Thiophene sulfoxides as reactive metabolites: Formation upon microsomal oxidation of a 3-aroylthiophene and fate in the presence of nucleophiles in vitro and in vivo

Article abstract of DOI:10.1021/tx9601622

Oxidative metabolism of a 3-aroylthiophene, 1, by rat liver microsomes in the presence of mercaptoethanol as a trapping agent led to the isolation of four main compounds, 2-5, which have been isolated and characterized by UV, ¹H NMR, and mass spectroscopy. They all derive from two primary metabolites, 2 and 3, which result from the nucleophilic addition of mercaptoethanol to a reactive, very electrophilic intermediate formed by sulfoxidation of the thiophene ring of 1. Further reactions of diastereoisomers 2 and 3 with mercaptoethanol led to compound 4 that is opened at the level of its thiophene ring and, eventually, to a final metabolite 5 resulting formally from the addition of mercaptoethanol on the 4, 5-double bond of the thiophene ring of 1. Compound 5 is very stable even in the presence of a large excess of mercaptoethanol. Similar reactions were observed upon microsomal oxidation of I in the presence of another thiol, N- acetylcysteine. Final metabolites 8a and 8b equivalent to 5 except for the replacement of its mercaptoethanol substituent with an N-acetylcysteinyl group were isolated and characterized by UV, ¹H NMR, and mass spectroscopy. Interestingly, after treatment of rats with 1, metabolites 8a and 8b could be detected in urine, indicating that the successive reactions, that were observed in vitro after microsomal oxidation of 1 in the presence of a thiol- containing trapping agent, also occur in vivo, glutathione acting as a nucleophile in that case. These data provide clear evidence for the intermediate formation of a reactive, electrophilic thiophene sulfoxide in metabolic oxidation of 1 in vitro and in vivo. They also provide the first data on the complex reactivity of such thiophene sulfoxides, whose chemistry is poorly known, and on their fates in living organisms.

Full text of DOI:10.1021/tx9601622

Chem. Res. Toxicol. 1996, 9, 1403-1413
1403
Th iop h en e Su lfoxid es a s Rea ctive Meta bolites:
F or m a tion u p on Micr osom a l Oxid a tion of a
3-Ar oylth iop h en e a n d F a te in th e P r esen ce of
Nu cleop h iles in Vitr o a n d in Vivo
Philippe Valadon, Patrick M. Dansette, J ean-Pierre Girault, Claudine Amar, and
Daniel Mansuy*
Laboratoire de Chimie et Biochimie Pharmacologiques et Toxicologiques, URA 400 CNRS,
Universite´ Paris V, 45 Rue des Saints-Pe`res, 75270 Paris Cedex 06, France
Received September 11, 1996X
Oxidative metabolism of a 3-aroylthiophene, 1, by rat liver microsomes in the presence of
mercaptoethanol as a trapping agent led to the isolation of four main compounds, 2-5, which
have been isolated and characterized by UV, 1H NMR, and mass spectroscopy. They all derive
from two primary metabolites, 2 and 3, which result from the nucleophilic addition of
mercaptoethanol to a reactive, very electrophilic intermediate formed by sulfoxidation of the
thiophene ring of 1. Further reactions of diastereoisomers 2 and 3 with mercaptoethanol led
to compound 4 that is opened at the level of its thiophene ring and, eventually, to a final
metabolite 5 resulting formally from the addition of mercaptoethanol on the 4,5-double bond
of the thiophene ring of 1. Compound 5 is very stable even in the presence of a large excess
of mercaptoethanol. Similar reactions were observed upon microsomal oxidation of 1 in the
presence of another thiol, N-acetylcysteine. Final metabolites 8a and 8b equivalent to 5 except
for the replacement of its mercaptoethanol substituent with an N-acetylcysteinyl group were
isolated and characterized by UV, 1H NMR, and mass spectroscopy. Interestingly, after
treatment of rats with 1, metabolites 8a and 8b could be detected in urine, indicating that the
successive reactions, that were observed in vitro after microsomal oxidation of 1 in the presence
of a thiol-containing trapping agent, also occur in vivo, glutathione acting as a nucleophile in
that case. These data provide clear evidence for the intermediate formation of a reactive,
electrophilic thiophene sulfoxide in metabolic oxidation of 1 in vitro and in vivo. They also
provide the first data on the complex reactivity of such thiophene sulfoxides, whose chemistry
is poorly known, and on their fates in living organisms.
of proteins would result in covalent binding of tienilic acid
to proteins. Intermediate formation of a thiophene
sulfoxide metabolite has been recently established in the
oxidative metabolism of two thiophene derivatives (13,
14). Oxidation of the isomer 1 of tienilic acid with liver
microsomes in the presence of a good nucleophile such
as mercaptoethanol has led to the isolation of a 2,5-
dihydrothiophene sulfoxide as a major metabolite which
clearly derived from the addition of mercaptoethanol to
the thiophene sulfoxide of 1 (13). The major urinary
metabolite of thiophene itself in rats has been shown only
recently to be a mercapturate deriving formally from the
addition of N-acetylcysteine to thiophene sulfoxide (14)
(Figure 1). These results suggest that thiophene sulfox-
ides, a new class of reactive metabolites, could play a
central role in the oxidative metabolism of thiophene
compounds.
In tr od u ction
Although there is evidence that several thiophene
derivatives cause toxic effects (1-4), very little is known
not only on the molecular mechanisms of these effects
but also, in a more general manner, on the oxidative
metabolism of the thiophene ring in mammals (5, 6).
Oxidation of tienilic acid, a diuretic drug that was
involved in the appearance of rare cases of immunoaller-
gic hepatitis (7, 8), leads to 5-hydroxytienilic acid as a
major metabolite in vivo and in vitro (9). This 5-hy-
droxylation of the thiophene ring of tienilic acid is mainly
catalyzed by cytochrome P450 2C9 in human liver (10,
11) and leads to a suicide inactivation of this cytochrome
(12). It has been recently proposed (12) that the thiophene
sulfoxide of tienilic acid could be the reactive, electrophilic
metabolite responsible both for the formation of 5-hy-
droxytienilic acid and for the covalent binding of tienilic
acid to liver proteins including P450 2C9 itself. Accord-
ing to this mechanism (Figure 1), tienilic acid sulfoxide
would rapidly react with various nucleophiles by a
Michael-type addition at position 5 of the thiophene ring;
reaction with H2O would lead eventually to 5-hydroxy-
tienilic acid whereas reaction with nucleophilic residues
However, it is noteworthy that the chemistry of
thiophene sulfoxides is still poorly known (15). In fact,
only very few thiophene sulfoxides have been prepared
and identified so far. Most of them bear two bulky
substituents at positions 2 and 5 (16), and it is only very
recently that a complete X-ray structure has been re-
ported for one of them (17). Very few data are available
on their chemical reactivity if one excepts the Diels-
Alder reactivity of thiophene sulfoxide itself and of a few
substituted derivatives (15, 18-22). As far as sulfoxides
of thiophenes bearing a keto substituent, such as tienilic
acid and its isomer 1, are concerned, almost nothing was
* Address correspondence to Dr. Daniel Mansuy, URA 400, 45, Rue
des Saints-Pe`res, 75270 Paris Cedex 06, France. Phone: (33) 01
42.86.21.87; FAX: (33) 01 42.86.83.87; E. Mail: dmjccjcd@bisance.citi2.fr.
X Abstract published in Advance ACS Abstracts, November 1, 1996.
S0893-228x(96)00162-2 CCC: $12.00 © 1996 American Chemical Society

1404 Chem. Res. Toxicol., Vol. 9, No. 8, 1996
Valadon et al.
F igu r e 1. Involvement of thiophene sulfoxides as key intermediates in the in vitro microsomal oxidation of tienilic acid, and in the
in vivo metabolism of thiophene in rats.
addition of 75 µL of acetonitrile or in acidic conditions by
addition of 80 µL of CH3CN/CH3COOH, 72.5/7.5 v/v. After 10
min at 4 °C, proteins were separated by microcentrifugation and
CH3CN was evaporated under a nitrogen flux. Acetate buffer
(0.1 M, pH 4.6) was then added to the samples prior to storage
at -80 °C and analytical chromatography. Large-scale incuba-
tions of 10-50 mL were performed for preparative chromatog-
raphy according to a similar protocol with gentle stirring to
known about their chemical reactivity and possible
behavior in a biological medium. This article describes
in a complete manner the formation of a thiophene
sulfoxide during metabolism of 1 in vitro, its trapping
by thiol-containing nucleophiles, and the complex reac-
tivity of the derived thiophene sulfoxide-nucleophile
adducts as a function of the incubation conditions. This
work provides insight into the reactions that occur in vivo
after formation of this thiophene sulfoxide and its reac-
tion with glutathione whose concentration is very high
in many cells, particularly in the liver. It also provides
a first view on the chemical reactivity of aroylthiophene
sulfoxides and of their adducts with thiol-containing
nucleophiles.
overcome microsome aggregation. After
a 10 min, 2000g
centrifugation, incubation supernatants were further purified
on a SepPack C18 column (Waters-Millipore) and metabolites,
as well as remaining 1, eluted with 2 mL of methanol. High-
pressure liquid chromatography (HPLC) was done on an Altex
320 gradient system using a C8 5 µm MOS Hypersil column (1
mL/min flow rate) with two main elution gradients: gradient
A (solution C: 0.1 M ammonium acetate, pH 4.6; solution D:
H2O/CH3CN, 50/50 v/v; 0-100% D in C in 20 min) and gradient
B (same solutions, starting with 33% D in C isocratic for 6 min,
45% D in C isocratic for 7 min, followed by a gradient 45-100%
in 7 min). UV and visible spectra were recorded on-line by a
Spectra-Focus system (Spectra-Physics); 0.5 mL fractions were
collected in 3 mL propylene tubes (LKB collector), and radio-
activity was counted after addition of 2 mL of Picofluor 30
(Packard) on a Packard Tri-carb 4500 scintillation counter.
