The Seleno Bis(S-glutathionyl) Arsinium Ion Is ... - ACS Publications

Chem. Res. Toxicol. 2006, 19, 601-607

601

The Seleno Bis(S-glutathionyl) Arsinium Ion Is Assembled in Erythrocyte Lysate Shawn A. Manley,† Graham N. George,‡ Ingrid J. Pickering,‡ Richard S. Glass,§ Elmar J. Prenner,| Raghav Yamdagni,† Qiao Wu,† and Ju¨rgen Gailer*,† Departments of Chemistry and Biological Sciences, UniVersity of Calgary, 2500 UniVersity DriVe Northwest, Calgary, Alberta, T2N 1N4, Canada, Department of Geological Sciences, UniVersity of Saskatchewan, 114 Science Place, Saskatoon, Saskatchewan, S7N 5E2, Canada, and Department of Chemistry, UniVersity of Arizona, Tucson, Arizona 85721, USA ReceiVed December 13, 2005

Approximately 75 million people are currently exposed to arsenic concentrations in drinking water, which is associated with the development of internal cancers. One way to ameliorate this undesirable situation is to remove arsenic (arsenite and arsenate) from drinking water. An alternative approach is the development of an inexpensive palliative dietary supplement that promotes the excretion of intestinally absorbed arsenite from the body. To this end, the simultaneous administration of New Zealand white rabbits with arsenite and selenite resulted in the biliary excretion of the seleno-bis (S-glutathionyl) arsinium ion, [(GS)2AsSe]-. This apparent detoxification mechanism has been recently extended to environmentally relevant doses [Gailer, J., Ruprecht, L., Reitmeir, P., Benker, B., and Schramel, P. (2004) Appl. Organometal. Chem. 18, 670-675]. The site of formation of this excretory product in the organism, however, is unknown. To investigate if [(GS)2AsSe]- is formed in rabbit blood, we added arsenite and selenite and analyzed blood aliquots using arsenic and selenium X-ray absorption spectroscopy. The characteristic arsenic and selenium X-ray absorption spectra of [(GS)2AsSe]- were detected within 2 min after addition and comprised 95% of the blood selenium 30 min after addition. To elucidate if erythrocytes are involved in the biosynthesis of [(GS)2AsSe]- in blood, arsenite and 77Se-selenite were added to rabbit erythrocyte lysate and the obtained solution was analyzed by 77Se NMR spectroscopy (273 K). This resulted in a 77Se NMR signal with a chemical shift identical to that of synthetic [(GS)2AsSe]added to lysate. Combined, these results demonstrate that [(GS)2AsSe]- is rapidly formed in blood and that erythrocytes are an important site for the in vivo formation of this toxicologically important metabolite. Introduction Arsenic and selenium are widely distributed in the Earth’s crust (1, 2). Because of its high affinity for sulfur, arsenic is predominantly found in the form of sulfidic minerals, such as orpiment (As2S3) and realgar (As4S4). Selenium is also associated with sulfur butsin contrast to arsenicsisomorphically substitutes sulfur in the lattice structure of sulfur-containing minerals. Although selenium is an essential trace element for all higher animals including humans (1), selenium compounds will become toxic when larger doses are ingested (2). Conversely, inorganic arsenic (arsenite + arsenate) displays both acute and chronic toxicity and is known to cause cancer of the skin, lung, urinary bladder, liver, and kidney in humans (3). Natural sources and anthropogenic activities release arsenic and selenium compounds to environmental waters, where the most commonly found species are the oxo-anions arsenite, arsenate, selenite, and selenate (4, 5). Among these environmentally abundant molecular forms, arsenite and selenite are the most toxic (4, 5). On the basis of the drinking water quality guidelines of the World Health Organization, it is estimated that up to 75 million people are currently exposed to concentrations of inorganic arsenic in drinking water that are associated with the aetiology of internal cancers (6-9). In Bangladesh, for instance, * To whom correspondence should be addressed. † Department of Chemistry, University of Calgary. ‡ University of Saskatchewan. § University of Arizona. | Department of Biological Sciences, University of Calgary.

the initiation of a “safe drinking water program” has inadvertently resulted in the contamination of the public drinking water supply with geogenic arsenic on a massive scale (7). An unexpected detoxification between two individually highly toxic metalloid compounds was discovered in the 1930s, when arsenite added to the drinking water of rats abolished the typical symptoms of selenium poisoning caused by the ingestion of seleniferous wheat (10). Subsequent studies revealed that the coadministation of arsenite can also overcome the toxicity of selenite in cattle, dogs, and swine (11), which suggested a common underlying mammalian detoxification mechanism. Investigations aimed at elucidating the molecular basis of this striking mammalian mineral antagonism revealed that arsenite dramatically increased the biliary excretion of selenium (given as selenite) (12) and that, conversely, selenite promoted the biliary excretion of arsenic (given as arsenite) (13). These results prompted speculations about the in vivo formation and biliary excretion of an arsenic and selenium-containing compound (13). The actual existence and the molecular structure of this hypothetic compound, however, was not revealed until more than 60 years after the initial discovery of this mineral antagonism and revealed the excretion of the seleno-bis (S-glutathionyl) arsinium ion, [(GS)2AsSe]-, in rabbit bile (14-16). Evidence in favor of a biliary excretion of [(GS)2AsSe]- in rabbits after the simultaneous administration of environmentally relevant doses of arsenite and selenite suggests that the in vivo formation of this metabolite represents a mechanism of considerable toxicological importance (17).

