference, the spectra of
peaks b and d show a bathochromic shift in the position
of one of the maxima from 289 to 300 nm and a slight
ipsochromic shift of the PAP maximum at 243 nm.
Comparison of these spectra with those from synthetic
standards 2, 3b, and 4a ,b (Chart 1) (17) suggests that b
and d could correspond to phenolic PAP metabolites. On
the other hand, the UV spectrum of peak c shows only a
maximum at 243 nm and a shoulder close to 280 nm,
which are similar to what is observed for standards that
contain an acylated aniline moiety like acetanilide,
paracetamol, or the N-acetyl derivative of PAP (17); these
compounds, however, exhibited longer retention times
under our chromatographic conditions. Peaks e, h , and
i show PAP-like spectra with minor shifts in the position
of their maxima, which suggests that these metabolites
bear an intact aniline moiety and that the modifications
should be located in the alkyl chain, as in acid 1.
when the concentration of quinoneimine 8 was maximal,
was assayed in mild acid media (HCl/MeOH), a complex
HPLC profile was obtained from which the presence of
quinoneimine 10 could not be inferred. Likewise, silyl-
ation of this crude reaction mixture yielded no peak by
GC/MS with the mass spectrum expected for the silylated
imine. However, the HPLC profile again revealed the
presence of 4-aminophenol.
Attempts to generate quinoneimine 10 by direct oxida-
tion of the phenolic PAP derivative 2 with Ag2O in
deuteriochloroform were also made (data not shown).
Although it was possible to detect by 1H NMR the initial
formation of a product with a set of signals in the
aromatic region similar to those described above for
quinoneimine 8, the more complex nature of this crude
reaction mixture prevented us from carrying out further
analysis.
14C-Ra d ioa ctivity Clea r a n ce. Table 1 depicts the
radioactivity clearance observed in both mouse strains
induced with â-NF after a single dose of each PAP
enantiomer spiked with [14C]-rac-PAP. As shown, we
collected 52-60% of the radioactive dose in the urine
samples over a 24 h period. In contrast, radioactivity
clearance in fecal samples was in all cases less than 2%.
This clearance profile was consistent with those previ-
ously reported for the racemate PAP (17). Similarly, the
radioactivity content measured in the analyzed tissues
was less than 1% of the total administered dose. There-
fore, the differences in the urine radioactivity clearance
observed between mice treated with each PAP enanti-
omer could be considered negligible, suggesting that both
isomers are metabolized and eliminated to a similar
extent.
We were confident of the reliable separation of urine
and feces samples in the metabolic cages used. However,
remaining radioactivity was encountered in the cage

1102 Chem. Res. Toxicol., Vol. 14, No. 8, 2001
Bujons et al.
F igu r e 1. Typical HPLC chromatograms of methanol-diluted urine samples from â-NF-induced A/J and C57BL/6 mice treated
with either (S)- or (R)-PAP spiked with [14C]-rac-PAP. 14C-radioactivity tracings are plotted 0.33 min to the left to correct for the
dead volume between the on-line UV and radioactivity detectors. HPLC conditions are described under Materials and Methods.
D of Figure 1 provide information about the relative
abundances of different metabolites. Thus, urine samples
from mice treated with (R)-PAP contained a higher
proportion (relative to the main single peak g) of me-
tabolites that give rise to peaks b-e and h -i than those
from mice treated with (S)-PAP. Mice treated with (R)-
PAP show also a higher proportion, relative to g, of the
peak assigned to PAP (k ). The differences among peaks
a , f, and j are less significant since, as stated above, their
UV spectra indicate overlapping of two or more peaks.
On the other hand, the profile differences observed
between the A/J and the C57BL/6 mouse strains treated
with either (R)- or (S)-PAP are less important.
