Sample Preparation and Storage Can Change Arsenic Speciation in Human Urine

Transcription

1 Clinical Chemistry 45: (1999) Drug Monitoring and Toxicology Sample Preparation and Storage Can Change Arsenic Speciation in Human Urine Jörg Feldmann, 1 Vivian W-M. Lai, 2 William R. Cullen, 2 Mingsheng Ma, 3 Xiufen Lu, 3 and X. Chris Le 3* Background: Stability of chemical speciation during sample handling and storage is a prerequisite to obtaining reliable results of trace element speciation analysis. There is no comprehensive information on the stability of common arsenic species, such as inorganic arsenite [As(III)], arsenate [As(V)], monomethylarsonic acid, dimethylarsinic acid, and arsenobetaine, in human urine. Methods: We compared the effects of the following storage conditions on the stability of these arsenic species: temperature (25, 4, and 20 C), storage time (1, 2, 4, and 8 months), and the use of additives (HCl, sodium azide, benzoic acid, benzyltrimethylammonium chloride, and cetylpyridinium chloride). HPLC with both inductively coupled plasma mass spectrometry and hydride generation atomic fluorescence detection techniques were used for the speciation of arsenic. Results: We found that all five of the arsenic species were stable for up to 2 months when urine samples were stored at 4 and 20 C without any additives. For longer period of storage (4 and 8 months), the stability of arsenic species was dependent on urine matrices. Whereas the arsenic speciation in some urine samples was stable for the entire 8 months at both 4 and 20 C, other urine samples stored under identical conditions showed substantial changes in the concentration of As(III), As(V), monomethylarsonic acid, and dimethylarsinic acid. The use of additives did not improve the 1 University of Aberdeen, Department of Chemistry, Old Aberdeen, AB24 3UE Scotland, UK. 2 University of British Columbia, Department of Chemistry, Vancouver, British Columbia, Canada V6T 1Z1. 3 University of Alberta, Department of Public Health Sciences, Edmonton, Alberta, Canada T6G 2G3. Portions of this work were presented at the 3rd International Conference on Arsenic Exposure and Health Effects, July 12 15, 1998, San Diego, CA. *Address for correspondence to this author at: Environmental Health Sciences Program, Department of Public Health Sciences, CSB, Faculty of Medicine, University of Alberta, Edmonton, Alberta, Canada T6G 2G3. Fax ; xc.le@ualberta.ca. Received June 8, 1999; accepted August 24, stability of arsenic speciation in urine. The addition of 0.1 mol/l HCl (final concentration) to urine samples produced relative changes in inorganic As(III) and As(V) concentrations. Conclusions: Low temperature (4 and 20 C) conditions are suitable for the storage of urine samples for up to 2 months. Untreated samples maintain their concentration of arsenic species, and additives have no particular benefit. Strong acidification is not appropriate for speciation analysis American Association for Clinical Chemistry Chronic exposure to inorganic arsenic compounds is a major concern in many parts of the world, including India, Bangladesh, China, Argentina, New Mexico, and Chile because of the high cancer risk and neurologic implications associated with the ingestion of increased amounts of arsenic (1, 2). Much research has been aimed at providing a better dose response relationship for health risk assessment from exposure to small amounts of arsenic. In this connection, chemical speciation of arsenic plays an important role in understanding human health effects and arsenic metabolism. Urinary excretion of arsenic metabolites is the primary pathway for the elimination of arsenic from human body (3 5). Determination of arsenic in urine is commonly used as a measure of recent exposure to arsenic. Most of the inorganic arsenic, As(III) and As(V), is metabolized to dimethylarsinic acid (DMA) 4 and monomethylarsonic acid (MMA) before excretion into the urine. The proportion of these arsenic species in urine is typically 60 80% DMA, 10 20% MMA, and 10 20% inorganic arsenic (5 8) in individuals who do not eat food of marine origin, such as fish, shellfish, and algae. Most populations have other arsenic species in their urine, which is assumed to be 4 Nonstandard abbreviations: DMA, dimethylarsinic acid; MMA, monomethylarsonic acid; AsB, arsenobetaine; ICPMS, inductively coupled plasma mass spectrometry; SRM, standard reference material; and HGAFS, hydride generation atomic fluorescence spectrometry. 1988

2 Clinical Chemistry 45, No. 11, arsenobetaine (AsB) that is present as a result of eating food of marine origin. Because the relative acute toxicity of these arsenic compounds decreases from inorganic arsenite and arsenate (LD 50, mg/kg) to MMA (LD 50, mg/kg) and DMA (LD 50, mg/ kg), it has been suggested that the methylation of arsenic in the body is a natural detoxification pathway (9 11). More recent research argues that although the acute toxicity is decreased by methylation, the genotoxic effects of these arsenic compounds are not well understood and may not follow the same decreasing order. Several studies have suggested that DMA may be more harmful than the parent inorganic arsenic compounds (12 17). It is possible that arsenic methylation can alter the methylation of DNA (18, 19) because the methylation of both arsenic and