Arch Toxicol (2005) 79: 183191 DOI 10.

1007/s00204-004-0620-x

I NO R G A N IC CO M PO U ND S

Toru Hayakawa Yayoi Kobayashi Xing Cui Seishiro Hirano

A new metabolic pathway of arsenite: arsenicglutathione complexes are substrates for human arsenic methyltransferase Cyt19

Received: 5 August 2004 / Accepted: 21 September 2004 / Published online: 4 November 2004 Springer-Verlag 2004

Abstract The metabolism of arsenic is generally accepted to proceed by repetitive reduction and oxidative methylation; the latter is mediated by arsenic methyltransferase (Cyt19). In human urine, the major metabolites of inorganic arsenicals such as arsenite (iAsIII) and arsenate (iAsV) are monomethylarsonic acid (MMAV) and dimethylarsinic acid (DMAV). On the other hand, in rat bile, the major metabolites of iAsIII have been reported to be arsenicglutathione (As-GSH) complexes. In the present study we investigate whether these AsGSH complexes are substrates for arsenic methyltransferase by using human recombinant Cyt19. Analyses by high-performance liquid chromatographyinductively coupled plasma mass spectrometry suggested that arsenic triglutathione (ATG) was generated nonenzymatically from iAsIII when GSH was present at concentrations 2 mM or higher. Human recombinant Cyt19 catalyzed transfer of a methyl group from Sadenosyl-L-methionine to arsenic and produced monomethyl and dimethyl arsenicals. The methylation of arsenic was catalyzed by Cyt19 only when ATG was present in the reaction mixture. Moreover, monomethylarsonic diglutathione (MADG) was a substrate of Cyt19 for further methylation to dimethylarsinic glutathione (DMAG). On the other hand, monomethylarsonous acid (MMAIII), a hydrolysis product of MADG, was not methylated to dimethyl arsenical by Cyt19. These results suggest that As-GSH complexes such as ATG and MADG were converted by Cyt19 to MADG and DMAG, respectively. Both MADG and DMAG were unstable in solution when the GSH concentration
T. Hayakawa Y. Kobayashi X. Cui S. Hirano (&) Environmental Health Sciences Division, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, 305-8506 Ibaraki, Japan E-mail: seishiro@nies.go.jp Tel.: +81-29-8502512 Fax: +81-29-8502512 T. Hayakawa Faculty of Pharmaceutical Sciences, Chiba University, Yayoi, Inage, 263-8522 Chiba, Japan

was lower than 1 mM, and were hydrolyzed and oxidized to MMAV and DMAV, respectively. Metabolism of iAsIII to methylated arsenicals by Cyt19 was via ATG and MADG rather than by oxidative methylation of iAsIII and MMAIII. Keywords Arsenic Glutathione Cyt19 Metabolism HPLC ICP

Introduction
Inorganic arsenicals are worldwide environmental contaminants, and chronic exposure to arsenicals is known to cause skin lesions, vascular diseases, and cancers (Engel and Smith 1994; Hu et al. 2003). Methylated pentavalent arsenicals such as monomethylarsonic acid (MMAV) and dimethylarsinic acid (DMAV) are major metabolites of inorganic arsenicals in human urine (Crecelius 1977; Tam et al. 1979). Recently, it has been reported that dimethylarsinous acid (DMAIII) and monomethylarsonous acid (MMAIII) were found in the urine of arsenism patients in West Bengal (Mandal et al. 2001) and Romania (Aposhian et al. 2000). It is generally accepted that inorganic arsenicals are metabolized by repetitive reduction and oxidative methylation with DMAV being the nal metabolite in humans as shown in Fig. 1. In some animal species, DMAV is further methylated and converted to trimethylarsine oxide (TMAO), possibly via DMAIII (Vahter et al. 1984; Yamauchi and Yamamura 1984). In the metabolism of arsenicals, S-adenosyl-L-methionine (SAM) and reduced glutathione (GSH) are requisites as a methyl group donor and a reducing agent, respectively (Buchet and Lauwerys 1985, 1987). Purine nucleoside phosphorylase (PNP) has been reported as one of the arsenate reductases and was observed in vitro to reduce arsenate (iAsV) to arsenite (iAsIII), but could not catalyze the conversion of MMAV to MMAIII (Nemeti and Gregus 2002; Radabaugh et al. 2002).

