Effect of arsenosugar ingestion on urinary arsenic speciation

^a Author for correspondence. Fax 403-492-0364; e-mail xc.le{at}ualberta.ca.

	Abstract

Top Abstract Introduction Materials and Methods Results and Discussion References

 
We developed and evaluated a method for the determination of µg/L concentrations of individual arsenic species in urine samples. We have mainly studied arsenite [As(III)], arsenate [As(V)], monomethylarsonic acid (MMAA), and dimethylarsinic acid (DMAA) because these are the most commonly used biomarkers of exposure by the general population to inorganic arsenic and because of concerns over these arsenic species on their toxicity and carcinogenicity. We have also detected five unidentified urinary arsenic species resulting from the metabolism of arsenosugars. We combined ion pair liquid chromatography with on-line hydride generation and subsequent atomic fluorescence detection (HPLC/HGAFS). Detection limits, determined as three times the standard deviation of the baseline noise, are 0.8, 1.2, 0.7, and 1.0 µ/L arsenic for arsenite, arsenate, MMAA, and DMAA, respectively. These correspond to 16, 24, 14, and 20 pg of arsenic, respectively, for a 20-µL sample injected for analysis. The excellent detection limit enabled us to determine trace concentrations of arsenic species in urine samples from healthy subjects who did not have excess exposure to arsenic. There was no need for any sample pretreatment step. We used Standard Reference Materials, containing both normal and increased concentrations of arsenic, to validate the method. Interlaboratory studies with independent techniques also confirmed the results obtained with the HPLC/HGAFS method. We demonstrated an application of the method to the determination of arsenic species in urine samples after the ingestion of seaweed by four volunteers. We observed substantial increases of DMAA concentrations in the samples collected from the volunteers after the consumption of seaweed. The increase of urinary DMAA concentration is due to the metabolism of arsenosugars that are present in the seaweed. Our results suggest that the commonly used biomarkers of exposure to inorganic arsenic, based on the measurement of arsenite, arsenate, MMAA, and DMAA, are not reliable when arsenosugars are ingested from the diet.

	Introduction

Top Abstract Introduction Materials and Methods Results and Discussion References

 
Ingestion from food and water is the major route of exposure to arsenic by the general population (1)(2)(3)(4). After ingestion, arsenic compounds can be readily absorbed through the gastrointestinal tract into the bloodstream (5)(6)(7)(8). They are subsequently metabolized in the human body (9)(10)(11)(12)(13)(14)(15)(16)(17)(18). Both parent arsenic compounds and their metabolites are further excreted into the urine (9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24). The extent of metabolism and excretion depends on chemical forms of arsenic ingested. Nonetheless, urinary excretion is the major pathway for the elimination of arsenic compounds from the body (9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23). Arsenic concentrations in urine have a short half-life and reflect recent exposure. Therefore, quantitative determination of arsenic species in human urine samples has been commonly used as a biomarker of recent exposure to arsenic. Arsenic in hair is a useful indicator for longer-term exposure because arsenic is believed to accumulate in hair and fingernails because of the high content of keratin.

Arsenite [As(III)], arsenate [As(V)], monomethylarsonic acid (MMAA), and dimethylarsinic acid (DMAA) are widely present in the natural environment (25)(26)(27)(28)(29).¹ The toxicity varies dramatically with different chemical forms of arsenic; the median lethal dose (LD₅₀) values in rats for some arsenic compounds (29)(30)(31) are (in mg/kg): potassium arsenite 14, calcium arsenate 20, MMAA 700–1800, DMAA 700–2600, and arsenobetaine >10 000. Therefore, it is essential to identify and quantify individual chemical forms of arsenic (i.e., chemical speciation) to assess health risks associated with arsenic exposure.

Although several analytical techniques, such as HPLC with inductively coupled plasma mass spectrometry (ICPMS) detection, are capable of quantifying arsenic species at µg/L concentrations, these methods are expensive and consequently are not suitable for routine analysis of large numbers of samples.

Furthermore, ingestion of arsenic from all sources contributes to the urinary arsenic concentration. To obtain a reliable assessment of arsenic exposure from a particular source, e.g., drinking water, one must ensure that the ingestion of arsenic from other sources can be differentiated and confounding factors can be identified.

Most seafoods contain µg/g (ppm) concentrations of arsenic, arsenobetaine being the major arsenic species in crustaceans and arsenosugars in seaweeds (26)(27)(28)(32)(33)(34). Both arsenobetaine and arsenosugars have been found in bivalves (35)(36)(37). Other organoarsenicals, such as arsenocholine, tetramethylarsonium ion, trimethylarsine oxide, and trimethylarsine, have also been reported to be present in some seafoods at much lower concentations than arsenobetaine and arsenosugars (26). The ingestion of seafood could cause a considerable increase in the urinary concentration of arsenic. Fortunately, arsenobetaine is very stable and is rapidly excreted unchanged into the urine. It is not metabolized in the body and it does not form an arsine with sodium borohydride with the conventional hydride generation method (38)(39)(40)(41). Therefore, the ingestion of arsenobetaine and inorganic arsenic can be differentiated through the speciation of arsenic in urine.

Although the fate and behavior of arsenobetaine are well known, the metabolism of arsenosugars is not well understood and the effect of arsenosugar ingestion on urinary arsenic excretion is not widely recognized. Our previous studies have shown that arsenosugars were metabolized in the human body (18)(24). The ingestion of arsenosugars from the diet and the excretion of metabolites may affect the use of urinary arsenic as an indicator of exposure to inorganic arsenic. The objectives of the present study are: (a) to develop simple and inexpensive techniques for arsenic speciation; (b) to apply the techniques to the determination of arsenic species in human urine samples after the ingestion of arsenosugar-containing seafood; and (c) to examine possible confounding factors in assessing inorganic arsenic exposure when arsenosugars are ingested from the diet.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results and Discussion References

 
apparatus
For the speciation of individual arsenic compounds, we use HPLC separation followed by hydride generation and atomic fluorescence (HGAFS) detection. This combination takes advantages of both the separation power offered by HPLC and good sensitivity obtainable by using HGAFS. Our HPLC/HGAFS system is schematically shown in Fig. 1 . Effluent from the HPLC column directly meets at two T-joints, T1 and T2, with continuous flows of hydrochloric acid (A) and sodium borohydride (B) introduced by using a peristaltic pump. Upon mixing the HPLC effluent and acid and borohydride solutions, hydride generation takes place. We used a 50-cm reaction coil (RC) (0.8 mm i.d.) to ensure a complete reaction. Hydride generated from the reaction is separated from liquid waste in the gas/liquid separator apparatus and carried by a continuous flow of argon carrier gas (Ar) to the atomic fluorescence detector (AFS).

