Measurement of Arsenic Bioavailability in Soil Using a Primate Model

Stephen M. Roberts,1, William R. Weimar, J. R. T. Vinson, John W. Munson and Raymond J. Bergeron

Departments of Physiological Sciences and Medicinal Chemistry, J. Hillis Miller Health Science Center, University of Florida, Gainesville, Florida 32610

Received September 28, 2001; accepted January 17, 2002

ABSTRACT

Several studies have shown limited absorption of arsenic from soils. This has led to increased interest in including measurements of arsenic relative bioavailability from soils in the calculation of risks to human health posed by arsenic-contaminated sites. Most of the information in the literature regarding arsenic bioavailability from soils comes from studies of mining and smelter sites in the western United States. It is unclear whether these observations are relevant to other types of arsenic-contaminated sites. In order to obtain information regarding arsenic bioavailability for other types of sites, relative bioavailability of arsenic from selected soil samples was measured in a primate model. Sodium arsenate was administered to five male Cebus apella monkeys by the intravenous and oral routes, and blood, urine, and feces were collected. Pharmacokinetic behavior of arsenic after intravenous administration and the fractions of dose excreted in urine and feces after both intravenous and oral doses were consistent with previous observations in humans. Soil samples from five waste sites in Florida (one from an electrical substation, one from a wood preservative treatment site, two from pesticide sites, and one from a cattle-dip vat site) were dried and sieved. Soil doses were prepared from these samples and administered orally to the monkeys. Relative bioavailability was assessed based on urinary excretion of arsenic following the soil dose compared with excretion following an oral dose of arsenic in solution. Differences in bioavailability were observed for different sites, with relative bioavailability ranging from 10.7 ± 4.9% (mean ± standard deviation) to 24.7 ± 3.2% for the five soil samples. These observations, coupled with data in the literature, suggest limited oral bioavailability of arsenic in soils from a variety of types of arsenic-contaminated sites.

Key Words: arsenic; bioavailability; intestinal absorption; primate model; soil; urinary excretion.

Arsenic is both a naturally occurring substance and a common contaminant at hazardous waste sites in the United States. Mining and manufacturing activities, as well as the use of arsenic-containing pesticides, have resulted in a wide array of types of contaminated sites, including mine tailings, smelter facilities, cattle dip sites, electric substations, wood treatment (chromated copper arsenate) sites, pesticide treatment areas, railroad rights-of-way, golf courses, and dumps. Collectively, these sites number in the tens of thousands or more, and the management of these sites is a significant public health and economic problem.

Arsenic is classified by the U.S. Environmental Protection Agency (U.S. EPA) as a Group A carcinogen; that is, it is known to produce cancer in humans. Workers exposed to arsenic by inhalation have been found to be at increased risk of lung cancer, and a number of studies have indicated that ingestion of inorganic arsenic is associated with increased risk of cancer of the skin, bladder, and lung (Morales et al., 2000Go; NAS, 1999Go). The risk of cancer from arsenic is calculated using an estimate of arsenic exposure and an arsenic cancer slope factor. The cancer slope factor for arsenic ingestion developed by the U.S. EPA is based on a study of Taiwanese who were exposed to elevated arsenic concentrations in drinking water (Tseng et al., 1968Go, 1977). This cancer slope factor is used to estimate cancer risk from arsenic ingestion, not only from drinking water, but also from other environmental media including soils. The cancer risk that results from arsenic ingestion is dependent upon the dose of arsenic that is absorbed. Arsenic in drinking water is in a water-soluble form, and its absorption from the gastrointestinal tract is thought to be extensive. Arsenic contaminants in soils, however, may be incompletely absorbed because they are present in water-insoluble forms or because they interact with other soil constituents. Logically, the diminished absorption of arsenic from soil relative to water should be taken into consideration when evaluating the cancer risk posed by arsenic exposure. The problem lies in determining, for a given situation, the degree of reduction of arsenic absorption.

A number of studies have attempted to measure the relative bioavailability of arsenic from soils compared with water using animal models (Table 1Go). An important limitation of these studies, from the perspective of regulating arsenic-contaminated sites throughout the United States, is that most involve soil samples from mining and smelter sites. Although the data indicate diminished bioavailability from these soils, generalizing these results to other types of contaminated sites is questionable, as there is reason to suspect that differences in the type of arsenic contamination, as well as perhaps soil type, may influence arsenic bioavailability.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Literature Reports of Arsenic Relative Bioavailability in Soils
 
The need for arsenic soil bioavailability information for a broader array of contaminated sites prompted this study. A primate model was used to measure arsenic bioavailability in soil samples from sites with differing sources of arsenic contamination. The Cebus monkey was selected for this study because of our extensive experience with this species for gastrointestinal absorption studies and its demonstrated value as a model for humans in preclinical pharmacokinetic studies (Bergeron et al., 1990Go, 1999Go, 2000Go). Since the Cebus monkey has not been used previously for arsenic absorption studies, initial experiments were directed to characterizing absorption and excretion of sodium arsenate. The objective of these experiments was to ensure that the absorption and excretion of arsenic in the Cebus monkey is sufficiently similar to humans that these monkeys can serve as an effective model for bioavailability studies. Subsequent experiments used urinary excretion data to measure the oral bioavailability of arsenic in five soil samples relative to sodium arsenate in water.