184

However, PNP is suspected to play a role in vivo. Glutathione S-transferase x has been shown to catalyze the conversion of MMAV to MMAIII in various tissues (Sampayo-Reyes et al. 2000; Zakharyan et al. 2001). Besides GSH, dithiol compounds seem to enhance arsenic methylation (Buchet and Lauwerys 1988). Until recently, methylation was considered as a detoxication process for harmful inorganic arsenicals since the toxicity of MMAV and DMAV is much lower than that of inorganic arsenicals (Hirano et al. 2003, 2004). Recently, however, the methylation of arsenicals has been reported to be a bioactivation process because the toxicity of intermediate metabolites such as MMAIII and DMAIII was found to be much higher than that of iAsIII (Petrick et al. 2000; Styblo et al. 2000). New world animals such as marmosets, tamarines, squirrel monkeys and guinea pigs, and Old World animals such as chimpanzees, do not appear to have ecient arsenic methyltransferase activity and do not excrete either monomethyl or dimethyl arsenicals in urine (Aposhian 1997; Wildfang et al. 2001). However, this deciency in arsenic methyltransferase activity has not been associated with the sensitivity of animals to arsenic. Lin and colleagues puried SAM-dependent arsenic methyltransferase from rat liver cytosol and reported that rat arsenic methyltransferase was a homologue of human Cyt19 (Lin et al. 2002). They also reported that rat recombinant Cyt19 catalyzed the conversion of iAsIII to monomethyl and dimethyl arsenicals (Walton et al. 2003). Rat recombinant Cyt19 appears to possess both iAsIII methyltransferase and iAsV reductase activities (Waters et al. 2004). Kala et al. (2000) reported that iAsIII was metabolized in the rat liver and excreted as arsenicglutathione (As-GSH) complexes in bile. Although As-GSH complexes such as arsenic triglutathione (ATG) and monomethylarsenic diglutathione (MADG) are major metabolites in bile, these arsenicals are not depicted in the classical arsenic metabolism pathway as outlined in Fig. 1. In the present study, we postulate that As-GSH complexes are directly involved in the methylation of inorganic arsenicals. We generated recombinant human Cyt19 and investigated methylation of arsenic in vitro by measuring both unstable trivalent As-GSH complexes and stable pentavalent arsenic metabolites using highperformance liquid chromatographyinductively coupled plasma mass spectrometry (HPLC-ICP MS). We propose a new metabolic pathway of iAsIII in which AsGSH complexes are substrates for Cyt19.

Japan). GSH, tetrabuthylammonium hydroxide (TBAH) solution, malonic acid, oxalic acid, and other chemicals of analytical grade were purchased from Wako Pure Chemicals (Osaka, Japan). ATG and MADG were synthesized according to the methods described in the literature (Kala et al. 2000; Scott et al. 1993). Briey, iAsIII or MMAV was reacted with GSH in degassed water and the solution was incubated overnight under a nitrogen atmosphere at room temperature. The products were collected as white precipitates and dried in a desiccator under reduced pressure. MMAIII and DMAIII (without conjugation) were synthesized by reducing MMAV and DMAV using metabisultethiosulfate for 1 h at room temperature (Reay and Asher 1977), and were used as standards for HPLC-ICP MS analyses and binding assays. However, recent reports show that thioarsenicals were produced by the metabisultethiosulfate method (Hansen et al. 2004; Suzuki et al. 2004). The details about DMAIII are described later. Binding of arsenic compounds to the soluble rat liver cytosolic proteins The soluble rat liver cytosolic fraction was obtained from four male Sprague-Dawley rats (6 weeks old; CLEA Japan, Tokyo). The rat liver tissue was homogenized by a glass-Teon homogenizer with 4 volumes of 100 mM phosphate buer (pH 7.0). The cytosolic fraction was obtained by centrifugation of the homogenate at 100,500 g for 1 h at 4C. The supernatant was collected and kept frozen at 40C for the following experiments. All procedures were approved by Animal Care and Use Committee of the National Institute for Environmental Studies. We measured binding of each arsenic compound (iAsIII, iAsV, MMAIII, MMAV, DMAIII, and DMAV) to the rat liver cytosolic proteins. The reaction mixture (300 ll), containing 1 lM of each arsenic compound and 30 ll of the cytosolic fraction in 25 mM phosphate buer (pH 7.0), was incubated at 37C for 20 min. After incubation, the sample was boiled for 5 min, and centrifuged at 13,000 g for 20 min to precipitate denatured proteins. The precipitate was wet-digested with concentrated HNO3 and H2O2 (3:1 v/v) at 135C for 2 days. The arsenic concentrations in the digests were measured by ICP MS (HP4500; Yokokawa, Tokyo, Japan). Eects of H2O2 treatment on the recovery of arsenic metabolites

Experimental procedures
Chemicals Sodium arsenite (NaAsO2), sodium arsenate dibasic heptahydrate (Na2HAsO4.7H2O), and SAM were purchased from Sigma (St. Louis, MO, USA). DMAV and MMAV were purchased from TRI Chem. (Yamanashi, It is prerequisite to determine the recovery of arsenicals from the reaction mixture for HPLC-ICP MS analysis because trivalent arsenicals have strong anity to proteins. The reaction mixture (100 ll), containing 0.2 lM iAsIII, 1 mM SAM, 5 mM GSH, and 50 lg human recombinant Cyt19 in 25 mM phosphate buer (pH 7.0), was incubated at 37C for 3 h. The reaction was quen-

185 Fig. 1 The classical metabolic pathway of inorganic arsenic in mammals (iAsIII arsenite, iAsV arsenate, MMAIII monomethylarsonous acid, MMAV monomethylarsonic acid, DMAIII dimethylarsinous acid, DMAV dimethylarsinic acid, Cyt19 arsenic methyltransferase, GSH reduced glutathione, SAM S-adenosyl-L-methionine)
Cyt19