View larger version (18K): [in this window] [in a new window]   Figure 1. A schematic showing the HPLC/HGAFS technique for the speciation of arsenic. A, acid (3 mol/L HCl); Ar, argon carrier gas; B, 13 g/L NaBH4 in 0.1 mol/L NaOH; C and H, column and column heater; IN, sample injector; M, mobile phase; RC, reaction coil (50 cm long and 0.8 mm i.d.); T1 and T2, T-joints.

We used a commercial AFS (Model Excalibur 10.003, P.S. Analytical) for the detection of arsenic. The AFS detector consisted of an excitation source, a hydrogen diffusion flame, fluorescence collection optics, and a photomultiplier tube (PMT) (42). A boosted discharge arsenic hollow cathode lamp was used to excite arsenic atoms in the flame. Fluorescence radiation was collected at a right angle and detected with a solar blind PMT. An interference filter was used to reduce the background radiation. The hydrogen, produced as a hydride generation byproduct, was sufficient to maintain hydrogen diffusion flame. A Pentium computer with Varian Star Workstation ADC board and software was used to record and process signals from the AFS detector. A Hewlett Packard 3390A integrator with both peak area and peak height measurement capability was also used to record chromatograms.

A HPLC system consisted of a Gilson HPLC pump (Model 307) with a 5 mL/min stainless steel pump head, a Rheodyne 6-port sample injector (Model 7725i) with a 20-µL sample loop, and a HPLC column. A reversed-phase C18 column (250 x 4.6 mm, 5-µm particles) from Phenomenex was used for the separation. The analytical column and a guard column (30-mm long) packed with the same material were mounted inside a column heater (Model CH-30, Eppendorf) that was equipped with a temperature controller (Model TC-50, Eppendorf). Mobile phase was preheated to the temperature of the column by using a precolumn coil of 50-cm stainless steel capillary tubing, which was also placed inside the column heater. The temperature controller, according to the manufacturer, was able to provide a ±0.1 °C temperature stability and ±1 °C accuracy. Isocratic HPLC operation was performed under 1 mL/min flow rate. The detailed HPLC separation conditions are shown in Table 1 , method 1. For the determination of non-hydride-forming arsenic species, such as arsenosugars, we performed on-line microwave-assisted decomposition. We used a solution containing 0.1 mol/L potassium persulfate and 0.3 mol/L sodium hydroxide to decompose organoarsenicals to arsenate with the aid of microwave heating. The HPLC effluent (1 mL/min) and the decomposition reagent (4 mL/min) met at a T-joint. This solution mixture flowed through a polytetrafluoroethylene (PTFE) decomposition coil (3 m x 0.5 mm i.d.) located in a continuously operating microwave oven (650 W, 2450 MHz, General Electronics), where the decomposition took place. The solution from the microwave oven then met the continuous flows of acid and borohydride at two T-joints. Arsines produced upon the hydride generation were detected with the AFS detector as described above.

Another HPLC system consisted of a Model 510 solvent delivery pump (Waters), a U6K injector (Waters), and a Hamilton PRP X-100 anion exchange column (250 mm x 4.1 mm) or a GL Sciences Inertsil ODS-2 column (250 mm x 4.6 mm). HPLC separation conditions are shown in Table 1 , methods 2 and 3. An ICPMS was used for HPLC detection as described elsewhere (23)(40). Briefly, a VG PlasmaQuad 2 Turbo Plus ICPMS (VG Elemental, Fisons Instrument) equipped with a SX300 quadrupole mass analyzer, a standard ICP torch, and a conventional concentric nebulizer was used. The sampling position and ion lens voltages were optimized with respect to signal-to-noise ratio at m/z 75 by introducing a solution containing 30 µg/L arsenite in 10 mL/L nitric acid. The quadrupole mass analyzer was operated in the single ion monitoring mode. The instrumental operating conditions are the same as described previously (18)(23)(40). A PTFE tubing (20 cm x 0.4 mm i.d.) with appropriate fittings was used to connect the outlet of the HPLC column directly to the inlet of the ICP nebulizer. Signals at m/z 75 were monitored with a multichannel analyzer and data were automatically transferred to and stored in the VG data system. Chromatograms were plotted on an Epson FX-850 printer.

reagents
Deionized water from a Maxima ultrapure water system (Elga) was used for the preparation and dilution of all reagents, samples, and calibrators. Calibrator solutions of arsenite, arsenate, MMAA, and DMAA were prepared by appropriate dilution with deionized water from 1000 mg/L stock solutions, as described previously (18)(22)(39)(40). Calibrator solutions containing >1 mg/L arsenic were stable for several months. Calibrator solutions containing <10 µg/L arsenic were prepared fresh daily by serial dilution with deionized water from 1 mg/L arsenic calibrator solutions. Arsenic concentration in the stock solutions was calibrated against an atomic absorption arsenic calibrator solution containing 1000.0 mg/L arsenic (Aldrich) by using direct ICPMS analysis.

A Standard Reference Material, Toxic Metals in Freeze-Dried Urine SRM 2670, was obtained from NIST. The freeze-dried urine was reconstituted by the addition of 20.0 mL of deionized water as recommended by NIST. For two bottles containing normal concentrations of toxic metals, the concentration of arsenic is not certified and a reference value of 60 µg/L has been provided for information. Two other bottles contained increased concentrations of toxic metals, which were prepared by adding toxic metals to human urine. The certified value for total arsenic concentration in the increased-concentration SRM is 480 ± 100 µg/L. The concentration and metabolite pattern in this SRM are different from those in urine samples from the general population. Thus, this SRM is not ideal for method validation purposes. However, in the absence of alternative suitable reference materials, we chose this SRM for the present study.

