Microsomal oxidation of tribromoethylene and reactions of tribromoethylene oxide.

Article abstract of DOI:10.1021/tx020047t

Halogenated olefins are of interest because of their widespread use in industry and their potential toxicity to humans. Epoxides are among the enzymatic oxidation products and have been studied in regard to their toxicity. Most of the attention has been given to chlorinated epoxides, and we have previously studied the reactions of the mono-, di-, tri-, and tetrachloroethylene oxides. To further test some hypotheses concerning the reactivity of these compounds, we prepared tribromoethylene (TBE) oxide and compared it to trichloroethylene (TCE) oxide and other chlorinated epoxides. TBE oxide reacted with H(2)O about 3 times faster than did TCE oxide. Several hydrolysis products of TBE oxide were the same as formed from TCE oxide, i.e., glyoxylic acid, CO, and HCO(2)H. Br(2)CHCO(2)H was formed from TBE oxide; the yield was higher than for Cl(2)CHCO(2)H formed in the hydrolysis of TCE oxide. The yield of tribromoacetaldehyde was < 0.4% in aqueous buffer (pH 7.4). In rat liver microsomal incubations containing TBE and NADPH, Br(2)CHCO(2)H was a major product, and tribromoacetaldehyde was a minor product. These results are consistent with schemes previously developed for halogenated epoxides, with migration of bromine being more favorable than for chlorine. Reaction of TBE oxide with lysine yielded relatively more N-dihaloacetyllysine and less N-formyllysine than in the case of TCE oxide. This same pattern was observed in the products of the reaction of TBE oxide with the lysine residues in bovine serum albumin. We conclude that the proposed scheme of hydrolysis of halogenated epoxides follows the expected halide order and that this can be used to rationalize patterns of hydrolysis and reactivity of other halogenated epoxides.

Full text of DOI:10.1021/tx020047t

1
414
Chem. Res. Toxicol. 2002, 15, 1414-1420
Micr osom a l Oxid a tion of Tr ibr om oeth ylen e a n d
Rea ction s of Tr ibr om oeth ylen e Oxid e
Tadao Yoshioka, J oel A. Krauser, and F. Peter Guengerich*
Department of Biochemistry and Center in Molecular Toxicology, Vanderbilt University School of
Medicine, Nashville, Tennessee 37232-0146
Received J une 4, 2002
Halogenated olefins are of interest because of their widespread use in industry and their
potential toxicity to humans. Epoxides are among the enzymatic oxidation products and have
been studied in regard to their toxicity. Most of the attention has been given to chlorinated
epoxides, and we have previously studied the reactions of the mono-, di-, tri-, and tetrachloro-
ethylene oxides. To further test some hypotheses concerning the reactivity of these compounds,
we prepared tribromoethylene (TBE) oxide and compared it to trichloroethylene (TCE) oxide
and other chlorinated epoxides. TBE oxide reacted with H
TCE oxide. Several hydrolysis products of TBE oxide were the same as formed from TCE oxide,
i.e., glyoxylic acid, CO, and HCO H. Br CHCO H was formed from TBE oxide; the yield was
higher than for Cl CHCO H formed in the hydrolysis of TCE oxide. The yield of tribromo-
acetaldehyde was < 0.4% in aqueous buffer (pH 7.4). In rat liver microsomal incubations
containing TBE and NADPH, Br CHCO H was a major product, and tribromoacetaldehyde
2
O about 3 times faster than did
2
2
2
2
2
2
2
was a minor product. These results are consistent with schemes previously developed for
halogenated epoxides, with migration of bromine being more favorable than for chlorine.
Reaction of TBE oxide with lysine yielded relatively more N-dihaloacetyllysine and less
N-formyllysine than in the case of TCE oxide. This same pattern was observed in the products
of the reaction of TBE oxide with the lysine residues in bovine serum albumin. We conclude
that the proposed scheme of hydrolysis of halogenated epoxides follows the expected halide
order and that this can be used to rationalize patterns of hydrolysis and reactivity of other
halogenated epoxides.
In tr od u ction
Halogenated olefins have had many uses as solvents
Sch em e 1. P r op osed Mech a n ism of P 450-Ca ta lyzed
a
Oxid a tion of Vin yl Ha lid es
and as vinyl monomers in the plastics industry. Many of
these chemicals are used in large volumes, e.g., perchlo-
1
8
-1
roethylene (PCE) at 10 kg year in the United States
1). These chemicals can all be toxic at high doses, and
(
the study of their toxicities has been a matter of consid-
erable interest. Vinyl chloride (VC) is clearly a rodent
and human carcinogen (2, 3). The carcinogenicity of
trichloroethylene (TCE) and perchloroethylene (PCE,
tetrachloroethylene) is more controversial (1, 4, 5), but
these compounds can have other detrimental effects (6).
Vinylidene chloride (VDC) is relatively toxic in liver and
kidney (7, 8) but probably not carcinogenic (9-11).
Although GSH conjugation can contribute to bioacti-
vation of some of the more highly substituted vinyl
halides (12, 13), much of the existing literature is
centered on the formation and roles of the epoxides (14,
a
X ) halogen.
early studies on epoxides yielded ambiguous results in
terms of what the products were or whether the epoxide
was the principal electrophile that reacts with macro-
molecules (22). Recent work in this laboratory has led to
the hypothesis that a major feature of the reaction of
halooxiranes with nucleophiles is the intermediacy of acid
halide hydrolytic products, which behave as acylating
agents (20, 21, 23). However, the situation is complicated
by the possibility that P450 enzymes can generate acyl
halides in 1,2-migration reactions exclusive of epoxide
products (Scheme 1) (18, 19, 21). Thus, modification of
macromolecules may be difficult to understand in the
context of a single reactive species.
1
5). The chlorooxirane derivatives of the above com-
pounds are all relatively unstable, with t1/2 of ∼2 s to 2.6
min under physiological conditions (16-21). Some of the
*
Address correspondence to this author at the Department of
Biochemistry and Center in Molecular Toxicology, Vanderbilt Univer-
sity School of Medicine, 638 Robinson Research Building (MRB I),
Nashville, TN 37232-0146. Telephone: (615) 322-2261. FAX: (615) 322-
3
141. E-mail: guengerich@toxicology.mc.vanderbilt.edu.
1
Abbreviations: TBE, tribromoethylene; TCE, trichloroethylene;
VC, vinyl chloride; VDC, vinylidene chloride (1,1-dichloroethylene);
PCE, perchloroethylene (tetrachloroethylene); VB, vinyl bromide; CBZ,
carbobenzyloxy; MOPS, 3-(N-morpholino)propanesulfonate; Orn, orni-
thine.