OH
GSH

OH As OH
reduction

OH O As CH3 OH
MMAV GSH

SAM, GSH

As OH OH
iAsV

OH
iAsIII

oxidative methylation

reduction

OH
GSH

OH O
reduction

Cyt19

SAM, GSH

OH As CH3

As CH3 CH3
DMAIII

As CH3 CH3
DMAV
oxidative methylation

OH
MMAIII

ched by boiling the sample for 5 min. The arsenic concentration in the sample was measured by ICP MS before and after treating the sample with H2O2 at a nal concentration of 3% to convert all arsenic metabolites to pentavalency. The H2O2-treated samples were then boiled again for 5 min. All samples were ltered though a 0.22-lm pore membrane before arsenic measurement by HPLC-ICP MS. Preparation of human recombinant Cyt19 Arsenic methyltransferase Cyt19 has been reported to be expressed in the HepG2 human hepatoma cell line (Lin et al. 2002). HepG2 cells were obtained from Riken (Saitama, Japan) and total RNA was extracted from the cells using TRIZOL (Invitrogen, Carlsbad, CA, USA). The coding region of human Cyt19 cDNA was amplied by reverse transcriptionpolymerase chain reaction (RTPCR) using a RNA PCR kit (ABI, Branchburg, NJ, USA) according to the manufacturers instructions (forward primer 5-ATGGCTGCACTTCGTGACGC3, reverse primer 5-GCAGCTTTTCTTTGTGCCACAGCA-3). Histidine-tagged Cyt19 was generated by ligation of the PCR product into pBAD TOPO TA expression vector (Invitrogen). Briey, TOP10 cells were transformed by the ligated vector by heat shock (42C, 30 s) and the colonies were selected on standard ampicillin-containing agar plates. The transformed cells were incubated in Luria-Bertani medium containing 100 lg/ ml ampicillin. Expression of Cyt19 was induced by 2% arabinose for 5 h. The cells were lyzed by B-PER (Pierce, Rockford, IL, USA) and the histidine-tagged Cyt19 (V5 epitope and 6His in the C-terminal) was puried using a B-PER 6His Fusion Protein Purication Kit (Pierce) according to the manufacturers recommendations. The recombinant Cyt19 was dialyzed twice against PBS to remove imidazole. The puried Cyt19 yielded a single band on sodium dodecyl sulfate polyacrylamide gel electrophoresis as stained with

Coomassie brilliant blue. The protein concentration was determined with a BCA Protein Assay kit (Pierce) using bovine serum albumin as a standard. HPLC-ICP MS analysis of As-GSH complexes using an anion exchange column The reaction mixture (100 ll), containing 0.2 lM iAsIII and 05 mM GSH in 25 mM phosphate buer (pH 7.0), was incubated at 37C for 3 h. As-GSH complexes were analyzed by HPLC-ICP MS on an anion exchange column (JJ-50 4D, 1504.6 mm; Shodex, Tokyo, Japan) at a ow rate of 0.6 ml/min at room temperature. The pH of the mobile phase (20 mM oxalic acid) was adjusted to 2.3 by ammonia solution. The eluate was directly introduced to the ICP MS and the arsenic signal (m/z 75) was monitored every 2 s. HPLC-ICP MS analysis of methylated arsenic compounds using a reversed phase column The reaction mixture (100 ll), containing 0.2 lM iAsIII, 1 mM SAM, 05 mM GSH, and 50 lg human recombinant Cyt19 in 25 mM phosphate buer (pH 7.0), was incubated at 37C for 3 h. The reaction was quenched by boiling the samples for 5 min. The conditions for analysis of arsenic metabolites by HPLC-ICP MS have been described elsewhere (Shraim et al. 2002, 2003). Because trivalent arsenicals are known to bind to proteins (Styblo et al. 1996; Styblo and Thomas 1997), the reaction products of Cyt19 were treated with H2O2 at a nal concentration of 3% to oxidize all the arsenicals to pentavalency and to release the arsenic metabolites from Cyt19 as described above. Separation of iAsV, MMAV, and DMAV was performed by HPLC on a reversed phase column (ODS-3, 1504.6 mm, 3-lm particle size; GL Science, Tokyo, Japan) at a ow rate of 1.0 ml/min at 50C. The pH of the mobile phase containing 5 mM TBAH, 3 mM malonic acid, and 5% (v/v) methanol was

186

Methylation of MADG by Cyt19 As described in the results, MADG was unstable and hydrolyzed in phosphate buer in the absence of GSH. MADG (0.2 lM) was reacted with 1 mM SAM and Cyt19 (0.25 mg/ml) in 25 mM phosphate buer at 37C for 3 h. The reaction products were analyzed on a reversed phase column by HPLC-ICP MS after H2O2 treatment as described in the previous section. Methylation of MMAV by Cyt19 in the presence of GSH We examined whether MMAV is a precursor of MMAIII or DMAV in the metabolic pathway as shown in Fig. 1. The reaction mixture (100 ll), containing either 0.2 lM MMAVor 0.2 lM iAsIII, 5 mM GSH, 1 mM SAM, and 25 lg human recombinant Cyt19 in 25 mM phosphate buer (pH 7.0), was incubated at 37C for 3 h. The reaction was quenched by boiling the samples for 5 min. The samples were treated with H2O2, ltered and analyzed by HPLC-ICP MS on a reversed phase column.

Arsenic content (ng)

approximately 5.6. The eluate was directly introduced to the ICP MS as described earlier.