MATERIALS AND METHODS

Animals and animal care.
Five adult male Cebus (Cebus apella) monkeys, 2.5–3.0 kg body weight, were purchased from Osage Regional Primates, Inc. (Osage Beach, MO). Between experiments, they were housed individually in metal cages in a climate-controlled room with a population of other monkeys. During these periods, the animals were fed standard monkey chow. During the experimental period, the animals were transferred to nonmetal metabolic cages in another environmentally controlled room. While in the metabolic cages, the monkeys were fed a low-arsenic liquid diet. The diet consisted of casein, 180 g; sucrose, 194 g; dextrin, 194 g; dextrose, 194 g; fiber (cellulose), 90 g; vitamin mix, 5 g; methionine, 5 g; banana flavor, 2 g; choline chloride, 2 g; cholesterol, 1 g; corn oil, 45 g; coconut oil, 45 g; soy lecithin, 20 g; manganese sulfate, 0.1 g; calcium carbonate, 25.4 g; potassium hydrogen phosphate, 35.8 g; magnesium sulfate, 5.2 g; and deionized water, 1350 ml. Solids and oils for the liquid diet were obtained from Bio-Serv (Frenchtown, NJ); minerals (manganese sulfate, calcium carbonate, potassium dihydrogen phosphate, and magnesium sulfate) were obtained from Fisher Scientific (Norcross, GA). The liquid diet presented to the animals was replaced daily from stocks kept refrigerated for up to 7 days after preparation. The arsenic concentration of each batch of liquid diet was measured as described below and confirmed to be below detection limits (< 1 ppb). The housing conditions, environmental enrichment program, and all procedures involving the animals were approved by the Institutional Animal Care and Use Committee.

Drugs and chemicals.
Sodium arsenate heptahydrate was purchased from Sigma Chemical Co. (St. Louis, MO). Telazol was purchased from Fort Dodge Animal Health (Fort Dodge, IA). Atropine for injection (Fugisawa USA, Deerfield, IL), Zofran (GlaxoWellcome, Inc., Research Triangle Park, NC), ketamine (Elkins Simm, Inc., Cherry Hill, NJ), and isoflurane (Abbott Labs, Abbott Park, IL) were obtained through the University of Florida Animal Resources Stores.

Soil samples.
Surface soil samples from selected contaminated sites were collected and provided by the Florida Department of Environmental Protection (FDEP). The top six inches of ground were excavated from areas known through previous sampling to contain substantial arsenic contamination (i.e., > 100 mg As/kg soil). Soil samples were either shipped to the laboratory at the University of Florida by commercial carrier or delivered directly by FDEP personnel. They were delivered in 5 x 5 gallon buckets and stored in an air-conditioned room until processing. For processing, the soils were dried for at least 3 days at 30–38°C, sieved through a 2-mm screen, and then thoroughly mixed. A 250-ml aliquot of this soil was retained for future reference, and the remainder was sieved to 250 µm or less using a number 6 screen and pan sieve shaker apparatus (Meinzer Sieve Shaker, Fisher Scientific, Norcross, GA). The 250-µm sieved soil was stored in sealed containers at room temperature until utilized. The total arsenic concentration in an aliquot of the 250-µm sieved soil was measured by the Central Chemistry Laboratory, Florida Department of Environmental Protection (Tallahassee, FL) using EPA Method 6010.

Animal dosing and sampling.
At the beginning of each experiment, monkeys were sedated with ketamine (10 mg/kg, im) combined with atropine (0.01 mg/kg, im) and weighed, and a blood sample was taken for standard health assessment. The animals were then transferred to metal-free metabolic cages and fed a low-arsenic liquid diet beginning 48 h prior to the arsenic dose. Each monkey was fasted overnight before dosing, but the liquid diet was restored 6 h after the animal was dosed and returned to its metabolic cage. Feces produced during the 24 h prior to the dose were collected and designated as baseline samples.

For experiments involving oral dosing, the animal was first pretreated with the antiemetic Zofran (0.15 mg/kg, im) and then 30 min later given a dose of the short-term anesthetic Telazol (2 mg/kg, im). While the monkey was sedated, a bladder catheter was inserted, and the contents of the bladder were collected, including a 5-ml rinse with sterile saline. This was designated as the baseline (or "0 time") urine sample. Additionally, a baseline blood sample (2 ml) was taken from the saphenous vein. Blood samples were collected in a 2-ml Vacutainer tube containing 0.2 ml buffered citrate (Becton Dickinson and Company, Rutherford, NJ). A gastric tube was placed, and a measured dose of sodium arsenate solution or soil was introduced into the stomach. Soil doses were administered as a slurry in metal-free, deionized water. Sodium arsenate was administered as a 1.0 mg As/ml solution in deionized water, and the volume was adjusted to provide a dose of 1.0 mg As/kg body weight. The gastric tube was flushed with metal-free, deionized water to ensure complete transfer of the dose to the stomach. Both the gastric tube and bladder catheters were then removed, and the animal was returned to its metabolic cage for 1 h. During this period, the animal was allowed to recover from the sedative. Any urine or feces produced during this period were collected from the metabolic cage. After removal of urine and feces from the metabolic cage, urine collection surfaces of the cage were rinsed with metal-free water to remove any residual excreted arsenic. Cage rinsate was recovered and analyzed separately from urine and feces. For purposes of assessing arsenic excretion, the arsenic present in cage rinsate was considered to have come from urine produced during the collection interval.