10.1021/tx0503505 CCC: $33.50 © 2006 American Chemical Society Published on Web 03/18/2006

602 Chem. Res. Toxicol., Vol. 19, No. 4, 2006

The formation of [(GS)2AsSe]- involves the reaction of equimolar arsenite and selenite with 8 mole equivalents of glutathione (GSH) (14, 18) according to eq 1.

As(OH)3 + HSeO3- + 8GSH f [(GS)2AsSe]- + 3GSSG + 6H2O (1) Therefore, mammalian cells with particularly high endogenous concentrations of GSH may be predicted to be important sites for the in vivo formation of [(GS)2AsSe]- from arsenite and selenite. Because GSH is one of the most abundant intracellular low molecular weight thiols in erythrocytes (3.0 mM; ref 19) and hepatocytes (7.5 mM; ref 20), eq 1 couldsin principles proceed in these cells. One prerequisite for the assembly of [(GS)2AsSe]- in erythrocytes, however, is the individual import of arsenite and selenite. Because arsenite has been previously demonstrated to be accumulated by rat and rabbit erythrocytes (21-23) and because selenite is also accumulated by rat and human erythrocytes (24-26), all reactants for eq 1 are available in erythrocytes. The previous observation that the coadministration of arsenite significantly delayed the translocation of iv-administered selenite from the bloodstream to the liver in rats (27) and the apparent involvement of erythrocytes in this delay (28, 29) prompted the latter authors to speculate that these findings could be explained by the intraerythrocytic formation of selenium-arsenic complexes. According to eq 1, the formation of [(GS)2AsSe]- inside erythrocytes could therefore represent the molecular basis for these observations. To provide direct experimental evidence for an assembly of [(GS)2AsSe]from arsenite and selenite in rabbit blood andsmore specifically in rabbit erythrocytesswe employed two element-specific techniques that are ideally suited to identify metalloid containing metabolites in complex biological matrices, namely, arsenic and selenium K-edge X-ray absorption spectroscopy (XAS) and 77Se NMR spectroscopy.

Experimental Procedures Caution: Inorganic arsenic compounds are established human carcinogens and should be handled carefully (3). Ingestion of arsenite may cause cancer of the skin, urinary bladder, kidneys, lungs, and liVer, as well as disorders of the circulatory and nerVous systems. Chemicals. 5,5′-Dithiobis (2-nitro-benzoic acid), ethylenediaminetetraacetic acid disodium salt dihydrate (>99%), trizma base (99.9%), and trichloroacetic acid (>99%) were obtained by Sigma (St. Louis, MO). Diphenyl diselenide (>97%) was purchased from Fluka (Buchs, Switzerland), and NaAsO2 (>99%) was purchased from GFS Chemicals (Columbus, OH). Na2SeO3 (>98%), L-glutathione (reduced, >99%), and NaOH (>97%) were purchased from Sigma Chemicals. D2O (99.9 atom %) was purchased from CDN Isotopes (Pointe-Claire, QC, Canada). 77Selenium powder (gray; >98%; isotopic enrichment, 99.96%) was purchased from Cambridge Isotope Laboratories (Andover, MA). HNO3 (68%), methanol (>99.8%), and HCl (38%) were purchased from EMD Chemicals (Darmstadt, Germany). Krebs-Ringer bicarbonate buffer (pH 6.4) was prepared from dry powder pouches (Sigma Aldrich, St. Louis, MO) and deionized H2O (Simplicity, Millipore, Molsheim, France). A 20 mM concentration of PBS buffer (pH 7.4) was prepared by dissolving two phosphate buffered saline tablets (Sigma) in deionized H2O (200 mL). Treatment of Blood with Arsenite and Selenite. NaAsO2 (86.5 mg) and Na2SeO3‚5H2O (175.1 mg) were dissolved in distilled water and filled to the 50 mL mark in a volumetric flask. An aliquot (20 µL) of this solution was added to 2.0 mL of fresh blood collected from male New Zealand white rabbits (10 mL of blood was collected in a Vacutainer tube with EDTA as the anticoagulant).

