Kinetic analysis of uranium accumulation in the bivalve Corbicula fluminea: effect of pH and direct exposure levels

Aquatic Toxicology 68 (2004) 95–108

Kinetic analysis of uranium accumulation in the bivalve Corbicula fluminea: effect of pH and direct exposure levels Olivier Simon∗ , Jacqueline Garnier-Laplace a

Laboratoire de Radioécologie et Ecotoxicologie, Institut de Radioprotection et de Sˆureté Nucléaire, Cadarache, Bat. 186. BP3, 13115 Saint Paul-Lez-Durance Cedex, France Received 16 April 2003; received in revised form 27 January 2004; accepted 2 March 2004

Abstract The bioaccumulation of natural uranium in the freshwater bivalve Corbicula fluminea was investigated subsequent to the bivalve’s experimental waterborne exposures. A first experiment determined the accumulation rate (transfer efficiency, tissular distribution) and subcellular distribution of uranium in organs after over 42 days of uranium exposure (100 ␮g l−1 ; pH 7) and later following 60 days of depuration. Results showed that there was direct transfer of uranium to the bivalve organs ([U]organism /[U]water = 0.16, fresh weight, fw). The highest accumulation levels occurred in the visceral mass and remained constant throughout the exposure duration, although a linear increase in the U concentration in the gills was observed (2.98 ± 1.3–10.9 ± 3.7 ␮g g−1 between Days 2 and 42). A second set of experiments were performed in order to test the influence of the exposure levels (100; 500; 1500 ␮g l−1 ) and pH (7 and 8.1) on the bioaccumulation capacities. A marked difference of U distribution is observed as a function of exposure levels (gills were favoured in the case of high exposure levels-relative burden: 49.1 ± 3% (1500 ␮g l−1 ), whereas the visceral mass presented higher accumulation levels at environmentally relevant U concentrations). Uranium concentration in the insoluble fraction (80%) in the whole body does not depend upon exposure levels in the water column or upon duration. These experiments did not allow any link to be established between the free-metal ion concentration and the bioaccumulation efficiency. Results showed a significant pH effect and indicated a link between the exposure conditions and the distribution of uranium in the bivalve organs. © 2004 Elsevier B.V. All rights reserved. Keywords: Uranium; Freshwater bivalve; Accumulation; Cellular and organ distribution

1. Introduction Uranium, an element from the actinide series, is among the metals naturally present and/or released into freshwater ecosystems. Its concentration in ter∗ Corresponding author Tel.: +33-4-42-25-27-87; fax: +33-4-42-25-64-44. E-mail address: [email protected] (O. Simon).

restrial and aquatic ecosystems may be increased by means of various anthropogenic contributions, originating from the different stages of the nuclear fuel cycle (mines and waste storage sites, in particular), agricultural use (phosphate based fertilisers), research laboratories and military use of depleted uranium (Colle et al., 2001). For example, Swanson (1983) and Waite et al. (1988) reported that in Canadian lakes, where the main anthropogenic sources are from

0166-445X/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.aquatox.2004.03.002