Cova len t Bin d in g to Micr osom a l P r otein s. Covalent
binding of metabolites of 1 to microsomal proteins was detected
according to a previously described protocol (26-28) with the
following modifications (10): 50 µL incubation medium aliquots
were dropped on glass fiber filters (GF/B from Whatman) and
immediately dipped in a large excess of acid (10% trifluoroacetic
acid in water) in order to precipitate proteins. After two 10 min
washes with methanol and one with CH3COOEt, filters were
air-dried and radioactivity was counted after addition of 3 mL
of toluene scintillator (Packard).
Exp er im en ta l Section
All reagents were of the highest quality commercially avail-
able. Tienilic acid isomer 1 was a gift of Anphar-Rolland
Laboratories (Chilly-Mazarin, France). Its labeled derivative
(
14COAr, 25 Ci/mol, 98% radiochemical purity determined from
HPLC as described previously (10)) was prepared by Amersham
(Bucks, U.K.).
P r ep a r a tion of Ra t Liver Micr osom es. Sprague-Dawley
male rats (200 g) were treated with clofibrate (10% v/v in maize
oil) according to the following protocol: day 0: 250 mg/kg ip,
day 1: 500 mg/kg ip, and day 2: 1000 mg/kg ip. On day 4, rats
were sacrified and microsomes of pooled livers were prepared
as previously described (23). Protein concentrations were
estimated by the method of Lowry (24); cytochrome P450
concentrations were measured according to the method de-
scribed by Omura and Sato (25).
Micr osom a l In cu ba tion s a n d HP LC An a lysis. In ana-
lytical conditions, rat liver microsomal suspensions containing
0.15 nmol of cytochrome P450 (1.6 nmol of P450‚(mg of
protein)-1), 15 nmol of [14C]-1 (25 mCi/mmol), and mercapto-
ethanol were mixed at 4 °C in a hemolysis tube in 150 µL of 0.1
M phosphate buffer (pH 7.4), containing 1 mM diethylenetri-
aminepentaacetic acid (DETAPAC).1 After a 3 min preincuba-
tion step at 37 °C with stirring under air, the reaction was
started by the addition of an NADPH-generating system (2
U/mL glucose-6-phosphate dehydrogenase, 10 mM glucose
6-phosphate, and 1 mM NADP). Controls were done in the
absence of the NADPH generating system. After 20 min, the
reactions were stopped either in neutral conditions by the
Sp ectr oscop ic Stu d ies. 1H and 13C NMR spectra were
recorded on Bruker WM 250 and AM 400 spectrometers.
Chemical shifts are given in ppm relative to (CH3)4Si and J in
Hz. The peak of monodeuteriated water (HOD) at 4.80 ppm
was eliminated when necessary either by presaturation or by
differential relaxation water elimination Fourier transform
(WEFT) experiment, or displaced by heating at 70 °C. UV and
visible spectra were recorded on a Kontron Uvikon 820 spec-
trophotometer; IR spectra were recorded on a Perkin FT IR
spectrometer in Nujol. Mass spectra were recorded on a Riber-
Mag R10-10C spectrometer with chemical ionization (NH3); in
the text, data are given after CIMS for chemical ionization mass
spectrum.
P r ep a r a tion of Com p ou n d s 2, 3, 4, a n d 5. These com-
pounds were prepared from 35 min incubations at 37 °C of large
volumes (20-40 mL) of 0.1 M phosphate buffer (pH 7.4)
containing microsomal suspensions (1.2 µM P450), 70 µM 1, 1
1
Abbreviations: COSY: Correlation spectroscopy; DETAPAC: di-
ethylenetriaminepentaacetic acid; HOD: monodeuterated water;
NOE: nuclear Overhauser effect; WEFT: water elimination Fourier
transform.

Thiophene Sulfoxides as Reactive Metabolites
Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1405
mM DETAPAC, an NADPH-generating system, and proper
concentrations of mercaptoethanol (from Figure 3). After
centrifugation at 2000g for 10 min, incubation supernatants
were purified on a SepPack C18 column and metabolites eluted
with methanol. After evaporation of methanol under nitrogen,
the pH of the solution was adjusted to 3 by addition of HCl,
and metabolites were extracted with CH3COOEt. Their separa-
tion by HPLC gave pure compounds 2, 3, 4, or 5 as ammonium
salts after lyophilization, in sufficient amounts (ca. 1 mg for 2,
P r ep a r a tion of Com p ou n d s 8a a n d 8b. Incubation of 1
(100 µM) with microsomes containing 1.2 µM cytochrome P450
in the presence of N-acetylcysteine (5 mM) and an NADPH-
generating system in a total volume of 40 mL was done for 2 h
at 37 °C. After acidification to pH 4.5, centrifugation for 10 min
at 2000g, SepPak extraction, and elution with CH3OH, the
solvent was evaporated and the residue taken in 0.5 mL of H2O,
acidified to pH 2 with HCl, extracted in CH3COOEt, and
immediately methylated with CH2N2 in Et2O. Dimethyl esters
of compounds 8a and 8b were purified by HPLC on a MOS
hypersil column with isocratic elution (CH3CN/H2O, 45:55).
8a dimethyl ester: 1H NMR (250 MHz, CD2Cl2) δ ) 7.18 (d,
9 Hz, 1H, H2′), 7.12 (s, 1H, H2), 6.80 (d, 9 Hz, 1H, H3′), 6.48 (dd,
5.2 Hz, 1H, NH), 4.83 (dd, 5-2 Hz, 1H, H1′′), 4.78 (s, 2H, OCH2),
4.69 (dd, 8-1.5 Hz, 1H, H4), 3.94 (dd, 13-8 Hz, 1H, H5a), 3.80
(s, 3H, OCH3′), 3.78 (s, 3H, OCH3′′), 3.45 (dd, 13-1.5 Hz, 1H,
H5b), 3.23 (dd, 14-5 Hz, 1H, H2′′a), 3.14 (dd, 14-5 Hz, 1H, H2′′b),
and 2.00 (s, 3H, COCH3); CIMS (NH3) m/z (%): 345 (100), 362
(10), 522 (4, MH+), and 539 (0.25, M + NH4+). H1′′ and H2′′ are
the protons of CR and Câ of the cysteine residue.
8b dimethyl ester: 1H NMR (250 MHz, CD2Cl2) δ ) 7.43 (d,
10 Hz, NH), 7.21 (d, 9 Hz, 1H, H2′), 7.17 (s, 1H, H2), 6.81 (d, 9
Hz, 1H, H3′), 5.02 (dt, 10-5 Hz, 1H, H5a), 4.79 (s, 2H, OCH2),
4.70 (dd, 8-1 Hz, 1H, H4), 3.95 (dd, 13-8 Hz, 1H, H5a), 3.80 (s,
3H, OCH3′), 3.72 (s, 3H, OCH3′′), 3.48 (dd, 13-1 Hz, 1H, H5b),
3.27 (dd, 14-5 Hz, 1H, H2′′a), 3.01 (dd, 14-5 Hz, 1H, H2′′b), and
2.11 (s, 3H, COCH3); CIMS (NH3), m/z (%): 178 (100), 345 (80),
365 (5), 522 (35, MH+), and 539 (4, M + NH4+).
Meta bolism of 1 in Vivo in Ra ts. Two male Sprague-
Dawley rats (300 g) were injected ip with 1 (30 mg/kg, 0.5 mCi/
mmol, Tris salt in H2O), and urine was collected in metabolic
cages for 24 h. Each urine fraction (1-2, 2-4, 4-6, and 6-24
h) was extracted on a SepPak C18 with CH3OH and analyzed
by HPLC. The extracts were combined, evaporated under N2
and taken up in water, acidified to pH 2 with HCl, back-
extracted in CH3COOEt, and methylated with CH2N2 in Et2O.
Three major metabolites were isolated. The first two ones
exhibited 1H NMR, UV, and CIMS spectra identical to those of
8a and 8b dimethyl esters described above. The third one was
found to be relatively unstable, giving back to 1 methyl ester.
It had a UV spectrum identical to those of 8a and 8b dimethyl
esters. 1HNMR (250 MHz, CD2Cl2 after exchange with D2O) δ
) 7.28 (d, 8 Hz, 1H, H2′), 7.20 (s, 1H, H2), 6.81 (d, 8 Hz, 1H,
H3′), 4.89 (dd, 8-1.5 Hz, 1H, H4), 4.79 (s, 2H, OCH2), 4.54 (t, 5
Hz, H1′′), 3.96 (dd, 13-8 Hz, 1H, H5a), 3.80 (s, 3H, COOCH3′),
3.76 (s, 3H, COOCH3′′), 3.51 (dd, 13-1.5 Hz, 1H, H5b), 3.13 (dd,
14-5 Hz, 1H, H2′′a), 3.04 (dd, 14-5 Hz, 1H, H2′′b).
