Identification of a new urinary metabolite of carbon disulfide using an improved method for the determination of 2-thioxothiazolidine-4-carboxylic acid

Article abstract of DOI:10.1021/tx010085x

A new method is reported for the analysis of 2-thioxothiazolidine-4-carboxylic acid (TTCA) in urine that is amenable to automation and provides greatly simplified chromatograms. The method comprises the addition of tetrahydro-2-thioxo-2H-1,3-thiazine-4-carboxylic acid, which is chemically similar to TTCA, as internal standard, purification on an Oasis HLB solid-phase extraction column, and analysis by HPLC with UV detection. The limit of detection for TTCA was 40 pmol/mL of urine, recovery was 79.3 ± 1.0%, and detection was linear over at least 3 orders of magnitude. In addition, during the analysis of urine samples from workers exposed to CS₂, a novel urinary metabolite of CS₂ was recognized. The new metabolite demonstrated a dose response, was present at approximately 30% the level of TTCA, and was characterized to be 2-thioxothiazolidin-4-ylcarbonylglycine (TTCG). Administration of TTCG to rats resulted in excretion of TTCA suggesting that TTCG is a likely precursor of TTCA. Although urinary excretion of both TTCA and TTCG resulted from administration of captan, only TTCA was detected following administration of methyl isothiocyanate. The greater selectivity of TTCG suggests that co-analysis of TTCA and TTCG in urine may aid in differentiating exposures to CS₂, captan and isothiocyanates.

Full text of DOI:10.1021/tx010085x

Chem. Res. Toxicol. 2001, 14, 1277-1283
1277
Id en tifica tion of a New Ur in a r y Meta bolite of Ca r bon
Disu lfid e Usin g a n Im p r oved Meth od for th e
Deter m in a tion of 2-Th ioxoth ia zolid in e-4-ca r boxylic Acid
Venkataraman Amarnath,*,† Kalyani Amarnath,† Doyle G. Graham,† Qiping Qi,‡
Holly Valentine,† J ing Zhang,† and William M. Valentine†
Department of Pathology and Center in Molecular Toxicology, Vanderbilt University Medical
Center, Nashville, Tennessee 37232-2561, and Institute of Environment Health Monitoring,
Chinese Academy of Preventive Medicine, Bejing, 100021, China
Received April 25, 2001
A new method is reported for the analysis of 2-thioxothiazolidine-4-carboxylic acid (TTCA)
in urine that is amenable to automation and provides greatly simplified chromatograms. The
method comprises the addition of tetrahydro-2-thioxo-2H-1,3-thiazine-4-carboxylic acid, which
is chemically similar to TTCA, as internal standard, purification on an Oasis HLB solid-phase
extraction column, and analysis by HPLC with UV detection. The limit of detection for TTCA
was 40 pmol/mL of urine, recovery was 79.3 ( 1.0%, and detection was linear over at least 3
orders of magnitude. In addition, during the analysis of urine samples from workers exposed
to CS2, a novel urinary metabolite of CS2 was recognized. The new metabolite demonstrated
a dose response, was present at approximately 30% the level of TTCA, and was charaterized
to be 2-thioxothiazolidin-4-ylcarbonylglycine (TTCG). Administration of TTCG to rats resulted
in excretion of TTCA suggesting that TTCG is a likely precursor of TTCA. Although urinary
excretion of both TTCA and TTCG resulted from administration of captan, only TTCA was
detected following administration of methyl isothiocyanate. The greater selectivity of TTCG
suggests that co-analysis of TTCA and TTCG in urine may aid in differentiating exposures to
CS2, captan and isothiocyanates.
Even with these limitations, determination of urinary
TTCA has become the method of choice for monitoring
workers’ exposure to CS2 (5, 6), and a biological exposure
index of 5 mg of TTCA/g of creatinine in urine has been
established.
In tr od u ction
Identification of 2-thioxothiazolidine-4-carboxylic acid
(TTCA,1 raphanusamic acid) in the urine of workers
Recently, while undertaking to estimate the concentra-
tions of TTCA in more than 200 urine samples from
workers of a rayon factory in the Peoples Republic of
China, we found that the currently available assays
(Table 1) did not satisfy all our requirements of sensitiv-
ity, reproducibility, and ease of operation. Therefore, a
new HPLC based method was developed that includes a
chemically similar internal standard, removes many
closely eluting impurities before chromatography, and
gives highly reliable and reproducible values. The method
can be automated to process multiple samples obtained
when assessing occupational exposures. In addition, since
the chromatograms obtained using the new method
contained only a few well-separated peaks, it was possible
to recognize a new peak in addition to that of TTCA that
was unique to urine samples obtained from exposed
workers. This paper provides evidence to show that the
extra peak arose from a novel urinary metabolite of CS2.