MMAIII

DMAIII

MMAV

iAsV

Fig. 2 Binding anity of arsenic compounds to the rat liver cytosolic proteins. Arsenite (iAsIII), arsenate (iAsV), monomethylarsonous acid (MMAIII), monomethylarsonic acid (MMAV), dimethylarsinous acid (DMAIII), and dimethylarsinic acid (DMAV) were incubated with a cytosolic protein fraction of 20% rat liver homogenate. The reaction mixture (300 ll), containing 1 lM arsenic compound (22.5 ng As; the maximum expected value), and 30 ll of the cytosolic fraction, was incubated for 20 min at 37C in 25 mM phosphate buer (pH 7.0). The reaction was quenched by boiling and the sample was centrifuged to pellet down the aggregated proteins. Arsenic concentrations in the pellet were measured by inductively coupled plasma mass spectrometry after acid digestion. Data are presented as the means SE of four dierent experiments

Results
In the present study, we analyzed arsenic compounds by HPLC-ICP MS using two dierent columns: an anion exchange column for the analysis of As-GSH complexes and a reversed phase column for analysis of arsenic metabolites by Cyt19. The arsenic metabolites could be separated on the reversed phase column. However, AsGSH complexes were unstable under the elution conditions in the reversed phase column. On the other hand, As-GSH complexes were relatively stable under the elution conditions in the anion exchange column. The arsenic metabolites were analyzed by HPLC-ICP MS in the form of pentavalent arsenicals after treatment with H2O2 because a large amount of trivalent arsenicals bound to proteins, as shown in Fig. 2. One possible explanation for the low anity of DMAIII is that it is unstable and was oxidized quickly to DMAV. Although the metabisultethiosulfate method has been commonly used to prepare DMAIII, recent reports show that reduction of DMAV with metabisultethiosulfate produced pentavalent thioarsenicals as major products (Hansen et al. 2004; Suzuki et al. 2004). Thus, the low binding anity of DMAIII to proteins may be due to concomitant thioarsenicals in the DMAIII sample. In vitro, iAsIII was reacted with Cyt19 in the presence of 1 mM SAM and 5 mM GSH at 37C for 3 h. The reaction products were analyzed on a reversed phase column by HPLC-ICP MS before and after H2O2 treatment. Figure 3 shows that a larger amount of DMAV was recovered in the sample after treatment with H2O2.

MMAV and iAsV were not detected on the chromatogram unless the sample was treated with H2O2, suggesting the detected pentavalent arsenicals corresponded to trivalent arsenicals extracted from proteins by H2O2 treatment. We examined whether As-GSH complexes were produced in vitro in the presence of a physiological concentration of GSH. iAsIII (0.2 lM) was incubated with 05 mM GSH for 3 h at 37C, and the reaction products were analyzed on an anion exchange column. The generation of ATG increased with the concentration of GSH, and the conversion of iAsIII to ATG appeared to reach a plateau at 4 mM GSH (Fig. 4A). Under similar conditions as those for ATG formation (05 mM GSH), Cyt19 and SAM were added to the reaction mixture for in vitro arsenic methylation experiments. Figure 4B indicates that no methylated arsenicals were observed on the chromatogram at GSH concentrations of 2 mM or less. However, iAsIII was methylated at concentrations of GSH of 2 mM or more. The total methylation increased with the concentration of GSH, as shown by the increase in peak height for DMAV and the corresponding decrease for iAsV. The amount of MMAV appeared to stay almost constant regardless of the GSH concentration (25 mM), probably because the generated MMAV was further methylated to DMAV. A good correlation was observed between the production of ATG and methylation of iAsIII (Fig. 5). We next examined the conversion of MADG to DMAV in the presence or absence of 5 mM GSH. When MADG was dissolved in 25 mM phosphate buer (pH 7.0) in the absence of GSH, MADG was promptly hydrolyzed to MMAIII, and some of the MMAIII was oxidized to MMAV (Fig. 6A). On the other hand, in the presence of 5 mM GSH, the chemical form of MADG

DMAV

iAsIII

8 7 6 5 4 3 2 1 0

187 Fig. 3AC Eects of H2O2 treatment on the recovery of arsenic metabolites. A Standard solutions containing iAsIII, iAsV, MMAV, and DMAV (0.2 lM each) were analyzed on a reversed phase column by HPLC-ICP MS. The reaction mixture (100 ll), containing 0.2 lM iAsIII, 1 mM SAM, 5 mM GSH, and 50 lg Cyt19 in 25 mM phosphate buer (pH 7.0), was incubated for 3 h at 37C. B,C Arsenic metabolites were analyzed by HPLC-ICP MS before (B) and after treatment with H2O2 (C)

Intensity (cps)

4000 3000 2000 1000 0 0

iAsIII

(A)
DMAV MMAV iAs V

100

200

300

400

Time (sec)
Intensity (cps) Intensity (cps)
1500 1500 1000 500 0 0 100 200 300 400 0 100 200 300 400