1
3, and 4, and 2-7 mg for 5) for structure determination by H
NMR, UV, and mass spectroscopy, but not for elemental
analysis. The 1H NMR characteristics of their ammonium salts
are described in the tables. Therefore, only data corresponding
to the 1H NMR spectra of their methyl esters (prepared upon
reaction with CH2N2 in Et2O) and to the mass spectra of their
ammonium salts are reported in the experimental part.
Com p ou n d s 2 a n d 3. Metabolites 2 and 3 were obtained
in 60% yield from microsomal incubations in the presence of
100 µM mercaptoethanol. Preparative HPLC of metabolites 2
or 3 involved the use of ammonium acetate (50 mM) as solution
A in order to prevent degradation upon lyophilization. Succes-
sive slow gradients such as 25-33% solution B (50% CH3CN in
H2O) in A in 20 min allowed progressive enrichment and
isolation of each metabolite as hygroscopic orange powders: IR
(Nujol) ν ) 1610, 1480, 1425, 1405, 1340, 1290, 1270, 1055, and
1030 cm-1; CIMS (NH3) m/z (%): 331 (25), 407 (100, MH+
H2O), and 424 (10, M+ + NH4 - H2O).
-
Com p ou n d 4. Fractions containing 4 and 5 were first
isolated by HPLC (gradient A); 4 was further purified to
homogeneity in isocratic conditions (40% solution B). It was
obtained as a white powder. CIMS (NH3) m/z (%): 172 (100),
331 (80), 348 (40), 377 (5), 391 (12), 408 (10), 424 (1), 451 (2),
467 (2), 485 (5), 502 (1), 545 (0.5), and 563 (1, MH+). Methyl
ester of 4: 1H NMR (250 MHz, CD2Cl2) δ ) 7.29 (s, 1H, H2),
7.17 (d, 9 Hz, 1H, H2′), 6.80 (d, 9 Hz, 1H, H3′), 4.77 (s, 2H), 4.39
(dd, 10 and 6 Hz, 1H, H4), 3.85 (6H), 3.80 (s, 3H, CH3), 3.55
(dd, 10 Hz, 1H, H5), 3.27 (dd, 6 Hz, 1H, H5), 2.9 (4H), 2.84 (t, 6
Hz, 2H), 2.32 (1H, OH), and 2.16 (2H, OH); CIMS (NH3) m/z
(%): 345 (100), 362 (50), 391 (90), 423 (10), 499 (30), and 577
(5, MH+).
Com p ou n d 5. Compound 5 was purified by HPLC (gradient
A) and obtained as a faint yellow powder (ammonium salt):
CIMS (NH3) m/z (%): 172 (100), 331 (50), 348 (60), 391 (2), 409
(10, MH+), and 426 (10, M + NH4+). Methyl ester of 5: 1H NMR
(250 MHz, CD2Cl2) δ ) 7.20 (d, 9 Hz, 1H, H2′), 7.14 (s, 1H, H2),
6.80 (d, 9 Hz, 1H, H3′), 4.81 (dd, 8 and 1.5 Hz, 1H, H4), 4.78 (s,
2H), 3.98 (dd, 12 and 8 Hz, 1H, H5), 3.83 (m, 6 Hz, 2H), 3.80 (s,
3H, CH3), 3.50 (dd, 12 and 1.5 Hz, 1H, H5), and 2.85 (m, 6 Hz,
2H); CIMS (NH3) m/z (%): 345 (100), 362 (60), 423 (30, MH+),
and 440 (3, M + NH4+).
Resu lts a n d Discu ssion
Com p ou n d 6. A mixture of 2 and 3 (0.5 mg in 0.5 mL of
H2O) was treated with 0.5 mL of 2 M HCl for 2 h at 37 °C.
Metabolite 6 was then purified by HPLC in gradient A condi-
tions and obtained as a white powder (ammonium salt). CIMS
(NH3) m/z (%): 331 (40), 348 (5), 353 (45), 371 (3), 388 (3), 407
(80, MH+), and 424 (3, M + NH4+). Methyl ester of 6: 1H NMR
(250 MHz, CD2Cl2) δ ) 7.27 (d, 9 Hz, 1H, H2′), 7.13 (d, 5.5 Hz,
1H, H5), 6.97 (d, 5.5 Hz, 1H, H4), 6.83 (d, 9.0 Hz, 1H, H3′), 4.79
(s, 2H), 3.93 (q, 6 Hz, 2H), 3.81 (t, 6 Hz, 2H), and 3.25 (s, 3H,
CH3); CIMS (NH3) m/z (%): 367 (60), 385 (15), 421 (100, MH+),
and 438 (1, M + NH4+).
Com p ou n d 7. HPLC fractions coming from partial purifica-
tion of 2 and 3 were treated with 50 mM HCl for 45 min at 37
°C in the presence of 10 mM mercaptoethanol. HPLC analysis
(gradient A or B) showed formation of 7 (70% yield) as well as
little amounts of 6 (10% yield). Purification of 7 by HPLC
(gradient A) gave 0.3 mg of 7. CIMS (NH3) m/z (%): 172 (100),
331 (85), 348 (40), 391 (10), 407 (35), 424 (5), and 485 (5, MH+).
Methyl ester of 7: 1H NMR (250 MHz, CD2Cl2) δ ) 7.31 (d, 1.5
Hz, 1H, H2), 7.22 (d, 9 Hz, 1H, H2′), 6.80 (d, 9 Hz, 1H, H3′), 6.49
(d, 10 Hz, 1H, H5), 6.24 (dd, 10 and 1.5 Hz, 1H, H4), 4.79 (s,
2H), 3.80 (7H, H2′′ + CH3), 2.96 (t, 6 Hz, 2H), and 2.86 (t, 6 Hz,
2H); CIMS (NH3) m/z (%): 345 (100), 362(5), 389 (25), 421 (10),
and 499 (20, MH+).
Oxid a tion of 1 by Liver Micr osom es in th e P r es-
en ce or Absen ce of Mer ca p toeth a n ol. Oxidation of
compound 1 (14C-labeled on the keto group) with liver
microsomes from rats pretreated with clofibrate (as an
inducer (29)) in the presence of NADPH and O2, the usual
cofactors of cytochrome P450-dependent monooxygena-
ses, led to soluble metabolites (57% of starting 1 under
the used conditions) and to metabolites covalently bound
to microsomal proteins (21% of starting 1). The HPLC
chromatogram of the soluble fraction (Figure 2A) re-
vealed the formation of many metabolites less hydropho-
bic than 1, none of which represented more than 5% of
the total metabolism. Under identical conditions but in
the presence of 100 µM mercaptoethanol in the incuba-
tion, the total transformation of 1 increased to 93%, with
a marked decrease in the covalent binding of metabolites
of 1 to protein (down to 6%) and a dramatic change in
the HPLC chromatogram profile (Figure 2B). Three
adducts, called metabolites 2, 3, and 4, clearly appeared.
The ratio of these metabolites was greatly dependent on
the starting concentration of mercaptoethanol in the
incubation medium. Under identical incubation condi-

1406 Chem. Res. Toxicol., Vol. 9, No. 8, 1996
Valadon et al.
F igu r e 4. UV spectra of metabolite 3 at pH 7 and 10, and of
metabolite 6 (in H2O).
The aforementioned results suggested that microsomal
oxidation of 1 produced a very electrophilic intermediate
able to react with various nucleophiles present in the
medium leading to an intense covalent binding to pro-
teins and the formation of many reaction products in
small amounts. Trapping of this intermediate with
mercaptoethanol leads to initial metabolites 2 and 3
which are further transformed into metabolites 4 and 5
in the presence of mercaptoethanol in excess. The
following section will focus on the characterization and
chemical properties of these metabolites.
F igu r e 2. HPLC chromatograms from incubations of rat liver
microsomes with 70 µM 1 and 1 mM NADPH in the presence
(B) or absence (A) of mercaptoethanol (100 µM) for 20 min. Liver
microsomes from clofibrate-preteated rats contained 1 µM
cytochrome P450 (1.6 nmol of P450/mg of protein). (s) UV
detection at 310 or 280 nm; (-o-): radiochromatogram.