Structural characterization of the new metabolite sheds
light on the possible mechanism of formation of TTCA
exposed to carbon disulfide introduced a biomarker for
exposure to this important industrial solvent (1, 2).
Although the quantity of TTCA has been correlated with
the level of exposure to CS2, the details of the steps
leading to the formation of TTCA from CS2 are still not
completely understood. As one would expect TTCA is also
excreted in urine on exposure to those chemicals that
release CS2 in vivo, such as, disulfiram (used in rubber
manufacture or administered orally in alcohol aversion
therapy) and dithiocarbamates (utilized as pesticides) (3).
The fungicide Captan (N-trichloromethylthio-4-cyclohex-
ene-1,2-dicarboximide) that undergoes hydrolytic cleav-
age to thiophosgene also generates TTCA in vivo. Excre-
tion of TTCA has been reported following consumption
of cruciferous vegetables possibly as a result of preformed
TTCA or glucosinolates present in the vegetables (4).
1
Abbeviations: cys-gly, cysteinylglycine; DMF, N,N-dimethylfor-
mamide, GSH, glutathione; HPLC, high-performance liquid chroma-
tography; MITC, methyl isothiocyanate; PBS, physiologically buffered
saline, TTCA, 2-thioxothiazolidine-4-carboxylic acid; TTCG, 2-thioxothi-
azolidin-4-ylcarbonylglycine; T3CA, tetrahydro-2-thioxo-2H-1,3-thi-
azine-4-carboxylic acid; TLC, thin-layer chromatography; UV, ultra-
violet.
* To whom correspondence should be addressed. E-mail:
venkataraman.amarnath@mcmail.vanderbilt.edu.
† Department of Pathology and Center in Molecular Toxicology.
‡ Institute of Environment Health Monitoring.
10.1021/tx010085x CCC: $20.00 © 2001 American Chemical Society
Published on Web 08/22/2001

1278 Chem. Res. Toxicol., Vol. 14, No. 9, 2001
Ta ble 1. Com p a r ison of th e Meth od s Ava ila ble for th e Mea su r em en t of TTCA in Ur in e
Amarnath et al.
ref
method
advantages
disadvantages
4
HPLC with column switching
direct injection of urine
column switching is inconvenient;
dilution of sample and limited column life
requires derivatization with
diazomethane or silyl reagents
several closely spaced peaks
3
8
9
GC-MS of a volatile derivative
less interference from other
components of urine
a detection limit of 1.5 pmol
HPLC-UV detection with
gradient elution
HPLC-UV detection with
isocratic elution
lack of an internal standard requires
exhaustive extraction
and provides complimentary information to using TTCA
alone as a biomarker for assessing CS2 exposure.
Ma ter ia ls a n d Meth od s
Ch em ica ls. Carbon disulfide is volatile, flammable, toxic, and
a skin irritant, MITC is toxic and easily absorbed through the
skin, and Captan (Chem Service, West Chester, PA) is toxic, a
suspected carcinogen and an ocular irritant. Gloves and a fume
hood should be used when handling these compounds. Carbon
disulfide, HPLC-grade solvents, silica for column chromatog-
raphy, and silica coated aluminum sheets for thin-layer chro-
matography were obtained from EM Science (Gibbstown, NJ ).
Sigma Chemical Co. (St. Louis, MO) was the source for cys-gly,
corn oil, 1,2-propanediol, and premeasured powder for phos-
phate buffered saline and Fluka-Aldrich (Milwaukee, WI) for
the other chemicals used. The syntheses of TTCA (7) and T3CA
(3) have been published.
2-Th ioxoth ia zolid in -4-ylca r bon ylglycin e Eth yl Ester .
