Novel Thioarsenic Metabolites in Human Urine ... - ACS Publications

Reingard Raml,† Walter Goessler,† Pedro Traar,† Takafumi Ochi,‡ and. Kevin A. Francesconi*,†. Institute of Chemistry-Analytical Chemistry, K...
1 downloads 0 Views 139KB Size
1444

Chem. Res. Toxicol. 2005, 18, 1444-1450

Novel Thioarsenic Metabolites in Human Urine after Ingestion of an Arsenosugar, 2′,3′-Dihydroxypropyl 5-Deoxy-5-dimethylarsinoyl-β-D-riboside Reingard Raml,† Walter Goessler,† Pedro Traar,† Takafumi Ochi,‡ and Kevin A. Francesconi*,† Institute of Chemistry-Analytical Chemistry, Karl-Franzens University Graz, Universitaetsplatz 1, 8010 Graz, Austria, and Laboratory of Toxicology, Faculty of Pharmaceutical Sciences, Tokyo University, Sagamiko, Kanagawa 199-0195, Japan Received April 27, 2005

The presence of arsenic-containing carbohydrates, arsenosugars, in many seafoods raises questions of human health concerning the ingestion and metabolism of these compounds. A previous study investigating the metabolites in human urine after the ingestion of a common arsenosugar 2′,3′-dihydroxypropyl 5-deoxy-5-dimethylarsinoyl-β-D-riboside (oxo-arsenosugar) showed that the arsenic was rapidly excreted in the urine and was present as at least 12 metabolites, only three of which could be identified. In this repeat study with oxo-arsenosugar and using high-performance liquid chromatography/inductively coupled plasma mass spectrometry, we report the identification of seven arsenic metabolites, which together accounted for 88% of the total urinary arsenic collected over 61 h. The metabolites included previously reported human urinary arsenicals dimethylarsinate (DMA), oxo-dimethylarsenoethanol (oxoDMAE), and trimethylarsine oxide, in addition to new human metabolites oxo-dimethylarsenoacetate (oxo-DMAA), thio-dimethlyarsenoacetate (thio-DMAA), thio-dimethylarsenoethanol (thio-DMAE), and the thio-arsenosugar. Cytotoxicity testing of the major metabolites DMA, oxo-DMAE, thio-DMAE, oxo-DMAA, and thio-DMAA showed that they were nontoxic even at 10 mM, except for DMA, which showed some toxic effects at 1 mM.

Introduction Adverse health effects, resulting from chronic exposure to inorganic arsenic in drinking water, such as increased risk of cancer and effects on the circulatory and neurological system, have been well-studied. The research has led to worldwide concern over the long-term health effects of exposure to inorganic arsenic and has resulted in the recommendation and enforcement of more stringent maximum permissible levels for arsenic in drinking water (1). For example, in 2006, the United States will adopt a maximum permissible level of 10 µg As/L (down from 50 µg As/L) (2), and some states, fearing that this is still too high, will enforce a more stringent limit of 4 µg As/L. Food is also a significant source of arsenic to humans. Assessing possible human health effects from arsenic in food, however, is difficult because the arsenic exists in many different chemical forms, and little is known regarding their metabolism and toxicity. Research carried out so far has sensibly focused on the most abundant arsenic compounds, namely, arsenobetaine, which is the major arsenical in fish and many other seafoods, and arsenosugars, which are significant compounds in particular seafoods such as edible algae, mussels, and oysters (3). The case for arsenobetaine is straightforward since this compound possesses no known toxicity (4) and is rapidly excreted unchanged in the urine by man (5). The situation for arsenosugars, however, is considerably more complex as shown by several studies conducted so far. † ‡

Karl-Franzens University Graz. Tokyo University.

Figure 1. Structures and abbreviations of arsenic compounds.