The extraction of buffered urine samples with a
mixture of dichloromethane-isopropyl alcohol-ammonia
permitted the separation of PAP from most of its more
polar metabolites. HPLC analysis of these organic ex-
tracts revealed, apart from PAP and minor amounts of
acid 1 and other less abundant metabolites, a major
radioactive peak (Figure 3, arrowed peak) which showed
a counterpart in the DAD detector with maxima at 232
and 296 nm. The shape and relative abundance of this
peak appeared to be dependent on sample history, but
the peak could be detected in samples from both mouse
strains treated with each PAP enantiomer. Analysis by
GC/MS of the extracts treated with BSTFA permitted
detection of a product with an EI-MS spectrum (Figure
3, bottom panel) consistent with the structure of the
F igu r e 2. UV spectra extracted from the tracings shown in
bis(trimethylsilyl) derivative of 4-aminophenol. Confir-
mation of the identity of this metabolite as 4-aminophenol
was obtained from comparison of its chromatographic and
spectroscopic data with those of an authentic standard.
Figure 1 (panels C and D) for the peaks that show a correspond-
ing signal in the radioactivity detector. Spectra of peaks a , f,
and j are omitted since they show overlapping signals.
Although accurate quantitation of these metabolites in
the different urine samples was beyond the scope of the
present work, the chromatograms shown in panels C and
In agreement with our previously reported study, the
major metabolite in the urine samples identified as acid

Metabolism of (R)- and (S)-PAP in Mice
Chem. Res. Toxicol., Vol. 14, No. 8, 2001 1103
Ch a r t 2. P AP Der iva tive Str u ctu r es P ostu la ted
fr om th e An a lysis by GC/MS of Ur in e Sa m p les
fr om P AP -Tr ea ted C57BL/6- a n d A/J -Str a in Micea
aThe MS spectrum (m/z) main peak lists are given for each
structure.
detected in the experiments with rac-PAP (17), and at
that time, it was tentatively assigned to the quinoneimine
derivative 11 (Chart 2). In the same collected fraction,
two additional peaks were also detected which exhibited
spectra with molecular ions at m/z 251 and base peaks
at m/z 120 and 105, respectively, which are consistent
with the ketonic structures 12 and 13 (Chart 2).
Further attempts for conjugate cleavage were carried
out by treating urine samples with the glucuronidases
from E. coli and H. pomatia, followed by organic extrac-
tion, as described under Materials and Methods. HPLC
analysis of the aqueous phases showed the peak of acid
1, at 16.2 min, as the main radioactive peak, while the
polar radioactive components present in the untreated
urine in the 3-10 min region (Figure 1) were absent or
with considerably diminished intensity. In contrast, in
addition to the peak of PAP, acid 1, and 4-aminophenol
already observed in the extracts of buffered urine (Figure
3), analysis of the organic phases showed a new peak with
a retention time close to 14 min (Figure 4, arrowed peak)
and a UV spectrum with a maximum at 243 nm and a
shoulder at ∼280 nm, probably derived from the metabo-
lite that gave rise to peak c in Figure 1. This peak was
initially identified as paracetamol based on the compari-
son of its chromatographic mobility and UV spectrum
with those of a commercial standard. Final confirmation
of the structure of this metabolite was obtained by
derivatization of these organic phases with BSTFA and
analysis by GC/MS, which permitted the detection of a
peak with a retention time and an EI-MS spectrum
(Figure 4, bottom panel) identical to those exhibited by
the bis(trimethylsilyl) derivative of paracetamol (30).
These observations constitute strong evidence for the
presence of paracetamol conjugates, which are well
documented in the literature (31, 32), in the urine
samples of A/J and C57BL/6 mice treated with either of
the two PAP enantiomers. The organic extracts after
enzyme hydrolysis were enriched in a 15-20% radio-
activity count compared with nonhydrolyzed extracted
urine. This finding was consistent with the appearance
of 4-aminophenol and paracetamol in the chromato-
graphic profiles, and suggests glucuronide and sulfate
conjugates for these metabolites. The possibility that
these two metabolites could arise from the metabolism
F igu r e 3. Typical HPLC chromatograms [upper panel, (a) 14C-
monitoring; (b) UV-monitoring] of organic extracts of buffered
urine representative of both mouse strains treated with each
PAP enantiomer [tracings correspond to (R)-PAP-treated A/J
mice]. The PAP peak is labeled, and the UV spectrum of the
peak labeled with an arrow is shown in the inset. Analysis by
GC/MS, as described under Materials and Methods, permitted
the detection of one product with an EI-MS spectrum (bottom
panel) consistent with the bis(trimethylsilyl) derivative of
4-aminophenol.