TTCA (10 mmol) was dissolved in DMF (20 mL) and stirred with
dichloromethane (20 mL) and triethylamine (3.5 mL, 25 mmol)
before cooling in ice. Isobutyl chloroformate (1.35 mL, 10 mmol)
was added and the mixture was stirred for 1 h. Glycine ethyl
ester hydrochloride (1.54 g, 11 mmol) was added and the stirring
was continued as the ice melted and the reaction mixture
warmed to room temperature. After 16 h, solvents were removed
and the residue was dissolved in ethyl acetate (30 mL) that was
washed with brine containing 1 M NaHCO3 (20 mL), dried, and
evaporated. The crude product was purified by flash chroma-
tography (50-0% hexane in ethyl acetate). The fractions
containing the pure product (with Rf of 0.5 on TLC in ethyl
acetate) were combined and evaporated to a viscous oil; 1.74 g
(70%); 1H NMR (CDCl3) δ 1.28 (t, 3H, J ) 9.5 Hz, CH3), 3.79
(dd 1H, J ) 6.6 and 15.2 Hz, C5-H), 3.96 (dd, 1H, J ) 12.1 and
15.2, C5-H), 3.98 (m, 2 H, NCH2), 4.10 (q, 2H, J ) 9.5 Hz, OCH2),
4.91 (dd, 1H, J ) 6.6 and 12.1 Hz, C4-H); 13C NMR δ 14.5 (CH3),
37.6 (C-5), 42.1 (NH-CH2), 62.4 (OCH2), 65.6 (C-4), 170.0 and
170.2 (CdO), 202.9 (C-2).
F igu r e 1. Sequence of steps used to analyze for TTCA and
TTCG in urine samples.
TTCA An a lysis. Urine samples were centrifuged for 5 min
at 4000g and frozen at -20 °C prior to TTCA analysis. The solid-
phase extraction steps prior to the chromatographic analysis
(Figure 1) were performed on multiple samples using a LiChro-
lut sample preparation unit with drying attachment (EM
Science). A Waters 2690 LC with 996 diode array detector and
Millennium software was used for the chromatographic analysis.
Two columns (a Lichrosphere 100RP-8 10 µm 4 × 250 mm
column connected to a Whatman Partisil 5 ODS-3 µM 4.6 × 250
mm column) were used. The solvent systems were 3.0% aceto-
nitrile (human samples) or 2.5% acetonitrile (rat samples) with
1% acetic acid (A) and 95:4:1 methanol-water-acetic acid (B).
The elution profile consisted of running 100% A isocratically
for 18 min, 100% B for 6 min, and then 100% A again for 10
min before the next injection.
Recover y. TTCA, T3CA, and TTCG (10 nmol each) were
added to 2.0 mL of a 40% methanol in water solution or to 0.5
mL of urine. The former was analyzed by HPLC directly, while
the latter was processed with Oasis cartridge and the eluate
was diluted to 2 mL with water before analysis. The peak areas
of the two analytes were compared to arrive at the percentage
of recovery which was average of three measurements.
Sta n d a r d Cu r ves. To generate a standard curve for the
determination of TTCA and TTCG in human or rat urine
samples, 0.1-1.0 nmol of TTCA and TTCG were added to 0.5
mL aliquots of control urine. These were then processed using
the procedure in Figure 1. Peak area ratios of TTCA/T3CA and
2-Th ioxoth ia zolid in -4-ylca r bon ylglycin e. Ethyl ester of
TTCG (0.5 g, 2 mmol) was dissolved in dioxane (5 mL) and
stirred with 1 N NaOH (3 mL) for 2 h. The reaction mixture
was washed with ethyl acetate (2 × 5 mL), acidified to pH 2
with 2 N HCl, and extracted with ethyl acetate (3 × 10 mL).
The crude TTCG obtained from the combined extracts was
purified on a column of silica (0-30% methanol in ethyl acetate);
1
0.33 g (75%) of off-white solid; H NMR (D2O) δ 3.58 (dd 1H, J
) 5.5 and 11.8 Hz, C5-H), 3.89 (dd, 1H, J ) 9.4 and 11.8, C5-H),
3.86 (m, 2 H, NCH2), 4.90 (dd, 1H, J ) 5.5 and 9.4 Hz, C4-H);
13C NMR δ 37.3 (C-5), 43.4 (NH-CH2), 66.8 (C-4), 171.5 (CON),
175.9 (CO2H), 203.8 (C-2) MS (negative ion) m/z 219 (M - 1).
En vir on m en ta l Mon itor in g a n d Hu m a n Ur in e Sa m p le
Collection . Monitoring pumps were used to obtain static area
sampling of inhaled air by drawing air at a fixed rate through
a drying tube into a charcoal trap. The pumps were calibrated
before and after the work shift and a tandem charcoal trap used
to ensure that saturation of the initial trap did not occur. CS2
was desorbed from the charcoal and quantified by gas chroma-
tography using a flame photometric detector. Three areas,
clerical, packaging, and spinning, were monitored and urine
samples were obtained from individuals working in each of the
three areas immediately before and after a 12 h work shift.