The reagents used in HPLC mobile phases, including tetrabutylammonium hydroxide, malonic acid, NaH₂PO₄, and Na₂HPO₄, were obtained from Aldrich. HPLC-grade methanol was from Fisher. These mobile-phase solutions were prepared in deionized water and filtered through a 0.45-µm membrane before use. Sodium borohydride (Aldrich) solutions in 0.1 mol/L sodium hydroxide (Fisher) were prepared fresh daily. All reagents used were of analytical grade or better.

seaweed and urine samples
Commercial seaweed products, nori and yakinori, were purchased from a local supermarket in Edmonton, Canada. A subsample of the seaweed (2–5 g dry weight) was extracted by using a procedure similar to that described by Shibata and Morita (35). The sample was weighed into a test tube to which was added 20 mL of a methanol:water mixture (1:1 by vol). The tube was sonicated for 20 min. After centrifugation, the extract was removed and placed in a 150-mL beaker. The extraction process with the aid of sonication was repeated a further four times. The extracts were combined in the beaker, evaporated to dryness, and the residue dissolved in 10 mL of deionized water. After filtration through a 0.45-µm nylon membrane, the sample was subjected to HPLC/HGAFS analyses.

Four volunteers (34-, 35-, and 62-year-old men and a 56-year-old woman) refrained from eating any seafood for at least 72 h before commencing the seaweed ingestion experiment. Each volunteer collected at least one urine sample during the 12-h period before the consumption of seaweed. These samples were used to determine the background concentration of arsenic species in the urine resulting from a regular diet that excluded any seafood.

The volunteers then consumed 10 g (dry weight) of yakinori in one meal. The time of this meal was referred to as time zero. Six hours later, each volunteer consumed another portion of 10 g of yakinori. All urine was completely collected in separate 500-mL polyethylene containers for three consecutive days. The volunteers did not eat any other seafood during the experiment period. The urine samples were stored at 4 °C and were analyzed for arsenic speciation within 4 days.

First morning urine samples were also obtained from two other volunteers (34- and 41-year-old men) who refrained from eating any seafood for 72 h before the sampling.

The volunteers were aware of the experimental details and possible health effects concerning the ingestion of seaweed in this experiment. All procedures followed were in accordance with the ethical guidelines of the Research Ethics Board, Faculty of Medicine, University of Alberta.

determination of creatinine in urine samples
Creatinine in urine samples was determined by HPLC with UV/Vis absorption spectrophotometric detection, as described previously (23). Urine samples were diluted 50 times with deionized water and a 10-µL aliquot was injected onto a C18 column (Bondclone C18, 3.9 x 300 mm, Phenomenex). Sodium acetate (50 mmol/L, pH 6.5) in 98:2 (by vol) water:acetonitrile was used as the mobile phase with a flow rate of 1.0 mL/min. A system consisting of a Dionex Gradient Pump DX300, a Waters 712 WISP Autosampler, and a Waters 484 Tunable Absorbance Detector was used. Absorbance at 254 nm was measured and peak area was used for the quantification of creatinine.

	Results and Discussion

Top Abstract Introduction Materials and Methods Results and Discussion References

 
detection limit and calibration
Figure 2 shows four chromatograms obtained from the HPLC/HGAFS analyses of 2, 10, 20, and 50 µg/L arsenic species, respectively. Only 20 µL of sample is injected for each analysis. Detection limits, defined as the analyte concentration that produces a chromatographic peak having a height equal to three times the standard deviation of the baseline noise (43)(44), are 0.8, 1.2, 0.7, and 1.0 µg/L arsenic for arsenite, arsenate, MMAA, and DMAA, respectively. These correspond to 16, 24, 14, and 20 pg of arsenic, respectively, for a 20-µL sample injected for analysis. These detection limits are comparable with those obtained by using the more expensive HPLC/ICPMS techniques (45)(46)(47)(48)(49)(50). These excellent detection limits are achieved through the combination of AFS detection with hydride generation.

View larger version (23K): [in this window] [in a new window]   Figure 2. Chromatograms obtained from the HPLC/HGAFS analyses of 2, 10, 20, and 50 µg/L arsenite [As(III)], arsenate [As(V)], MMAA, and DMAA. A Phenomenex ODS-3 reversed-phase column (250 x 4.6 mm, 5-µm particle size) was used with a temperature maintained at 70 °C. The mobile phase contained 5 mmol/L tetrabutylammonium chloride, 4 mmol/L malonic acid, and 50 mL/L methanol (pH 6.0), with a flow rate of 1.0 mL/min. HGAFS was used for detection (method 1).

One of the most attractive features of fluorescence methods is their inherent high sensitivity. Detection limits on the order of a single molecule with laser-induced fluorescence detection for capillary electrophoresis have provided evidence for the excellent capability of fluorescence detection (51)(52)(53). Despite the success of molecular fluorescence, AFS has not shown the same advantage. This has been primarily due to the interference effects that occur in AFS when real samples are analyzed. Light scattering and background due to sample matrix are major problems. However, these problems are solved by separating analytes of interest (arsenic in the present study) from sample matrix through a hydride generation process.

Hydride generation is a chemical derivatization process that produces volatile hydrides upon chemical treatment of a sample with a reducing agent, typically sodium borohydride. Hydride generation techniques coupled with atomic absorption, atomic emission, and mass spectrometry have found wide application in the determination of trace amounts of several elements, including arsenic. As an efficient sample introduction method, hydride generation enhances sensitivity normally by 10- to 100-fold over the more commonly used liquid sample nebulization procedure. Also, the analyte element can be separated from almost all other accompanying materials in the sample through the hydride generation process. Only gaseous hydrides are introduced to the detector and sample matrix is left in the liquid waste. Thus spectral and chemical interferences encountered in the detection system are essentially eliminated. This is particularly beneficial to AFS detection where the interference had previously been the major problem. Without scattering and background interference from sample matrix, the detection limit by using AFS is dramatically improved. Our results are consistent with those of Stockwell and Corns (42), who have demonstrated that AFS detection improves the sensitivity by two orders of magnitude over absorption techniques.