The early studies (6, 7) with humans used natural edible algae as the source of arsenosugars and showed that dimethylarsinate (DMA; see Figure 1) was a significant urinary metabolite and that several other unknown arsenical metabolites were also excreted in the urine. In a later study, Van Hulle et al. (8) identified DMA and oxo-dimethylarsenoethanol (oxo-DMAE) in human urine after ingestion of algae. Related studies (912) with sheep, a particular breed from the Orkney Islands that feeds almost exclusively on algae washed up on the beach, produced similar results. The sheep studies, however, identified three additional arsenic metabolites, namely, oxo-dimethylarsenoacetate (oxo-

10.1021/tx050111h CCC: $30.25 © 2005 American Chemical Society Published on Web 08/06/2005

Thioarsenic Metabolites in Urine after Arsenosugar Ingestion

DMAA) (10) and two novel thio-arsenic species, thiodimethylarsenoacetate (thio-DMAA) (11) and thio-dimethylarsinate (thio-DMA) (12). Although studies with actual food samples are toxicologically relevant, the interpretation of the data can be confounded by the mixture of arsenicals naturally present in the original ingested algal sample. For this reason, we examined in an earlier study (13) the human metabolism of the pure (chemically synthesized) oxo-arsenosugar and showed that it was quickly and almost completely metabolized to at least 12 arsenic products, only three of which could be identified, namely, DMA, oxo-DMAE, and trimethylarsine oxide (TMAO). Many of the metabolites, including two major unknown arsenicals, were unstable, which exacerbated difficulties in their identification at that time. We now report a repeat experiment examining the human metabolism of a pure arsenosugar and identify by high-performance liquid chromatography/ inductively coupled plasma mass spectrometry (HPLC/ ICPMS) seven arsenic metabolites in urine that collectively accounted for 88% of the excreted arsenic.

Materials and Methods Nomenclature. Figure 1 provides the structures, names, and abbreviations for the various compounds discussed in this study. Until recently, arsenic compounds found naturally in biological samples comprised two major groups, namely, tetralkylarsonio compounds (e.g., arsenobetaine and arsenocholine) and trialkylarsine oxides (e.g., oxo-arsenosugars and oxo-DMAE), while inorganic arsenic species (arsenite and arsenate) and mono- and dimethylated arsenic acids (e.g., DMA) were also present but generally in lower amounts. Of these arsenical groups, the arsine oxides are the most numerous with about 20 representatives. In the last year, however, a new group of arsenicals has been reported whose members are analogous to the arsine oxides except that they contain As-S (thio) rather than As-O (oxo). Although only four of these thio arsenicals have been reported so far, it seems probable that other thio arsenicals, analogous to the known oxo forms, are also present in biological samples and will in time be discovered. The oxo- and thio forms are readily interconverted; hence, the relative quantities of each found in a particular sample could vary depending on the physiological and chemical environment. Because almost all of the oxo forms are dimethylated, they have been named as dimethylarsinoyl compounds where the “arsinoyl” represents As-O [e.g., (CH3)2As(O)CH2COO- has been termed dimethylarsinoyl acetate and abbreviated as DMAA]. To facilitate reference to thio and oxo arsenicals in the current study, we will use the generic term “arseno” for both types of compounds with the prefix “thio” or “oxo” to distinguish them. Thus, the species (CH3)2As(S)CH2COO- is termed thioDMAA, and (CH3)2As(O)CH2COO- is termed oxo-DMAA. Standards and Reagents. All solutions were prepared with Milli-Q water (18.2 MΩ cm). Standard solutions of methylarsonate (MA), DMA, and TMAO were prepared as described elsewhere (14). The oxo-arsenosugar (2′,3′-dihydroxypropyl 5-deoxy-5-dimethylarsinoyl-β-D-riboside) was synthesized according to McAdam et al. (15), oxo-DMAA was synthesized according to Francesconi et al. (16), oxo-DMAE was synthesized according to Edmonds et al. (17), and the thio analogue of the arsenosugar was obtained as described by Schmeisser et al. (18). For the synthesis of thio-DMAA and thio-DMAE, 500 mg of the oxo analogues was dissolved in water (5 mL) and H2S was bubbled through the solution for 1 h. The flask was stoppered, and the solution was stirred for a further 1 h. The product was obtained after evaporation and crystallization. Because these compounds were tested, as part of this study, for cytotoxicity, it was essential that they were pure; their purity was established by NMR spectroscopy and elemental analysis. Thio-