96

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

uranium mining, very high concentration of uranium (U) in water (200–3000 ␮g l−1 ) are detected as compared to a control lake (5.2 ␮g l−1 ). In natural waters, the concentration of U is very wide ranging, and varies from below 12 ng l−1 to in excess of 2 mg l−1 , primarily reflecting the concentration of uranium in the surrounding rocks (Betcher et al., 1988; WHO, 2001). Despite this wide variation in freshwater uranium concentration, little information is currently available regarding the ecological impact of natural uranium, and existing ecotoxicological data for freshwater biota (phytoplankton, crustaceans, cnidarians, fish) relates primarily to acute exposure toxicity tests. Exposure to natural uranium, whose main isotope is 238 U (99.3%), can theoretically induce both radiological (alpha particles) and chemical (heavy metal without any known biological function) effects. Due to its low specific activity, natural uranium in freshwater systems is very often considered to be a significant chemical hazard (Khune et al., 2002; Cooley et al., 2000), although the radiological effects cannot be ruled out, as shown by Miller et al. (2002). Despite these knowledge gaps, few authors have become interested in the uranium bioaccumulation capacities in aquatic organisms as compared to other metals such as cadmium, mercury or zinc. Therefore, the toxicity of uranium must be evaluated as a first step in explaining and predicting the biological effects of this element. The main set of experiments presented in this paper relates to the quantification of the direct exposure route via the water column, and the determination of uranium availability, bioaccumulation and target organs in a freshwater bivalve. During the experiments, the ionic composition and pH of the exposure medium were carefully controlled, since it is generally recognised that metal speciation has an effect on its availability to aquatic organisms (Campbell, 1995; Franklin et al., 2000). The freshwater bivalve Corbicula fluminea was selected due to the species abundance throughout the majority of European waterways and its high adaptability (Inza et al., 1998; Tran et al., 2001). Moreover, the bivalve model is very often used as bioindicator of aquatic pollution or as a biological model during experiments (Perceval et al., 2002; Serra et al., 1995; Vesk and Byrne, 1999). A second set of experiments was performed to test the influence on accumulation rate of two characteristics of the exposure conditions, namely pH and aqueous

uranium concentration combined with exposure duration. Additionally, the bivalve response was related to the aqueous speciation of uranium in the water used for the experiment, and the consistency of this response with the Free Ions Activity Model was discussed. As such, the aim of this article was three-fold: (i) to quantify the uranium bioaccumulation levels, as well as the uptake and depuration kinetics in the main tissues/organs of the bivalves following direct semi-chronic exposure (42 days) to concentration levels similar to those in the environment, (ii) to assess the influence of the aqueous speciation of uranium (while varying pH) and (iii) to determine the impact of the total dose (duration and exposure level) on the bioaccumulation response. Studies of the subcellular fractionation of uranium in the whole body and in target organs were also performed as an initial step to understanding of the storage processes involved. Results are discussed with comparisons drawn relative to the bioaccumulation capacities of others metals.

2. Materials and methods 2.1. Bivalves collection and maintenance C. fluminea (soft body mass = 0.30 ± 0.05, fresh weight (fw) n = 45) were manually collected from a reference site on the bank of the Sanguinet lake (Gironde, France) between the months of april and june. The background uranium level of the collected bivalves was below the analytical detection limit (3 ␮g l−1 , fw) of the ICP-AES equipment. Bivalves were acclimatised to laboratory conditions for at least 1 month prior to beginning experiments (ambient temperature:19–20 ◦ C; artificial water in mg l−1 ; Ca2+ = 11.5; Mg2+ = 8; Na+ = 11.6; K+ = 6.2; Cl− = 13.5; NO3 − = 6.3; SO4 2− = 8.1). Bivalves were maintained in a storage tank containing quartz sand (granulometry: 0.8–1.4 mm, Silaq, Mios, Gironde, France) and aerated water; the water was renewed weekly. Organisms were fed an algae suspension (Chlamydomonas reinhardii, total density of 2 × 105 cells ml−1 l−1 in the tank immediately following addition) twice a week. The daily period of light was fixed at 12 h. The ionic composition of the artificial water was identical during the acclimatisation period and the experiments.

