Speciation of Dimethylarsinous Acid and Trimethylarsine Oxide in Urine from Rats Fed with Dimethylarsinic Acid and Dimercaptopropane Sulfonate

Article abstract of DOI:10.1021/ac034868u

Speciation of arsenic in urine from rats treated with dimethylarsinic acid (DMA^V) alone or in combination with dimercaptopropane sulfonate (DMPS) were studied. Methods were developed for the determination of the methylarsenic metabolites, especially trace levels of dimethylarsinous acid (DMA^III) and trimethylarsine oxide (TMAO), in the presence of a large excess of DMA^V. Success was achieved by using improved ion-exchange chromatographic separation combined with hydride generation atomic fluorescence detection. Micromolar concentrations of DMA^III were detected in urine of rats fed with a diet supplemented with either 100 μg/g of DMA^V or a mixture of 100 μg/g of DMA^V and 5600 μg/g of DMPS. No significant difference in the DMA^III concentration was observed between the two groups; however, there was a significant difference in TMAO concentrations. Urine from rats fed with the diet supplemented with DMA^V alone contained 73 ± 30 μM TMAO, whereas urine from rats fed with the diet supplemented with both DMA ^V and DMPS contained only 2.8 ± 1.4 μM TMAO. Solutions containing mixtures of 100 μg/L DMA^V or TMAO and 5600 μg/L DMPS did not show reduction of DMA^V and TMAO. The significant decrease (p < 0.001) of the TMAO concentration in rats administered with both DMA^V and DMPS suggests that DMPS inhibits the biomethylation of arsenic.

Full text of DOI:10.1021/ac034868u

Anal. Chem. 2003, 75, 6463-6468
Speciation of Dimethylarsinous Acid and
Trimethylarsine Oxide in Urine from Rats Fed with
Dimethylarsinic Acid and Dimercaptopropane
Sulfonate
Xiufen Lu,† Lora L. Arnold,‡ Samuel M. Cohen,‡ William R. Cullen,§ and X. Chris Le*,†
Department of Public Health Sciences, University of Alberta, Edmonton, Alberta T6G 2G3, Canada, Department of
Pathology and Microbiology and the Eppley Institute for Cancer Research, University of Nebraska Medical Center,
983135 Nebraska Medical Center, Omaha, Nebraska 68198-3135, and Department of Chemistry,
University of British Columbia, Vancouver, British Columbia, Canada
Scheme 1. Pathway of Arsenic Methylation,
Showing Alternate Steps of Two-Electron
Speciation of arsenic in urine from rats treated with
dimethylarsinic acid (DMAV) alone or in combination with
dimercaptopropane sulfonate (DMPS) were studied. Meth-
ods were developed for the determination of the methy-
larsenic metabolites, especially trace levels of dimethyl-
arsinous acid (DMAIII) and trimethylarsine oxide (TMAO),
in the presence of a large excess of DMAV. Success was
achieved by using improved ion-exchange chromato-
graphic separation combined with hydride generation
atomic fluorescence detection. Micromolar concentrations
of DMAIII were detected in urine of rats fed with a diet
supplemented with either 1 0 0 µg/ g of DMAV or a mixture
of 1 0 0 µg/ g of DMAV and 5 6 0 0 µg/ g of DMP S. No
significant difference in the DMAIII concentration was
observed between the two groups; however, there was a
significant difference in TMAO concentrations. Urine from
rats fed with the diet supplemented with DMAV alone
contained 7 3 ( 3 0 µM TMAO, whereas urine from rats
fed with the diet supplemented with both DMAV and
DMP S contained only 2 .8 ( 1 .4 µM TMAO. Solutions
containing mixtures of 1 0 0 µg/ L DMAV or TMAO and
5 6 0 0 µg/ L DMP S did not show reduction of DMAV and
TMAO. The significant decrease (p < 0 .0 0 1 ) of the TMAO
concentration in rats administered with both DMAV and
DMP S suggests that DMP S inhibits the biomethylation of
arsenic.
Reduction (2e-) and Oxidative Addition of a
+
Methyl Group (CH3
)
methylated to monomethylarsonic acid (MMAV). MMAV is then
III
reduced to monomethylarsonous acid (MMA ) and further
methylated to dimethylarsinic acid (DMAV), Similarly, further
III
steps produce dimethylarsinous acid (DMA ) and trimethylarsine
oxide (TMAO), which can be reduced to trimethylarsine (TMAIII).
The metabolic pathway does not usually proceed beyond the
dimethylarsenic species stage for humans and most animals.3
Biomethylation of arsenic has previously been considered a
detoxification process. However, this notion is changing with the
findings of trivalent methylation metabolites, MMAIII and DMAIII
in human urine,4-10 and the numerous studies showing the higher
toxicity of these metabolites than the inorganic arsenic species.11-20
(4) Aposhian, H. V.; Zheng, B.; Aposhian, M. M.; Le, X. C.; Cebrian, M. E.;
Cullen, W. R.; Zakharyan, R. A.; Ma, M.; Dart, R. C.; Chang, Z.; Andrewes,
P.; Yip, L.; O’Malley, G. F.; Maiorino, R. M.; van Voorhies, W.; Healy, S.
