Chemosphere 75 (2009) 1065–1073

Contents lists available at ScienceDirect

Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere

Rapid determination of arsenic species in freshwater organisms

from the arsenic-rich Hayakawa River in Japan using HPLC-ICP-MS
Shinichi Miyashita a,*, Masahito Shimoya b, Yoshiaki Kamidate c, Takayoshi Kuroiwa d, Osamu Shikino e,
Shoko Fujiwara a, Kevin A. Francesconi f, Toshikazu Kaise a
a
School of Life Sciences, Tokyo University of Pharmacy and Life Sciences, 1432-1 Horinouchi, Hachioji, Tokyo 192-0392, Japan
b
GL Sciences Inc., 6-22-1 Nishishinjuku, Shinjuku, Tokyo 163-1130, Japan
c
Nichirei Food Safety Research Center, 9 Shinminato, Mihama, Chiba 261-8545, Japan
d
National Institute of Advanced Industrial Science and Technology, 1-1-1 Umezono, Tsukuba, Ibaraki 305-0045, Japan
e
PerkinElmer Japan Co., Ltd., 134 Koubemachi, Hodogaya, Yokohama 240-0005, Japan
f
Analytical Chemistry, Institute of Chemistry, Karl-Franzens-University, Universitaetsplatz 1, 8010 Graz, Austria

a r t i c l e i n f o a b s t r a c t

Article history: Speciation analyses of water-soluble arsenicals from freshwater and biological samples collected from
Received 31 July 2008 the Hayakawa River (Kanagawa, Japan), which contains a high concentration of arsenic, were performed
Received in revised form 9 January 2009 using high performance liquid chromatography/inductively coupled plasma mass spectrometry (HPLC-
Accepted 9 January 2009
ICP-MS). River water contained only arsenate, which is a pentavalent inorganic arsenical. The water
Available online 8 February 2009
bug Stenopsyche marmorata contained inorganic arsenicals accounting for 77% of the water-soluble arsen-
icals, followed by oxo-arsenosugar-glycerol, which is a type of dimethylarsinoylriboside (arsenosugar).
Keywords:
The freshwater green macroalga Cladophora glomerata contained oxo-arsenosugar-glycerol and oxo-
Arsenosugar
Arsenobetaine
arsenosugar-phosphate as 64% of the water-soluble arsenicals. Production of the same types of arsenosu-
Speciation analysis gars was confirmed in the freshwater green microalga Chlamydomonas reinhardtii CC125 experimentally
Freshwater ecosystem exposed to arsenate. The muscle tissues of all freshwater fish and crustaceans analyzed contained arsen-
Biotransformation obetaine, oxo-arsenosugar-glycerol, and/or oxo-arsenosugar-phosphate in various concentrations. Seven
Chlamydomonas reinhardtii freshwater fish (Cobitis biwae, Leuciscus hakonensis, Phoxinus lagowski steindachneri, Plecoglossus altivelis,
Rhinogobius sp. CB, Rhinogobius sp. CO, Sicyopterus japonicus) and the crustacean Macrobracbium nippone-
nese contained arsenobetaine in their muscle tissues as the predominant form, contributing up to 80% of
the water-soluble arsenicals, while the freshwater fish Anguilla japonica muscle tissues primarily con-
tained dimethylarsinic acid as 77% of the water-soluble arsenicals, followed by arsenobetaine. The fresh-
water fish Zacco platypus muscle tissues predominantly contained oxo-arsenosugar-phosphate,
accounting for 51% of the water-soluble arsenicals, followed by dimethylarsinic acid and arsenobetaine.
These biological samples possessed non-extractable arsenical(s) accounting for more than 50% of the
total arsenic concentration.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction
Arsenic is ubiquitous in the biosphere and undergoes uptake
and bioaccumulation through food chains, followed by alkylation
to form a variety of arsenicals in organisms. Arsenic circulation
in the marine environment has been extensively studied. Marine
organisms contain much higher concentrations of arsenic (<10–
100 mg kg1 dry weight [DW]) than seawater (1.0–1.8 lg L1),
and inorganic arsenicals are biotransformed into organoarsenicals
through marine food chains (Francesconi and Edmonds, 1997).
Arsenosugars are its predominant forms in marine algae (McShee-
hy et al., 2002), while arsenobetaine is most abundant in marine
* Corresponding author. Tel.: +81 42 676 1633; fax: +81 42 676 5084.
E-mail address: s037104@ls.toyaku.ac.jp (S. Miyashita).
fish and crustaceans (Peshut et al., 2008). In marine bivalves and
sponges, arsenosugars and arsenobetaine have been identified as
the primary arsenicals (Shibata and Morita, 1992; Yamaoka et al.,
2001).
In contrast to the large number of reports on arsenicals in mar-
ine organisms, there are fewer data on arsenic speciation in fresh-
water organisms. Certain river waters in Japan contain much
higher arsenic concentrations than seawater (0.25–7.7 lg L1), be-
cause hot-spring waters containing large amounts of arsenic (10–
5000 lg L1) mix with river water in some areas (Kanamori and
Sugawara, 1965). Mt. Hakone is famous as a hot-spring area in
Japan, and Owakudani Valley is one source of the hot springs.
The hot-spring waters containing high concentrations of arsenic
from the Owakudani Valley flow into the Hayakawa River. Kaise
et al. (1997) found that the major arsenic species in river water,

