Disulfiram generates a stable N,N-diethylcarbamoyl adduct on Cys-125 of rat hemoglobin β-chains in vivo

Article abstract of DOI:10.1021/tx990191n

Disulfiram (DSF) is a drug used in aversion therapy to treat alcoholics and acts by inhibiting mitochondrial low-K(m) aldehyde dehydrogenase. Investigations into the mechanisms for in vivo inactivation suggest that the DSF metabolite S-methyl-N,N-diethylthiocarbamate sulfoxide reacts irreversibly with an active site Cys. This work aimed to determine if DSF generates monothiocarbamate adducts on cysteine residues in vivo by examining hemoglobin. Sprague-Dawley rats were treated with DSF po for 2, 4, and 6 weeks. Rats have four different globin β-chains, of which three (β-1-3) contain two cysteine residues each. MALDI-TOF MS analysis of two new globin species from DSF-treated rats collected by HPLC revealed increments of 99 Da above the mass of the unmodified chains (β-2 and β-3). In a separate experiment, the globin mixture was digested for 2 h with Glu-C and reanalyzed by MALDI-TOF MS. Results showed a peptide at m/z 2716.3 having a mass 99 Da higher than a known Cys-containing peptide. Subsequently, the Glu-C digest was analyzed using Q-TOF tandem MS, enabling observation of the +4 charge state of the peptide with m/z 2716.3. This peptide was fragmented to produce y-sequence ions that located the modification to Cys-125 (present on both β- 2 and β-3). Cys-125 is the most reactive of two cysteine residues on these β-chains. To confirm the structure of the modification, globin was hydrolyzed with 6 N HCl at 110 °C for 18 h. The adduct survived these conditions so that S-(N,N-diethylcarbamoyl)cysteine was detected in the hydrolysates of treated rats on the basis of comparison with the tandem MS spectrum of a standard. These results extend the findings of others obtained using glutathione conjugates and demonstrate the ability of DSF to covalently modify Cys residues of proteins in a manner consistent with the production of S-methyl-N,N-diethylthiocarbamate sulfoxide, or sulfone, intermediates.

Full text of DOI:10.1021/tx990191n

Chem. Res. Toxicol. 2000, 13, 237-244
237
Disu lfir a m Gen er a tes a Sta ble N,N-Dieth ylca r ba m oyl
Ad d u ct on Cys-125 of Ra t Hem oglobin â-Ch a in s in Vivo
J ohn C. L. Erve,† Ole N. J ensen,‡ Holly S. Valentine,†
Venkataraman Amarnath,† and William M. Valentine*,†
Department of Pathology and Center in Molecular Toxicology, Vanderbilt University Medical
Center, Nashville, Tennessee 37232-2561, and Department of Molecular Biology, Odense University/
University of Southern Denmark, Odense 5230 M, Denmark
Received November 11, 1999
Disulfiram (DSF) is a drug used in aversion therapy to treat alcoholics and acts by inhibiting
mitochondrial low-Km aldehyde dehydrogenase. Investigations into the mechanisms for in vivo
inactivation suggest that the DSF metabolite S-methyl-N,N-diethylthiocarbamate sulfoxide
reacts irreversibly with an active site Cys. This work aimed to determine if DSF generates
monothiocarbamate adducts on cysteine residues in vivo by examining hemoglobin. Sprague-
Dawley rats were treated with DSF po for 2, 4, and 6 weeks. Rats have four different globin
â-chains, of which three (â-1-3) contain two cysteine residues each. MALDI-TOF MS analysis
of two new globin species from DSF-treated rats collected by HPLC revealed increments of 99
Da above the mass of the unmodified chains (â-2 and â-3). In a separate experiment, the globin
mixture was digested for 2 h with Glu-C and reanalyzed by MALDI-TOF MS. Results showed
a peptide at m/z 2716.3 having a mass 99 Da higher than a known Cys-containing peptide.
Subsequently, the Glu-C digest was analyzed using Q-TOF tandem MS, enabling observation
of the +4 charge state of the peptide with m/z 2716.3. This peptide was fragmented to produce
y-sequence ions that located the modification to Cys-125 (present on both â-2 and â-3). Cys-
125 is the most reactive of two cysteine residues on these â-chains. To confirm the structure
of the modification, globin was hydrolyzed with 6 N HCl at 110 °C for 18 h. The adduct survived
these conditions so that S-(N,N-diethylcarbamoyl)cysteine was detected in the hydrolysates
of treated rats on the basis of comparison with the tandem MS spectrum of a standard. These
results extend the findings of others obtained using glutathione conjugates and demonstrate
the ability of DSF to covalently modify Cys residues of proteins in a manner consistent with
the production of S-methyl-N,N-diethylthiocarbamate sulfoxide, or sulfone, intermediates.
3). On the basis of early in vitro work with ALDH, in
vivo DSF inhibition was thought to result from an
intramolecular disulfide bond involving an active site Cys
(4). However, evidence now suggests that metabolism of
DSF to a reactive metabolite(s) is required for the
irreversible inactivation observed in vivo (5-7).
In tr od u ction
Disulfiram [bis(diethyldithiocarbamoyl)disulfide](DSF),1
or Antabuse, has been used in alcohol aversion therapy
to treat alcoholism for almost 50 years since its imple-
mentation by Larsen et al. in 1948 (1). The basis of this
therapy relies on the series of unpleasant reactions,
known as the DSF ethanol reaction (2), which include
intense flushing of the face and/or neck, tachycardia,
nausea, and vomiting, that occur after ingestion of
ethanol (1, 3). Blood levels of acetaldehyde, the toxic
proximate metabolite of ethanol metabolism, is increased
in patients on Antabuse who drink ethanol (3) and is
thought responsible for many of these effects. It is
believed that acetaldehyde accumulation arises due to
inhibition of low-Km liver mitochondrial aldehyde dehy-
drogenase (EC 1.2.1.3, ALDH) brought about by DSF (1,
The disulfide bond of DSF (1) can be reduced in blood
(Scheme 1) presumably by glutathione (GSH) (8), or by
albumin (9). The resulting diethyldithiocarbamate (DEDC)
(2) can be methylated by thiol methyl transferase or
thiopurine methyltransferase (10, 11) to give rise to
methyl diethyldithiocarbamate (DEDC-Me) (3). Oxidation
of the sulfonyl group of DEDC-Me by a number of
different P450s such as P450 3A4/5 (7) produces methyl-
diethythiocarbamate (DETC-Me) (4), which can undergo
additional oxidation to generate the sulfoxide (DETC-
MeSO) (5). Further oxidation to a sulfone (DETC-MeSO2)
(6) may also occur, and the oxygenated metabolites have
been shown to be capable of inactivating ALDH in vitro
with the reactivity of DETC-MeSO2 being higher than
that of DETC-MeSO (6). GSH conjugates (7) that were
consistent with reaction between the Cys of GSH and
DETC-MeSO and/or DETC-MeSO2 were detected in bile
of rats treated with DSF or diethyldithiocarbamate (12).