New Urinary Metabolite of CS2
Chem. Res. Toxicol., Vol. 14, No. 9, 2001 1279
TTCG/T3CA were plotted against their molar ratios (x) to obtain
using UV detection, it is particularly important to
separate TTCA from several other UV absorbing compo-
nents of urine. Solvent extraction of TTCA, a carboxylic
acid, from acidified urine is commonly used. Recovery
with ethyl acetate was reported to be less than 50% (8),
although the yield can be increased by using multiple
extractions with ether (9).
straight lines given by the equations 0.1301 + 1.026x (R2
)
0.998) for TTCA and 0.011 + 1.257x (R2 ) 0.999) for TTCG.
From the peak area ratios of the analyzed samples TTCA and
TTCG in nmoles per 0.5 mL of urine were calculated.
Id en tifica tion of TTCG in Ur in e of Ra ts Tr ea ted w ith
CS2. Urine collected from rats treated with CS2 (see below) was
pooled (∼25 mL), washed with ethyl acetate (2 × 10 mL),
acidified to pH 2, and saturated with salt. It was vigorously
stirred with ethyl acetate (20 mL) for 10 min, centrifuged at
8000 rpm for 10 min, and the organic layer removed. The
supernatants from four such extractions were combined, dried,
and evaporated. The residue was coevaporated with absolute
ethanol (3 × 5 mL), dissolved in 2 mL of the same solvent, and
heated at 70 °C with BF3-OEt2 for 1 h. The residue after the
removal of solvent was purified on a column of silica (80-0%
hexanes-ethyl acetate). The fractions producing spots on TLC
(ethyl acetate) with an Rf of 0.2-0.8 were collected, concen-
trated, and analyzed by HPLC (column: Xterra MS column 2.1
× 100 mm (3.5 µm particle size) with a 2.1 × 10 mm guard
cartridge; flow rate 200 µL/min; isocratic elution 20% acetonitrile
in 5 mM formic acid). A peak with an identical retention time
and UV spectrum to the ethyl ester of TTCG was analyzed by
using a Finnigan TSQ 7000 mass spectrometer (with electro-
spray ionization) and found to produce the same molecular ion
of 249 (M + H)+ and daughter ions of 203 (M - OC2H5) and
175 (203 - CO) observed for the ethyl ester of TTCG.
An im a ls. This study was performed in accordance with the
National Institutes of Health’s Guide for Care and Use of
Laboratory Animals, and was approved by the Institutional
animal care committee. For animal experiments, male Sprague
Dawley rats, 7-9 weeks old and 230-300 g (Harlan, India-
napolis, IN) were used. Rats were housed in a room on a diurnal
light cycle, and while in metabolic cages given finely ground
rodent chow and water ad libitum.
Carbon Disulfide Exposures. Prior to CS2 administration, five
groups of male rats (n ) 3/group) were placed into metabolic
cages for 24 h. A 24 h urine collection was obtained as control
urine and was spun for 5 min at 4000g and frozen at -20 °C
prior to TTCA analysis. Next, each group was administered one
dose of CS2 in corn oil by gavage. Groups 1-5 received 0.04,
0.20, 1.0, 3.0, and 5.0 mmol of CS2/kg, respectively. The rats
were placed back in their metabolic cages and a 24 h urine
sample was collected the next day.
Captan, MITC, and TTCG Exposures. Male rats (n ) 3) were
placed in metabolic cages and control urine collected for 24 h.
A single dose of either MITC (0.1 mmol/kg by i.p. injection using
40 mM MITC in 1,2-propanediol), TTCG (0.4 mmol/kg by i.p.
injection using 50 mM TTCG in water), or Captan (0.05 mmol/
kg p.o. by gavage using 0.166 M Captan in 1:1 v/v 1,2-
propanediol-PBS) was administered. The rats were returned to
their metabolic cages and the urine collected for 24 h and
processed as described above.
Deter m in a tion of Cr ea tin in e. The determination of crea-
tinine in the urine samples was performed using the Sigma
Diagnostics creatinine kit 555-A (Sigma, St. Louis, MO). This
kit uses a modified J affe reaction and was scaled down for use
as a microplate assay on the SpectraMax 250 Plate Reader
(Molecular Devices). A standard curve for linearity and a known
3 mg/dL creatinine standard were run with each sample plate.