One hour after the oral dose, the animal was again sedated with Telazol and intubated. Anesthesia was maintained using isoflurane gas at 1.5%. An intravenous line was placed in the lower leg, and 100 ml of sterile saline was introduced by slow infusion for hydration. A bladder catheter was placed, again by standard technique, and urine samples were collected initially and every hour for the next 5 h. Blood samples (2 ml in Vacutainers with 0.2 ml buffered citrate) were collected 1, 1.5, 2, 2.5, 3, 4, 5, and 6 h after the dose. While anesthetized, supplemental heat was provided in the form of warm water blankets, and body temperature was closely monitored. At 6 h after the dose, the animal was allowed to regain consciousness and was returned to its metabolic cage, where urine and feces were collected daily over the next 4 days. As part of the collection procedure, the cage was rinsed with metal-free water as described above to ensure complete recovery of arsenic excreted in urine. At the end of the collection period, the monkey was returned to its home cage for a minimum of 2 weeks before another dosing experiment was conducted.

Initially, an experiment was conducted in which the monkeys were administered a single intravenous dose of arsenic (1 mg As [as sodium arsenate] per kg body weight in sterile saline). For this experiment, each monkey was sedated with Telazol and placed under isoflurane anesthesia as described above. Intravenous and urinary catheters were then placed. The bladder was rinsed prior to administration of the intravenous dose. The arsenic dose was introduced through the intravenous line over a period of about 5 min. Blood and urine samples were taken at 30 min and 1, 1.5, 2, and 2.5 h. Two milliliters of blood were taken at each time point, and all urine produced was collected via catheter. After 6 h, the anesthetic was withdrawn, intravenous and bladder catheters were removed, and the animals were returned to their metabolic cages after regaining consciousness. Urine and feces were then collected for 4 days.

Sample preparation.
Urine samples were collected in 1-l polycarbonate bottles containing 10 ml of 65% nitric acid and then stored in smaller, sealed polycarbonate bottles at room temperature until processed for analysis. A 1.0-ml aliquot of urine was added to 1.0 ml 65% nitric acid in a 15-ml pressure tube. The tube was sealed, placed in a 140°C oil bath for 3 h to digest, and then allowed to cool. Hydrogen peroxide (0.75 ml of a 30% solution) was then added to the tube, which was placed in a 100°C oil bath for 45–60 min while loosely capped. After cooling, the contents of the tube and five successive washes with metal-free water were transferred to a 5-ml volumetric flask for analysis.

Feces were collected and weighed. Nitric acid (65%) was then added in an amount equal to 10% of the fecal weight, and the samples were homogenized. A 3-g sample of homogenate (actual weight was recorded to the nearest 0.01 g) was added to 20 ml of 65% nitric acid. The mixture was refluxed at 100°C in an extractor for 24–48 h until the entire sample was dissolved. The sample was allowed to cool and 5 ml of hydrogen peroxide (30%) was added. Heat was reapplied, and the mixture was allowed to reflux for 1 h. After cooling, the contents of the flask and five successive washes with metal-free water were transferred to a 50-ml volumetric flask for analysis. Samples of each batch of liquid diet were processed using the same procedure as for feces.

Whole blood was centrifuged just after collection to separate plasma. The plasma layer was stored at –80°C until analyzed. A 0.25-ml aliquot of plasma was added to 1.0 ml of 65% nitric acid in a 15-ml pressure tube. As with the urine assay described above, the sample was digested in a 140°C oil bath for 3 h and allowed to cool. Hydrogen peroxide (0.75 ml) was then added to the acidified plasma, and the mixture was heated at 100°C for 45–60 min. The contents of the tube and five successive washes with metal-free water were transferred to a 5-ml volumetric flask for analysis.

Quantification of arsenic in plasma, urine, and feces.
All samples were analyzed with a Perkin-Elmer Model 5000 atomic absorption spectrophotometer using a modification of Method 7060A, Revision A, September 1994. The instrument used was a graphite furnace unit with Zeeman background correction. A sample was placed into the L'vov platform of a graphite tube via an auto sampler. Matrix modifiers and diluents were placed into the graphite tube. For urine samples, the following additions were made to the tube: 20 µl of sample/standard, 5 µl of diluent (6.5% HNO3), and 10 µl of palladium nitrate working solution (500 ppm). For fecal samples, the additions were as follows: 20 µl of sample/standard, 5 µl of diluent, 10 µl of palladium nitrate working solution, and 5 µl of magnesium nitrate/ascorbic acid solution (0.5%/1.0%). All samples were ashed at 1200–1300°C and atomized at 2200–2300°C. Feces/liquid diet samples required preinjection of the matrix modifiers and ashing to 1200°C. Blanks, duplicates, spikes, and lab control samples were run at a rate of 5%, or minimally one, per batch. The blank was run with all components minus the sample matrix. Metal-free water was added at the same volume as the sample matrix. Spikes were added at 10 ppb for urine and plasma samples and at 20 ppb for feces or food samples. Laboratory control samples were prepared in plasma, urine, or feces samples collected prior to dosing (T0 time point). Plasma and urine samples had 5 ppb arsenic added; feces samples had 25 ppb added. Spikes and lab control samples were prepared using commercially obtained arsenic standards (SPEX CertiPrep, Inc, Metuchen, NJ).

Standard curves were run at the start of each analytical batch. At least five standards from 2.5 µg/l to 100 µg/l were analyzed for the calibration, requiring a minimal correlation coefficient of 0.995. Samples outside the range were diluted and rerun. The mean detection limit (MDL) for each matrix was determined as the lowest concentration that falls within 2–5 times the calculated MDL (3.14 x standard deviation, n = 7).