M.; Titcomb, A. Toxicol. Appl. Pharmacol. 2 0 0 0 , 165, 74-83.
(5) Aposhian, H. V.; Gurzau, E. S.; Le, X. C.; Gurzau, A.; Healy, S. M.; Lu, X.;
Ma, M.; Yip, L.; Zakharyan, R. A.; Maiorino, R. M.; Dart, R. C.; Tirus, M.
G.; Gonzalez-Ramirez, D.; Morgan, D. L.; Avram, D.; Aposhian, M. M. Chem.
Res. Toxicol. 2 0 0 0 , 13, 693-697.
Biomethylation of arsenic is the major metabolic pathway for
inorganic arsenic. The full pathway as seen for many fungi involves
first the reduction of a pentavalent arsenic species to a trivalent
arsenic species followed by the addition of a methyl group to the
trivalent arsenic.1-3 Specifically, as shown in Scheme 1, inorganic
arsenate (AsV) is reduced to arsenite (AsIII). AsIII is oxidatively
(6) Sampayo-Reyes, A.; Zakharyan, R. A.; Healy, S. M. Aposhian, H. V. Chem.
Res. Toxicol. 2 0 0 0 , 13, 1181-1186.
(7) Le, X. C.; Ma, M.; Lu, X.; Cullen, W. R.; Aposhian, V.; Zheng, B. Environ.
Health Perspect. 2 0 0 0 , 108, 1015-1018.
* Tel.: (780) 492-6416. Fax: (780) 492-7800. E-mail: xc.le@ualberta.ca.
† University of Alberta.
(8) Le, X. C.; Lu, X.; Ma, M.; Cullen, W. R.; Aposhian, H. V.; Zheng, B. Anal.
Chem. 2 0 0 0 , 72, 5172-5177.
(9) Del Razo, L. M.; Styblo, M.; Cullen, W. R.; Thomas, D. J. Toxicol. Appl.
Pharmacol. 2 0 0 1 , 174, 282-293.
‡ University of Nebraska Medical Center.
§ University of British Columbia.
(1) Challenger, F. Chem. Rev. 1 9 4 5 , 36, 315-361.
(2) Cullen, W. R.; McBride, B. C.; Reglinski, J. J. Inorg. Biochem. 1 9 8 4 , 21,
179-194.
(10) Mandal, B. K.; Ogra, Y.; Suzuki, K. T. Chem. Res. Toxicol. 2 0 0 1 , 14, 371-
378.
(11) Styblo, M.; Serves, S. V.; Cullen, W. R.; Thomas, D. J. Chem. Res. Toxicol.
1 9 9 7 , 10, 27-33.
(3) Cullen, W. R.; Reimer, K. J. Chem. Rev. 1 9 8 9 , 89, 713-764.
10.1021/ac034868u CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/18/2003
Analytical Chemistry, Vol. 75, No. 23, December 1, 2003 6463

It is now believed that the methylation of inorganic arsenic may
not be a detoxification mechanism and could be an activation
process.21-23 As a consequence, much attention has been paid to
studies of the toxic effects of these metabolites. There is an
increasing demand for analytical techniques that are capable of
speciating these arsenic metabolites in human and experimental
animal samples.
Although there are differences between human and animals
and between various animal species,24-28 the majority of experi-
mental animals have been found to excrete arsenic efficiently,
usually as DMAV. There is a much lower concentration of MMAV
in animal urine compared with human urine. Animals such as the
marmoset monkey,24 chimpanzee,25 and guinea pig26 are reported
to be extremely poor methylators of inorganic arsenic. These
animals have very little MMAV excreted in their urine presumably
because of a lack of arsenic methyl transferase enzymes. Rats and
mice, however, are efficient methylators of arsenic.27,28
Urinary excretion is the major pathway for elimination of
arsenic compounds from the human and animal body.24-32 While
inorganic AsIII, AsV, MMAV, and DMAV have been frequently
detected in animal urine samples, the presence of the trivalent
methylarsenic metabolites has only been reported recently.28 Rats
were treated with either 100 µg/ g of DMAV or a mixture of 100
µg/ g of DMAV and 5600 µg/ g of dimercaptopropane sulfonate
(DMPS).28 Urine samples from the rats were collected for detailed
analysis of arsenic speciation. Speciation of trace levels of
intermediate arsenic metabolites in the presence of a large excess
of DMAV presented an analytical challenge. This paper describes
a method to deal with this challenge, with an emphasis on the
speciation of TMAO and DMAIII in rat urine samples. DMAIII is a
key metabolic intermediate in the pathway from DMAV to TMAO.
The results on speciation of these arsenicals suggest an inhibitory
effect of DMPS on the biomethylation of arsenic.