1
0.1021/tx020047t CCC: $22.00 © 2002 American Chemical Society
Published on Web 10/10/2002

Tribromoethylene Oxide
Chem. Res. Toxicol., Vol. 15, No. 11, 2002 1415
Brominated olefins are used less commonly than the
chlorinated analogues in industry and have been studied
less. Vinyl bromide (VB) has been studied in several
contexts and exhibits behavior generally similar to VC,
in regard to both carcinogenicity (24) and mechanism of
bioactivation (25-29). Despite the limited industrial use
of bromine compounds, they can be highly useful in
studying mechanisms of reactions of halooxiranes through
insights in the halide order. A key model for our research
on vinyl halides has been TCE and the reactions of TCE
oxide (19, 20, 23, 30). We extended this research project
to tribromoethylene (TBE) in order to address some
general hypotheses about vinyl halides and their ep-
oxides. The literature contains some studies on the
chemistry of peracid oxidation of TBE (31) and the use
of TBE as a model substrate for bacterial methane
monooxygenase (32). Interestingly, TBE has also been
demonstrated to be a volatile natural product produced
by algae (33) and is a metabolite of 1,1,2,2-tetrabromo-
ethane in rats (34). TBE has also been shown to be toxic
to algae, crustaceans, and fish (35).
signals. From these measurements, the concentration of TBE
oxide in the TBE was 4.9% (v/v) in the sample used, and the
ꢀ
560 of the 4-(4-nitrobenzyl)pyridine complex (37) was calculated
-
1
-1
to be 19 800 M cm under these conditions (20, 21). This
extinction coefficient was subsequently used in other assays.
The residual TBE did not interfere with any of the subsequent
measurements, and it did not yield any of the analyzed products
of TBE oxide. TBE oxide was stored in the presence of desiccant
at -80 °C.
6
Syn th esis of N -Dibr om oa cetylLys. Ethyl dibromoacetate
(0.88 g, 1.2 equiv) was added to 3.0 mmol of L-Lys‚HCl in 1.5
mL of 2 N NaOH (3.0 mmol), and the reaction was stirred at
room temperature (2 h). The addition of 1 mL of 1 N HCl yielded
a white precipitate, which was collected by suction filtration,
washed with 50% (v/v) aqueous C
over P to give the product (203 mg, 20%). Recrystallization
from hot aqueous C OH gave 170 mg of crystalline material
2 5
H OH, and dried in vacuo
2
O
5
2
H
5
1
(mp 190 °C, decomp): H NMR (D
2
O) δ 1.40 (m, 2H, γ-H), 1.58
(
3
m, 2H, â-H), 1.87 (m, 2H, δ-H), 3.25 (t, 2H, J ) 6.7 Hz, ꢀ-H),
13
.69 (t, 1H, J ) 6.3 Hz, R-H), 6.17 (s, 1H, -COCHBr
2
); C NMR
(D
2
O) δ 22.3 (γ), 28.3 (δ), 30.7 (â) 37.0 (-CHBr ), 40.5 (ꢀ), 53.3
2
(R), 168.2 (-COCHBr ), 175.4 (CO H); MS m/z 345, 347, 349
(MH ) (plus MNa ions at 367, 369, 371).
2
2
+
+
2
We synthesized TBE oxide and compared it with TCE
oxide in terms of rates of hydrolysis, hydrolytic products,
and reactions with free and protein Lys. With this
information about TBE oxide and previous results with
VC oxide, VDC oxide, TCE oxide, and PCE oxide, we
postulate general mechanisms for the formation of ha-
logenated epoxides and their reactivity.
Syn th esis of N -Dibr om oa cetylLys. Br
2
CHCO
2
H (5.6 g,
2
6 mmol) was dissolved in 5.0 mL of SOCl
2
and heated at reflux
for 2 h. Excess SOCl was removed in vacuo. A portion of the
2
resulting dibromoacetyl chloride (0.19 mL, 1.2 equiv) was added
6
to 0.81 g (2.0 mmol) of N -CBZ-L-Lys benzyl ester‚HCl (21, 38)
2 5 3
in a mixture of 8 mL of ethyl acetate/1 mL of (C H ) N at 0 °C,
and the mixture was stirred for 2 h, yielding a pale yellow solid
1
(
0.86 g, 90% yield): H NMR (CDCl
3
) δ 1.24-1.48 (m, 4H, â-
and γ-Η), 1.77-201 (m, 2H, â-Η), 3.14 (m, 2H, ꢀ-H), 4.58 (m,
1
5
H, R-H), 4.73 (br s, 1H, NH-), 5.09 (s, 2H, PhCH
.24 (each d, total 2H, J ) 12.1 Hz, PhCH -), 5.81 (s, 1H,
), 7.08 (app d, 1H, J ) 6.6 Hz, NH), 7.26-7.36 (m, 10
H, aromatic).
2
-), 5.17 and
Exp er im en ta l P r oced u r es
2
Ca u tion : TBE oxide and the acyl halides used here are
reactive acylating agents and should be handled carefully.
Gloves and other protective clothing should be used appropri-
ately.
-CHBr
2
The above compound (740 mg) was hydrolyzed with 1 N
NaOH (2 mL) in CH OH (4 mL) for 1.5 h at room temperature.
3
Ma ter ia ls. TBE was purchased from TCI America (Portland,
OR) and distilled under reduced pressure (bp 48-52 °C, 10
Torr). TCE oxide was prepared as described previously (20).
The mixture was acidified with 1 N HCl, and the resulting
carboxylic acid was extracted into ethyl acetate (3×). The solvent
was removed in vacuo, and the residue was treated with 3 mL
Br
3
CCHO and Br
2
CHCO
2
H were purchased from Aldrich Chemi-
of HBr in CH
followed by the addition of 200 mL of (C
3 2
CO H (30% w/v) at room temperature for 1 h,
2
cal Co. (Milwaukee, WI).