Derivation of descriptive pharmacokinetic parameters.
Nonlinear regression (PCNONLIN, V4.2, Pharsight, Mountain View, CA) was used to estimate the terminal elimination rate constant (ß) for disappearance of arsenic from plasma over time. The elimination half-life (t1/2ß) was calculated from 0.693/ß. The area under the plasma concentration-versus-time curve (AUC) between time 0 and the last data point was derived using the linear trapezoidal rule. In order to estimate the AUC from time 0 to infinity, the area beyond the last time point in the study was estimated by dividing the last measurable plasma concentration by ß. Mean residence time (MRT) was calculated by dividing the first moment of the plasma concentration versus time profile (AUMC) by the AUC. The volume of distribution at steady state (Vdss) was calculated as Dose x AUMC/(AUC)2. Clearance (Cl) was derived from the Dose/AUC.

Urinary excretion data following intravenous and oral administration of sodium arsenate were used to calculate the absolute bioavailability (FA) of the oral dose using the following relationship:


where U refers to the amount of arsenic excreted in urine over the observation period, and D is the dose of arsenic. Urinary excretion data were also used to calculate the relative bioavailability of orally administered arsenic in soil versus sodium arsenate in solution. Relative bioavailability (FR) was calculated as follows:


Statistical analyses.
An analysis of variance was used to analyze the relationship between soil sample, animal, and relative bioavailability. Main effects for soil sample and animal were included in the model. Statistical significance was determined if p values were < 0.05. In the case of multiple comparisons for differences in relative bioavailability, 10 comparisons were calculated. Ten comparisons were also calculated for animal effect. To preserve the overall Type I error rate for the entire comparison, the Bonferroni adjustment was used and significance was determined for p values < 0.0005 (resulting in an overall Type I error rate of 0.05).

RESULTS

Blood samples were collected following intravenous administration of arsenic, permitting pharmacokinetic analysis. Intravenous dose data were available for four monkeys.2 Figure 1Go shows the disappearance of arsenic from plasma following the 1 mg As/kg body weight dose. Plasma concentrations were remarkably consistent among the animals, with almost superimposable plasma concentration versus time profiles. Descriptive pharmacokinetic parameters were derived from the plasma concentration data for each monkey and are shown in Table 2Go. Mean values are also presented.



View larger version (18K):
[in this window]
[in a new window]
 
FIG. 1. Plasma concentrations over time after a single intravenous dose of sodium arsenate. Each monkey received a single intravenous dose of sodium arsenate (1 mg As/kg body weight). Plasma concentrations for each animal are shown. Descriptive pharmacokinetic parameters derived from these data are presented in Table 2Go.

 

View this table:
[in this window]
[in a new window]
 
TABLE 2 Descriptive Pharmacokinetics after Intravenous Administration of Sodium Arsenate
 
Urinary and fecal excretion were also measured after the intravenous arsenic dose. Urinary excretion of arsenic was rapid, with nearly half of the arsenic dose appearing in the urine within a few hours after the dose (time course not shown). Collection of urine over 4 days recovered, on average, 66.8% ± 6.5% (± standard deviation) of the arsenic dose. As expected, only a very small fraction of the intravenous arsenic dose appeared in feces over 4 days of collection (0.5–0.6%), indicating little biliary excretion.

Following oral (intragastric) administration of the same dose of sodium arsenate to five monkeys, 48.8 ± 2.3% (mean ± standard deviation) of the dose was recovered in urine within the 4-day collection period. Most of the dose was recovered within the first 24 h after the dose (time course not shown). Consistent with extensive gastrointestinal absorption of sodium arsenate in solution, fecal excretion was low—only 1.9 ± 1.3% of the dose appeared in the feces over the 4-day recovery period. Fecal excretion was, however, generally higher than that observed with intravenous administration, indicating that absorption was not complete. An estimate of the absolute oral bioavailability of sodium arsenate solution in the monkey can be obtained by comparing the urinary excretion (cumulative percent of dose excreted over the collection period) following the oral dose with that obtained in the same animal following administration of the same dose intravenously. From such comparisons, the absolute bioavailability of sodium arsenate solution was found to be 74.4 ± 4.8% (mean ± standard deviation; n = 4). Absolute bioavailability could not be calculated for one animal because data for urinary recovery after intravenous administration were not obtained.

Funding was available for measurement of arsenic bioavailability from five soil samples. Each soil sample included in the study was obtained from a different arsenic-contaminated site, and the arsenic concentrations in the samples ranged from 101 to 743 mg/kg. For purposes of assessing relative bioavailability, an attempt was made to use an arsenic dose in soil as close as possible to the comparison dose of sodium arsenate in solution (1 mg As/kg body weight). However, it was also considered important to use soil doses that were not excessive in volume. In order to keep the dose of soil itself to 12 g or less, the arsenic in the soil doses ranged from 0.3 to 1.0 mg As/kg body weight (see Table 3Go for the arsenic dose for each soil sample).


