|Home | About | Journals | Submit | Contact Us | Français|
Arsenic (As) is a toxic trace element found in groundwater due to natural and industrial processes. Exposure has been linked to cancers of the bladder, lungs, skin, kidneys, nasal passages, liver, and the prostate. Arsenic in drinking water is a problem in many countries, notably Bangladesh and Taiwan. The purpose of this research was to utilize binding isotherms, a simulated gastrointestinal (GI) model, and the adult Hydra bioassay to evaluate ferrihydrite’s potential to bind As and serve as a potential enterosorbent for As found in drinking water. A variety of clay minerals and synthesized iron oxides including ferrihydrite were screened for their ability to bind As(III), as sodium arsenite, and As(V), as sodium arsenate. After ferrihydrite was demonstrated to be the most effective sorbent for both As species, adsorption isotherms were performed. All isotherm data were fit to the Langmuir equation to determine adsorption capacity (Qmax). Ferrihydrite bound 96% of As(III) and 97% of As(V) in the screening studies and had a Qmax of 1.288 mol/kg for As(III) and 0.744 mol/kg for As(V). Using a simulated GI model, ferrihydrite was found to effectively adsorb As(V) and As(III) in the stomach and intestine. Ferrihydrite (0.25% w/w) protected adult hydra at levels up to 200 times the minimal effective concentration (MEC) for As(III) and up to 2.5 times the MEC for As(V). These experiments confirm that ferrihydrite is a high capacity sorbent of As, and that it is effective at removing As in a simulated GI model. These results suggest that ferrihydrite could be used as a potential enterosorbent for As found in drinking water. Future work will focus on verifying ferrihydrite’s safety and efficacy in vivo.
Arsenic (As) is a toxic trace element found in groundwater due to natural and industrial processes. Health effects of chronic As exposure from groundwater include skin lesions, skin cancer, bladder cancer, kidney cancer, lung cancer, neurological effects, hypertension and cardiovascular disease, pulmonary disease, peripheral vascular disease, and diabetes mellitus (1). Inorganic As, the most common form of As in groundwater, is usually found in one of two forms, arsenate or arsenite, hereafter referred to as As(V) and As(III), respectively, with As(V) being more prevalent in oxidized environments and As(III) more common in reduced environments like those found in groundwater. Relatively slow redox transformations allow both species to commonly be found in both environments (2). The current WHO guideline and EPA legal limit for As is 10 ug/L in drinking water.
Surface adsorption is important to As mobility and bioavailability in the environment. Iron oxides and oxy-hydroxides, particularly ferrihydrite (also referred to as hydrous ferric oxide) have a high capacity for adsorption of both As(V) and As(III). Ferrihydrite’s adsorption capacity for both arsenic species is higher than more crystalline iron oxides and oxy-hydroxides like goethite and soil minerals such as kaolinite, montmorillonite, and illite (3, 4). The As binding reaction with ferrihydrite is pH dependent with As(V) being adsorbed in larger quantities at lower pHs, and As(III) at higher pHs. The two species are equally adsorbed at approximately pH 6–7.5 (5). At higher pHs, As(V) was found to steadily decrease in amount adsorbed while As(III) bound remained high at increasing pH’s up to about 10 (5).
Geophagy, the intentional consumption of earthy substances (e.g. clay), has been observed in animals and humans from numerous continents (6). Incidental consumption of soil has been investigated as a potential source of As exposure, particularly in children (7–9). Currently As in soil is assessed by using total digests of soils and using As concentration values to determine risks of exposure from contaminated soils (10). Recent research, however, suggests that arsenic found in soil has limited bioavailability (11–13) and, importantly, studies on arsenate loaded ferrihydrite showed that over 95% of the arsenate remained adsorbed to ferrihydrite throughout a simulated GI model (14).