The chromatograms shown in Fig. 2 also illustrate a linear increase of peak intensity with the increase of arsenic concentration. Calibration curves for each of the arsenic species, obtained from the determination of 0, 1, 2, 5, 10, 20, 50, 100, 150, and 200 µg/L arsenic, are linear, with linear regression coefficients (r) of better than 0.99 for all four arsenic species. Furthermore, these chromatograms demonstrate the reproducibility of chromatographic retention time for each of the four arsenic species.

recovery and interference
The stability of retention time is essential to a reliable identification of analytes because the identity of chromatographic peaks is usually obtained by comparing the retention time of the analyte in the sample with those of the calibrators. We further confirm the peak identity by an analysis of the coinjected arsenic calibrator compound and the sample. Coelution of an analyte and the arsenic calibrator is evidence that the sample contains that arsenic compound.

Figure 3 shows chromatograms from the HPLC/HGAFS analysis of a urine sample (top) and the urine sample supplemented with 10 µg/L each of the four target arsenic species (bottom). The identical retention time of arsenic species in the sample and in the calibrator solution suggests that the urine sample matrix does not cause interference on the chromatographic separation.

View larger version (19K): [in this window] [in a new window]   Figure 3. Chromatograms obtained from HPLC/HGAFS analyses of a urine sample (top) and the urine sample supplemented with 10 µg/L of the four arsenic species (bottom). The urine sample was obtained from a 35-year-old male volunteer who did not have excess exposure to arsenic. Same HPLC/HGAFS conditions as shown in Fig. 2 were used.

We also obtained the recovery of arsenic species by comparing the amount of arsenic added to the sample and the amount detected. We used this approach to check whether there was any interference with quantitative determination. The urine sample shown in Fig. 3 was obtained from a 35-year-old male volunteer who did not have excess exposure to arsenic. The concentration of arsenic species in this urine sample is: As(III) 2.6 µg/L, DMAA 16.2 µg/L, and MMAA 1.6 µg/L. The recovery of the arsenic species that are added to the urine sample is in the range of 85–100%, indicating that there is no significant interference.

We have previously experienced that chromatography retention time of arsenic species in the sample may be different from that in calibrator solutions when a mobile phase is not suitable for ion pair chromatography. This can cause a problem in identifying analyte species, which is usually based on a match of retention time of the analyte with that of the calibrator. We found that appropriate buffer concentration and organic solvent are required in ion pair chromatographic separation of arsenic species. In addition to tetrabutylammonium ion pair reagent, we used malonic acid as a buffer and 50 mL/L methanol as an organic modifier. Using this mobile phase, we were able to analyze undiluted urine samples directly without interference from the sample matrix. Eliminating malonic acid or lowering the methanol concentration has resulted in interference from the urine sample matrix.

This mobile phase may not be suitable for HPLC techniques with ICPMS detection because of the high content of methanol. The incomplete combustion of organic solvent results in a blockage of the sampling cone of the interface between the inductively coupled plasma and the mass spectrometer (45)(46)(48). This in turn leads to inaccurate measurement due to a downward drifting problem. Thus, only <1% of organic solvent is commonly used in ICPMS protocols. When this lower concentration of organic modifier is used in ion pair chromatography, interference from complex sample matrix can be a problem. Therefore, many HPLC/ICPMS techniques for arsenic speciation in urine require a dilution of urine samples before analysis. However, a dilution of sample is not desirable when the concentration of the analyte in the sample is already very low, usually approaching the detection limit. Our technique is not prone to interference and does not require the dilution of urine sample. Direct analysis of urine samples without dilution enables us to maintain a good detection limit of arsenic in actual urine samples.

srm
We use SRMs (from NIST) to validate our results and to eliminate any systematic errors. Urine SRM 2670 consists of two sets of urine samples, one containing an increased concentration of arsenic and the other containing a normal concentration of arsenic. The certified total arsenic concentration in the increased-concentration urine is 480 ± 100 µg/L. Arsenic in the normal-concentration urine is not certified; a reference value of 60 µg/L has been provided by NIST for information. Our eight replicate analyses of these SRMs give arsenic concentrations of 460 ± 10 and 56 ± 3 µg/L, respectively, which are in good agreement with the certified and reference values.

There is no certified value available from any of the SRMs on arsenic speciation. This is primarily because of a lack of analytical methods that could provide the necessary detection limit and specificity to quantify trace amounts of individual arsenic species. As a consequence, most published work on the speciation of arsenic in SRM 2670 have only dealt with the increased-concentration urine. Our HPLC/HGAFS method has an excellent detection limit that is sufficient to determine MMAA and DMAA in the normal-concentration urine SRM. Fig. 4 shows chromatograms obtained from replicate HPLC/HGAFS analyses of urine SRM 2670 (both normal concentration and increased concentration of arsenic). These results demonstrate that both the retention time and the intensity of the arsenic peaks are reproducible between replicate analyses. DMAA (48 ± 2 µg/L) and MMAA (7.4 ± 0.7 µg/L) are the only arsenic species present at detectable concentrations in the normal-concentration urine sample, and the ratio of these arsenic metabolites is 6.5:1 (DMAA:MMAA). Similar concentrations of DMAA (49 ± 2 µg/L) and MMAA (8.1 ± 0.7 µg/L) are also detected in the increased-concentration urine SRM, with a similar ratio of DMAA:MMAA (6:1) to that of the normal-concentration urine sample. A much higher concentration of As(V) (403 ± 8 µg/L) is also present in the increased-concentration urine sample. This reflects the amount of As(V) added to the normal urine in the preparation of this SRM. The concentration of individual arsenic species from eight replicate determinations is summarized in Table 2 . The SRM containing normal concentrations of trace metals resembles typical human urine samples obtained from the general population that do not involve excess exposure to arsenic. These results clearly demonstrate the applicability of the method to the speciation of arsenic in urine samples from the general population.