EXPERIMENTAL SECTION
Standards and Reagents. The source of MMAIII was the solid
oxide (CH3AsO) and that of DMAIII the iodide [(CH3)2AsI]. The
precursors were prepared following literature procedures33,34 and
were kept at +4 or -20 °C when not in use. Dilute solutions of
the precursors were prepared fresh in deionized water to form
III
CH3As(OH)2 (MMA ) and (CH3)2AsOH (DMAIII), respectively.
III
III
MMA and DMA solutions were prepared immediately prior
to use. TMAO [(CH3)3AsO] was obtained from Tri Chemical
Laboratory (Yamanashi, Japan). Solutions of other standard
arsenic compounds, AsIII, AsV, DMAV (Aldrich, Milwaukee, WI),
and MMAV (Chem Service, West Chester, PA), were prepared
by appropriate dilutions with deionized water from 1000 mg/ L
stock solutions, as described previously.35-38
DMPS was purchased from Heyltex (Houston, TX). Sodium
dihydrogen phosphate and sodium borohydride were obtained
from Aldrich (Milwaukee, WI). High-performance liquid chroma-
tography (HPLC) grade methanol, sodium hydroxide, hydrochlo-
ric acid, and nitric acid were from Fisher (Pittsburgh, PA). A
sodium borohydride solution (1.3%) in 0.1 M sodium hydroxide
was prepared fresh daily. These reagents were of analytical grade
or better. All solutions were prepared using deionized water (18
MΩ; Millipore, Bedford, MA).
Instrument and Methods. Arsenic speciation was carried out
by using HPLC separation with hydride generation atomic
fluorescence detection. The HPLC system consisted of a Gilson
(Middleton, WI) HPLC pump (model 307), a Rheodyne six-port
sample injector (model 7725i) with a 20 µL sample loop, and an
analytical column. HPLC separation was carried out at room
temperature. Mobile-phase solutions were filtered through a 0.45
µM membrane and sonicated in an ultrasonic bath for 15 min prior
to use.
(12) Lin, S.; Cullen, W. R.; Thomas, D. J. Chem. Res. Toxicol. 1 9 9 9 , 12, 924-
930.
(13) Lin, S.; Del Razo, L. M.; Styblo, M.; Wang, C. Q.; Cullen, W. R.; Thomas, D.
J. Chem. Res. Toxicol. 2 0 0 1 , 14, 305-311.
(14) Thomas, D. J.; Styblo, M.; Lin, S. Toxicol. Appl. Pharmacol. 2 0 0 1 , 176, 127-
144.
(15) Petrick, J. S.; Ayala-Fierro, F.; Cullen, W. R.; Carter, D. E.; Aposhian, H. V.
Toxicol. Appl. Pharmacol. 2 0 0 0 , 163, 203-207.
(16) Petrick, J. S.; Bhumasamudram, J.; Mash, E. A.; Aposhian, H. V. Chem. Res.
Toxicol. 2 0 0 1 , 14, 651-656.
(17) Mass, M. J.; Tennant, A.; Roop, R. C.; Cullen, W. R.; Styblo, M.; Thomas,
D. J.; Kligerman, A. D. Chem. Res. Toxicol. 2 0 0 1 , 14, 355-361.
(18) Vega, L.; Styblo, M.; Patterson, R.; Cullen, W. R.; Wang, C.; Germolec, D.
Toxicol. Appl. Pharmacol. 2 0 0 1 , 172, 225-232.
A polymer-based, strong anion-exchange column (150 × 4.1
mm; PRP-X100, Hamilton, Reno, NV) was used to separate TMAO
from other arsenic species. A mobile phase contained 5 mM
phosphate buffer (pH 8.2) and 5% methanol in deionized water.
The flow rate of the mobile phase was 0.8 mL/ min.
Two silica-based anion-exchange columns (250 × 4.60 mm;
Phenosphere, 5 µm SAX 80R, Phenomenex, Torrance, CA), after
modification with DMPS, were used for separation of DMAIII from
the large excess of DMAV. One of the columns was previously
used, and the other was new. A mobile phase consisted of 20 mM
phosphate and 5% methanol (pH 5.8), and its flow rate was 0.8
mL/ min. The columns were pretreated with DMPS, by 10
repeated injections of a 50 µM DMPS solution as a sample onto
the column, followed by washing with 50 mL of 5% methanol in
deionized water. The treated column provided the desired separa-
(19) Chen, G.-Q.; Zhou, L.; Styblo, M.; Walton, F.; Jing, R.; Weinberg, R.; Chen,
Z.; Waxman, S. Cancer Res. 2 0 0 3 , 63, 1853-1859.
(20) Andrewes, P.; Kitchin, K. T.; Wallace, K. Chem. Res. Toxicol. 2 0 0 3 , 16, 994-
1003.
(21) Kitchin, K. T. Toxicol. Appl. Pharmacol. 2 0 0 1 , 172, 249-261.
(22) Hughes, M. F. Toxicol. Lett. 2 0 0 2 , 133, 1-16.