DNA requires the same methyl donor, S-adenosylmethionine. Although the effect of arsenic methylation on the genotoxicity of arsenic species is not clear, the relative concentrations of the methylated arsenic metabolites in the excreted urine have been used to compare methylation capacity between individuals and between populations (20 33). For example, a much lower portion (2.2%) of urinary MMA was found in native Andean women (34) compared with 10 20% of urinary MMA in other populations. In another study of a population in northern Argentina, children were found to have a substantially higher percentage of inorganic arsenic (50%) in their urine samples than the women (32%) (35). A crucial requirement for obtaining relative concentration of these arsenic species is maintaining the concentration of the original chemical species in the sample before analysis. This is a special requirement for speciation analysis. For determining total element concentrations, the main considerations for sample collection and storage are to prevent contamination and to minimize loss of trace amounts of analytes. Polyethylene containers usually are preferred to glass containers because the former is less adsorptive for arsenic (36). Traditionally, samples are acidified to reduce potential adsorption of trace elements onto the sample container surface. Little consideration has been given to the stability of chemical forms of the element. In the case of speciation analysis, obtaining reliable information requires the concentration of individual species of the element to be unchanged by sample handling and treatment. Many of the urinary arsenic speciation studies involved the collection of urine samples from populations in remote areas, often in a foreign country. Urine samples were then shipped to a laboratory several days later for arsenic speciation analysis. Various sampling and storage protocols have been reported in the literature, including acidification, centrifugation, refrigeration, and freezing. However, little is known about how these procedures affect the concentration of individual arsenic species. Larsen et al. (37) observed that the concentrations of DMA, MMA, and AsB were relatively constant. However rapid oxidation of As(III) to As(V) was observed. Palacios et al. (38) found that As(V), MMA, DMA, and AsB (200 g/l each) in urine were stable for the entire testing period of 67 days at 4 C. The present study provides a systematic investigation into the stability of arsenic speciation in urine. Sample storage conditions, including temperature, duration, acidification, and the use of additives, on the stability of arsenic species are examined. The results provide the basis for designing appropriate urine sample storage conditions that are suitable for arsenic speciation analysis. Materials and Methods reagents and standards An atomic absorption arsenic standard solution (Sigma) containing mg As/L as arsenite in 20 g/l KOH was used as the primary arsenic standard. Sodium arsenate [As(O)OH(ONa) 2 7 H 2 O] and sodium cacodylate [(CH 3 ) 2 As(O)ONa] were obtained from Sigma, and monomethylarsonate [CH 3 As(O)OHONa] was obtained from Chem Service. AsB was synthesized as described previously (39). Stock solutions (1000 mg As/L) of these arsenicals were prepared by dissolving appropriate amounts of the corresponding arsenic compounds in 0.01 mol/l HCl, and calibrators were prepared by serial dilution with deionized water. Concentrations of arsenic in sodium arsenate, sodium cacodylate, sodium monomethylarsonate, and AsB solutions were standardized against the atomic absorption arsenic standard solution using both inductively coupled plasma mass spectrometry (ICPMS) and flame atomic absorption spectrometry analyses. The reagents used in HPLC mobile phases, including tetrabutylammonium hydroxide, malonic acid, NaH 2 PO 4, and Na 2 HPO 4, were obtained from Aldrich. HPLC-grade methanol was from Fisher. These mobile phase solutions were prepared in deionized water and filtered through a 0.2 m membrane before use. Sodium borohydride (Aldrich) solutions (30 g/l) in 0.1 mol/l sodium hydroxide (Fisher) were prepared fresh daily. All reagents used were of analytical grade or better. A stock solution (0.4 mol/l) of benzoic acid (BDH) was prepared by dissolving an appropriate amount of solid benzoic acid in methanol (HPLC grade; Fisher). Stock solutions of sodium azide (MC&B), benzyltrimethylammonium chloride (Hexcel), and cetylpyridinium chloride (BDH) were prepared by dissolving appropriate amounts of these reagents in water. HCl (37%; Fisher), formic acid (BDH), sodium hydroxide (BDH), and ammonia (BDH) were used for the ph adjustment. urine samples A Standard Reference Material (SRM), Toxic Metals in Freeze-Dried Urine SRM 2670, was obtained from NIST (Gaithersburg, MD). The freeze-dried urine was reconstituted by the addition of 20.0 ml of deionized water as recommended by the supplier. The certified value for total arsenic concentration is g/l in two