View larger version (27K): [in this window] [in a new window]   Figure 4. Chromatograms obtained from replicate HPLC/HGAFS analyses of SRM 2670 normal-concentration (a) and increased-concentration (b) urines. Same conditions as shown in Fig. 2 were used.

interlaboratory comparison
Another laboratory has independently used anion-exchange chromatography with ICPMS detection to confirm the results obtained with HPLC/HGAFS. A typical chromatogram from the analysis of SRM 2670 urine containing the increased concentration of arsenic is shown in Fig. 5 . The order in which arsenic species are eluted off the anion exchange column (Fig. 5 ) is the same as those from a reversed-phase column (Fig. 4 ). This is expected because we use tetrabutylammonium hydroxide as an anion pairing reagent in the mobile phase. The anion pairing process results in similar effects to anion exchange for the separation of the four arsenic species. The advantage of using ion pair chromatography compared with anion-exchange chromatography is that other neutral and cationic arsenic species can also be separated by ion pair chromatography (23)(24)(35)(45)(47).

View larger version (15K): [in this window] [in a new window]   Figure 5. A chromatogram obtained from the HPLC/ICPMS analysis of SRM 2670 increased-concentration urine. A Hamilton PRP X-100 (250 x 4.1 mm) anion exchange column was used for separation, which was carried out at ambient temperature (method 3). The mobile phase contained 30 mmol/L H2PO4-/HPO42- buffer (pH 6, adjusted with NH4OH) and its flow rate was 1.3 mL/min. ICPMS was used to monitor arsenic at m/z 75. AsB, arsenobetaine.

Consistent with HPLC/HGAFS results (Fig. 4 ), the HPLC/ICPMS analysis (Fig. 5 ) of the SRM increased-concentration urine sample also reveals the presence of DMAA, MMAA, and As(V), with As(V) being the major arsenic species in the sample. The concentration of these samples, obtained by the HPLC/ICPMS method, are 49 ± 3 (DMAA), 7 ± 1 (MMAA), 443 ± 20 [As(V)], 15 ± 3 (arsenobetaine), and 514 ± 23 µg/L (total arsenic). A difference from HPLC/HGAFS analysis is that the HPLC/ICPMS technique also detects arsenobetaine in the sample. The concentration of arsenobetaine detected is 15 ± 3 µg/L As, which is in good agreement with the value (11 ± 3) determined by HPLC with electrospray mass spectrometry (54). Arsenobetaine is a predominant arsenic species present in many marine organisms (26)(27)(28). Consumption of arsenobetaine-containing seafood results in the excretion of arsenobetaine into the urine.

Under the given mobile-phase conditions (pH 5.9–6.0), arsenobetaine is present as a zwitterion [(CH₃)₃AsCH₂COO^-] and arsenite is present as a neutral species [As(OH)₃]. These two species do not retain on the anion-exchange column and they coelute at the void volume.

Although anion-exchange chromatography with ICPMS detection is unable to differentiate arsenobetaine from As(III), the HPLC/HGAFS technique has the ability to differentiate these species by two different approaches. First, the ion pair chromatography used in the HPLC/HGAFS technique is able to resolve arsenobetaine from arsenite (23)(24)(40). Second, arsenobetaine does not form hydride and therefore is not detected directly. It can be detected by using an on-line microwave digestion procedure as we have described previously (24)(40).

speciation of arsenic in urine samples
We have applied the HPLC/HGAFS technique to the determination of arsenic speciation in urine samples from six volunteers. The volunteers stopped eating any seafood for 3 days before collecting a first morning urine sample. The concentrations of arsenic species in these samples, obtained from eight replicate HPLC/HGAFS analyses, are summarized in Table 2 . DMAA is the major arsenic species in the urine samples from all volunteers, and the arsenic concentration in these samples is within the range reported by others for the general population. Our results demonstrate that the HPLC/HGAFS technique is capable of determining arsenic speciation in urine samples from the general population. Epidemiology studies involving a survey of larger numbers of urine samples from people who are exposed to different concentrations of arsenic may help to establish a dose/response relation for health risk assessment.

urinary arsenic species after ingestion of arsenosugars
Figure 6 shows chromatograms obtained from HPLC/HGAFS analyses of a yakinori extract (a) and two urine samples collected from volunteer one (34-year-old man) 29 h after (b) and 15 h before (c) the consumption of yakinori, respectively. The seaweed preparation, yakinori, contains two arsenosugars (X and XI) as the major arsenic species (Fig. 6a ). After the consumption of 10 g (dry weight) of yakinori, the arsenosugars in the original yakinori are not detected in the urine samples. However, two new metabolites are found in the urine sample collected 20–33 h after the ingestion of yakinori (Fig. 6b ). These metabolites are not found in urine samples collected before the ingestion of yakinori (Fig. 6c ).

View larger version (14K): [in this window] [in a new window]   Figure 6. Chromatograms obtained from the analyses of yakinori extract (a) and urine samples collected 29 h after (b) and 15 h before (c) the consumption of 10 g of yakinori by volunteer one (male, age 34 years). HGAFS with microwave digestion was used for detection (method 2). A Phenomenex ODS-3 column (4.6 x 250 mm) was used for the separation, and the column temperature was 70 °C. The two major arsenic species in the seaweed are two arsenosugars (X and XI). Peaks marked * are two new arsenic metabolites.

We have further confirmed the metabolism of arsenosugars by HPLC/ICPMS analyses of urine samples collected from another volunteer (62-year-old man) before and after the ingestion of 9.5 g of a commercial seaweed product nori. HPLC/ICPMS analyses of an extract of the nori reveals that an arsenosugar (X) is the major arsenic species in the nori sample (Fig. 7 c) and its content is approximately 21 µg/g (dry weight). Other forms of arsenic in the nori are approximately 0.7 µg/g (dry weight). Urine samples collected encompassing the ingestion of nori were analyzed for arsenic speciation. Chromatograms obtained from urine samples collected 13 h before (Fig. 7a ) and 13.5 h after (Fig. 7b ) the volunteer ate 9.5 g of nori are compared with that of an extract of the nori (Fig. 7c ). It clearly shows that the arsenosugar is not present in the urine samples. Instead, five new arsenic compounds, at retention times of 5.7, 6.6, 8.2, 9.9, and 14.7 min, are present in the urine sample collected 13.5 h after the ingestion of nori (Fig. 7b ). These compounds do not correspond to the arsenosugar present in nori on the basis of their retention times (Fig. 7c ). The identity of these new metabolites from the arsenosugar are not yet clear. However, they are arsenic-containing compounds because the ICPMS and HGAFS were tuned to selectively detect arsenic. The retention times of these metabolites do not correspond to those of a dozen arsenic calibrators currently available to us. The chromatographic behavior of the metabolites is also significantly different from that of the original arsenosugar present in the seaweed. Coinjection of the urine sample and the seaweed extract onto HPLC for analysis showed that the metabolites were distinct from the original arsenosugar.