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

2.2. Experimental system and exposure conditions The experimental system consisted of two connected tanks (resin polyester): the first (40 cm × 70 cm × 40 cm), held 108 l of synthetic water at constant pH (regulated by addition of HCl, 0.1M, Merks, NJ, USA) and with a constant uranium level, and was fed into the second tank, namely the experimental unit (40 cm × 30 cm × 50 cm), containing 25 kg of sand, aerated water (108 l aerated by air bubbles) and bivalves (150 individuals). The flow rate from the feeding tank to the experimental unit was selected to ensure the complete renewal of the second unit within 48 h and to maintain a constant ionic composition. Prior to introducing the organisms, the water pH in the entire experimental system was brought to the selected fixed value (7.0 ± 0.1 and 8.1 ± 0.2 by addition of HCl via a peristaltic pump (Gilson, Villiers-le-Bel, France)), which was controlled by a pH-Stat (Consort R305, Turnhout, Belgium). The bivalves were then introduced and acclimatised to the experimental conditions during a 7-day period and they were not fed until the end of the experiments. The experiments were performed at 20.0 ± 0.5 ◦ C. Uranium added to the feeding tank was taken from a stock solution of uranyl nitrate (UO2 (NO3 )2 ·6H2 O, Sigma-Aldrich, Saint-Quentin, France, 1 g l−1 concentration of U). The uranium concentration was maintained constant at the selected nominal value throughout the experiment by the flow-through system. Over the course of each 48 h renewal period, 15 ml water samples were taken regularly (5 min, 1, 24 and 48 h after the uranium addition) to measure the concentration of total uranium. The uranium content was determined by ICP-AES. All water samples were acidified by 2% HNO3 (65.3%, Merks, USA), and maintained in a dark environment at 4 ◦ C prior to analysis. Preliminary experiments showed that the aqueous uranium concentration in the experimental unit remained relatively constant as of the first hour following the uranium addition and until the end of the 48 h feeding tank renewal period. The main experiment was performed using a nominal uranium concentration of 100 ␮g l−1 at pH 7 during 42 days of exposure, followed by a 60-day depuration phase (identical conditions but without any added uranium). Three complementary experiments were performed in order to test the influence of three

97

main exposure condition characteristics on the bioaccumulation process, notably the pH, aqueous uranium concentration and exposure duration. These additional experiments involving one low uranium concentration (100 ␮g l−1 , pH 8.1, 30-day exposure duration) and two high concentrations (500; 1500 ␮g l−1 , pH 7, 14, and 7-day exposure durations, respectively) were performed using the same experimental conditions as during the main experiment. The respective durations of each experiment were adapted to the waterborne uranium concentrations in order to compare the bioaccumulation rates and tissue distributions after a short exposure to a high contamination level and a semi-chronic exposure to a low contamination level. A specific notation was used throughout the paper to indicate the aqueous uranium concentration (␮g l−1 ), pH and exposure duration (days), e.g., the main experimental conditions are abbreviated “100; 7; 42d”. 2.3. Tissue analysis For each experimental treatment, the changes in uranium bioaccumulation and tissue distribution (gills, visceral mass, mantle, foot and adductor muscle) over time were studied in five individuals based on measurements taken prior to beginning the exposure and after 2, 7, 14, 21 and 42 days. Immediately following sampling, the animals were blotted dry using tissue paper, weighed and stored in a freezer (−80 ◦ C). After dissection to organ levels, each sample was mineralised in a polypropylene tube using nitric acid (3 ml, Merck, 65%) and perchloric acid (2 ml, Merck, 33%) digestion at 105 ◦ C for 3 h. Samples were then diluted in 10 ml of acidified (2% nitric acid) ultra-pure water and spiked with a well-known quantity of yttrium for analysis using a Multi-componant Spectral Fitting correction treatment and taking into account the matrix effects on the ICP-AES signal. Results were obtained by averaging the data obtained at three different wavelengths (409; 417; 424 nm). Uranium concentrations and burdens at the whole body level were calculated by combining the results obtained from each organ. The quality control sample was prepared by addition of a known concentration of uranium to a mineralised C. fluminea tissue; this preparation was necessary as no certified biological material was available.

98

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

2.4. Subcellular fractionation procedure Following the exposure phase of each experiment, subcellular fractionation of the soft body tissue of C. fluminea was carried out on five animals as follows. The whole bodies removed were homogenised using a tissue-grinder (Ultra turrax T25, Ika labortechnik) in 25mM Tris–buffer containing pefabloc SC (1 mM), dithiothreitol (DTT, 2 mM), and sodium azide (0.2%) at pH 7.2. The isolation procedure used in the main experiment involved four successive differential centrifugations (Beckman Coulter, rotor JS 24 15, USA) in order to obtain four types of cellular structures (at 4 ◦ C, nuclei, cellular debris and granules: 900 × g for 10 min; mitochondria and lysosomes: 12,000 × g for 15 min; membranes: 45,000 × g for 30 min; microsomes: 103,000 × g for 70 min) along with the cytosol (Paquet, 1993). The same procedure was also carried out on the gills and visceral mass of the C. fluminea. The uranium concentration was then measured on the digestates of theses different subcellular fractions by ICP-AES using the tissues analysis procedure described previously. As regards the complementary experiments, the fractionation procedure was simpler and involved only two successive centrifuga-