1 (peak g, Figure 1), as well as the polar radioactive
components appearing in the 3-4 min region and most
of the other minor metabolites, remained in the aqueous
phase. With the exception of the disappearance of acid 1
and the consequent formation of the corresponding
methyl ester (17), treatment of lyophilized aliquots of
these aqueous phases with HCl/MeOH caused only minor
changes in the HPLC profiles.
Semipreparative HPLC injections of the urine samples
and collection of individual peaks or groups of peaks were
also carried out; these fractions were lyophilized, deriva-
tized with BSTFA, and analyzed by GC/MS. As occurred
in the experiments carried out with rac-PAP (17), the
O-silylated derivative of acid 1 and the corresponding
O-silylated derivatives of phenols 2 and 3a could be
detected in urine samples of mice treated with either of
the two PAP enantiomers in the 3-10 min fraction.
Moreover, a minor peak was detected in the 15-19 min
fraction of urine samples from (R)-PAP-treated A/J mice,
with an EI-MS spectrum showing a molecular ion at m/z
325 and a base peak at m/z 120. This product was already

1104 Chem. Res. Toxicol., Vol. 14, No. 8, 2001
Bujons et al.
cursors of quinoneimine derivatives which could be
formed through cytochrome P-450 or peroxidase-medi-
ated oxidations, as described for the case of the known
hepatotoxics paracetamol (22, 33) and amodiaquine
(23, 25). In this respect, the detection of minor products
which showed EI-MS spectral data consistent with the
quinoneimine structures 11 and 12 is noteworthy.
The detection of such quinoneimine compounds in
urine samples seems inconsistent with our current
observations concerning the instability of the synthetic
analogue 8, which suggests that these products in aque-
ous medium could verify a rapid isomerization to the
corresponding imines, followed by hydrolytic cleavage
with ultimate release of 4-aminophenol. However, this
possibility cannot be completely ruled out since the
observation of a base peak at m/z 120 is consistent with
fragmentation of the C2-C3 bond, which has been
systematically observed for all PAP derivatives studied
(17). In this respect, the detection of 4-aminophenol in
the organic extracts of buffered urine samples constitutes
additional indirect evidence of the presence of quinone-
imine species in those urine samples. We also detected
minor amounts of 4-aminophenol in samples of the
purified standards 2, 3b, and 4a (data not shown), which
suggests that nonenzymatic formation of this product can
be expected from molecules that contain such a 4-alkyl-
aminophenol moiety.
Similar observations have been reported during the
tyrosinase-catalyzed oxidation of N-(4-hydroxyphenyl)-
glycine 4a (28), as well as for the quinoneimine deriva-
tives of 4-methylaminophenol and other alkylamino-
phenols (29). It has been described that the formation of
4-aminophenol occurs through prototropic rearrangement
of the corresponding quinoneimines to give an imine
which undergoes hydrolytic cleavage to yield 4-amino-
phenol and the corresponding aldehyde (Scheme 2). Our
results indicate that such a mechanism can be extended
to the PAP phenolic derivatives considered here.
Taken together, these observations suggest that dif-
ferent quinoneimine PAP derivatives could be generated
in â-NF PAP-treated mouse strains. The possibility that
these quinoneimine compounds are an artifact due to the
spontaneous air oxidation of phenolic species such as 2
or 3a , followed by a transformation pathway such as that
indicated in Scheme 2, can be ruled out because of the
detection of paracetamol in the same urine samples. The
occurrence of paracetamol should be attributed to in vivo
acetylation of 4-aminophenol (34); thus, it strongly
indicates that these transformations must be part of the
in vivo metabolic fate of PAP in mice.
