Immunochemical detection of protein adducts in mice treated with trichloroethylene

Article abstract of DOI:10.1021/tx950171v

Trichloroethylene has been shown to produce tumors in rodents and is a suspect human carcinogen. In addition, a number of case reports raise the possibility that trichloroethylene can induce an autoimmune disorder known as systemic sclerosis. To investigate whether covalent binding of reactive trichloroethylene metabolites may be involved in the mechanisms underlying these toxic responses, we have developed a polyclonal antibody that can recognize trichloroethylene-protein adducts in tissues. The antibody was prepared by immunizing a rabbit with dichloroacetic anhydride-modified keyhole limpet hemocyanin. Enzyme-linked immunosorbent assay data indicated that the serum antibody recognized dichloroacetic anhydride-modified rabbit serum albumin, but not unmodified protein. In addition, N(ε)-dichloroacetyl- L-lysine was the most potent inhibitor of antibody binding to dichloroacetic anhydride-modified rabbit serum albumin, indicating that the antibody recognizes primarily dichloroacetylated lysine residues. Immunoblots revealed the presence of two major trichloroethylene adducts at 50 and 100 kDa in liver microsomal fractions from male B6C3/F1 mice treated with trichloroethylene. The formation of trichloroethylene adducts was both dose and time dependent. Furthermore, the 50-kDa adduct was found to comigrate on a polyacrylamide gel with cytochrome P450 2E1. These data show that reactive metabolites of trichloroethylene are formed in vivo and bind covalently to discrete proteins in mouse liver. The data also suggest that one of the protein targets is cytochrome P450 2E1. Further studies will be necessary to elucidate the relationship between covalent binding of trichloroethylene and trichloroethylene toxicity.

Full text of DOI:10.1021/tx950171v

Chem. Res. Toxicol. 1996, 9, 451-456
451
Im m u n och em ica l Detection of P r otein Ad d u cts in Mice
Tr ea ted w ith Tr ich lor oeth ylen e
†
‡
‡
N. Christine Halmes, David C. McMillan, J ohn E. Oatis, J r., and
,
†
Neil R. Pumford*
Division of Toxicology, University of Arkansas for Medical Sciences, Little Rock,
Arkansas 72205-7199, and Department of Pharmacology, Medical University of South Carolina,
Charleston, South Carolina 29425
Received October 5, 1995X
Trichloroethylene has been shown to produce tumors in rodents and is a suspect human
carcinogen. In addition, a number of case reports raise the possibility that trichloroethylene
can induce an autoimmune disorder known as systemic sclerosis. To investigate whether
covalent binding of reactive trichloroethylene metabolites may be involved in the mechanisms
underlying these toxic responses, we have developed a polyclonal antibody that can recognize
trichloroethylene-protein adducts in tissues. The antibody was prepared by immunizing a
rabbit with dichloroacetic anhydride-modified keyhole limpet hemocyanin. Enzyme-linked
immunosorbent assay data indicated that the serum antibody recognized dichloroacetic
ꢀ
anhydride-modified rabbit serum albumin, but not unmodified protein. In addition, N -
dichloroacetyl-L-lysine was the most potent inhibitor of antibody binding to dichloroacetic
anhydride-modified rabbit serum albumin, indicating that the antibody recognizes primarily
dichloroacetylated lysine residues. Immunoblots revealed the presence of two major trichlo-
roethylene adducts at 50 and 100 kDa in liver microsomal fractions from male B6C3/F1 mice
treated with trichloroethylene. The formation of trichloroethylene adducts was both dose and
time dependent. Furthermore, the 50-kDa adduct was found to comigrate on a polyacrylamide
gel with cytochrome P450 2E1. These data show that reactive metabolites of trichloroethylene
are formed in vivo and bind covalently to discrete proteins in mouse liver. The data also suggest
that one of the protein targets is cytochrome P450 2E1. Further studies will be necessary to
elucidate the relationship between covalent binding of trichloroethylene and trichloroethylene
toxicity.
In tr od u ction
liver and lung tumors in mice and kidney tumors in male
rats (2, 4, 5). On the basis of these animal studies, TRI
has been classified as a suspect human carcinogen. In
addition, clinical case reports have indicated an associa-
tion between TRI exposure and autoimmune disease,
such as systemic sclerosis, systemic lupus erythematosus,
and fasciitis (6-9).
1
Trichloroethylene (1,1,2-trichloroethene, TRI) has been
used extensively in industry as a solvent for the degreas-
ing of metals. TRI has also been used as a dry cleaning
agent, a lubricant, and an extractant in food processing
and has had limited use as an inhalation anesthetic. The
annual production of TRI in the United States is about
Several studies have shown that metabolic activation
is a requirement for TRI-induced cytotoxicity and carci-
nogenicity (1, 2, 5); however, the molecular mechanisms
underlying these toxicities are not known. Although the
production of liver tumors in B6C3/F1 mice is thought
to be related to the ability of the TRI metabolite trichlo-
roacetic acid to induce peroxisome proliferation (4), a role
for covalent binding of reactive metabolites has not been
entirely ruled out. TRI has been shown to initiate or
accelerate an autoimmune response in MRL mice (10).