* To whom correspondence should be addressed.
† Vanderbilt University Medical Center.
‡ Odense University/University of Southern Denmark.
1
Abbreviations: ACN, acetonitrile; CID, collision-induced dissocia-
tion; DSF, disulfiram; DTT, dithiothreitol; DTNB, dithiobis(2,2′-
nitrobenzoic) acid; DETC-MeSO, S-methyl diethylthiocarbamate sul-
foxide; DETC-MeSO2, S-methyl diethylthiocarbamate sulfone; DETC-
Cys, diethylcarbamoylcysteine; DEDC, diethyldithiocarbamate; DEDC-
Me, S-methyl diethyldithiocarbamate; ESI, electrospray ionization;
GSH, glutathione; MALDI-TOF MS, matrix-assisted laser desorption
ionization time-of-flight mass spectrometry; MeOH, methanol; Q-TOF,
quadrupole orthogonal time-of-flight; SA, sinapic acid; TFA, trifluo-
roacetic acid.
The reactivity of DSF metabolites with proteins in vivo
has not been investigated in detail. For example, al-
10.1021/tx990191n CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/16/2000

238 Chem. Res. Toxicol., Vol. 13, No. 4, 2000
Erve et al.
Carbonyl sulfide evaporated and condensed into the reaction
mixture. After 4 h, the condenser was removed and iodomethane
(0.7 mL) was added. The stirring was continued for 2 h before
ethanol was removed, and the residue was distributed between
water (10 mL) and chloroform (20 mL) layers. The latter was
separated, dried, and evaporated.
Sch em e 1
(2) N,N-Dieth ylth ioca r ba m a te Meth yl Su lfon e (DETC-
MeSO2). A solution of DETC-Me (1.1 g, 7.5 mmol) in dichlo-
romethane (15 mL) was cooled in ice, and 3-chloroperoxybenzoic
acid (3.75 g, 15 mmol) in the same solvent (20 mL) was added
with stirring. The reaction mixture was stirred in ice for 30 min
and at room temperature for 4 h. The reaction was checked by
TLC (a 9:1 hexane/ethyl acetate mixture). The solid was filtered,
and the filtrate was washed with saturated NaHCO3 (2 × 25
mL), dried, and concentrated. Purification was accomplished
with flash chromatography (hexane, followed by a 5:1 hexane/
ethyl acetate mixture). The presence of the sulfone in the
fractions was checked by GC: 1H NMR δ 1.23 and 1.30 (t, 6H,
J ) 7.1 Hz, CH3), 3.14 (s, 3H, CH3S), 3.42 and 3.75 (q, 4H, J )
7.1 Hz, CH2); 13C NMR δ 12.1 and 13.7 (CH3), 39.5 (CH3S), 41.4
and 42.5 (CH2), 160.5 (CdO).
(3) S-(N,N-Dieth ylca r ba m oyl)cystein e (DETC-Cys). The
pH of a solution of Cys (0.52 g, 4 mmol) in water (20 mL) was
adjusted to 8 with 1 N NaOH and the mixture stirred with
DETC-MeSO2 in methanol (20 mL) under Ar for 2 h. The
precipitated solid was filtered and washed with 1:1 MeOH/H2O
(20 mL). The filtrate was concentrated before purification over
a column of silica (40-200 mesh) with 20% H2O in ACN. The
fractions exhibiting a UV maximum around 220 nm were
combined and concentrated to obtain the cysteinyl adduct (TLC
with 20% H2O in ACN; Rf ) 0.4): 13C NMR δ 13.2 and 13.6
(CH3CH2), 31.4 (SCH2), 43.5 and 44.0 (CH3CH2), 55.8 (CH),
168.8 (NCdO), 173.0 (CO2H); ESI-MS m/z 221 (M + H)+.
though irreversible inhibition of human ALDH by DETC-
MeSO has been demonstrated in vitro (13, 14), detection
of the adduct on ALDH in vivo has not been reported.
DSF hepatotoxicity is a rare consequence of DSF therapy
but is associated with high mortality rates (15). Recently,
a case of DSF hepatitis in which autoantibodies were
formed against P450 was reported (16). Such findings
suggest that proteins are modified following DSF therapy,
at least in some individuals, although the nature of the
hapten has not been elucidated. The primary purpose of
this study was to determine if administration of DSF
resulted in covalent modification of hemoglobin in vivo.
To achieve this goal, we used a combination of mass
spectrometric techniques to characterize an adduct on
Cys-125 of the â-chain of hemoglobin following oral
administration of DSF to rats.
In Vivo Exp osu r es. This study was performed in accordance
with the NIH Guide for Care and Use of Laboratory Animals
and was approved by the Vanderbilt animal care committee.
Male Sprague-Dawley rats (Harlan Sprague Dawley, Inc.) (n )
6; 2 controls), 6-7 weeks old, were used after
a 2 week
acclimation period. The rats were housed under a diurnal light
cycle with water (ad libitum) and fed 1% DSF in a nutritionally
complete pelleted diet for 14, 28, or 40 days ad libitum. Body
weights were checked weekly to assess food intake and weight
gain.
Globin P u r ifica tion . Rats were bled under general anes-
thesia (15 mg of Xylazine/kg of body weight, combined with 100
mg of Ketamine/kg of body weight) by exsanguination from the
abdominal aorta. Blood was collected in a heparinized syringe
and centrifuged to separate the plasma from the red cells. A 5
mM phosphate-buffered saline solution (pH 7.4) containing 150
mM NaCl was added in an equal volume to the red cells, and
the mixture resuspended and centrifuged at 4000g. The super-
natant and buffy coat were discarded, and the washing proce-
dure was repeated twice. The washed red cells were lysed with
twice their volume of 5 mM phosphate buffer (pH 7.4) and
centrifuged at 14000g for 25 min. The hemolysate was mixed
with 1 mL of 1 M ascorbic acid and added dropwise to 50 mL of
cold 2.5% oxalic acid in acetone, and globin was allowed to
precipitate for 15 min. Samples were then centrifuged at 12000g
for 10 min; the supernatant was aspirated, and the globin was
washed by adding 5-6 mL of acetone, resuspended with a metal
spatula, and centrifuged at 12000g for 10 min. The supernatant
was removed and the globin dried under a stream of nitrogen
and stored at -78 °C.