(23) Styblo, M.; Drobna, Z.; Jaspers, I.; Lin, S.; Thomas, D. J. Environ. Health
Perspect. 2 0 0 2 , 110 (Supplement 5), 767-771.
(24) Zakharyan, R. A.; Wildfang, E.; Aposhian, H. V. Toxicol. Appl. Pharmacol.
1 9 9 6 , 140, 77-84.
(25) Vahter, M.; Couch, R.; Nilsson, R. Toxicol. Appl. Pharmacol. 1 9 9 5 , 133,
262-268.
(26) Healy, S. M.; Zakharyan, R. A.; Aposhian, H. V. Mutat. Res. 1 9 9 7 , 386,
229-239.
(27) Peng, B.; Sharma, R.; Mass, M. J.; Kligerman, A. D. Environ. Mol. Mutagen.
2 0 0 2 , 39, 323-332.
(33) Cullen, W. R.; McBride, B. C.; Manji, H.; Pickett, A. W.; Reglinski, J. Appl.
Organomet. Chem. 1 9 8 9 , 3, 71-78.
(28) Cohen, S. M.; Arnold, L. L.; Uzvolgyi, E.; Cano, M.; St. John, M.; Yamamoto,
S.; Lu, X.; Le, X. C. Chem. Res. Toxicol. 2 0 0 2 , 15, 1150-1157.
(29) Buchet, J. P.; Lauwerys, R.; Roels, H. Int. Arch. Occup. Environ. Health 1981,
48, 111-118.
(30) Vahter, M. Acta Pharmacol. Toxicol. 1 9 8 6 , 59, 31-34.
(31) Le, X. C.; Cullen, W. R.; Reimer, K. J. Clin. Chem. 1 9 9 4 , 40, 617-624.
(32) Vahter, M. Toxicology 2 0 0 2 , 181/ 182, 211-217.
(34) Burrows, G. J.; Turner, E. E. J. Chem. Soc., Trans. 1 9 2 0 , 117, 1373-1383.
(35) Le, X. C.; Ma, M. Anal. Chem. 1 9 9 8 , 70, 1926-1933.
(36) Le, X. C.; Ma, M.; Wong, N. A. Anal. Chem. 1 9 9 6 , 68, 4501-4506.
(37) Feldmann, J.; Lai, V. W.-M.; Cullen, W. R.; Ma, M.; Lu, X.; Le, X. C. Clin.
Chem. 1 9 9 9 , 45, 1988-1997.
(38) Jiang, G.; Gong, Z.; Li, X.-F.; Cullen, W. R.; Le, X. C. Chem. Res. Toxicol.
2 0 0 3 , 16, 873-880.
6464 Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

Table 1. Concentration of DMAIII in Rat Urine Samples Collected on Day 2 and TMAO and DMAV in 24 h Rat Urine
Samples Collected on Day 8 of the Experimenta
treatment group
DMAIII concn (µM)
TMAO concn (µM)
DMAV concn (µM)
control
nd (n ) 8)
1.38 ( 0.88 (n ) 8) (0.6-3.0)
0.93 ( 0.49 (n ) 9) (0.5-2.1)
0.2 ( 0.3 (n ) 10)
73 ( 30 (n ) 10) (34-124)
2.8 ( 1.4 (n ) 10) (1.2-4.9)
0.2 ( 0.1 (n ) 10)
100 µg/ g DMAV
66 ( 3 (n ) 10) (51-77)
100 µg/ g DMAV and 5600 µg/ g DMPS
507 ( 31 (n ) 10) (376-688)
a
nd: not detectable (concentration below the detection limit of 0.027 µM for DMAIII). n: number of rats. The results of DMAIII and TMAO are
mean ( standard deviation from replicate analyses of urine samples from these rats. The values in the parentheses represent the range. The spot
urine samples for the analysis of DMAIII were collected at 7:00-9:00 am on day 2 of the experiment. The 24 h urine samples for the analysis of
TMAO and DMAV were collected on day 8 of the experiment.
tion and was stable for 50 analyses before repeated treatment was
needed.
analysis. The frozen urine samples were thawed at room temper-
ature and were analyzed for TMAO and DMAV concentrations
within 1 month of sample collection (results are shown in Table
1).
A hydride generation atomic fluorescence detector (model
Excalibur 10.003, P.S. Analytical, Kent, U.K.) was used for arsenic
detection. HPLC effluent directly met with a flow of hydrochloric
acid (10%, 10 mL/ min) and sodium borohydride (1.3%, 3 mL/
min). Volatile arsines generated were carried by a continuous flow
of argon (250 mL/ min) to the atomic fluorescence detector. A
computer with Varian Star Workstation software and an analog-
to-digital converter board was used for data acquisition and
processing. For the determination of TMAO, a lower concentration
of hydrochloric acid (5%) was used because a higher sensitivity
was achieved for TMAO at the lower acid concentration.