A current strategy for reducing dietary exposure to toxicants such as mycotoxins is the addition of sorbents or binding agents in the diet. These sorbents when added to food or feed serve to sorb the toxins in the GI tract and thus reduce the bioavailability of the toxin (15–18). Previous work has shown that NovaSil clay, a calcium montmorillonite, can reduce the adverse effects of exposure to aflatoxin in a variety of animal models. NovaSil and similar montmorillonites and smectite clays have been shown to protect against the adverse effects of dietary aflatoxins in a variety of animals including rodents, chicks, turkey poults, ducklings, lambs, pigs, mink and trout (reviewed in 19). More recently, NovaSil has been shown to be safe and e ffective in an aflatoxin-exposed human population in the West African country of Ghana (20, 21). The fact that much of the As adsorbed to contaminated soil is limited in its bioavailability suggests that soil minerals or iron oxides have the potential to protect humans from As found in drinking water by using them as enterosorbents. To our knowledge, the potential use of soil minerals or iron oxides and iron oxy-hydroxides as enterosorbents for protection from As has not been tested. Given the success of a similar strategy for mycotoxins, we feel it warrants investigation. For this reason, the objectives of this work were to identify the best potential binding agent for As and to evaluate its potential use as an enterosorbent for As in vitro.
All chemicals were purchased from Sigma Chemical Co. (St. Louis, MO) and were at least ACS reagent grade. Goethite was obtained from Sigma, while kaolinite, muscovite, clinoptilolite, attapulgite, halloysite, and SWy-2 were obtained from the Clay Mineral Repository (Purdue University, Indiana). Low pH montmorillonite (LPHM) and NovaSil were obtained from BASF corporation, Jackson, MS. Goethite, 2-line ferrihydrite (hereafter referred as ferrihydrite), and magnetite were prepared according to Schwertmann and Cornell, 2000 (22). All iron oxides were air dried in a fume hood and ground and sieved to < 100 µm. Iron oxides were verified by X-ray diffraction, and their surface areas were measured after degassing at 110°C for 2 hours (5) using a multi-point BET analysis with N2 adsorption on a Micromeritics ASAP 2020. CP-LPHM was prepared according to Lemke et al. (23). SWy-2 was exchanged at 100% of its CEC with L-cysteine ethyl ester, L-cysteine dimethyl ester, and thiamine to make SW-CYSTE, SW-CYSTI, and SW-THIAM, respectively, using previously established methods (24).
Stock solutions of 9.375 and 37.5 ppm As(V) and As(III), respectively, were prepared in water from sodium metarsenite and sodium arsenate, respectively. The final pH of each solution was adjusted to 7. Into 15 mL polypropylene tubes, 4 mL of the stock solution of each specie of arsenic was added along with 1 mL of a 5 mg/mL slurry of sorbent for a final concentration of 1 mg/mL sorbent and 7.5 ppm As(V) or 30 ppm As(III). The final concentrations were based on previous work showing that 1 mg/mL ferrihydrite in a total of 5 mL solution bound 98% of 7.5 ppm As(V) in water (25) and from Raven et al. (5) who found the adsorption maxima of As(III) on ferrihydrite to be nearly 4 times higher than As(V) at higher pH values (5). The tubes were then agitated at 1000 rpm on a shaker at 25°C for 4 hours. They were then centrifuged at 2500 rpm, and 2.5 mL of the supernatant was saved for arsenic analysis. Each test was done in triplicate. All As analysis was done by inductively-coupled plasma-optical emission spectroscopy (ICP-OES).
Sample extracts were diluted 1:1 with a solution containing deionized water, 2% nitric acid (v/v) and 10% hydrochloric acid (v/v) in order to match the sample matrix to the standards used to calibrate the instrument. Samples were then analyzed for As and Iron (Fe) using a Spectro CirOS axial ICP-OES (Spectro AI, Fitchburg, MA) employing a modified version of USEPA Method 200.7 (26). The diluent was analyzed as an independent sample, which showed no traces of As or Fe at the levels measured in the samples. The instrument was calibrated using a series of standards (As 100 ppm, Fe 500 ppm) at approximately twice the concentration in solution observed in the highest test sample for As and above all Fe samples. The calibration was verified against independent standards and the calibration blank before the analytical run, approximately every ten samples during the run, and at the end of the analytical run. Ytterbium was employed as an internal standard.