2
5 2
H ) O to give N -
dibromoacetylLys‚HBr. The free base was obtained by neutral-
ization with NaOH, followed by decolorization with activated
charcoal and concentration in vacuo [yield 210 mg (47%); mp
HP LC a n d Sp ectr oscop y. HPLC analysis was done using
a Hitachi 7100 or a Spectra-Physics 8700 pumping system. Mass
spectra were measured with Finnigan TSQ7000 instruments,
using either direct injection (synthetic chemicals) or HPLC
introduction (albumin products), in an electrospray/positive ion
mode. NMR spectra were collected on Bruker 300 and 400 MHz
instruments in the Vanderbilt facility. The HOD signal was
1
72 °C (from C
2
H
5
OH, decomp)]: 1H NMR (D
and δ-H), 1.90 (m, 1H, â-H
t, 2H, J ) 7.5 Hz, ꢀ-Η), 4.19 (dd, 1H, J ) 5.0 and 8.1 Hz, R-H),
2
O) δ 1.43 (m, 2H,
), 2.99
γ-Η), 1.63-1.82 (m, 3H, â-Η
(
6
1
2
1
3
2 2
.28 (s, 1H, -CHBr ); C NMR (D O) δ 22.6 (γ), 26.9 (δ), 31.5
(
â), 37.0 (-CHBr
2
), 39.9 (ꢀ), 56.2 (R), 167.3 (-COCHBr
2
), 178.4
adjusted to δ 4.79 in D
2
O solutions. CH
3
OH (49.5 ppm) was used
O solutions).
solutions.
+
as an internal reference for 3C NMR spectra (D
1
(-CO
2
H); MS m/z 345, 347, 349 (MH ).
2
5
(
CH Si was used as an internal standard in CDCl
3
)
4
3
Syn th esis of N -Dibr om oa cetylOr n . (This compound was
used as an internal standard in the analysis of Lys residues in
albumin modified with TBE oxide.) The procedure used was the
Syn th esis of TBE Oxid e. Distilled TBE was oxidized with
m-chloroperbenzoic acid using the same procedure described for
TCE oxide (20, 36), for 100 min. Analysis of the reaction for
TBE oxide using 4-(4-nitrobenzyl)pyridine reaction (37) indi-
cated that TBE oxide formation had peaked at this time. The
product was distilled in vacuo (bp 60-66 °C, 15 Torr) to yield a
colorless liquid, in residual TBE (the most concentrated fraction
of TBE oxide was obtained at 60-66 °C). The product was stable
6
same as for preparations of N -dibromoacetylLys (vide supra).
L-Orn‚HCl (5.5 nmol) was stirred with 2.1 equiv of NaOH and
1
.3 equiv of Br
2
CHCOCl in 3.4 mL of H
temperature. The reaction was neutralized with 1 N HCl and
concentrated in vacuo. C OH was added to the residue.
Cooling at 4 °C yielded a precipitate, which was collected and
2
O for 1.6 h at room
2
H
5
1
dried over P
NMR (D O) δ 1.66 (m, 2H, γ-Η), 1.88 (m, 2H, â-Η), 3.32 (t, 2H,
J ) 6.6 Hz, δ-Η), 3.75 (t, 1H, J ) 6.0 Hz, R-Η), 6.16 (s, 1H,
2 5
O (320 mg, 18% yield): mp 164 °C (decomp); H
in CDCl
remove acid and acyl bromide impurities. The H NMR spectrum
CDCl ) showed a proton singlet at δ 5.42 (along with the TBE
singlet at δ 7.00) [cf. δ 5.31 for TCE oxide (19)]. The C NMR
spectrum (in CDCl ) yielded peaks at 55.5 and 64.7 ppm, which
are assigned to the CHBr and CBr
respectively (with the CHBr and CBr
3
and could be washed (quickly) with cold 6 N NaOH to
1
2
(
3
1
3
1
3
-CHBr
2
); C NMR (D
2
O) δ 24.5 (γ), 28.4 (â), 36.9 (-CHBr
2
), 40.2
(
3
δ), 55.0 (R), 168.2 (-COCHBr
2
), 175.0 (-CO
35 (MH ), 353, 355, 357 (MNa ).
2
H); MS m/z 331, 333,
3
+
+
2
carbons of TBE oxide,
carbon signals of TBE
An a lysis of TBE a n d TCE Oxid e Hyd r olysis Kin etics
2
found at 92.8 and 110.4 ppm, respectively) [cf. TCE and TCE
oxide shifts in ref (19)]. The concentration of the TBE oxide in
the NMR tube was calculated with the addition of a known
amount of TCE (δ 6.46 singlet), with integration of the proton
a n d P r od u ct s. The general methods were as previously
described (20, 21).
The kinetics of hydrolysis of TBE oxide were examined using
the following procedure. To 1.0 mL of 100 mM potassium MOPS

1
416 Chem. Res. Toxicol., Vol. 15, No. 11, 2002
Yoshioka et al.
(
pH 7.4) in an ice bath was added 80 µL of TBE oxide (in dry
Ta ble 1. Decom p osition P r od u cts of TCE a n d
TBE Oxid esa
CH
3
CN), with stirring (final concentration of TBE oxide ∼0.8
mM). At varying times, 40 µL aliquots of the solution were added
to 500 µL of a solution of 4-(4-nitrobenzyl)pyridine reagent (16,
yield (µmol/µmol of epoxide)
product
TCE oxide
TBE oxide
1
7, 19, 20). The mixture was reacted at room temperature for 5
min, and then 500 µL of a mixture of (C N/acetone (1:1, v/v)
was added. A560 measurements were used to determine the
b
2
H
5
)
3
CHX CO H
2
2
0.18
0.53
OHCCO2H
0.08
0.08
0.16
0.37
<0.004
-
1
-1
HCO H
0.382
remaining TBE oxide (ꢀ560 ) 19 800 M cm ), fitting the data
to a plot for (pseudo) first-order decomposition.