Rats and Urine Samples. Thirty female F344 rats, 4 weeks
old, were purchased from Charles River Breeding laboratories
(Raleigh, NC). The animals were housed in polycarbonate cages
(5/ cage) on dry corncob bedding in a room with a targeted
temperature of 22 °C, humidity of 50%, and 12 h light/ dark cycle
as described previously.28 They were fed pelleted Certified Rodent
Diet 5002 (PMI Nutrition International, Inc., St. Louis, MO). Food
and water were available ad libitum throughout the study.
Following quarantine, rats were randomized into three groups of
10 each: group 1 was fed basal diet, group 2 was fed the basal
diet supplemented with 100 µg/ g of DMAV, and group 3 was fed
the diet supplemented with 100 µg/ g of DMAV and 5600 µg/ g of
DMPS. The rat diets supplemented with DMAV and DMPS were
obtained from Dyets, Inc. (Bethlehem, PA), that mixed the desired
amounts of DMAV and DMPS into Certified Purina 5002 and then
pelleted the diets.
RESULTS AND DISCUSSION
Determination of DMAIII. We have previously developed a
method for the determination of the six individual arsenic species,
III
III
AsV, AsIII, MMAV, MMA , DMAV, and DMA , in human urine
samples.8 It involved ion-pair HPLC separation of the arsenic
species followed by hydride generation atomic fluorescence
spectrometry (HGAFS) detection of trace amounts of arsenic. The
method has been successfully used for the speciation of arsenic
in urine samples from highly exposed populations in Guizhou
(China),7,8 Romania,5 and Inner Mongolia (China).4 The six target
arsenic species were well resolved, and the elution order was AsIII,
III
III
MMA , DMAV, MMAV, DMA , and AsV.
To investigate the possibility that DMAIII might be present in
the urine of animals, rats were fed with a diet containing 100 µg/ g
of DMAV. The concentrations of DMAV in the urine from these
rats peaked at 5000 µg/ L when the samples were analyzed by
using the HPLC-HGAFS method as described previously.8 This
high concentration of DMAV presents a problem for the determi-
III
nation of DMA at trace levels because the small DMAIII peak
was masked by the tail of the much larger DMAV peak. One
approach to solving this problem would be to modify the
III
chromatographic separation so that DMA (trace amounts) is
eluted before the large excess of DMAV.
We have previously observed that the chromatographic reten-
III
tion behavior of MMA , DMAIII, and DMAV could be affected by
Fresh void urine samples were collected separately from each
micromolar levels of DMPS.39 In addition, DMPS is present in
the urine samples from the rats fed with a diet that contained
DMPS. It is necessary to achieve a consistent chromatographic
retention time for the arsenic species in urine samples regardless
of whether DMPS is present in the sample. Therefore, we decided
III
rat for the quantification of DMA . After the three groups of rats
were given the specified diet for 1 day, urine samples were
collected from each rat at 7:00-9:00 a.m. the following morning.
All urine samples were immediately frozen in liquid nitrogen and
kept frozen on dry ice during transportation. The frozen urine
samples were thawed at room temperature and were analyzed for
DMAIII within 48 h of sample collection (results are shown in Table
1). To confirm the presence of DMAIII in rat urine, repeated fresh
void rat urine samples were collected for analysis on study day
71 and day 175. The small volume of urine samples from each rat
III
to use DMPS to modify the column for the separation of DMA
in the presence of a large excess of DMAV. The pretreatment of
the column with DMPS involves 10 replicate injections (20 µL
each) of 50 µM DMPS. With this DMPS treatment, the silica-
based anion-exchange column enables consistent separation of
arsenic species. The column is stable for approximately 50
analyses before another DMPS treatment is needed. Importantly,
III
(∼200 µL) was only sufficient for the quantification of DMA , and
other arsenic species in these samples were not quantified.
Another set of urine samples were collected on day 8 after
the rats had been administered with the specified diets. The rats
were housed in individual metabolic cages. A 24 h urine sample
was collected from each rat. These urine samples were kept frozen
on dry ice during transportation and stored at -20 °C until
III
the desired elution order (DMA before DMAV) is achieved,
III
allowing for the determination of trace levels of DMA in the
presence of much higher concentrations of DMAV in the same
(39) Gong, Z.; Jiang, G.; Cullen, W. R.; Aposhian, H. V.; Le, X. C. Chem. Res.
Toxicol. 2 0 0 2 , 15, 1318-1323.
Analytical Chemistry, Vol. 75, No. 23, December 1, 2003 6465

Figure 1. HPLC-HGAFS chromatograms of a standard solution
containing six arsenic species in deionized water (a) and arsenic
species in a rat urine sample (b). The urine sample was collected on
day 2 of the experiment and was analyzed for DMAIII within 48 h after
collection. A previously used anion-exchange column (250 × 4.60
mm, Phenosphere, 5 µm, SAX) was modified with DMPS and was
used for separation. The mobile phase contained 20 mM phosphate
and 5% methanol (pH 5.8), and its flow rate was 0.8 mL/min. Peak 2
is DMAIII, and peak 4 is DMAV. Peak 1 represents MMAIII and AsIII.