3 1990 Feldmann et al.: Stability of Arsenic Speciation in Human Urine Table 1. Concentrations of arsenic species ( g/l) in SRM 2670 urine (increased concentrations). AsB As(III) DMA MMA As(V) Total As Reference 15 3 ND a b ND c (46) (47) (48) d (49) 11 3 (50) (51) (52) (53) (54) e a ND, below detection limit;, not available or not determined. b HPLC-ICPMS method described in this report. c HPLC-HGAFS method described in this report. d The sum of As(III) and As(V). e Certified value given by the supplier (NIST). bottles containing increased concentrations of toxic metals. In the other two bottles containing normal concentrations of toxic metals, the concentration of arsenic is not certified and a reference value of 60 g/l has been provided. No arsenic speciation information was given for the SRM. Human urine standard (lot nos and 43182) was obtained from Quantimetrix Corp. It is a ready-to-use liquid and is prepared from human urine. Sodium azide is present in the urine as a preservative. This is referred to as standard urine in the present study. It is used as a urine matrix and supplemented with arsenic calibrators before storage studies to examine the stability of solution of arsenic species in this urine matrix. First morning urine specimens from a female and a male volunteer were collected and were also used in the stability study. These specimens are referred to as volunteer urine. The volunteers are healthy students who refrained from eating any seafood for 4 days before the collection of the urine specimens. Total arsenic concentrations in these urine specimens were 10 g/l. The volunteer urine was supplemented with known amounts of arsenic species before the stability of arsenic species over storage time was examined. Several first morning specimens were also obtained from a study population in Utah (40). The arsenic concentration in the drinking water of this population ranged from 8 to 680 g/l. Arsenic concentrations in urine samples from those who ingested large amounts of arsenic from drinking water were higher than those from low-exposure populations. Four urine samples shown later in Fig. 6 contained the following arsenic concentrations: inorganic arsenite (3 12 g/l), DMA (7 120 g/l), MMA (4 34 g/l), and inorganic arsenate (up to 5 g/l). Creatinine concentrations in these samples ranged from 0.7 to 2.9 g/l. No arsenic species were added into these samples. The original arsenic species present in the samples were monitored over time to examine the stability of arsenic species in representative urine samples. Fig. 1. Chromatograms obtained from the HPLC-ICPMS analyses of a urine sample to which 50 g/l of each of four arsenic species had been added. A Hamilton PRP X-anion-exchange column 100 ( mm) was used for separation, which was carried out at ambient temperature. The mobile phase contained 50 mmol/l H 2 PO 4 /HPO 2 4 buffer (ph 6, adjusted with NH 4 OH), and its flow rate was 1.3 ml/min. ICPMS was used to monitor arsenic, m/z 75. The sample had been stored at 4 C for 2 months before HPLC-ICPMS analysis. (A), no preservative was added. The recovery was 95% for the arsenic species added into the urine sample. (B), HCl was added to the sample to a final concentration of 0.1 mol/l before sample storage.

4 Clinical Chemistry 45, No. 11, Fig. 3. Chromatograms obtained from the HPLC-ICPMS analyses of a volunteer urine sample to which 50 g/l of each of four arsenic species had been added. The sample had been stored at 20 C for 8 months before HPLC-ICPMS analysis. (A), no preservative was added; (B), HCl was added to the sample to a final concentration of 0.1 mol/l before sample storage. The solid trace was from the analysis of the sample itself. The dotted trace corresponds to co-injection of the acidified sample with MMA and DMA calibrators for HPLC analysis. For clarity, the dotted trace was manually shifted on vertical axis. U, uncharacterized arsenic species. HPLC-ICPMS conditions were the same as shown in the legend for Fig. 1. Fig. 2. Recovery of four arsenic species in the standard urine (US) and volunteer urine (UR) samples after storage for 2 months at different temperatures. (A), all samples contained 0.1 mol/l HCl; (B), no HCl was added to the samples. HPLC-ICPMS was used for the determination of arsenic species. sample storage experiments One set of standard urine and volunteer urine samples were supplemented with arsenite, arsenate, MMA, and DMA (50 g/l as arsenic). Another set of standard urine and volunteer urine samples were supplemented with AsB (50 g/l as arsenic). Replicate aliquots of these samples were placed in separate polyethylene bottles. They were stored in the dark for up to 8 months at three temperature conditions to simulate field sampling situations: 20 C (freezer), 4 C (cool box or refrigerator), and 25 C (room temperature). After a desired storage time (1, 2, 4, and 8 months), aliquots of the samples were subjected to HPLC/ICPMS analyses. Samples stored at 20 C were thawed at room temperature before analysis. All samples were filtered through 0.2 m membrane filters before injection onto the HPLC column for analysis. To study the effect of acidification on arsenic stability, appropriate volumes of HCl was added to a set of standard urine and volunteer urine samples to make the final HCl concentration of 0.1 mol/l. The samples were supplemented with the same arsenic species and stored under the conditions stated above. Another set of standard urine and volunteer urine samples were tested for the effect of other possible preservatives. Sodium azide, benzoic acid, benzyltrimethylammonium chloride, and cetylpyridinium chloride were added to separate urine samples to make the final concentration in the sample 0.01 mol/l. The samples were supplemented with the same arsenic species and stored under the conditions stated above. The urine samples from the Utah population were stored at 20 C without any additives. Most of the samples contained all four arsenic species, and therefore, no arsenic was added into these samples. hplc separation with icpms detection HPLC separation with ICPMS detection was used for arsenic speciation (5, 41, 42). The system consisted of a Waters Model 510 solvent delivery pump, a Waters U6K injector, and an appropriate column. Separation of As(III),