Covalent binding of reactive TRI metabolites to protein
may be responsible for initiating an immune response,
directed against either the hapten (TRI), the protein
1
30 000 metric tons, and it has been estimated that 3.5
million people in this country are occupationally exposed
to TRI (1). Because of its widespread use, TRI has
become a major air pollutant and groundwater contami-
nant and is the most frequently detected organic pollut-
ant at many hazardous waste sites (2).
Acute TRI exposure is known to induce a number of
toxic effects in experimental animals, including CNS
depression, impaired cardiac function, mild hepatotox-
icity, and nephrotoxicity (2); similar effects have been
observed in occupationally exposed humans (3). Chronic
exposure to TRI is known to increase the incidence of
(autoantigen), or a novel epitope on the protein that forms
as a consequence of TRI binding to the protein (11).
TRI is known to undergo initial biotransformation by
both oxidative (i.e., cytochrome P450) and conjugative
(i.e., GSH transferase) pathways, either of which has the
potential to generate reactive metabolites capable of
binding covalently to protein (1, 2, 5). On the basis of in
vitro studies, Miller and Guengerich (12, 13) suggested
that TRI forms an oxygenated intermediate complex with
cytochrome P450 which, on chemical grounds, could be
expected to rearrange to form a number of reactive
*
To whom correspondence should be addressed at University of
Arkansas for Medical Sciences, 4301 W. Markham St., Slot 638, Little
Rock, AR 72205; e-mail address: npumford@biomed.uams.edu.
†
University of Arkansas for Medical Sciences.
Medical University of South Carolina.
Abstract published in Advance ACS Abstracts, February 1, 1996.
Abbreviations: TRI, trichloroethylene; DCA, dichloroacetyl; KLH,
‡
X
1
keyhole limpet hemocyanin; RSA, rabbit serum albumin; ELISA,
enzyme-linked immunosorbent assay; DCA-RSA, dichloroacetylated
rabbit serum albumin; EDC, 1-ethyl-3-[3-(dimethylamino)propyl]car-
bodiimide hydrochloride; SDS-PAGE, sodium dodecyl sulfate-poly-
acrylamide gel electrophoresis; DQ-COSY, double quantum filtered
correlation spectroscopy.
0
893-228x/96/2709-0451$12.00/0 © 1996 American Chemical Society
+
+

4
52 Chem. Res. Toxicol., Vol. 9, No. 2, 1996
Halmes et al.
metabolites including TRI oxide, dichloroacetyl chloride,
and chloral. In addition, conjugation of TRI by GSH
followed by â-lyase-mediated metabolic activation is
thought to result in the generation of reactive thioketene
or thionoacyl chloride derivatives (5), which may be
responsible for the nephrotoxic effects of TRI observed
in male rats.
Anti-DCA serum antibodies were detected by an ELISA as
described previously (18), with the use of dichloroacetylated
rabbit serum albumin (DCA-RSA) as the test antigen. Serial
dilutions of DCA-RSA (50 µL) in 0.06 M carbonate buffer (pH
9
.6) were applied to microtiter plate wells and incubated
overnight at 4 °C. The wells were washed, and rabbit serum
100 µL, diluted 1:500 to 1:1 093 500, v/v) was added and allowed
(
to incubate for 2 h at room temperature. Specific antibody
binding was detected with the use of goat anti-rabbit alkaline
phosphatase (1:2000, v/v) and p-nitrophenyl phosphate as
substrate. The absorbance was measured at 410 nm using a
Dynatech plate reader.
A competitive ELISA was performed essentially as described
above, except that 1 µg of solid-phase antigen was applied to
each microtiter plate well. Rabbit serum (1:2500, v/v), incubated
overnight in the presence of an equal volume of various
concentrations of inhibitors, was then added to the plates, and
the assay was carried out as described above.
Although a variety of studies have reported that TRI
undergoes metabolic activation and protein covalent
binding in vitro and in vivo (13-17), confirmation of
1
4
covalent binding in vivo using [ C]TRI has been difficult
since most of the label associated with protein arises as
a consequence of metabolic incorporation of TRI metabo-
2
lites into the amino acid (C ) pool of proteins (17). Thus,
the identity of the adduct(s) and the reactive metabolites
from which they are formed remain to be established.
The present studies were undertaken to determine
whether TRI adducts could be detected in vivo using an
immunochemical approach (11). This was accomplished
by raising polyclonal antibodies against dichloroacetyl-
ated proteins, which were assumed to have the highest
probability of detecting a variety of TRI-derived adducts
containing from one to three chlorine atoms. Further-
more, dichloroacetylated proteins would be expected to
arise from dichloroacetyl chloride, a putative reactive
metabolite of TRI (2, 13). We report that anti-dichloro-
acetyl (anti-DCA) antibodies, collected from the serum
of a rabbit immunized with dichloroacetylated keyhole
limpet hemocyanin (KLH), recognized dichloroacetylated
rabbit serum albumin (RSA) but not unmodified protein.
Of importance, these antibodies were able to detect TRI-
protein adducts in the livers of TRI-treated mice, which
confirms that TRI undergoes metabolic activation and
covalent binding to protein in vivo.