View this table:
[in this window]
[in a new window]
 
TABLE 3 Urinary and Fecal Excretion of Arsenic Dose after Oral Administration of Contaminated Soil
 
Only a small fraction of the dose, generally less than 15%, was recovered in urine when arsenic was administered in soil (Table 3Go). Consistent with incomplete gastrointestinal absorption, the vast majority of the dose in nearly all cases was eliminated with the feces. The urinary excretion of arsenic from soil was compared with excretion following an oral dose of sodium arsenate in solution to generate relative bioavailability estimates for each soil sample in each animal (Table 3Go). Mean relative bioavailability values for the five soil samples ranged from 10.7 ± 4.9% (mean ± standard deviation) to 24.7 ± 3.2%. As expected, some variability in relative bioavailability was observed among subjects within each soil-treatment group, and the average coefficient of variation was about 39%. Although the results suggested that some of the animal subjects tended to have higher arsenic bioavailability from soils than others, differences among animals were not statistically significant, as determined through an analysis of variance. Relative bioavailabilities from the highest sample (from the cattle-dip site) and lowest sample (from pesticide site #1) were significantly different from each other. No statistically significant differences in relative bioavailability were detected between other samples.

Few of the blood samples collected after administration of arsenic in soil had concentrations above the minimum quantitation limit. Arsenic could be quantified in only one or two samples per soil per animal (data not shown), making it impossible to calculate a meaningful AUC for bioavailability measurement. Consequently, relative bioavailability was assessed based on urinary excretion only.

DISCUSSION

In developing bioavailability information relevant to human health risk assessment, it is important to assess, to the extent possible, the degree to which absorption and elimination behavior in animal models resembles that in humans. Arsenic administered intravenously to the Cebus monkey disappeared rapidly from blood, with an apparent half-life of about 1 h (Fig. 1Go). This is consistent with the initial, very rapid disappearance of arsenic from blood observed in four human subjects administered an intravenous dose of As74 (Mealey et al., 1959Go). By using a radiolabeled arsenic dose, plasma arsenic concentrations in the human subjects could be followed for 10 days, revealing a much slower rate of disappearance after the first few hours. The terminal elimination rate, which appeared after about 7 days, corresponded to a half-life of approximately 86 h. Collection of urine from these subjects over 9 days recovered 57–90% of the intravenous dose. The urinary recovery of intravenous arsenic in the Cebus monkey (66.8% on average) falls within this range. Ducoff et al. (1948) administered sodium arsenite intravenously to two subjects and collected urine and feces over the next 7 days. Their recovery of 65.7 and 59.1% of the dose in urine for the two subjects, and 0.9 and 0.5% of the dose from feces, matches quite closely the urinary and fecal recoveries of arsenic in the Cebus monkeys (66.8 ± 6.5% of the dose in urine and 0.6 ± 0.1% of the dose in feces). Urinary and fecal recoveries were also consistent with previous observations in other nonhuman primates. Vahter et al. (1995) administered arsenic intravenously to two chimps and recovered 52 and 63% of the dose in urine and 1.2% and 1.4% of the dose in feces over 4 days. Another 5.1 and 5.4% of the dose was recovered in cage wash and was presumed to reflect primarily urinary excretion. Freeman et al. (1995) administered arsenic intravenously to three cynomolgus monkeys, and recovered 76.5 ± 2.5% of the dose in urine and 3.2 ± 1.9% in feces.

Urinary and fecal excretion of sodium arsenate after oral administration in the Cebus monkey was also consistent with previous observations in humans and other primates (Bettley and O'Shea, 1975Go; Buchet et al., 1981Go; Crecelius, 1977Go; Mappes, 1977Go; Pomroy et al., 1980Go; Tam et al., 1979Go). The studies by Bettley and O'Shea (1975) and Pomroy et al. (1980) are among the few to present data for both urinary and fecal excretion of arsenic after oral administration in human volunteers. The results for three volunteers reported by Bettley and O'Shea (52% of the dose excreted in urine and <= 3.5% excreted in feces) are nearly identical to those observed here with the Cebus monkey. Dose recovery from six subjects was slightly greater in the study of Pomroy et al. (1980), with 62.3 ± 4.0% of the dose excreted in urine and 6.1 ± 2.8% of the dose eliminated in feces.

Humans methylate inorganic arsenic to monomethylarsonic acid (MMA) and dimethylarsinic acid (DMA) metabolites, which are excreted in the urine along with unmetabolized inorganic arsenic (Buchet et al., 1981Go; Crecelius, 1977Go; Vahter, 1986Go)). Chimpanzees and marmoset monkeys do not appear to methylate arsenic, although MMA and DMA are formed from hepatic metabolism of arsenate in cynomolgus and rhesus monkeys, both Old World species (Vahter et al., 1995Go; Zakharyan et al., 1996Go). The metabolism of arsenic has not been examined in the Cebus monkey, which like the marmoset is a New World species. It is possible that the hepatic metabolism of arsenic may be different in Cebus monkeys than humans. This would be an important consideration in a metabolism study and also in a study of the toxicity of arsenic. However, the focus of this study is arsenic bioavailability. Given the way in which the relative bioavailability of arsenic is operationally defined and measured, differences in arsenic metabolism, even if they exist, would not be a confounding factor. The bioavailability of arsenic, for risk assessment purposes, is evaluated in terms of total arsenic rather than the appearance in the systemic circulation of any particular form. Bioavailability is consequently independent of metabolism in this situation. To the extent that urinary and fecal excretion of total arsenic are used as means to measure bioavailability, metabolism could affect bioavailability measurement indirectly by influencing excretion. The most straightforward means of assessing this is through comparisons of total arsenic excretion behavior among species. The studies of urinary and fecal excretion of arsenic following intravenous and oral administration of sodium arsenate, described above, strongly suggest that the fundamental absorption and excretion of arsenic are the same in the Cebus monkey and humans.