2
CO
0.422
<0.001
CX3CHOb
3
In the analysis of hydrolysis products, 40 µL of a CH CN stock
rate of hydrolysisc
0.0042 s-1 d
0.0114 s-1
solution of TBE or TCE oxide (1.8 µmol) was added to 1.0 mL
of potassium 3-(N-morpholino)propanesulfonate (MOPS) buffer
at 0 °C. TBE was removed by extraction of the solution with
(0.0002
(0.0018
a
Incubations were done at 0 °C for 10 min in 0.10 M potassium
1
.0 mL of (C
2
H
5
)
2
O, and 30 µL of the aqueous phase was injected
b
MOPS buffer (pH 7.4). X ) Cl for TCE oxide, X ) Br for TBE
onto a 4.6 × 250 mm octadecylsilane column (5 µm, Beckman,
oxide. c Estimated using 4-(4-nitrobenzyl)pyridine reagent. d At 0
-
1
San Ramon, CA) equilibrated with 10 mM tetra-n-butylammo-
°C. Compare with 0.0069 s reported previously (20).
nium sulfate (pH 6.0) (for analysis of HCO
2
H and glyoxylic acid)
or the same buffer plus 20% CH CN (v/v) for the analysis of
3
-
Rea ction of TBE Oxid e w ith Albu m in . The indicated
amounts of TBE oxide, diluted to 50 µL in dry CH CN (to give
3
final concentrations of 8 and 20 mM TBE oxide), were added to
a solution of bovine serum albumin (1.0 mg in 200 µL of 250
mM potassium MOPS buffer, pH 7.4). The samples were
incubated at 0 °C for 30 min and then evaporated to dryness
1
Br
2
CHCO
2
H (flow rate 1.2 mL min , A210 detection in all cases).
Br CCHO (tribromoacetaldehyde, bromal) was analyzed by
3
treating 1.0 mL aliquots of the hydrolyzed samples with 2,4-
dinitrophenylhydrazine and extraction as described for HCHO
previously (39). The derivatized extracts, including a set of
standards of known concentration, were analyzed by HPLC
2
under an N stream. The residue was dissolved in 250 µL of
[
6.2 × 80 mm Zorbax octadecylsilane column, 3 µm (MacMod,
potassium MOPS buffer (pH 7.4). To the mixture was added 50
Chadds Ford, PA), 54% aqueous CH CN (v/v), flow rate 2.5 mL
min , A360].
3
5
µL of a solution containing both 4 mM N -formylOrn (23) and 4
-
1
5
mM N -dibromoacetylOrn (vide supra). The albumin was di-
CO was analyzed in a closed gas-phase circuit using an
electrochemical sensor (ERMF-0503, Draeger, Pittsburgh, PA),
as described earlier (20), with calibration using standardized
amounts of CO.
gested with proteinase K overnight at 37 °C under Ar, and the
digest was derivatized with phenylisothiocyanate as described
elsewhere (23).
Aliquots of the modified amino acid digest were analyzed by
combined HPLC/electrospray MS in the general procedure
described for TCE oxide products (23). Internal standard
Micr osom a l In cu ba tion s. Liver microsomes prepared from
-
1
phenobarbital-treated rats (4.6 mg of protein mL ) (40) were
incubated with an NADPH-generating system (40) and 10 mM
TBE (added without organic solvent) in 0.10 M potassium MOPS
buffer (pH 7.4). The flask was shaken in air at 37 °C, and 1.0
mL aliquots were withdrawn at various time points, quenched
6
calibration curves were prepared using varying amounts of N -
6
formylLys and N -dibromoacetylLys, carried through the de-
5
rivatization and HPLC/MS steps with fixed amounts of N -
5
formylOrn and N -dibromoacetylOrn. The HPLC and MS
with the addition of 0.15 mL of 25% HClO
Insoluble protein was removed by centrifugation at 3000g for
0 min, and 1.0 mL of the supernatant was analyzed for
Br CCHO, using 2,4-dinitrophenylhydrazone as described above.
In separate incubations, 5.0 mL aliquots of a similar incuba-
tion were withdrawn after varying incubation times and quenched
by the addition of H SO to 2% (w/v). Precipitated protein was
4
, and chilled.
conditions were identical to those described for PCE (21), using
a Zorbax octylsilane column (4.6 × 250 mm) connected to a
Finnegan TSQ7000 instrument operating in the electrospray
1
+
3
mode, with single ion monitoring (MH ion in each case). A 10
mM NH
4 3 2
CH CO buffer (pH 6.5) was used with an increasing
CH OH gradient. The appropriate m/z peaks were integrated,
3
and the internal standard curves (based on Orn derivatives)
were used to do calculations.
2
4
removed by centrifugation at 3000g for 10 min, and the
supernatants (pH <2) were extracted 5× with 5 mL of
3 2 2
(CH CH ) O. The pooled organic extracts were concentrated to
Resu lts
dryness under an N
ion-pair HPLC (A210, vide supra) was unsuccessful because of
interfering peaks. Br CHCO H in the extracts was derivatized
2 2 2
stream. Direct analysis of Br CHCO H by
Hyd r olysis of TBE Oxid e. TBE oxide was prepared
as described, using a general approach used previously
for VDC and TCE oxides (18, 20, 36). We were unable to
obtain a mass spectrum due to the instability of the
2
2
using 4-nitrophenacyl bromide (Aldrich) according to Morozo-
wich and Douglas (19, 41). A 0.10 M solution of 4-nitrophenacyl
bromide (200 µL) and 1 µL of N,N-diisopropyl-N-ethylamine was
added to each sample, and the p-nitrophenacyl ester product
was analyzed after 15 min at room temperature, using HPLC
1
13
compound, but the H and C NMR spectra and reactiv-
ity are consistent with the nature of the product.
The rate of hydrolysis (at 0 °C and pH 7.4) was
estimated using 4-(4-nitrobenzyl)pyridine reagent and
[
6.2 × 80 mm Zorbax octadecylsilane column, 3 µm, linear
gradient of 18-45% CH CN in H
2 2
The analyses of Br CCHO and Br CHCO H were done using
3
2
O, A260].
-
1
was 0.0114 ((0.0018) s under these conditions (Table
1) (20). This rate is about 3 times faster than that of TCE
oxide (measured in a parallel experiment) (Table 1). The
rate corresponds to a t1/2 of 61 s (pseudo-first-order
reaction at 0 °C) and would be considerably faster at
higher temperatures (estimated ∼5 s at 23 °C, based on
behavior of TCE oxide) (19, 20).
3
comparisons with standards prepared using the same procedure.
Benzene extracts of NADPH-fortified microsomal incubations
were analyzed for TBE oxide using 4-(4-nitrobenzyl)pyridine as
described elsewhere (19).
Rea ction of TBE Oxid e w ith Lys. In the reaction with Lys,
1
3
00 µL of 100 mM TBE oxide (in dry CH CN) was added to 900
µL of an aqueous solution of (10 mM) L-Lys (free base) on ice,
and reaction was allowed to occur for 30 min. The samples were
The products were identified and quantified by HPLC
with the exception of CO, which was measured with a
gas-selective electrode (Table 1). The products of TCE
oxide hydrolysis have been measured previously (19, 20)
and were analyzed again here in parallel experiments
(Table 1). More than 95% of the TBE oxide products were
washed with an equal volume of (C
was removed under an N stream. The aqueous samples were
diluted 5-fold with 10 mM NH CH CO buffer (pH 6.5) and
analyzed using a 4.6 × 250 mm Beckman octadecylsilane HPLC
column (5 µm), with an (isocratic) buffer of 10 mM NH CH CO
pH 6.5) at a flow rate of 0.7 mL min-1 (A210).
2 5 2 2 5 2
H ) O, and residual (C H ) O
2
4
3
2
4
3
2
(
2
accounted for, except for the low recovery of HCO H. Only

Tribromoethylene Oxide
Chem. Res. Toxicol., Vol. 15, No. 11, 2002 1417
F igu r e 1. Products of P450-catalyzed oxidation of TBE. Liver
F igu r e 2. Products of reaction of TBE oxide with Lys. TBE
microsomes prepared from phenobarbital-treated rats (4.6 mg
-
1
oxide and Lys (100 mM each) were mixed in H O, and the
of protein mL ) were incubated with TBE (10 mM) and an
NADPH-generating system in 0.10 M potassium MOPS buffer
2
reaction proceeded at 0 °C for 30 min. An aliquot of the reaction
was diluted with the HPLC solvent and analyzed by HPLC as
described under Experimental Procedures. The identities of the
peaks were identified by co-chromatography with standard
reference materials and HPLC/MS in a similar chromatography
system. The first two peaks (near void volume) are unidentified.
(
pH 7.4) at 37 °C; the products were analyzed as described under
Experimental Procedures. (9) Br CCHO; (O) Br CHCO H.
3
2
2
trace Br CCHO was detected (<0.4%), and the possibility
3
that this had been formed slowly in the TBE oxide by
rearrangement prior to hydrolysis cannot be excluded.
The yield of glyoxylic acid from TBE oxide is very similar
as from TCE oxide. CO and HCO
as in the case with TCE oxide. In the case of TCE oxide,
the yield of HCO H was nearly equivalent to that of CO
in this (Table 1) and previous work. However, here we
2
H were both formed,
2
2
recovered only 50% HCO
2
H relative to CO. The yield of
CHCO
Br CHCO H was clearly higher than that of Cl
2
2
2
2
H
recovered here in the parallel experiment or in any of
the previous studies in this laboratory (19, 20).
Micr osom a l Oxid a tion of TBE. Liver microsomes
(
prepared from rats treated with phenobarbital) were
used as a source of P450 and incubated with TBE (Figure
). Br CHCO H, detected as the p-nitrophenacyl ester,
was the major product detected (Figure 1). Br CCHO was
1
2
2
3
formed, detected, and measured as the 2,4-dinitrophen-
ylhydrazone. The rates of formation of these products are
F igu r e 3. Formation of Lys adducts in bovine serum albumin
modified with TBE oxide. Reactions were done at 0 °C for 30
min in 250 mM potassium MOPS buffer (pH 7.4) with bovine
much lower than for formation of Cl
other products, Cl CHCO H, not detected) from TCE
measured under similar conditions (19) and less than
that of Cl CCO H formed from PCE (21).
3
CCHO (major and
-
1
serum albumin (4 mg mL ) and the indicated concentrations
of TBE oxide. Lys adducts were quantified (followed proteinase
K digestion and derivatization with phenylisothiocyanate) using
2
2
5
5
HPLC/MS. N -FormylOrn and N -dibromoacetylOrn were used
3
2
6
6
to generate internal standard curves for N -formylLys and N -
dibromoacetylLys. Results are shown for the formation of N -
dibromoacetylLys (9) and N -formylLys (O).
A benzene extract of an incubation containing a higher
concentration of microsomal protein yielded a 4-(4-
nitrobenzyl)pyridine adduct, as judged by the character-
istic λmax of the purple adduct observed in the presence
of (C H ) N under nonaqueous conditions (19). However,
2 5 3
under the conditions used in the assays of Figure 1, the
6
6
adducts were analyzed (HPLC/MS) after proteinase K
digestion and derivatization with phenylisothiocyanate
(Figure 3). In the protein, only the N6 amino groups can
concentration of TBE oxide formed (measured over 15-
be modified (barring the possibility of an N-terminal Lys
1
50 s) was <1 µM (corresponding to <0.01 A520 in the
6
N2 amino group in a protein). The amount of N -
assays).
Rea ction of TBE Oxid e w ith Lys. The reactions of
TBE oxide with free Lys in solution yielded a mixture of
acylated Lys derivatives (Figure 2)(plus Br CHCO
6
dibromoacetylLys was 2-3× greater than that of N -
formylLys.
2
2
H
Discu ssion
resulting from hydrolysis of TBE oxide). The major Lys
products were the N2 derivatives, consistent with the
lower pK of the Lys N2 nitrogen relative to the N6 atom
a
A central focus of the work described in this report was
the examination of the halide order in reactions of
polyhalooxiranes. Specifically, reactions involving TBE
oxide were compared to those with TCE oxide. The
results, along with previous work, support some rela-
tively general hypotheses regarding the reactions of
chlorinated and brominated epoxides.
(
42). The amount of the dibromoacetyl product was ∼2×
greater than that of the formyl.
Rea ction of TBE Oxid e w ith Albu m in . TBE oxide
was reacted with bovine serum albumin, and the Lys
2
An error of (5% is estimated to be associated with the precision
The results obtained with NADPH-fortified microsomal
incubations of TBE contrast with those with TCE, with
Br CCHO as a minor product and Br CHCO H as a major
3 2 2
one (Figure 1). The rate of oxidation of TBE was much
less than that of TCE (in liver microsomes prepared from
2
of measurement of both CO and HCO H. However, accurate compari-
son of each relative to each other is probably more difficult due to the
inherent error in the calibration of each assay. In the present
experiments with TCE oxide (Table 1), the CO:HCO
unity. However, in previous work, the ratio has varied from 1.4 to 0.6
19, 20).