View larger version (22K): [in this window] [in a new window]   Figure 7. HPLC/ICPMS chromatograms of a nori extract (c) and of urine samples from volunteer three (male, age 62 years) 13 h before (a) and 13.5 h after (b) the ingestion of 9.5 g of nori. An Inertsil ODS-2 column (4.6 x 250 mm) was used for the separation, under ambient temperature (method 4). The major arsenic species in the nori is an arsenosugar (X). Peaks marked * are five new arsenic metabolites.

The most commonly used biomarkers of exposure (or internal dose) to inorganic arsenic are based on the measurement of urinary arsenite, arsenate, MMAA, and DMAA because urinary excretion is the main pathway for the elimination of arsenic. These biomarkers have also been claimed to "exclude other ingested forms of arsenic which are much less toxic" (55). However, this statement is valid only if "other ingested forms of arsenic" are not metabolized, such as arsenobetaine. Ingestion of arsenobetaine, which is the predominant arsenic species in crustacean seafood, results in urinary excretion of this species unmodified. The presence of arsenobetaine does not affect the determination of inorganic arsenic, MMAA, and DMAA when appropriate speciation techniques are used. Therefore, ingestion of crustacean seafood does not affect the validity of the traditional biomarker of exposure to inorganic arsenic.

However, in the case of the ingestion of arsenosugar-containing food, arsenosugars are metabolized. It is important to determine whether the ingestion of arsenosugars affects the urinary concentration of arsenite, arsenate, MMAA, and DMAA. For this reason, we have determined the concentration of these four hydride-forming arsenic species in urine samples collected from four volunteers encompassing the ingestion of yakinori. The four volunteers (34-, 35- and 62-year-old men, and a 56-year-old woman) repeatedly ingested two portions (10 g each) of yakinori, and the urine samples collected for four consecutive days were analyzed for arsenic by flow injection analysis (FIA)/HGAFS. The time of the first ingestion was referred to as time zero, and the second ingestion was 6 h after the first ingestion. The concentrations of these arsenicals in urine, obtained by FIA/HGAFS, were normalized against urinary creatinine concentration and are shown in Fig. 8 . The relative concentration of arsenic over creatinine in the urine is a measure in which the uncertainty due to the volume change is taken into account. Therefore, the normalized results in Fig. 8 roughly illustrate the rate of excretion of the arsenic compounds. The excretion rate of arsenic after the ingestion of arsenosugars is much slower than that for the ingestion of arsenobetaine (18)(19)(22). This is understandable because arsenobetaine is excreted unchanged (18)(19)(20)(21)(22)(23), whereas arsenosugars are metabolized in the body.

View larger version (29K): [in this window] [in a new window]   Figure 8. Urinary arsenic/creatinine concentrations of four volunteers before and after the ingestion of 9.5 g of nori. Determination of arsenic was carried out by flow injection HGAFS. Volunteers: 1, , male, age 34 years; 2, , male, 35; 3, , male, 62; and 4, , female, 56.

Figure 8 also demonstrates that urinary arsenic excretion patterns are different for the four individuals who ingested the same amounts of seaweed products. The reasons for this interindividual variability are not yet clear. However, we have observed the substantial increases of urinary arsenic concentrations after these volunteers ingested the seaweed, which contains two major arsenosugars (Fig. 6a ). These results imply that the use of urinary concentrations of arsenite, arsenate, MMAA, and DMAA as biomarkers to assess inorganic arsenic exposure is not valid when arsenosugar ingestion is also involved.

Figure 9 shows chromatograms from the analyses of several urine samples collected from the four volunteers before and after the ingestion of the seaweed yakinori. DMAA is the major species in most of these urine samples collected after the consumption of seaweed. As much as 90 µg/L DMAA is present in urine samples collected 26.5 h after volunteer two (35-year-old man) ingested 10 g of the seaweed yakinori, compared with <20 µg/L in the samples collected before the ingestion. The seaweed products used do not contain high concentrations of DMAA (Fig. 6a and Fig. 7c ). The increased concentration of DMAA in urine samples collected after the ingestion of seaweed is a result of the metabolism of arsenosugars.

View larger version (35K): [in this window] [in a new window]   Figure 9. Chromatograms obtained from HPLC/HGAFS analyses of urine samples collected from four volunteers before and after the consumption of seaweed yakinori. A Phenomenex ODS-3 reversed-phase column (250 x 4.6 mm, 5-µm particle size) was used with a temperature maintained at 70 °C. The mobile phase contained 5 mmol/L tetrabutylammonium chloride, 4 mmol/L malonic acid, and 50 mL/L methanol (pH 5.9), with a flow rate of 1.0 mL/min. (a) Volunteer 1, male, age 34 years; (b) volunteer 2, male, 35; (c) volunteer 3, male, 62; (d) volunteer 4, female, 56.

Therefore, when arsenosugar-containing foods are ingested, the commonly used biomarkers of exposure to arsenic are not reliable. Arsenosugars are abundant in seaweeds, mussels, clams, and oysters (35)(36)(37), which are common for human consumption. The ingestion of arsenosugars in the human diet will invalidate the use of the conventional biomarkers of exposure to inorganic arsenic. Arsenic species from seafood sources are usually eliminated from the body within 3 days after ingestion (18)(19)(23)(38). For this reason, subjects should not eat any seafood for at least three days before urine samples are taken for the assessment of exposure to inorganic arsenic. In addition, speciation information of the metabolites may be useful as a potential marker of exposure to arsenosugars. If arsenosugar metabolites are present in a urine sample, any increased concentration of DMAA in the urine should not be attributed entirely to the exposure to inorganic arsenic. One needs to consider the possible contribution from arsenosugar metabolism to urinary DMAA concentration.