From the nmoles of TTCA or TTCG in 0.5 mL of urine samples
and the creatinine values, TTCA and TTCG were expressed as
nmoles per milligram of creatinine in urine.
The sensitivity and ease of chromatography of the last
two methods in Table 1 are usable but both involve time-
consuming and relatively labor-intensive solvent extrac-
tion. Replacing solvent extraction with solid-phase ex-
traction (SPE) to allow the handling of multiple samples
simultaneously using a manifold would expedite sample
preparation. As noted previously (8), a reversed-phase
C-18 silica cartridge did not adsorb TTCA preferentially
to effect purification. Oasis HLB sorbent (from Waters)
exhibiting both hydrophilic and hydrophobic properties
seemed to be a good candidate for isolating TTCA.
Because TTCA is a polar carboxylic acid with a lipophilic
-S-C(dS)- group, it seemed reasonable that through
changing the pH the ability of TTCA to bind to the Oasis
cartridge could be controlled. Preliminary experiments
indicated that nonionized TTCA was readily adsorbed to
Oasis under acidic conditions and that it could be eluted
using an 80% methanol-water mixture, suitable for
concentration.
To account for small variations during extraction and
concentration, an internal standard was added. In previ-
ous studies tetrahydro-2-thioxo-2H-1,3-thiazine-4-car-
boxylic acid (T3CA), a six membered analogue of TTCA,
was used (3). Also for the present analysis T3CA was
chosen, rather than 2-methylhippuric acid (8) as internal
standard, because of its chemical similarity to TTCA.
Further, under the present chromatographic conditions,
it was well separated from TTCA and the UV maxima of
272 and 282 nm for TTCA and TT3CA, respectively, allow
for single wavelength monitoring at 278 nm. On the basis
of preliminary experiments we arrived at the sample
preparation protocol outlined in Figure 1. The recovery
of TTCA was 79.3 ( 1.0% while that of T3CA was 78.5 (
0.3%.
Resu lts of An a lysis. The urine samples from workers
of a rayon factory were analyzed by the new method and
Figure 2 compares the chromatograms of three samples
of subjects exposed to differing ambient levels of CS2. The
chromatograms contain very few peaks other than those
for TTCA and the internal standard T3CA, and the peak
areas could be measured accurately at quantities less
than 100 pmol/0.5 mL of urine. Since the determination
of TTCA concentration was based on standard curves
using T3CA as internal standard, the values were highly
reproducible. Day to day variation was 2.66% and same
day variation was 0.29%. The sensitivity of determination
was 20 pmol of TTCA in 0.5 mL of urine (or 40 nmol/L)
making it applicable to even low-level exposures (10). A
comparison of the chromatograms points to the estab-
lished relationship between TTCA levels and CS2 expo-
sure levels.
One outcome of the new method that was not antici-
pated during method development was the recognition
of an additional peak (appearing ∼1.6 min after TTCA)
in the samples from workers exposed to CS2 (Figure 2).
This peak was not present in the urine samples obtained
from control subjects. In addition, the area of the peak
appeared to be dependent on the exposure level. To
establish that the new peak originated from CS2, rats
In Vitr o Stu d ies. Cysteine or cys-gly (1 mM) was treated
with either MITC or thiophosgene (2 mM) in phosphate buffer
(pH 8.0) at 37 °C for 18 h and analyzed by HPLC using the
conditions given above.
Resu lts a n d Discu ssion
Assa y Meth od . The first step in most methods of
analysis (Table 1) is extraction of TTCA from urine. When

1280 Chem. Res. Toxicol., Vol. 14, No. 9, 2001
Amarnath et al.
F igu r e 3. Chromatograms of urine samples obtained from a
control (A) and CS2 exposed (B) rat. The HPLC conditions were
the same as specified in Figure 2 except
a lower (2.5%)
concentration of acetonitrile in the eluent was found to give a
better separation of T3CA and TTCG. A peak for TTCA (11.3
min) as well as a peak for TTCG (13.3 min) were present in
urine of rats treated with 5 mmol/kg of CS2 but not controls.
Although two small peaks were observed in control urine
samples eluting at 11.2 and 13.3 min, their UV maxima 280
and 260 nm, respectively, and elution times relative to the
internal standard were different from TTCA (272 nm) and TTCG
(274 nm). The peak at 12.1 (A) and 12.4 (B) min represents
added T3CA.