tions (at 4 ◦ C, 20,000 × g for 30 min then 100,000 × g for 60 min) in order to separate the cytosol from the other cellular components and from the whole body of the bivalve. 2.5. Data treatment and statistical analysis To estimate the bioavailability of the accumulated U, an accumulation factor (AF) was determined using the following equation: AF = [U]organism (␮g g−1 )/[U]water (␮g ml−1 ). This calculation takes into account the average aqueous uranium concentration measured during the experiment. The relative burden represents the ratio between the burden measured in organs and that measured in the whole body. This parameter allowed the contribution of each organ to the U tissue distribution to be evaluated. Through regulation of abiotic parameters, the chemical U speciation was determined. Due to the lack of practical techniques to directly measure the individual uranium chemical species in solution, the thermodynamic geochemical speciation code J-Chess (Van der Lee and de Windt, 2000) was used to predict the speciation of uranium in the experimental aqueous environment. A consistent database

Table 1 Experimental conditions (U concentrations in water, pH) and contribution of major (>0.1%) aqueous chemical species of uranium and the free ion UO2 2+ to the total concentration of the metal (expressed in percentage of [U]w ) in the main and complementary experiments U concentration (␮g l−1 ) Nominal Measured pH Regulated

100 63 ± 13

500 482 ± 23

1500 1477 ± 237

7 ± 0.1

7 ± 0.1

7 ± 0.1

Contribution of dissolved U species (%) (UO2 )2 CO3 (OH)3 − 78.1 2.5 UO2 (CO3 )2 2− UO2 (CO3 )3 4− 0 5.5 UO2 CO3 (aq)− 9.4 UO2 (OH)2 (aq) UO2 (OH)3 − 1.4 0 UO2 (CO3 )3 4− Ca2 UO2 (CO3 ) aq 0 0 CaUO2 (CO3 )3 2− 6.4 × 10−2 UO2 2+ Notation [U]w ; pH; exposure

100; 7; 42d

88.8 1.3 0 2.6 4.5 0.7 0 0 0 3.1 × 10−2 500; 7; 14d

92.4 0.8 0 1.5 2.6 0.4 0 0 0 1.8 × 10−2 1500; 7; 7d

100 115 ± 17 8.1 ± 0.2 5.9 34.0 10.1 0.4 0.7 1.4 10.1 39.6 7.9 3.2 × 10−5 100; 8.1; 30d

The values were predicted using J-Chess software assuming equilibrium with the gas mixtures, and based on a revised thermodynamic database.

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

was compiled from the OECD-NEA Thermochemical Data Base, and modified to include more recent data. A critical literature review was compiled by Denison (2002). The input parameters required by J-Chess code were based on measured physico-chemical data (temperature, pH, ion concentrations, as mentioned in Table 1), and all calculations, including alkalinity, were constrained to a fixed input pH (7 or 8.1) for the artificial water at equilibrium with the atmosphere (PCO2 of 0.03 kPa). Results of U bioaccumulation (concentration and burden) are expressed as a mean of five individual data values and associated standard error. Paired comparison were carried out using the Student’s t-test with P < 0.05. The effect of exposure duration on metal bioaccumulation in the organs was analysed by multiple linear regression.

99

3. Results

available surfaces, including tanks wall. Predicting the speciation of uranium is complex as it is highly dependent on pH and total U concentration. The concentration of the free ion UO2 2+ , predicted to represent a minor species, and was divided by a factor that reached up to 2000 with increasing pH (6.4 × 10−2 % at pH 7 and 3.2 × 10−5 % at pH 8.1). At pH 7, the uranium carbonate complexes represented up to 86% of the total uranium for the various aqueous uranium concentrations used. In all of the experiments, the main species among the various polymeric uranyl species was the carbonate-complexes of uranium (UO2 CO3 (OH)3 − , 78 to 92%, Table 1). At pH 7, the increase in uranium concentration slightly affected the relative contributions of the different species. Other significant changes to the contributions of the uranium species between the two pH values were predicted for Ca2 UO2 (CO3 ) (aq) and CaUO2 (CO3 )3 2− .