F igu r e 4. Typical HPLC chromatograms [upper panel, (a) 14C-
monitoring; (b) UV-monitoring] of organic extracts of H. pomatia
â-glucuronidase hydrolysis representative of urine samples from
both mice strains treated with each PAP enantiomer [tracings
correspond to (R)-PAP-treated A/J mice]. The PAP peak is
labeled as well as the peak of 4-aminophenol (asterisk). The UV
spectrum of the peak labeled with an arrow is shown in the
inset. Analysis by GC/MS, as described under Materials and
Methods, permitted the detection of one product with an EI-
MS spectrum (bottom panel) consistent with the bis(trimethyl-
silyl) derivative of paracetamol.
of residual aniline present in the PAP doses administered
to mice was ruled out through careful analysis of PAP
injection.
Discu ssion
The toxicological implications of these findings are
important because they provide evidence, for the first
time, that PAP metabolism can give rise to the formation
of products, such as paracetamol and 4-aminophenol,
with well-known toxicity. Furthermore, they suggest the
formation of highly reactive quinoneimines which might
be considered potentially toxic compounds, as well as
other metabolites derived from further transformation
of the alkyl chain, i.e., the R-keto acid precursor of the
tentatively identified 13. The range of PAP metabolites
may be even wider if we consider the alkyl derivatives,
probably aldehydes (Scheme 2), which should be released
during the cleavage of the carbon-nitrogen bond that
gives rise to the formation of 4-aminophenol. Additional
structures are also plausible if we consider that similar
transformations to those described in the oxidation of
Metabolism and clearance of intraperitoneally admin-
istered (R)- and (S)-PAP in A/J and C57BL/6 mice were
studied, and several metabolites excreted in urine were
identified. As previously reported for analogous experi-
ments with rac-PAP (17), approximately 60% of the total
radioactivity dose was found in urine samples of both
mouse strains; thus, urine elimination is the main PAP
clearance route. Analysis of urine samples from A/J and
C57BL/6 strains revealed qualitatively similar composi-
tions, with acid 1 as the major PAP derivative formed
from both enantiomers; however, animals treated with
(R)-PAP showed a higher proportion of other metabolites
(Figure 1). Among these metabolites, phenolic derivatives
of PAP (2) and the above-mentioned acid 1 (3a ) were
identified. These phenolic compounds are potential pre-

Metabolism of (R)- and (S)-PAP in Mice
Chem. Res. Toxicol., Vol. 14, No. 8, 2001 1105
N-(4-hydroxyphenyl)glycine (4a ) (28), to give for instance
muconic acid derivatives, may be taking place.
tion of a quinoneimine intermediate, suggests the pos-
sible generation of some of these species in a nonenzy-
matic way in different tissues, and their possible reaction
with macromolecules in a widespread manner after
higher, or accumulation of, PAP doses.
These findings have potential toxicological implications
for TOS since patients could have had an enzyme activity
impairment on pathways dealing with detoxification of
PAP derivatives, resulting in the accumulation of reac-
tive, unstable species similar to those reported here. In
this respect, we recently reported that TOS survivors
presented a specific profile derived from genetic poly-
morphisms on xenobiotic metabolism enzymes; in fact,
impaired NAT-2 (acetylation) may have mediated sus-
ceptibility in those patients compared with their geo-
graphical controls (16). Moreover, reactive metabolites
derived from xenobiotic metabolism have been associated
with autoimmune diseases related to xenobiotic exposure
(35, 36), with immune activation being a crucial charac-
teristic for TOS and EMS (37). In particular, aromatic
oxidized species and quinoneimines have proved to
induce some type of lymphocyte transformation (35, 36,
38).
To obtain an animal model in the future for in-depth
study of TOS pathogenic mechanisms, the data of Berk-
ing et al. (39) add interest since they reported a Th2-
lymphocyte pattern in the same mouse strains used in
this study after OLA dosage given intraperitoneally. This
Th2 pattern has also been reported in TOS patients (40).