Oasis sorbent only under acidic conditions. Because GSH
has been implicated in the formation of TTCA (2) and
the new metabolite may have been a precursor of TTCA,
we treated GSH or cysteinylglycine (cys-gly) with CS2
followed by oxidation with lead nitrate, the conditions
that produce TTCA from cysteine. When the products
F igu r e 2. Chromatograms of human urine samples processed
as described in Figure 1 and analyzed by HPLC. Separations
were performed using a Lichrosphere 100RP-8 10 µm 4 × 250
mm column connected to a Whatman Partisil 5 ODS-3 µM 4.6
× 250 mm column, 3% acetonitile, with 1% acetic acid and
monitoring at 278 nm. The range of exposure levels to carbon
disulfide (A) less than 0.03-0.4 ppm (clerical area); (B) 0.2-
2.0 ppm (packaging area); (C) 8.6-43 ppm (spinning area). The
peak at 11.9 min representing added T3CA is present in all the
chromatograms. The level of TTCA (eluting at 10.7 min) is
undetectable in panel A, 19.2 nmol/mg creatinine in B, and 48.6
nmol/mg creatinine in panel C. An additional peak with a
retention time of 12.5 min could be seen as a shoulder in panel
B and as a distinct peak in panel C.
were exposed to CS2, urine samples collected and ana-
lyzed, and the results compared to those obtained for
control rats (Figure 3). Similar to the human samples,
the extra peak appearing after the internal standard
along with the peak for TTCA was observed only in the
urine of exposed rats.
were analyzed by HPLC, only cys-gly subjected to the
above conditions gave a peak with the retention time
similar to that of the new metabolite suggesting a
structure related to both TTCA and cys-gly (2-thioxothi-
azolidin-4-ylcarbonylglycine, TTCG), for the unidentified
metabolite.
To gain further insight into the structure of the
metabolite, an independent synthesis of TTCG was
undertaken. The formation of the dipeptide bond in
The UV spectrum obtained for the late eluting peak
using the diode array detector had a maximum of 274
nm and it was similar to that of TTCA. The presence of
a carboxyl group was suggested by the fact that the
compound was extracted by ethyl acetate or retained by

New Urinary Metabolite of CS2
Chem. Res. Toxicol., Vol. 14, No. 9, 2001 1281
TTCG using the mixed anhydride method (11) starting
with TTCA and glycine gave a mixture of products from
which the isolation of pure TTCG was very difficult.
However, when the ethyl ester of glycine was used, the
resulting ester of TTCG could be easily purified and
subsequently hydrolyzed under basic conditions to gener-
ate TTCG whose structure was verified by NMR and
negative ion mass spectroscopy [m/z 219 (M - 1)]. Its
retention time was also identical to that of the sample
obtained from cys-gly (data not shown).
Confirmation for the structure of the new urinary
metabolite was achieved utilizing LC/MS/MS. Urine from
rats treated with CS2 was pooled, acidified, extracted
with ethyl acetate, and concentrated. The crude mixture
was purified by column chromatography and by HPLC
for analysis by mass spectrometry, but obtaining a
positive molecular ion for TTCG proved difficult. To
circumvent this problem the extract was treated with
BF3-OEt2 in ethanol to obtain the ester of the carboxylic
acid which was purified and found to produce the
identical parent ion and daughter ions to those obtained
for synthetic TTCG ethyl ester (Figure 4).
F igu r e 4. Tandem mass spectra of the [M + 1] parent
(m/z 249) of TTCG ethyl ester synthetic standard (A) and
urine extract (B). Many fragment ions including those at 203
(M-OC2H5) and 175 (203-CO) were observed in both samples.
The spectrum was obtained on a Finnigan TSQ 7000 triple-
quadrupole mass spectrometer.
The potential use of TTCG as a biomarker was ex-
plored by treating rats with a single dose of CS2 (0.04-
5.0 mmol/kg) and collecting urine during the following
24 h period. The amounts of TTCA and TTCG present in
the urine were determined using standard curves estab-
lished with T3CA as the internal standard. When ex-
pressed per mg of creatinine, both TTCA and TTCG
showed linear increases with the dose of CS2 (Figure 5),
although the concentration of the latter was ∼30% of that
of the former. The recovery of TTCG when using the
sample preparation of Figure 1 was 80.8 ( 0.9% and the
sensitivity of determination was 20 pmol. The day to day
variation was 3.7% while the determinations in a single
day were within 2.7%.