TMAO, MMAIII, and AsV coelute under peak 3.
Figure 2. Chromatograms showing speciation analyses of DMAIII
in a rat urine sample (dotted line) and the urine sample spiked with
a DMAIII standard (solid line). A new anion-exchange column (Phe-
nosphere, 5 µm, 250 × 4.60 mm) modified with DMPS was used for
separation. The mobile phase contained 20 mM phosphate and 5%
methanol (pH 5.8), and its flow rate was 0.8 mL/min. The rat was fed
with a diet supplemented with 100 µg/L DMAV. The urine sample was
collected from the rat on day 2 of the experiment. The amount of
DMAIII spiked to the urine sample (solid line) is equivalent to a final
concentration of ∼10 µg/L.
urine sample. Furthermore, the retention time of arsenic species
on the DMPS-treated column is not affected by the presence of
DMPS in the urine sample.
DMPS. An objective of the study was to examine how DMPS
affects the metabolism and toxicity of DMAV. DMPS is a chelating
agent that has been used for the treatment of acute arsenic and
heavy-metal poisoning.40,41
Figure 1 shows chromatograms from preliminary analyses of
seven arsenic species in water (Figure 1a) and in a rat urine
sample (Figure 1b). DMAIII is clearly resolved from a large excess
of DMAV. The chromatogram of the rat urine sample (Figure 1b)
was acquired from the direct analysis of the original urine sample
without any sample treatment or dilution. The fact that DMAIII is
eluted before DMAV makes it possible to determine trace levels
of DMAIII in the presence of excess DMAV. The concentration of
DMAV in the experimental rat urine samples was as high as 1000-
fold that of DMAIII. The high DMAV concentration did not show
Figure 3 shows representative chromatograms from the
analyses of urine samples collected from rats on day 2 of the
experiment. Chromatogram a was from the analysis of a urine
sample from a control rat under the basal diet. DMAIII was not
detectable in this group of rats. Chromatogram b was obtained
from a urine sample from the second group of rats that were fed
III
with 100 µg/ g of DMAV in the diet. DMA was detected in this
urine sample and in urine samples from the other seven rats in
this group (data not shown). Chromatogram c was obtained from
a urine sample from the third group of rats. Their diet was
supplemented with 100 µg/ g of DMAV and 5600 µg/ g of DMPS.
DMAIII was also present in urine samples from this group of rats.
The procedure of urine collection and analysis was repeated
three times during study day 1, day 71, and day 175 from rats fed
the control diet, the diet with 100 µg/ g of DMAV, and the diet
with the mixture of 100 µg/ g of DMAV and 5600 µg/ g of DMPS.
DMAIII was present in all urine samples from the DMAV treatment
group and the DMAV plus DMPS treatment group (data not
shown).
III
interference with the determination of DMA . The detection limit
III
of DMA was 2.0 µg/ L (∼0.027 µM).
III
A DMA standard was spiked into urine samples to confirm
the chromatographic peak identity of DMAIII in the urine sample.
Figure 2 shows chromatograms obtained from the analyses of a
urine sample (dotted trace) and the same urine sample spiked
III
with a DMA standard (solid trace). The urine sample was
collected from a rat that was fed with the mixture of 100 µg/ g of
DMAV and 5600 µg/ g of DMPS in the diet. Co-injection of the
III
urine samples with a DMA standard (∼10 µg/ L) showed the
III
same retention time for DMA in the urine and standard,
III
suggesting the presence of DMA in the urine sample.
The concentrations of DMAIII in urine samples from the rats
are summarized in Table 1. The concentrations of DMAIII in the
DMAV and the DMAV plus DMPS treatment groups were at the
micromolar levels. There was no significant difference (p ) 0.3
Having established a method for the speciation of DMAIII in
the presence of excess DMAV, we further demonstrated its
application to studies of the arsenic metabolism of experimental
animals. Thirty rats were randomly divided into three groups (10
in each group). The control rats were fed with a basal diet. The
diet for the second group of rats was supplemented with 100 µg/ g
of DMAV. The third group of rats was fed with the diet
supplemented with both 100 µg/ g of DMAV and 5600 µg/ g of
(40) Aposhian, H. V. Annu. Rev. Pharmacol. Toxicol. 1 9 8 3 , 23, 193-215.
(41) Aposhian, H. V.; Arroyo, A.; Cebrian, M. E.; DelRazo, L. M.; Hrulbut, K.
M.; Dart, R. C.; Gonzalez-Ramirez, D.; Kreppel, H.; Speisky, H.; Smith, A.;
Gonsebatt, M. E.; Ostrosky-Wegman, P.; Aposhian, M. M. J. Pharmacol.
Exp. Ther. 1 9 9 7 , 282, 192-200.