5 1992 Feldmann et al.: Stability of Arsenic Speciation in Human Urine Fig. 4. Effect of additives on the recovery of four arsenic species in urine samples stored at 4 C (A) or 20 C (B) for 1 month. Sodium azide (NaN 3 ), benzyltrimethylammonium chloride (BzMe 3 NH 4 Cl ), benzoic acid (BzAcid), cetylpyridinium chloride (Cetyl.), and methanol (MeOH) were tested as preservatives. HPLC-ICPMS was used for the determination of arsenic species. DMA, MMA, and As(V) species was carried out on a strong anion-exchange column with 30 mmol/l phosphate as the mobile phase (ph 6.0). The ph was adjusted with ammonia, purged with helium, and filtered through a 0.2 m membrane filter. The HPLC effluent was directly introduced to a DeGalan nebulizer of the ICPMS system (PlasmaQuad 2 Turbo Plus; VG Elemental; Fisons Instrument) via a PTFE tube (20 cm 0.4 mm) and appropriate fittings. The time-resolved analysis mode was used to monitor multiple ions. Signal intensity (cps) at m/z 75 was monitored for the quantification of arsenic. Signal intensity at m/z 77 was also monitored and used to correct for interference from 40 Ar 37 Cl. Cation-exchange chromatography with ICPMS detection was used for the determination of AsB. A Supelcosil LC-SCX column (4.6 mm 250 mm, 5 m particle size; Supelco) with 20 mmol/l pyridine (Fisher) as the mobile phase was used for the separation. The ph of the mobile phase was adjusted to 2.7 with formic acid (BDH). Fig. 5. Effect of storage duration on the recovery of four arsenic species in volunteer urine (A) and standard urine (B) samples stored at 4 C. HPLC-ICPMS was used for the determination of arsenic species. hplc separation with hydride generation atomic fluorescence spectrometry (hgafs) detection The second method of quantifying arsenic species was based on ion-pair chromatographic separation with hydride generation atomic fluorescence detection as described previously (43, 44). The HPLC system consisted of a Gilson Model 370 pump with a 5 ml/min stainless steel pump head, a Rheodyne six-port sample injector (Model 7725i) with a 20- L sample loop, and a reversed-phase C 18 column (ODS-3, 150 mm 4.6 mm, 3- m particle size; Phenomenex). For HPLC column temperature control, the separation column was mounted inside a column heater (Model CH-30; Eppendorf) that was controlled by a temperature controller (Model TC-50; Eppendorf). The mobile phase was preheated to the temperature of the column by a 50-cm precolumn coil of stainless steel capillary tubing, which was also placed inside the column heater. The temperature controller, according to the manufacturer, provides a 0.1 C temperature stability and 1 C

6 Clinical Chemistry 45, No. 11, Table 2. Recovery and relative concentrations of arsenic species in five replicate samples of standard urine stored at 20 C for 8 months. Recovery, % As(III) DMA MMA As(V) Total As A. Recovery (%) of arsenic species (50 g/l) Sample A B C D E Mean SD RSD, a % B. Relative concentrations of arsenic species Relative concentration, % Sample A B C D E Mean SD RSD, a % a RSD, relative standard deviation. accuracy. The temperature of the column was maintained at either 30 or 50 C. A solution (ph 5.8) containing 5 mmol/l tetrabutylammonium hydroxide, 4 mmol/l malonic acid, and 50 ml/l methanol, was used as the HPLC mobile phase at a flow rate of 1.5 ml/min. Effluent from the HPLC column was mixed at two T-joints, with continuous flows of HCl (2 mol/l) and sodium borohydride (13 g/l). Any arsines generated were separated from liquid waste and carried by a continuous flow of argon to an atomic fluorescence detector (Excalibur ; P.S. Analytical) for quantification. A hydride generation atomic fluorescence spectrometer (model Excalibur ; PS Analytical) was used as an HPLC detector. The atomic fluorescence detector consisted of an excitation source, an atom cell, fluorescence collection optics, a photomultiplier tube, and a data collection unit. A quartz tube with argon/hydrogen diffusion flame was used as the atom cell for atomization. An arsenic hollow cathode lamp was used for fluorescence excitation. Atomic fluorescence (193.7 nm) was collected at a right angle with respect to the excitation light, filtered with a multireflectance filter to reduce scattering and background noise, and detected with a solar blind photo multiplier tube. A Pentium computer with Varian Star Workstation software and an ADC board was used to acquire and process signals from the atomic fluorescence detector. A Hewlett Packard 3390A integrator with both peak area and peak height measurement capability was also used to record chromatograms. method comparison The quantification of arsenic species in urine samples was compared between two laboratories, one