Manley et al. The metalloid concentrations in the blood were 10 mg/L and were chosen to obtain good signal-to-noise arsenic and selenium nearedge spectra in a reasonable amount of time. The temperature of the blood was maintained at 37 °C (water bath), and aliquots of blood (120 µL) were withdrawn after 2, 10, 30, and 60 min, immediately mixed with glycerol (80 µL), and frozen in liquid nitrogen until analyzed by XAS. To ascertain that the concentration of both metalloid compounds in blood did not adversely affect the integrity of the erythrocytes (e.g., lysis), this experiment was repeated once more together with a control (blood without no metalloid compounds added). After 60 min at 37 °C, the erythrocytes were pelleted (centrifuge). Because the plasma of the treated blood (with metalloid compounds added) was indistinguishable from that of the control blood (without metalloid compounds added), the used metalloid concentrations did not result in significant lysis within 60 min of treatment. XAS. K-edge arsenic and selenium X-ray absorption near-edge spectra were measured at the Stanford Synchrotron Radiation Laboratory on Beamline 7-3 using a Si(220) double-crystal monochromator with an upstream vertical aperture of 1 mm. Harmonic rejection was accomplished by detuning one monochromator crystal to approximately 50% off-peak, and no specular optics were present. Spectra were measured as the X-ray fluorescence K R excitation spectra using a Canberra 30 element Ge detector. The incident X-ray intensity was monitored using a N2-filled ionization chamber, and energy calibration was performed by simultaneous measurement of elemental gray arsenic and hexagonal selenium foils, the lowest energy inflection points of which were assumed to be 11867.0 and 12658.0 eV, respectively. Samples were maintained at 10 K during data collection using an Oxford Instruments liquid helium flow cryostat. X-ray absorption spectroscopic data were analyzed using the EXAFSPAK suite of computer programs (http://ssrl.slac. stanford.edu/exafspak.html). Preparation of Erythrocyte Lysate. The Animal Care Committee of the University of Calgary approved all animal experimental procedures (Protocol Approval #BI 2005-27). Male New Zealand white rabbits were purchased from Casey Vandermeer (Edmonton, AB, Canada) and fed ad libitum on a “high-fiber” diet (LabDiet 5321, Canadian Lab Diets, Leduc, AB, Canada). Following halothane anesthesia, blood was collected by heart puncture (June 10, 2005, 10:15 am) into ice-chilled heparinized Vacutainer tubes (Becton Dickinson, Franklin Lakes, NJ), and all subsequent steps were performed at 4 °C. After centrifugation (1000g, 20 min), plasma and buffy coat were removed, the erythrocytes were slurried with an equal volume of ice-cold Krebs-Ringer buffer, and centrifugation was repeated. After the removal of the supernatant, the washing procedure with the Krebs-Ringer buffer was repeated once more, and the erythrocytes were centrifuged (2000g, 10 min). After the supernatant was removed, the washed erythrocytes were lysed with an equal volume of D2O (of room temperature) and centrifuged at 20000g (40 min) to remove cell debris following a published procedure (30). The obtained clear lysate (∼42 mL) was stored in aliquots (2.0 mL) at -30 °C. 77Se NMR Spectroscopy. All 77Se NMR spectra were acquired on a Bruker AMX 300 at 57.25 MHz, using D2O as the lock signal, and a 1.0 M diphenyl diselenide solution in CHCl3 (sealed in a glass capillary and placed inside a 10 mm i.d. glass NMR tube) corresponding to a 77Se NMR chemical shift of 463.0 ppm (31). Every 77Se NMR spectrum was obtained using a total sample volume of 2.5 mL. A 10 mm multinuclear NMR probe was used without proton decoupling, and scans were accumulated using a 90° pulse width of 12 µs, a spectral width of 45045 Hz, and a relaxation delay of 1.5 s. All samples were measured without spinning at 298, 280, or 273 K. Nitrogen gas was used to purge the headspace above the sample in the NMR tube because [(GS)2AsSe]- is oxygen sensitive (14). All spectra were plotted with a line-broadening factor of 20 Hz, except the ones obtained of synthetic [(GS)2AsSe]- in D2O (8 Hz). Effect of Temperature on 77Se NMR Shift of [(GS)2AsSe]-. [(GS)2AsSe]- was synthesized by the addition of equimolar Na2SeO3 and NaAsO2 to a solution of GSH in D2O (adjusted to

Seleno Bis(S-glutathionyl) Arsinium Ion

Chem. Res. Toxicol., Vol. 19, No. 4, 2006 603

Figure 1. Comparison of arsenic and selenium X-ray absorption near-edge spectra of rabbit blood 60 min after the addition of arsenite and selenite (10 ppm per metalloid) with spectra of relevant arsenic and selenium model compounds. (A) Arsenic K near-edge spectra of (a) whole blood, (b) [(GS)2AsSe]-, and (c) arsenite (broken line) and (GS)3As (solid line). (B) Selenium K near-edge spectra of (a) whole blood, (b) [(GS)2AsSe]-, and (c) selenite (broken line) and GS-Se-SG (solid line).

pH 7.4 by 4.0 M NaOH) following a previously reported procedure (18) but using D2O instead of Tris buffer. After the collection of 1860 scans at 298 K, the sample was cooled to 280 K (inside the spectrometer) and another 1860 scans were collected. 77Se NMR Spectrum of Erythrocyte Lysate after the Addition of Synthetic [(GS)2AsSe]-. An aliquot of synthetic [(GS)2AsSe](200 µL, 909 µg of Se) was added to 2.45 mL of erythrocyte lysate, and 34132 scans were collected at 280 K. GSH Assay in Erythrocyte Lysate. The concentration of GSH in the obtained lysate was determined using an established spectrophotometric assay (32) and was 0.967 ( 0.003 mM (the correlation coefficient of the standard curve was 0.9996). 77Se NMR Spectra of Erythrocyte Lysate after the Addition of Na277SeO3 and NaAsO2. Because the preparation of erythrocyte lysate resulted in an approximate 1:1 dilution of the GSH concentration [D2O:erythrocytes (v/v)], exogenous GSH (50 µL of a 0.2 M solution adjusted to pH 7.82 with 4.0 M NaOH) was added to 2.5 mL of lysate in order to increase the GSH concentration to levels that have been reported for mammalian erythrocytes (3.0 mM; ref 19). This lysate will be referred to as “GSH-fortified” lysate. NaAsO2 (4.12 mg) was dissolved in 824 µL of PBS buffer, and the pH of this solution was adjusted to 7.47 with 185 µL of 10% HNO3 and 6 µL of 4.0 M NaOH. A 100 µL amount of this NaAsO2 solution was added to a solution obtained by dissolving 77Se powder (2.498 mg, Sartorius MC5, Goettingen, Germany) in 50 µL of concentrated HNO3 and 69 µL of 8.0 M NaOH. After the addition of 1.688 mL of PBS buffer, the pH was adjusted to 7.42 (4.0 M NaOH). An aliquot of this solution (57 µL) was added to the GSHfortified lysate at room temperature (total 77Se added, 74 µg). After it was stirred for 2 min, the sample was cooled to 277 K in an ice-water bath, and the accumulation of 77Se NMR spectra was initiated (33294 scans at 273 K). An aliquot of the 77Se selenite and arsenite solution (17 µL; total 77Se added, 22 µg) was added to 2.5 mL of “GSH-unfortified” lysate at room temperature (following the same sample preparation), and 88666 scans were collected at 273 K (this experiment was performed in duplicate).