2
H ratio was nearly
(

1
418 Chem. Res. Toxicol., Vol. 15, No. 11, 2002
Yoshioka et al.
phenobarbital-treated rats) (19), as judged by the rates
of formation of the characterized products. It is of interest
Sch em e 2. P r op osed Mech a n ism of Hyd r olysis of
TBE Oxid e (20)
to note that the rate of oxidation of TBE to Br
and HCO
3
CCHO
(
2
H) was reported to be ∼1% that of TCE for
the bacterial enzyme methane monooxygenase (32). A
possible explanation is that the bulkier bromine atoms
restrict access of the (FeO complex of the) oxygenases to
the olefin. In the case of TCE (19), we were able to
measure steady-state levels of the epoxide in microsomal
incubations and found that these were kinetically incon-
sistent with an obligatory role of TCE epoxide in the
3
formation of Cl CCHO (19). With TBE such an argument
cannot be presented because of the lower rates of forma-
tion of products (Figure 1) plus the higher rate of
hydrolysis of TBE oxide (Table 1). Thus, although we
were only able to provide qualitative evidence for the
microsomal oxidation of TBE to TBE oxide, we could
not quantify it (below a level of 1 µM), and the possi-
bility that it is an intermediate in the formation of
Br
did not form Br
TBE to Br CCHO is attributed to pathway a in Scheme
. Br CHCO H could be formed by pathway c or d or by
2
CHCO
2
H cannot be dismissed. However, TBE oxide
3
CCHO, and the enzymatic oxidation of
3
1
2
2
hydrolysis of the epoxide.
TBE oxide was synthesized in this work, apparently
for the first time. The same method we employed for the
synthesis of TCE oxide (20, 36) was used, and the spectral
and other properties are in order with those expected for
such a halooxirane. TBE oxide hydrolyzed ∼3× faster in
2
H O than did TCE oxide (Table 1). The effect of substitu-
tion of bromine for chlorine was similar to that seen with
VC and VB (17, 29). The pattern of hydrolysis products
seen with TBE oxide was shifted from 1-carbon products
Sch em e 3. Hyd r olysis P r od u cts of Ha looxir a n es
2 2
to Br CHCO H, relative to TCE oxide, as judged by a
direct comparison to work done either with TCE oxide
in parallel experiments or with previous analyses in this
2
laboratory (19, 20). The results reflect some error in the
difficulty of making several difficult measurements. The
fraction of the total products recovered as X
X ) halogen) was 27% in the case of TCE oxide and 62%
in the case of TBE oxide, with normalization for the
number of carbons (Table 1). The HCO H production was
2 2
CHCO H
(
2
decreased relative to CO, although the difference may
reflect some error in the CO measurements.2
2 2
The shift from 1-carbon products to Br CHCO H as-
sociated with substitution of bromine for chlorine appears
to be consistent with the scheme developed earlier
(
labeling patterns (20). Pathway B of Scheme 2 may be
facilitated due to the properties of bromide as a leaving
pH, C-C bond scission is a dominant reaction. The major
1
-carbon products obtained with the halooxiranes are
Scheme 2) for TCE oxide on the basis of 18O and 2
H
shown in Scheme 3. CO is a product in all cases and can
be rationalized with Scheme 4, generalized to accom-
modate the observations with tri- and tetrahalooxiranes
2 2
group, yielding more Br CHCO H.
(18-21). Thus we propose that the gem-dihalo-substi-
The patterns of Lys adducts are also shifted to a
predominance of dibromoacetyl adducts relative to TCE
oxide (Figures 2 and 3). The change is most clearly seen
tuted carbon yields CO and that the valence of the second
carbon dictates the course of the other product and
determines its oxidation state, yielding HCHO, HCO
2
H,
6
6
in the case of albumin, where the N -dihaloacetylLys:N -
formylLys ratio is shifted from the value of ∼0.1 observed
for TCE oxide (23) to 2-3 (Figure 3).
or CO . The leaving group ability of the halogen can alter
2
the balance between C-C scission and halide migration
(e.g., see Table 1).
This pathway can be generalized for other haloox-
iranes, at least for the gem-substituted ones. At very low
pH values, hydrolysis of TCE and VDC oxides is domi-
nated by formation of 2-carbon carboxylic acids (18, 19).
However, throughout most of the pH range, the composi-
tion of TCE oxide products is relatively invariant, when
phosphate trapping is avoided (20, 21). At neutral to high
VDC oxide requires some extra explanation because
both HCHO and HCO H are produced, along with CO
2
(18). In terms of oxidation state, CO and HCO H are
2
equivalent and in one sense differ only in their hydration.
We previously proposed formyl chloride as a precursor
of HCO
2
H (19) and subsequently rejected it as an
intermediate (20). Formyl chloride is generally dehydro-

Tribromoethylene Oxide
Chem. Res. Toxicol., Vol. 15, No. 11, 2002 1419
Sch em e 4. P r op osed Gen er a l Mech a n ism of
Hyd r olysis of Tr i- a n d Tetr a -h a looxir a n es
Sch em e 6. P r op osed Gen er a l Mech a n ism of
Rea ction of Ha logen a ted Ep oxid es w ith Lys (20)
a
their reactivity (Scheme 6) (20). One issue that first
surfaced in our first work with TCE (19) is the reaction
of intermediates with phosphate (Scheme 5). The pres-
ence of phosphate in aqueous systems had the effect of
diminishing the production of some products of TCE oxide
(19), VDC oxide (18), and PCE oxide (21). The effect was
explained in the case of PCE oxide, where oxalyl phos-
phate was identified in 0.1 M phosphate reactions (21).
Oxalyl phosphate is not stable indefinitely but slowly
hydrolyzes to oxalic acid, without reacting with nucleo-
philes such as Lys (21). We have not identified any
organic phosphates generated from VDC oxide or TCE
oxide but presume that the reactions are similar; i.e., one
might expect to find products such as glyoxyl phosphate
from TCE oxide or TBE oxide (Schemes 2 and 4). These
reactions with phosphate should be considered experi-
mental artifacts, and phosphate and other potential
nucleophilic buffers should be avoided in experiments
with these halooxiranes. The relevance of the phosphate
reaction to reactions with the phosphate groups of nucleic
acids is not clear. In the case of TCE, we have shown
that unstable products are formed in the reaction of TCE
oxide with oligonucleotides, but the possibility that the
reaction is with phosphates instead of the nucleic acid
bases has not been addressed (30).
a
R1 ) H or X (halogen); Pi ) phosphate ion. See (20, 21).