0045-6535/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2009.01.029
1066 S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073
freshwater algae, and freshwater fish from the Hayakawa River
were inorganic arsenicals, dimethylated arsenicals, and trimethy-
lated arsenicals, respectively. Shiomi et al. (1995) and Slejkovec
et al. (2004) reported that arsenobetaine was a major arsenical in
freshwater fish, while Zheng and Hintelmann (2004) found that
arsenobetaine was present only in trace amounts in freshwater
fish, and Lawrence et al. (1986) did not detect arsenobetaine at
all in freshwater fish. Koch et al. (2001) and Soeroes et al. (2005)
reported that arsenosugars predominated in some freshwater fish,
and that arsenobetaine was present as only a minor arsenical.
Schaeffer et al. (2006) found that the dominant arsenic species in
freshwater algae, mussels, and fish from a river were oxo-arse-
nosugars, and that the freshwater mussels and fish contained only
traces of arsenobetaine. Thus, there is little consensus on the major
arsenic species present in freshwater organisms, and it is impor-
tant to understand arsenic behavior in freshwater ecosystems from
the standpoint of environmental assessment.
In the present study, freshwater and biological samples were
collected from the Hayakawa River, which contains a high level
of arsenic from the hot springs at Mt. Hakone. The chemical forms
of water-soluble arsenicals were analyzed by high performance li-
quid chromatography/inductively coupled plasma mass spectrom-
etry (HPLC-ICP-MS). Experimentally, the freshwater green
microalga Chlamydomonas reinhardtii CC125 (C. reinhardtii CC125)
was also exposed to arsenate and arsenic metabolites in the cells
were analyzed to confirm biotransformation of arsenic. The results
suggest that the hot-spring water and river water contained only
arsenate, and also that the water bug collected from the river con-
tained arsenate as the predominant water-soluble arsenical. On the
other hand, the freshwater green algae, crustaceans, and fish con-
tained oxo-arsenosugar-glycerol and/or oxo-arsenosugar-phos-
phate at variety of concentrations and, moreover, the crustaceans
and fish contained arsenobetaine as a primary water-soluble
arsenical.
2. Experimental
2.1. Materials
Water was prepared with a Milli-Q water system
(18.2 MX cm1). Methanol (HPLC grade), hydrogen peroxide (spe-
cial grade), and nitric acid (ultrapure grade) were purchased from
Kanto Chemical Co., Inc. (Tokyo, Japan). Sodium 1-butane sulfonate
and tetramethylammonium hydroxide pentahydrate were ob-
tained from the Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan).
O O
H3C As O OH H3C As
O O
CH3 OH CH3
Oxo-arsenosugar-glycerol
OH OH
O O
H3C As O SO3- H3C As
O O
CH3 OH CH3
Oxo-arsenosugar-sulfonate
OH OH
Fig. 1. Structures of oxo-arsenosugars relevant to this study.
Malonic acid (analytical grade) was purchased from Wako Pure
Chemical Industries, Ltd. (Osaka, Japan). Sodium arsenate
(Na2HAsO4), sodium arsenite (NaAsO2), methylarsonic acid
(CH3AsO(OH)2), dimethylarsinic acid ((CH3)2AsO(OH)), arsenobe-
taine ((CH3)3As+CH2COO), trimethylarsine oxide ((CH3)3AsO),
tetramethylarsonium iodide ((CH3)4As+I), and arsenocholine bro-
mide ((CH3)3As+CH2CH2OHBr) were obtained from Tri Chemical
Laboratories Inc. (Yamanashi, Japan). Oxo-arsenosugar-glycerol
was synthesized following the procedures described in McAdam
et al. (1987). Brown macroalga Fucus extract containing four types
of oxo-arsenosugars was prepared as described by Madsen et al.
(2000). The structures of these oxo-arsenosugars are shown
in Fig. 1. Certified reference materials (CRMs) such as NMIJ 7402-
a cod-fish tissue and NIES No.9 Sargasso were utilized to evaluate
analytical methods. Freshwater green microalga C. reinhardtii
CC125 (wild-type mt+) strain was provided by Dr. Mikio Tsuzuki
(Tokyo University of Pharmacy and Life Sciences, Tokyo, Japan).
2.2. Collection of water and biological samples from the Hayakawa
River
Sample collection was carried out in the Hayakawa River
(Kanagawa, Japan) in May and June, 2005. The river water contains
a much higher arsenic concentration than the environmental water
quality standard in Japan (10 lg L1) (Kaise et al., 1998). The col-
lected samples were: hot-spring water, river water, a water bug
(Stenopsyche marmorata), a green macroalga (Cladophora glomera-
ta), an omnivorous crustacean (Macrobracbium nipponenese), a her-
bivorous fish (Plecoglossus altivelis), and eight omnivorous fish
(Anguilla japonica, Cobitis biwae, Leuciscus hakonensis, Phoxinus
lagowski steindachneri, Rhinogobius sp. CB, Rhinogobius sp. CO, Sicy-
opterus japonicus, Zacco platypus).
Water samples were cryopreserved at 84 °C. Biological sam-
ples were rinsed with water and green macroalga and water bug
were freeze-dried. The muscle tissues of fish and crustaceans were
isolated by scalpel from the skin, viscera, bones, and carapaces and
freeze-dried. Freeze-dried samples were homogenized by pestle
and kept in a powdered state until extraction.
2.3. Experimental exposure of freshwater algal cells to arsenate
The green microalga C. reinhardtii CC125 was cultured mixotro-
phically in tris-acetate-phosphate (TAP) medium. Cultures in flasks
were agitated on a gyratory shaker (120 rpm) at 27 °C with a 12 h
light and 12 h dark cycle at 35 lmol photons m2 s1 of fluorescent
O O-
P
O O O OH
OH OH
Oxo-arsenosugar-phosphate
OH OH
O OSO3-
OH
Oxo-arsenosugar-sulfate
OH OH
 S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073 1067

light. The cells were cultivated until the stationary phase, and the Table 1
culture medium was replaced with fresh medium. After the addi- Analytical conditions for HPLC-ICP-MS: (a) ICP-MS conditions applied for specific
determination of arsenic, (b) HPLC conditions used for separation of water-soluble
tion of arsenate to the cell suspension at a final concentration of arsenicals.
200 lg L1, the cells were cultured for 3 d. The exposure concen-
tration was selected based on the range of total arsenic concentra- Parameter Setting