using HPLC/ ICPMS methodology and the other using HPLC/HGAFS. Arsenic speciation analysis of the SRM urine, SRM 2670, by the two methods showed good agreement with the reference values as summarized in Table 1. Variations in arsenic speciation obtained by others may in part be the result of differences in sample storage and treatment. Results and Discussion effects of acid on the stability of four arsenic species Acidification of samples is a common practice in sample treatment and handling for trace element analysis. The primary purpose of acidification of samples in the determination of trace element content has been to stabilize the concentration of the total amounts of trace elements of interest, regardless of the specific chemical form of the element. The effect of acidification on chemical species has not been investigated systematically. To address this issue, we carried out a series of studies to examine how acidification affects the concentration of arsenic species. Fig. 1 shows chromatograms obtained from an untreated urine sample (Fig. 1A) and an acidified urine

7 1994 Feldmann et al.: Stability of Arsenic Speciation in Human Urine Fig. 6. Chromatograms obtained from the HPLC-HGAFS analyses of four arsenic species in urine samples within 2 weeks of collection (1A 4A) and after eight months of storage at 20 C (1B 4B). A 15-cm column (Phenomenex ODS-3, 3 m particle size) was used for separation. The mobile phase contained 5 mmol/l tetrabutylammonium hydroxide, 4 mmol/l malonic acid, and 50 ml/l methanol (ph 5.8), and its flow rate was 1.5 ml/min. Peaks 1, 2, 3, and 4 correspond to As(III), DMA, MMA, and As(V), respectively. sample (Fig. 1B) after 2 months storage in a refrigerator (4 C). The urine sample was from a volunteer who refrained from eating any seafood for 4 days before collection of the first morning void. It was supplemented with 50 g/l arsenite, arsenate, MMA, and DMA species. The concentrations of the four arsenic species in the untreated urine sample were essentially unchanged for 2 months (Fig. 1A), with recoveries ranging from 87% to 108%. The concentrations of the same arsenic species in the parallel samples containing 0.1 mol/l HCl, however, were not stable at the same storage temperature and duration (Fig. 1B). Inorganic arsenate was partially reduced to arsenite, the recovery of arsenate was reduced to 22%, and the recovery of arsenite was increased to 260%. The recoveries for MMA and DMA were 50% and 63%, respectively. The overall recovery of total arsenic was 99%. Fig. 2 shows recoveries for the four arsenic species in acidified samples (Fig. 2A) and untreated samples (Fig. 2B) stored for 2 months at three temperature conditions: 20 C (frozen), 4 C (refrigerated), and 25 C (room temperature). Urine samples from volunteer (UR) and standard (US) urines, both with 50 g/l of each of the four arsenic species added, are compared. In the acidified samples, although there were variations among various storage conditions, changes of arsenic speciation are apparent in all cases (Fig. 2A). None of these storage temperatures provided the needed arsenic speciation stability when 0.1 mol/l HCl was present in the samples. After a longer storage time, the effect of 0.1 mol/l HCl (final concentration) added for acidification was more severe. In several samples stored for 8 months at the three temperature conditions, DMA, MMA, and arsenate were completely lost. Only inorganic arsenite and an unknown arsenic species were observed. Fig. 3 shows an example of untreated (Fig. 3A) and acidified (Fig. 3B) sample, stored for 8 months at 20 C. In the acidified sample, the arsenite concentration increased to 94 g/l from 50 g/l initially added to the sample. No inorganic arsenate, MMA, or DMA was detected, although 50 g/l of each of these species was added to the samples 8 months earlier. Instead, an unknown arsenic species was present at an arsenic concentration of 30 g/l. The overall recovery of total arsenic was 60%. The loss may be attributable to coprecipitation of arsenic with urine sample matrix. Coinjection of this sample with freshly prepared arsenic calibrators for HPLC analysis confirmed that the unknown arsenic species did not coelute with any of the other arsenic calibrators available to us, which include trimethylarsine oxide, AsB, arsenocholine, tetramethylarsonium, and three arsenosugars. Five replicate preparations of the acidified sample consistently showed the presence of this uncharacterized arsenic species after storage under identical conditions. We concluded that the addition of HCl (final concentration, 0.1 mol/l) to urine samples for acidification produces substantial changes in the concentration of arsenic species and therefore is not a suitable sample treatment if species information is required. As(III) is thermodynamically favored in acidic solution (E V). Thus, the conversion of As(V) to As(III) in acidic media is understandable. The converse is true in alkaline ph (E V). use of possible preservatives Because acidification was unsuccessful for speciation preservation, several candidate preservatives, including sodium azide, benzoic acid, benzyltrimethylammonium chloride, and cetylpyridinium chloride, were tested. These have known antibacterial activities (45); for example, benzoic acid has been used as a common food preservative.