Isola tion a n d Ch a r a cter iza tion of Com p ou n d s 2
a n d 3. In agreement with the data of Figure 3, prepara-
tive incubations of 1 with rat liver microsomes (contain-
ing 2 µM cytochrome P450) in the presence of 1 mM
NADPH and 100 µM mercaptoethanol mainly gave
compounds 2 and 3. Both compounds were isolated by
preparative HPLC. The major metabolite 3 exhibited a
UV band at 287 nm (ꢀ ) 7200 M-1‚cm-1) in H2O at
neutral pH. At basic pHs (8.5-13), its UV spectrum was
characterized by an intense red-shifted band at 345 nm
(ꢀ ) 21 200 M-1‚cm-1) (Figure 4). The species absorbing
at 287 and 345 nm were in equilibrium as a function of
pH. 1H NMR of 3 in D2O at pH 4 (CD3COOD) exhibited
a set of signals expected for the protons of the aroyl
moiety of 1 and the protons corresponding to a S(CH2)2-
OH residue, and four additional signals: a singlet at 5.44
ppm and three signals at 7.02, 4.57, and 3.83 ppm
corresponding to a CHCHaHb system (Table 1). The two
geminal protons absorbing at 3.83 and 4.57 ppm could
be exchanged with D2O at different rates. The latter was
rapidly exchanged already at pH above 3 whereas the
former was only exchanged at pH above 7. The irrevers-
ible transformation of 3 into 6 under acidic conditions
(0.5-1 M HCl) was of great help for the determination
of its structure. Actually, the structure of 6, a very stable
compound, could be easily deduced from its 1H NMR
spectrum, which exhibited the signals characteristic for
the protons of the aroyl moiety of 1 and an S(CH2)2OH
group, and two doublets at 7.12 and 7.30 ppm with a
coupling constant of 5.8 Hz as expected for two vicinal
thiophene protons (30) (Table 1), and its mass spectrum
displaying a molecular peak at 407 (M + H)+. These data
clearly showed that 6 was formally derived from 1 by
introduction of a mercaptoethanol residue at position 2
of its thiophene ring (Figure 5). The UV spectrum of 6
(Figure 4) with an intense absorption at 358 nm (ꢀ ) 8200
M-1‚cm-1), which was found in thiophene derivatives
such as 1, was in good agreement with the presence of a
thiophene ring in 6.
F igu r e 3. Proportions of microsomal metabolites of 1 as a
function of mercaptoethanol concentration used in the incubate.
Conditions as in Figure 2 except for mercaptoethanol concentra-
tion used in the incubate. CB: level of covalent binding to
microsomal proteins. Yields are given in % relative to starting
1.
tions (NADPH and protein concentrations, and incuba-
tion time), conversion of 1 slightly decreased when the
starting concentration of mercaptoethanol increased from
1 to 20 mM (Figure 3). The covalent binding of metabo-
lites of 1 to microsomal proteins greatly decreased in the
presence of mercaptoethanol as a trapping agent in the
incubation medium. It was almost completely sup-
pressed in the presence of mercaptoethanol concentra-
tions higher than 1 mM. Metabolites 2 and 3 appeared
when small concentrations of mercaptoethanol were used
(Figure 3). They reached a maximum level for a starting
mercaptoethanol concentration around 100 µM. At 1 mM
mercaptoethanol, they were almost absent while metabo-
lite 4 now represented 57% of the total metabolism of 1.
At higher concentrations of mercaptoethanol, the amounts
of 4 decreased while a new metabolite 5 began to appear.
It was almost the only metabolite of 1 detected in the
presence of 20 mM mercaptoethanol.

Thiophene Sulfoxides as Reactive Metabolites
Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1407
Ta ble 1. 1H NMRa of Meta bolites 2, 3, a n d 6
2
3
6
H2
H4
H5
5.67 (s, 1H)
5.44 (s, 1H)
6.74 (broadened, 1H)
under multiplet of H1′′
4.28 (dd, 3.4-18, 1H)
7.42 (d, 8.8, 1H)
6.95 (d, 8.8, 1H)
4.66 (s, 2H)
7.02 (d, 3.4, 1H)b
3.83 (dd, 20-3.4, 1H)d
4.57 (dd, 20-2, 1H)d
7.45 (d, 8.8, 1H)
6.96 (d, 8.8, 1H)
4.66 (s, 2H)
7.12 (d, 5.8, 1H)
7.3 (d, 5.8, 1H)
c
H2′
H3′
OCH2
H1′′
H2′′
7.37 (d, 8.8, 1H)
6.97 (d, 8.8, 1H)
4.66 (s, 2H)
3.93 (t, 6, 2H)
3.33 (t, 6, 2H)
3.78 (t, 6, 2H)
2.89 (m, 2H)
3.89 (t, 6, 2H)
3.02 (m, 2H)
a
+
+
Metabolites 2 and 3 as NH4 salt in D2O/CD3COOD, pH 4.0 (400 MHz), or D2O at pH 6.5 (250 MHz); 6 as NH4 salt in D2O; δ in
b
ppm/Si(CH3)4; J in Hz. Doublet because of partial exchange of the H absorbing at 4.57 ppm at pH >3, but quadruplet (dd, J ) 3.7 and
1.5) in CD3CN-D2O-DCl mixtures in which no exchange of H5 occurred. c Irradiation at 3.8 ppm led to the collapse of the signal at 4.28
d
ppm to a doublet (J ) 3.4). Slow exchange with D2O of H absorbing at 3.83 and fast exchange of H at 4.57 at pH 6.5.
F igu r e 5. Main products formed upon metabolic oxidation of 1 in the presence of rat liver microsomes, NADPH, and mercaptoethanol.
2 and 3 are diastereoisomers. Compound 6 is not formed under these conditions; it was obtained afterward upon acidic treatment
of 2 and 3.
Metabolite 2 showed characteristics very similar to
those previously described for 3: (i) both compounds
underwent an irreversible transformation to 6 in almost
quantitative yields under acidic treatment; (ii) both
compounds exhibited almost identical UV spectra at
neutral (λ ∼290 nm) and basic (λ ∼345 nm) pH, with a
reversible passage from the neutral to the basic form as
a function of pH; (iii) the 1H NMR spectrum of 2 was very
similar to that of 3, with the same kind of signals
(identical multiplicity and integration, slightly different
chemical shifts for some of them, as shown in Table 1).
All these data are in agreement with a 2,5-dihy-
drothiophene sulfoxide structure substituted at position
2 by a S(CH2)2OH group, for compounds 2 and 3 which
are diastereoisomers with a different relative configura-
tion of their O and S(CH2)2OH substituents (Figure 5).
The presence of a band at 1030 cm-1 in the IR spectra of
2 and 3 could be assigned to the presence of a sulfoxide
function in these compounds. Dehydratation of 2 and 3
with formation of 6 under acidic conditions (pH <1) is
well-known for dihydrothiophene sulfoxides (31, 32); it
could explain the mass spectra (CI, NH3) of 2 and 3 which
both exhibit a peak at 407 corresponding to [MH - H2O]+
with the expected isotopic cluster (for the presence of two
Cl) and fragments identical with those of the mass
spectrum of 6. The structures indicated for 2 and 3 also
explain why the two geminal protons detected by 1H
NMR spectroscopy are easily exchanged in D2O; their
acidic character is due to their position R to two strong
electron-withdrawing groups (SO and CHdCHCOAr).
Finally, the presence of these acidic protons allows one
to understand the pH-dependent UV spectra of 2 and 3
which are those of a conjugated ketone at neutral pH and
of a highly conjugated anion at basic pH.
Rea ctivity of Com p ou n d s 2 a n d 3 w ith Mer ca p -
toeth a n ol. The data described in the preceding section
showed that compounds 2 and 3 were reactive toward
an excess of mercaptoethanol (Figure 3) as well as toward
strong acids (formation of 6). They could be isolated only
after incubation of 1 with microsomes in the presence of
low concentrations of mercaptoethanol and careful sepa-
ration using HPLC at neutral pH. Upon reaction of
freshly purified 2 and 3 with mercaptoethanol at 37 °C,
it was found that the nature of the products was greatly
dependent on the buffer pH.
(A) Rea ction of 2 a n d 3 a t p H betw een 4.5 a n d
6.5 a n d Isola tion of Com p ou n d 4. In this pH range,
2 and 3 were quantitatively transformed into compound

1408 Chem. Res. Toxicol., Vol. 9, No. 8, 1996
Valadon et al.
Ta ble 2. 1H NMRa of 1 a n d Meta bolites 4, 5, a n d 7 (a n d Th eir Deu ter ia ted Equ iva len ts Obta in ed fr om 2 Bisd eu ter ia ted
a t P osition 5)
1b
4
5
7
H2
H4
H5
8.12 (t, 1H)b
7.56b
7.73 (s, 1H)
7.71 (s, 1H)
7.70 (s, 1H)
4.42 (dd, 10-6, 1H) [s]c
3.31 (dd, 14-10, 1H) [D]c
3.53 (dd, 14-6, 1H) [D]c
7.30 (d, 8.8, 1H)
6.96 (d, 8.8, 1H)
4.66 (s, 2H)
4.80 (dd, 8.5-1.5, 1H)
3.58 (dd, 13-1.5, 1H) [D]c
4.03 (dd, 13-8.5, 1H) [D]c
7.29 (d, 8.5, 1H)
6.96 (d, 8.5, 1H)
4.66 (s, 2H)
6.29 (d, 10, 1H) [s]c
6.65 (d, 10, 1H) [D]c
7.56b
H2′
H3′
7.44 (d, 9, 1H)
6.97 (d, 9, 1H)
4.67 (s, 2H)
7.34 (d, 8.8, 1H)
6.94 (d, 8.8, 1H)
4.65 (s, 2H)
3.73 (t, 6, 2H)
3.01 (t, 6, 2H)
3.79 (t, 6, 2H)
2.89 (t, 6, 2H)
OCH2
H1′′a
H2′′a
H1′′b
H2′′b
H1′′c
H2′′c
3.73 (t, 6, 2H)
3.00 (t, 6, 2H)
3.82 (m, 2H)
2.87 (m, 2H)
3.87 (t, 6, 2H)
2.94 (t, 6, 2H)
3.82 (t, 6, 2H)
2.91 (m, 6, 2H)
a
b
NH4+ salt in D2O (250 MHz, and 400 MHz for 4); δ in ppm/Si(CH3)4; J in Hz. In D2O, as NH4 salt H2 appears as a collapsed triplet
(J ≈ 3 Hz), H4 and H5 are not separated; in DMSO-d6 all signals are separated: 8.09 (dd, 3-1.5, 1H, H2) [d, 1.5 Hz],c 7.68 (dd, 5-3, 1H,
H5) [D],c 7.46 (dd, 5-1.5, 1H, H4) [d, 1.5 Hz],c 7.38 (d, 8.5, 1H, H2′), 6.92 (d, 8.5, 1H, H3′), and 4.39 (s, 2H, OCH2). c Data in brackets
correspond to deuteriated compounds obtained from 2 (or 3) bisdeuteriated at position 5 (see text); [D] indicates the presence of a deuterium
at the corresponding position.