The incomplete recovery of the arsenic dose in urine and feces during the experiment indicated that a substantial fraction of the dose was being retained in the body. Under such circumstances, care must be taken that the chemical does not accumulate to toxic levels with repeated dosing, and that residual chemical in the body does not affect interpretation of absorption data from a subsequent dose. With a 2-week minimum washout period between doses, we did not observe significant residual arsenic in blood, urine, or feces. That is, despite the fact that each monkey had received over time several doses of arsenic, baseline blood, urine, and fecal measurements taken before dosing were consistently below detection levels. This is not surprising. Assuming that the terminal elimination rate in the monkeys is similar to that observed in humans (ca. 86 h, as discussed above), the 3-week minimum collection and washout period allowed six elimination half-lives to pass between doses. Typically, the dosing interval was 1 month or more, corresponding to eight or more half-lives. Under these circumstances, significant accumulation would not be expected.

To minimize stress to the animal subjects, it was necessary to anesthetize them during the period of most frequent blood and urine sample collection. It was also considered important to include in the experimental protocol a period without anesthesia immediately after the dose to allow normal gastric emptying. Extensive preliminary experimentation (not presented in Results) established 1 h as the optimum interval between the dose and the beginning of sampling. When sampling was initiated more than 1 h after the dose, peak blood concentrations and urinary excretion rate were missed, and shorter intervals resulted in reduced recovery of arsenic dose, perhaps due to interference by the anesthetic with normal gastric contractions and emptying. A half-life for gastric emptying of 30 min has been reported for fasted male Red Patas monkeys (Franklin, 1977Go). If this emptying rate is applicable also to the Cebus monkeys in this study, it would suggest that most of the oral dose was emptied from the stomach before induction of anesthesia with the 1-h interval protocol. Perhaps the best argument for the acceptability of this protocol is empirical; that is, that the absorption and excretion of arsenic appear to match closely that observed in unanesthetized human volunteers, as discussed above.

Relative oral bioavailability can be measured using either blood concentration or urinary excretion data, and it was the original intent of this study to use both approaches. However, with the analytical methods employed, it was very difficult to develop an accurate characterization of the blood concentration versus time profile after oral dosing, particularly for arsenic in soils. The amount absorbed was simply too small to produce blood concentrations that could be measured at more than one or two time points. A previous study of arsenic bioavailability from soil in cynomolgus monkeys also measured arsenic in blood and urine, and reported estimates of relative bioavailability based on both approaches (Freeman et al., 1995Go). However, as we observed with Cebus monkeys, blood arsenic concentrations after the soil doses were only marginally above quantitation limits at a few time points. In this situation, the ability to capture the complete arsenic blood concentration versus time profile from the soil dose is compromised, resulting in an underestimation of absorption. This would explain why these investigators obtained a relative bioavailability estimate from blood data that was only about one-half that based on urinary excretion data. For the purposes of assessing arsenic bioavailability from soils, the use of urinary excretion data would appear to be much more reliable, at least in nonhuman primates.

The objective of this study was to begin to generate data on bioavailability of arsenic from soils covering a broader range of types of arsenic-contaminated sites than currently exists in the literature. This study offers some of the first measurements of the relative bioavailability of arsenic in soil at cattle dip, wood treatment, electrical substation, and pesticide sites. The limitations in generalizing these results are obvious. It would be inappropriate, for example, to use the single value reported here from a wood treatment facility as representative of all wood-treatment sites. There are a variety of factors (e.g., soil characteristics, arsenic formulation, manner of release of arsenic to the soil) that could conceivably affect arsenic soil bioavailability from site to site, and even different areas within a site. This is illustrated well by the recent report (Casteel et al., 2001Go) of the relative bioavailability of arsenic from five samples of soil contaminated with the arsenical herbicide PAX. All five samples were taken from residential yards at the Vasquez Boulevard and I-70 Superfund site and all were presumably contaminated with arsenic from the same product. The relative arsenic bioavailability among the five samples, as measured in a swine model, ranged from 18 to 45%.

Despite the expectation of some variability in arsenic bioavailability from specific types of sites, the results of this study are nonetheless consistent with the concept that arsenic bioavailability from soils is generally reduced compared with that from water. For the sites included in this study, the extent of reduction ranged from 4- to 10-fold (based on relative bioavailabilities ranging, on average, from about 25 to 10%). Typically, when evaluating risks from arsenic-contaminated soils, the predominant route of arsenic intake is assumed to be through incidental ingestion of soil, and the relative bioavailability of arsenic is assumed to be 100%. That is, the bioavailability of arsenic from soil is assumed to be equivalent to the bioavailability of soluble arsenic in water. Because incidental ingestion is the dominant route of exposure, any adjustment in the oral relative bioavailability from the default 100% assumption will have an essentially proportional effect on the overall dose (and risk) estimate. For example, a relative bioavailability adjustment of 4-fold (i.e., incorporation of a relative bioavailability of 0.25) reduces the estimated risk for a given concentration of arsenic in soil 4-fold. At some sites, a correction of this magnitude in the risk estimate to improve accuracy can have important economic consequences in terms of the resources required for cleanup. Consequently, even though the reduction in relative bioavailability measured in this study and others is not large, usually an order of magnitude or less, its recognition can be of enormous practical value in some situations.