tions in river water from the Hayakawa River (6.2–98 lg L1) (a)
(Kaise et al., 1998). The cells were collected in 15-mL polypropyl- RF power 1550 W
Nebulizer gas flow 1.04 L min1
ene centrifuge tubes, and the tubes were centrifuged at 1600 g Auxiliary gas flow 1.28 L min1
for 5 min. The supernatants were removed, and the cells were Plasma gas flow 17.8 L min1
rinsed with 10 mL arsenic-free medium three times and finally Monitored ion m/z 91 (AsO+)
rinsed with 10 mL water. After rinsing, the cells were transferred Reaction gas O2 (purity: 99.999%)
Cell gas flow 0.8 mL min1
into 1.5-mL polypropylene tubes and frozen for 1 h at 20 °C.
RPq 0.5
The tubes were sonicated for 10 min at 4 °C using a sonication Axial field voltage 325–350 V
bath, and the cells were freeze-dried and homogenized by pestle. (b)
Column Shiseido CAPCELL PAK C18 MGII
2.4. Sample preparation for total arsenic analysis by ICP-MS 4.6 mm id  250 mm, 5 lm
Guard column Shiseido CAPCELL PAK C18 MGII
4.0 mm id  10 mm, 5 lm
Nitric acid (1.0 mL) was added to weighed powdered samples in Column temperature 40 °C
15-mL centrifuge tubes and allowed to stand for 10 min. The sus- Mobile phase 10 mol m3 sodium 1-butansulfonate
pensions were heated at 110 °C for 30 min in a heating block 4 mol m3 tetramethylammonium hydroxide
4 mol m3 malonic acid
(HF-200; Yamato Scientific Co., Ltd., Tokyo, Japan). Hydrogen per-
0.5% methanol
oxide (500 lL) was added to the tubes at room temperature, and pH 3.0 adjusted with HNO3
the suspensions were heated again at 110 °C for 30 min. Solubi- Injection volume 10 lL
lized solutions were diluted with water to 50 mL in DigiTUBEs Flow rate 1.0 mL min1
(SCP SCIENCE, Baie D’Urfé, Quebec, Canada) and filtered using
0.45-lm membrane filters (GL Sciences Inc., Tokyo, Japan).
Table 2
2.5. Sample preparation for arsenic speciation analysis by HPLC-ICP- Analytical results of arsenic species for the CRMs.
MS CRM Certified value Measured value
(lg g1) (lg g1)
Hot-spring water and river water were filtered using 0.45-lm Total As AsB Total As AsB
membrane filters. Water-soluble arsenicals in freeze-dried biolog-
NMIJ CRM 7402-a cod-fish tissue 36.7 ± 1.8 33.1 ± 1.5 39.5 ± 1.3 30.3 ± 1.7
ical samples were extracted as follows. Freeze-dried samples (10– NIES CRM No. 9 Sargasso 115 ± 9.0 128 ± 4.4
20 mg) were weighed in 1.5-mL tubes, and methanol/water (1:1, v/
Values are means ± standard deviations of triplicates.
v) solutions (400 lL) were added. After standing for 10 min, the
suspensions were sonicated for 10 min and centrifuged at 5200 g
for 2 min, and the initial extracts were transferred to new tubes.
Methanol/water (1:1, v/v) solutions (200 lL) were added to the arsenobetaine (AsB), trimethylarsine oxide (TMAO), tetramethy-
residues, and the water-soluble arsenicals were reextracted two larsonium (TMA), and arsenocholine (AsC) were included in quali-
more times. Combined extracts for each sample were dried under tative and quantitative analyses by HPLC-ICP-MS. Fucus extract
vacuum. The dried extracts were dissolved in water, followed by containing oxo-arsenosugar-glycerol, oxo-arsenosugar-phosphate,
filtration using 0.45-lm membrane filters. oxo-arsenosugar-sulfonate, and oxo-arsenosugar–sulfate was used
for arsenic qualitative analyses by HPLC-ICP-MS.
2.6. Total and speciation analyses of arsenic using ICP-MS and HPLC- These analytical methods were validated through analyses of
ICP-MS NMIJ CRM 7402-a cod-fish tissue and NIES CRM No.9 Sargasso,
and the measured values of the total arsenic and AsB concentra-
ICP-MS (ELAN DRC-e; PerkinElmer SCIEX Inc., Ontario, Canada) tions acceptably agreed with the certified values for the CRMs.
was used to specifically detect arsenic. High levels of chloride The certified and measured values are shown in Table 2.
can interfere with arsenic due to the formation of argon chloride
ion (ArCl+) in plasma, which has the same mass-to-charge ratio 3. Results and discussion
as arsenic (m/z 75). Therefore, a dynamic reaction cell (DRC) tech-
nique was applied to oxidize arsenic under an oxygen atmosphere 3.1. Total arsenic analyses of water and biological samples from the
and generate arsenic oxide (AsO+, m/z 91). The calibration curve for Hayakawa River
total arsenic analysis was based on arsenate. A signal intensity of
10 times the standard deviation of the blank sample was used to Total arsenic concentrations in water and biological samples
estimate the minimum limits of quantitation (LOQs). Analytical from the Hayakawa River are shown in Table 3. The arsenic con-
conditions for the ICP-MS analyses are shown in Table 1a. centration in the hot-spring water (750 lg L1) that flows into
An HPLC connected to an ICP-MS was used to separate and de- the Hayakawa River was about 45 times higher than the river
tect arsenicals for speciation analysis. The HPLC system was com- water (17 lg L1). All of the biological samples contained much
posed of a carrier reservoir (CR670), a metal free pump (PU611; GL higher arsenic concentrations than did river water, although there
Sciences, Inc., Tokyo, Japan), an autosampler with a built-in column was a considerable range of concentrations among the various
oven (MIDAS; Spark Holland, Emmen, Netherlands), and an ODS organisms. In particular, the water bug and green macroalga sam-
column (Shiseido CAPCELL PAK C18 MGII; Shiseido Co., Ltd., Tokyo, ples contained high concentrations of total arsenic (18 000 and
Japan). Analytical conditions for the HPLC analyses are shown 18 000 lg kg1 DW, respectively), followed by crustacean and fish
in Table 1b. The standard arsenic species arsenate (iAsV), arsenite muscle tissues (2600 ± 90 and 150–2100 lg kg1 DW, respec-
(iAsIII), methylarsonic acid (MAA), dimethylarsinic acid (DMAA), tively). Total arsenic concentrations in the same types of organisms
 1068
Table 3
Analytical results for arsenic species in water and biological samples from the Hayakawa River.