DMAV

(B)
1000

(C)
iAsV MMAV

DMAV
500 0

Time (sec)
Fig. 4A,B HPLC-ICP MS analyses of A nonenzymatic production of arsenic triglutathione (ATG), and B glutathione (GSH)-dependent production of arsenic metabolites by human Cyt19. A Samples of 0.2 lM iAsIII were incubated with various concentrations of GSH in 25 mM phosphate buer (pH 7.0) for 3 h at 37C. ATG and other arsenicals were separated on an anion exchange column. B The assay mixtures (100 ll), containing 0.2 lM iAsIII, 1 mM S-adenosyl-L-methionine, various concentrations of GSH, and 50 lg Cyt19 in 25 mM phosphate buer (pH 7.0), were incubated for 3 h at 37C. The sample was treated with H2O2 and arsenic metabolites, iAsV, MMAV, and DMAV were separated on a reversed phase column. GSH concentrations used in A,B were a,g 0 mM, b,h 1 mM, c,i 2 mM, d,j 3 mM, e,k 4 mM, f,l 5 mM

Time (sec)

(A) Anion Exchange Column

ATG iAsIII 1000 cps

(B) Reversed Phase Column

DMAV 1000 cps iAsV MMA
V

(f)

5 mM

(l)

(e)

4 mM

(k)

(d)

3 mM

(j)

(c)

2 mM

(i)

(b)

1 mM

(h)

(a)

0 mM
0 100 200 300 400 500 0

(g)

100

200

300

400

500

Time (sec)
remained unchanged (Fig. 6B). When MADG was reacted with 1 mM SAM and Cyt19 (0.25 mg/ml) in the absence (Fig. 6C) and presence of 5 mM GSH (Fig. 6D), further methylation of MADG to dimethyl arsenicals only occurred in the latter case.

Time (sec)

In the classical metabolic pathway (Fig. 1), iAsIII and MMAV (or MMAIII) are precursors of MMAV and DMAV, respectively. We tested either iAsIII or MMAV as an initial substrate under in vitro arsenic methylation to examine whether iAsIII could be converted to DMAV

188
120 iAsV, MMAV, DMAV ( % of the sum of 3 arsenic species ) 100 35 80 60 20 40 20 5 0 0 1 2 3 GSH (mM) 4 5 0 15 10 30 25 ATG ( % of total As ) 45 40

Fig. 5 Correlation between glutathione (GSH)-related production of arsenic triglutathione (ATG) and the arsenic metabolites. The HPLC-ICP MS chromatograms in Fig. 4 were quantitatively analyzed using Plasma Chromatographic Software (Yokokawa, Japan). Columns show proportions of iAsV (open columns), MMAV (closed), and DMAV (hatched) as percentages of the total of the three arsenic species. Closed circles with a solid line show GSHdependent production of ATG

via MMAV. When iAsIII was reacted with 1 mM SAM, 5 mM GSH, and Cyt19 (0.25 mg/ml) at 37C for 3 h, a large amount of dimethyl arsenical was produced (Figs. 7A and 4). On the other hand, when MMAV was used, only a small amount of dimethyl arsenical was produced (Fig. 7B). These results suggest that a much larger amount of dimethyl arsenical was produced from iAsIII than from MMAV, although Fig. 1 indicates that MMAV is an intermediate product between iAsIII and dimethyl arsenicals. ATG was also generated nonenzymatically from iAsIII in the presence of 5 mM GSH (Figs. 7C and 4). Figure 7D shows that a small amount of MADG was produced by incubating MMAV with 5 mM GSH for 3 h. Thus, DMAV in Fig. 7B was probably produced from MADG rather than from oxidative methylation of MMAIII.

Discussion
The present study demonstrated that As-GSH complexes such as ATG and MADG were substrates for human arsenic methyltransferase Cyt19. Kala and coworkers reported that ATG and MADG were major metabolites in the bile of rats injected (intravenously) with iAsIII (Kala et al. 2000). Our in vitro study revealed good correlation between the production of ATG and methylated arsenic metabolites (Fig. 5). Similarly, monomethyl arsenicals were further methylated to dimethyl arsenicals by Cyt19 only in the presence of MADG (Fig. 6). In the classical pathway, arsenic is thought to be metabolized by repetitive reduction and oxidative methylation (Fig. 1). However, a larger amount of DMAV was produced from iAsIII (Fig. 7A) than from

Fig. 6AD HPLC-ICP MS analyses of the stability of monomethylarsenic diglutathione (MADG) and methylation of MADG. MADG was dissolved at 0.2 lM in 25 mM phosphate buer (pH 7.0) in the absence (A) or presence (B) of 5 mM glutathione (GSH) and the sample was analyzed immediately by HPLC-ICP MS using an anion exchange column. The methylation reaction was performed by adding S-adenosyl-L-methionine (1 mM) and Cyt19 (0.25 mg/ml) to the reaction mixtures of A and B and the samples were incubated at 37C for 3 h. The reaction products in the absence of GSH (C, corresponding to A) and presence of 5 mM GSH (D, corresponding to B) were analyzed by reversed phase column after treatment of the sample with H2O2

Anion Exchange Column

15000

Intensity (cps)

MMA

III

15000

MADG
(B)

(A)
10000 5000 0 0 100 200 300 400 10000

MMAV

5000 0 0 100 200 300 400

Time (sec)

Time (sec)

Cyt19
20000

Reversed Phase Column

4000 3000 2000 1000 0

Cyt19
DMAV
(D)