Results Figure 1 shows the As and Se near-edge X-ray absorption spectra obtained by treating blood with arsenite and selenite after 60 min. These spectra were essentially identical to those of [(GS)2AsSe]- and quite distinct from the spectra of the starting materials (arsenite and selenite). Fitting the experimental spectra with linear combinations of the spectra of

Figure 2. Estimated precentage of arsenic (O) and selenium (b) in the form of [(GS)2AsSe]- in blood following the addition of arsenite and selenite, as described in the text. Similar kinetics can be seen for both elements.

model species (33) allowed us to obtain quantitative information on the arsenic and selenium species that were present in the samples at the various time points. The percentage of total As and total Se that is present as [(GS)2AsSe]- is shown as a function of time in Figure 2. The fitting indicates that the remainder of the arsenic and selenium in blood appears to be essentially arsenite and selenite, respectively (not illustrated). This suggests that the formation of [(GS)2AsSe]- in blood is not the rate-limiting step because other intermediates would build up if it were. Using 77Se NMR to detect [(GS)2AsSe]- in rabbit erythrocyte lysate required the reproduction of the previously reported 77Se NMR chemical shift for synthetic [(GS) AsSe]- of -5.7 2 ppm (14). The 77Se NMR chemical shift obtained at room temperature was -4.9 ppm (Figure 3A) and shifted to -6.9 ppm when the spectra were measured at 280 K (Figure 3B, note the concomitant peak sharpening). The 77Se NMR spectra of all lysate samples were measured at 280 or 273 K to prevent the decomposition of [(GS)2AsSe]- during the accumulation of the spectra. Figure 4 depicts the 77Se NMR spectra that were obtained after the addition of (A) synthetic [(GS)2AsSe]- to lysate, (B) NaAsO2 and Na277SeO3 to GSH-fortified lysate, and (C) NaAsO2 and Na277SeO3 to GSH-unfortified lysate. The observed chemical shifts were -8.0, -8.7, and -8.9 ppm, respectively (duplication of experiment C resulted in a 77Se NMR signal at -8.7 ppm).

604 Chem. Res. Toxicol., Vol. 19, No. 4, 2006

Figure 3. 77Se NMR spectra of synthetic [(GS)2AsSe]- obtained at room temperature (A) and 280 K (B).

Figure 4. 77Se NMR spectra of lysate after the addition of (A) synthetic [(GS)2AsSe]- to erythrocyte lysate (350 ppm Se, 34132 scans), (B) NaAsO2 and Na277SeO3 to “GSH-fortified” erythrocyte lysate (28 ppm 77 Se, 33294 scans), and (C) NaAsO2 and Na277SeO3 to “GSHunfortified” erythrocyte lysate (9 ppm 77Se, 88666 scans).

Discussion Despite significant advances toward a better understanding of the mammalian metabolism of arsenitesparticularly over the past 15 yearssthe biomolecular basis of its chronic toxicity and its carcinogenicity in humans remains elusive (3, 34-36). This undesirable situation must be largely attributed to our limited understanding of the molecular biotransformations of arsenite after its absorption into the mammalian bloodstream. Even though arsenite has been demonstrated to readily accumulate in rat and rabbit erythrocytes (21-23) where it will form labile (GS)xAs(OH)(3-x) complexes (x ) 1-3) (37, 38), it is not understood which arsenic species is translocated from the blood

Manley et al.