Sample Concentration (lg L1 or lg kg1 DW) Extraction

yield (%)
Total As Sum of iAsV iAsIII MAA DMAA AsB TMAO TMA AsC Oxo-arsenosugar Oxo-arsenosugar Unknown a

species -glycerol -phosphate

Water
Hot-spring water b 750 750 ND ND ND ND ND ND ND ND ND ND
River water b 17 17 ND ND ND ND ND ND ND ND ND ND
Water bug Stenopsyche 18 000 6300 4600 190 81 310 20 ND ND 32 500 ND 530 35
marmorata b,c (100) (74) (3) (1) (5) (<1) (1) (8) (8)
Green macroalga Cladophora 18 000 2800 620 ND 53 350 ND ND ND ND 1700 170 <0.43 16
glomerata b (100) (22) (2) (12) (58) (6) (<1)

S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073

Crustacean
Omnivorous Macrobracbium 2600 ± 90 460 ± 170 16 ± 23 5.7 ± 8.0 0.13 ± 0.18 28 ± 0.21 280 ± 76 3.8 ± 5.4 22 ± 17 6.5 ± 9.1 39 ± 6.5 44 ± 16 14 ± 8.2 17
nipponenese (100) (4) (1) (<1) (6) (61) (1) (5) (1) (9) (10) (3)
Fish
Herbivorous
Plecoglossus altivelis 860 ± 410 240 ± 240 7.5 ± 10 ND ND 34 ± 37 120 ± 88 64 ± 90 ND ND ND 8.1 ± 11 ND 28
(100) (3) (14) (52) (27) (3)
Omnivorous
Anguilla japonica 150 ± 15 37 ± 8.8 0.25 ± 0.00 ND 0.25 ± 0.00 29 ± 5.7 7.8 ± 2.5 ND ND ND 0.16 ± 0.28 0.16 ± 0.28 ND 25
(100) (1) (1) (77) (21) (<1) (<1)
Cobitis biwaec 660 200 <0.25 ND 23 <0.25 140 ND ND ND ND 36 ND 31
(100) (<1) (11) (<1) (71) (18)
Leuciscus hakonensis 300 ± 26 64 ± 36 6.2 ± 8.4 ND ND 13 ± 7.1 19 ± 10 11 ± 3.0 ND ND 15 ± 7.4 ND ND 21
(100) (10) (20) (29) (17) (24)
Phoxinus lagowski 310 ± 140 130 ± 120 15 ± 29 ND 0.13 ± 0.14 13 ± 19 54 ± 12 22 ± 27 23 ± 20 ND 2.0 ± 3.6 4.1 ± 7.9 ND 42
steindachneri (100) (11) (<1) (10) (41) (16) (18) (2) (3)
Rhinogobius sp. CB 720 ± 120 150 ± 67 6.1 ± 6.2 ND 0.15 ± 0.14 2.6 ± 3.3 100 ± 26 1.1 ± 2.2 ND 4.9 ± 4.8 8.5 ± 14 24 ± 10 ND 21
(100) (4) (<1) (2) (69) (1) (3) (6) (16)
Rhinogobius sp. CO 700 ± 120 210 ± 130 1.3 ± 2.4 ND 0.10 ± 0.14 1.1 ± 2.0 170 ± 84 0.17 ± 0.24 ND 5.6 ± 7.4 3.0 ± 4.1 28 ± 20 2.0 ± 4.4 30
(100) (1) (<1) (1) (80) (<1) (3) (1) (13) (1)
Sicyopterus japonicus 2100 ± 490 440 ± 120 5.1 ± 4.6 ND ND 44 ± 26 290 ± 53 18 ± 17 ND 0.086 ± 0.19 ND 78 ± 23 ND 20
(100) (1) (10) (67) (4) (<1) (18)
Zacco platypus 1000 ± 500 180 ± 90 0.25 ± 0.00 ND ND 35 ± 26 31 ± 14 13 ± 19 ND ND 7.3 ± 10 89 ± 21 ND 17
(100) (<1) (20) (18) (8) (4) (51)