8 Clinical Chemistry 45, No. 11, Table 3. Recovery of AsB (50 g/l) from volunteer and standard urine samples stored at three temperatures for up to 8 months. Recovery, g/l 20 C 4 C 25 C Storage time, months No HCl With HCl No HCl With HCl No HCl With HCl Volunteer urine Standard urine Fig. 4 shows the recovery of the four arsenic species in volunteer urine samples stored for 1 month at 4 C (Fig. 4A) and 20 C (Fig. 4B). Five additives were tested for their potential ability to preserve chemical speciation; however, no marked improvement in the recovery of the arsenic species was observed as a result of the addition of these agents. To simplify sample procedures and to eliminate potential contamination related to the introduction of any additives, no preservative was used in subsequent studies. no additives Concentrations of all four arsenic species were relatively constant in both the volunteer urine (UR) and standard urine (US) matrices (Fig. 2B) for up to 2 months of storage at either 4 or 20 C. Recovery for the four arsenic species ranged from 84% to 115%. The arsenic speciation was much less stable in samples stored at room temperature for the same duration of 2 months; the recovery of the four arsenic species ranged from 57% for DMA to 150% for As(III). Storage for longer than 2 months produced differences in recovery from different urine sample matrices, as demonstrated in Fig. 5. Recoveries of four arsenic species in two sample matrices, volunteer urine (Fig. 5A) and standard urine (Fig. 5B), stored at 4 C for up to 8 months were compared. Whereas the arsenic speciation in the volunteer urine sample was relatively stable for the entire 8 months, 30% of As(III) was oxidized to As(V) in the standard urine after storage for longer than 4 months. Recovery of arsenic species in samples stored at 20 C also varied with sample matrices. In some cases, recoveries for the four arsenic species were as low as 30% (Table 2A) after the volunteer urine was stored at 20 C for 4 months. The low recovery may be caused by adsorption of the arsenic species to the surface of the sample container and/or precipitation of arsenic. Precipitates were visible after the urine samples were frozen at 20 C and later thawed for HPLC analysis. However, the relative ratio among the four arsenic species did not change significantly, as shown in Table 2B. The concentrations of MMA and DMA were less variable than inorganic As(III) and As(V) throughout the entire storage period, a finding consistent with earlier studies (37, 38). The possibility of low recovery should be considered when HPLC-ICPMS or HPLC-HGAFS is used because samples are filtered before HPLC analysis. If the sample is not filtered, as in the case of selective hydride generation atomic absorption spectrometry or hydride generation followed by cold-trap gas chromatography-atomic absorption spectrometry, any arsenic species adsorbed to particulate matter will be analyzed. Adjusting the ph of urine to 4.5 with dilute nitric acid before sample storage at 4 C did not show adverse effects on the stability of arsenic speciation. application to urinary arsenic speciation studies It can be concluded from the present research that the appropriate storage conditions for obtaining quantitative recovery are either 4 C (refrigeration) or 20 C (freezing), without the use of any additives. This sample storage protocol was applied to an epidemiological survey of urinary arsenic speciation resulting from arsenic exposure (40). Fig. 6 shows four pairs of chromatograms obtained from untreated urine samples. The concentration of creatinine in these samples was 2.95, 2.08, 0.86, and 0.70 g/l, respectively. The left column represents analyses of urine samples 2 weeks after sample collection. The samples were stored at 20 C and reanalyzed 8 months later; the chromatograms for the samples stored for 8 months are shown in the right column. Arsenic speciation profiles are similar in most samples before (Fig. 6, 1A 4A) and after (Fig. 6, 1B 4B) storage for 8 months at 20 C. There was some reduction of As(V) to As(III) in sample 4 after storage for 8 months; the concentrations of DMA and MMA are unchanged, confirming the findings shown in Fig. 5. stability of AsB Table 3 summarizes the recovery of 50 g/l arsenic as AsB added to volunteer urine and standard urine samples. Both acidified and untreated urine samples were