Isotherms were run using modifications of methods previously reported (27, 5). Stock solutions of 125 ppm As(III) and As(V) were prepared in water from sodium metarsenite and sodium arsenate, respectively. The final pH of each solution was adjusted to 7. From these stock solutions of As(III) or As(V), a series of dilutions were added to 15 mL polypropylene tubes along with 1 mL of a 5 mg/mL suspension of sorbent for a final volume of 5 mL and final sorbent concentration of 1 mg/mL. Final concentrations of the respective arsenic species were 100, 50, 25, 10, 5, 2.5, and 1 ppm. Each dilution was prepared in triplicate. The tubes were then agitated at 1000 rpm on a shaker at 25°C for 4 hours. They were then centrifuged at 2500 rpm for 20 min, and 2.5 mL of the supernatant was saved for arsenic analysis via ICP-OES. . Binding data were fit to the Langmuir equation using Table 2 Curve software (27).
The adult Hydra bioassay was carried out as previously described (28, 29). Hydra vulgaris were obtained as a gift from E. Marshall Johnson, Jefferson Medical College (Philadelphia, PA) and were cultured in shallow trays containing media consisting of 1.0 mM calcium chloride, 0.45 mM TES buffer, and 0.012 mM EDTA, pH 7 at 18°C. These conditions ensure asexual reproduction of Hydra, which aids in the reliability and reproducibility of this assay. For assays, only adult Hydra were used. For each concentration of As(III) or As(V), 4.0 mL of media, with or without As, was placed in a small Petri dish at 18°C; all assays with As were done without EDTA in the media. Adult Hydra were observed for signs of toxicity at 0, 4, 20, 28, 44, 68 and 92 hours. The tulip stage has been validated as the toxic endpoint of the assay (28,29). The minimal effective concentration (MEC) at 92 h was determined for As(III) and As(V) through a series of assays, consisting of three phases. The first phase involved exposing adult Hydra to whole log concentrations of each As specie to determine the range of toxicity apparent within 92 h. The lowest concentration of As resulting in the toxic endpoint, MEC, was carried forward to the second phase of testing. In phase II of testing, the highest and lowest concentrations of As were within one log unit of one another. The MEC from the second phase of testing was used as the highest concentration in the phase III testing. The results from the phase III test were used to confirm those of the previous assays. To evaluate ferrihydrite’s ability to protect Hydra, 0.25% ferrihydrite was added to each dish 24 hours prior to the addition of Hydra, and the assay was carried out as previously indicated.
As adsorption on ferrihydrite was evaluated in a simulated GI tract. We used a previously described method with slight modifications as follows (14, 30). To simulate the binding of As to ferrihydrite in the stomach, 1.5g of ferrihydrite was added to 600 mL of water in a water-jacketed beaker at 37°C containing 0, 10, 25, 50 or 100 ppm As(V) or As(III) at pH 1.8 using concentrated HCl. The solution was stirred with an overhead stirrer at approximately 100 rpm while argon was bubbled through the solution. The total time for the stomach phase was 2 hours. To simulate the conditions in the intestine, the pH of the solution was raised to 6.5 by the addition of saturated NaHCO3. The total time in the intestine phase was 4 hours with the total time for both phases equaling 6 hours. Aliquots (40 mL) were taken at the end of each phase, centrifuged at 10,000 rpm for 10 minutes, and the supernatant was filtered through a 0.45 µm nylon filter and analyzed for Fe and As by ICP-OES.