Values are means ± standard deviations of the analytical results for at least two samples per organism, and the values in parentheses are the percentages of each arsenic species of the sum of species.
LOQs: 0.25 lg L1 for iAsV, iAsIII, MAA, DMAA, and AsB; 0.43 lg L1 for TMAO, AsC, and TMA; 0.49 lg L1 for oxo-arsenosugar-glycerol and oxo-arsenosugar-phosphate.
ND: not detected.
a
Arsenic concentrations were estimated by comparison to the peak areas of the nearest-neighbor standard arsenic species.
b
Arsenic concentrations were determined only once.
c
Only one sample was collected from the Hayakawa River.
 S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073
decreased with each step up in both experimental food chain mod-
els, one of which used a freshwater green microalga (Chlorella vul-
garis), a herbivorous shrimp (Neocaridina denticulata), and a
carnivorous fish (Oryzias latipes), and the other using the same
kinds of alga and shrimp and a carnivorous killifish (Tilapia mos-
sambica) (Kuroiwa et al., 1994; Suhendrayatna et al., 2002). These
results indicate that total arsenic concentrations in freshwater
organisms may decrease by an order of magnitude with each
higher trophic level.
3.2. Arsenic speciation analyses and extraction yields for water and
biological samples from the Hayakawa River
Speciation analyses of water-soluble arsenicals were carried out
using HPLC-ICP-MS. Eight standard arsenic species were rapidly
separated into each peak within 10 min without interference with
ArCl+, and also four oxo-arsenosugars in the Fucus extract were
separated under the same analytical conditions. The chromato-
grams for the standard arsenic species and the Fucus extract are
shown in Fig. 2, although oxo-arsenosugar-glycerol is not visible
in Fig. 2b because of dilution.
Analytical results for arsenic species in water and biological
samples from the Hayakawa River are summarized in Table 3,
and representative chromatograms are shown in Fig. 3. Oxo-
arsenosugar-sulfonate and oxo-arsenosugar-sulfate were not de-
tected in any sample. Extraction yields for water-soluble arsenicals
in marine organisms are usually above 70% and often above 95%,
while those for the freshwater organisms analyzed in this study
were quite low (16–42%) (Schaeffer et al., 2005). Notably, the green
macroalga C. glomerata had the lowest extraction yield. In green al-
gae from hot springs, more than 50% of the arsenic was not ex-
tracted using methanol/water (1:1, v/v) solution (Koch et al.,
1999). Furthermore, the water extraction yields for algae, plants,
a 30000
DMAA
iAsIII MAA
Pulse Intensity
AsB
20000
iAsV
10000
0
0 100 200
30000
b
Oxo-arsenosugar-sulfonate
Pulse Intensity
20000
Oxo-arsenosugar-phosphate
Oxo-arsenosugar-sulfate
10000
0
0 100 200
Fig. 2. HPLC-ICP-MS chromatograms of (a) standard arsenic species and (b) Fucus extract containing oxo-arsenosugars.
1069
and animals from a river were below 50% (mean 36%) (Schaeffer
et al., 2006). These results imply that substantial arsenical(s) not
extractable by aqueous solution may exist in water bugs, green al-
gae, and muscle tissues of crustaceans and fish from freshwater
environments.
3.2.1. Hot-spring water and river water
Speciation analyses of hot-spring water and river water samples
showed only the presence of arsenate (Fig. 3a), which is expected
to exist in most waters (Cullen and Reimer, 1989). Koch et al.
(1999) reported only the presence of arsenate in the water from
hot springs. However, arsenite made up as much as 80% of the total
arsenic in a hot creek (Wilkie and Hering, 1998). These results im-
ply that if arsenite initially dissolves from minerals, oxidation to
arsenate may take place after outflow of the water from the source.
3.2.2. Water bug
The water bug S. marmorata contained 18 000 lg kg1 DW total
arsenic, and most of the extractable arsenic was present as arse-
nate at the high concentration of 4600 lg kg1 DW (Fig. 3b). In
addition, the small amounts of several arsenicals such as arsenite,
MAA, DMAA, AsB, AsC, oxo-arsenosugar-glycerol, and unknown
arsenicals were also detected. These results imply that the water
bug may take up arsenate in the river water or prey on plankton
or algae containing these arsenicals, and also retain incorporated
arsenate rather than actively metabolizing it.
3.2.3. Freshwater green macroalga
The green macroalga C. glomerata contained 18 000 lg kg1 DW
total arsenic, and the dominant water-soluble arsenical was oxo-
arsenosugar-glycerol at a concentration of 1700 lg kg1 DW
(Fig. 3c). Additionally, oxo-arsenosugar-glycerol and oxo-arseno-
sugar-phosphate accounted for 64% of the water-soluble arsenicals.
Standard arsenic species
10 µg L-1
TMAO
TMA
AsC
Oxo-arsenosugar-glycerol
300 400 500 600 700
Time (sec)
Fucus extract
300 400 500 600 700
Time (sec)
1070 S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073

a 20000

Pulse Intensity
10000 iAsV

0
0 100 200 300 400 500 600 700
Time (sec)

b 70000
iAsV
60000
Pulse Intensity

50000

40000

30000

20000
Oxo-arsenosugar-glycerol
10000
0
0 100 200 300 400 500 600 700
Time (sec)

c 20000
Oxo-arsenosugar-glycerol
Pulse Intensity

iAsV
10000
Oxo-arsenosugar-phosphate
DMAA

0
0 100 200 300 400 500 600 700
Time (sec)

20000
d
Pulse Intensity

AsB

10000

Oxo-arsenosugar-phosphate
Oxo-arsenosugar-glycerol
iAsV

0
0 100 200 300 400 500 600 700

Time (sec)

Fig. 3. HPLC-ICP-MS chromatograms of (a) river water, (b) the water bug Stenopsyche marmorata, (c) the green macroalga Cladophora glomerata, (d) the omnivorous
crustacean Macrobracbium nipponenese muscle tissue, (e) the herbivorous fish Plecoglossus altivelis muscle tissue, and (f) the omnivorous fish Sicyopterus japonicus muscle
tissue from the Hayakawa River; and (g) the freshwater green microalga C. reinhardtii CC125 exposed to 200 lg L1 of arsenate for 3 d.
 S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073 1071

2000
e
DMAA AsB

Pulse Intensity 1000

Oxo-arsenosugar-phosphate

TMAO

0
0 100 200 300 400 500 600 700
Time (sec)

f 8000
7000 AsB
Pulse Intensity

6000
5000
4000
3000 DMAA Oxo-arsenosugar-phosphate

2000 TMAO
V
1000 iAs
0
0 100 200 300 400 500 600 700
Time (sec)

g 40000
Oxo-arsenosugar-glycerol
DMAA
30000
Pulse Intensity

20000 Oxo-arsenosugar-phosphate

10000
iAsV
0
0 100 200 300 400 500 600 700

Time (sec)

Fig. 3 (continued)