There are uncertainties in the measurement of bioavailability as conducted in this and similar studies that should be acknowledged. One such uncertainty is whether the animal model used in the study serves as a sufficiently valid predictor of human response. We attempted to address this by comparing pharmacokinetic and excretion behavior of sodium arsenate in the Cebus monkey with previous studies in human volunteers. Although the results were quite similar, it is always possible that there could be species differences when arsenic is present in a soil matrix. Unfortunately, there are no reliable measurements of arsenic bioavailability from soil samples in human subjects to serve as a basis for comparison, making true validation of animal models difficult. A second uncertainty relates to the possible effect of arsenic concentration in soil on bioavailability. Existing measurements of relative bioavailability of arsenic from soils, including those presented here, use soil samples with arsenic concentrations in the hundreds of ppm. These concentrations are needed to present to the experimental animal an arsenic dose sufficient for detection in a reasonable soil volume. This raises the question of whether bioavailability measurements at these concentrations are predictive of the bioavailability in lesser-contaminated soils (i.e., with arsenic concentrations < 100 ppm.). Conceivably, arsenic bioavailability from soils could be influenced by concentration if critical interactions with soil constituents are capacity-limited, but this issue has not been well studied. In the present study, there was no significant relationship between soil arsenic concentration and bioavailability, but the range of concentrations tested (ca. 100 to 750 mg/kg) was not particularly large. Another area of uncertainty pertains to the feeding status of the animals during bioavailability measurement. In this study, monkeys received a soil dose after an overnight fast. The previous bioavailability study using monkeys (Freeman et al., 1995Go) and two studies using swine (Battelle, 1996Go; Lorenzana et al., 1996Go) also fasted the animals overnight before the dose. Other studies using the swine model (Casteel et al., 1997Go, 2001Go) did not fast the animals per se, but timed the presentation of food to the animals such that the dose was always given on an empty stomach. Administering the dose on an empty stomach no doubt aids in reducing variability associated with the bioavailability measurements, but probably does not mimic well the circumstances under which humans ingest soil. If the presence of food diminishes the bioavailability of arsenic from soil, as might be expected, then the measurements as conducted in these studies are upper-bound estimates and therefore useful for regulatory purposes. However, this is speculation, and basic information on the effects of food on bioavailability from soils is lacking.

The results presented here extend considerably the types of arsenic-contaminated sites for which quantitative soil bioavailability data are available. As with soils from other types of sites, those investigated in this study showed substantially diminished arsenic bioavailability. The consistency of this observation highlights the importance of explicit, quantitative consideration of bioavailability when assessing the risks from arsenic-contaminated soils.

ACKNOWLEDGMENTS

This study was funded by a contract with the Florida Department of Environmental Protection.

NOTES

1 To whom correspondence should be addressed at Box 110885, University of Florida, Gainesville, FL 32611. Fax: (352)392-4707. E-mail: smr{at}ufl.edu. Back

2 One of the original five monkeys purchased for the project was not in good health, and was therefore not included in the study. By the time a replacement animal was obtained, the intravenous dosing portion of the study was completed. Since obtaining intravenous data from each animal was not essential to the objectives of the study, an intravenous-dose experiment was not conducted using this animal. Back

REFERENCES

Battelle Laboratories (1996). Determination of the Bioavailability of Arsenic in Slag Following Oral Administration in Microswine. Battelle Laboratories, Columbus, OH.

Bergeron, R. J., Merriman, R. L., Olson, S. G., Wiegand, J., Bender, J., Streiff, R. R., and Weimar, W. R. (2000). Metabolism and pharmacokinetics of N1,N11-diethylnorspermine in a Cebus apella primate model. Cancer Res. 60, 4433–4439.[Abstract/Free Full Text]

Bergeron, R. J., Streiff, R. R., Wiegand, J., Vinson, J. R., Luchetta, G., Evans, K. M., Peter, H., and Jenny, H. B. (1990). A comparative evaluation of iron clearance models. Ann. N.Y. Acad. Sci. 612, 378–393.[Abstract]

Bergeron, R. J., Weimar, W. R., and Wiegand, J. (1999). Pharmacokinetics of orally administered desferrithiocin analogs in Cebus apella primates. Drug Metab. Dispos. 27, 1496–1498.[Abstract/Free Full Text]

Bettley, F. R., and O'Shea, J. A. (1975). The absorption of arsenic and its relation to carcinoma. Br. J. Dermatol. 92, 563–568.[ISI][Medline]

Buchet, J. P., Lauwerys, R., and Roels, H. (1981). Urinary excretion of inorganic arsenic and its metabolites after repeated ingestion of sodium metaarsenite by volunteers. Int. Arch. Occup. Environ. Health. 48, 111–118.[ISI][Medline]

Casteel, S. W., Brown, L. D., Dunsmore, M. E., Weis, C. P., Henningsen, G. M., Hoffman, E., Brattin, W. J., and Hammon, T. L. (1997). Relative bioavailability of arsenic in mining wastes. U.S. Environmental Protection Agency, Region 8, Denver, CO.

Casteel, S. W., Evans, T., Dunsmore, M. E., Weis, C. P., Lavelle, B., Brattin, W. J., and Hammon, T. L. (2001). Relative bioavailability of arsenic in soils from the VBI170 site. Final Report. U.S. Environmental Protection Agency, Region 8, Denver, CO.