	Acknowledgments

 
This work was supported in part by the American Water Works Association Research Foundation. We gratefully acknowledge that the Foundation is the joint owner of the technical information upon which this manuscript is based. We thank the Foundation for its financial, technical, and administrative assistance in funding and managing the project through which this information was discovered. Without the Foundation's support and encouragement, this publication would not be possible. We also thank Joerg Feldmann of the University of British Columbia for providing the interlaboratory comparison results shown in Fig. 5 .

	Footnotes

 
Environmental Health Sciences Program, Department of Public Health Sciences, 13–103 CSB, Faculty of Medicine, University of Alberta, Edmonton, AB T6G 2G3, Canada.

¹ Nonstandard abbreviations: MMAA, monomethylarsonic acid; DMAA, dimethylarsinic acid; ICPMS, inductively coupled plasma mass spectrometry; HGAFS, hydride generation atomic fluorescence spectrometry; AFS, atomic fluorescence detector; PMT, photomultiplier tube; PTFE, polytetrafluoroethylene; SRM, Standard Reference Material; and FIA, flow injection analysis.

	References

Top Abstract Introduction Materials and Methods Results and Discussion References

 

1. . World Health Organization. Environmental health criteria 18: arsenic 1981 WHO Geneva. .
2. . National Research Council. Arsenic 1977 US National Academy of Sciences Washington, DC. .
3. Chappell WR, Abernathy CO, Cothern CR, eds. Arsenic exposure and health. Sci Technol Lett: Northwood, 1994..
4. Potius FW, Brown KG, Chen C-J. Health implications of arsenic in drinking water. J Am Water Works Assoc 1994;86:52-63.
5. Smith TJ, Crecelius EA, Reading JC. Airborne arsenic exposure and excretion of methylated arsenic compounds. Environ Health Perspect 1977;19:89-93. [ISI][Medline] [Order article via Infotrieve]
6. Buchet JP, Lauwerys R. Evaluation of exposure to inorganic arsenic in man. Facchetti S eds. Analytical techniques for heavy metals in biological fluids 1983:75-90 Elsevier Amsterdam. .
7. Foa V, Colombi A, Maroni M, Buratti M, Calzaferri G. The speciation of the chemical forms of arsenic in the biological monitoring of exposure to inorganic arsenic. Sci Total Environ 1984;34:241-259. [Medline] [Order article via Infotrieve]
8. Stoeppler M, Vahter M. Arsenic. Herber RFM Stoeppler M eds. Trace element analysis in biological specimens 1994:291-320 Elsevier Amsterdam. .
9. Aposhian HV. Biochemical toxicology of arsenic. Rev Biochem Toxicol 1989;10:265-299.
10. Hindmarsh JT, McCurdy RF. Clinical and environmental aspects of arsenic toxicity. CRC Crit Rev Clin Lab Sci 1986;23:315-347.
11. Tam GKH, Charbonneau SM, Bryce F, Pomroy C, Sandi E. Metabolism of inorganic arsenic (⁷⁴As) in humans following oral ingestion. Toxicol Appl Pharmacol 1979;50:319-322. [ISI][Medline] [Order article via Infotrieve]
12. Valentine JL, Kang KH, Spivey G. Arsenic levels in human blood, urine and hair in response via drinking water. Environ Res 1979;20:24-32. [Medline] [Order article via Infotrieve]
13. Pomroy C, Charbonneau SM, McCullough RS, Tam GKH. Human retention studies with ⁷⁴As. Toxicol Appl Pharmacol 1980;53:550-556. [ISI][Medline] [Order article via Infotrieve]
14. Buchet JP, Lauwerys R, Roels H. Comparison of the urinary excretion of arsenic metabolites after a single oral dose of sodium arsenite, monomethylarsonate, or dimethylarsinate in man. Int Arch Occup Environ Health 1981;48:71-79. [ISI][Medline] [Order article via Infotrieve]
15. Buchet JP, Lauwerys R, Roels H. Urinary excretion of inorganic arsenic and its metabolites after repeated ingestion of sodium metaarsenite by volunteers. Int Arch Occup Environ Health 1981;48:111-118. [ISI][Medline] [Order article via Infotrieve]
16. Mahieu P, Buchet JP, Roels HA, Lauwerys R. The metabolism of arsenic in humans acutely intoxicated by As₂O₃. Its significance for the duration of BAL therapy. Clin Toxicol 1981;18:1067-1082. [ISI][Medline] [Order article via Infotrieve]
17. Vahter M, Friberg L, Rahnster B, Nygren A, Nolinder P. Airborne arsenic and urinary excretion of metabolites of inorganic arsenic among smelter workers. Int Arch Occup Environ Health 1986;57:79-91. [ISI][Medline] [Order article via Infotrieve]
18. Le XC, Cullen WR, Reimer KJ. Human urinary arsenic excretion after one-time ingestion of seaweed, crab, and shrimp. Clin Chem 1994;40:617-624. [Abstract/Free Full Text]
19. Crecelius EA. Changes in the chemical speciation of arsenic following ingestion by man. Environ Health Perspect 1977;19:147-150. [ISI][Medline] [Order article via Infotrieve]
20. Freeman HC, Uthe JF, Fleming RB, Odense PH, Ackman RG, Landry G, et al. Clearance of arsenic ingested by man from arsenic contaminated fish. Bull Environ Contam Toxicol 1979;22:224-229. [ISI][Medline] [Order article via Infotrieve]
21. Yamauchi H, Takahashi K, Mashiko M, Saitoh J, Yamanura Y. Intake of different chemical species of dietary arsenic by the Japanese, and their blood and urinary arsenic levels. Appl Organomet Chem 1992;6:383-388.
22. Le XC, Cullen WR, Reimer KJ. Determination of urinary arsenic and impact of dietary arsenic intake. Talanta 1993;40:185-193.