Sch em e 5. Sch em es for Hyd r olysis of VDC Oxid e
To Exp la in Obser ved P r od u cts (18): (A) HCHO +
2
CO; (B) HCHO + HCO H
halogenated to CO except at high pH (>13) (43). Further,
the isotopic labeling patterns we found with TCE oxide
rejected this possibility (20). An explanation for the
hydrolysis of VDC oxide is presented in Scheme 5. In part
A, the formation of CO and HCHO is explained by the
general mechanism presented in Schemes 2-4. A revers-
ible hydration step is proposed in part B, yielding HCHO
In conclusion, we have extended studies of haloox-
iranes to TBE oxide, an apparent oxidation product of
TBE. We considered the halide order in the context of
several previous studies with chlorooxiranes and inter-
pret reactivity in the context of a general mechanism first
proposed for TCE oxide (20). The results are generally
consistent with a set of general paradigms (Schemes 2
and 4) for gem-substituted dihalooxiranes but probably
do not apply to monohalooxiranes. This paradigm leads
to general predictions about the reactions of some vinyl
halides. However, the point should be emphasized that
generation of acyl halides in P450 enzymes in paths
independent of halooxirane formation (Scheme 1) will
alter the balance of reactive products, and therefore the
halooxiranes cannot be considered the only factors in the
prediction of reactivity of vinyl halides.
2
and HCO H.
The above considerations apply to the gem-dihaloox-
iranes discussed here and should apply to others. In the
case of monohalogenated oxiranes (from VC and VB), the
principles do not appear to be operative. CO has not been
detected as a product, and hydrolysis appears to involve
simple halide migration (17). In the context of the
enzymatic oxidations, we have no proof that a 1,2-shift
occurs with a P450-VC (or -VB) intermediate but
cannot rule this out. Our current view is that the
paradigm of Schemes 2 and 4 will only apply to gem-
dihalooxiranes such as VDC, TCE, TBE, and PCE.
The reactions of TBE oxide with Lys are assumed to
be similar to those previously defined by isotopic labeling
Ack n ow led gm en t. This work was supported in part
by U.S. Public Health Service Grants R01 ES10546 and
P30 ES00267. J .A.K. was supported in part by U.S.
Public Health Service Training Grant T32 ES07028. We
thank M. Voehler and L. Manier for technical assistance
with several aspects of NMR and mass spectrometry,
respectively.
6
6
with TCE oxide, yielding N -formyl- and N -dihalo-
acetylLys. The increased fraction of dihaloacetyl adducts
with TBE oxide (relative to N-formyl adducts, more
abundant with TCE oxide) is postulated to reflect the
concentrations of the individual acylhalides rather than

1
420 Chem. Res. Toxicol., Vol. 15, No. 11, 2002
Yoshioka et al.
(
24) Benya, T. J ., Busey, W. M., Dorato, M. A., and Berteau, P. E.
Refer en ces
(1982) Inhalation carcinogenicity bioassay of vinyl bromide in rats.
(
1) United States Department of Health and Human Services (1993)
Toxicological profile for tetrachloroethylene. Public Health Ser-
vice, Agency for Toxic Substances and Disease Registry.
2) Apfeldorf, R., and Infante, P. F. (1981) Review of epidemiologic
study results of vinyl chloride-related compounds. Environ. Health
Perspect. 41, 221-226.
3) Spirtas, R., and Kaminski, R. (1978) Angiosarcoma of the liver
in vinyl chloride/poly(vinyl chloride) workers: 1977 update of the
NIOSH register. J . Occup. Med. 20, 427-429.
Toxicol. Appl. Pharmacol. 64, 367-379.
(25) Ottenw a¨ lder, H., Laib, R. J ., and Bolt, H. M. (1979) Alkylation of
RNA by vinyl bromide metabolites in vitro and in vivo. Arch.
Toxicol. 41, 279-286.
26) Ottenw a¨ lder, H., and Bolt, H. M. (1980) Metabolic activation of
vinyl chloride and vinyl bromide by isolated hepatocytes and
hepatic sinusoidal cells. J . Environ. Pathol. Toxicol. 4, 411-417.
(
(
(
(
(
27) Bolt, H. M., Filser, J . G., and Hinderer, R. K. (1978) Rat liver
1
4
microsomal uptake and irreversible protein binding of [1,2- C]-
vinyl bromide. Toxicol. Appl. Pharmacol. 44, 481-489.
(4) National Cancer Institute (1977) Bioassay of tetrachloroethylene
for possible carcinogenicity. Natl. Toxicol. Program Tech. Rep.,
28) Barbin, A., Br e´ sil, H., Croisy, A., J acquignon, P., Malaveille, C.,
Montesano, R., and Bartsch, H. (1975) Liver-microsome-mediated
formation of alkylating agents from vinyl bromide and vinyl
chloride. Biochem. Biophys. Res. Commun. 67, 596-603.
(29) Guengerich, F. P., Mason, P. S., Stott, W. T., Fox, T. R., and
Watanabe, P. G. (1981) Roles of 2-haloethylene oxides and
2-haloacetaldehydes derived from vinyl bromide and vinyl chloride
in irreversible binding to protein and DNA. Cancer Res. 41, 4391-
4398.
(30) Cai, H., and Guengerich, F. P. (2001) Reaction of trichloroethylene
oxide with proteins and DNA: instability of adducts and modula-
tion of functions. Chem. Res. Toxicol. 14, 54-61.
31) Shakhnazaryan, G. M. (1974) Molecular rearrangements. XI.
Factors influencing the oxidation of tribromoethylene to dibro-
moacetic acid during the oxidation of peracetic acid. Arm. Khim.
Zh. 27, 177-181.
7
7-813.
(
(
(
(
5) National Cancer Institute (1986) Carcinogenesis bioassay of
tetrachloroethylene. Natl. Toxicol. Program Tech. Rep. 232.
6) Reichert, D. (1983) Biological actions and interactions of tetra-
chloroethylene. Mutat. Res. 123, 411-429.
7) Haley, T. J . (1975) Vinylidene chloride: a review of the literature.
Clin. Toxicol. 8, 633-643.
8) Szabo, S., J aeger, R. J ., Moslen, M. T., and Reynolds, E. S. (1977)
Modification of 1,1-dichloroethylene hepatotoxicity by hypothy-
roidism. Toxicol. Appl. Pharmacol. 42, 367-376.
(
9) Quast, J . F., Humiston, C. G., Wade, C. E., Ballard, J ., Beyer, J .