Ferrihydrite proved to be the most effective sorbent in the screening experiments, binding 95.9% and 97.2% of the As(III) and As(V) in solution (Figure 1). Attapulgite and clinoptilolite both have very large surface areas with small pores, but neither is an effective sorbent for As(III) or As(V) when compared to the iron oxides and ferrihydrite. The two montmorillonite minerals, NovaSil and SWy-2, adsorbed virtually none of the As(V), but bound 9.93% and 9.34%, respectively of the As(III). The exchanging of SWy-2 with sulfur containing organic cations greatly increased the adsorption of both As species. SW-CYSTE, SW-CSYSTI, and SW-THIAM all showed a ffinity for As(III) and As(V) suggesting that they could be effective in removing As from water. They have previously been shown to be effective at removing heavy metals from water including mercury (24). However, this is the first report of these types of clay-based composites having efficacy for As. The exchanging of low pH montmorillonite (LPHM) with cetyl pyridinium (CP) increased the binding of As(V) when compared to the other montmorillonites. Since we exceeded a 1:1 stoichiometry for exchange, this effect could be due to an excess of the organic cations on the clay surface which have been shown to result in a positively charged clay. The positively charged clay may then attract As(V), which would be in the H2AsO4− or HAsO42− state at neutral pH. As(III) exists as H3AsO3 at neutral pH which could explain CP-LPHM’s lack of affinity for As(III). It was clear from these screening experiments that ferrihydrite had a much higher adsorption capacity for both As species than any of the other sorbents tested, including two goethites and magnetite.
Binding isotherms for As(III) and As(V) further illustrated ferrihydrite’s ability to adsorb As (Figure 2). Ferrihydrite had a maximum adsorption capacity (Qmax) and affinity (Kd) of 1.288 and 2.7×103 mol/kg and 0.745 and 5.27×104 mol/kg for As(III) and As(V), respectively. Ferrihydrite’s high capacity for As adsorption has been demonstrated previously (5). As(V) and As(III) both form innersphere (or bidentate) complexes on the surface of ferrihydrite (31–34). The large binding capacity of ferrihydrite is due to a high surface area, calculated as 273 m2 /g by multi-point BET. This finding is consistent with previous characterization of this material indicating that the surface area of synthetic ferrihydrite ranges between 200–320 m2 /g (22).
Due to the pH dependent nature of the As-Fe binding reaction, As binding to ferrihydrite was demonstrated using a simulated GI model. This GI model has been used to evaluate the potential bioavailability of As from contaminated soil and As loaded ferrihydrite (14, 30). Moreover, it has been positively correlated with in vivo results in pigs (30). In the present study, experiments were conducted to evaluate ferrihydrite’s ability to sorb As from the GI tract to simulate an exposure to contaminated drinking water with subsequent, or prior treatment with ferrihydrite as an enterosorbent. The results showed that ferrihydrite retains significant affinity for both As(III) and As(V) with observed maximum binding of 0.212 mol/kg and 0.194 mol/kg, respectively, in the simulated stomach (Figure 3). These values were lower than what was found in isotherms which could be a result of the optimal pH for As(V) binding being around 4, while the optimal pH for As(III) is above 7 (5). Small changes in sorption could result from As(III) being oxidize to As(V) by the conditions of the simulated stomach. However, oxidation of As(III) in the simulated stomach (pH 1.8 under anoxic conditions) would be unlikely given the duration of our study, i.e., 2 hr. Nakazawa and Hareyama noticed no observable As(III) oxidation in aqueous media at pH 1.8 after 5 days(35). Observed maximum bound values for As binding to ferrihydrite in the simulated intestine were 0.335 mol/kg and 0.266 mol/kg for As(III) and As(V), respectively (Figure 3). Raven et al. showed that ferrihydrite has similar affinities for both As(III) and As(V) from pH 6–7.5. These values are also similar to those found for mycotoxin binders which have been shown to be effective in vivo (15, 16).