to the liver, where toxicologically important enzymatic methylation reactions occur (39, 40). Our understanding of the mammalian metabolism of arsenite took an unexpected turn when a previously unknown arsenite metabolite, [(GS)2AsSe]-, was detected in bile after the iv injection of rabbits with selenite followed by arsenite (14, 15). The in vivo formation of this metabolite demonstrates that the mammalian biochemistry of arsenite is not only driven by interactions with endogenous thiols, such as GSH, but is also fundamentally interconnected with the essential trace element selenium (14, 17, 41). This metabolic interrelatedness of the mammalian biochemistry of arsenite and selenite was substantiated when [(GS)2AsSe]- was also detected in bile (within 25 min) after the injection of rabbits with arsenite and selenate with the identification of hepatocytes as the actual site of formation of [(GS)2AsSe]- (42). Hence, in vivo interactions between arsenite and inorganic selenium compounds could be fundamentally involved in establishing the molecular basis for the toxicological response of arsenite (e.g., cancer) at the organ level (41). Arsenic and selenium XAS were used to investigate if [(GS)2AsSe]- can be formed in mammalian blood after the addition of arsenite and selenite. For a description of the fundamental principles of this synchrotron-based spectroscopic technique, the reader is referred to a previously published article (43). The addition of arsenite and selenite to rabbit blood and the subsequent examination of the near-edge portion of the obtained X-ray absorption spectra revealed the rapid formation of an arsenic-selenium compound (within 2 min) with the characteristic spectrum of [(GS)2AsSe]-. This species accounted for 95% of the detected blood selenium after 30 min and virtually 100% after 60 min (Figures 1 and 2). Neither X-ray absorption near edge spectroscopy nor 77Se NMR (see below) can explicitly identify the sulfur donor (RS) in the detected arsenic-selenium species in blood, and it is therefore referred to as [(RS)2AsSe]-. On the basis of the concentration of GSH in mammalian erythrocytes (3 mM; ref 19) and the more than 100-fold lower concentrations of GSH and L-cysteine in rat and human plasma (44), however, GSH most likely represents the sulfur donor in the arsenic-selenium compound [(RS)2AsSe]detected in blood. This in turn strongly suggests that [(GS)2AsSe]was rapidly formed in blood. XAS is ideally suited to investigate the in vivo chemistry of arsenic and selenium compounds but has the significant disadvantage that access to synchrotron beamtime is required. We therefore used 77Se NMR spectroscopy as an alternative approach that is more readily available to investigate if [(GS)2AsSe]- is assembled in erythrocyte lysate. 77Se NMR spectroscopy has already been successfully applied to identify selenium compounds in a complex biological matrix, such as yeast extracts (45). Because of the extraordinarily wide range of 77Se NMR chemical shifts of ∼3300 ppm reported for selenium compounds (31), the previously reported 77Se NMR chemical shift of synthetic [(GS)2AsSe]- essentially provides a fingerprint of this structurally unique selenium compound. 77Se NMR spectroscopy is therefore also suited to detect [(GS)2AsSe]- in complex biological matrices, such as erythrocyte lysate. The only prerequisites for the 77Se NMR spectroscopic detection of [(GS)2AsSe]- in a biological matrix are that (i) the compound itself does not decompose during the accumulation of the 77Se NMR spectra and (ii) sufficient 77Se is present in the form of [(GS)2AsSe]- to be detected. The 77Se NMR chemical shift of [(GS)2AsSe]- has been previously reported at -5.7 ppm (14) and was reproduced at

Seleno Bis(S-glutathionyl) Arsinium Ion

Figure 5. Proposed mechanism for the formation of [(GS)2AsSe]- in erythrocytes after the individual influx of arsenite and selenite. The nucleophilic attack of HSe- on (GS)2As-OH to yield [(GS)2AsSe]- has been demonstrated in aqueous solution (16). Export of [(GS)2AsSe]to plasma could be mediated by ATP-driven GS-X conjugate export pumps located in the erythrocyte membranes (47-49).

-4.9 ppm (Figure 3A), which is within the experimental error of 77Se NMR spectroscopy (31). Accumulation of the 77Se NMR spectra of synthetic [(GS)2AsSe]- at 280 K (instead of room temperature) resulted in a significant upfield shift to -6.9 ppm (Figure 3B) along with considerable peak sharpening (Figure 3A,B). The addition of synthetic [(GS)2AsSe]- to lysate and the subsequent accumulation of the 77Se NMR spectra at 280 K resulted in a 77Se NMR signal at -8.0 ppm (Figure 4A). This small shift of the 77Se signal of synthetic [(GS)2AsSe]from -6.9 (in D2O) to -8.0 ppm (in lysate) can be rationalized in terms of the different sample matrices in which the spectra were obtained. This demonstrates that 77Se NMR spectroscopy can be employed to identify [(GS)2AsSe]- in erythrocyte lysate. To elucidate if [(GS)2AsSe]- is assembled in lysate after the addition of arsenite and selenite (following the overall stoichiometry depicted in eq 1), we had to decrease the detection limit for the 77Se NMR spectroscopic detection of [(GS)2AsSe]-. Because selenium has six natural isotopes and because 77Se accounts for only 7.6% of the total selenium, we synthesized 77Se selenite from elemental 77Se. After increasing the GSH concentration of the prepared lysate from 0.967 (lysis decreased the GSH concentration as compared to that in intact erythrocytes) to 3.0 mM [the GSH level of mammalian erythrocytes (19)], equimolar arsenite and 77Se selenite were added to this GSH-fortified lysate and the solution was analyzed by 77Se NMR spectroscopy. The detection of a single 77Se NMR signal at -8.7 ppm (Figure 4B) indicatedsfor the first timesthat [(GS)2AsSe]- had been assembled in GSH-fortified lysate (Figure 4A,B). The most relevant proof for the assembly of [(GS)2AsSe]in lysate required the detection of the “fingerprint” of this compound after the addition of arsenite and 77Se selenite to GSH-unfortified lysate and resulted in a 77Se NMR signal at -8.9 ppm (Figure 4C), which was confirmed in a repeat experiment (77Se NMR signal at -8.7 ppm). This result provides direct experimental evidence for the formation of [(GS)2AsSe](or a structurally very similar species) in rabbit erythrocyte lysate