Ch r om a togr a p h y. Separation of globin chains was ac-
complished by dissolving globin from control or DSF-treated rats
in HPLC solvent A to a final concentration of 1 mg/mL and
injecting 20 µL onto a C4 RP column (4.6 mm × 150 mm, 300
Å, 5 µm, Vydac, Hesperia, CA). Proteins were eluted at a rate
of 1 mL/min with a linear gradient from 44 to 57% B over the
course of 40 min [solvent A is ACN/H2O/TFA (20:60:0.01 v/v/v);
solvent B is ACN/H2O/TFA (60:40:0.008 v/v/v)]. The eluant was
monitored at 214 nm. Protein fractions were collected and
Ma ter ia ls a n d Meth od s
Ch em ica ls. Endoproteinase Glu-C was purchased from Boe-
hringer Mannheim (Mannheim, Germany). Sinapic acid (SA)
was purchased from Fluka Chemie (Buchs, Switzerland). DSF
in feed was prepared by Bio-Serve (Frenchtown, NJ ). Acetoni-
trile (ACN), methanol (MeOH), and water were all HPLC grade.
Constantly boiling (6 N) HCl was from Pierce (Rockford, IL).
Nano-electrospray (ESI) needles were obtained from Protana
A/S (Odense, Denmark). Myoglobin, human insulin, n-octyl â-D-
glucopyranoside, and dithiothreitol (DTT) were from Sigma (St.
Louis, MO). SelfPack 20-R2 and 50-R1 POROS reverse phase
packing material was purchased from PerSeptive Biosystems
Inc. (Framingham, MA). Carbonyl sulfide was obtained from
Aldrich Chemical Co. (Milwaukee, WI).
Ch em ica l Syn t h esis. (1) N,N-Diet h ylt h ioca r b a m a t e
Meth yl Ester (DETC-Me). Diethylamine (1.1 mL, 10 mmol),
ethanol (10 mL), and 10 N NaOH (1 mL) were taken in a 25
mL two-neck flask, and the flask was fitted with a Dewar
condenser containing dry ice and 2-propanol. The flask was
stirred and cooled in a dry ice/2-propanol bath (-70 °C), and
carbonyl sulfide (∼1 mL) from a gas cylinder was condensed
into the flask. The second neck was stoppered, and the tem-
perature of the bath was allowed to rise to 10 °C and maintained
at that temperature while the Dewar condenser was kept cold.

Disulfiram Metabolite Adduct Hemoglobin in Vivo
Chem. Res. Toxicol., Vol. 13, No. 4, 2000 239
lyophilized prior to subsequent mass analysis. The HPLC
system consisted of dual LKB 2150 pumps with a diode array
detector (Shimadzu, SpectraChrom SPD-M10A).
Ma ss Sp ectr om etr y. MALDI-TOF MS analyses of pro-
teolytic digests (both on probe and solution) were performed on
a Voyager Elite (PerSeptive Biosystems Inc.) time-of-flight
(TOF) mass spectrometer equipped with delayed extraction. The
ion acceleration was 20.5 kV, the grid voltage 97.5%, and the
guide wire voltage 0.3%. The instrument was operated with a
delay time of 350 ns in the linear mode. Mass spectra were
acquired as the sum of ions generated by irradiation of the target
with 100-300 laser pulses (337 nm N2 laser) and processed
using GRAMS/386 software (Galactic Industries Corp., Salem,
NH) and GPMAW (LightHouse Data, Odense, Denmark).
Internal calibration was accomplished using masses of the two
known peptide fragments derived from â-3, [27-43] and [27-
90] appearing at m/z 2098.5 and 7047.1, respectively. MALDI-
TOF MS of intact proteins was performed on a Bruker Reflex
II mass spectrometer (Bruker-Franzen Analytik GmbH, Bre-
men, Germany). The matrix solution was prepared by saturating
ACN/H2O/TFA (50:49:0.1 v/v/v) with SA. For analysis of peptide
mixtures generated by enzymatic digestion in solution, 0.6 µL
of sample was mixed with 0.6 µL of SA matrix, and the mixture
was applied to the sample plate by the dried-droplet technique
and air-dried. For intact protein analysis, 1 µL of micropurified
DSF globin (1 mg/mL in 20% ACN) was mixed with the internal
standards insulin and myoglobin and the mixture applied by
the dried droplet technique with the SA matrix. LaserOne
software (European Molecular Biology Laboratory, Heidelberg,
Germany) was used to analyze data acquired on the Bruker
mass spectrometer.
F igu r e 1. MALDI/TOF MS spectra of new globin peaks
isolated from DSF-treated rats (6 week DSF administration in
feed at 1% w/w). The peaks at m/z 15 948 (A) and 15 962 (B)
correspond to diethylcarbamoyl adducts (+99 Da) on â-3 and
â-2, respectively. Samples were dissolved in SA matrix and
spectra acquired as described in Materials and Methods.
(pH 8.5) to which 4 µL of 50 mM DTT was added. Samples were
incubated for 2 h at 37 °C prior to microcolumn desalting (see
below) and MALDI-TOF MS analysis. The identical procedure
was repeated at pH 10.5 with incubation at either 37 or 56 °C.
Micr op u r ifica tion of Globin a n d Globin Digests. Prior
to ESI analysis of Glu-C digests, or MALDI-TOF MS analysis
of intact globin, samples were desalted using a micropurification
column containing 20-R2, or 50-R1, POROS reverse phase
material, respectively. The micropurification column consisted
of a Gel-Loader pipet tip prepared according to the method of
Gobom et al. (17). The micropurification column was washed
once with 70% ACN and 0.1% TFA and then equilibrated with
5% formic acid. Then 5-10 µL of digest, or 4 µL globin, was
loaded onto the column. Peptide, or globin, was washed with
15 µL of 5% formic acid, while proteins had an additional wash
of 15 µL of a mixture of 5% MeOH and 5% formic acid. Peptides,
or globin, were eluted with 10 µL of a 50:45:5 (v/v/v) MeOH/
H2O/formic acid mixture into a 0.6 mL Eppendorf tube and
stored at -20 °C prior to analysis. For some globin analyses,
proteins were eluted with 3 µL of SA matrix and applied directly
to the MALDI-TOF MS target in small spots.
Tandem mass spectrometry was performed on a quadrupole-
orthogonal time-of-flight instrument (Q-TOF) (Micromass,
Manchester, U.K.) equipped with a nano-ESI ion source. Ap-
proximately 3 µL of peptide mixture, or 1 µL of acid hydrolysate,
was loaded into the sample needle. The needle voltage was 900
V, the cone voltage at 30-40 V, and the focus at 0 V. The
aperture was adjusted between 2 and 20 V, and resolutions (LM
and HM) were between 0 and 8 V. Collision energy varied
between 15 and 20 eV (12 eV for acid hydrolysate), and Ar was
used as a collision gas (analyzer reading of 1 × 10-4 Torr). Data
were collected over 2-5 min and peptide tandem MS data
deconvoluted using MaxEnt sequencing software provided with
the instrument.