Intensity (cps)

15000 10000 5000 0 0 100

MMAV

(C)

200

300

400

500

100

200

300

400

500

Time (sec)

Time (sec)

189

Reversed Phase Column

Intensity (cps)
2000 1500 1000 500 0 0 100 200 300 400

DMAV

Intensity (cps)

2000 1500 1000 500 0 0 100

MMAV

(A)
iAsV MMAV

(B)

DMAV

200

300

400

Time (sec)

Time (sec)

Anion Exchange Column

Intensity (cps)
4000

ATG

Intensity (cps)

(C)
3000 2000 1000 0 0 100 200 300 400 500

3000 2500 2000 1500 1000 500 0 0 100

MMAV

(D)

iAsIII

MADG

200

300

400

Time (sec)

Time (sec)

Fig. 7AD Absence of MMAV in conversion from iAsIII to DMAV in the Cyt19-mediated metabolic process. A Methylation of iAsIII. The reaction mixture (100 ll), containing 0.2 lM iAsIII, 1 mM Sadenosyl-L-methionine (SAM), 5 mM glutathione (GSH), and 50 lg Cyt19 in 25 mM phosphate buer (pH 7.0), was incubated for 3 h at 37C. B Methylation of MMAV. The reaction mixture (100 ll), containing 0.2 lM MMAV, 1 mM SAM, 5 mM GSH, and 50 lg Cyt19 in 25 mM phosphate buer (pH 7.0), was incubated for 3 h at 37C. The reaction products were analyzed by HPLC-ICP MS equipped with a reversed phase column. C The reaction mixture of 0.2 lM iAsIII and 5 mM GSH in 25 mM phosphate buer (pH 7.0) was incubated for 3 h at 37C. The reaction products were analyzed by HPLC-ICP MS equipped with an anion exchange column. D The reaction mixture of 0.2 lM MMAV and 5 mM GSH in 25 mM phosphate buer (pH 7.0) incubated for 3 h at 37C and analyzed by HPLC-ICP MS equipped with an anion exchange column. Note that a small amount of monomethylarsenic diglutathione (MADG), which can be a substrate of Cyt19, was generated

MMAV (Fig. 7B), suggesting that iAsIII is a more preferred substrate for Cyt19 than MMAV in the catalysis of dimethyl arsenical production. These results contradict the metabolic pathway shown in Fig. 1 in which MMAV is depicted as a metabolite of iAsIII. Thus, it is unlikely that iAsIII is metabolized to dimethyl arsenicals through MMAV. Zakharyan and Aposhian (1999) reported that iAsIII was methylated nonenzymatically in the presence of methylcobalamin and GSH, and that methylation did not occur when either was absent. Methylation of iAsIII by methylcobalamin is thought to proceed by nucleophilic attack of the As-GSH complex on cobalt (Fig. 8A). The present study suggests that GSH is a requisite for the methylation of arsenic by Cyt19. We propose a similar reaction mechanism for the methylation of arsenic by Cyt19 whereby a lone pair from a thiol group of ATG attacks the cationic sulfur of SAM and a methyl group of SAM is transferred to ATG (Fig. 8B). The present study indicates that in vitro methylation of arsenic by Cyt19 proceeds according to the mecha-

nism shown in Fig. 9. After reduction of iAsV to iAsIII, iAsIII forms a complex with GSH producing ATG, and ATG is methylated to MADG by Cyt19 in the presence of SAM. MADG and MMAIII are in equilibrium depending on the GSH concentration and some of the MMAIII is oxidized to MMAV. We suppose that MADG is further methylated to DMAG by Cyt19. However, DMAG was not detected in the reaction mixture by HPLC-ICP MS, probably because DMAG was not stable and decomposed easily in solution. Since DMAIII is very easily oxidized to DMAV (Kato et al. 2003; Yamanaka et al. 2003), it is plausible that DMAG was immediately converted to DMAV in solution. The synthesized ATG and MADG were hydrolyzed to iAsIII and MMAIII, respectively, in phosphate buer (pH 7.0) in the absence of GSH, whereas both ATG and MADG were stable in the presence of 5 mM GSH. We suppose that As-GSH complexes (ATG, MADG, and DMAG) and trivalent arsenic compounds (iAsIII, MMAIII, and DMAIII) are in equilibrium depending on the GSH concentration. Moreover, the same metabolic process could occur in the liver, because ATG and MADG were found as major metabolites in bile and the concentration of GSH exceeded 10 mM in iAsIII-injected rats (Kala et al. 2000). In the conventional metabolic pathway of arsenic (Fig. 1), pentavalent metabolites are thought to be reduced to the more toxic trivalent compounds. However, in the new metabolic pathway that we describe (Fig. 9), trivalent metabolites are converted to the less toxic pentavalent compounds. This new pathway is in accordance with the concept that oxidation is detoxication of arsenic as suggested by Aposhian et al. (2003). In summary, both ATG and MADG are substrates for human arsenic methyltransferase Cyt19. Arsenite is metabolized and converted to DMAV by Cyt19 via formation of As-GSH complexes rather than via repetitive reduction and oxidative methylation.

