Health impact assessment by ingestion of polluted soil/sediment

Potentially toxic elements (PTEs) pose a threat to human health as they can easily enter the human body via ingestion of polluted soil/sediment. In order to estimate the bioavailability and assess the health impact on people, measurement of the oral bioaccessibility of a pollutant is crucial. Various laboratory based in vitro tests which mimic human gastrointestinal tract conditions can be used. In order to set up the method for analysing the bioaccessibility of pollutants in soil samples in the Laboratory for the analysis of geological materials at the Department of Mineralogy, Petrology and Mineral resources (Faculty of Mining, Geology and Petroleum Engineering, University of Zagreb - RGNF), with regards to the available equipment, an orientation survey was carried out in collaboration with the Department of Biochemical Engineering (Faculty of Food Technology and Biotechnology, University of Zagreb - PBF). The digestion of two different samples in synthetic fluids (gastric and intestinal fluid) was performed simultaneously at the RGNF laboratory and PBF laboratory under different extraction conditions according to each laboratory’s ability. Prior to the analysis of bioaccessibility, detailed mineralogical and chemical characterization of the samples was performed. The comparison of two experiments showed that there is a relatively good correlation between concentrations obtained after digestion of the samples in different laboratories, under different conditions. As a result of this study, an efficient and relatively inexpensive method for determining bioaccessibility was set up at Faculty of Mining, Geology and Petroleum Engineering, which makes this kind of test more accessible and enables a new approach in risk assessment


Introduction
One of the major exposure routes for potentially toxic elements (PTEs) in soil/sediment is through ingestion.A crucial step in the estimation of bioavailability and evaluation of the risk to human health is the measurement of the oral bioaccessibility of a pollutant, defined as the fraction that is soluble in the gastrointestinal tract and is available for absorption (Paustenbach, 2000).Bioaccessibility is commonly confused in the literature with bioavailability, which is defined by toxicologists to be the fraction of an administered dose of a toxicant that is absorbed via an exposure route, reaches the bloodstream, and is transported in the body to a site of toxicological action (Plumlee et al., 2006).A number of laboratory based in vitro tests which mimic the biochemical conditions in the human gastrointestinal tract are used to assess the degree of metal solubility i.e. oral bioaccessibility (DIN, 2000, Rodriguez et al., 1999, Hamel et al., 1999, Hamel, 1998, Medlin, 1997, Rotard et al., 1995, Ruby et al., 1993, Molly et al., 1993, XII, 1990) (reviewed in (Dean and Ma, 2007, Intawongse and Dean, 2006, Wragg and Cave, 2003).The human gastrointestinal tract consists of a number of compartments where ingested soil/sediment undergoes a series of reactions in which fluid composition, pH and reaction time all vary.The analogue digestive tract methods use extractions that mimic a combination of one or more of the compartments (Oomen et al., 2002).These methods can be divided into two types: 1) the batch test, where the sample and reagents are reacted in a single container with additional chemicals being added and samples removed as The Mining-Geology-Petroleum Engineering Bulletin, 2016, pp.29-39 © The Author(s), DOI: 10.17794/rgn.2016.2.3 stated in each specified protocol, and 2) the flow through a reactor system, in which the reaction takes place under flowing conditions, designed to simulate the actual transit characteristics of the gastrointestinal tract (Wragg and Cave, 2003) (Rodriguez et al., 1999); the Dutch National Institute for Public Health and the Environment method (RIVM) (Versantvoort and Rompelberg, 2004) and the relative bioaccessibility leaching procedure (RBALP) (Drexler andBrattin, 2007, Drexler, 1999).The Bioaccessibility Research Group of Europe (BARGE) developed in vitro physiologically based ingestion bioaccessibility procedure for soils, based on the RIVM methodology, known as the Unified BARGE Method (UBM) (Wragg et al., 2011).In vitro gastrointestinal protocols provide a rapid and inexpensive means to determine the bioaccessibility of a given potentially toxic element present in soil and represent appropriate alternatives to animal testing.In order to set up the method for analysing bioaccessibility of pollutants in soil samples in the Laboratory for the analysis of geological materials at the Department of Mineralogy, Petrology and Mineral resources (Faculty of Mining, Geology and Petroleum Engineering, University of Zagreb -RGNF), with regards to the available equipment, an orientation survey was carried out in collaboration with the Department of Biochemical Engineering (Faculty of Food Technology and Biotechnology, University of Zagreb -PBF).The digestion of two different samples in synthetic fluids (gastric and intestinal fluid) was performed simultaneously at the RGNF laboratory and the PBF laboratory under different extraction conditions according to each laboratory's ability.Prior to the analysis of bioaccessibility, detailed mineralogical and chemical characterization of the samples was performed.