En zym a tic Digestion . Enzymatic digestion of protein on the
MALDI probe sample plate was performed by redisolving the
protein-SA matrix spot with 1 µL of NH4HCO3 (100 mM, pH
7.8) and adding 0.3 µg of Glu-C. The sample plate was then
placed in a humidity chamber and the chamber maintained at
37 °C for 1.5 h. After digestion had occurred, 0.5 µL of SA matrix
and 0.5 µL of 1% formic acid were added to the digest and the
mixture was allowed to recrystallize before reanalysis by
MALDI-TOF MS. In solution, digestion was performed in 0.6
mL Eppendorf tubes by adding 5 µL of a globin mixture (1 mg/
mL), 10 µL of a NH4HCO3 buffer (pH 6.5), and 1.5 µg of Glu-C.
Digestion was performed for 2 h at 37 °C. For some samples,
more extensive digestion of the â-chains was desired. This was
accomplished by adding, after the initial 2 h, 5 µL of n-octyl
â-D-glucopyranoside (20 mM) to aid in denaturing the globin
and an additional 1 µg of Glu-C, followed by incubation for 45
min at 37 °C.
Resu lts
DSF Globin An a lysis by MALDI-TOF MS. Two new
peaks in the chromatograph of globin from DSF-treated
rats were collected and lyophilized prior to MALDI-TOF
MS. The peak with a retention time of 44 min exhibited
a prominent molecular ion with m/z 15 948.2 ( 0.3
(Figure 1A), while the peak with a retention time of 42
min exhibited a molecular ion with m/z 15 962.2 ( 0.6
(Figure 1Β). Both these species were 99.1 ( 0.6 Da higher
in mass than â-2 and â-3, respectively. SA matrix adducts
were observed about 200 Da higher in mass than the
protein signals.
DSF Globin a n d DTT. Incubations (2 h) of DSF
globin in the presence of DTT (pH 8.5 or 10.5) at either
37 or 56 °C produced no apparent loss of the adduct from
either â-2 or â-3, as judged by MALDI-TOF MS. The DSF
globin spectra before and after incubation with DTT
appeared to be identical (data not shown).
Acid Hyd r olysis. Approximately 1 mg of globin was added
to an Eppendorf tube and placed in a hydrolysis vial. A portion
of 6 N HCl (180 µL) was placed in the bottom of the vial which
was flushed with Ar gas before being placed in an oven at 110
°C for 18 h. Upon completion of hydrolysis, samples were
reconstituted with 100 µL of a MeOH/H2O/formic acid mixture
(50:49:1) and filtered using a spin filter with a pore size of 0.2
µm. Samples were stored at -20 °C before being analyzed by
Q-TOF MS as described above.
Glu -C Digestion of Globin an d An alysis by MALDI-
TOF MS. After on-probe digestion for 1 h followed by
mass analysis using MALDI-TOF MS, 20 peptide frag-
ments could be assigned to peptides derived from â-3,
for which the amino acid sequence was available in the
data banks (Swiss Prot accession number P02091) (Table
Rea ction w ith DTT. Globin (4 µL, 1 mg/mL) isolated from
rats exposed to DSF was mixed with 8 µL of a NH4HCO3 buffer

240 Chem. Res. Toxicol., Vol. 13, No. 4, 2000
Erve et al.
Ta ble 1. Obser ved Glu -C P ep tid es in DSF -Tr ea ted Ra ts
a n d Con tr ol Globin
with a 99 Da heavier peptide. A small peptide at m/z
2467.9 could be assigned to Glu-C fragment [115-138],
which also contains Cys-125. However, no peptide with
a mass 99 Da higher than this could be observed. The
control globin digest had many of the same Glu-C
fragments as the DSF-treated globin, and the overall
appearance was similar with a few exceptions. For
example, the high-intensity peptide at m/z 4967.0 corre-
sponding to [44-90] in the control spectrum was much
smaller in the DSF-treated globin spectrum, possibly due
to suppression. Peptide fragments seen in the control but
not in the DSF-treated sample include [57-90] and [27-
56], while fragment [91-101] was seen in the DSF-
treated, but not in the control, digest. Solution phase
enzymatic digestion was performed to effect more com-
plete digestion with the objective to generate more of the
peptide Mod[122-146]. For the solution digests, fewer
large peptides were observed by MALDI-TOF MS and
the incompletely digested Mod[102-146] disappeared
over time to generate Mod[122-146] whose abundance
increased, as judged by MALDI-TOF MS (data not
shown).
Glu -C Digestion of HP LC-P u r ified P r otein F r a c-
tion s. Protein-containing fractions collected following
separation by RP-HPLC were assessed by MALDI-TOF
MS. The fractions corresponding to â-2 or â-3 isolated
from control animals were then separately digested
enzymatically on-probe with Glu-C. In both of these
digests, no sign of peptide signals corresponding to Mod-
[122-146] or Mod[102-146] was observed. Similarly,
fractions corresponding to adducted â-2 or â-3 from DSF-
treated rats were also separately digested. In both
digests, a signal at m/z 2716.4, corresponding to Mod-
[122-146], was observed, while no m/z 2617.4 was
identified. However, although a signal at m/z 4959.1
corresponding to Mod[102-146] was observed in both
digests from exposed animals, a small signal at m/z
4859.8 representing the corresponding unmodified pep-
tide was also observed.
fragment
observed
predicted
∆Da
∆ppm
91-101a
1306.5
1927.4
2098.5
2261.8
2467.9
2505.0
2603.2
2617.4
2716.4
3322.8
3363.1
3549.1
3743.1
3769.8
4007.1
4413.2
4561.1
4577.2
4859.8
4959.1
4967.7
5184.8
5848.9
7047.1
8955.1
1306.5
1927.1
2098.5
2261.7
2467.9
2504.9
2603.2
2617.1
2716.2
3322.8
3362.9
3549.2
3743.1
3769.3
4006.6
4413.2
4561.1
4577.3
4859.8
4958.9
4967.0
5185.0
5848.7
7047.1
8955.2
-0.01
-0.3
-0.02
-0.10
0.00
-0.18
-0.13
-0.25
-0.25
-0.02
-0.13
0.07
-9
-128
-10
-44
0
-70
-50
-96
-89
-6
-38
20
51
130
-137
-57
28
80-96
27-43b
102-121
115-138
27-47 or 80-101
99-121
122-146
Mod[122-146]a
27-56c
48-79
91-121
57-90c
0.19
44-79
-0.49
-0.55
-0.25
0.13
8-43
8-47
48-90
48-90d
0.10
22
102-146
Mod[102-146]a
44-90
-0.05
-0.22
0.64
-10
-44
128
26
27-73
0.13
27-79 or 48-101
-0.15
-0.01
-0.10
-26
27-90b
-2
8-90
7
a
b
DSF globin digest only. Used as an internal standard.
c Control digest only. Contains oxidized Met.
d
Ta n d em MS of Glu -C Digests. The DSF-treated
globin Glu-C digest was analyzed by nano-ESI. The
peptide corresponding to Mod[122-146] ionized to pro-
duce a +4 ion at m/z 679.3. This peptide was selected by
Q1 and subjected to collision-induced dissociation (CID)
in the collision cell, resulting in extensive fragmentation.