Chem. Res. Toxicol., Vol. 18, No. 9, 2005 1445 DMAA (crystallized from dichloromethane in 90% yield): 1H NMR (400 MHz, DMSO): δ 3.11 (2H, s, CH2), 1.79 (6H, s, SAsMe2). 13C NMR (100 MHz, DMSO): δ 168.9 (CO), 40.1 (CH2, overlapped by DMSO), 19.2 (SAsMe2). C, 24.6; H, 4.6%. C4H9O2SAs requires C, 24.5; H, 4.6%. Thio-DMAE (crystallized from acetone in 91% yield): 1H NMR (400 MHz, DMSO): δ 3.75 (2H, t, CH2, J 6.4 Hz), 2.16 (2H, t, CH2, J 6.4 Hz), 1.68 (6H, s, SAsMe2). 13C NMR (100 MHz, DMSO): δ 55.9 (CH2), 36.6 (CH2), 19.0 (SAsMe2). C, 26.6; H, 6.1%. C4H11OSAs requires C, 26.4; H, 6%. An “analytical” synthesis of thio-DMA was carried out by adding 2 mL of a saturated H2S solution to 8 mL of a 10 µg As/L solution of DMA; after 2 days, argon was purged through the solution to remove excess H2S. Analysis with HPLC/ICPMS and HPLC/ESMS revealed a single peak accounting for 60% of the initial arsenic with the following mass fragments: 155 [Me2As(S)OH2]+, 137 [Me2AsS]+, 107 [AsS]+, and 91 [AsO]+. We note that although this arsenical has not been fully characterized, it was used as a tentative standard for thio-DMA. We are currently investigating this reaction in attempt to account for the missing 40%, which may represent one or more other reaction products not eluting from the column. The commercial products were used as follows: formic acid (p.a.), ammonium dihydrogen phosphate (p.a.), ammonium hydrogen carbonate (p.a), ammonium nitrate, and ortho phosphoric acid (p.a.) from Fluka (Buchs, Switzerland); methanol (p.a.) from Fisher Scientific (Leicestershire, United Kingdom); hydrogen sulfide from Messer Griesheim (Krefeld, Germany); and pyridine (p.a.), hydrogen peroxide 30% (p.a.), aqueous ammonia 25% (suprapure), and nitric acid (p.a.) from Merck (Darmstadt, Germany). Chemicals were used without further purification except for the nitric acid, which was distilled in a quartz sub-boiling distillation unit. Collection of Urine Samples. A volunteer (male, 38 years of age) refrained from eating food known to contain significant concentrations of arsenic (e.g., seafood or mushrooms) for 3 days before and during the experiment. The volunteer gave informed consent and was aware of the experimental details and possible effects of ingesting the arsenic compound. This experiment with a single individual is the first part of a larger study planned to investigate individual variability in arsenosugar metabolism conducted with the approval (12-203 ex 01/02) of the Ethics Commission of the Medical University Graz. The quantity of arsenic ingested was equivalent to about 4-fold the average daily amount of arsenic ingested by Japanese through edible algal products (19) (which contain mostly arsenosugars). On the first day of the experiment, a morning urine sample was collected to give the background concentration of arsenic. A stock solution of the oxo-arsenosugar was prepared in water, and its precise concentration was determined by ICPMS as 626 ( 11 mg As/L (n ) 3). A portion (1.51 g) of this solution (containing 945 µg of As) was ingested in about 200 mL of water at 09:15. The urine samples were individually collected in 500 mL polyethylene bottles during the following 4 days. The masses of the urine samples were recorded, and specific gravity was measured with a Leica TS 400 total solids refractometer (Leica Microsystems, Buffalo, NY); urine volumes were then calculated from the masses of the samples. In total, 20 individual samples were collected; they were stored briefly at -18 °C (domestic freezer) before being transferred to a -80 °C freezer until the analysis. Total Arsenic Determination. Arsenic concentrations were determined by ICPMS after mineralization of the samples with microwave-assisted acid digestion. Portions of urine (1.00 mL), nitric acid (2.00 mL), and water (2.00 mL) were transferred to 12 mL quartz vials of an autoclave digestion system (ultraClave 2, EMLS Leukirch, Germany). The system was closed and loaded with argon to 4 × 106 Pa, and the mixture was heated for 30 min at 250 °C. After digestion, the samples were diluted to 10.0 mL with water before analysis with an Agilent 7500c ICPMS (Agilent, Waldbronn, Germany). To test the accuracy of the measurement, two standard reference materials were included in the analysis process: DORM-2 (Dogfish Muscle,

1446

Chem. Res. Toxicol., Vol. 18, No. 9, 2005

Raml et al.