Affin ity P u r ifica tion of Ra bbit Ser u m . Rabbit antiserum
was affinity purified using a column consisting of dichloroacetic
acid conjugated to diaminodipropylamine immobilized on aga-
rose gel. The affinity column was prepared by incubating
dichloroacetic acid (2 mM) for 2 min with 10 mM EDC in 12.5
mM sodium phosphate (pH 5). This mixture was then added
to a 50% suspension of diaminodipropylamine immobilized on
4
% beaded agarose (Pierce) in 200 mM sodium phosphate (10
mL, pH 8) and incubated at room temperature for 24 h. The
beads were washed with 10 mM Tris-HCl buffer (pH 7.5) and
incubated with rabbit serum (1 mL) at room temperature for 1
h. The suspension was transferred to a column and washed
with 2 column vol each of Tris-HCl buffer and Tris-HCl buffer
containing 0.5 M NaCl. The antibody was then eluted with 0.2
M glycine (pH 2.6) and immediately neutralized with 1 M Tris
(
pH 8).
ꢀ
Syn th esis of An tibod y In h ibitor s. N -Monochloroacetyl-
ꢀ
ꢀ
L-lysine, N -dichloroacetyl-L-lysine, and N -trichloroacetyl-L-
lysine were synthesized as described previously (19). Briefly,
preparation of each lysine derivative was carried out by reacting
L-lysine monohydrochloride (10 mmol) with the corresponding
S-ethyl monochloro-, dichloro- and trichlorothiolacetate (15
mmol) in 10 mL of aqueous 1 N sodium hydroxide for 2 h at
room temperature. The dichloro- and trichloroacetyl lysine
derivatives precipitated from the reaction mixture and were
collected by vacuum filtration. These lysine derivatives were
then purified by recrystallization from absolute ethanol. The
monochloroacetyl lysine derivative was purified by HPLC using
a Vydac reverse-phase C18 semi-preparative column (10 mm ×
25 cm; Separations Group, Hesperia, CA). Monochloroacetyl
lysine was eluted using an isocratic solvent system consisting
of 0.1% aqueous trifluoroacetic acid/0.084% trifluoroacetic acid
in acetonitrile (99:1, v/v), at a flow rate of 3 mL/min.
Exp er im en ta l P r oced u r es
ꢀ
Ma ter ia ls. TRI (99% purity), dichloroacetic anhydride, N -
acetyl-L-lysine, and mono-, di-, and trichloroacetyl chloride were
obtained from Aldrich Chemical Company (Milwaukee, WI).
L-Lysine monohydrochloride was purchased from NovaBiochem
(La J olla, CA). Anti-cytochrome P450 2E1 was a generous gift
of Dr. Magnus Ingelman-Sundberg, Department of Medical
Biochemistry and Biophysics, Karolinska Institutet, Stockholm,
Sweden. KLH, 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide
hydrochloride (EDC), and the ImmunoPure IgG Orientation kit
were obtained from Pierce Chemical Company (Rockford, IL).
The alkaline phosphatase substrate kit was obtained from Bio-
Rad Laboratories (Hercules, CA). Enhanced Chemilumines-
cence Western blotting detection reagents were from Amersham
Life Science (Buckinghamshire, England). All other chemicals
were reagent, electrophoretic, or HPLC grade and used without
further purification.
Liquid secondary ion mass spectrometry of the chloroacetyl
lysine derivatives was performed in the first stage of a J eol
HX110/HX110 tandem mass spectrometer in the Medical Uni-
versity of South Carolina Mass Spectrometry Research Resource
Facility. Approximately 1 µg of each compound dissolved in
water (1 µL) was loaded onto the probe tip; glycerol was used
as the matrix. Sample ionization was achieved by bombardment
with 9 keV of cesium ions.
P r ep a r a t ion a n d Det ect ion of An t i-DCA An t ib od ies.
Dichloroacetic anhydride (80 µL; 250 mM final concentration)
was added dropwise to the immunogenic carrier protein KLH
Proton NMR spectra of the chloroacetyl lysine derivatives
(40 mg) in 2 mL of sodium phosphate (pH 10) with vigorous
dissolved in Me SO-d (Sigma, St. Louis, MO) were acquired in
2
6
mixing. The reaction mixture was incubated at room temper-
ature for 1 h and then dialyzed three times with 4 L of 10 mM
phosphate-buffered saline (pH 7.2). A nonrelated carrier pro-
tein, RSA, was derivatized in a similar manner for use as the
solid-phase antigen in an enzyme-linked immunosorbent assay
the Medical University of South Carolina NMR Research
Resource Facility on a Varian VXR 400 NMR spectrometer
operating at 400 MHz with conventional quadrature detection.
Spin-spin correlations were determined by employing a double
quantum filtered correlation spectroscopy (DQ-COSY) pulse
sequence collected in the phase-sensitive format. A total of 512
1024-point free induction decays were collected, resolution
enhanced and converted from Lorenzian-to-Gaussian transfor-
mation functions, and then Fourier transformed to a 2 K × 2 K
data matrix.
(ELISA).
One female New Zealand White rabbit was immunized sc at
1
5 sites along the back and im in the right and left hindquarters
with 1.5 mg of dichloroacetylated KLH suspended in 5 vol of
Freund’s complete adjuvant. The rabbit was boosted 4 weeks
later with 500 µg of dichloroacetyl KLH in 2 vol of Freund’s
incomplete adjuvant.