6466 Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

Figure 3. Typical chromatograms showing the analysis of DMAIII
in urine samples from a control rat (a), a rat fed with 100 µg/g DMAV
(b), and a rat fed with 100 µg/g DMAV in combination with 5600 µg/g
DMPS (c). The analysis of samples a and b was carried out using
the same conditions as those shown in Figure 1. The analysis of
sample c was carried out using the same conditions as those shown
in Figure 2.
Figure 4. Typical chromatograms showing the presence of TMAO
in urine samples from three groups of rats (b-d) in comparison with
standard arsenic species in water (a). The control rats were fed a
basal diet (b). The second group of rats were fed the diet supple-
mented with 100 µg/g DMAV (c). The third group of rats were fed the
diet supplemented with 100 µg/g DMAV and 5600 µg/g DMPS (d).
Urine samples were collected for 24 h from the rats on day 8 of the
experiment. An anion-exchange column (PRP-X100, 150 × 4.1 mm)
was used for separation. The mobile phase contained 5 mM
phosphate and 5% methanol (pH 8.2), and its flow rate was 0.8 mL/
min. Peak 1 represents TMAO. Peaks 2, 4, and 5 correspond to AsIII,
MMAV, and AsV, respectively. DMAV, DMAIII, and MMAIII coelute under
peak 3.
III
using Mann Whitney test) in the DMA concentration between
the rats in the DMAV alone group and the DMAV plus DMPS
treatment group. No DMAIII was detected in any of the control
urine samples (below detection limit of 2 µg/ L, or ∼0.027 µM).
To confirm that the DMAIII detected in the rat urine is a
consequence of metabolism by the rats, we have analyzed the
DMAV that was added to the rat diet. There was no DMAIII in the
rat diet. In addition, there was no detectable DMAIII in a solution
containing 100 µg/ L DMAV and 5600 µg/ L DMPS.
exchange column (PRP-X100) with 5 mM phosphate buffer (pH
8.2) and 5% methanol as the mobile phase. Detection of TMAO
at trace levels is achieved by HGAFS. Figure 4 shows the analyses
of arsenic standards (a) and rat urine samples (b-d) from three
treatment groups. Chromatograms reveal the presence of TMAO
in all of the rats tested.
Unlike DMAIII, TMAO is much more stable in the urine
samples. TMAO remained unchanged in the urine samples at -20
°C for 1 month, the storage duration tested in this study. Thus,
the measured concentrations of TMAO represent the actual
concentrations of TMAO in the urine samples.
III
DMA was only found in fresh void urine analyzed within 48
h after collection. As demonstrated previously,42 DMAIII in human
urine is oxidatively unstable. It is readily oxidized to DMAV. Thus,
III
the amounts of DMA detected in these rat urine samples
probably represent a small fraction of DMAIII that was in the fresh
void urine.
Our recent work indicates that DMAIII readily binds to rat
hemoglobin.43 The binding of DMAIII to hemoglobin in red blood
cells may explain that only a small amount of DMAIII is excreted
into the urine.
Concentrations of TMAO and DMAV in 24 h rat urine samples
collected on study day 8 are summarized in Table 1. These are
the major arsenic species found in the rat urine samples. In the
rats fed with the diet supplemented with 100 µg/ g of DMAV, their
urine samples contained 34-124 µM (73 ( 30 µM) TMAO.
However, in the rats fed with the diet supplemented with both
100 µg/ g of DMAV and 5600 µg/ g of DMPS, the concentration of
TMAO in their urine decreased to 1.2-4.9 µM (2.8 ( 1.4 µM).
The concentrations of TMAO in the two treatment groups are
significantly different (p < 0.001).
III
Determination of TMAO. Further methylation of DMA
would result in the formation of TMAO (Scheme 1). To examine
whether TMAO is formed in the rats, we further developed a
method for the determination of TMAO. This method involves
the separation of TMAO from other arsenic species using an anion-
(42) Gong, Z.; Lu, X.; Cullen, W. R.; Le, X. C. J. Anal. At. Spectrom. 2 0 0 1 , 16,
1409-1413.
(43) Jiang, G.; Lu, M.; Le, X. C. Mass spectrometry studies of interactions between
arsenicals and proteins. Presented at the 51st American Society for Mass
Spectrometry Conference, Montreal, Canada, June 8-12, 2003.
The concentration of DMAV in the combined DMAV and DMPS
treatment group is much higher than that in the DMAV treatment
Analytical Chemistry, Vol. 75, No. 23, December 1, 2003 6467

group. This reflects the increased total urinary arsenic excretion
due to the administration of DMPS, a finding consistent with the
observed increases in the total arsenic concentration in human
urine samples following the administration of DMPS.4 The
concentrations of AsV, AsIII, and DMAV in the urine samples from
all three groups of rats are minor (2 orders of magnitude lower
than the concentration of DMAV).