Along with As, the quantity of Fe that passed through the 0.45 µm filter was measured. Similarly to Beak et al (14), we chose to use a 0.45 µm filter to mimic the 0.5 µm size cutoff of particles capable of entering epithelial cells. This would approximate the Fe exposure to cells that is either dissolved Fe (likely in the form of Fe3+) or colloidal ferrihydrite remaining in suspension after centrifugation. The levels of Fe in the simulated stomach decreased with increasing As(V) concentration, similar to what was previously reported (14); however, there was no clear trend of Fe concentration with respect to increasing As(III) concentration (Figure 4 and Figure 5). The maximum Fe levels seen in the stomach phase reached 7.57 ppm and 8.97 ppm with initial As(V) concentration of 10 ppm and initial As(III) concentration of 10 ppm, respectively. Concentrations of Fe seen in the simulated intestine were very low for all concentrations tested and for both As(III) and As(V). The highest Fe concentration measured in the intestine was 0.12 ppm. The recommended daily intake of Fe is 18 mg, with the maximum tolerable daily Fe intake being 40 mg for children and 45 mg for adults (36). The maximum expected Fe available for absorption from each test with a total of 1.5g of ferrihydrite was 5.4 mg. This dose (1.5g) of ferrihydrite was chosen for the GI model based on prior work with NovaSil clay in an aflatoxin-exposed group from Ghana (20). This ferrihydrite level throughout the GI model experiments represented the dose if enterosorption therapy is given twice daily for a total of 3g/day. Any enterosorption therapy including ferrihydrite would have to be closely monitored, as some individuals are sensitive to Fe or suffer from hemochromatosis which could cause a toxic buildup of excess iron (37).
Ferrihydrite was also evaluated for its ability to protect a small aquatic organism, Hydra vulgaris, that is very sensitive to As. The genetic homogeneity of this organism has been strictly maintained in culture to facilitate the reproducibility of the MEC for diverse toxins including chlorinated phenols (28), heavy metals (38), estrogenic compounds (39), and organophosphate nerve agents (29). The body wall of Hydra consists of a trilaminar structure and a hollow tube called the gastric cavity or gut. The Hydra gut contains a mouth (hypostome) and aboral pore that expels digested materials and fluids which is in equilibrium with the surrounding media and toxins. The Hydra assay has been reported to accurately predict the safety and efficacy of potential enterosorbents (40, 41) prior to animal studies and has been utilized along with an in vitro GI model (42) for screening purposes.
The toxic endpoint of the Hydra bioassay has been validated as the tulip stage, which represents the point at which the Hydra will not recover if placed in control media. The minimal effective concentration (MEC) for As(V) and As(III) were 90 ppm and 1 ppm, respectively (Figure 6). The addition of 0.25% w/w ferrihydrite protected the Hydra at As levels of 250 ppm and 200 ppm for As(V) and As(III), respectively. Additionally, ferrihydrite was not toxic to the hydra, unlike other sorbents that are lethal to this sensitive organism (40,41). Ferrihydrite also protected the Hydra from As toxicity at levels that were predicted from isothermal analysis.
The current treatments for arsenic from either an acute or chronic exposure include chelation therapy with analogs of dimercaprol (BAL), oral binders such as charcoal, or agents that promote methylation and elimination of As via the urine. Other than remediating current water sources or finding alternative sources, no current treatment exists for reducing exposure to As from drinking water (43). A previous study found no difference in placebo and groups given 2,3-dimercaptosuccinic acid for treatment of chronic arsenicosis from drinking water (44). A recent toxicity profile for As by ATSDR suggested “phosphate” binders and compounds with sulfhydryl groups need further investigation for their potential as treatment for As toxicity (43). Our work demonstrates that ferrihydrite has a high capacity for both species of inorganic As, was effective in a simulated GI model, and protects a susceptible organism from As toxicity. Previous work that focused on evaluating As bioavailability from contaminated soil, suggests limited bioavailability from the soil, possibly due in part to surface adsorption to soil minerals (12, 13). Concentrations of As found in contaminated wells from Bandgladesh and Taiwan have been shown to range from 50–6700 ppb (45); our tests demonstrate that ferrihydrite may offer protection from As toxicity at even higher levels of contamination. Our work shows that ferrihydrite could offer protection from As, either long or short term, but future work is needed to verify safety and efficacy in vivo. Studies are ongoing to evaluate ferrihydrite’s safety and efficacy in a rodent model.
This research was supported by NIEHS P42-ES04917.