The resolution was sufficient to determine the charge
states of the individual fragment ions. The resulting
spectrum was deconvoluted using MaxEnt sequence
software that converts multiply charged ion signals to
singly charged ion signals (Figure 3). Following decon-
volution, a nearly complete sequence of yn fragment ions
could be identified, consistent with the sequence Thr-Pro-
Cys-Ala-Gln-(Ala)2-Phe-Gln-Lys-(Val)2-Ala-Gly-Val-Ala-
Ser-Ala-Leu-Ala-His-Lys-Tyr-His (Table 2). The fragment
ions at m/z 2369, 2466, and 2567 were shifted 99 Da
higher than the predicted y22-y24 ions, respectively.
These latter yn ions all contain Cys-125 near the N-
terminus. The fragment at m/z 584.3 was consistent with
either y4 or b4 shifted by 99 Da. Other shifted bn ions
that were observed were b7-b9, b12-b14, b20, b23, and b24.
The low-mass region of the spectrum did not contain any
ion signals at m/z 72, 100, or 175 as observed previously
in the tandem MS spectrum for GSH conjugates of DSF
reactive metabolites (12), or the modified octapeptide
previously reported for DETC-MeSO2-modified ALDH
(14).
F igu r e 2. MALDI/TOF MS spectra of Glu-C digests of rat
globin from DSF-treated rats (6 week DSF administration in
feed at 1% w/w) (A) and control rat (B). Peptides are labeled
with the corresponding sequence of the â-3 globin chain.
Peptides [27-43] (m/z 2098.5) and [27-90] (m/z 7047.1, not in
spectrum) were used for internal calibration. The peptides Mod-
[122-146] and Mod[102-146] have a 99 Da mass increment
above those of their corresponding unmodified peptides. NI
denotes not identified. Spectra were acquired on a PerSeptive
Voyager Elite mass spectrometer as described in Materials and
Methods.
1). These fragments had mass errors ranging from -0.55
to 0.13 Da and covered 100% of the amino acid sequence.
Two peptide signals with high intensities were observed
at m/z 2716.4 and 4959.1 in globin from a rat exposed to
DSF (Figure 2A) that were not present in digests from
control rat globin (Figure 2B). Both new peptides had
molecular masses 99 Da higher than the peptides [122-
146] and [102-146] observed at m/z 2617.4 and 4859.8,
respectively. Both of these peptides contain Cys-125 and
were designated as the modified peptides, Mod[122-146]
and Mod[102-146]. Peptides at m/z 1306.5, 1927.4, and
3549.1 were identified as fragments [91-101], [80-96],
and [91-121], respectively, all of which contain Cys-93.
None of these Cys-93-containing peptides could be paired

Disulfiram Metabolite Adduct Hemoglobin in Vivo
Chem. Res. Toxicol., Vol. 13, No. 4, 2000 241
F igu r e 3. Charge-deconvoluted tandem MS spectrum of the
peptide designated Mod[122-146] in which its +4 charge state
(m/z 709) was fragmented. The spectrum was acquired on a
Micromass Q-TOF instrument equipped with a nano-electro-
spray source. The charge-deconvoluted spectrum exhibits singly
charged sequence ions. The gap of 202 Da between y21 and y22
is caused by the adduct on Cys-125.
Ta ble 2. Obser ved a n d P r ed icted yn F r a gem en t Ion s
F ollow in g CID of Mod [122-146]
yn
amino acid
observed
predicted
∆Da
∆ppm
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
His
Tyr
Lys
His
Ala
Leu
Ala
Ser
Ala
Val
Gly
Ala
Val
Val
Lys
Gln
Phe
Ala
Ala
Gln
Ala
Cys
Pro
Thr
156.09
319.10
447.23
584.30
655.34
768.44
na
925.48
997.52
1096.57
1153.61
1224.69
1323.72
1422.81
1550.91
naa
1826.02
1897.07
1968.11
2096.21
2167.23
2369.30b
2466.36b
2567.39b
156.08
319.14
447.24
584.30
655.33
768.42
839.45
926.49
997.52
1096.59
1153.61
1224.65
1323.72
1422.79
1550.88
1678.94
1826.01
1897.05
1968.08
2096.14
2167.18
2369.26b
2466.31b
2567.36b
0.01
-0.14
0.01
0.005
0.01
0.02
-
-0.01
-0.002
-0.02
-0.005
0.04
0
96.1
-128
-6.7
8.6
16.8
28.6
-
-10.8
-2.0
-21.9
-4.3
36.7
0
F igu r e 4. Spectrum of DETC-Cys (m/z 221) acquired on a
Micromass Q-TOF instrument equipped with a nano-ESI source
as described in Materials and Methods. Spectra are from a
standard (A), a control rat (B), a rat after the 2 week DSF
administration (C), and a rat after the 4 week DSF administra-
tion (D). The fragment ion at m/z 100 corresponds to cleavage
of the diethyl carbamoyl adduct and is observed in the standard
and hydrolysates prepared from exposed rats (C and D), but is
not in the control (B).
0.02
0.03
-
17.6
20.0
-
0.01
0.02
0.03
0.07
0.05
0.04
0.05
0.03
4.9
reduction of DSF (1) (Scheme 1) by GSH, or albumin, to
DEDC (2) which can be methylated by several enzymes
to DEDC-Me (3). Oxidation of the thiocarbonyl carbon
by P450s of DEDC-Me generates DETC-Me (4). Further
thioxidation of the sulfur can generate the sulfoxide,
DETC-MeSO (5), or the sulfone, DETC-MeSO2 (6). Either
5 or 6 may react with Cys on GSH (7), or sulfhydryl-
containing proteins (8), to generate an N,N-diethyl-
carbamoyl Cys residue. In work reported here, following
acid hydrolysis of intact globin, DETC-Cys (9) could be
released and quantified.