E., Schwetz, R. W., and Norris, J . M. (1983) A chronic toxicity
and oncogenicity study in rats and subchronic toxicity study in
dogs on ingested vinylidene chloride. Fundam. Appl. Toxicol. 3,
(
(
(
(
5
5-62.
(
(
(
10) J ones, R. B., and Mackrodt, W. C. (1983) Structure-genotoxicity
relationship for aliphatic epoxides. Biochem. Pharmacol. 32,
32) Fox, B. G., Borneman, J . G., Wackett, L. P., and Lipscomb, J . D.
2
359-2362.
(1990) Haloalkene oxidation by the soluble methane monooxyge-
11) J acobson-Kram, D. (1986) The reproductive Effects Assessment
Group’s review of the mutagenicity of vinylidene chloride. Envi-
ron. Mutagen. 8, 161-169.
12) Dekant, W., Vamvakas, S., Koob, M., K o¨ chling, A., Kanhai, W.,
M u¨ ller, D., and Henschler, D. (1990) A mechanism of haloalkene-
induced renal carcinogenesis. Environ. Health Perspect. 88, 107-
nase from Methylosinus trichosporium OB3b: mechanistic and
environmental implications. Biochemistry 29, 6419-6427.
33) Marshall, R. A., Harper, D. B., McRoberts, W. C., and Dring, M.
J . (1999) Volatile bromocarbons produced by Falkenbergia stages
of Asparagopsis spp. (Rhodophyta). Limnol. Oceanogr. 44, 1348-
1
352.
1
10.
34) Kennedy, C. H., Cohen, K. B., Bechtold, W. E., Chang, I. Y.,
Eidson, A. F., Dahl, A. R., and Henderson, R. F. (1993) Effect of
dose on the metabolism of 1,1,2,2-tetrabromoethane in F344/N
rats after gavage administration. Toxicol. Appl. Pharmacol. 119,
(
13) Anders, M. W., Dekant, W., and Vamvakas, S. (1992) Glutathione-
dependent toxicity. Xenobiotica 22, 1135-1145.
14) Greim, H., Bonse, G., Radwan, Z., Reichert, D., and Henschler,
D. (1975) Mutagenicity in vitro and potential carcinogenicity of
chlorinated ethylenes as a function of metabolic oxirane forma-
tion. Biochem. Pharmacol. 24, 2013-2017.
(
2
3-33.
(35) Canton, J . H., and Wegman, R. C. C. (1983) Studies on the toxicity
of tribromoethene, cyclohexene and bromocyclohexane to different
freshwater organisms. Water Res. 17, 743-747.
(
(
(
15) Henschler, D. (1977) Metabolism and mutagenicity of halogenated
olefinssa comparison of structure and activity. Environ. Health
Perspect. 21, 61-64.
(36) Iyer, R. S., and Harris, T. M. (1993) Preparation of aflatoxin B
1
16) Guengerich, F. P., and Watanabe, P. G. (1979) Metabolism of [14C]-
and [36Cl]-labeled vinyl chloride in vivo and in vitro. Biochem.
Pharmacol. 28, 589-596.
8,9-epoxide using m-chloroperbenzoic acid. Chem. Res. Toxicol.
6, 313-316.
37) Epstein, J ., Rosenthal, R. W., and Ess, R. J . (1955) Use of p-(4-
nitrobenzyl)pyridine as analytical reagent for ethylenimines and
alkylating agents. Anal. Chem. 27, 1435-1439.
(
17) Guengerich, F. P., Crawford, W. M., J r., and Watanabe, P. G.
(1979) Activation of vinyl chloride to covalently bound metabo-
lites: roles of 2-chloroethylene oxide and 2-chloroacetaldehyde.
Biochemistry 18, 5177-5182.
(
38) Coughlin, D. J ., and Wood, R. (1994) Preparation of glycyltyrosyl-
(N-diethylenetriaminepentaacetic acid)lysine hydrochloride (gyk-
(
18) Liebler, D. C., and Guengerich, F. P. (1983) Olefin oxidation by
cytochrome P-450: evidence for group migration in catalytic
intermediates formed with vinylidene chloride and trans-1-
phenyl-1-butene. Biochemistry 22, 5482-5489.
dtpa). U. S. Patent Appl. 92-890519, cited as Statutory Invent.
Regist.
(
39) Miller, G. P., and Guengerich, F. P. (2001) Binding and oxidation
of alkyl 4-nitrophenyl ethers by rabbit cytochrome P450 1A2:
evidence for two binding sites. Biochemistry 40, 7262-7272.
40) Guengerich, F. P. (2001) Analysis and characterization of enzymes
and nucleic acids. In Principles and Methods of Toxicology (Hayes,
A. W., Ed.) pp 1625-1687, Taylor & Francis, New York.
(
19) Miller, R. E., and Guengerich, F. P. (1982) Oxidation of trichlo-
roethylene by liver microsomal cytochrome P-450: evidence for
chlorine migration in a transition state not involving trichloro-
ethylene oxide. Biochemistry 21, 1090-1097.
(
(
(
(
20) Cai, H., and Guengerich, F. P. (1999) Mechanism of aqueous
decomposition of trichloroethylene oxide. J . Am. Chem. Soc. 121,
(41) Morozowich, W., and Douglas, S. L. (1975) Resolution of pros-
taglandin p-nitrophenacyl esters by liquid chromatography and
conditions for rapid, quantitative p-nitrophenzylation. Prosta-
glandins 10, 19-33.
1
1656-11663.
21) Yoshioka, T., Krauser, J . A., and Guengerich, F. P. (2002)
Tetrachloroethylene oxide: hydrolytic products and reactions with
phosphate and lysine. Chem. Res. Toxicol. 15, 1096-1105.
22) Miller, R. E., and Guengerich, F. P. (1983) Metabolism of
trichloroethylene in isolated hepatocytes, microsomes, and re-
constituted enzyme systems containing cytochrome P-450. Cancer
Res. 43, 1145-1152.
(42) Mahler, H. R., and Cordes, E. H. (1971) Biological Chemistry, 2nd
ed., 45, Harper & Row, New York.
(
43) Dowideit, P., Mertens, R., and von Sonntag, C. (1996) Non-
hydrolytic decay of formyl chloride into CO and HCl in aqueous
solution. J . Am. Chem. Soc. 118, 11288-11292.
(
23) Cai, H., and Guengerich, F. P. (2000) Acylation of protein lysines
by trichloroethylene oxide. Chem. Res. Toxicol. 13, 327-335.
TX020047T