This interesting study indicated the possibility of reactive
metabolites from oleanilides. In this respect, amidase
hydrolysis of OLA can occur to yield aniline, and the
aniline moiety possibly undergoes conjugation either by
microsomal carboxylesterase enzymes (41, 42) or by
acetylation (34). The latter acetanilide product can be
metabolized to yield 4-aminophenol and paracetamol (34).
Thus, a metabolic interaction between OLA and PAP
derivatives could be postulated, and the metabolic find-
ings (16) observed in TOS survivors (impaired acetyl-
ation) would have toxicological significance. Therefore,
A/J and C57BL/6 mice could be considered good candi-
dates for TOS-animal models, and our studies add a
further step to determine the toxicokinetics necessary for
future research seeking toxicity targets of these poten-
tially toxic metabolites.
In relation to other clinically related diseases, a link
between EMS and TOS has been suggested in the
literature owing to the chemical similarities of PAA and
PAP (43, 44). Even though we have not been able to
detect PAA in urine in any of our experiments, our results
suggest that PAA could suffer similar transformations
to PAP to yield quinoneimine metabolites and generate
4-aminophenol and paracetamol. In this respect, phenolic
metabolites from PAA-orally-administered rats had been
reported (19). Confirmation of this point would establish
an additional connection between the toxicokinetics of
both compounds, which does not require necessarily their
metabolic interconversion.
In summary, the results presented here indicate that
(R)- and (S)-PAP are metabolized in A/J and C57BL/6
â-NF-induced mice to yield a wide range of polar deriva-
tives, with higher rates observed for the (R)-PAP enan-
tiomer. Some of these polar metabolites are easily cleared
from the organism in urine, i.e., acid 1, and may
constitute a major detoxifying route in those mice,
whereas some others are highly reactive compounds. The
observation that samples of some of the phenolic com-
pounds considered here (2, 3b, and 4a ) spontaneously
generate 4-aminophenol, presumably through the forma-
Ack n ow led gm en t. This work was supported by
Grant FIS-97/1317-SAT of the Fondo de Investigacio´n
Sanitaria of the Spanish Ministry of Health. Ms. C.
O’Hara kindly revised the English version of the manu-
script.
Refer en ces
(1) Gradjean, P., and Tarkowski, S., Eds. (1984) Toxic Oil Syn-
drome: mass food poisoning in Spain. Report of a WHO Meeting:
Madrid, 21-25 March, 1983, World Health Organisation Regional
Office for Europe, Copenhagen, Denmark.
(2) Tabuenca, J . M. (1981) Toxic allergic syndrome caused by
ingestion of rapeseed oil denatured with aniline. Lancet 2, 567-
568.
(3) Diaz de Rojas, F., Castro-Garc´ıa, M., Abaitua-Borda, I., Alonso-
Gordo, J . M., Posada de la Paz, M., Kilbourne, E. M., and
Tabuenca-Oliver, J . M. (1987) The association of oil ingestion with
Toxic Oil Syndrome in two convents. Am. J . Epidemiol. 125, 907-
911.
(4) Can˜as, R., and Kilbourne, E. M. (1987) Oil ingestion and the Toxic
Oil Syndrome. Results of a survey of residents of the Orcasur
neighborhood in Madrid Spain. Int. J . Epidemiol. 16, 3-6.
(5) Rigau-Perez, J . G. (1984) Summary of case-control studies and
case or cluster investigations. In Toxic Oil Syndrome: mass food
poisoning in Spain. Report of a WHO Meeting: Madrid, 21-25
March, 1983 (Gradjean, P., and Tarkowski, S., Eds.) pp 47-51,
World Health Organisation Regional Office for Europe, Copen-
hagen, Denmark.
(6) Rigau-Perez, J . G., Pe´rez-Alvarez, L., Duen˜as-Castro, S., Choi,
K., Thacker, S. B., Germain, J . L., Gonza´lez-de-Andre´s, G.,
Canada-Royo, L., and Pe´rez-Gallardo, F. (1984) Epidemiologic
investigation of an oil-associated pneumonic paralytic eosinophilic
syndrome in Spain. Am. J . Epidemiol. 119, 250-260.