Following administration of DSF to rats, MALDI-TOF
MS analysis of HPLC-purified proteins demonstrated the
presence of two new protein species with molecular
masses 99 Da higher than those of â-2 and â-3 that were
not present in samples from control rats. The 99 Da
molecular mass increase suggested the presence of a
diethylcarbamoyl adduct (calculated as 99 Da) on Cys.
Both of these globin chains contain a Cys-125 (19), which
is a highly reactive nucleophile (20). Although â-1 also
contains a Cys-125, no evidence for adduct formation was
observed on this chain. This apparent lack of reactivity
could be due to slight differences in structure that
prevent accessibility of Cys-125 by the reactive metabo-
lite. It is also possible, however, that low levels of
adduction took place, but could not be detected. The
15.3
15.2
32.4
21.7
18.2
20.3
14.0
a
b
Not observed. Contains adduct.
Ta n d em MS of DETC-Cys a n d Globin Acid Hy-
d r olysa tes. The parent ion of authentic DETC-Cys at
m/z 221.0 was subjected to CID to produce fragment ions
at m/z 204.0, 175.0, 134.0, 100.0, 76.0, 74.0, and 72.0
(Figure 4A). The control hydrolysate produced many
fragment ions with m/z values close to those seen in the
standard, but ions at m/z 204.0, 100.0, and 76.0 were
absent (Figure 4B). In addition, ions were observed at
m/z 132.0, 118.0, 116.0, 90.0, and 86.0. Acid hydrolysates
from globin obtained from treated rats (2 and 4 weeks)
displayed all the ions observed in the standard (Figure
4C,D) and control globin hydrolysates.
Discu ssion
The metabolism of DSF to the electrophilic metabolites
that have been suggested to inhibit ALDH in vivo has
been described in detail (2, 5, 18). The first step is the

242 Chem. Res. Toxicol., Vol. 13, No. 4, 2000
Erve et al.
major R-chain, and a slightly smaller protein that could
represent the minor R-chain, were observed. However,
no evidence of adduction on either of these chains was
detected by MALDI-TOF MS. Although two Cys residues
are present on the R-chain, they are both masked and
hence not reactive (20).
Others have reported the reversibility of the diethyl-
carbamoyl Cys adduct following incubation of adducted
yeast ALDH with the reducing agent DTT at 37 °C (21).
In our work, the adduct could not be removed with DTT
(pH 8.5 or 10.5) after 2 h at either 37 or 56 °C as judged
by MALDI-TOF MS analysis following incubation. Such
an observation may be explained by some unique feature
of ALDH, such as the three adjacent active site Cys
residues (14), which make the adduct labile compared to
the adduct on globin. Another factor may relate to
differences in how the protein was handled prior to
incubation with DTT. In our work, the globin had been
precipitated and stored lyophilized for several months
prior to analysis.
To locate the adduct observed on â-2 and â-3, the
proteins were enzymatically digested into smaller pep-
tides. The endoproteinase Glu-C has been used success-
fully to digest a variety of â-chains (22) which at pH 7.8
in ammonium bicarbonate buffer cleave at the carboxy
side of Glu (23). Adding Glu-C directly to the unfraction-
ated globin on the MALDI-TOF MS probe (see Materials
and Methods for details) resulted in extensive digestion
of the â-chains with minimal digestion of the R-chains.
This finding allowed mass analysis of the â-chains
without the prior need to separate the individual sub-
units using chromatography. Following Glu-C digestion,
two peptides were observed whose molecular masses were
99 Da higher than those of two peptide fragments known
to contain Cys-125. The larger of these peptides was an
incomplete digestion product in which one Glu-C cleavage
site had not been cleaved. Although several â-chains had
been digested simultaneously, because of sequence ho-
mology, many of the predicted Glu-C peptide fragments
were identical, including the peptide containing Cys-125.
It was apparent from the MALDI-TOF MS spectrum that
these two modified peptides were of particularly high
intensity even though they only represented a small
fraction of the total peptides present. This fortuitous
occurrence may be explained by the increased ionizing
potential of the peptide following addition of the less polar
diethylcarbamoyl moiety to the polar sulfhydryl group
of Cys.
Individual â-2 and â-3 chains were HPLC purified and
digested with Glu-C to show that neither these chains
gave rise to any modified peptides. Likewise, individual
adducted â-2 and â-3 chains were purified prior to Glu-C
digestion, and both modified chains gave rise exclusively
to Mod[122-146] at m/z 2617.4 with no sign of unmodi-
fied [122-146]. Regarding the larger Mod[102-146],
resulting from one missed cleavage, both of the modified
chains gave rise to a peptide at m/z 4859.8, but a small
amount of unmodified [102-146] was also observed
(<5%). This may have arisen due to slight contamination
of the unmodified protein.
sample is sprayed (on the order of several micrometers),
which leads to sample flow rates on the order of tens of
nanoliters per minute (24). Thus, the unseparated digest
was analyzed, allowing the peptide of interest at m/z
679.3 to be identified and an optimized fragmentation
spectrum recorded. Interpretation of the resulting peptide
fragment spectrum facilitated determination of the amino
acid sequence and location of the adduct at Cys-125.
However, careful examination of the mass spectrum did
not reveal any fragment ions at low m/z values charac-
teristic of diethylthiocarbamoyl Cys that had previously
been reported (12, 14). Consequently, although the loca-
tion of a 99 Da adduct at Cys-125 was unambiguous on
the basis of these sequence results, the nature of the
adduct could not be defined with certainty from this
tandem MS spectrum alone. A possible explanation for
the lack of these diagnostic fragment ions might lie in
the larger size of our peptide, which contained 24 amino
acids. For such a large peptide, the energy imparted
during collision may not have been efficiently distributed
to an extent sufficient to cause fragmentation of the
adduct. Previous work in which fragment ions charac-
teristic of DETC-Cys were obtained on smaller analytes
such as on GSH conjugates (12), or a small octapeptide
(14), has been reported.