Koch et al. (1999) reported that small amounts of arsenosugars
were detected in green algae from hot springs, whose waters pri-
marily contained arsenate and arsenite, implying that green algae
are capable of producing arsenosugars from arsenate or arsenite.
The production of arsenosugars after experimental exposure to
arsenate was confirmed in the marine brown alga Fucus gardneri
and the freshwater green microalga C. vulgaris (Granchinho et al.,
2001; Murray et al., 2003). These results indicate that these primary
producers may take up arsenate and transform arsenate into
arsenosugars.
3.2.4. Freshwater crustaceans
The omnivorous crustacean M. nipponenese muscle tissues con-
tained 2600 ± 90 lg kg1 DW total arsenic, and AsB was present as
the primary extractable arsenical at a concentration of 280 ± 76 lg
kg1 DW (Fig. 3d). Intriguingly, almost all of the standard arsenic
species were also detected as minor arsenicals, including arsenate,
arsenite, MAA, DMAA, TMAO, TMA, AsC, oxo-arsenosugar-glycerol,
and oxo-arsenosugar-phosphate. Devesa et al. (2002) reported that
the freshwater crayfish Procambarus clarkii from a polluted river
primarily contained inorganic arsenicals and lower amounts of
AsB. The freshwater herbivorous crustacean Moina macrocopa took
up more arsenic from arsenic-containing food than from arsenic-
containing water (Suhendrayatna et al., 2002). Furthermore, the
freshwater herbivorous shrimp Neocaridina denticulate, which
preyed on arsenate-exposed freshwater green microalga C. vulgaris,
contained small amounts of trimethylated arsenicals (Kuroiwa
et al., 1994). These data imply that the various arsenicals detected
may result from predation on plankton, algae, or small fish con-
taining these arsenicals. The crustaceans may also internally pro-
duce AsB from precursor arsenical(s), but these pathways have
not been demonstrated.
1072 S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073
3.2.5. Freshwater fish
The herbivorous fish P. altivelis muscle tissues contained
860 ± 410 lg kg1 DW total arsenic, and AsB was the main arseni-
cal detected at a concentration of 120 ± 88 lg kg1 DW (Fig. 3e).
Small amounts of arsenate, DMAA, and oxo-arsenosugar-phos-
phate were also detected. Muscle tissues from eight omnivorous
fish contained a wide range of total arsenic concentrations
(150 ± 15 to 2100 ± 490 lg kg1 DW), and the predominating
water-soluble arsenical was also AsB with the exception of two
species, ranging from 19 ± 10 to 290 ± 53 lg kg1 DW (Fig. 3f).
Interestingly, the omnivorous fish A. japonica muscle tissues pri-
marily contained DMAA at a concentration of 29 ± 5.7 lg kg1
DW, and another omnivorous fish Z. platypus muscle tissues pre-
dominantly contained oxo-arsenosugar-phosphate at a concentra-
tion of 89 ± 21 lg kg1 DW. AsB was detected in all fish samples
and made up relatively large percentages of the water-soluble
arsenicals (18–80%). Furthermore, arsenosugars were also detected
in all fish samples at various concentrations contributing up to 55%
of the water-soluble arsenicals.
AsB is the most abundant and stable organoarsenical in marine
animals. The Japanese freshwater smelt Hypomesus nipponensis
also contained AsB as a major arsenical (Shiomi et al., 1995). In
addition, the freshwater salmonids Salmo marmoratus, Oncorhyn-
chus mykiss, and Salmo trutta m. fario from several rivers predomi-
nantly contained AsB (Slejkovec et al., 2004). On the other hand,
AsB was found at minor concentrations in freshwater fish from
the arsenic-rich lake and moreover was not detected in the fresh-
water fish families Esocidae, Cyprinidae, and Percidae (Zheng and
Hintelmann, 2004). Fish may take up arsenate in river water and
metabolize incorporated arsenate to AsB. Indeed, the freshwater
carnivorous fish T. mossambica accumulated arsenate in culture
medium and transformed small amounts of it into trimethylarsenic
species (Suhendrayatna et al., 2002). However, the freshwater
omnivorous fish Cyprinus carpio exposed to arsenate were unable
to produce AsB (Maeda et al., 1993). Accordingly, the presence of
AsB in freshwater fish may result from predation on other organ-
ism(s). The freshwater carnivorous killifish O. latipes, which preys
on the herbivorous shrimp N. denticulata containing dimethylated
arsenicals and small amounts of trimethylated arsenicals, predom-
inantly possessed trimethylated arsenicals (>80%) (Kuroiwa et al.,
1994). The internal production of AsB by fish after predation on
other organism(s) containing precursor arsenical(s) of AsB has
not been effectively proven.
3.3. Similarity of arsenic speciation patterns between freshwater
crustaceans and fish
Most of the extractable arsenicals were present as AsB in both
freshwater crustaceans and fish samples from the Hayakawa River,
although little is known regarding the biosynthesis of AsB. It is now
speculated that AsB is formed via degradation of arsenosugars by
microbial activity (Ritchie et al., 2004). Water extracts of fish and
mussel samples from a river contained only traces of AsB, and
oxo-arsenosugar-phosphate was the major arsenical (Schaeffer
et al., 2006). Arsenosugars were also the major arsenicals in fresh-
water mussels (Koch et al., 2001). Crustacean and fish muscle
tissues analyzed in this study contained, to some extent, oxo-
arsenosugar-glycerol or oxo-arsenosugar-phosphate. These oxo-
arsenosugars presumably came from predation on algae, because
the freshwater green macroalga C. glomerata, collected from the
Hayakawa River, contained the same types of oxo-arsenosugars.