(7) Bernert, J . T., Kilbourne, E. M., Akins, J . R., Posada de la Paz,
M., Meredith, N. K., Abaitua-Borda, I., and Wages, S. (1987)
Compositional analysis of oil samples implicated in the Spanish
toxic oil syndrome. J . Food Sci. 52, 1562-1569.
(8) Va´zquez-Roncero, A., J aner del Valle, C., Maestro-Duran, R., and
Graciane-Constante, E. (1983) New aniline derivatives in cooking
oils associated with the toxic oil syndrome. Lancet 2, 1024-1025.
(9) Hill, R. H., Schurz, H. H., Posada de la Paz, M., Abaitua-Borda,
I., Philen, R. M., Kilbourne, E. M., Head, S. L., Bailey, S. L.,
Driskell, W. J ., Barr, J . R., and Needham, L. L. (1995) Possible
etiologic agents for Toxic Oil Syndrome: fatty acid esters of 3-(N-
phenylamino)-1,2-propanediol. Arch. Environ. Contam. Toxicol.
28, 259-264.
(10) Posada de la Paz, M., Philen, R. M., Schurz, H. H., Hill, R. H.,
Gimenez-Ribota, O., Go´mez de la Ca´mara, A., Kilbourne, E. M.,
and Abaitua-Borda, I. (1999) Epidemiologic evidence for a new
class of compounds associated with Toxic Oil Syndrome. Epi-
demiology 10, 130-134.
(11) Posada de la Paz, M., Philen, R. M., Abaitua Borda, I., Bernert,
J . T., Bada-Gancedo, J . C., DuClos, P. J ., and Kilbourne, E. M.
(1996) Toxic Oil Syndrome: traceback of the toxic oil and evidence
for a point source epidemic. Food Chem. Toxicol. 34, 251-257.
(12) Aldridge, W. N. (1992) Experimental studies. In Toxic Oil
Syndrome: current knowledge and future perspectives (Nadal, J .,
and Tarkowski, S., Eds.) pp 63-93, WHO Regional Publications:
European Series 42, World Health Organization Regional Office
for Europe, Copenhagen, Denmark.
(13) Va´zquez-Roncero, A., Maestro-Duran, R., and Ruiz-Gutierrez, V.
(1984) New aniline derivatives in oils related to the Toxic
Syndrome. Toxicity in mice of 3-phenylamino-1,2-propanediol and
its fatty acid mono- and diesters. Grasas Aceites (Sevilla) 35, 330-
331.
(14) Maestro-Duran, R., Ruiz-Gutierrez, V., and Va´zquez-Roncero, A.
(1986) Toxic Oil Syndrome in Spain: toxicological study in mice
of new aniline derivatives present in toxic oils. Grasas Aceites
(Sevilla) 37 (4), 207-214.
(15) Ruiz-Gutierrez, V., and Maestro-Duran, R. (1992) Lymphatic
absorption of 3-phenyalmino-1,2-propanediol and its esters. Exp.
Toxicol. Pathol. 44, 29-33.

1106 Chem. Res. Toxicol., Vol. 14, No. 8, 2001
Bujons et al.
(16) Ladona, M. G., Izquierdo, M., Posada de la Paz, M., De la Torre,
R., Ampurdane´s, C., Segura, J ., and Sanz, E. (2001) Pharmaco-
genetic profile of xenobiotic enzyme metabolism in survivors of
the Spanish Toxic Oil Syndrome. Environ. Health Perspect. 109,
369-375.
plasma and urine using mass fragmentography. Biomed. Mass
Spectrom. 3, 60-63.
(31) Andrews, R. S., Bond, C. C., Burnett, J ., Saunders, A., and
Watson, K. (1976) Isolation and identification of paracetamol
metabolites. J . Int. Med. Res. 4, 34-39.