Mech a n ism of TTCA F or m a tion by CS2. Identifica-
tion of TTCG in urine raises two questions: (1) is TTCG
a precursor of TTCA and (2) does TTCG provide ad-
ditional exposure information to that derived from TTCA
alone? The proposed mechanism for the generation of
TTCA from GSH and CS2 is presented in the top of
Scheme 1. The first step is the formation of trithiocar-
bonate from GSH and this reaction has been demon-
strated in vitro (12). Next, removal of γ-glutamate
followed by cyclization leads to TTCG and the final step
is hydrolysis of the peptide bond to yield TTCA. To
determine if TTCG could be metabolized to TTCA in vivo,
rats were given a single dose of TTCG and the urine
collected during the next 24 h was analyzed. TTCA was
detected in the urine and from the relative peak areas
F igu r e 5. Dose response of TTCA and TTCG in the urine of
rats administerd CS2 per os. Urinary excretion of both TTCA
(slope ) 17.2, R2 ) 0.998)) and TTCA (slope ) 5.2, R2 ) 0.994)
demonstrated a linear response over the exposure range of 0.04-
5.0 mmol/kg of CS2.
of TTCA and TTCG the extent of conversion was calcu-
lated to be 88%. Although these results suggest that
TTCG is likely to be an immediate precursor to TTCA,

1282 Chem. Res. Toxicol., Vol. 14, No. 9, 2001
Amarnath et al.
Sch em e 1
the pathway leading to TTCG itself still remains to be
established. Both addition of CS2 to GSH followed by
cleavage of γ-glutamate or addition of CS2 to cys-gly could
result in production of TTCG.
In addition to direct exposure to CS2 urinary excretion
of TTCA has been reported following intake of dithiocar-
bamates that release CS2 in vivo, exposure to the non-
CS2-generating compound captan, and consumption of
cruciferous vegetables (13). The possible generation of
TTCG following exposure to captan was explored. If
generation of TTCA from captan occurs exclusively
through addition of a captan derived thiophosgene me-
tabolite to cysteine, no TTCG excretion would be ex-
pected. Thus, the observation of TTCG following admin-
istration of captan in the present study (Figure 6B)
suggests that captan derived thiophosgene can react with
cys-gly to form TTCG or with GSH to form a relatively
stable product that can undergo cyclization with the
removal of glutamic acid. When cysteine and cys-gly were
treated with thiophosgene at pH 8.0, the formation of
TTCA and TTCG, respectively, were observed (data not
shown).
To explore the contribution of isothiocynates derived
from dietary glucosinolate and from industrial sources
to TTCA formation, rats were administered methyl
isothiocyanate and the urinary metabolites analyzed. In
contrast to captan, methyl isothiocyanate generated
TTCA and an unidentified metabolite of MITC but not
TTCG (Figure 6C). The absence of TTCG implies that
cyclization proceeds following derivatization of free cys-
teine but not after derivatization of cys-gly. Consistent
with this interpretation in vitro experiments indicated
that while cysteine readily cyclized to TTCA on treatment
F igu r e 6. Chromatograms obtained after processing urine
samples collected from control (A), captan exposed (0.05 mmol/
Kg per os) (B), or methylisothiocyanate (0.1 mmol/kg (C) rats.
T3CA internal standard eluting at 12 min and an unidentified
peak at 9 min is present in all three samples. The HPLC
conditions were the same as in Figure 3. Whereas TTCA can
be identified at 11 min following exposure to either captan (56.2
( 5.0 nmol/mg of creatinine) or methisothiocyanate (27 ( 10
nmol/mg of creatinine) only captan exposure resulted in mea-
surable amounts of TTCG (4.9 ( 2.3 nmol/mg of ceatinine)
eluting at 12.4 m. A large peak eluting at 10 min in the urine
of methylisothiocyanate exposed rats was observed but not
characterized. Low levels of absorbance were detected in
samples prepared from controls at 11.0 min but exhibited a
different UV maxima (280 nm) than TTCA (272 nm).
with MITC, the formation of TTCG was very slow with
cys-gly under similar conditions. These results suggest
that urinary measurement of both TTCA and TTCG will
be a better indication of exposure to CS2 than the
determination of TTCA alone. The HPLC method of
analysis presented in this paper can accomplish this in
a single chromatogram.

New Urinary Metabolite of CS2
Con clu sion
Chem. Res. Toxicol., Vol. 14, No. 9, 2001 1283
urine of workers exposed to carbon disulfide. Arch. Environ.