The crustaceans and six kinds of fish analyzed in this study con-
tained small amounts of TMAO. In the silver carp Hypophthamich-
thys molitrix, collected from a river, TMAO was the only arsenical
detected (Schaeffer et al., 2006). Moreover, the marine teleost fish
silver drummer Kyphosus sydneyanus muscle tissue contained
TMAO as the primary extractable arsenical (Edmonds et al.,
1997). Hanaoka (2004) reported that AsB could be degraded into
TMAO by microorganisms. However, the common marine shrimp
Crangon crangon exposed to TMAO in food contained unmetabo-
lized TMAO as a minor arsenical (Hunter et al., 1998). Conse-
quently, the TMAO present in analyzed crustaceans and fish may
come from internal production from precursor arsenical(s) of
TMAO, microbial degradation of AsB, or predation on other organ-
ism(s) containing TMAO.
3.4. Total arsenic and speciation analyses of arsenate-exposed
freshwater green microalga C. reinhardtii CC125
C. reinhardtii CC125 cells contained 11 000 ± 290 lg kg1 DW
total arsenic after exposure to 200 lg L1 arsenate for 3 d. The con-
centration of the water-soluble arsenicals was 3400 ± 790 lg kg1
DW, and the extraction yield was quite low (32%), similar to the
biological samples from the Hayakawa River. Oxo-arsenosugar-
phosphate was the primary water-soluble arsenical detected at a
concentration of 2400 ± 500 lg kg1 DW (Fig. 3 g). In addition,
traces of arsenate and DMAA and oxo-arsenosugar-glycerol
(620 ± 120 lg kg1 DW) were also detected. These results indicate
that the freshwater green microalga C. reinhardtii CC125 took up
arsenate from the culture medium and transformed it into oxo-
arsenosugar-glycerol and oxo-arsenosugar-phosphate.
3.5. Comparison of arsenic speciation patterns between freshwater and
marine organisms
While arsenosugars are most abundant in marine algae, AsB is
the predominant form of arsenic in marine fish and crustaceans,
along with traces of MAA, DMAA, TMAO, and TMA (Hirata and
Toshimitsu, 2007). Interestingly, arsenic speciation patterns in
marine and freshwater organisms have some similarities, because
the freshwater green macroalga and fish samples in the present
study primarily contained arsenosugars and AsB, respectively.
However, AsB was identified as a minor arsenical in some marine
algae (Nischwitz and Pergantis, 2005), although the presence of
AsB was not confirmed in our freshwater green algal samples. Fur-
thermore, total arsenic concentrations in the freshwater green
macroalga, crustaceans, and fish from the Hayakawa River
(18 000, 2600 ± 90, and 150–2100 lg kg1 DW, respectively) were
much lower than those in same types of marine samples (1400–
62 000, 14 000–68 000, and 3500–60 000 lg kg1 DW, respectively)
(Francesconi and Edmonds, 1997). In addition, the extraction yields
of the water-soluble arsenicals from the freshwater samples ana-
lyzed in this study were quite low (16–42%) compared with those
from marine organisms (>70%) (Schaeffer et al., 2005). These differ-
ences imply that freshwater organisms possess non-extractable
arsenical(s) to some extent, which might be important for under-
standing arsenic behavior in freshwater ecosystems.
4. Conclusions
In this study, arsenic species in water and biological samples
from the arsenic-rich Hayakawa River and freshwater green micro-
alga C. reinhardtii CC125 exposed to arsenate were analyzed using a
rapid determination method with HPLC-ICP-MS. The speciation
analyses revealed that the hot-spring water and river water con-
tained only arsenate, and the water bug collected from the river
also contained arsenate as a major water-soluble arsenical, while
the freshwater green algae, crustaceans, and fish contained
oxo-arsenosugar-glycerol and/or oxo-arsenosugar-phosphate in
various concentrations and, furthermore, the crustaceans and fish
contained arsenobetaine as a major water-soluble arsenical. In
 S. Miyashita et al. / Chemosphere 75 (2009) 1065–1073
contrast to marine organisms, these organisms also possessed non-
extractable arsenical(s) contributing more than 50% of the total ar-
senic concentrations. Future detailed studies of non-extractable
arsenical(s) in freshwater organisms will provide valuable infor-
mation about arsenic behavior in freshwater ecosystems.
Acknowledgements
The authors would like to acknowledge B.Sc. Hiroyuki Hirashi-
ma and B.Sc. Seiko Fujita for their analytical support. We also
would like to express our thanks to Kanagawa Prefectural Fisheries
Technology Center and Kanagawa Environmental Research Center
for their cooperation in sampling in the Hayakawa River.
References
Cullen, W.R., Reimer, K.J., 1989. Arsenic speciation in the environment. Chem. Rev.
89, 713–764.
Devesa, V., Súñer, M.A., Lai, V.W.-M., Granchinho, S.C.R., Martínez, J.M., Vélez, D.,
Cullen, W.R., Montoro, R., 2002. Determination of arsenic species in a
freshwater crustacean Procambarus clarkii. Appl. Organomet. Chem. 16, 123–
132.
Edmonds, J.S., Shibata, Y., Francesconi, K.A., Rippingale, R.J., Morita, M., 1997.
Arsenic transformations in short marine food chains studied by HPLC-ICP MS.
Appl. Organomet. Chem. 11, 281–287.
Francesconi, K.A., Edmonds, J.S., 1997. Arsenic and marine organisms. Adv. Inorg.
Chem. 44, 147–189.
Granchinho, S.C.R., Polishchuk, E., Cullen, W.R., Reimer, K.J., 2001. Biomethylation
and bioaccumulation of arsenic(V) by marine alga Fucus gardneri. Appl.
Organomet. Chem. 15, 553–560.
Hanaoka, K., 2004. Studies on arsenic circulation in marine ecosystems. Nippon
Suisan. Gakk. 70, 284–287.