9 1996 Feldmann et al.: Stability of Arsenic Speciation in Human Urine stored for up to 8 months at three temperature conditions: 20, 4, and 25 C. Partial transformation of AsB was observed only in several cases. Up to 3 g/l DMA and 1 g/l MMA were observed in separate samples stored for 8 months, which correspond to 6% and 2% conversion of AsB, respectively. Acceptable recoveries were obtained in most cases. These results are consistent with previous observations, indicating that AsB is very stable. In conclusion, low temperature (4 and 20 C) conditions are suitable for the storage of urine samples for up to 2 months without substantial changes of arsenic speciation. For longer storage times, the stability of arsenic species varies with sample matrix. Accurate measurement of inorganic As(III) and As(V) separately is more difficult because the concentrations of these arsenic species in urine samples are more variable over storage time. Untreated samples maintain their concentration of arsenic species, and additives have no particular benefit. Strong acidification of samples leads to changes of arsenic speciation and thus is not suitable for arsenic speciation analysis, although dilute acetic, hydrochloric, and nitric acids have traditionally been added to samples to minimize possible adsorption of trace elements to sample containers. Depending on field sampling logistics, storage and shipping of samples at either 4 or 20 C may be chosen. The concentration of AsB is essentially unchanged for up to 8 months storage. This work was supported in part by the American Water Works Association Research Foundation (AWWARF) and the Natural Sciences and Engineering Research Council of Canada (NSERC). We thank AWWARF for its financial, technical, and administrative assistance in funding and managing the project (RFP 287). We thank Dr. R. L. Calderon and colleagues at the US Environmental Protection Agency for providing the Utah urine samples. References 1. Chappell WR, Abernathy CO, Cothern CR, eds. Arsenic: exposure and health effects. Northwood, UK: Sci Technol Lett, 1994:318 pp. 2. Abernathy CO, Calderon RL, Chappell WR, eds. Arsenic. exposure and health effects. London: Chapman & Hall, 1997:429 pp. 3. Buchet JP, Lauwerys R. Evaluation of exposure to inorganic arsenic in man. In: Facchetti S, ed. Analytical techniques for heavy metals in biological fluids. Amsterdam: Elsevier, 1983: Vahter M. Species differences in the metabolism of arsenic compounds. Appl Organomet Chem 1994;8: Le XC, Cullen WR, Reimer KJ. Human urinary arsenic excretion following one-time ingestion of arsenosugars present in seaweed and arsenobetaine present in crab and shrimp. Clin Chem 1994; 40: 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: Vahter M. Environmental and occupational exposure to inorganic arsenic. Acta Pharmacol Toxicol 1986;59: Vahter M. What are the chemical forms of arsenic in urine, and what can they tell us about exposure? Clin Chem 1994;40: Vahter M. Metabolism of arsenic. In: Fowler BA, ed. Biological and environmental effects of arsenic. Amsterdam: Elsevier, 1983: Buchet JP, Lauwerys R, Roels H. Comparison of the urinary excretion of arsenic metabolites after a single dose of sodium arsenite, monomethylarsonate or dimethylarsinate in man. Int Arch Occup Environ Health 1981;48: Aposhian HV. Biochemical toxicology of arsenic. Rev Biochem Toxicol 1989;10: Styblo M, Delnomdedieu M, Thomas DJ. Biological mechanisms and toxicological consequences of the methylation of arsenic. In: Goyer RA, Cherian G, eds. Toxicology of metals biochemical aspects. Handbook of experimental pharmacology. Berlin: Springer-Verlag, 1995: Tezuka M, Hanioka K, Yamanaka K, Okada S. Gene damage induced in human alveolar type II (L-132) cells by exposure to dimethylarsinic acid. Biochem Biophys Res Commun 1993;191: Yamanaka K, Okada S. Induction of lung-specific DNA damage by metabolically methylated arsenics via the production of free radicals. Environ Health Perspect 1994;102: Ochi T, Nakajima F, Sakurai T, Kaise T, Oya-Ohya Y. Dimethylarsinic acid causes apoptosis in HL-60 cells via interaction with glutathione. Arch Toxicol 1996;70: Wanibuchi H, Yamamoto S, Chen H, Yoshida K, Endo G, Hori T, Fukushima S. Promoting effects of dimethylarsinic acid on N-butyl- N-(4-hydroxybutyl)nitrosamine-induced urinary bladder carcinogenesis in rats. Carcinogenesis 1996;17: Brown JL, Kitchin KT, George M. Dimethylarsinic acid treatment alters six different rat biochemical parameters: relevance to arsenic carcinogenesis. Teratog Carcinog Mutagen 1997;17: Zhao CQ, Young M, Diwan BA, Coogan TP, Waalkes MP. Association of arsenic-induced malignant transformation with DNA hypomethylation and aberrant gene expression. Proc Natl Acad Sci U S A 1997;94: Mass MJ, Wang L. Arsenic alters cytosine methylation patterns of the promoter of the tumor suppressor gene p53 in human lung cells: a model for a mechanism of carcinogenesis. Mutat Res 1997;386: 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: Hopenhayn-Rich C, Biggs ML, Smith AH, Kalman DA, Moore LE. Methylation study of a population environmentally exposed to arsenic in drinking water. Environ Health Perspect 1996;104: Hopenhayn-Rich C, Biggs ML, Kalman DA, Moore LE, Smith AH. Arsenic methylation patterns before and after changing from high to lower concentrations of arsenic in drinking water. Environ Health Perspect 1996;104: Foà V, Colombi A, Maroni M, Burrati 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: Kalman DA, Hughes J, van Belle G, Burbacher T, Bolgiano D, Coble K, et al. The effect of variable environmental arsenic contamination on urinary concentrations of arsenic species. Environ Health Perspect 1990;89: Yamauchi H, Takahashi K, Mashiko M, Saitoh J, Yamamura Y.