Health 36, 289-297.
An improved method for the determination of TTCA
in urine suitable for monitoring exposure to CS2 is
presented. The method is amenable to automation, uses
tetrahydro-2-thioxo-2H-1,3-thiazine-4-carboxylic acid (T3-
CA) as an internal standard, Oasis HLB cartridge for
sample preparation, and HPLC with UV detection for
measurement. The sample preparation provided a high
recovery and resulted in chromatograms with very few
extra peaks thereby improving the accuracy of measure-
ment of TTCA. Use of an internal standard increased the
reproducibility while the diode array detector aided in
verifying the identity of metabolites. A novel urinary
metabolite was identified, 2-thioxothiazolidin-4-ylcarbo-
nylglycine (TTCG), for which the output was found to be
dependent on the exposure level of CS2. TTCG appears
to be a precursor to TTCA and presents a potential
biomarker to be used in conjunction with TTCA when
the presence of TTCA in urine may result from exposure
to sources other than CS2.
(3) J ohnson, D. J ., Graham, D. G., Amarnath, V., Amarnath, K., and
Valentine, W. M. (1996) The measurement of 2-thiothiazolidine-
4-carboxylic acid as an index of the in vivo release of CS2 by
dithiocarbamates. Chem. Res. Toxicol. 9, 910-916.
(4) Simon, P., Nicot, T., and Dieudonne, M. (1994) Dietary habits, a
nonnegligible source of 2-thiothiazolidine-4-carboxylic acid and
possible overestimation of carbon disulfide exposure. Int. Arch.
Occup. Environ. Health 66, 85-90.
(5) Cox, C., Que Hee, S. S., and Tolos, W. P. (1998) Biological
monitoring of workers exposed to carbon disulfide. Am. J . Ind.
Med. 33, 48-54.
(6) Omae, K. (1998) Cross sectional observation of the effects of
carbon disulfide on arteriosclerosis in rayon manufacturing
workers. Occup. Environ. Med. 55, 468-472.
(7) Kopecky, J ., and Smejkal, J . (1984) A simple preparation of Pure
2-thiothiazolidine-4-carboxylic acid (TTCA) as a reference stan-
dard for carbon disulfide exposure tests. Bull. Soc. Chim. Belg.
93, 231-232.
(8) Lee, B. L., Yang, X. F., New, A. L., and Ong, C. N. (1995) Liquid
chromatographic determination of urinary 2-thiothiazolidine-4-
carboxylic acid, a biomarker of carbon disulfide exposure. J .
Chromatogr. B. 668, 265-272.
(9) Cox, C., and Que Hee, S. S. (1996) Synthesis of 2-thiothiazolidine-
4-carboxylic acid and its chromatography in rat and human urine.
J . Chromatgr. B. 679, 85-90.
(10) Drexler, H., Goen, T., and Angerer, J . (1995) Carbon disulfide.
II. Investigation on the uptake of CS2 and the excretion of its
metabolie 2-thiothiazolidine-4-carboxylic acid after occupational
exposure. Int. Arch. Occup. Environ. Health 67, 5-10.
Ack n ow led gm en t. These studies were supported by
NIEHS Grant ES06387 and the Center in Molecular
Toxicology Grant P30 ES00267. We thank Mr. Rob Buck
of Waters Corporation for the initial supply of Oasis HLB
extraction cartridges.
(11) Sheehan, J . C., and Yang, D. J . (1958) A new synthesis of cysteinyl
peptides. J . Am. Chem. Soc. 80, 1158-1164.
(12) Valentine, W. M., Amarnath, V., Graham, D. G., and Anthony,
D. C. (1992) Covalent cross-linking of proteins by carbon disulfide.
Chem. Res. Toxicol. 5, 254-262.
(13) Kivisto, H. (2000) TTCA measurements in biomonitoring of low-
level exposure to carbondisulfide. Arch. Occup. Environ. Health
73, 263-269.
Refer en ces
(1) van Doorn, R., Delbressine, L. P. C., Leijdekkers, C. M., Vertin,
P. G., and Henderson, P. T. (1981) Identification and determina-
tion of 2-thiothiazolidine-4-carboxylic acid in urine of viscose
workers exposed o carbon disulfide. Arch. Toxicol. 47, 51-58.
(2) van Doorn, R., Leijdekkers, C. M., Henderson, P. T., Vanhoorne,
M., and Vertin, P. G. (1981) Determination of thio compounds in
TX010085X