Crecelius, E. A. (1977). Changes in the chemical speciation of arsenic following ingestion by man. Environ. Health Perspect. 19, 147–150.[ISI][Medline]

Ducoff, H. S., Neal, W. B., Straube, R. L., Jacobson, L. O., and Brues, A. M. (1948). Biological studies with arsenic76. II. Excretion and tissue localization. Proc. Soc. Exp. Biol. Med. 69, 548–554.

Ellickson, K. M., Meeker, R. J., Gallo, M. A., Buckley, B. T., and Lioy, P. J. (2001). Oral bioavailability of lead and arsenic from a NIST standard reference soil material. Arch. Environ. Contam. Toxicol. 40, 128–135.[ISI][Medline]

Franklin, R. A. (1977). The influence of gastric emptying on plasma concentrations of the analgesic, meptazinol. Br. J. Pharmacol. 59, 565–569.[Abstract]

Freeman, G. B., Johnson, J. D., Killinger, J. M., Liao, S. C., Davis, A. O., Ruby, M. V., Chaney, R. L., Lovre, S. C., and Bergstrom, P. D. (1993). Bioavailability of arsenic in soil impacted by smelter activities following oral administration in rabbits. Fundam. Appl. Toxicol. 21, 83–88.[ISI][Medline]

Freeman, G. B., Schoof, R. A., Ruby, M. V., Davis, A. O., Dill, J. A., Liao, S. C., Lapin, C. A., and Bergstrom, P. D. (1995). Bioavailability of arsenic in soil and house dust impacted by smelter activities following oral administration in cynomolgus monkeys. Fundam. Appl. Toxicol. 28, 215–222.[ISI][Medline]

Groen, K. Vaessgen, H. A., Kliest, J. J. G., de Boer, J. L. M., van Ooik, T., Timmerman, A., and Vlug, R. F. (1994). Bioavailability of inorganic arsenic from bog ore-containing soil in the dog. Environ. Health Perspect. 102, 182–184.[ISI][Medline]

Lorenzana, R. M., Duncan, B., Ketterer, M., Lowry, J., Simon, J., Dawson, M., and Poppenga, R. (1996). Bioavailability of arsenic and lead in environmental substrates. 1. Results of an oral dosing study of immature swine. EPA 910/R-96-002. U.S. Environmental Protection Agency, Region 10, Seattle, WA.

Mappes, R. (1977). (Experiments on excretion of arsenic in urine. [Author's transl.]) Int. Arch. Occup. Environ. Health. 40, 267–272.[ISI][Medline]

Mealey, J., Brownell, G. L., and Sweet, W. H. (1959). Radioarsenic in plasma, urine, normal tissues, and intracranial neoplasms. A.M.A. Arch. Neurol. Psychiatry. 81, 310–320.[Medline]

Morales, K. H., Ryan, L., Kuo, T.-L., Wu, M.-M., and Chen, C.-J. (2000). Risk of internal cancers from arsenic in drinking water. Environ. Health Perspect. 108, 655–661.[ISI][Medline]

NAS (National Academy of Science) (1999). Arsenic in Drinking Water. National Research Council, National Academy Press, Washington, DC.

Ng, J. C., Kratzmann, S. M., Qi, L., Crawley, H., Chiswell, B., and Moore, M. R. (1998). Speciation and absolute bioavailability: Risk assessment of arsenic-contaminated sites in a residential suburb in Canberra. Analyst 123, 889–892.[ISI][Medline]

Pomroy, C., Charbonneau, S. M., McCullough, R. S., and Tam, G. K. H. (1980). Human retention studies with 74As. Toxicol. Appl. Pharmacol. 53, 550–556.[ISI][Medline]

Tam, G. K. H., Charbonneau, S. M., Bryce, F., Pomroy, C., and Sandi, E. (1979). Metabolism of inorganic arsenic (74As) in humans following oral ingestion. Toxicol. Appl. Pharmacol. 50, 319–322.[ISI][Medline]

Tseng, W. P. (1977). Effects and dose-response relationships of skin cancer and blackfoot disease with arsenic. Environ. Health Perspect. 19, 109–119.[ISI][Medline]

Tseng, W. P., Chu, H. M. How, S. W., Fong, J. M., Lin C. S., and Yeh, S. (1968). Prevalence of skin cancer in an endemic area of chronic arsenicism in Taiwan. J. Natl. Cancer. Inst. 40, 453–463.[ISI][Medline]

Vahter, M. (1986) Environmental and occupational exposure to inorganic arsenic. Acta Pharmacol. Toxicol. (Copenh). 59(Suppl.), 31–34.[Medline]

Vahter, M., Couch, R., Nermell, B., and Nilsson, R. (1995). Lack of methylation of inorganic arsenic in the chimpanzee. Toxicol. Appl. Pharmacol. 133, 262–268.[ISI][Medline]

Zakharyan, R. A., Wildfang, E., and Aposhian, H. V. (1996). Enzymatic methylation of arsenic compounds. III. The marmoset and tamarin, but not the rhesus, monkeys are deficient in methyltransferases that methylate inorganic arsenic. Toxicol. Appl. Pharmacol. 140, 77–84.[ISI]





This Article
Abstract
FREE Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (6)
Disclaimer
Request Permissions
Google Scholar
Articles by Roberts, S. M.
Articles by Bergeron, R. J.
PubMed
PubMed Citation
Articles by Roberts, S. M.
Articles by Bergeron, R. J.