F igu r e 6. 1H NMR spectra of compounds 4, 5, and 7 (D2O, 250 MHz). The 4-6 ppm region involving the signals of OCH2 and
(HOD)1 has been omitted for the sake of clarity. Labeling of protons are those shown in Table 2.
4. The reaction rate increased upon increasing the pH
and/or mercaptoethanol concentration. For instance, at
pH 4.6 and in the presence of 5 mM mercaptoethanol,
formation of 4 was almost complete after 20 min whereas
only 10% 4 was obtained under identical conditions in
the presence of 0.5 mM mercaptoethanol. With this
mercaptoethanol concentration but at pH 6.5, the forma-
tion of 4 was again quantitative in less than 20 min.
grating for one proton at 3.31 (J ) 10 and 14 Hz), 3.53
(J ) 6 and 14 Hz), and 4.42 ppm (J ) 6 and 10 Hz) (Table
2 and Figure 6). The intense UV absorption of 4 (λ )
317 nm, ꢀ ) 20 800 M-1‚cm-1) and the presence of a low-
field singlet at ca. 7.7 ppm that could be assigned to a
vinylic proton indicated the presence of a double bond
conjugated with the COAr group. The opened structure
shown for compound 4 in Figure 5 and Table 2 is in
agreement with its mass spectrum (m/z ) 563 for MH+)
and a detailed analysis of its 1H NMR spectrum. The
presence of the ArCOCdCHS(CH2)2OH moiety was in-
dicated by an intense nuclear Overhauser effect (NOE)
between the vinylic proton H2 and the aromatic proton
1
Compound 4 was isolated by preparative HPLC. Its H
NMR spectrum in D2O (or in CD2Cl2 for its methyl ester,
see Experimental Section) revealed a set of signals
corresponding to the intact aroyl moiety of 1, the signals
of three different S(CH2)2OH groups that were clearly
separated at 400 MHz, a singlet at low field (1H, 7.73
ppm), and a system of three coupled signals each inte-
4
H2′ in ortho position to the CO group and a J coupling
between H2 and the proximal protons H2′′ of a mercap-

Thiophene Sulfoxides as Reactive Metabolites
Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1409
Ta ble 3. 13C NMRa of 1 a n d Com p ou n d 5
1 (DMSO-d6)
1 (D2O)
5 (D2O)
C2
C3
C4
C5
137.35 (CH)b
141.3
128.42 (CH)
127.18 (CH)
187.02
141.92 (CH)
143.52
161.33 (CH)
139.85
51.37 (CH)
45.32 (CH2)
191.53
F igu r e 7. Summary of the various reactions observed upon
treatment of 2 and 3 with mercaptoethanol as a function of pH.
Formations of 1 from 5 and 6 from 2 (or 3) occur without
mercaptoethanol.
131.16 (CH)d
129.92 (CH)d
186.37
C6
c
C1′
C2′
C3′
C4′
C5′
C6′
C7′
C8′
131.62
134.5
134.35
127.71 (CH)
111.95 (CH)
157.06
121.09
129.41
130.86 (CH)
113.83 (CH)
159.29
123
133.36
129.80 (CH)
113.65 (CH)
158.44
124.50
132.42
70.15 (CH2)
177.70
63.32
OH residue for compound 5 (Figure 5). This structure,
which formally derives from the addition of mercapto-
ethanol to one double bond of the thiophene ring of 1,
was further confirmed by the mass spectrum of 5 exhibit-
ing a MH+ ion at m/z ) 409, and by the transformation
of 5 into 1 in the presence of NaOH (pH ) 14). This
reaction corresponds to a base-catalyzed loss of mercap-
toethanol from 5.
68.31 (CH2)
169.26
70.53 (CH2)
178.03
C1′′
C2′′
35.10
a
+
1 (10 mg) and 5 (7 mg) as NH4 salts in D2O or DMSO-d6
(C) Rea ction of 2 a n d 3 w ith Mer ca p toeth a n ol a t
p H ca . 1. Reaction of compounds 2 and 3 with 10 mM
mercaptoethanol in the presence of 0.05 M HCl at 37 °C
led after 45 min to minor amounts of 5 and 6 (∼10% and
2%) and to a major new product 7 (70% yield). Compound
b
(250 MHz); δ in ppm/Si(CH3)4, J in Hz. Multiplicity determined
by DEPT experiments (41). c Signals of the phenyl ring were
d
assigned by selective INEPT experiments (42). Assignment of C4
and C5 signals was only done in DMSO-d6 (see text).
1
7 was isolated by HPLC. Its H NMR spectrum (D2O)
toethanol residue (see Table 2). The two geminal protons
were assigned by their negative coupling constant (J )
-14 Hz) determined from a correlation spectroscopy
(COSY) 45 experiment (H5, see Table 2). The presence
exhibited, in addition to the signals corresponding to the
protons of the aroyl moiety, a singlet at 7.7 ppm (1H), 2
coupled doublets at 6.29 and 6.65 ppm (2 × 1H), and the
signals corresponding to two nonequivalent S(CH2)2OH
groups (Table 2 and Figure 6). These NMR data and the
mass spectrum of 7 exhibiting a peak MH+ at m/z ) 458
are in agreement with the opened structure shown in
Table 2. The coupling constant of 10 Hz between the
protons absorbing at 6.29 and 6.65 ppm indicated their
cis relative position (33). The signal of H4 was assigned
1
of three S(CH2)2OH groups in 4 was shown by H NMR
and mass spectroscopy. Moreover, the presence of a
dithioether function was suggested by the easy transfor-
mation of 4 into 5 in the presence of mercaptoethanol (1
mM, 30 min at 37 °C and pH 7.4, vide infra), which could
be replaced with either dithiothreitol or SnCl2.
At that point, the opened structure shown in Figure 5
for 4 appeared likely and in agreement with the spec-
troscopic data; however, it was not possible to exclude
another structure obtained by permutation of the S(CH2)2-
OH and SS(CH2)2OH substituents on positions 4 and 5.
(B) Rea ction of 2 a n d 3 a t Ba sic p H a n d Isola tion
of Com p ou n d 5. Reaction of pure compounds 2 and 3
with 0.5 mM mercaptoethanol in phosphate buffer (pH
7.4) led to a 90:10 mixture of 4 and 5. However, at pH
12 and in the presence of 5 mM mercaptoethanol, 2 and
3 were quantitatively transformed into compound 5 in
less than 3 min. This reaction allowed us to prepare 5
3
by the presence of a J coupling (1.5 Hz) with H2 (NMR
study on the methyl ester of 7), which also confirmed that
position 4 of 7 was substituted by a proton. The presence
of a dithioether function was indicated by the easy
transformation of 7 into 1 (with loss of two S(CH2)2OH
groups) in the presence of low concentrations of mercap-
toethanol or dithiothreitol. The dithioether function
could have been located either on carbon 2 or on carbon
5. However, the almost identical chemical shifts of H2,
of the aroyl group protons, and of the protons of one
S(CH2)2OH group in 7 and 4 argued in favor of the
dithioether function at position 5 rather than at position
2.
1
which was purified by HPLC and characterized by H
and 13C NMR, UV, and mass spectroscopy. Its H NMR
1
Com plem en tar y Str u ctu r e Deter m in ation by Deu -
ter ia tion Exp er im en ts. As shown in Figure 5, meta-
bolic oxidation of 1 with liver microsomes in the presence
of mercaptoethanol leads to compounds 2 and 3 as
primary metabolites. Further reaction of 2 and 3 with
an excess of mercaptoethanol leads successively to the
opened compound 4 then to the dihydrothiophene 5.
Interestingly, treatment of 5 with a strong base gives
back to 1. Moreover, treatment of 2 and 3 at pH ca. 1,
in the presence or in the absence of mercaptoethanol,
respectively, leads to two other products 7 and 6 (Figure
7).
The spectroscopic data described above allowed us to
establish the structures indicated in Figure 5 for com-
pounds 2, 3, and 6, without any ambiguity. In the case
of compounds 4 and 5, the available data did not allow
us to exclude structures with an S(CH2)2OH group at
spectrum (in D2O) exhibited the set of signals corre-
sponding to the aroyl moiety of 1, a vinyl singlet at 7.71
ppm (showing an intense (NOE)1 with the H2′ of the aroyl
group) integrating for 1H, a system of three coupled
signals at 3.58 (1H, dd, 13 and 1.5 Hz), 4.03 (1H, dd, 13
and 8.5 Hz), and 4.80 ppm (1H, dd, 8.5 and 1.5 Hz), and
the signals expected for one S(CH2)2OH group (Table 2
and Figure 6). Comparison of the 13 C NMR spectrum of
5 with that of 1 showed almost identical signals for the
carbons of the aroyl group. Supplementary signals at
35.1 and 63.32 ppm corresponding to the two carbons of
the S(CH2)2OH group were found in the spectrum of 5.
This spectrum also exhibited signals for the carbons
derived from the thiophene ring of 1, as a quaternary
carbon, 2 CH carbons, and a CH2 carbon (Table 3). All
these data were in agreement with a 4,5-dihydrothiophene
structure substituted at position 4 (or 5) by an S(CH2)2-

1410 Chem. Res. Toxicol., Vol. 9, No. 8, 1996
Valadon et al.
position 5 instead of position 4, as indicated in Figure 5.
Therefore, supplementary experiments have been per-
formed on compound 3 deuteriated at position 5 of its
dihydrothiophene ring. Exchange of the two hydrogens
H5 of 3 was followed by 1H NMR in D2O containing
NaOD. After complete disappearance of the H5 signals,
deuteriated 3 was isolated by HPLC and treated with
50 mM mercaptoethanol at pH 4.5 (in D2O containing
H2O and neutralization of the medium. This rich chem-
istry of 2 and 3 may be interpreted if one takes into
account two starting characteristics of these com-
pounds: (i) their ability to act as Michael acceptors for
nucleophiles because of the presence of their conjugated
double bond, and (ii) the acidity of their H5 protons whose
abstraction leads to a stabilized anion. Interestingly,
both situations should lead to the appearance of an
intermediate conjugated anion located in part on carbon
3. These considerations have helped us to propose
mechanisms for the formation of 4, 5, and 7.
P ossible Mech a n ism for th e F or m a tion of 4 a n d
5 fr om 2 a n d 3 (F igu r e 8). Nucleophilic addition of the
thiol group of mercaptoethanol on the conjugated double
bond of 2 (or 3) should lead to an intermediate anion at
C3, which may evolve by cleavage of the C2-S bond in
the â-position. After protonation, the sulfenic acid B
derived from that reaction may either recyclize to give
back to 2 or react with mercaptoethanol to give 4. The
latter reaction is likely since sulfenates are known to
rapidly react with thiols in excess to give dithioethers
(35). In fact, the formation of a dithioether function
should drive the reaction toward 4. At neutral and basic
pH, the SH group of mercaptoethanol should reduce the
dithioether function of 4, leading, as shown in Figure 8,
to an S- (or SH) intermediate able to add to the
conjugated double bond in an intramolecular manner.
The resulting cyclic anion should then undergo a â-elim-
ination of a -S(CH2)2OH residue eventually leading to 5.
Transformation of 5 into 1 in the presence of a base
simply results from a base-catalyzed 1,2-elimination of
HSCH2CH2OH.
P ossible Mech a n ism s for th e In ter con ver sion of
2 a n d 3 a n d F or m a tion of 7. Interconversion between
2 and 3, which was observed after treatment of these
compounds with NaOH followed by a quick neutralization
of the medium, should not be due to a change of the
sulfoxide stereochemistry, as it is known that H/D
exchanges in position R to dialkyl sulfoxides occur with
complete retention of the stereochemistry of the sulfoxide
function (36-38). Thus, interconversion between 2 and
3 requires a bond to be broken. A likely mechanism for
this reaction is depicted in Figure 9. It also involves the
intermediate formation of an anion, similar to that
proposed in Figure 8. Here this anion is not formed by
addition of a nucleophile to the conjugated double bond
of 2 (or 3) (as in Figure 8), but by simple abstraction of
a proton at C5. Anion C (Figure 9) also could undergo a
cleavage of the C-S bond in â-position leading to the
opened sulfenate anion D. Recyclization of D by in-
tramolecular Michael-type addition of -SO- to the con-
jugated double bond should lead back to the starting
anion C or to its isomer E. Upon neutralization of the
medium, protonation of anions C, D, and E, that are in
equilibrium, should lead to a mixture of 2 and 3 as major
products and to a very minor formation of the opened
isomer DH which has never been detected. The final 2:3
ratio was found to mainly depend on the pH of the
“neutralization” medium, but was independent on the
nature of the starting product (2 or 3). It was maximum
(0.2) at pH 7 and lower at higher pH (0.01 at pH 13, 0.05
at pH 10, and 0.16 at pH 8). Such openings of thiophene
rings have been proposed as intermediate steps before
sulfenate methylation (39).
1
DCl) to give 4. The H NMR spectrum of deuteriated 4
obtained by this method showed an almost complete loss
of the signals assigned to H5, whereas the signal assigned
to H4 remained and now appeared as a singlet. Deute-
riated 4 was then transformed into 5 in H2O in the
1
presence of excess mercaptoethanol. The H NMR spec-
trum of deuteriated 5 also showed an almost complete
disappearance of the signals at 3.58 and 4.03 ppm
assigned to H5. Finally, aromatization of deuteriated 5
into 1 in H2O containing NaOH required heating at 70
°C to get to completion. This more difficult reaction with
deuteriated 5 than with 5 itself would be in agreement
with the removal of a deuteron (instead of a proton) at
position 5 as a first, rate limiting step of elimination of
mercaptoethanol from 5. The 1H NMR spectrum of
deuteriated 1 in DMSO-d6 (a solvent in which all the
thiophene proton signals are separated) showed the
almost complete disappearance of one thiophene signal
(at 7.68 ppm).
1
Assignment of the H NMR signals of the thiophene
protons of nondeuteriated 1 was done as follows. The
H2 signal was identified because of its downfield shift at
8.09 ppm (in DMSO) and the existence of W couplings
with the two other thiophene signals. After irradiation
of the H3′ signal, an NOE transferred to the H2′ signal
and the signal at 7.48 ppm at 70 °C. This signal was
thus assigned to H4 as H5 would not have been suf-
ficiently close to H2′ in space for such an NOE effect. In
fact, assignment of the signals of H5 and H2 was further
confirmed by experiments in which 1 was treated for 24
h at 105 °C with CF3CF2CF2COOD. Such conditions
were known to lead to a partial H/D exchange at positions
2 and 5 of 3-aroylthiophenes (34). Accordingly, this
treatment led to a partial disappearance of the signals
at 8.1 (90%) and 7.68 (50%) ppm (spectrum in DMSO).
These results allowed the assignment of the thiophene
signals of 1 indicated in Table 2. Thus, the signal (at
7.68 ppm) that has almost completely disappeared in
deuteriated 1 prepared from deuteriated 3 was that of
H5. These data clearly showed that the S(CH2)2OH
moiety present in compound 5 is located on carbon 4 and
confirmed the structures shown in Figure 5 for com-
pounds 4 and 5.
Mech a n ism s for th e F or m a tion of th e Va r iou s
Meta bolites of 1. The structure found for compounds
2 and 3 strongly suggests that they derive from a
Michael-type addition of mercaptoethanol to the most
reactive position (C2) of the thiophene sulfoxide metabo-
lite A formed by microsomal monooxygenation of 1
(Figure 5). Thiophene sulfoxide derived from 1 should
be unstable as most thiophene sulfoxides (15), the pres-
ence of a COAr substituent making it especially electro-
philic in nature. Its adducts with mercaptoethanol, 2 and
3, could be isolated under selected conditions since they
appeared to be unstable in the presence of acids (forma-
tion of 6) or thiols such as mercaptoethanol in excess.
Several compounds are formed upon reaction of 2 and 3
with mercaptoethanol as a function of pH (Figure 7).
Moreover, it was also observed that 2 and 3 are inter-
converted into each other after treatment by NaOH in
Formation of 7, which was observed upon treatment
of 2 (or 3) with mercaptoethanol at acidic pH, could be
due to a trapping of the opened isomer DH by mercap-
toethanol after protonation of its OH group, loss of H2O,

Thiophene Sulfoxides as Reactive Metabolites
Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1411
F igu r e 8. Possible mechanisms for the formation of 4 and 5 from reaction of 2 and 3 with mercaptoethanol. Anionic intermediates
are indicated in order to illustrate similarities of mechanisms in Figures 8 and 9 (similar openings of the thiophene ring in the case
of an anionic intermediate at C-3, for instance). However, there is no direct evidence for such intermediates; concerted mechanisms
are as likely.
F igu r e 9. Possible mechanism for the interconversion between 2 and 3 after basic treatment, and for the formation of 7 upon
reaction of 2 and 3 with mercaptoethanol at pH 1. Absolute and relative stereochemistry of the sulfoxide and C-2 is presently unknown
for compounds 2 and 3, and for the derived intermediates C and E. Formulas used for 2, 3, C, and E only indicate that 2 and 3 (as
well as C and E) have different relative stereochemistry.
and formation of a dithioether function. Formation of 1
upon treatment of 7 with mercaptoethanol should involve
a mechanism very similar to that proposed for reaction
of 4 with mercaptoethanol that leads to 5 (Figure 8): (i)
nucleophilic attack of the S-S bond of 7 by the S atom
of mercaptoethanol, (ii) intramolecular addition of the S
center to the conjugated double bond, and (iii) elimination
of HSCH2CH2OH.
in Vivo. A complete view of the microsomal metabolism
of 1 in the presence of mercaptoethanol is shown in
Figure 5. In microsomal incubations performed under
usual conditions (pH 7.4) and in the presence of mercap-
toethanol in large excess (10-20 mM), almost only one
product is formed (with an overall yield based on 1 up to
70%). This product, 5, is the final stable metabolite
resulting from the cascade of reactions occurring on 2 and
3 (Figure 5). Under in vivo conditions, the thiol-contain-
ing nucleophile that is present in large amounts in many
Micr osom a l Meta bolism of 1 in th e P r esen ce of a
Th iol in Excess: Rela tion sh ip w ith Its Meta bolism

1412 Chem. Res. Toxicol., Vol. 9, No. 8, 1996
Valadon et al.
Ta ble 4. Com p a r ison of th e 1H NMR Ch a r a cter istics of
th e Meth yl Ester s of Com p ou n d s 8a , 8b, a n d 5a
8a
8b
5
H2
H4
7.12 (s, 1H)
4.69 (dd,
7.17 (s, 1H)
4.70 (dd,
7.14 (s, 1H)
4.81 (dd,
8-1.5, 1H)
3.94 (dd,
8-1, 1H)
3.95 (dd,
8-1.5, 1H)
3.98 (dd,
H5
13-8, 1H)
3.45 (dd,
13-1.5, 1H)
7.18 (d, 9, 1H)
6.80 (d, 9, 1H)
4.78 (s, 2H)
13-8, 1H)
12-8, 1H)
3.50 (dd,
12-1.5, 1H)
7.20 (d, 9, 1H)
6.80 (d, 9, 1H)
4.78 (s, 2H)
3.48 (dd,
F igu r e 10. Formation of 8 as a metabolite of 1 in vivo in rats,
and in vitro after incubation with rat liver microsomes, NADPH,
and N-acetylcysteine. GSH is used for glutathione and Ar for
4-(2,3-dichlorophenoxy)-2-acetic acid. Compound 8 is a mixture
of two diastereoisomers, 8a and 8b.
13-1, 1H)
7.21 (d, 9, 1H)
6.81 (d, 9, 1H)
4.79 (s, 2H)
H2′
H3′
OCH2
a
Methyl ester of 5 and dimethyl esters of 8a and 8b in CD2Cl2
(250 MHz); δ in ppm/Si(CH3)4; J in Hz. The signals of the
thioether substituent at position 4 and of COOCH3 groups are not
indicated; they are given in the Experimental Section. Labeling
of protons as for 5 in Table 2.
cells is glutathione. In the hepatocytes, which are greatly
involved in the metabolism of xenobiotics, the glutathione
concentration is high (1-5 mM). It was thus tempting
to propose that the in vivo metabolism of 1 in rats led to
an equivalent of 5 in which the mercaptoethanol residue
was replaced with a glutathione residue. It is well-known
that such glutathione adducts are very often further
transformed into the corresponding mercapturates (40).
Thus, formation of a mercapturate corresponding to 5 in
which the mercaptoethanol group is replaced with an
N-acetylcysteine group (8 in Figure 10) appeared very
likely, as a potential in vivo metabolite of 1. In order to
verify this proposition based on the scheme established
for the microsomal metabolism of 1 in the presence of a
thiol-containing nucleophile, two series of experiments
were performed. In the first one, rats were treated with
1 and their urinary metabolites were studied after
separation by HPLC. In the second one, the expected
mercapturate 8 was prepared from reaction of 1 with rat
liver microsomes in the presence of NADPH, O2, and an
excess of N-acetylcysteine under conditions similar to
those used to prepare 5.
esters). They showed HPLC retention times, UV, 1H
NMR, and mass spectra identical to those previously
found for dimethyl esters of 8a and 8b. They represented
1.3% and 0.8% of administered 1, and 10% and 5% of
the total radioactivity found in the urine. The fourth
isolated compound (1% of starting 1) showed a UV
1
spectrum identical to those of 8a and 8b. Its H NMR
spectrum was very similar to that of 8a and 8b except
for the absence of the COCH3 singlet. We did not try to
completely identify this compound; however, its UV and
1H NMR characteristics suggested that it could be 8a (or
8b) with a free NH2 function (instead of NHCOCH3).
These results show that the relatively complex series
of reactions occurring after S-oxidation of 1 and trapping
of 1 sulfoxide with a thiol, that was established in vitro
using liver microsomes and mercaptoethanol, also occurs
in vivo in rats. The final stable metabolite is a 3-aroyl,
4-S-alkyl 4,5-dihydrothiophene both in vitro and in vivo
(5 and 8, respectively). Thus, it is very likely that
compounds 8 detected in the urine of rats treated with 1
are derived from (i) a monooxygenase-dependent S-
oxidation of 1, (ii) a nucleophilic addition of glutathione
at position 2 of 1 sulfoxide, (iii) a transformation of the
resulting equivalent of 2 into 8 bearing a glutathione
residue instead of an N-acetylcysteinyl residue, and (iv)
the usual processing of glutathione adducts (40) into
corresponding mercapturates 8.
(A) P r ep a r a tion of 8 u p on Micr osom a l Tr a n sfor -
m a tion of 1 in th e P r esen ce of N-Acetylcystein e. As
expected, this reaction led to two major products which
were isolated by preparative HPLC and whose UV
spectra were almost identical to those of 4 and 5,
respectively. Compound 8 which exhibited a UV spec-
trum identical to that of 5 (λ ) 334 nm) was in fact a
mixture of two diastereoisomers, 8a and 8b, that were
separated by HPLC after methylation of their carboxylic
acid functions with CH2N2. Their 1H NMR and mass
spectra were in complete agreement with a structure
derived from that of 5 upon replacement of its mercap-
toethanol substituent with an N-acetylcysteinyl group
(under the form of its methyl ester). Their mass spec-
trum (CI, NH3) exhibited peaks at m/z ) 522 (M + H)+
Con clu sion
The aforementioned results provide a clear indirect
evidence for the intermediate formation of a reactive,
electrophilic thiophene sulfoxide in metabolic oxidation
of 1 in vitro and in vivo. They also indicate how to trap
such intermediates using low concentrations of a thiol
trapping agent and a careful control of pH during
isolation of substituted 2,5-dihydrothiophene sulfoxides
such as 2. Moreover, they give a first idea of the various
reactions that may occur when such 3-aroylthiophene
sulfoxides react with a thiol in excess (Figure 5). This
allowed us to explain the global transformation of 1 into
the final stable metabolite 5 in vitro in the presence of
excess mercaptoethanol, and of 1 into 8 in vivo. Trans-
formation of 1 into 5 with liver microsomes and mercap-
toethanol in excess, that we had observed in the begin-
ning of this study, was very difficult to explain without
isolation of intermediates such as 2 and 4 and study of
their reactivity toward thiols.
1
and 539 (M + NH4)+, and their H NMR spectra showed
signals expected for a 3-aroyl-4,5-dihydrothiophene struc-
ture, as in 5, in addition to those corresponding to an
N-acetylcysteinyl substituent at position 4. This was
1
particularly clear when comparing the H NMR spectra
of 8a (and 8b) and 5 (Table 4).
(B) In Vivo Meta bolism of 1 in Ra ts. For the in
vivo experiments, rats were treated with radiolabeled 1
(30 mg/kg, 14C at the keto group, 0.5 mCi/mmol), and
their urine was collected during 24 h. 20% of the dose
was recovered in the urine with a maximum extracted
between 2 and 4 h. The 2-4 h fraction was studied by
HPLC; four radioactive peaks were observed. After
extraction with ethyl acetate and methylation with
CH2N2, the four compounds were studied by UV, 1H
NMR, and mass spectroscopy. One of them was un-
changed 1 (as methyl ester), whereas two of those
metabolites could be identified as 8a and 8b (dimethyl
Recent data on the oxidative metabolism of thiophene
itself and of 2-aroylthiophenes have indicated the key role
of thiophene S-oxides as reactive intermediates (12, 14).

Thiophene Sulfoxides as Reactive Metabolites
Chem. Res. Toxicol., Vol. 9, No. 8, 1996 1413
(19) Naperstkow, A. M., Macaulay, J . B., Newlands, M. J ., and Fallis,
A. G. (1989) Π-facial diastereoselectivity in Diels-Alder reactions
of 2,5-dimethylthiophene oxide. Tetrahedron Lett. 30, 5077-5080.
(20) Melles, J . L., and Backer, H. J . (1953) Oxydation de thiophe`nes
en sulfones. Proprie´te´s du groupe sulfonyle XXXV. Rec. Trav.
Chim. 72, 491-505.
The results described above provide the first data on the
reactivity of such thiophene S-oxides, whose chemistry
is poorly known (15), and on their possible fates in living
organisms.
(21) Davies, W., Gamble, N., J ames, F. C. J ., and Savige, W. (1952)
The direct conversion of thiophene into dicyclic and tetracyclic
derivatives. Chem. Ind., 804.
(22) Davies, W., and J ames, F. C. J . (1954) The conversion of thiophen
into 4:7:8:9:-tetrahydro-4:7-sulphinylthianaphthen. J . Chem. Soc.,
15-18.
(23) Kremers, P., Beaune, P., Cresteil, T., Degraeve, J ., Columelli, S.,
Leroux, J . P., and Gielen, J . E. (1981) Cytochromes P-450
monooxygenases activities in human and rat liver microsomes.
Eur. J . Biochem. 118, 599-606.
(24) Lowry, O. H., Rosebrough, N. J ., Farr, A. L., and Randall, R. J .
(1951) Protein measurement with the folin phenol reagent. J . Biol.
Chem. 193, 265-275.
Refer en ces
(1) McMurtry, R. J ., and Mitchell, J . R. (1977) Renal and hepatic
necrosis after metabolic activation of 2-substituted furans and
thiophenes. Toxicol. Appl. Pharmacol. 42, 285-291.
(2) Imler, M., Chabrier, G., Cherfran, J ., and Simon, C. (1987)
He´patite imputable au thiophe`ne. Gastroenterol. Clin. Biol. 11,
173-174.
(3) Beaune, P. H., Dansette, P. M., Mansuy, D., Kiffel, L., Finck, M.,
Amar, C., Leroux, J . P., and Homberg, J . C. (1987) Human anti-
endoplasmic reticulum autoantibodies appearing in
a drug-
induced hepatitis are directed against a human liver cytochrome
P-450 that hydroxylates the drug. Proc. Natl. Acad. Sci. U.S.A.
84, 551-555.
(25) Omura, T., and Sato, R. (1962) The carbon monoxide binding
pigment of liver microsomes. J . Biol. Chem. 237, 1375-1376.
(26) Fenselau, A. (1970) Nicotinamide dinucleotide as an active site
director in glyceraldehyde-3-phosphate dehydrogenase modifica-
tion. J . Biol. Chem. 245, 1239-1246.
(4) Hastier, P., Bianchi, D., Kuhdorf, H., Arpurt, J . P., Chichmanian,
R. M., Saintpaul, M. C., and Delmont, J . (1993) He´patite au
thiophe`ne: un nouveau cas. Gastroenterol. Clin. Biol. 17, 613-
614.
(27) Wallin, A., Schelin, C., Tunek, A., and J ergil, B. (1981) A rapid
and sensitive method for determination of covalent binding of
benzo[a]pyrene to proteins. Chem.-Biol. Interact. 38, 109-118.
(28) Dansette, P. M., Amar, C., Smith, C., Pons, C., and Mansuy, D.
(1990) Hydroxylation and formation of electrophilic metabolites
during metabolism of tienilic acid and its isomer by rat liver
microsomes. Biochem. Pharmacol. 39, 911-918.
(29) Gibson, G. G. (1992) Co-induction of cytochrome P450 4A1 and
peroxisome proliferation. Xenobiotica 22, 1101-1109.
(30) Cagniant, P., Cagniant, D., Paquer, D., and Kirsch, G. (1985) In
The Chemistry of heterocyclic compounds, thiophene and its
derivatives (Gronowitz, S., Ed.) Vol. 44, Part 2, pp 119-123, J ohn
Wiley and Sons, New York.
(31) Cava, M. P., and Pollack, M. J . (1960) A Novel Synthesis of
Isothianaphthenes. J . Am. Chem. Soc. 88, 4112-4116.
(32) Takaya, T., Kosaka, S., Otsuji, Y., and Imoto, E. (1968) Prepara-
tion and oxidation of dihydrothiophenes. Bull. Chem. Soc. J pn.
41, 2086-2095.
(33) J ackman, L. M., and Sternhell, L. S. (1969) Application of the
chemical shift. In Applications of nuclear magnetic resonance
spectroscopy in organic chemistry (Barton, D. H. R., and Doering,
W., Eds.) pp 159-266, Pergamon Press, London.
(5) Rance, D. J . (1989) Metabolism of sulphur functional groups. In
Sulphur-containing drugs and related organic compounds (Da-
mani, Ed.) Vol. 1B, pp 227-268, Ellis Horwood Limited Publish-
ers, Chichester, U.K.
(6) Galtier, P. (1989) Antihelminticssphenothiazine, thiabendazole
and related compounds. In Sulphur-containing drugs and related
organic compounds (Damani, Ed.) Vol. 3A, pp 108-141, Ellis
Horwood Limited Publishers, Chichester, U.K.
(7) Homberg, J . C., Andre´, C., and Abuaf, N. (1984) A new anti-liver-
kidney-microsome antibody (anti-LKM2) in tienilic acid-induced
hepatitis. Clin. Exp. Immunol. 55, 561-570.
(8) Beaune, P., Pessayre, D., Dansette, P., Mansuy, D., and Manns,
M. (1994) Autoantibodies against cytochrome P450: Role in
human diseases. In Advances in Pharmacology (August, J . T.,
Anders, M. W., Murad, F., and Coyle, J . T., Eds.) Vol. 30, pp 199-
245, Academic Press, San Diego, CA.
(9) Mansuy, D., Dansette, P. M., Foures, C., J aouen, M., Moinet, G.,
and Bayer, N. (1984) Metabolic hydroxylation of the thiophene
ring: isolation of 5-hydroxytienilic acid as the major urinary
metabolite of tienilic acid in man and rat. Biochem. Pharmacol.
33, 1429-1435.
(10) Dansette, P. M., Amar, C., Valadon, P., Pons, C., Beaune, P. H.,
and Mansuy D. (1991) Hydroxylation and formation of electro-
philic metabolites of tienilic acid and its isomer by human liver
microsomes. Biochem. Pharmacol. 41, 553-560.
(34) Meth-Cohn, O. (1979) Thiophens. In Comprehensive Organic
Chemistry (Barton, D., and Ollis, D. W., Eds.) Vol. 4, pp 789-
838, Pergamon Press, London.
(35) Hogg, D. R. (1979) Sulfenic acids and their derivatives. In
Comprehensive Organic Chemistry (Barton, D., and Ollis, D. W.,
Eds.), Vol. 3, pp 261-310, Pergamon Press, London.
(36) Hutchinson, B. J ., Andersen, K., and Katritsky, A. R. (1969) The
conformational analysis of satured heterocycles. XX. The stere-
ochemistry of base-catalysed hydrogen-deuterium exchange of
methylene protons R to sulfinyl group. J . Am. Chem. Soc. 91,
3839-3844.
(37) Kingsbury, C. A., and Cram, D. J . (1960) Studies in stereochem-
istry. XXXII. Mechanism of elimination of sulfoxides. J . Am.
Chem. Soc. 82, 1810-1819.
(38) Patai, S., Rappoport, Z., and Stirling, C. (1988) The chemistry of
sulphones and sulphoxides. In The chemistry of sulphones and
sulphoxides (Patai, S., Ed.) pp 459-832, J ohn Wiley and Sons,
New York.
(39) Crumbie, R. L., and Riddley, D. D. (1981) Reactions of 2,5-
dihydrothiophen and S-substitued derivatives with bases. Prepa-
rations of 1-(alkylthio)buta-1,3-dienes and the corresponding
sulfoxides and sulfones. Aust. J . Chem. 34, 1017-1026.
(40) Chassaud, L. F. (1976) Mercapturic acids. In Glutathione: me-
tabolism and function (Arias, I. M., and J akoby, W. B., Eds.) pp
77-114, Raven Press, New York.
(41) Pegg, D. T., Doddrell, D. M., and Bendall, M. R. (1982) Proton-
polarization transfer enhancement of a heteronuclear spin mul-
tiplet with preservation of phase coherency and relative compo-
nent intensities. J . Chem. Phys. 77, 2745-2752.
(11) Lopez-Garcia, M. P., Dansette, P. M., Valadon, P., Amar, C.,
Beaune, P. H., Guengerich, F. P., and Mansuy, D. (1993) Human
liver cytochromes P450 expressed in yeast as tools for reactive
metabolite formation studies. Eur. J . Biochem. 213, 223-232.
(12) Lopez-Garcia, M. P., Dansette, P. M., and Mansuy, D. (1994)
Thiophene derivatives as new mechanism-based inhibitors of
P450: Inactivation of yeast-expressed human liver P450 2C9.
Biochemistry 33, 166-175.
(13) Mansuy, D., Valadon, P., Erdelmeier, I., Lopez-Garcia, P., Amar,
C., Girault, J .-P., and Dansette, P. (1991) Thiophene S-oxide as
new reactive metabolites: Formation by cytochrome P450 de-
pendent oxidation and reaction with nucleophiles. J . Am. Chem.
Soc. 113, 7825-7826.
(14) Dansette, P. M., Thang, D. C., El Amri, H., and Mansuy, D. (1992)
Evidence for thiophene-S-oxides as primary reactive metabolite
of thiophene in vivo: formation of a dihydrothiophene sulfoxide
mercapturic acid. Biochem. Biophys. Res. Commun. 186, 1624-
1630.
(15) Raasch, M. S. (1985) Thiophene sulfoxides, sulfones and sesqui-
oxides. In The chemistry of heterocyclic compounds, thiophene and
its derivatives (Gronowitz, S., Ed.) Vol. 44, Part 1, pp 571-627,
J ohn Wiley and Sons, New York.
(16) Mock, W. L. (1970) Stable thiophene sulfoxides. J . Am. Chem.
Soc. 92, 7610-7612.
(17) Pouzet, P., Erdelmeier, I., Ginderow, D., Mornon, J . P., Mansuy,
D., and Dansette, P. M. (1995) Convenient preparation, first
complete structural characterization and unexpected dimerization
of one of them, 2,5-Diphenylthiophenesulfoxide. J . Chem. Soc.,
Chem. Commun., 473-474.
(42) Bax, A. (1984) Structure determination and spectral assignment
by pulsed polarization transfer via long-range proton-carbon-13
couplings. J . Magn. Reson. 57, 314-318.
(18) Procha˜zka, M. (1965) Uber Thiophen-1-oxid. Collect. Czech. Chem.
Commun. 30, 1158-1168.
TX9601622