Materials and Methods
Two different types of samples were used in this study: (1) residential soil; and (2) tailings.The analysed samples were collected in the area of an abandoned lead-zinc-vanadium mine Berg Aukas in Namibia.Samples were air dried and sieved to fraction < 250 μm, because this particle size is representative of that which adheres to children's hands (Agency, 2008).Mineral content of the samples was determined using X-ray diffraction (Philips diffractometer PW1710).In addition to the body fluids extraction, pseudo-total metal content (aqua regia extraction) as well as sequential extraction were performed.All extracts were analysed using flame atomic absorption spectroscopy (AAnalyst 700 Perkin Elmer).Samples with analytical methods performed are listed in Table 1.

Aqua regia extraction
The pseudo-total Cd, Cu, Pb and Zn contents were obtained by a standardized aqua regia extraction protocol according to ISO Standard 11466 (ISO, 1995).Aqua regia soluble metal concentrations in analysed soil and tailings were compared with the Canadian Soil Quality Guidelines (CCME, 1999).

Sequential extraction
Different binding sites of metals in samples were analysed using sequential extraction, whereby a series of single reagents areused to extract operationally-defined phases (the selectivity depends on such factors as chemicals employed, the time and nature of contact, and the sample to volume ratio) in a defined sequence from 1 g of the sample.BCR (Community Bureau of Reference) sequential extraction procedure (Rauret et al., 2001) was used to give four fractions (Table 2): (1) acid (bound to carbonates); ( 2) reducible (bound to iron and manganese oxides, hydroxides and oxyhydroxides); ( 3) oxidisable (bound to organic matter and/or sulphides); and ( 4) residual (bound to chemically resistant minerals).

Body fluids extractions
Simulated gastric and intestinal fluids were prepared according to Kos and Goreta (2000), with some modifications.Briefly, simulated gastric fluid was prepared by dissolving pepsin (AppliChem, Germany) (3 g/L) in a sodium chloride solution (0.5%) adjusting the pH to 1 ± 0.2 with concentrated HCl.Simulated intestinal fluid was prepared by suspending pancreatin (AppliChem, Germany) (1 g/L) and bile salts (Difco, USA) (3 mg/mL ox gall) in a sodium chloride solution (0.5%) and the pH was adjusted to 8.0 with 30% NaOH.Both gastric and intestinal fluids were prepared fresh.Digestion in synthetic fluids was performed simultaneously at the RGNF laboratory and the PBF laboratory under different extraction conditions, as described below and shown in Table 3, along with an overview of the different in vitro digestion models described in the literature.For each sample, two extracts were collected: one at the end of the gastric phase and another at the end of the gastro-intestinal phase.In this study, the gastric phase (G) is a digestive extract collected after 1 hour of agitation with simulated gastric fluid.The gastro-intestinal phase (GI) is a digestive extract collected after 1 hour of agitation with gastric fluid followed by 4 hours agitation with simulated intestinal fluid.The digestion was started by suspending 0.5 g of dry soil sample in 18 mL of gastric fluid.This mixture was rotated endover-end at room temperature (25-27°C) for 1 h at the RGNF laboratory, while agitated in a benchtop water bath shaker at 37°C for 1 h at the PBF laboratory.For the replicate samples, this was followed by an addition of 29 mL of intestinal fluid and agitation for another 4 hours.Subsequently, the samples of gastric and gastro-intestinal phase were centrifuged at 3800 rpm for 10 min and filtered, and 67% HNO3 was added (0.5 mL and 1 mL, respectively) to the supernatants before they were stored for analysis.The pH of the extracts was measured at the end of digestion for each phase.Bioaccessibility was calculated per digestion according to the following equation: (  −1 ) × 100 (1) The Mining-Geology-Petroleum Engineering Bulletin, 2016, pp.29-39 © The Author(s), DOI: 10.17794/rgn.2016.2.3