Table 1. Chromatographic Conditions Used in This Study method

column

mobile phase

flow rate (mL/min)

temp (°C)

A B C D E F

PRP-X100 PRP-X100 Zorbax 300 SCX Ionospher 5C Shodex NN-614 Atlantis

20 mM NH4H2PO4, pH 6.0, 3% methanol 20 mM NH4HCO3, pH 9.0, 3% methanol 20 mM pyridine, pH 2.6 10 mM pyridine, pH 3.0, 3% methanol 6 mM NH4NO3 and 5 mM HNO3, pH 2.2 20 mM NH4H2PO4, pH 3.0

1.5 1.5 1.5 1.5 0.4 1

40 40 30 40 25 30

National Research Council, Ottawa, Canada), certified [As] ) 18.0 ( 1.1 mg/kg, determined [As] ) 18.4 ( 0.6 mg/kg (n ) 4); and NIES CRM no. 18 (Human Urine, National Institute for Environmental Studies, Ibaraki, Japan), certified [As] ) 137 ( 11 µg/L, determined [As] ) 142 ( 11 µg/L (n ) 5). The % relative standard deviation (%RSD) of replicates (n ) 3) of the urine samples was 5-10%, except for the samples taken at 51, 57, and 59 h, which were close to our level of quantification (5 µg As/L), where the %RSD was 15-20%. The total arsenic concentrations were normalized (Cnorm) to the mean specific gravity using the equation:

Cnorm )

(spec.grav.mean - 1) × Csample spec.grav.sample - 1

Determination of Arsenic Species. The analytical system consisted of a Hewlett-Packard 1100 series HPLC (equipped with a quaternary pump, a vacuum degasser, column oven, and an autosampler with a variable 100 mm3 injection loop) connected to a Hewlett-Packard 4500 ICPMS (equipped with a Babington nebulizer) with 0.125 mm PEEK (polyetheretherketone) tubing (Upchurch Scientific, Oak Harbor, United States). The signal was detected at m/z 75 and m/z 77. The data evaluation was carried out with chromatographic software G1824C Version C.01.00. Quantification was performed with external calibration using MA or DMA as calibrants. The samples were filtered through 0.22 µm Nylon filters (Berger, Weisskirchen, Austria) and analyzed directly. HPLC separations were performed with five chromatographic columns: a PRP-X100 polymer based anion exchange column, 4.1 mm × 100 mm (Hamilton, Bonaduz, Switzerland); a Zorbax 300-SCX silica based cation exchange column, 4.6 mm × 150 mm (Agilent, Waldbronn, Germany); an Ionopsher 5C silica based cation exchange column, 3 mm × 100 mm (Varian, Middelburg, Netherlands); a Shodex NN-614 RSpack column, 6 mm × 150 mm (Showa Denko, Kawasaki, Japan); and an Atlantis C18 column, 4.6 mm × 150 mm (Waters, MA). The chromatographic conditions are summarized in Table 1. Chemical Conversions of Arsenicals in Urine. To portions of 1.0 mL of urine, 100 µL of water, H2O2 solution, or saturated aqueous solution of H2S was added. After about 1 h, the samples were filtered (0.22 µm) and analyzed. Preparative HPLC of the Unknown Arsenical A5. A subsample of the urine sample at t ) 9 h that had been stored about 10 months at -18 °C (by which time species prone to oxidation were almost completely transformed) was chromatographed with method C (Table 1) except that the injection volume was increased to 80 µL. The elution time of the major peak was recorded, and the eluate during this time period was subsequently collected in 10 consecutive runs. The collected solution was then evaporated under reduced pressure in a centrifugal lyophilizer (Heto Holton, Allerød, Denmark). The residue was dissolved in 200 µL of water and analyzed before and after addition of a saturated aqueous H2S solution (5 µL added to 50 µL of sample solution). Cytotoxicity Tests of the Five Major Urinary Metabolites. The five major urinary metabolites in this study, namely, DMA, oxo-DMAA, thio-DMAA, oxo-DMAE, and thio-DMAE, were investigated for cytotoxicity. Details of the general procedure employed for our cytotoxicity tests have been previously reported (20). In brief, HepG2 cells derived from human hepatocarcinoma were grown in a monolayer in Dulbecco’s