(17) Ladona, M. G., Bujons, J ., Messeguer, A., Ampurdane´s, C.,
Morato´, A., and Corbella, J . (1999) Biotransformation and clear-
ance of 3-(phenylamino)propane-1,2-diol, a compound present in
samples related to Toxic Oil Syndrome, in C57BL/6 and A/J mice.
Chem. Res. Toxicol. 12, 1127-1137.
(18) Dahlin, D. C., and Nelson, S. D. (1982) Synthesis, decomposition
kinetics, and preliminary toxicological studies of pure N-acetyl-
p-benzoquinone imine, a proposed toxic metabolite of acetami-
nophen. J . Med. Chem. 25, 885-886.
(19) Adachi, J ., Mio, T., Ueno, Y., Naito, T., Nishimura, A., Fujiwara,
S., Sumino, K., and Tatsuno, Y. (1994) Identification of four
metabolites of 3-(phenylamino)alanine, a constituent in L-tryp-
tophan products implicated in eosinophilia-myalgia syndrome,
in rats. Arch. Toxicol. 68, 500-505.
(20) Islam, I., Skibo, E. B., Dorr, R. T., and Alberts, D. S. (1991)
Structure-activity studies of antitumour agents based on pyrrolo-
[1,2-a]benzimidazoles: New reductive alkylating DNA cleaving
agents. J . Med. Chem. 34, 2954-2961.
(32) Watari, N., Iwai, M., and Kaneniwa, N. (1983) Pharmacokinetic
study of the fate of paracetamol and its conjugates in rats. J .
Pharmacokinet. Biopharm. 11, 245-272.
(33) Potter, D. W., and Hinson, J . A. (1987) The 1- and 2-electron
oxidation of paracetamol catalyzed by prostaglandin H synthase.
J . Biol. Chem. 262, 974-980.
(34) Gorrod, J . W., and Manson, D. (1986) The metabolism of aromatic
amines. Xenobiotica 16, 933-955.
(35) Griem, P., Wulferink, M., Sachs, B., Go´nzalez, J . B., and Gleich-
mann, E. (1998) Allergic and autoimmune reactions due to
xenobiotics: how do they arise? Immunol. Today 19, 133-141.
(36) Hess, D. A., and Rieder, M. J . (1997) The role of reactive drug
metabolites in immune-mediated adverse drug reactions. Ann.
Pharmacother. 31, 1378-1387.
(37) Philen, R. M., and Posada de la Paz, M. (1993) Toxic Oil Syndrome
and Eosinophilia Myalgia Syndrome: May 8-10, 1991, World
Health Organization Meeting Report. Semin. Arthritis Rheum.
23, 104-124.
(38) Kubicka-Muranyi, M., Goebels, R., Goebel, C., Uetrecht, J ., and
Gleichmann, E. (1993) T Lymphocytes ignore procainamide, but
respond to its reactive metabolites in peritoneal cells: demonstra-
tion by the adoptive transfer popliteal lymph node assay. Toxicol.
Appl. Pharmacol. 122, 88-94.
(39) Berking, C., Hobbs, M. V., Chatelain, R., Meurer, M., and Bell,
S. A. (1998) Strain-dependent cytokine profile and susceptibility
to oleic acid anilide in a murine model of the Toxic Oil Syndrome.
Toxicol. Appl. Pharmacol. 148, 222-228.
(40) del Pozo, V., de Andre´s, B., Gallardo, S., Cardaba, B., de Arruda-
Chaves, E., Cortegano, I., J urado, A., Palomino, P., Oliva, H.,
Aguilera, B., Posada, M., and Lahoz, C. (1997) Cytokine mRNA
(21) Zimmer, H., Lankin, D. C., and Horgan, S. W. (1970) Oxidations
with potassium nitrosodisulfonate (Fremy’s radical). The Teuber
reaction. Chem. Rev. 71, 229-246.
(22) Potter, D. W., and Hinson, J . A. (1987) Mechanisms of paraceta-
mol oxidation to N-acetyl-p-benzoquinone imine by horseradish
peroxidase and cytochrome P-450. J . Biol. Chem. 262, 966-973.