Chem. Res. Toxicol., Vol. 19, No. 4, 2006 605

(Figure 4A-C). We note that due to the time it took to accumulate the 77Se NMR spectra (∼48 h), it is possible that the initially formed [(GS)2AsSe]- may have undergone thiol exchange with other intraeryhtrocytic thiols, such as hemoglobin, cysteinylglycine, or γ-glutamylcysteine. The accumulation of the 77Se NMR spectra at 273 K, however, likely minimized secondary thiol exchange. Because it has been previously demonstrated that arsenite and selenite can be individually accumulated by erythrocytes (21, 22, 24), a chemically feasible mechanism for the intraerythrocytic assembly of [(GS)2AsSe]- is depicted in Figure 5. After selenite enters the erythrocyte, it will be reduced in three consecutive two-electron reduction steps to HSe- (46). Conversely, arsenite will react with GSH to form (GS)2AsOH under the physicochemical conditions inside erythrocytes (37). On the basis of the previously reported nucleophilic attack of HSeon (GS)2AsOH in aqueous solution to give [(GS)2AsSe]- in virtually quantitative yield (16) and according to the results reported in this study (Figure 4B,C), we propose that this reaction essentially proceeds in eryhrocytes following eq 1. In view of the fact that ATP-driven GSH-conjugate export pumps have been identified in mammalian erythrocyte membranes (47-49), it is conceivable that the formed [(GS)2AsSe]- will be subsequently expelled to the plasma. Because this process takes time, it could explain the observed “delayed processing” of selenite in blood (regarding its translocation to the liver) when arsenite is coadministered (27, 28). Using 77Se NMR spectroscopy, we have demonstrated that the addition of arsenite and 77Se selenite to erythrocyte lysate results in the formation of [(GS)2AsSe]-. This finding reinforces the previously expressed notion that erythrocytes may be regarded as the first line of defense against ingested toxic metals and metalloid compounds (41, 50) and has important repercussions with regard to the toxicity of arsenite in mammals. Because free selenite has been detected in human plasma (51), the intraerythrocytic formation of [(GS)2AsSe]- may be the first mechanism by which mammals detoxify ingested arsenite. According to the proposed mechanism depicted in Figure 5, arsenite will react with equimolar selenite in erythrocytes to form [(GS)2AsSe]-, which is then expelled (from the erythrocyte), transported to the liver, and ultimately excreted in the bile. Any leftover arsenite (after the formation of [(GS)2AsSe]in erythrocytes) will then be translocated to the liver, where the established enzymatic methylation reactions will convert it to the molecular forms, which have been identified in urine (52). Combined, these results extend the previously established mechanism for the in vivo formation of [(GS)2AsSe]- from the liver (42) to the erythrocyte. It remains to be determined, however, if [(GS)2AsSe]- is formed in the bloodstream after the oral exposure of mammals to arsenite and selenite. The intraerythrocytic assembly of [(GS)2AsSe]- has important repercussions with regard to a better understanding of the toxicity of arsenite since its formation, followed by its biliary excretion and thus the removal of toxic arsenite from the systemic circulation, could explain the observed prevention of cytotoxic effects of arsenite in mice after the dietary supplementation with selenite (53) and may represent the molecular basis for the systemic toxicity of arsenite in mammals (38). Acknowledgment. We thank the National Science and Engineering Research Council (NSERC) of Canada for support of this research. Wayne Jansen and Dr. Douglas Morck (Animal Health Unit, LESARC, University of Calgary) are gratefully acknowledged for the collection of rabbit blood.