190 Fig. 8 A Non-enzymatic mechanism of As methylation by methylcobalamin as proposed by Zakharyan and Aposhian (1999) (AGT arsenic triglutathione, MADG monomethylarsenic diglutathione). B A putative enzymatic mechanism of As methylation by arsenic methyltransferase (Cyt19) and S-adenosyl-L-methionine (SAM) in the present study
(A)

GS

Me

Me
As GS SG

Co
As GS SG

GS

+ Co

ATG

Methylcobalamin

MADG

Glutathionyl cobalamin

(B)

H2N

H C CH2 CH2

COOH N

NH2 N N GS

H2N

H C CH2 CH2

COOH N

NH2 N N

Cyt19
GS

GS

+ S O H H OH H

S + H

N O H H OH H OH

GS

As
GS

Me

H OH

As
GS

Me

ATG

S-adenosylmethionine (SAM)

MADG

Fig. 9 A new metabolic pathway of inorganic arsenic via arsenicGSH complexes. (iAsIII arsenite, iAsV arsenate, GSH reduced glutathione, AGT arsenic triglutathione, SAM Sadenosyl-L-methionine, Cyt19 arsenic methyltransferase, MADG monomethylarsenic diglutathione, MMAIII monomethylarsonous acid, MMAV monomethylarsonic acid, DMAG dimethylarsinic glutathione, DMAIII dimethylarsinous acid, DMAV dimethylarsinic acid)

OH OH HO As O
iAsV iAsIII SAM Cyt19 GSH

SG
ATG

OH HO As OH

GS

As

SG

OH OH HO As O
MMAV MMAIII SAM Cyt19 GSH

SG GS As CH3

CH3

MADG

HO

As

CH3

CH3 HO As O
DMAV

CH3

CH3 HO As
DMAIII

GSH

CH3 CH3 GS As CH3

DMAG

Acknowledgements This study was partially supported by Grantin-aid for Scientic Research (14390058) from The Japan Society for Promotion of Science.

References
Aposhian HV (1997) Enzymatic methylation of arsenic species and other new approaches to arsenic toxicity. Annu Rev Pharmacol Toxicol 37:397419 Aposhian HV, Gurzau ES, Le XC, Gurzau A, Healy SM, Lu X, Ma M, L Yip, Zakharyan RA, Maiorino RM, Dart RC, Tircus MG, Gonzalez Ramirez D, Morgan DL, Avram D,

Aposhian MM (2000) Occurrence of monomethylarsonous acid in urine of humans exposed to inorganic arsenic. Chem Res Toxicol 13:693697 Aposhian HV, Zakharyan RA, Avram MD, Kopplin MJ, Wollenberg ML (2003) Oxidation and detoxication of trivalent arsenic species. Toxicol Appl Pharmacol 193:18 Buchet JP, Lauwerys R (1985) Study of inorganic arsenic methylation by rat-liver in vitrorelevance for the interpretation of observations in man. Arch Toxicol 57:125129 Buchet JP, Lauwerys R (1987) Study of factors inuencing the in vivo methylation of inorganic arsenic in rats. Toxicol Appl Pharmacol 91:6574 Buchet JP, Lauwerys R (1988) Role of thiols in the in vitro methylation of inorganic arsenic by rat-liver cytosol. Biochem Pharmacol 37:31493153