The purity of the chloroacetyl lysine derivatives was deter-
mined on a Waters HPLC system (Waters Associates, Milford,
+
+

Detection of Trichloroethylene Adducts
Chem. Res. Toxicol., Vol. 9, No. 2, 1996 453
Ta ble 1. 400 MHz 1H-NMR Sp ectr a l P a r a m eter s of
MA) consisting of a Vydac C18 reverse-phase analytical column
E
(
4.6 mm × 25 cm) and a UV absorbance detector (214 nm). The
N -Ch lor oa cetyla ted Lysin e Der iva tives
products were eluted using a 60-min linear gradient of 99%
solvent A (aqueous 0.1% trifluoroacetic acid) to 99% solvent B
chloroacetyl Lysa
a
a
d
i
c
h
l
o
r
o
a
c
e
t
y
l
L
y
s
t
r
i
c
h
l
o
r
o
a
c
e
t
y
l
L
y
s
shift multiplicity shift multiplicity shift multiplicity
(0.084% trifluoroacetic acid in acetonitrile) at a flow rate of 1
proton (δ)b
b
b
(J , Hz)
(δ)
(J , Hz)
(δ)
(J , Hz)
mL/min. HPLC analysis indicated a UV purity for each lysine
derivative of >99%.
An im a l Tr ea tm en t a n d Im m u n och em ica l Detection of
TRI Ad d u cts. The presence of TRI-protein conjugates was
determined in a time course and dose responsive manner with
NH
8.21
t (5.5)
8.59
8.21
3.88
1.74
1.75
1.35
1.43
3.11
6.44
t (5.5)
bs
m
m
m
m
m
q (6.8)
s
8.99
8.20
3.87
1.75
1.76
1.36
1.50
3.18
t (5.5)
bs
m
m
m
m
m
q (6.6)
H
H
3
N+ n.o.c
R
3.85
1.74
1.75
1.34
1.42
3.08
4.03
t (6.2)
m
m
m
m
q (6.5)
s
d
d
d
d
d
d
d
d
b
d
d
d
d
Hâ
H
H
H
H
â′
1
2-wk-old B6C3/F1 male mice. Various doses of TRI were
δ
γ
ꢀ
administered ip in corn oil (5 mL/kg). Mice dosed with corn oil
only and sacrificed at 6 h served as controls. At designated
intervals following TRI administration, mice were anesthetized
x y
CH Cl
a
with CO
2
and sacrificed by cervical dislocation. A blood sample
The test compounds were fully protonated as the trifluroac-
etate salts. b In Me2SO-d6 downfield from TMS (δ 0.00 ppm). c Not
observed due to the presence of a strong water peak in the
spectrum. d Chemical shift determination based on the center of
the DQ-COSY cross peak. Abbreviations: s, singlet; t, triplet; q,
quartet; m, multiplet; bs, broad singlet.
was removed from the orbital sinus of each mouse for determi-
nation of serum alanine aminotransferase levels using a colo-
rimetric assay kit from Sigma.
Livers from mice in each treatment group were excised,
minced, and homogenized in 3 w/v of 0.25 M sucrose buffer (pH
7
.5) containing 10 mM HEPES, 1 mM EDTA, 0.2 mM phenyl-
methanesulfonyl fluoride, 1 µM leupeptin, and 1 µM pepstatin,
utilizing a Dounce homogenizer, 6 strokes. An aliquot of each
individual homogenate was reserved; then tissues from each
group were pooled, frozen in liquid nitrogen, and stored at -80
°
C until fractionation procedures were carried out.
Crude subcellular fractions were prepared using the liver
homogenates as previously described (20). Protein content of
tissues was determined according to Lowry et al. (21), using
bovine serum albumin as a standard. Proteins from the 25000g
pellet, cytosolic, and microsomal fractions were separated by
SDS-PAGE under reducing conditions and then transferred to
nitrocellulose as described previously (22). Adducts were
detected with affinity-purified anti-DCA serum antibodies using
an alkaline phosphatase detection system, following reported
procedures (18, 22).
F igu r e 1. Competitive inhibition of binding of anti-DCA serum
antibodies to the solid phase antigen, DCA-RSA (1 µg/well).
Serial dilutions of inhibitors were incubated overnight with
affinity purified rabbit antiserum (1:2500, v/v). The inhibitors
The potential that the 50-kDa TRI-protein adduct is cyto-
chrome P450 2E1 was explored by comigration following electro-
phoresis. Proteins separated by SDS-PAGE were transferred
to nitrocellulose; parallel lanes were immunochemically probed
with anti-DCA or anti-cytochrome P450 2E1 and visualized
using chemiluminescence.
ꢀ
used in the competitive ELISA were N -(dichloroacetyl)-L-lysine
ꢀ
ꢀ
(b), N -(trichloroacetyl)-L-lysine (2), N -(chloroacetyl)-L-lysine
ꢀ
(9), N -(acetyl)-L-lysine (0), and L-lysine monohydrochloride (4).