To confirm the structure of the proposed adduct, the
globin mixture was hydrolyzed to produce an amino acid
hydrolysate. Although it was not known if DETC-Cys was
stable to such harsh conditions, the model peptide Arg-8
vasopressin, which contains two free Cys residues fol-
lowing reduction, was modified with DETC-MeSO2. Fol-
lowing overnight acid hydrolysis at 110 °C, DETC-Cys
could be detected (data not shown). Therefore, globin from
control and treated rats was hydrolyzed and examined
for the presence of DETC-Cys. In particular, m/z 221.0,
corresponding to DETC-Cys, was selected in Q1 and a
tandem MS spectrum was acquired. The tandem MS
spectrum corresponding to authentic DETC-Cys produced
a readily interpretable spectrum which included the
characteristic fragment ions previously reported, such as
m/z 100 which arises from loss of the diethylcarbamoyl
adduct [(CH3CH2)2NdCdO] +. Performing the same ex-
periment on globin from treated rats (2, 4, and 6 weeks,
n ) 2) resulted in the observation of these identical
fragment ions, indicating that DETC-Cys was indeed
present in the hydrolysate. The fragment ion at m/z 100
was the base peak in both the spectrum of the standard
globin and the spectrum of the treated globin. In contrast,
globin from control rats contained only some of these
fragment ions and, most importantly, did not contain m/z
100. A number of fragment ions were observed in the
globin samples that were not observed in the standard,
most likely arising from interfering ions with m/z values
very close to 221.0. Although these results are not
quantitative and do not demonstrate a dose-response
relationship between the adduct and the duration of
treatment, qualitatively the adduct can be detected as
early as 2 weeks following administration.
Of the three Cys residues on the R-chains and up to
two on the â-chains, we found evidence for adduction only
at Cys-125. Recent work has classified the reactivity of
the sulfhydryl group of Cys-125 as “fast reacting” because
it reacts 4000 times faster with dithiobis(2,2′-nitrobenzoic
acid) (DTNB) than the other free sulfhydryl groups at
Cys-93 (â-chain) or Cys-13 (R-chain) (20). In fact, that
work also reported that Cys-125 reacts 100 times faster
To verify that Mod[122-146] did contain DETC-Cys
at position 125, sequence information was obtained. A
solution digestion of the globin mixture was performed
and the resulting digest mixture analyzed using a nano-
ESI source. The unique feature of nano-ESI is the very
tiny diameter of the glass capillary from which the

Disulfiram Metabolite Adduct Hemoglobin in Vivo
Chem. Res. Toxicol., Vol. 13, No. 4, 2000 243
with DTNB than GSH, which in light of the greater
concentration of Cys-125 in erythrocytes relative to GSH
(∼8 vs 2 mM) suggests that Cys-125 on hemoglobin is
more efficient than GSH in trapping reactive species in
the erythrocyte. Two other investigations have shown
that Cys-125 was the major site of adduction for the diol
epoxide metabolite of fluoranthene (25) and the triazene
herbicide metabolite, symmetrin sulfoxide (26). In con-
trast, Hamboek et al. (27) in investigating the reactivity
of p-mercuribenzoate with hemoglobins observed this
reagent to react 100-fold faster with Cys-93, suggesting
that the relative reactivity of a particular cysteinyl
residue may be dependent on the nature of the reactant.
It may be possible to consider the reaction between
hemoglobin and the DSF metabolites DETC-MeSO or
DETC-MeSO2 observed in this work as a detoxification
reaction. Since the lifetime of the stable protein adduct
would be expected to equal that of hemoglobin (60 days
in rats, 120 days in humans), and is much longer than
that of the metabolites (28), the protein adduct might be
applied to biomonitoring human exposure, or forensic
applications. However, because human hemoglobin lacks
the reactive Cys-125 that is present in rats, additional
experiments will be required to investigate the reactivity
of DETC-MeSO or DETC-MeSO2 toward human hemo-
globin. Further work will also be required to identify
sulfhydryl-containing proteins, in addition to ALDH, that
could become compromised following carbamylation by
DSF metabolites which may lead to the idiosyncratic
hepatotoxic reactions associated with DSF and Schwann
cell toxicity observed in rats treated with disulfiram (29,
30).
Refer en ces
(1) Hald, J ., J acobsen, E., and Larsen, V. (1948) The sensitizing effect
of tetraethylthiuram disulfide (Antabuse) to ethyl alcohol. Acta
Pharmacol. Toxicol. 4, 285-296.
(2) J ohansson, B. (1992) A review of the pharmacokinetics and
pharmacodynamics of disulfiram and its metabolites. Acta Psy-
chiat. Scand. 86, 15-26.
(3) Hald, J ., and J acobsen, E. (1948) The formation of acetaldehyde
in the organism after ingestion of Antabuse (Tetraethylthiuram-
disulfide) and alcohol. Acta Pharmacol. and Toxicol. 4, 305-310.
(4) Vallari, R. C., and Pietruszko, R. (1982) Human aldehyde dehy-
drogenase: Mechanisms of inhibition by disulfiram. Science 216,
637-639.
(5) Hart, B. W., and Faiman, M. D. (1993) Bioactivation of S-methyl
N,N-diethylthiolcarbamate to S-methyl N,N-diethylthiocarbamate
sulfoxide. Biochem. Pharmacol. 46, 2285-2290.
(6) Hu, P., J in, L., and Baillie, T. A. (1997) Studies on the metabolic
activation of disulfiram in rat. Evidence for electrophilic S-
oxygenated metabolites as inhibitors of aldehyde dehydrogenase
and precursors of urinary N-acetylcysteine conjugates. J . Phar-
macol. Exp. Ther. 281, 611-617.
(7) Madan, A., Parkinson, A., and Faiman, M. D. (1998) Identification
of the human P-450 enzymes responsible for the sulfoxidation and
thiono-oxidation of diethyldithiocarbamate methyl ester: Role of
P-450 enzymes in disulfiram bioactivation. Alcohol.: Clin. Exp.
Res. 22, 1212-1219.
(8) Cobby, J ., Mayerson, M., and Selliah, S. (1977) The rapid
reduction of disulfiram in blood and plasma. J . Pharmacol. Exp.
Ther. 202, 724-731.
(9) Agarwal, R. P., McPhearson, R. A., and Phillips, M. (1983) The
rapid reduction of disulfiram by serum albumin. Res. Commun.
Chem. Pathol. Pharmacol. 42, 293-310.
(10) Gessner, T., and J akubowski, M. (1972) Diethyldithiocarbamic
acid methyl ester. A metabolite of disulfiram. Biochem. Pharma-
col. 21, 219-230.
(11) Glauser, T. A., Nelson, A. N., Zembower, D. E., Lipsky, J . J ., and
Weinshilboum, R. M. (1993) Diethyldithiocarbamate S-methyla-
tion: evidence for catalysis by human liver thiol methyltrans-
ferase and thiopurine methyltransferase. J . Pharmacol. Exp.
Ther. 266, 23-32.
(12) J in, L., Davis, M. R., Hu, P., and Baillie, T. A. (1994) Identification
of novel glutathione conjugates of disulfiram and diethyldithio-
carbamate in rat bile by liquid chromatography-tandem mass
spectrometry. Chem. Res. Toxicol. 7, 526-533.
(13) Mays, D. C., Nelson, A. N., Lam-Holt, J . P., Fauq, A. H., and
Lipsky, J . J . (1996) S-Methyl-N,N-diethylthiocarbamate sulfoxide
and S-methyl-N,N-diethylthiocarbamate sulfone, two candidates
for the active metabolite of disulfiram. Alcohol.: Clin. Exp. Res.
20, 595-600.
(14) Tomlinson, A. J ., J ohnson, K. L., Lam-Holt, J ., Mays, D. C.,
Lipsky, J . J ., and Naylor, S. (1997) Inhibition of human mito-
chondrial aldehyde dehydrogenase by the disulfiram metabolite
S-methyl-N,N-diethylthiocarbamoyl sulfoxide. Biochem. Phar-
macol. 54, 1253-1260.
(15) Berlin, R. G. (1989) Disulfiram hepatotoxicity: A consideration
of its mechanism and clinical spectrum. Alcohol Alcohol. 24, 241-
246.
(16) Eliasson, E., Stahl, P., Oksaanen, A., and Lytton, S. (1998)
Expression of autoantibodies to specific cytochromes P450 in a
case of disulfiram hepatitis. J . Hepatol. 29, 819-825.
(17) Gobom, J ., Nordhoff, E., Mirgorodskya, E., Ekman, R., and
Roepstorff, P. (1999) Sample purification and preparation tech-
niques based on nano-scale reversed-phase columns for the
sensitive analysis of complex peptide mixtures by matrix-assisted
laser desorption/ionization mass spectrometry. J . Mass Spectrom.
34, 105-116.
(18) Mays, D. C., Nelson, A. N., Fauq, A. H., Schriver, Z. H., Veverka,
K. A., Naylor, S., and Lipsky, J . J . (1995) S-methyl N,N-
diethylthiocarbamate sulfone, a potential metabolite of disulfiram
and potent inhibitor of low Km mitochondrial aldehyde dehydro-
genase. Biochem. Pharmacol. 49, 693-700.
(19) Ferranti, P., Carbone, V., Sannolo, N., Fiume, I., and Malorni, A.
(1993) Mass spectrometric analysis of rat hemoglobin by FAB-
overlapping. Int. J . Biochem. 25, 1943-1950.
(20) Rossi, R., Barras, D., Bellelli, A., Boumis, G., Canofeni, S.,
Simplicio, P. D., Lusini, L., Pascarella, S., and Amiconi, G. (1998)
Fast-reacting thiols in rat hemoglobins can intercept damaging
species in erythrocytes more efficiently than glutathione. J . Biol.
Chem. 271, 19198-19206.
Con clu sion
Detection of hemoglobin modifications can help to
elucidate mechanisms of toxicity by clarifying the nature
of reactive metabolites formed in vivo, in addition to
serving as a means of biomonitoring exposure to chemi-
cals. The major finding of this investigation is the
identification of an N,N-diethylcarbamoyl adduct on Cys-
125 of rat globin following administration of DSF. The
adducted amino acid, DETC-Cys, most likely was formed
through reaction with DETC-MeSO, although DETC-
MeSO2 would generate the same adduct and, thus, cannot
be ruled out as the source of the adduct. Nevertheless,
observation of this adduct supports the potential rel-
evance of protein modification by metabolites of DSF in
vivo and suggests that other Cys-containing proteins
besides ALDH may be targeted, resulting in altered
function. The ability to liberate DETC-Cys following acid
hydrolysis may provide the basis for the development of
a quantitative assay, which in turn could allow the
determination of the dose-response relationship for this
adduct, analyzing adducts on target proteins, and moni-
toring compliance of disulfiram therapy.
Ack n ow led gm en t. J .C.L.E. thanks all members of
the Protein Research Group, Odense University, espe-
cially Drs. Katya Mirogordskya, J ohan Gobom, and Dario
Kalume for expert technical assistance and valuable
discussions during the time this work was being per-
formed. These studies were supported by NIH Grant
ESO6387 and Center in Molecular Toxicology Grant P30
ES00267. In addition, J .C.L.E. was supported, in part,
by NRSA Grant ES05764.
(21) Veverka, K., J ohnson, K., Mays, D., Lipsky, J ., and Naylor, S.
(1997) Inhibition of aldehyde dehydrogenase by disulfiram and

244 Chem. Res. Toxicol., Vol. 13, No. 4, 2000
Erve et al.
its metabolite methyl diethyldithiocarbamoyl-sulfoxide. Biochem.
Pharmacol 53, 511-518.
(22) J ensen, O. N., Højrup, P., and Roepstorff, P. (1991) Plasma
desorption mass spectrometry as a tool in characterization of
abnormal proteins. Application to variant human hemoglobins.
Anal. Biochem. 199, 175-183.
(23) Houmard, J ., and Drapeau, G. R. (1972) Staphylococcal Pro-
tease: A proteolytic enzyme specific for glutamoyl bonds. Proc.
Natl. Acad. Sci. U.S.A. 69, 3506-3509.
a specific thiol group in rat hemoglobin. Biochem. J . 199, 61-67.
(27) Hamboek, H., Fischer, R. W., Di J orio, E. E., and Winterhalter,
K. H. (1981) The binding of s-triazine metabolites to rodent
hemoglobins appears irrelevant to other species. Mol. Pharmacol.
20, 579-584.
(28) Petersen, E. N. (1992) The pharmacology and toxicology of
disulfiram and its metabolites. Acta Psychiatr. Scand., Suppl. 86,
7-13.
(29) Anzil, A. P. (1985) Morphological assessment of neurotoxicity:
Disulfiram neuropathy as an animal model of human toxic
axonopathies. In Neurotoxicology (Blum, K., and Manzo, L., Eds.)
pp 535-558, Marcel Dekker, New York.
(30) Tonkin, E. G., Amarnath, K., Valentine, H., and Valentine, W.
M. (1999) Schwann cells are the initial target in disulfiram-
mediated axonal degeneration. ISSX Proceedings of the 9th North
American ISSX Meeting, Vol. 15, p 165 (no. 328).
(24) Wilm, M., and Mann, M. (1996) Analytical properties of the
nanoelectrospray ion source. Anal. Chem. 68, 1-8.
(25) Hutchins, D. A., Skipper, P. L., Naylor, S., and Tannenbaum, S.
R. (1988) Isolation and characterization of the major fluoranthene-
hemoglobin adducts formed in vivo in the rat. Cancer Res. 48,
4756-4761.
(26) Hughes, G. J ., De J ong, C., Fisher, R. W., Winterhalter, K. H.,
and Wilson, K. J . (1981) Modification by simetrin sulphoxide of
TX990191N