Results and Discussion
The dominant mineral in the tailings sample is dolomite, followed by calcite, quartz, willemite (zinc silicate mineral -Zn2SiO4), illitic material and traces of K-feldspar.There is indication of the presence of descloizite (lead and zinc vanadate -(Pb,Zn)2(OH)VO4).The soil sample contains dominantly quartz, followed by calcite, dolomite and K-feldspar.
Potentially toxic elements measured in aqua regia extracts of the samples are Cd, Cu, Pb and Zn.Pseudo-total concentrations of PTEs in the sample of tailings are 290 mg/kg, 153 mg/kg, 10,636 mg/kg and 43,783mg/kg, while those in sample of soil are 23 mg/kg, 25 mg/kg, 1,199 mg/kg and 5,919 mg/kg, respectively.Metal concentrations in the analysed soil (except Cu concentration) and tailings exceed Canadian Soil Quality Guidelines (CCME, 1999) which are defined as the concentrations recommended to provide a healthy, functioning ecosystem capable of sustaining the existing and likely future uses of the site by ecological receptors and humans (Table 4).

Bioaccessibility of Cd, Cu, Pb and Zn in analysed samples
Bioaccessible concentrations (mg/kg) of Cd, Cu, Pb and Zn measured in the samples after digestion in simulated gastric and intestinal fluids, as part of the orientation study simultaneously carried out at RGNF and PBF, are given in Table 5.
For each digestive phase, there is a relatively good correlation between concentrations obtained after digestion of samples in different labs, under different conditions.This is also obvious from Figure 2, which shows relative bioaccessibilities (%) of measured potentially toxic elements in the samples.The concentration of lead in the gastrointestinal phase of the tailings sample is significantly lower than in the gastric phase, possibly due to poor sample homogeneity.Even so, the values obtained in different labs differ only slightly.Hence, the reported results show that it is likely that differences in the experimental design of the tests i.e. agitation procedure and temperature conditions, do not cause significant variation in the bioaccessibility values of the measured elements.The acidic environment of the stomach compartment causes metals to mobilize from soil, which usually causes high bioaccessibility values in this phase.However, it can be observed that values in the intestinal phase for the tailings sample are higher compared to the gastric phase, except for Pb.This can be explained with more neutral pH levels of the intestinal compartment, where the metals can form new complexes.Furthermore, the pH values of the compartments are not fixed so they are dependent on the matrix of ingestion i.e. tailings/soil.The comparison of results of these two experiments enabled the decision to perform bioaccessibility determination at RGNF.The bioaccessibility test was repeated on samples as described previously and they were characterised in regards of their potential health risk as follows.
To estimate the potential risk of PTEs on humans, the Maximum Permissible Risk (MPR) values according to Baars et al. ( 2001) were quantified, herein expressed as tolerable daily intake (TDI).TDI is defined as the estimated amount of the chemical that humans can ingest daily during their entire lifetime without resultant adverse health effects.Values for the elements analysed in this study are as follows: Cd -0.5 g/kg tt/day; Cu -140 μg/kg tt/day; Pb -3.6 μg/kg tt/day; Zn -500 μg/kg tt/day; furthermore recalculated for body weight of 5 kg, 10 kg, 20 kg, 30 kg and 40 kg (see Table 6).An impact of PTEs is calculated based on the assumption that a child ingests 100 mg of soil per day by frequent hand to mouth activity, as suggested by Van Wijnen et al. (1990).Calculated values (see Table 7) were compared to the maximum daily intake.Health risk assessment of potentially toxic elements showed that the elements of greatest concern in the tailings sample are lead and cadmium, exceeding the maximum allowable daily intake for children weighing up to 40 kg and up to 10 kg, respectively.The soil sample has a high bioaccessible concentration of lead which could pose a risk for children weighing up to 10 kg.