23. Le XC, Ma M.. Speciation of arsenic compounds using ion pair chromatography with atomic spectrometry and mass spectrometry detection. J Chromatogr A 1997;764:55-64.
24. Le XC, Ma M, Wong NA. Speciation of arsenic compounds using high performance liquid chromatography at elevated temperature and selective hydride generation atomic fluorescence detection. Anal Chem 1996;68:4501-4506.
25. Braman RS, Foreback CC. Methylated forms of arsenic in the environment. Science 1973;182:1247-1249. [Abstract/Free Full Text]
26. Cullen WR, Reimer KJ. Arsenic speciation in the environment. Chem Rev 1989;89:713-764.
27. Francesconi KA, Edmonds JS. Arsenic in the sea. Oceanogr Marine Biol Annu Rev 1993;31:111-151.
28. Shibata Y, Morita M, Fuwa K. Selenium and arsenic in biology: their chemical forms and biological functions. Adv Biophys 1992;28:31-80. [Medline] [Order article via Infotrieve]
29. Andreae MO. Organoarsenic compounds in the environment. Craig PJ eds. Organometallic compounds in the environment. Principles and reactions 1986:198-228 John Wiley New York. .
30. Penrose WR. Arsenic in the marine and aquatic environment. Analysis, occurrence, and significance. CRC Crit Rev Environ Control 1974;4:465-482.
31. Kaise T, Fukui S. The chemical form and acute toxicity of arsenic compounds in marine organisms. Appl Organomet Chem 1992;6:155-160.
32. Edmonds JS, Francesconi KA, Cannon JR, Raston CL, Skelton BW, White AH. Isolation, crystal structure and synthesis of arsenobetaine, the essential constituent of the western rock lobster Panulilrus longipes cygnus George. Tetrahedron Lett 1977;18:1543-1546.
33. Edmonds JS, Francesconi KA. Arseno-sugars from brown kelp (Ecklonia radiata) as intermediates in cycling of arsenic in a marine ecosystem. Nature 1981;289:602-604.
34. Irgolic KJ. Arsenic. Stoeppler M eds. Hazardous metals in the environment 1992:287-350 Elsevier Amsterdam. .
35. Shibata Y, Morita M. Characterization of organic arsenic compounds in bivalves. Appl Organomet Chem 1992;6:343-349.
36. Le XC, Cullen WR, Reimer KJ. Speciation of arsenic compounds in some marine organisms. Environ Sci Technol 1994;28:1598-1604.
37. Larsen EH. Speciation of dimethylarsinyl-riboside derivatives (arsenosugars) in marine reference materials by HPLC-ICP-MS. Fresenius J Anal Chem 1995;352:582-588.
38. Vahter ME. Clarkson TW Friberg L Nordberg GF Sager PR eds. Biological monitoring of toxic metals 1988:303-321 Plenum Press New York. .
39. Le XC, Cullen WR, Reimer KJ. Decomposition of organoarsenic compounds by using a microwave oven and subsequent determination by flow injection hydride generation atomic absorption spectrometry. Appl Organomet Chem 1992;6:161-171.
40. Le XC, Cullen WR, Reimer KJ. Speciation of arsenic compounds by using HPLC with hydride generation atomic absorption spectrometry and inductively coupled plasma mass spectrometry detection. Talanta 1994;41:495-502.
41. Howard AG, Hunt LE. Coupled photo-oxidation–hydride AAS detector for the HPLC of arsenic compounds. Anal Chem 1993;65:2995-2998.
42. Stockwell PB, Corns WT. Environmental sensors based on atomic fluorescence. Analyst 1994;119:1641-1645.
43. . ACS Committee on Environmental Improvement. Guidelines for data acquisition and data quality evaluation in environmental chemistry. Anal Chem 1980;52:2242-2249.
44. Knoll JE. Estimation of the limit of detection in chromatography. J Chromtogr Sci 1985;23:422-425.
45. Beauchemin D, Siu KWM, McLaren JW, Berman SS. Determination of arsenic species by HPLC inductively coupled plasma mass spectrometry. J Anal At Spectrom 1989;4:285-289.
46. Heitkemper DT, Creed J, Caruso JA, Fricke FL. Speciation of arsenic in urine using HPLC with ICPMS detection. J Anal At Spectrom 1989;4:279-284.
47. Shibata Y, Morita M. Speciation of arsenic by reversed-phase high performance liquid chromatography–inductively coupled plasma mass spectrometry. Anal Sci 1989;5:107-109.
48. Vela NP, Olson LK, Caruso JA. Elemental speciation using inductively coupled plasma mass spectrometry. Anal Chem 1993;65:585A-597A. [Medline] [Order article via Infotrieve]
49. Uden PC, ed. Element-specific chromatographic detection by atomic emission spectroscopy. ACS Symposium Series 479; Washington, DC: American Chemical Society, 1992..
50. Branch S, Ebdon L, O'Neill P. Determination of arsenic species in fish by directly coupled high-performance liquid chromatography–inductively coupled plasma mass spectrometry. J Anal At Spectrom 1994;9:33-37.
51. Castro A, Shera EB. Single-molecule electrophoresis. Anal Chem 1995;67:3181-3186.
52. Haab BB, Mathies RA. Single molecule fluorescence burst detection of DNA fragments separated by capillary electrophoresis. Anal Chem 1995;67:3253-3260. [Medline] [Order article via Infotrieve]
53. Chen D, Dovichi NJ. Single-molecule detection in capillary electrophoresis: molecular shot noise as a fundamental limit to chemical analysis. Anal Chem 1996;68:690-696.
54. Pergantis SA, Winnik W, Betowski D. Determination of ten organoarsenic compounds using microbore high performance liquid chromatography coupled with electrospray mass spectrometry–mass spectrometry. J Anal At Spectrom 1997;12:531-536.
55. Hopenhayn-Rich C, Smith AH, Goeden HM. Human studies do not support the methylation threshold hypothesis for the toxicity of inorganic arsenic. Environ Res 1993;60:161-177. [Medline] [Order article via Infotrieve]