191 Crecelius EA (1977) Changes in the chemical speciation of arsenic following ingestion by man. Environ Health Perspect 19:147 150 Engel RR, Smith AH (1994) Arsenic in drinking-water and mortality from vascular-diseasean ecologic analysis in 30 counties in the United States. Arch Environ Health 49:418427 Hansen HR, Raab A, Jaspars M, Milne BF,Feldmann J (2004) Sulfur-containing arsenical mistaken for dimethylarsinous acid [DMA(III)] and identied as a natural metabolite in urine: major implications for studies on arsenic metabolism and toxicity. Chem Res Toxicol 17:10861091 Hirano S, Cui X, Li S, Kanno S, Kobayashi Y, Hayakawa T, Shraim A (2003) Dierence in uptake and toxicity of trivalent and pentavalent inorganic arsenic in rat heart microvessel endothelial cells. Arch Toxicol 77:305312 Hirano S, Kobayashi Y, Cui X, Kanno S, Hayakawa T, Shraim A (2004) The accumulation and toxicity of methylated arsenicals in endothelial cells: important roles of thiol compounds. Toxicol Appl Pharmacol 198:458467 Hu CJ, Zhang AH, Huang HH (2003) Molecular pathology of skin carcinogenesis due to arsenicalism from coal-burning. Arch Environ Health 58:9296 Kala SV, Neely MW, Kala G, Prater CI, Atwood DW, Rice JS, Lieberman MW (2000) The MRP2/cMOAT transporter and arsenic-glutathione complex formation are required for biliary excretion of arsenic. J Biol Chem 275:3340433408 Kato K, Yamanaka K, Hasegawa A, Okada S (2003) Active arsenic species produced by GSH-dependent reduction of dimethylarsinic acid cause micronuclei formation in peripheral reticulocytes of mice. Mutat Res 539:5563 Lin S, Shi Q, Nix FB, Styblo M, Beck MA, Herbin-Davist KM, Hall LL, Simeonsson JB, Thomas DJ (2002) A novel S-adenosyl-L-methionine: arsenic(III) methyltransferase from rat liver cytosol. J Biol Chem 277:1079510803 Mandal BK, Ogra Y, Suzuki KT (2001) Identication of dimethylarsinous and monomethylarsonous acids in human urine of the arsenic-aected areas in West Bengal, India. Chem Res Toxicol 14:371378 Nemeti B, Gregus Z (2002) Reduction of arsenate to arsenite in hepatic cytosol. Toxicol Sci 70:412 Petrick JS, Ayala-Fierro F, Cullen WR, Carter DE, Aposhian HV (2000) Monomethylarsonous acid (MMA(III)) is more toxic than arsenite in Chang human hepatocytes. Toxicol Appl Pharmacol 163:203207 Radabaugh TR, Sampayo-Reyes A, Zakharyan RA, Aposhian HV (2002) Arsenate reductase II. Purine nucleoside phosphorylase in the presence of dihydrolipoic acid is a route for reduction of arsenate to arsenite in mammalian systems. Chem Res Toxicol 15:692698 Reay PF, Asher CJ (1977) Preparation and purication of 74Aslabeled arsenate and arsenite for use in biological experiments. Anal Biochem 78:557560 Sampayo-Reyes A, Zakharyan RA, Healy SM, Aposhian HV (2000) Monomethylarsonic acid reductase and monomethylarsonous acid in hamster tissue. Chem Res Toxicol 13:11811186 Scott N, Hatlelid KM, Mackenzie NE, Carter DE (1993) Reactions of arsenic(III) and arsenic(V) species with glutathione. Chem Res Toxicol 6:102106 Shraim A, Sekaran NC, Anuradha CD, Hirano S (2002) Speciation of arsenic in tube-well water samples collected from West Bengal, India, by high-performance liquid chromatographyinductively coupled plasma mass spectrometry. Appl Organomet Chem 16:202209 Shraim A, Cui X, Li S, Ng JC, Wang HP, Jin YL, Liu YC, Guo L, Li DS, Wang SQ, Zhang RZ, Hirano S (2003) Arsenic speciation in the urine and hair of individuals exposed to airborne arsenic through coal-burning in Guizhou, PR China. Toxicol Lett 137:3548 Styblo M, Thomas DJ (1997) Binding of arsenicals to proteins in an in vitro methylation system. Toxicol Appl Pharmacol 147:1-8 Styblo M, Delnomdedieu M, Thomas DJ (1996) Mono- and dimethylation of arsenic in rat liver cytosol in vitro. Chem Biol Int 99:147164 Styblo M, Del Razo LM, Vega L, Germolec DR, LeCluyse EL, Hamilton GA, Reed W, Wang C, Cullen WR, Thomas DJ (2000) Comparative toxicity of trivalent and pentavalent inorganic and methylated arsenicals in rat and human cells. Arch Toxicol 74:289299 Suzuki KT, Mandal BK, Katagiri A, Sakuma Y, Kawakami A, Ogra Y, Yamaguchi K, Sei Y, Yamanaka K, Anzai K, Ohmichi M, Takayama H, Aimi N (2004) Dimethylthioarsenicals as arsenic metabolites and their chemical preparations. Chem Res Toxicol 17:914921 Tam GKH, Charbonneau SM, Bryce F, Pomroy C, Sandi E (1979) Metabolism of inorganic arsenic (74As) in humans following oral ingestion. Toxicol Appl Pharmacol 50:319322 Vahter M, Marafante E, Dencker L (1984) Tissue distribution and retention of As-74-dimethylarsinic acid in mice and rats. Arch Environ Contam Toxicol 13:259264 Walton FS, Waters SB, Jolley SL, LeCluyse EL, Thomas DJ, Styblo M (2003) Selenium compounds modulate the activity of recombinant rat As-III-methyltransferase and the methylation of arsenite by rat and human hepatocytes. Chem Res Toxicol 16:261265 Waters SB, Devesa V, Del Razo LM, Styblo M, Thomas DJ (2004) Endogenous reductants support the catalytic function of recombinant rat Cyt19, an arsenic methyltransferase. Chem Res Toxicol 17:404409 Wildfang E, Radabaugh TR, Aposhian HV (2001) Enzymatic methylation of arsenic compounds. IX. Liver arsenite methyltransferase and arsenate reductase activities in primates. Toxicology 168:213221 Yamanaka K, Mizoi M, Tachikawa M, Hasegawa A, Hoshino M, Okada S (2003) Oxidative DNA damage following exposure to dimethylarsinous iodide: the formation of cis-thymine glycol. Toxicol Lett 143:145153 Yamauchi H, Yamamura Y (1984) Metabolism and excretion of orally-administered dimethylarsinic acid in the hamster. Toxicol Appl Pharmacol 74:134140 Zakharyan RA, Aposhian HV (1999) Arsenite methylation by methylvitamin B-12 and glutathione does not require an enzyme. Toxicol Appl Pharmacol 154:287291 Zakharyan RA, Sampayo-Reyes A, Healy SM, Tsaprailis G, Board PG, Liebler DC, Aposhian HV (2001) Human monomethylarsonic acid (MMAV) reductase is a member of the glutathione-Stransferase superfamily. Chem Res Toxicol 14:10511057