Hirata, S., Toshimitsu, H., 2007. Determination of arsenic species and arsenosugars
in marine samples by HPLC-ICP-MS. Appl. Organomet. Chem. 21, 447–454.
Hunter, D.A., Goessler, W., Francesconi, K.A., 1998. Uptake of arsenate,
trimethylarsine oxide, and arsenobetaine by the shrimp Crangon crangon.
Mar. Biol. 131, 543–552.
Kaise, T., Ogura, M., Nozaki, T., Saitoh, K., Sakurai, T., Matsubara, C., Watanabe, C.,
Hanaoka, K., 1997. Biomethylation of arsenic in an arsenic-rich freshwater
environment. Appl. Organomet. Chem. 11, 297–304.
Kaise, T., Sakurai, T., Saitoh, T., Matsubara, C., Sakai, T., Inoue, Y., 1998. Chemical
species of arsenic in the river water at Mt. Hakone. Jpn. J. Tox. Environ. Health
44, 7.
Kanamori, S., Sugawara, K., 1965. Geochemical study of arsenic in natural waters. I.
Arsenic in rain and snow. J. Earth Sci. Nagoya Univ. 13, 23–35.
Koch, I., Feldmann, J., Wang, L., Andrewes, P., Reimer, K.J., Cullen, W.R., 1999. Arsenic
in the Meager Creek hot springs environment, British Columbia, Canada. Sci.
Total Environ. 236, 101–117.
Koch, I., Reimer, K.J., Beach, A., Cullen, W.R., Gosden, A., Lai, V.W.-M., 2001. Arsenic
speciation in freshwater fish and bivalves. In: Chappell, W.R., Abernathy, R.L.,
Calderon, R.L. (Eds.), Arsenic Exposure and Health Effects IV. Elsevier,
Amsterdam, pp. 115–123.
Kuroiwa, T., Ohki, A., Naka, K., Maeda, S., 1994. Biomethylation and
biotransformation of arsenic in a freshwater food chain: green alga (Chlorella
vulgaris) ? shrimp (Neocaridina denticulata) ? killifish (Oryzias latipes). Appl.
Organomet. Chem. 8, 325–333.
1073
Lawrence, J.F., Michalik, P., Tam, G., Conacher, H.B.S., 1986. Identification of
arsenobetaine and arsenocholine in Canadian fish and shellfish by high-
performance liquid chromatography with atomic absorption detection and
confirmation by fast atom bombardment mass spectrometry. J. Agr. Food Chem.
34, 315–319.
Madsen, A.D., Goessler, W., Pedersen, S.N., Francesconi, K.A., 2000. Characterization
of an algal extract by HPLC-ICP-MS and LC-electrospray MS for use in
arsenosugar speciation studies. J. Anal. Atom. Spectrom. 15, 657–662.
Maeda, S., Mawatari, K., Ohki, A., Naka, K., 1993. Arsenic metabolism in a freshwater
food chain: blue-green alga (Nostoc sp.) ? shrimp (Neocaridina
denticulata) ? carp (Cyprinus carpio). Appl. Organomet. Chem. 7, 467–476.
McAdam, D.P., Perera, A.M.A., Stick, R.V., 1987. The synthesis of (R)-20 , 30 -
dihydroxypropyl 5-deoxy-5-dimethylarsinyl-ß-D-riboside, a naturally
occurring arsenic-containing carbohydrate. Aust. J. Chem. 40, 1901–1908.
McSheehy, S., Pohl, P., Vélez, D., Szpunar, J., 2002. Multidimensional liquid
chromatography with parallel ICP MS and electrospray MS/MS detection as a
tool for the characterization of arsenic species in algae. Anal. Bioanal. Chem.
372, 457–466.
Murray, L.A., Raab, A., Marr, I.L., Feldmann, J., 2003. Biotransformation of arsenate to
arsenosugars by Chlorella vulgaris. Appl. Organomet. Chem. 17, 669–674.
Nischwitz, V., Pergantis, S.A., 2005. First report on the detection and quantification
of arsenobetaine in extracts of marine algae using HPLC-ES-MS/MS. Analyst
130, 1348–1350.
Peshut, P.J., Morrison, R.J., Brooks, B.A., 2008. Arsenic speciation in marine fish and
shellfish from American Samoa. Chemosphere 71, 484–492.
Ritchie, A.W., Edmonds, J.S., Goessler, W., Jenkins, R.O., 2004. An origin for
arsenobetaine involving bacterial formation of an arsenic–carbon bond. FEMS
Microbiol. Lett. 235, 95–99.
Schaeffer, R., Soeroes, C., Ipolyi, I., Fodor, P., Thomaidis, N.S., 2005. Determination of
arsenic species in seafood samples from the Aegean Sea by liquid
chromatography-(photo-oxidation)-hydride generation-atomic fluorescence
spectrometry. Anal. Chim. Acta 547, 109–118.
Schaeffer, R., Francesconi, K.A., Kienzl, N., Soeroes, C., Fodor, P., Váradi, L., Raml, R.,
Goessler, W., Kuehnelt, D., 2006. Arsenic speciation in freshwater organisms
from the river Danube in Hungary. Talanta 69, 856–865.
Shibata, Y., Morita, M., 1992. Characterization of organic arsenic compounds in
bivalves. Appl. Organomet. Chem. 6, 343–349.
Shiomi, K., Sugiyama, Y., Shimakura, K., Nagashima, Y., 1995. Arsenobetaine as the
major arsenic compound in the muscle of two species of freshwater fish. Appl.
Organomet. Chem. 9, 105–109.
Slejkovec, Z., Bajc, Z., Doganoc, D.Z., 2004. Arsenic speciation patterns in freshwater
fish. Talanta 62, 931–936.
Soeroes, C., Goessler, W., Francesconi, K.A., Kienzl, N., Schaeffer, R., Fodor, P.,
Kuehnelt, D., 2005. Arsenic speciation in farmed Hungarian freshwater fish. J.
Agr. Food Chem. 53, 9238–9243.
Suhendrayatna Ohki, A., Nakajima, T., Maeda, S., 2002. Studies on the accumulation
and transformation of arsenic in freshwater organisms II. Accumulation and
transformation of arsenic compounds by Tilapia mossambica. Chemosphere 46,
325–331.
Wilkie, J.A., Hering, J.G., 1998. Rapid oxidation of geothermal arsenic(III) in
streamwaters of the eastern Sierra Nevada. Environ. Sci. Technol. 32, 657–
662.
Yamaoka, Y., Carmona, M.L., Oclarit, J.M., Jin, K., Shibata, Y., 2001. Arsenic
compounds in marine sponge (Haliclona permolis, Halichondria japonica,
Halichondria okadai and Haliclona sp. white) from Seto Inland Sea, Japan. Appl.
Organomet. Chem. 15, 261–265.
Zheng, J., Hintelmann, H., 2004. Hyphenation of high performance liquid
chromatography with sector field inductively coupled plasma mass
spectrometry for the determination of ultra-trace level anionic and cationic
arsenic compounds in freshwater fish. J. Anal. Atom. Spectrom. 19, 191–195.