Ch a r a cter iza tion of An tibod y Sp ecificity. In an
ELISA, serum from a rabbit immunized with dichloro-
acetylated KLH was found to react with the solid-phase
antigen (DCA-RSA) but not with unmodified RSA (data
not shown). In a competitive ELISA, using affinity
Resu lts
Syn th esis of An tibod y In h ibitor s. To characterize
the structural specificity of the serum antibodies (see
ꢀ
ꢀ
below), a series of N -(chloroacetyl)-L-lysine derivatives
purified antiserum (Figure 1), the N -(chloroacetyl)-L-
ꢀ
ꢀ
(
i.e., N -[chloroacetyl]-L-lysine, N -[dichloroacetyl]-L-lysine,
lysine derivatives were at least 3 orders of magnitude
more effective as inhibitors of antibody binding to DCA-
ꢀ
and N -[trichloroacetyl]-L-lysine) were synthesized as
ꢀ
described by Birner et al. (19). Structure and purity of
the chloroacetyl lysine derivatives were confirmed by
HPLC, mass, and NMR spectral analyses. Chromato-
graphic analysis of the purified monochloro-, dichloro-,
and trichloroacetyl lysine derivatives (214 nm) showed
single peaks with retention times of 9.2, 14.0, and 19.4
min, respectively. Mass spectral analysis of the chloro-
acetyl lysine derivatives showed molecular ions (MH )
at m/z 222.7 (monochloro), 256.9 (dichloro), and 290.6
trichloro), and chlorine isotopic patterns consistent with
RSA than was L-lysine (IC50 >1 M) or N -(acetyl)-L-lysine
(IC50 0.6 M). Dichloroacetyl lysine was the most potent
inhibitor tested, with an IC50 of about 15 µM. Trichlo-
roacetyl and monochloroacetyl lysine were less potent,
with IC50 values of 0.6 and 10 mM, respectively. TRI,
trichloroacetic acid, and dichloroacetic acid, which were
also tested for inhibitory activity, had no effect on
antibody binding below concentrations that resulted in
antibody inactivation due to nonspecific (i.e., denaturing)
effects.
+
(
the number of chlorine atoms for each derivative. In
addition, DQ-COSY analysis confirmed that chloroacetyl-
ation had occurred on the ꢀ-amino group of lysine (Table
Im m u n och em ica l Detection of TRI Ad d u cts in
Vivo. In order to confirm the existence of TRI-protein
adducts in vivo and also to determine whether the anti-
DCA antibody could detect such adducts, mice were given
a single dose of TRI (1000 mg/kg) and sacrificed 1, 3, 6,
and 9 h later. The livers were then removed and
homogenized, and subcellular fractions were prepared.
Microsomal, 25000g pellet, and cytosolic proteins were
then separated by SDS-PAGE, transferred to nitrocel-
lulose, and probed with anti-DCA antibodies (Figure 2).
1
). Of interest, monochloroacetyl lysine was found to
react with ethanethiol (liberated from S-ethyl monochlo-
rothiolacetate) during its synthesis to form the corre-
sponding thioether as a major byproduct of the reaction.
This observation raises the possibility that a monochlo-
roacetylated lysine adduct, if formed in vivo, may not be
stable.
+
+

4
54 Chem. Res. Toxicol., Vol. 9, No. 2, 1996
Halmes et al.
F igu r e 4. Inhibition of immunoblot detection of protein adducts
in the microsomal (M) and 25000g pellet (P) fractions from liver
homogenates of B6C3/F1 mice given TRI (1000 mg/kg). Mice
were sacrificed 6 h after TRI administration. anti-DCA antise-
rum was incubated overnight in the absence (no inhibitor) or
F igu r e 2. Immunoblot detection of liver protein adducts using
affinity purified anti-DCA antiserum following SDS-PAGE of
(
left panel) microsomal, (center panel) 25000g pellet, and (right
panel) cytosolic proteins (200 µg of protein/lane) from male
B6C3/F1 mice given TRI (1000 mg/kg, ip in corn oil). Mice were
sacrificed 1 (lane 2), 3 (lane 3), 6 (lane 4), and 9 h (lane 5) after
TRI administration. Mice treated with corn oil and sacrificed
at 6 h served as controls (lane 1).
presence of 500 µM N
ꢀ
-(trichloroacetyl) lysine (TCA-Lys), N
-(chloroacetyl) lysine (MCA-
-(acetyl) lysine (AC-Lys) prior to immunoblot
ꢀ
-
ꢀ
(dichloroacetyl) lysine (DCA-Lys), N
ꢀ
Lys), and N
analysis.
F igu r e 3. Effect of dose of TRI on the amount of liver protein
adducts detected by the anti-DCA antibody in (left panel)
microsomal, (center panel) 25000g pellet, and (right panel)
cytosolic proteins (200 µg of protein/lane) from male B6C3/F1
mice. Lane 1: corn oil control; lane 2: 250 mg/kg; lane 3: 500
mg/kg; lane 4: 1000 mg/kg; lane 5: 2000 mg/kg TRI.
In the microsomal fraction of corn oil-treated mice
(
Figure 2, lane 1), a single 70-kDa protein reacted
F igu r e 5. Immunoblot detection of protein adducts following
SDS-PAGE of liver microsomal protein (200 µg protein/lane).
Mice were sacrificed 6 h after administration of TRI (1000 mg/
kg) One lane was stained with anti-DCA antiserum (anti-DCA),
and the other was stained with anti-cytochrome P450 2E1 (anti-
CYP2E1). One lane of the SDS-PAGE gel was stained with
Coomassie blue to visualize all proteins.
immunochemically with the anti-DCA antibody; however,
this reaction was not observed consistently, and its
significance is not yet known. In contrast, two major
adducts, at 50 and 100 kDa, were detected in liver
proteins from TRI-treated mice. As shown in Figure 2
(
left panel, lanes 2-4), the amount of the 50- and 100-
kDa adducts in the microsomal fraction increased up to
h following treatment and then declined by 9 h. Similar
inhibitors. As shown in Figure 4, all of the chloroacetyl
lysine derivatives were inhibitory. Dichloroacetyl lysine
completely inhibited antibody binding under the de-
scribed experimental conditions, whereas trichloroacetyl
and monochloroacetyl lysine were less potent inhibitors,
respectively; this pattern was also seen in the competitive
ELISA. Acetyl lysine was the least effective inhibitor
tested.
6
results were observed in the 25000g pellet (Figure 2,
center panel, lanes 2-4); however, an additional protein
band was observed in this fraction at approximately 95
kDa. In the cytosol (Figure 2, right panel), the 50- and
1
00-kDa adducts were observed; however, the intensity
of staining in this fraction was much less than that
observed in microsomes.
Im m u n oblot of Liver Micr osom a l P r otein Usin g
An t i-CYP 2E 1 An t iser a . Since TRI has been shown
previously to be a mechanism-based inhibitor of cyto-
chrome P450 (12), the possibility that the 50-kDa protein
was cytochrome P450 was investigated. Liver microso-
mal protein from mice treated with 1000 mg/kg TRI and
sacrificed at 6 h was divided into two aliquots, each of
which was separated on a polyacrylamide gel, and
transferred to nitrocellulose. One lane was stained with
the anti-DCA antibody, and the other was stained with
an antibody raised against cytochrome P450 2E1, an
isoform known to be responsible for TRI oxidation in rats
(23) and humans (24). As shown in Figure 5, a single
50-kDa protein was detected in the lane probed with anti-
cytochrome P450 2E1. This protein band comigrated
Adduct formation in the liver was also dose-dependent
from 250 to 1000 mg/kg (Figure 3). In the microsomes
and the 25000g pellet, the 50-kDa adduct exhibited the
highest intensity of staining when using the anti-DCA
antibody, with the 100-kDa adduct staining less in-
tensely. Similar bands are observed at all dose levels,
with the most concentrated staining in the 50- and 100-
kDa protein fractions. In the cytosolic fraction, staining
was much less intense; however, both the 50- and 100-
kDa adducts are detectable.
In order to confirm the specificity of TRI adduct
detection, microsomal and 25000g proteins from TRI-
treated mice were probed with the anti-DCA antibody
in the presence of 500 µM of the chloroacetyl lysine
+
+

Detection of Trichloroethylene Adducts
Chem. Res. Toxicol., Vol. 9, No. 2, 1996 455
with the 50-kDa protein adduct detected in the lane
probed with anti-DCA antibodies.
(29). Thus, the status of acyl chlorides in general, and
dichloracetyl chloride in particular, as metabolites is in
doubt. In the present studies, the high specificity of the
anti-DCA antibody toward the dichloroacetyl moiety is
significant because it suggests that some, if not all, of
the TRI adducts are formed by acetylation of hepatic
protein by a dichloroacetyl chloride metabolite of TRI.
The validity of this approach is supported by the studies
of Pohl and colleagues, who showed that a polyclonal
antibody raised against trifluoroacetylated protein was
able to recognize protein adducts formed in rats exposed
to halothane (30).
Since metabolic incorporation of TRI metabolites into
the amino acid pool does not interfere with immune
recognition of chemical-protein adducts, it will be pos-
sible to use the anti-DCA antibody as a tool to investigate
the role of covalent binding in the mechanisms underly-
ing acute and chronic TRI toxicity. The antibody also
has the potential for use as a dosimeter for TRI, which
may be useful for assessing the risk associated with
environmental exposure. Another application is the
potential to determine whether anti-DCA antibodies are
present in the sera of systemic sclerosis patients. This
type of response has been observed previously with serum
from halothane hepatitis patients, which recognized
trifluoroacetylated proteins (31). Additionally, the pro-
teins shown to bind covalently to reactive metabolites of
TRI can be purified, through immunoaffinity or other
forms of chromatography (32), and can be examined for
changes in structure and/or activity. The ability to detect
and identify the proteins covalently modified by TRI will
enhance our understanding of the mechanism(s) of
chemical-induced cytotoxicity and may elucidate the
potential role of adduct formation in the production of a
sclerosis-like disease.
TRI-In d u ced Hep a totoxicity. Serum alanine ami-
notransferase levels were determined on all mice from
each of the treatment groups. Alanine aminotransferase
levels for mice treated with TRI were not found to be
statistically different from those of control mice (data not
shown).
Discu ssion
These studies describe the preparation of a rabbit
polyclonal antibody against dichloroacetylated KLH,
which is highly specific for the dichloroacetyl moiety on
protein (Figure 1). When the anti-DCA antibody was
used to probe liver protein from TRI-treated mice (Fig-
ures 2 and 3), two major protein adducts, at 50 and 100
kDa, were detected. Adduct formation was found to be
both dose and time dependent in all subcellular fractions
examined; however, the amount of TRI adduct formation
was highest in the microsomal fraction. In addition, anti-
DCA specific binding to TRI adducts on immunoblots was
ꢀ
inhibited completely by N -(dichloroacetyl)-L-lysine (Fig-
ure 4).
Of interest, the formation of TRI adducts appeared to
be highly selective, in that most proteins in the liver did
not react with the anti-DCA antibodies. This observation
is in contrast to the situation seen with acetaminophen,
where the reactive metabolite N-acetyl-p-benzoquinone
imine has been shown by immunochemical methods to
covalently bind first to a small, selective group of hepatic
proteins, and later in the course of toxicity (after 2-4
h), to a relatively large number of proteins (22, 25). The
reason for the high degree of selectivity observed with
TRI covalent binding is not yet known. However, since
the presumed target for TRI reactive metabolites (i.e.,
the ꢀ-amino groups of lysine in cellular protein) is widely
available in the cytosol and within cellular organelles,
the lack of general binding indicates that the TRI reactive
metabolite is not kinetically free in the cell. One pos-
sibility is that the selectivity of binding is related to the
ability of TRI to inactivate cytochrome P450 by a mech-
anism-based process (12, 13, 26). That is, a TRI-oxygen-
ated cytochrome P450 intermediate complex rearranges
to a reactive intermediate that binds covalently to the
apoprotein and/or the heme prosthetic group. Since
hepatic microsomal cytochrome P450 2E1 is thought to
be the major isoform responsible for the oxidation of TRI
in rodents (23, 27), we postulate that the 50-kDa TRI-
adducted protein is cytochrome P450 2E1. This postulate
is supported by the data presented in Figure 5, which
shows that the 50-kDa TRI adduct comigrates electro-
phoretically with a protein recognized by an antibody
raised against cytochrome P450 2E1. The identity of the
Ack n ow led gm en t. The authors thank Dr. Magnus
Ingelman-Sundberg for kindly providing anti-cytochrome
P450 2E1 antiserum. This work was supported by a
grant from the Department of Energy (DE-FG01-
92EW50625).
Refer en ces
(
(
(
1) Bruckner, J . V., Davis, B. D., and Blancato, J . N. (1989)
Metabolism, toxicity, and carcinogenicity of trichloroethylene.
Crit. Rev. Toxicol. 20, 31-50.
2) Davidson, I. W., and Beliles, R. P. (1991) Consideration of the
target organ toxicity of trichloroethylene in terms of metabolite
toxicity and pharmacokinetics. Drug Metab. Rev. 23, 493-599.
3) Phoon, W. H., Chan, M. O., Rajan, V. S., Tan, K. J ., Thirumoorthy,
T., and Goh, C. L. (1984) Stevens-J ohnson syndrome associated
with occupational exposure to trichloroethylene. Contact Derma-
titis 10, 270-276.
(4) Green, T. (1990) Chloroethylenes: a mechanistic approach to
human risk evaluation. Annu. Rev. Pharmacol. Toxicol. 30, 73-
8
9.
1
00-kDa adduct is not yet known; however, it seems
(
5) Goeptar, A. R., Commandeur, J . N., van Ommen, B., van
Bladeren, P. J ., and Vermeulen, N. P. (1995) Metabolism and
kinetics of trichloroethylene in relation to toxicity and carcino-
genicity. Relevance of the mercapturic acid pathway. Chem. Res.
Toxicol. 8, 3-21.
unlikely that it is a dimer of the 50-kDa adduct, since it
did not react immunochemically with anti-cytochrome
P450 2E1.
The intermediary role of acyl chlorides in the metabo-
lism of haloalkenes and haloalkanes has been widely
assumed based primarily on the appearance of haloacetic
acid metabolites in microsomal incubations and as uri-
nary metabolites (19, 28). However, recent evidence
suggests that haloacetic acid metabolites can arise by
alternate metabolic pathways, as has been shown for the
TRI metabolite, dichloroacetic acid, which can arise in
mice via reductive dechlorination of trichloroacetic acid
(
6) Lockey, J . E., Kelly, C. R., Cannon, G. W., Colby, T. V., Aldrich,
V., and Livingston, G. K. (1987) Progressive systemic sclerosis
associated with exposure to trichloroethylene. J . Occup. Med. 29,
493-496.
(7) Yanez Diaz, S., Moran, M., Unamuno, P., and Armijo, M. (1992)
Silica and trichloroethylene-induced progressive systemic scle-
rosis. Dermatology 184, 98-102.
(
8) Kilburn, K. H., and Warshaw, R. H. (1992) Prevalence of
symptoms of systemic lupus erythematosus (SLE) and of fluo-
rescent antinuclear antibodies associated with chronic exposure
+
+

4
56 Chem. Res. Toxicol., Vol. 9, No. 2, 1996
Halmes et al.
to trichloroethylene and other chemicals in well water. Environ.
Res. 57, 1-9.
9) Waller, P. A., Clauw, D., Cupps, T., Metcalf, J . S., Silver, R. M.,
and Leroy, E. C. (1994) Fasciitis (not scleroderma) following
prolonged exposure to an organic solvent (trichloroethylene). J .
Rheumatol. 21, 1567-1570.
(21) 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.
(22) Pumford, N. R., Hinson, J . A., Benson, R. W., and Roberts, D. W.
(1990) Immunoblot analysis of protein containing 3-(cystein-S-
yl)acetaminophen adducts in serum and subcellular liver fractions
from acetaminophen-treated mice. Toxicol. Appl. Pharmacol. 104,
521-532.
(23) Nakajima, T., Wang, R. S., Elovaara, E., Park, S. S., Gelboin, H.
V., and Vainio, H. (1992) A comparative study on the contribution
of cytochrome P450 isozymes to metabolism of benzene, toluene
and trichloroethylene in rat liver. Biochem. Pharmacol. 43, 251-
257.
(
(
10) Khan, M. F., Kaphalia, B. S., Prabhaker, B. S., Kanz, M. F., and
Ansari, G. A. S. (1995) Trichloroethene-induced autoimmune
response in female MRL +/+ mice. Toxicol. Appl. Pharmacol. 134,
1
55-160.
(
11) Pohl, L. R. (1993) An immunochemical approach of identifying
and characterizing protein targets of toxic reactive metabolites.
Chem. Res. Toxicol. 6, 786-793.
(
12) 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.
(24) Guengerich, F. P., Kim, D. H., and Iwasaki, M. (1991) Role of
human cytochrome P-450 IIE1 in the oxidation of many low
molecular weight cancer suspects. Chem. Res. Toxicol. 4, 168-
179.
(
(
(
13) 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.
(25) Bartolone, J . B., Sparks, K., Cohen, S. D., and Khairallah, E. A.
(1987) Immunochemical detection of acetaminophen-bound liver
proteins. Biochem. Pharmacol. 36, 1193-1196.
(26) Allemand, H., Pessayre, D., Descatoire, V., De Gott, C., Feldmann,
G., and Benhamou, J . P. (1978) Metabolic activation of trichlo-
roethylene into a chemically reactive metabolite toxic to the liver.
J . Pharmacol. Exp. Ther. 204, 714-723.
(27) Nakajima, T., Wang, R. S., Elovaara, E., Park, S. S., Gelboin, H.
V., and Vainio, H. (1993) Cytochrome P450-related differences
between rats and mice in the metabolism of benzene, toluene and
trichloroethylene in liver microsomes. Biochem. Pharmacol. 45,
1079-1085.
(28) Sipes, I. G., Gandolfi, A. J ., Pohl, L. R., Krishna, G., and Brown,
B. R., J r. (1980) Comparison of the biotransformation and
hepatotoxicity of halothane and deuterated halothane. J . Phar-
macol. Exp. Ther. 214, 716-720.
(29) Larson, J . L., and Bull, R. J . (1992) Metabolism and lipoperoxi-
dative activity of trichloroacetate and dichloroacetate in rats and
mice. Toxicol. Appl. Pharmacol. 115, 268-277.
(30) Satoh, H., Gillette, J . R., Davies, H. W., Schulick, R. D., and Pohl,
L. R. (1985) Immunochemical evidence of trifluoroacetylated
cytochrome P-450 in the liver of halothane-treated rats. Mol.
Pharmacol. 28, 468-474.
14) Parchman, L. G., and Magee, P. N. (1982) Metabolism of
1
4
14
[
C]trichloroethylene to CO
with liver DNA in rats and mice. J . Toxicol. Environ. Health 9,
97-813.
2
and interaction of a metabolite
7
15) Stott, W. T., Quast, J . F., and Watanabe, P. G. (1982) The
pharmacokinetics and macromolecular interactions of trichloro-
ethylene in mice and rats. Toxicol. Appl. Pharmacol. 62, 137-
1
51.
(
16) Bergman, K. (1983) Interactions of trichloroethylene with DNA
in vitro and with RNA and DNA of various mouse tissues in vivo.
Arch. Toxicol. 54, 181-193.
(
17) Stevens, D. K., Eyre, R. J ., and Bull, R. J . (1992) Adduction of
hemoglobin and albumin in vivo by metabolites of trichloroeth-
ylene, trichloroacetate, and dichloroacetate in rats and mice.
Fundam. Appl. Toxicol. 19, 336-342.
(
18) Pumford, N. R., Myers, T. G., Davila, J . C., Highet, R. J ., and
Pohl, L. R. (1993) Immunochemical detection of liver protein
adducts of the nonsteroidal antiinflammatory drug diclofenac.
Chem. Res. Toxicol. 6, 147-150.
(
19) Birner, G., Richling, C., Henschler, D., Anders, M. W., and
Dekant, W. (1994) Metabolism of tetrachloroethene in rats:
(
31) Satoh, H., Gillette, J . R., Takemura, T., Ferrans, V. J ., J elenich,
S. E., and Pohl, L. R. (1986) Investigation of the immunological
basis of halothane-induced hepatotoxicity. Adv. Exp. Med. Biol.
ꢀ
ꢀ
identification of N -(dichloroacetyl)-L-lysine and N -(trichloro-
acetyl)-L-lysine as protein adducts. Chem. Res. Toxicol. 7, 724-
1
97, 657-673.
7
32.
20) Pumford, N. R., Roberts, D. W., Benson, R. W., and Hinson, J . A.
1990) Immunochemical quantitation of 3-(cystein-S-yl)acetami-
(
32) Pumford, N. R., Martin, B. M., and Hinson, J . A. (1992) A
metabolite of acetaminophen covalently binds to the 56 kDa
selenium binding protein. Biochem. Biophys. Res. Commun. 182,
(
(
nophen protein adducts in subcellular liver fractions following a
hepatotoxic dose of acetaminophen. Biochem. Pharmacol. 40,
1
348-1355.
5
73-579.
TX950171V
+
+