606 Chem. Res. Toxicol., Vol. 19, No. 4, 2006

References (1) Schwarz, K., and Foltz, C. M. (1957) Selenium as an integral part of factor 3 against dietary necrotic liver degeneration. J. Am. Chem. Soc. 70, 3292-3293. (2) Finkelman, R. B., Belkin, H. E., and Zheng, B. (1999) Health impacts of domestic coal use in China. Proc. Natl. Acad. Sci. U.S.A. 96, 34273431. (3) Kitchin, K. Y. (2001) Recent advances in arsenic carcinogenesis: modes of action, animal model systems, and methylated arsenic metabolites. Toxicol. Appl. Pharmacol. 172, 249-261. (4) Conde, J. E., and Alaejos, M. S. (1997) Selenium concentrations in natural and environmental waters. Chem. ReV. 97, 1979-2003. (5) Cullen, W. R., and Reimer, K. J. (1989) Arsenic speciation in the environment. Chem. ReV. 89, 713-764. (6) Nordstrom, D. K. (2002) Worldwide occurrences of arsenic in drinking water. Science 296, 2143-2145. (7) Nickson, R., McArthur, J., Burgess, W., Ahmed, K. M., Ravenscroft, P., and Rahman, M. (1998) Arsenic poisoning of Bangladesh groundwater. Nature 395, 338. (8) Islam, F. S., Gault, A. G., Boothman, C., Poyla, D. A., Charnock, J. M., Chatterjee, D., and Lloyd, J. R. (2004) Role of metal-reducing bacteria in arsenic release from Bengal delta sediments. Nature 430, 68-71. (9) Rahman, M. M., Chowdhury, U. K., Mukherjee, S. C., Mondal, B. K., Paul, K., Lodh, D., Biswas, B. K., Chanda, C. R., Basu, G. K., Saha, K. C., Roy, S., Das, R., Palit, S. K., Quamruzzaman, Q., and Chakraborti, D. (2001) Chronic arsenic toxicity in Bangladesh and West Bengal, IndiasA review and commentary. Clin. Toxicol. 39, 683-700. (10) Moxon, A. L. (1938) The effect of arsenic on the toxicity of seleniferous grains. Science 88, 81. (11) Levander, O. A. (1977) Metabolic interrelationships between arsenic and selenium. EnViron. Health Perspect. 19, 159-164. (12) Levander, O. A., and Baumann, C. A. (1966) Selenium metabolism. V. Studies on the distribution of selenium in rats. Toxicol. Appl. Pharmacol. 9, 98-105. (13) Levander, O. A., and Baumann, C. A. (1966) Selenium metabolism. VI. Effect of arsenic on the excretion of selenium in the bile. Toxicol. Appl. Pharmacol. 9, 106-115. (14) Gailer, J., George, G. N., Pickering, I. J., Prince, R. C., Ringwald, S. C., Pemberton, J. E., Glass, R. S., Younis, H. S., DeYoung, D. W., and Aposhian, H. V. (2000) A metabolic link between arsenite and selenite: The seleno-bis(S-glutathionyl) Arsinium Ion. J. Am. Chem. Soc. 122, 4637-4639. (15) Gailer, J., Madden, S., Buttigieg, G. A., Denton, M. B., and Younis, H. S. (2002) Identification of [(GS)2AsSe]- in rabbit bile by sizeexclusion chromatography and simultaneous multielement-specific detection by inductively coupled plasma atomic emission spectroscopy. Appl. Organomet. Chem. 16, 72-75. (16) Gailer, J., George, G. N., Pickering, I. J., Buttigieg, G. A., Denton, M. B., and Glass, R. S. (2002) Synthesis, X-ray absorption spectroscopy and purification of the seleno-bis (S-glutathionyl) arsinium anion from selenide, arsenite and glutathione. J. Organomet. Chem. 650, 108-113. (17) Gailer, J., Ruprecht, L., Reitmeir, P., Benker, B., and Schramel, P. (2004) Mobilization of exogenous and endogenous selenium to bile after the intravenous administration of environmentally relevant doses of arsenite to rabbits. Appl. Organomet. Chem. 18, 670-675. (18) Gailer, J., Madden, S., Burke, M. F., Denton, M. B., and Aposhian, H. V. (2000) Simultaneous multielement-specific detection of a novel glutathione-arsenic-selenium ion [(GS)2AsSe]- by ICP-AES after micellar size-exclusion chromatography. Appl. Organomet. Chem. 14, 355-363. (19) Griffith, O. W. (1981) Glutathione turnover in human erythrocytes. J. Biol. Chem. 256, 4900-4904. (20) Bellomo, G., Vairetti, M., Stivala, L., Mirabelli, F., Richelmi, P., and Orrenius, S. (1992) Demonstration of nuclear compartmentalization of glutathione in hepatocytes. Proc. Natl. Acad. Sci. U.S.A. 89, 44124416. (21) Winski, S. L., and Carter, D. E. (1995) Interactions of rat red blood cell sulfhydryls with arsenate and arsenite. J. Toxicol. EnViron. Health 46, 379-397. (22) Delnomdedieu, M., Styblo, M., and Thomas, D. J. (1995) Time dependence of accumulation and binding of inorganic and organic arsenic species in rabbit erythrocytes. Chem.-Biol. Interact. 98, 69-83. (23) Shiobara, Y., Ogra, Y., and Suzuki, K. T. (2001) Animal species difference in the uptake of dimethylarsinous acid (DMAIII) by red blood cells. Chem. Res. Toxicol. 14, 1446-1452. (24) Suzuki, K. T., Shiobara, Y., Itoh, M., and Ohmichi, M. (1997) Selective uptake of selenite by red blood cells. Analyst 123, 63-67. (25) Rohling, O., and Neidhart, B. (1999) Use of erythrocyte ghosts for preconcentration in element speciation. Anal. Chem. 71, 1077-1082.