10 Clinical Chemistry 45, No. 11, Intake of different chemical species of dietary arsenic by the Japanese, and their blood and urinary arsenic levels. Appl Organomet Chem 1992;6: Styblo M, Yamauchi H, Thomas DJ. Comparative in vitro methylation of trivalent and pentavalent arsenicals. Toxicol Appl Pharmacol 1995;135: Hakala E, Pyy L. Assessment of exposure to inorganic arsenic by determining the arsenic species excreted in urine. Toxicol Lett 1995;77: Lin TH, Huang YL. Chemical speciation of arsenic in urine of patients with blackfoot disease. Biol Trace Elem Res 1995;48: Buchet JP, Lison D, Ruggeri M, Foa V, Elia G. Assessment of exposure to inorganic arsenic, a human carcinogen, due to the consumption of seafood. Arch Toxicol 1996;70: Yager JW, Hicks JB, Fabianova E. Airborne arsenic and urinary excretion of arsenic metabolites during boiler cleaning operations in a Slovak coal-fired power plant. Environ Health Perspect 1997; 105: Del Razo LM, Garcia-Vargas GG, Vargas H, Albores A, Gonsebatt ME, Montero R, et al. Altered profile of urinary arsenic metabolites in adults with chronic arsenicism. A pilot study. Arch Toxicol 1997;71: Ng JC, Johnson D, Imray P, Chiswell B, Moore MR. Speciation of arsenic metabolites in the urine of occupational workers and experimental rats using an optimised hydride cold-trapping method. Analyst 1998;123: Kavanagh P, Farago ME, Thornton I, Goessler W, Kuehnelt D, Schlagenhaufen C, Irgolic KJ. Urinary arsenic species in Devon and Cornwall residents, UK. A pilot study. Analyst 1998;123: Vahter M, Concha G, Nermell B, Nilsson R, Dulout F, Natarajan AT. A unique metabolism of inorganic arsenic in native Andean women. Eur J Pharmacol Environ Toxicol Pharmacol 1995;293: Concha G, Vogler G, Lezcano D, Nermell B, Vahter M. Exposure to inorganic arsenic metabolites during early human development. Toxicol Sci 1998;44: Schaller KH, Fleischer M, Angerer J, Lewalter J. Arsenic determination in urine. In: Angerer J, Schaller KH, eds. Analyses of hazardous substances in biological materials. Weinheim, Germany: VCH Verlag, 1991: Larsen EH, Pritzl G, Hansen SH. Speciation of eight arsenic compounds in human urine by high-performance liquid chromatography with inductively coupled plasma mass spectrometric detection using antimonate for internal chromatographic standardization. J Anal Atom Spectrom 1993;8: Palacios MA, Gomez M, Camara C, Lopez MA. Stability studies of arsenate, monomethylarsonate, dimethylarsinate, arsenobetaine and arsenocholine in deionized water, urine and clean-up dry residue from urine samples and determination by liquid chromatography with microwave-assisted oxidation-hydride generation atomic absorption spectrometric detection. Anal Chim Acta 1997; 340: Edmonds JS, Francesconi KA, Cannon JR, Raston CL, Skelton BW, White A. Isolation, crystal structure and synthesis of arsenobetaine, the arsenical constituent of the lobster Panulirus longipes cygnus george. Tetrahedron Lett 1977;18: Calderon R, Hudgens EE, Le XC, Schreinemachers DM, Thomas, DJ. Excretion of arsenic in urine as a function of arsenic exposure in drinking water. Environ Health Perspect 1999;107: Le XC, Cullen WR, Reimer K. Speciation of arsenic compounds in some marine organisms. Environ Sci Technol 1994;28: Le XC, Li X-F, Lai V, Ma M, Yalcin S, Feldmann J. Simultaneous speciation of selenium and arsenic using elevated temperature liquid chromatography separation with inductively coupled plasma mass spectrometry detection. Spectrochim Acta Part B At Spectrosc 1998;53: Le XC, Ma M. Short-column liquid chromatography with hydride generation atomic fluorescence detection for the speciation of arsenic. Anal Chem 1998;70: Ma M, Le XC. Effect of arsenosugar ingestion on urinary arsenic speciation. Clin Chem 1998;44: Lynn B, Hugo WB. Chemical disinfectants, antiseptics and preservatives. In: Hugo WB, Russell AD, eds. Pharmaceutical microbiology, 3rd ed. London: Blackwell Scientific Publications, 1983: Zheng J, Goessler W, Kosmus W. Speciation of arsenic compounds by coupling high-performance liquid chromatography with inductively coupled plasma mass spectrometry. Mikrochim Acta 1998;130: Goessler W, Kuehnelt D, Irgolic KJ. Determination of arsenic compounds in human urine. In: Abernathy CO, Calderon RL, Chappell WR, eds. Arsenic: exposure and health effects. London: Chapman and Hall, 1997: Ritsema R, Dukan L, Roig T, Navarro I, van Leeuwen W, Oliveira N, et al. Speciation of arsenic compounds in urine by LC-ICPMS. Appl Organomet Chem 1998;12: Crecelius E, Yager J. Intercomparison of analytical methods for arsenic speciation in human urine. Environ Health Perspect 1997; 105: 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 Atom Spectrom 1997;12: Heitkemper D, Creed J, Caruso J. Speciation of arsenic in urine using high-performance liquid-chromatography with inductively coupled plasma mass-spectrometric detection. J Anal Atom Spectrom 1989;4: Sheppard BS, Shen WL, Caruso JA, Heitkemper DT, Fricke FL. Elimination of the argon chloride interference on arsenic speciation in inductively coupled plasma mass-spectrometry using ion chromatography. J Anal Atom Spectrom 1990;6: Story WC, Caruso JA, Heitkemper DT, Perkins L. Elimination of the chloride interference on the determination of arsenic using hydride generation inductively coupled plasma mass-spectrometry. J Chromatogr Sci 1992;30: Ding H, Wang JS, Dorsey JG, Caruso JA. Arsenic speciation by micellar liquid-chromatography with inductively-coupled plasmamass spectrometric detection. J Chromatogr A 1995;694: