Chlorothioketene, the ultimate reactive intermediate formed by cysteine,conjugate β-lyase-mediated cleavage of the trichloroethene metabolite S-(1,2-dichlorovinyl)-L-cysteine, forms cytosine adducts in organic solvents, but not in aqueous solution

Article abstract of DOI:10.1021/tx980084d

Chlorothioketene has been suggested as a reactive intermediate formed by the cysteine conjugate β-lyase-mediated cleavage of S-(1,2-dichlorovinyl)- L-cysteine, a minor metabolite of trichloroethene. Halothioketenes are highly reactive, and their intermediate formation may be confirmed by reactions such as cycloadditions and thioacylations of nucleophiles. A precursor of chlorothioketene, S-(1,2-dichlorovinyl)thioacetate, is readily accessible by the reaction of dichloroethyne with thioacetic acid. In presence of base, S- (1,2-dichlorovinyl)thioacetate is cleaved to chlorothioketene. Chlorothioketene is not stable at room temperature and was characterized after transformation to stable products by reaction with compounds such as cyclopentadiene, N,N-diethylamine, and ethanol. In organic solvents, the cleavage of S-(1,2-dichlorovinyl)thioacetate in the presence of cytosine results in N⁴-acetylcytosine, N⁴-(chlorothioacetyl)cytosine; and small amounts of 3-(N⁴-thioacetyl)cytosine. No reaction products were seen with guanosine, adenosine, and thymidine under identical conditions. When cytosine was reacted with S-(1,2-dichlorovinyl)thioacetate in aqueous solutions, only N⁴-acetylcytosine was formed. N⁴-(Chlorothioacetyl)cytosine and 3-(N⁴- thioacetyl)cytosine were not detected even when using a very sensitive method, derivatization with pentafluorobenzyl bromide and electron capture mass spectrometry with a detection limit of 50 fmol/μL of injection volume. Aqueous solutions of DNA cleave S-(1,2-dichlorovinyl)thioacetate to give N⁴- acetyldeoxycytidine in DNA, but chlorothioketene adducts of deoxynucleosides were also not detected in these experiments. These results confirm the electrophilic reactivity of chlorothioketene toward nucleophilic groups of DNA constituents in inert solvents but also demonstrate that the formation of DNA adducts under physiological conditions likely is not efficient. Therefore, DNA adducts may not represent useful biomarkers of exposure and biochemical effects for trichloroethene.

Full text of DOI:10.1021/tx980084d

1082
Chem. Res. Toxicol. 1998, 11, 1082-1088
Ch lor oth iok eten e, th e Ultim a te Rea ctive In ter m ed ia te
F or m ed by Cystein e Con ju ga te â-Lya se-Med ia ted
Clea va ge of th e Tr ich lor oeth en e Meta bolite
S-(1,2-Dich lor ovin yl)-L-cystein e, F or m s Cytosin e Ad d u cts
in Or ga n ic Solven ts, bu t Not in Aqu eou s Solu tion
Wolfgang Vo¨lkel and Wolfgang Dekant*
Department of Toxicology, University of Wu¨rzburg, Versbacherstrasse 9, 97078 Wu¨rzburg, Germany
Received April 27, 1998
Chlorothioketene has been suggested as a reactive intermediate formed by the cysteine
conjugate â-lyase-mediated cleavage of S-(1,2-dichlorovinyl)-L-cysteine, a minor metabolite of
trichloroethene. Halothioketenes are highly reactive, and their intermediate formation may
be confirmed by reactions such as cycloadditions and thioacylations of nucleophiles. A precursor
of chlorothioketene, S-(1,2-dichlorovinyl)thioacetate, is readly accessible by the reaction of
dichloroethyne with thioacetic acid. In presence of base, S-(1,2-dichlorovinyl)thioacetate is
cleaved to chlorothioketene. Chlorothioketene is not stable at room temperature and was
characterized after transformation to stable products by reaction with compounds such as
cyclopentadiene, N,N-diethylamine, and ethanol. In organic solvents, the cleavage of S-(1,2-
dichlorovinyl)thioacetate in the presence of cytosine results in N4-acetylcytosine, N4-(chlo-
rothioacetyl)cytosine, and small amounts of 3-(N4-thioacetyl)cytosine. No reaction products
were seen with guanosine, adenosine, and thymidine under identical conditions. When cytosine
was reacted with S-(1,2-dichlorovinyl)thioacetate in aqueous solutions, only N4-acetylcytosine
was formed. N4-(Chlorothioacetyl)cytosine and 3-(N4-thioacetyl)cytosine were not detected even
when using a very sensitive method, derivatization with pentafluorobenzyl bromide and electron
capture mass spectrometry with a detection limit of 50 fmol/µL of injection volume. Aqueous
solutions of DNA cleave S-(1,2-dichlorovinyl)thioacetate to give N4-acetyldeoxycytidine in DNA,
but chlorothioketene adducts of deoxynucleosides were also not detected in these experiments.
These results confirm the electrophilic reactivity of chlorothioketene toward nucleophilic groups
of DNA constituents in inert solvents but also demonstrate that the formation of DNA adducts
under physiological conditions likely is not efficient. Therefore, DNA adducts may not represent
useful biomarkers of exposure and biochemical effects for trichloroethene.
compounds from this group such as S-(1,2-dichlorovinyl)-
L-cysteine and S-(trichlorovinyl)-L-cysteine showed â-lyase-
dependent mutagenicity in the Ames test (1) and did
induce some DNA damage in mammalian cells in culture
(5, 6). Since nephrotoxic cysteine S-conjugates are minor
metabolites of several haloalkenes, which induce a low
incidence of renal tumors in rats after long administra-
tion of high doses (7-10).
In this study, we examined the potential of chlo-
rothioketene generated in situ from a novel precursor,
S-(1,2-dichlorovinyl)thioacetate (1), to react with DNA
bases and DNA using a highly sensitive GC/MS tech-
nique for identification and detection of reaction products
under different conditions.
In tr od u ction
Thioketenes have been suggested as reactive interme-
diates formed by the cysteine conjugate â-lyase-mediated
cleavage of several halovinyl cysteine S-conjugates such
as S-(1,2-dichlorovinyl)-L-cysteine (1). Halothioketenes
are highly reactive electrophiles and are difficult to
synthesize and to handle experimentally (2); however,
several lines of evidence indicate that the R-chloroenethi-
ols formed by cysteine conjugate â-lyase from halovinyl
cysteine S conjugates are converted to halothioketenes.
For example, dichloroacetylated proteins are formed by
the interaction of a reactive intermediate, likely dichlo-
rothioketene, formed by the cysteine conjugate â-lyase-
dependent metabolism of the perchloroethene metabolite
S-(trichlorovinyl)-L-cysteine (3). Moreover, conversion
products characteristic for thioketenes were observed
when R-chloroenethiols were released from precursors in
chemical systems (4).
Exp er im en ta l P r oced u r es
Syn th esis of S-(1,2-Dich lor ovin yl)th ioa ceta te. Dichlo-
roethyne etherate was synthesized as described (Ca u t ion :
Dichloroethyne is highly flammable, explosive, and highly toxic.
Do not attempt to generate pure dichloroethyne. Use an efficient
hood!) (11). To generate S-(1,2-dichlorovinyl)thioacetate, 45
mmol of thioacetic acid was added dropwise to 45 mmol of
dichloroethyne etherate in 10 mL of n-hexane (Scheme 1). The
reaction mixture was kept at room temperature and stirred for
An interaction of metabolites formed from halovinyl
cysteine S-conjugates with DNA may occur since several
* To whom correspondence should be addressed: Department of
Toxicology, University of Wu¨rzburg, Versbacherstrasse 9, 97078
Wu¨rzburg, Germany. Telephone: +49(0)931/201-3449. Fax: +49(0)-
931/201-3865. E-mail: dekant@toxi.uni-wuerzburg.de.
S0893-228x(98)00084-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/25/1998

Reactivity of Chlorothioketene with Nucleosides
Chem. Res. Toxicol., Vol. 11, No. 9, 1998 1083
PFBnBr (NCI): m/z (relative intensity) 347 (15%) (M-), 346
(100%) (M- - H), 326 (50%) (M- - 2H - F). Mass spectra after
derivatization with PFBnBr (EI): m/z (relative intensity) 181
(100%) (PFBn+), 166 (40%) (M+ - PFBn).
Sch em e 1. Syn th esis of
S-(1,2-Dich lor ovin yl)th ioa ceta te
In cu b a t ion of S-(1,2-Dich lor ovin yl)t h ioa cet a t e w it h
DNA Ba ses, Deoxyn u cleosid es, a n d DNA in Aqu eou s
Solu tion s. Incubation of S-(1,2-dichlorovinyl)thioacetate with
DNA bases, deoxynucleosides, and calf thymus (ct) DNA in
buffered aqueous solutions at pH 5-7.5 was performed in the
same stoichiometric amounts as described above. Consumption
of S-(1,2-dichlorovinyl)thioacetate was determined by GC/MS
after extraction into n-hexane.
DNA Hyd r olysis. ctDNA from the reaction mixtures was
either precipitated with 2-propanol or enzymatically hydrolyzed
if incubations were performed in buffered solutions at pH 5. The
precipitated residues were dissolved in buffered solutions at pH
5 and cleaved enzymatically using 1000 units of nuclease S1
(EC 3.1.30.1), 1 unit of acid phosphodiesterase II (EC 3.1.16.1),
and 2 units of acid phosphatase (EC 3.1.3.2) per 1 mg of DNA
within 12 h to minimize hydrolysis of potential labile adducts
(16).
Sa m p le P r ep a r a tion for GC/MS. The method is based on
a publication of Fedtke et al. (14) with the following modifica-
tions. Dry residues of reaction mixtures (≈1 mg) were dissolved
in 2 mL of dimethylformamide, 3 mg of K2CO3, and 6 µL of
pentafluorobenzyl bromide (Ca u tion : Pentafluorbenzyl bromide
is a potent lacrymator. Use an efficient hood!). The mixtures
were then stirred for 12 h at room temperature and concentrated
in vacuo. The obtained residues were dissolved in 200 µL of
dichloromethane, and 2 µL of the obtained solution was injected
into the gas chromatograph.
Gas Ch r om atogr aph y/Mass Spectr om etr y. GC/MS analy-
sis was performed on a MD 800 mass spectrometer equipped
with a Carlo Erba 8000 series gas chromatograph (Fisons
Instruments, Mainz, FRG). For all separations, a DB-5 (J &
W Scientific, Folsom, CA) fused silica capillary GC column (30
m, 0.25 mm i.d., 0.1 µm film thickness) with helium (average
linear velocity of 35 mL/min) as the carrier gas was used.
Injections were made splitless (valve time of 1.0 min); a
temperature gradient starting at an oven temperature of 60 °C
and a heating rate of 10 °C/min used to reach 290 °C were used
for separation. The transfer line was kept at a temperature of
280 °C. The injector temperature was 250 °C, and the electron
source of the mass spectrometer was adjusted to 200 °C both in
the electron impact ionization mode and in the chemical
ionization mode. Chemical ionization was performed with
methane as the reactant gas. The solvent delay was 8 min.
Mass spectra (m/z 100-600) were recorded from 8 to 17 min
with a scan time of 0.5 s and an interchannel delay of 0.05 s.
The detection limit was measured in SIR mode using charac-
teristic fragments of the relevant compounds (m/z 347 and 383).
Dwell times were 30 ms.
1 h. A mixture of (E)- and (Z)-S-(1,2-dichlorovinyl)thioacetate
was isolated by vacuum distillation after removal of the solvent.
The purity of the product was >98% as checked by GC/MS. The
boiling point was 70 °C at 0.1 Torr.
Ch a r a cter iza tion of th e Tw o Isom er s of S-(1,2-Dich lo-
r ovin yl)th ioa ceta te. Mass spectra (identical for both isomers)
(EI, 35Cl): m/z (relative intensity) 170 (2Cl, 2%) (M+), 135 (1Cl,
11%) (M+ - Cl), 127 (2Cl, 6%) (M+ - COCH3), 92 (1Cl, 100%)
(M+ - COCH3 - Cl). 1H NMR for isomer I (250 MHz, CDCl3):
δ 2.30 (s, 3H, CH3), 6.90 (s, 1H, CHCl). 13C NMR for isomer I
(63 MHz, CDCl3): δ 30.2 (s, CH3), 124.1 (s, CClS), 129.2 (s,
CHCl), 188.2 (s, CO). 1H NMR for isomer II (250 MHz, CDCl3):
δ 2.40 (s, 3H, CH3), 6.80 (s, 1H, CHCl). 13C NMR for isomer II
(63 MHz, CDCl3): δ 29.2 (s, CH3), 124.1 (s, CClS), 129.0 (s,
CHCl), 188.2 (s, CO).
Syn th esis of (Ch lor oa cetylth ion o)a cetic Acid Eth yl
Ester . (Chloroacetylthiono)acetic acid ethyl ester was synthe-
sized by the reaction of 32 mg (0.26 mmol) of chloroacetic acid
ethyl ester and 136 mg (0.33 mmol) of Lawessons reagent (12)
in 1 mL of xylene. After being stirred for 12 h at 140 °C, the
mixture was cooled to room temperature and the formed
precipitate removed by centrifugation. GC/MS analysis of the
supernatant demonstrated the exclusive formation of the thiono
ester (stench). The thiono ester was not isolated and character-
ized by GC/MS. Mass spectrum (EI, 35Cl): m/z (relative
intensity) 138 (1Cl, 100%) (M+), 110 (1Cl, 75%) (MH+ - Et),
103 (18%) (M+ - Cl), 93 (1Cl, 33%) (M+ - OEt), 77 (1Cl, 68%)
(M+ - Et - S).
Syn th esis of N4-Acetylcytosin e. An excess of cytosine (36
µmol) was stirred with acetic acid anhydride (23 µmol) in 2 mL
of anhydrous dimethylformamide. The obtained solution was
analyzed by HPLC with diode array detection and with HPLC/
MS. The obtained product had an electronic spectrum identical
to that described in the literature (13) and an electrospray mass
spectrum consistent with the proposed structure. Electronic
spectrum (MeOH/H2O, pH 2): λmax (ꢀ in mAU) 217 (100), 246
(90), 302 (60). Electrospray mass spectrum: m/z 153 (M+).
In cu ba t ion of S-(1,2-Dich lor ovin yl)t h ioa cet a t e w it h
DNA Ba ses. Cytosine (36 µmol) and a 1:2 mixture (based on
the GC/MS response) of both isomers of S-(1,2-dichlorovinyl)-
thioacetate (23 µmol) were stirred in a mixture of dimethylform-
amide and ethyl acetate (1:3, v/v) at room temperature. After
4 h, ethyl acetate was evaporated in a stream of nitrogen and
the residue (5 µL) was analyzed by HPLC or subjected to
derivatization with pentafluorobenzyl bromide for GC/MS/NCI
(14). Reactions with adenine, guanine, and thymine were
performed under identical conditions.
GC/MS data for detection of thioketene trapping products
were aquired with a HP 5890 gas chromatograph coupled with
a HP 5970 mass-selective detector (Hewlett-Packard, Avondale,
PA) with electron-impact ionization (70 eV). For gas chromato-
graphic separations, a DB 1 fused silica capillary column (40
m, 0.18 mm i.d., 0.4 µm film thickness; J &W Scientific) with
helium as the carrier gas was used. Injections were made
splitless; a temperature gradient starting at an oven tempera-
ture of 40 °C and a heating rate of 15 °C/min used to reach 280
°C were used. The injector and transfer line temperatures were
kept at 280 °C.
Electr osp r a y Ma ss Sp ectr om etr y (ES/MS). ES/MS was
performed on a Trio 2000 mass spectrometer (VG Biotech,
Manchester, England) coupled with a HP 1090 HPLC system
(Hewlett-Packard) using H2O/TFA (pH 2)/AcCN (30:70) as a
mobile phase with a flow rate of 10 µL/min.
Characterization ofN4-(Chlorothioacetyl)cytosine Formed
in Th ese Rea ction s. Electronic spectrum (MeOH/H2O, pH
2): λmax (ꢀ in mAU) 215 (150), 244 (80), 298 (30). Mass spectra
after derivatization with PFBnBr (NCI, 35Cl): m/z (relative
intensity) 383 (1Cl, 100%) (M-), 364 (1Cl, 1%) (M- - F), 203
(1Cl, 8%) (MH- - PFBn). Mass spectra after derivatization with
PFBnBr (EI, 35Cl): m/z (relative intensity) 349 (8%) (MH+
-
Cl), 203 (1Cl, 30%) (MH+ - PFBn), 181 (100%) (PFBn+), 92 (1Cl,
18%) (CHCl - CS).
1H NMR spectra were identical to those reported for N4-
(chloroacetyl)cytosine (15).
Cycliza tion of N4-(Ch lor oth ioa cetyl)cytosin e. N4-(Chlo-
rothioacetyl)cytosine was stirred at room temperature with
catalytic amounts of K2CO3 in DMF for 1 h. The obtained
solution (10 µL) was directly analyzed by HPLC, and 3-(N4-
thioacetyl)cytosine was characterized by its electronic and by
its mass spectrum (15). UV (MeOH/H2O, pH 2): λmax (ꢀ in mAU)
202 (600), 302 (1200). Mass spectra after derivatization with
H igh -P er for m a n ce Liq u id Ch r om a t ogr a p h y (H P LC).
HPLC was performed using two Waters M-6000 A pumps
(Millipore GmbH, Eschborn, FRG) coupled with a gradient
control unit. Steel columns (250 mm × 4 mm i.d.) filled with

1084 Chem. Res. Toxicol., Vol. 11, No. 9, 1998
Vo¨lkel and Dekant
F igu r e 1. Transformation of the intermediate chlorothioketene to stable reaction products for characterization and reaction of
chlorothioketene with cytosine.
Partisil ODS III (5 µm particle size, Bischoff, Leonberg, FRG)
were used for separation. Solvents were as follows: (i) solvent
A, water with trifluoroacetic acid adjusted to pH 2; solvent B,
CH3OH; 0 to 100% B in 40 min at a flow rate of 1 mL/min. The
effluent was passed through a HP 1090 A (Hewlett-Packard)
diode array detector.
Resu lts
S-(1,2-Dichlorovinyl)thioacetate (1, Figure 1) may be
easily obtained by synthesis (Scheme 1), is relatively
stable in aqueous solution (t1/2 at pH 7 determined to be
approximately 12 h), and is cleaved under basic condi-
tions to give S-(1,2-dichlorovinyl)thiolate (1a , Figure 1)
in much higher yields compared to those of previously
used precursors for vinyl thiols (4).
The formation of chlorothioketene (2, Figure 1) from
thiol 1a was confirmed by the detection of characteristic
reaction products. Without agents to transform chlo-
rothioketene to stable products, a low concentration of
dithiethane 3 was detected using GC/MS (4, 17). In
F igu r e 2. GC separation of an incubation of S-(1,2-dichlorovi-
addition, oligomers of the highly reactive thioketene
seemed to be formed (18). In the presence of cyclopen-
tadiene, 5-thioxo-6-chlorobicyclo[2.2.1]hept-2-ene 4 (2, 4,
18) was formed as described for structurally similar
thioketenes such as di-tert-butylthioketene (19). In the
presence of alcohols, thiono ester 5 [mass spectrum
identical to that of synthetic (chloroacetylthiono)acetic
acid ethyl ester] was formed, and in the presence of N,N-
diethylamine, thioamide 6 was formed (20). An example
of the results of these trapping reactions is shown in
Figure 2. In the presence of diethylamine, which also
functions as a base to catalyze the cleavage of S-(1,2-
dichlorovinyl)thioacetate to thiol 1a and chlorothioketene
2, both N,N-diethylacetamide (peak at 10.3 min in the
chromatogram in Figure 2) and N,N-diethylchlorothio-
acetamide 6 (peak at 15.0 min in the chromatogram in
Figure 2) were formed. These observations suggest that
hydrolytic cleavage of S-(1,2-dichlorovinyl)thioacetate
represents a useful route for generating high concentra-
nyl)thioacetate in organic solution in the presence of N,N-
diethylamine. Mass spectrum A represents N,N-diethylaceta-
mide, and mass spectrum
B
represents N,N-diethyl-
chlorothioacetamide (20).
tions of chlorothioketene with only one simple byproduct,
acetate. The previous method used to generate halovinyl
thiols had a much lower yield (as compared in the GC/
MS response) and produced a variety of byproducts
interfering with further analysis (4).
When S-(1,2-dichlorovinyl)thioacetate was reacted with
DNA bases in organic solvents such as dimethylforma-
mide, formation of base modifications was only observed
with cytosine by HPLC. Incubations of S-(1,2-dichlorovi-
nyl)thioacetate with guanosine, adenosine, and thymi-
dine did not result in the formation of new peaks, even
in the presence of additional non-nucleophilic bases such
as 1,4-diazabicyclo[2.2.2]octane which promotes hydroly-
sis of the thioacetate. The formed cytosine derivatives

Reactivity of Chlorothioketene with Nucleosides
Chem. Res. Toxicol., Vol. 11, No. 9, 1998 1085
F igu r e 3. HPLC chromatogram and the electronic spectrum of N4-acetylcytosine (A) and N4-(chlorothioacetyl)cytosine (peak B)
after incubation of S-(1,2-dichlorovinyl)thioacetate in the presence of cytosine in dimethylformamide.
F igu r e 4. Mass spectra of pentafluorobenzyl derivatives of 8 after chemical ionization (negative ion detection) (A) and electron
impact ionization (B).
had characteristic electronic spectra (Figure 3); the
compound representing peak A (10.9 min) was identified
as N4-acetylcytosine by comparison of the electronic
spectrum with that of a reference compound. Peak B
(14.7 min) exhibited an electronic spectrum identical to
that of N4-(chlorothioacetyl)cytosine (15). The presence
of this compound in the incubations was also confirmed
by GC/MS/NCI. Two peaks were obtained by gas chro-
matographic separation of the pentafluorobenzylated
reaction mixtures. One peak contained one chlorine atom
(m/z 383) and represents the pentafluorobenzyl derivative
of 8 (Figure 1). Figure 4 shows the mass spectra of 8
(the top shows the electron capture mode and the bottom
was recorded in the electron impact mode). The spectra
are consistent with the proposed structure of 8. A value
of m/z 383 (1Cl) represents M- in the electron capture
mode, consistent with the low degree of fragmentation
in this ionization process. After electron impact ioniza-
tion, the molecule fragments intensively and major
fragments represent m/z 181 (F5C6CH2+) and m/z 203
(1Cl) (MH - F5C6CH2+). The other peak with a single
mass fragment (m/z 347) likely represents the pentafluo-
robenzyl derivative of 9.
N4-(Chlorothioacetyl)cytosine is rapidly hydrolyzed in
water at pH 8 back to cytosine and, presumably, chloro-
(thiono)acetic acid, and ring closure to 9 occurs only in
low yield (≈10%) in water (data not shown). The half-
life of the hydroslysis of N4-(chlorothioacetyl)cytosine in
water was similar to that of N4-(chloroacetyl)cytosine (15,
21). In organic solvents in the presence of base, ring
closure of 8 to give 9 occurs in high yields. Thus, under
the conditions of the derivatization reaction which is
performed in dimethylformamide in the presence of base,
ring closure may be preferred, thus explaining the higher
yields.
On the basis of the results of the reaction of S-(1,2-
dichlorovinyl)thioacetate in organic solvents, a GC/MS/
NCI method was developed to demonstrate a reaction of
the intermediate chlorothioketene with DNA bases and
DNA under biologically more relevant conditions in

1086 Chem. Res. Toxicol., Vol. 11, No. 9, 1998
Vo¨lkel and Dekant
tidine 7 and had an identical electronic spectrum, sug-
gesting that thiol 1a is released and that 2 is formed in
the reaction mixture. However, peaks indicative of the
formation of the deoxy derivatives of 8 or 9 were not
observed by HPLC (Figure 5); GC/MS/NCI analysis of the
reaction mixtures also was unable to detect the formation
of 8 or 9 in these incubations. A GC/MS trace of an
incubation of S-(1,2-dichlorovinyl)thioacetate with cy-
tosine in water after pentafluorobenzylation by monitor-
ing representative ions of the two products formed from
N4-(chlorothioacetyl)cytosine is shown in Figure 6 (chro-
matogram B) and compared with the chromatogram
(Figure 6A) obtained with approximately 10 pmol/µL of
N4-(chlorthioacetyl)cytosine treated identically. Whereas
a clear signal for the synthetic reference compound was
obtained, the concentration of N4-(chlorothioacetyl)cy-
tosine formed by the reaction of chlorothioketene with
cytosine in aqueous solution was below the limit of
detection. Consistent with these observations, incuba-
tions of S-(1,2-dichlorovinyl)thioacetate with DNA fol-
lowed by DNA hydrolysis and HPLC separation also
demonstrated the formation of N4-acetyldeoxycytidine
(Figure 5B). The small peak in the chromatogram at tR
) 23.4 min had retention and electronic spectra identical
to those of N4-acetyldeoxycytidine, indicating cleavage of
the thioacetate and chlorothioketene formation; attempts
to detect N4-(chlorothioacetyl)deoxycytidine 8 or 9 in
these incubations by UV HPLC and/or GC/MS/NCI were
alsounsuccessful. However,the detection ofN4-acetyldeox-
ycytidine in DNA treated with the thioacetate demon-
strates that chlorothioketene is released in close prox-
imity to the DNA.
F igu r e 5. HPLC chromatograms and electronic spectra of N4-
acetyldeoxycytidine formed by incubations of S-(1,2-dichlorovi-
nyl)thioacetate 1 with deoxycytidine (A) and DNA after enzy-
matic hydrolysis (B). The peaks marked with a V have electronic
spectra identical to those of N4-acetyldeoxycytidine and coelute
with the reference compound.
aqueous solution. The method based on selected ion
monitoring had a limit of detection of 50 pmol/mL of N4-
(chlorothioacetyl)cytosine.
Incubations of S-(1,2-dichlorovinyl)thioacetate with
deoxycytidine or cytosie in aqueous solutions showed a
rapid decrease of the concentration of the thioacetate by
GC/MS analysis of hexane extracts and the formation of
N4-acetylcytosine (7, Figure 1), respectively, the deoxy
derivative in high yield (Figure 5, chart A). The peak at
tR ) 23.7 min coeluted with synthetic N4-acetyldeoxycy-
Discu ssion
The results show that chlorothioketenes, most likely
representing the toxic metabolites formed from halovinyl
cysteine S-conjugates, are highly reactive and short-lived.
On the basis of the results with trapping agents, we
conclude S-(1,2-dichlorovinyl)thioacetate is an easily
accessible precursor for the in situ generation of chlo-
rothioketene for studying its reactivity with potential
biologically relevant target molecules (2). The previous
F igu r e 6. GC/MS traces (m/z 383) of 10 pmol/µL of N4-(chlorothioacetyl)cytosine (after pentafluorobenzylation) (A) and an incubation
of cytosine with S-(1,2-dichlorovinyl)thioacetate in buffered solution at pH 8 treated identically (B). For details of the sample
preparation and GC/MS conditions, see Experimental Procedures.

Reactivity of Chlorothioketene with Nucleosides
Chem. Res. Toxicol., Vol. 11, No. 9, 1998 1087
precursors, halovinyl 2-nitrophenyl disulfides, although
useful for generating structurally different halothioketenes,
require a more complex synthesis and purification pro-
cedure. Moreover, their cleavage results in byproducts,
which may be chemically reactive and interfere during
analytical procedures. The novel chlorothioketene pre-
cursor used in this study in contrast is simple to
synthesize, relatively stable in water at neutral pH, and
cleaved by base to dichlorovinyl thiol. This vinyl thiol is
further converted to chlorothioketene. Due to its high
reactivity, isolation or direct characterization of this
molecule is not possible, but the stable products were
detected in the presence of trapping agents often used
to demonstrate the intermediate formation of a thioketene
(2, 17, 18). Under favorable conditions, chlorothioketene
may interact with DNA bases or deoxynucleosides. In
polar organic solvent in the presence of base (cytosine,
which catalyzes the release of chlorovinyl thiol as indi-
cated by the formation of N4-acetylcytosine), relatively
high concentrations of chlorothioketene may be present
and available for reaction with the DNA base since water
is not a completing nucleophile. In presence of methanol,
1 was cleaved at about 50 °C by cytosine to chlo-
rothioketene which immediately reacts with methanol
quantitatively to form the corresponding thionoester.
Both the thionoester and N4-acetylcytosine were detected
in the methanolic solution either by GC/MS or by UV
HPLC (data not shown). Moreover, cytosine shows a
higher nucleophilicity in dimethylformamide than in
water due to the absence of solvation. Therefore, chlo-
rothioketene reacts in relatively good yields (as compared
with the reaction of other electrophiles with DNA con-
stituents). The results presented here show that 1 was
also cleaved by cytosine in the presence of water adjusted
to pH 6-8 as confirmed by formation of the relatively
stable N4-acetylcytosine (13) resulting from the nucleo-
philic attack of the exocyclic amino group of cytosine on
the electrophilic carbonyl function of 1, but the generation
of N4-(chlorothioacetyl)cytosine by the reaction of cytosine
with chlorothioketene did not occur. This may occur for
two reasons. (i) The highly reactive chlorothioketene is
rapidly hydrolyzed in the presence of water. (ii) Small
amounts of N4-(chlorothioacetyl)cytosine formed may be
rapidly hydrolyzed under these conditions like those
described for N4-(chloroacetyl)cytosine (13). Our results
are confirmed by the observation that chloro(thio)acety-
lation of cytosine only occurs in anhydrous dimethylform-
amide (15). In the presence of water, chlorothioketenes
are very rapidly hydrolyzed and react with DNA con-
stituents, if at all, only in very low yields. On the basis
of theoretical considerations (hard and soft electrophiles),
a reaction of chlorothioketene, a soft electrophile, with
nucleophilic sites in DNA constituents (hard nucleo-
philes) should be less favored (22-24). Thus, proteins
(softer nucleophilic centers) may be favored targets for
interactions of halothioketenes in biological systems. This
assumption is supported by the in vivo formation of
protein adducts of an intermediate thioketene, formed
by â-lyase-mediated cleavage of the perchloroethene
metabolite S-(trichlorovinyl)-L-cysteine. These protein
adducts are formed in relatively good yields. Due to the
very low yield of the reactions of halothioketenes with
DNA constituents and the low concentrations of hal-
othioketene precursors formed in rodents in vivo after
administration of trichloroethene and perchloroethene
(25, 26), experimental demonstration of DNA adduct
formation in the kidney after administration of tri- or
perchloroethene to rodents has to be considered very
difficult, if not impossible.
Refer en ces
(1) Dekant, W., Vamvakas, S., Berthold, K., Schmidt, S., Wild, D.,
and Henschler, D. (1986) Bacterial â-lyase mediated cleavage and
mutagenicity of cysteine conjugates derived from the nephrocar-
cinogenic alkenes trichloroethylene, tetrachloroethylene and
hexachlorobutadiene. Chem.-Biol. Interact. 60, 31-45.
(2) Schaumann, E. (1988) The chemistry of thioketenes: Tetrahedron
Report No. 231. Tetrahedron 44, 1827-1871.
(3) Birner, G., Richling, C., Henschler, D., Anders, M. W., and
Dekant, W. (1994) Metabolism of tetrachloroethene in rats:
identification of Nꢀ-(dichloroacetyl)-L-lysine and Nꢀ-(trichloro-
acetyl)-L-lysine as protein adducts. Chem. Res. Toxicol. 7, 724-
732.
(4) Dekant, W., Urban, G., Go¨rsman, C., and Anders, M. W. (1991)
Thioketene formation from R-haloalkenyl 2-nitrophenyl disul-
fides: models for biological reactive intermediates of cytotoxic
S-conjugates. J . Am. Chem. Soc. 113, 5120-5122.
(5) Vamvakas, S., Dekant, W., and Henschler, D. (1989) Genotoxicity
of haloalkene and haloalkane glutathione S-conjugates in porcine
kidney cells. Toxicol. in Vitro 3, 151-156.
(6) Vamvakas, S., Dekant, W., and Henschler, D. (1989) Assessment
of unscheduled DNA synthesis in a cultured line of renal epithelial
cells exposed to cysteine S-conjugates of haloalkenes and haloal-
kanes. Mutat. Res. 222, 329-335.
(7) National Toxicology Program, National Cancer Institute (1977)
Bioassay of tetrachloroethylene for possible carcinogenicity.
Technical Report No. 13, U.S. Department of Health and Human
Services, Washington, DC.
(8) National Toxicology Program, National Cancer Institute (1986)
Carcinogenesis bioassay of tetrachloroethylene. Technical Report
No. 232, U.S. Department of Health and Human Services,
Washington, DC.
(9) National Toxicology Program, National Cancer Institute (1986)
Carcinogenesis bioassay of trichloroethylene. Technical Report No.
311, U.S. Department of Health and Human Services, Washing-
ton, DC.
(10) National Toxicology Program, National Cancer Institute (1988)
Toxicology and carcinogenesis studies of trichloroethylene in four
strains of rats (ACI, August, Marshall, Osborne-Mendel) (gavage
studies). Technical Report No. 273, U.S. Department of Health
and Human Services, Washington, DC.
(11) Kanhai, W., Dekant, W., and Henschler, D. (1989) Metabolism
of the nephrotoxin dichloroacetylene by glutathione conjugation.
Chem. Res. Toxicol. 2, 51-56.
(12) Pedersen, B. S., and Lawesson, S.-O. (1979) Studies on Organo-
phosphorus Compounds-XXVIII. Tetrahedron 35, 2433-2437.
(13) Goody, R. S., and Walker, R. T. (1967) Some N6 acylated
cytosines. Tetrahedron Lett., 289-291.
(14) Fedtke, N., Boucheron, J . A., Turner, M. J ., and Swenberg, J . A.
(1990) Vinyl chloride-induced DNA adducts. I: quantitative
determination of N2,3-ethenoguanine based on electrophore label-
ing. Carcinogenesis 11, 1279-1285.
(15) Mu¨ller, M., Birner, G., and Dekant, W. (1998) Reactivity of
haloketenes and halothioketenes with nucleobases: chemical
characterization of reaction products. Chem. Res. Toxicol. 11,
454-463.
(16) Mu¨ller, M., Birner, G., Sander, M., and Dekant, W. (1998)
Reactivity of haloketenes and halothioketenes with nucleobases:
reactions in vitro with DNA. Chem. Res. Toxicol. 11, 464-470.
(17) Raasch, M. S. (1970) Bis(trifluoromethyl)thioketene. I. Synthesis
and cycloaddition reactions. J . Org. Chem. 35, 3470-3483.
(18) Seybold, G., and Heibl, C. (1975) Blitzhermolyse von 1,2,3-
Thiadiazolen: ein einfacher Weg zu Thioketenen (Flash ther-
molysis of 1,2,3-thiadiazoles. Simple route to thioketenes). Angew.
Chem. 87, 171-172.
(19) Elam, E. U., Rash, F. H., Dougherty, J . T., Goodlett, V. W., and
Brannock, K. C. (1968) Di-tert-butylthioketene. J . Org. Chem. 33,
2738-2741.
(20) Dekant, W., Berthold, K., Vamvakas, S., Henschler, D., and
Anders, M. W. (1988) Thioacylating intermediates as metabolites
of S-(1,2-dichlorovinyl)-L-cysteine and S-(1,2,2-trichlorovinyl)-L-
cysteine formed by cysteine conjugate â-lyase. Chem. Res. Toxicol.
1, 175-178.
(21) Goody, R. S., and Walker, R. T. (1971) The preparation and
properties of some cytosine derivatives. J . Org. Chem. 36, 727-
730.

1088 Chem. Res. Toxicol., Vol. 11, No. 9, 1998
Vo¨lkel and Dekant
(22) Anders, M. W. (1988) Bioactivation mechanisms and hepatocel-
lular damage. In The Liver: Biology and Pathology (Arias, I. M.,
J akoby, W. B., Popper, H., Schachter, D., and Shafritz, D. A., Eds.)
2nd ed., pp 389-400, Raven Press, Ltd., New York.
(25) Dekant, W., Metzler, M., and Henschler, D. (1986) Identification
of S-1,2-dichlorovinyl-N-acetyl-cysteine as a urinary metabolite
of trichloroethylene: A possible explanation for its nephrocarci-
nogenicity in male rats. Biochem. Pharmacol. 35, 2455-2458.
(26) Dekant, W., Metzler, M., and Henschler, D. (1986) Identification
of S-1,2,2-trichlorovinyl-N-acetylcysteine as a urinary metabolite
of tetrachloroethylene: bioactivation through glutathione conju-
gation as a possible explanation of its nephrocarcinogenicity. J .
Biochem. Toxicol. 1, 57-72.
(23) Nelson, S. D., and Pearson, P. G. (1990) Covalent and noncovalent
interactions in acute lethal cell injury caused by chemicals. Annu.
Rev. Pharmacol. Toxicol. 30, 169-195.
(24) Hinson, J . A., and Roberts, D. W. (1992) Role of covalent and
noncovalent interactions in cell toxicity: effects on proteins. Annu.
Rev. Pharmacol. Toxicol. 32, 471-510.
TX980084D