Impact of binding sites of metals on bioaccessibility
The bioaccessibility of measured potentially toxic elements in samples was plotted in relation to the sum of their concentrations in soil/tailings fractions, obtained by sequential extraction (see Figure 3).It must be noted that the concentration of cadmium extracted by simulated body fluids from the soil sample (4590) was higher than its concentration extracted by sequential extraction, and therefore the percentage of bioaccessible Cd was calculated as > 100% and lines excluded from the graph.It can be observed that the bioaccessible concentration of all measured elements is higher than their concentration in a carbonate fraction of samples.This leads to the conclusion that the risk assessment code (RAC) cannot be used to assess the potential risk of elements to human health, as it gives lower values compared to concentrations which are extractable in simulated body fluids.Furthermore, there is no obvious correlation between the concentration extracted by gastric fluids and the concentration in soil/tailings sequential fractions of samples.Hence, it appears that the use of the sequential extraction analysis for assessing the risk of soil/sediment ingestion on health would be a difficulty.

Conclusion
In conclusion, the comparison of these two experiments enabled the decision to perform bioaccessibility determination at RGNF.As a result of this study, an efficient and relatively inexpensive method for determining the bioaccessibility was set up at the Faculty of Mining, Geology and Petroleum Engineering, which makes this kind of test more accessible and enables a new approach to risk assessment studies.
Metal concentrations in aqua regia extracts cannot be the sole indicator of the environmental risk caused by polluted soil/sediment, since they can be tightly bound in a mineral lattice of minerals and although present in significant amounts do not affect the environment and human health.Hence, sequential extraction analysis was performed to determine the binding sites of the analysed metals.Distribution of metals in different soil phases, determined by sequential extraction analysis, is shown in Figure1.It appears that approximately 20-40% of total concentration in the sample is bound in organic as well as in residual fraction for most of the analysed metals in the sample of tailings, whereas in the soil sample nearly 50% of metals are bound to organic fraction, with cadmium as an exception.The main binding sites for Cd are carbonate and reducible fraction.

Figure 2
Figure 2 Relative bioaccessibility (%) of potentially toxic elements (PTEs) in the (a) tailings sample (4578), and (b) soil sample (4590).Legend: RGNF -measured at the Faculty of Mining, Geology and Petroleum Engineering, PBF -measured at the Faculty of Food Technology and Biotechnology, G -gastric phase, GI -gastrointestinal phase.

Figure 3
Figure 3 Bioaccessibility of potentially toxic elements (a) cadmium, (b) copper, (c) lead, and (d) zinc, following extraction in body fluids (BF) for a tailings (4578) and soil (4590) sample, in relation to the sum of their concentrations in soil fractions (CC -carbonate fraction, FEMN -reducible fraction (iron and manganese oxides, hydroxides and oxyhydroxides), OR -oxidisable fraction (organic matter/sulphides), RES -residual fraction), obtained by sequential extraction (SEA).Red line -bioaccessibility in the gastric phase (G), black dashed line -bioaccessibility in the gastrointestinal phase (GI).*Concentration of Cd extracted by BF from the soil sample was higher than its concentration extracted by SEA, therefore the % of bioaccessible Cd is > 100% and lines are excluded from the graph.

.
As reported in several reviews (Dean and

Ma, 2007, Intawongse and Dean, 2006, Wragg and Cave, 2003), the
most commonly used batch extraction methods are as follows: the physiologically based extraction test (PBET) originally developed by Ruby et al. (1993); the in vitro gastrointestinal method (IVG)

Table 4
Soil quality guidlines for the protection of environmental and human health(CCME, 1999) To assess the possible environmental risk posed by samples, they were classified according to the Risk Assessment Code (RAC) introduced by Perin et al.(1985).It gives an indication of the available fraction of the elements by taking into account the percentage of element concentrations present in the exchangeable and carbonate fractions (< 1% no risk; 1 -10% low risk, 11 -30% medium risk, 31 -50% serious risk and > 50% high risk).The tailings pose a medium risk to the environment in regards to all measured elements, while the risk of the soil sample is low for Cu, medium for Pb and serious for Cd and Zn.

Table 5
Values of bioaccessible concentrations (mg/kg) of metals in the gastric (G) and gastrointestinal (GI) phases for the analysed samples measured at the Faculty of Mining, Geology and Petroleum Engineering (RGNF) and at the Faculty of Food Technology and Biotechnology (PBF).

Table 7
Concentration of bioaccessible fraction of PTEs in 100 mg of soil/tailings