Environ. Sci. Technol. 2006, 40, 2812-2818
Comparison of Soil Pb in Vitro Bioaccessibility and in Vivo Bioavailability with Pb Pools from a Sequential Soil Extraction B E R N D M A R S C H N E R , * ,† P E T E R W E L G E , ‡,§ A L F O N S H A C K , ‡,| JU ¨ RGEN WITTSIEPE,‡ AND MICHAEL WILHELM‡ Department of Soil Science and Soil Ecology, Geographical Institute, Ruhr-University Bochum, Universitaetsstrasse 150, 44780 Bochum, Germany, and Department of Hygiene, Social and Environmental Medicine, Ruhr-University Bochum, Universitaetsstrasse 150, 44780 Bochum, Germany
In this study, 15 soils ranging in Pb content from 32 to 6330 mg kg-1 were subjected to in vitro gastrointestinal extractions with and without added powdered milk. Before and after treatment, Pb in the soils was fractionated according to a 7-step sequential extraction procedure. A subset of five soils and Pb acetate was used for a minipig dosing study. The amount of bioaccessible Pb determined with the in vitro system ranged from 3 to 20% without powdered milk and from 11 to 56% with powdered milk. The higher bioaccessibility of Pb in the in vitro model with addition of powdered milk was related to a depletion of Pb in the organic Pb pool and indicates that soluble milk constituents compete with soil organic ligands for Pb. The absolute and relative bioavailabilities of Pb in the minipig dosing experiment were not related to bioaccessible Pb determined in any of the two in vitro systems. However, relative bioavailabilities in liver, kidney, and total uptake were highly correlated to Pb in the third fraction of the sequential extraction that is attributed to easily reducible Mn oxides. These results indicate that reductive processes in the intestine may be more relevant for Pb absorption than the initial solubilization in the acidic stomach.
Introduction Lead is one of the most widespread environmental hazards for children’s health and may have adverse effects on cognitive development if blood levels are above 10 µg dL-1. The risk associated with the ingestion of Pb contaminated dusts and soils is a major concern for European and U.S. legislation. In Germany, the threshold values for such Pb hazards in soils from children playgrounds (200 mg kg-1) and residential areas (400 mg kg-1) were derived from epidemiological studies that showed a relationship between * Corresponding author e-mail:
[email protected]; phone: +49-234-3222108; fax: +49-234-3214469. † Department of Soil Science and Soil Ecology, Geographical Institute. ‡ Department of Hygiene, Social and Environmental Medicine. § Present address: BG Research Institute for Occupational Medicine (BGFA) at Ruhr-University Bochum, Buerkle-de-la-Camp-Platz 1, 44789 Bochum, Germany. | Hessian State Health Office, Wolframstr. 33, 35683 Dillenburg, Germany. 2812
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 8, 2006
topsoil Pb-concentrations and blood Pb levels in children (1). The U.S. Environmental Protection Agency has based its definition of hazardous Pb levels in soils of playgrounds (400 mg kg-1) and residential gardens (1200 mg kg-1) on the assumption that 30% of ingested soil Pb is bioavailable (2). This assumption seems appropriate, since an isotope study with adults has shown that 26% of Pb was absorbed from a contaminated soil if swallowed on an empty stomach, while the uptake was only 2.5% if the soil was ingested after a meal (3). In comparison, uptake of ingested soluble Pb-salts ranges from 45 to 71% after 12 h of fasting and it is reduced to 3-8% when administered with meals (4). In rat feeding experiments the absolute bioavailability of Pb in soils is generally lower and varies from 0.7% to 3.0% (5, 6). In most other studies, relative rather than absolute bioavailabilities are determined, relating Pb concentrations in blood or selected organs from animals dosed with contaminated soils to the respective Pb levels in animals dosed with a soluble Pb-salt. Depending on the experimental conditions, soils, animals, or analyzed organs, these relative bioavailabilities vary greatly (6-9). In an extensive single study with 19 different soils fed to juvenile swines, Schroder et al. (7) report relative bioavailabilities of Pb ranging from 1 to 87% in blood or 1 to 94% in bone, thus indicating that soil properties or Pb mineralogy greatly influence Pb uptake. Since such in vivo studies are costly, less expensive extraction methods have been developed for the estimation of the bioavailability of metals and other contaminants (6, 7, 10-12). These methods generally mimic the biochemical and physical environment of the gastrointestinal system and the amount of solubilized contaminant is termed “bioaccessible” since it is assumed to be available for absorption. Most of these so-called in vitro gastrointestinal (IVG) methods are two-step extractions with an acidic gastric phase followed by an intestinal phase at near neutral pH. In a comparative study using five different methods for the determination of bioaccessible As, Cd, and Pb it was shown that the pH of the gastric phase largely determined the amount of extractable metals in the in vitro systems (13). While the presence of dough reduced the bioaccessible fraction of Pb in the gastric and intestinal extracts of the system used by Schroder et al. (7), the presence of powdered milk increased heavy metal solubility in the system developed by Hack and Selenka (13, 14), which has become the standardized tool for the assessment of heavy metal bioaccessibility in Germany (15). The bioaccessibility of Pb determined with any of these systems is generally in the range of 15-70% of soil Pb and thus much higher than absolute bioavailabilities determined with animal models. Only a few studies have directly compared in vivo bioavailability with in vitro bioaccessibility. Schroder et al. (7) report good correlations between relative Pb bioavailability in various organs of juvenile swines and Pb extracted only with a gastric phase system. Similarily, the positive effects of soil P amendments on Pb bioavailability in rats were reflected in reduced bioaccessibilities determined with only the gastric phase of an in vitro system (9). As mentioned above, the high variability in bioavailability or bioaccessibility is generally attributed to differences in soil properties and Pb mineralogy. There is some evidence that Pb carbonates and Pb associated with Fe- or Mn-oxides are more bioavailable than Pb in other mineral forms such as PbS or lead phosphates (7, 9, 16, 17), but so far no study has provided quantitative data to support these relationships. In the past, various sequential extraction procedures have been developed to characterize heavy metals in soils. The 10.1021/es051617p CCC: $33.50
2006 American Chemical Society Published on Web 03/18/2006
most widely used methods are the 5-step procedure of Tessier et al. (18) and the modified 4-step version of the European Community Bureau of Reference (19). In Germany, Zeien and Bru ¨ mmer (20) developed a 7-step procedure based on extraction methods for soil minerals. Its main distinguishing feature from the other procedures is that it differentiates among three fractions of reducible metals and assigns them to associations with Mn-oxides, amorphous and crystalline Fe-oxides. The method is quite specific and reproducible (21) and has been used in various studies (22-24). All these methods do not allow the determination of certain binding forms, as the fractions are only operationally defined by the extractants and artifacts may result from redistribution of metals after solubilization (25, 26). Still, these procedures are a useful tool to study qualitative differences in heavy metal content of soils. For the current study, such a tool was used to investigate the relationships between these operationally defined soil Pb fractions and the bioaccessibility and bioavailability of Pb determined with IVG and in vivo test systems. It was expected that mainly exchangeable and organically bound Pb would be solubilized in the IVG system and that the sequential extraction procedure might therefore provide a fairly simple tool for risk assessment of Pb contaminated soils.
Experimental Section Soils. The soils were obtained from all over Germany. The samples were selected for being representative of a wide range of large-scale contaminations occurring in urban areas and around former mine or smelting industries. All samples were air-dried and sieved to