While this and previous studies have shown that rats produce
TMAO,28,44-47 much less is known about the formation of TMAO
in humans. Only two studies have reported the presence of TMAO
in human urine.45,48 Francesconi et al.48 detected traces of TMAO
in urine collected from a volunteer after a single ingestion of 1220
µg of arsenic in the form of a dimethylated arsenosugar. Arsenic
speciation analysis was carried out using ion-exchange chroma-
tography with inductively coupled plasma mass spectrometry
detection. TMAO accounted for only 0.5% of the total arsenic
excreted. Marafante et al.45 studied arsenic species in urine
samples from a volunteer following the ingestion of a high dose
of DMAV (8 mg of arsenic). Arsenic species were determined
using hydride generation followed by cold-trapping gas chroma-
tography with atomic absorption detection. They found small
amounts of TMAO in the urine samples up to 3 days after the
ingestion of DMAV. Le et al.7,8 did not find TMAO in human urine
samples collected from people who were exposed to arsenic in
their drinking water, although their method was able to detect
low levels of TMAO (detection limit of 2 µg/ L). There is no other
report on the determination of TMAO in human samples from
the general population.
To examine whether the in vitro reduction of TMAO by DMPS
could be responsible for the decrease in the TMAO concentration,
an aqueous solution containing 100 µg/ L TMAO and 5600 µg/ L
DMPS was analyzed repeatedly for arsenic speciation. This ratio
of TMAO and DMPS is the same as that used in the rat diet.
Speciation analysis did not show a reduction of TMAO over a
period of 48 h. Likewise, no reduction of DMAV was observed in
a solution containing 100 µg/ L DMAV and 5600 µg/ L DMPS.
Therefore, the dramatic decrease of the TMAO concentration in
the rats that were treated with both DMAV and DMPS suggests
that DMPS may inhibit the formation of TMAO in vivo. It is
possible that DMPS inhibits the pathway of arsenic methylation
from DMAV to TMAO.
Two steps are involved in the methylation of DMAV to TMAO
III
(Scheme 1): the reduction of DMAV to DMA , which may be
While the previous arsenic speciation techniques have assisted
studies of arsenic metabolism, the analytical techniques that we
report here for speciation of DMAIII and TMAO also enable the
study of arsenic methylation in rats. The dosing of rats with DMAV
alone or in combination with DMPS provides an opportunity to
observe DMPS inhibiting the methylation of DMAV to TMAO.
Results of this study along with previous findings on the methy-
lation of MMAV to DMAV suggest that DMPS inhibits arsenic
methylation. Understanding the mechanism of arsenic methylation
and the factors affecting arsenic methylation is important for a
better interpretation of arsenic toxicity and carcinogenicity be-
cause of the dramatic differences in toxicity of various arsenic
methylation products.11-19,49
mediated by a reductase, and the oxidative addition of a methyl
group to DMAV, which may require a methyl transferase. DMPS
could presumably interfere with these enzymatic mediated reac-
tions. Alternatively, DMPS could form complexes with DMAIII,
reducing the availability of DMAIII as a substrate to accept a methyl
group. Both of these actions would result in the decrease of TMAO
formation.
Aposhian et al.4,41 reported that administration of DMPS to
people who were exposed to high levels of arsenic in drinking
water resulted in large decreases in the concentration and
percentage of urinary DMAV. They pointed out that DMPS
inhibited the methylation of MMAV to DMAV. Our results, focusing
on the subsequent step of arsenic methylation, from DMAV to
TMAO, are consistent with these earlier findings.
ACKNOWLEDGMENT
A striking difference between humans and rats in arsenic
methylation is the differences in the end methylation products.
This work was supported by the Natural Sciences and
Engineering Research Council of Canada (NSERC), the Canada
Research Chairs Program, and the Canadian Water Networks
NCE. The authors thank Mr. Gian Jhangri for assistance on
statistical analysis.
(44) Yamauchi, H.; Yamamura, Y. Toxicol. Appl. Pharmacol. 1 9 8 4 , 74, 134-
140.
(45) Marafante, E.; Vahter, V.; Norin, H.; Envall, J.; Sandstom, M.; Christako-
poulos, A.; Ryhage, R. J. Appl. Toxicol. 1 9 8 7 , 7, 111-117.
(46) Yoshida, K.; Chen, H.; Inoue, Y.; Wanibuchi, H.; Fukushima, S.; Kuroda,
K.; Endo, G. Arch. Environ. Contam. Toxicol. 1 9 9 7 , 32, 416-421.
(47) Yoshida, K.; Kuroda, K.; Inoue, Y.; Chen, H.; Date, Y.; Wanibuchi, H.;
Fukushima, S.; Endo, G. Appl. Organomet. Chem. 2 0 0 1 , 15, 539-547.
(48) Francesconi, K. A.; Tanggaar, R.; McKenzie, C. J.; Goessler, W. Clin. Chem.
2 0 0 2 , 48, 92-101.
Received for review July 28, 2003. Accepted September
19, 2003.
(49) Le, X. C.; Lu, X.; Li, X.-F. Anal. Chem., in press.
AC034868U
6468 Analytical Chemistry, Vol. 75, No. 23, December 1, 2003