Manley et al. (26) Galanter, W. L., Hakimian, M., and Labotka, R. J. (1993) Structural determinants of substrate specificity of the erythrocyte anion transporter. Am. J. Physiol. 265, C918-C926. (27) Palmer, I. S., and Bonhorst, C. W. (1957) Selenium poisoning. Modification of selenite metabolism by arsenite. J. Agric. Food Chem. 5, 928-930. (28) Gregus, Z., Gyurasics, A., and Koszorus, L. (1998) Interactions between selenium and group Va-metalloids (arsenic, antimony and bismuth) in the biliary excretion. EnViron. Toxicol. Pharmacol. 5, 8999. (29) Csanaky, I., and Gregus, Z. (2003) Effect of selenite on the disposition of arsenate and arsenite in rats. Toxicology 186, 33-50. (30) Dodge, J. T., Mitchell, C., and Hanahan, D. J. (1963) The preparation and chemical characteristics of hemoglobin-free ghosts of human erythrocytes. Arch. Biochem. Biophys. 100, 119-130. (31) Duddeck, H. (1995) Selenium-77 nuclear magnetic resonance spectroscopy. Prog. NMR Spectrosc. 27, 1-323. (32) Sedlak, J., and Lindsay, R. H. (1968) Estimation of total, proteinbound, and nonprotein sulfhydryl groups in tissue with Ellman’s reagent. Anal. Biochem. 25, 192-205. (33) Pickering, I. J., Prince, R. C., Divers, T. C., and George, G. N. (1998) K-edge X-ray absorption spectroscopy for determining the chemical speciation of sulfur in biological systems. FEBS Lett. 441, 11-14. (34) Goering, P. L., Aposhian, H. V., Mass, M. J., Cebrian, M., Beck, B. D., and Waalkes, M. P. (1999) The enigma of arsenic carcinogenesis: Role of Metabolism. Toxicol. Sci. 49, 5-14. (35) Liu, S. X., Athar, M., Lippai, I., Waldren, C., and Hei, T. K. (2001) Induction of oxyradicals by arsenic: Implication for mechanism of genotoxicity. Proc. Natl. Acad. Sci. U.S.A. 98, 1643-1648. (36) Abernathy, C. O., Liu, Y.-P., Longfellow, D., Aposhian, H. V., Beck, B., Fowler, B., Goyer, R., Menzer, R., Rossman, T., Thompson, C., and Waalkes, M. (1999) Arsenic: health effects, mechanisms of action, and research issues. EnViron. Health Perspect. 107, 593-597. (37) Gailer, J., and Lindner, W. (1998) On-column formation of arsenicglutathione species detected by size-exclusion chromatography in conjunction with arsenic-specific detectors. J. Chromatogr. B 716, 8393. (38) Thomas, D. J., Styblo, M., and Lin, S. (2001) The cellular metabolism and systemic toxicity of arsenic. Toxicol. Appl. Pharmacol. 176, 127144. (39) Aposhian, H. V. (1997) Enzymatic methylation of arsenic species and other new approaches to arsenic toxicity. Annu. ReV. Pharmacol. Toxicol. 37, 397-419. (40) Lin, S., Shi, Q., Nix, F. B., Styblo, M., Beck, M. A., Herbin-Davis, K. M., Hall, L. L., Simeonsson, J. B., and Thomas, D. J. (2002) A novel S-adenosyl-L-methionine:Arsenic(III) methyltransferase from rat liver cytosol. J. Biol. Chem. 277, 19795-10803. (41) Gailer, J. (2002) Reactive selenium metabolites as targets of toxic metals/metalloids in mammals: A molecular toxicological perspective. Appl. Organomet. Chem. 16, 701-707. (42) Gailer, J., George, G. N., Pickering, I. J., Prince, R. C., Younis, H. S., and Winzerling, J. J. (2002) Biliary excretion of the seleno-bis (S-glutathionyl) arsinium ion after intravenous injection of rabbits with arsenite and selenate. Chem. Res. Toxicol. 15, 1466-1471. (43) Gailer, J., George, G. N., Pickering, I. J., Madden, S., Prince, R. C., Yu, E. Y., Denton, M. B., Younis, H. S., and Aposhian, H. V. (2000) Structural basis of the antagonism between inorganic mercury and selenium in mammals. Chem. Res. Toxicol. 13, 1135-1142. (44) Lash, L. H., and Jones, D. P. (1985) Distribution of oxidized and reduced forms of glutathione and cysteine in rat plasma. Arch. Biochem. Biophys. 240, 583-592. (45) Block, E., Glass, R. S., Jacobsen, N. E., Johnson, S., Kahakachchi, C., Kaminski, R., Skowronska, A., Boakye, H. T., Tyson, J. F., and Uden, P. C. (2004) Identification and synthesis of a novel seleniumsulfur amino acd found in selenized yeast: Rapid indirect detection NMR methods for characterizing low-level organoselenium compounds in complex matrices. J. Agric. Food Chem. 52, 3761-3771. (46) Hsieh, H. S., and Ganther, H. E. (1975) Acid-volatile selenium formation catalyzed by glutathione reductase. Biochemistry 14, 16321636. (47) Kondo, T., Dale, G. L., and Beutler, E. (1995) Thiol transport from human red blood cells. Methods Enzymol. 252, 72-82. (48) Board, P., Nishida, T., Gatmaitan, Z., Che, M., and Arias, I. M. (1992) Erythrocyte membrane transport of glutathione conjugates and oxidized glutathione in the Dubin-Johnson syndrome and in rats with hereditary hyperbilirubinemia. Hepatology 15, 722-725. (49) Srivastava, S. K., and Beutler, E. (1969) The transport of oxidized glutathione from human erythrocytes. J. Biol. Chem. 244, 9-16. (50) Zalups, R. K., and Lash, L. H. (1996) Interactions between glutathione and mercury in the kidney, liver and blood. In Toxicology of Metals (Chang, L. W., Ed.) pp 145-163, CRC Lewis Publishers, Boca Raton.

Seleno Bis(S-glutathionyl) Arsinium Ion (51) Michalke, B. (2002) Selenspeziation mit SAX/ICP-MS und RPLCICP-MS. In Moderne Techniken der Ionenanalyse (Jensen, K. F. u. D., Ed.) pp 50-60, ECOMED Verlagsgesellschaft AG+Co. KG, Landsberg, Germany. (52) Suzuki, K. T. (2002) Speciation of arsenic in body fluids. Talanta 58, 111-119.

Chem. Res. Toxicol., Vol. 19, No. 4, 2006 607 (53) Biswas, S., Talukder, G., and Sharma, A. (1999) Prevention of cytotoxic effects of arsenic by short-term dietary supplementation with selenium in mice in vivo. Mutat. Res. 441, 155-160.

TX0503505