(23) Thompson, C. D., Vachaspati, P. R., Kolis, S. P., Gulden, P. H.,
Garst, M. E., Wiese, A., Munk, S. A., Harman, W. D., and
Macdonald, T. L. (1997) A proposed mechanism for p-aminocloni-
dine allergenicity based on its relative oxidative lability. Chem.
Res. Toxicol. 10, 1032-1036.
(24) Bolton, J . L., Trush, M. A., Penning, T. M., Dryhurst, G., and
Monks, T. J . (2000) Role of quinones in toxicology. Chem. Res.
Toxicol. 13, 135-160.
expression in lung tissue from Toxic Oil Syndrome patients:
a
(25) Harrison, A. C., Kitteringham, N. R., Clarke, J . B., and Park, B.
K. (1992) The mechanism of bioactivation and antigen formation
of amodiaquine in the rat. Biochem. Pharmacol. 43, 1421-1430.
(26) Bessems, J . G. M., Te Koppele, J . M., van Dijk, P. A., van Stee,
L. L. P., Commandeur, J . N. M., and Vermeulen, N. P. E. (1996)
Rat liver microsomal cytochrome P450-dependent oxidation of 3,5-
disubstituted analogues of paracetamol. Xenobiotica 26, 647-666.
(27) Zeller, K. P., and Mu¨ller, R. (1988) Mass spectra of quinones. In
The Chemistry of functional groups. The Chemistry of quinonoid
compounds (Patai, S., and Rappoport, Z., Eds.) Vol. 2, pp 1231-
1292, J . Wiley & Sons, Chichester, U.K.
Th2 immunological mechanism. Toxicology 118, 61-70.
(41) Kaphalia, B. S., Fritz, R. R., and Ansari, G. A. S. (1977)
Purification and characterization of rat liver microsomal fatty acid
ethyl and 2-chloroethyl ester synthase and their relationship with
carboxylesterase (pI 6.1). Chem. Res. Toxicol. 11, 211-218.
(42) Kaphalia, B. S., Green, S. M., and Ansari, G. A. S. (1999) Fatty
acid ethyl and methyl ester synthases, and fatty acid anilide
synthase in HepG2 and AR42J cells: interrelationships and
inhibition by tri-o-tolyl phosphate. Toxicol. Appl. Pharmacol. 159,
134-141.
(28) Mascagna, D., Costantini, C., d’Ischia, M., and Prota, G. (1992)
Tyrosinase-catalyzed oxidation of the new melanocytotoxic agent
N-4-hydroxyphenylglycine: unusual cleavage of the 4-amino-
phenol ring to muconic semialdehyde derivatives. Tetrahedron
48, 8309-8316.
(29) Brown, E. R. (1988) Quinonediimines, monoimines and related
compounds. In The Chemistry of functional groups. The Chemistry
of quinonoid compounds (Patai, S., and Rappoport, Z., Eds.) Vol.
2, pp 1231-1292, J . Wiley & Sons, Chichester, U.K.
(43) Mayeno, A. N., Belongia, E. A., Lin, F., Lundy, S. K., and Gleich,
G. J . (1992) 3-(Phenylamino)alanine, a novel aniline-derived
amino acid associated with the eosinophilia-myalgia syndrome:
a link to the toxic oil syndrome? Mayo Clin. Proc. 67, 1134-1139.
(44) Mayeno, A. N., Benson, L. M., Naylor, S., Colberg-Beers, M.,
Puchalski, J . T., and Gleich, G. J . (1995) Biotransformation of
3-(phenylamino)-1,2-propanediol to 3-(phenylamino)alanine:
a
chemical link between Toxic Oil Syndrome and Eosinophilia-
Myalgia Syndrome. Chem. Res. Toxicol. 8, 911-916.
(30) Baty, J . D., Robinson, P. R., and Wharton, J . (1976) A method
for the estimation of acetanilide, paracetamol and phenacetin in
TX010001K