modified Eagle’s medium and supplemented with 10% fetal bovine serum. The cells were cultured in an incubator in an atmosphere of 5% CO2 in humidified air. HepG2 cells, seeded at a density of 1 × 104 cells/well in 96 well plates (IWAKI, Japan) and incubated for 24 h, were placed in fresh medium that contained test chemicals and incubated for 48 h. Cytotoxicity was assayed by the method of WST-8 (Cell Counting Kit8, Wako, Japan), in which WST-8 was converted to soluble formazan by the action of mitochondrial dehydrogenase in viable cells. The cultures in 96 well plates were placed in 100 mm3 of medium that contained WST-8 and incubated for 1 h at 37 °C. The absorbance at 450 nm was determined by a multiplate reader.

Results and Discussion Excretion of Total Arsenic. After the ingestion of the oxo-arsenosugar, the volunteer quickly excreted arsenic in the urine (Table 2). Increased urinary arsenic concentrations were already evident 3 h after ingestion (t ) 3 h), and the peak arsenic concentration was recorded for the sample collected at t ) 9 h. The concentrations then steadily decreased and reached almost background levels after 48 h by which time about 81% of the ingested arsenic was accounted for in the urine samples. Although the peak concentration was observed considerably earlier in the current study as compared with the earlier study (13) (9 vs about 27 h), the total arsenic excreted (about 80%) was similar in both cases. The arsenic excretion rate in the current study was also somewhat faster than that observed in related studies with arsenosugars in algae where peak arsenic concentrations were recorded between 22 and 38 (7), 1525 (8), or 10-60 h (6). Arsenic Metabolites in Urine. Speciation analysis with HPLC/ICPMS showed that the urine samples collected between 3 and 45 h had a qualitatively similar pattern of arsenic species. For urine collected after 45 h, the arsenic concentrations were close to background levels, and most of the species were no longer detectable. The following discussion deals with the urine sample collected at t ) 9 h, which had the highest concentration of arsenic. Anion exchange HPLC/ICPMS (method A) analysis revealed the presence of six significant (A1-A6) and several minor arsenic peaks (Figure 2a), a pattern very similar to that reported in the earlier study with another volunteer (13). The peak A1 was a front peak and presumably contained cationic and neutral arsenic species. The chromatogram of the urine sample performed under cation exchange conditions was dominated by a front peak but also revealed the presence of several cationic species. Thus, oxo-DMAE (6% of total) and unchanged oxo-arsenosugar (1%) were identified by spiking experiments (oxo-DMAE, methods C-E; oxo-arsenosugar, methods C and D). In addition, chromatography with method D indicated the presence of a trace amount of TMAO. These results are entirely consistent with the earlier study (13).

Thioarsenic Metabolites in Urine after Arsenosugar Ingestion

Chem. Res. Toxicol., Vol. 18, No. 9, 2005 1447

Table 2. Urinary Excretion of Total Arsenic after Ingestion of Oxo-arsenosugar (≡ 945 µg As) time (h)

volume (mL)

specific gravity

arsenic concna (µg As/L)

sum of speciesb (µg As/L)

0 1 3 9 13 14 16 21 24 26 32 36 39 45 48 51 56 57 59 61

545 341 492 481 476 485 285 476 484 507 297 293 216 343 625 496 426 485 532

1.027 1.009 1.008 1.012 1.011 1.005 1.006 1.020 1.010 1.005 1.014 1.020 1.012 1.025 1.012 1.006 1.015 1.005 1.005 1.005

10