3.1. Water chemistry and uranium exposure conditions

3.2. U accumulation kinetics in the bivalve soft body and organs

As regards the complementary experiments involving high aqueous uranium concentrations (500; 1500 ␮g l−1 ), the average concentration measured daily was close to the nominal concentration (482±23 and 1477±237 ␮g l−1 , respectively, see Table 1). Nevertheless, for the main experiment referred to as the “100; 7; 42d” experiment, the average concentrations measured in the experimental unit (63.2 ± 13 ␮g l−1 , n = 63) differed significantly from the nominal value. Despite the additional uranium, the nominal uranium concentration in the experimental unit was not reached, due to adsorption the element on the

Throughout the main experiment, the mass (0.29 ± 0.05 g, fw, n = 25) and the relative mass of organs presented no significant difference (t-test, P < 0.05) between Days 2 and 42 (Table 2). The U accumulation pattern (␮g g−1 , fw) in the four organs and throughout the body is illustrated in Fig. 1A. The standard deviation was considerable following 2 days of exposure, and associated with a standard deviation to mean ratio (S.D./mean) superior to 0.4 for all organs. At the end of the 42nd day of the exposure period, this ratio was inferior to 0.2 for the gills, visceral mass and whole body. The increase in

Table 2 Evolution of the whole body mass (g, fw) and of the contribution of the main organs to the total body mass for C. fluminea throughout the main experiment (mean ± 1 S.D., five replicates) Exposure duration (days)

Mass (g, fw)

Relative mass (%)

Whole body

Gills

2 7 14 21 42 42 + 60

0.324 0.333 0.295 0.259 0.295 0.228

± ± ± ± ± ±

0.058 0.039 0.042 0.047 0.082 0.082

16.7 15.1 13.8 18.6 19.1 19.3

Mantle ± ± ± ± ± ±

0.9 2.1 3.9 2.8 2.8 3.4

19.1 21.3 26.3 23.1 24.5 24.5

± ± ± ± ± ±

Visceral mass 3.3 2.7 3.8 1.5 2.6 2.6

43.3 42.5 37.9 39.1 35.9 41.9

± ± ± ± ± ±

4.1 4.8 4.6 2.1 5.5 3.9

Foot 20.8 21.2 21.9 19.2 20.4 19.9

± ± ± ± ± ±

3.1 2.9 3.1 2.1 2.1 1.8

100

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

(A)

(B)

Fig. 1. (A) Average U concentrations (␮g g−1 , fw) in the soft body of the bivalve C. fluminea after a 42-day exposure to aqueous U concentration (63 ␮g l−1 ) and followed by a 60-day period of depuration. (B) Average U concentrations (␮g g−1 , fw) and relative burdens (%) measured in the four organs of the bivalve C. fluminea. Symbols represent the mean measured values (five replicates) ± 1 S.D. The curved lines represent a regression model.

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

101

Table 3 Contribution (%) of the five subcellular fractions to the U accumulation in the whole body after 21 and 42 days and in the gills and visceral mass of C. fluminea after 42 days of waterborne U exposition during the main experiment (mean ± 1 S.D., five replicates) Whole body 21a Contribution (%) of Subcellular fraction Nuclei and cellular debris Lysosomes and mitochondria Membranes Microsomes Cytosol a

39 14 21 14 12

42a ± ± ± ± ±

5 4 2 3 4

30 16 22 19 13

± ± ± ± ±

7 4 2 3 5

Gills

Visceral mass

42a

42a

49 17 11 18 4

± ± ± ± ±

7 7 3 3 2

36 12 17 9 25

± ± ± ± ±

9 5 3 4 8

Exposure duration (day).

exposure duration significantly reduced the individual variability. As regards the organism as a whole, the accumulation levels remained constant and near 10 ␮g g−1 , fw (not taking Day 2 into account), and led to an accumulation factor of 160. In the case of the individual organs, two patterns of uranium accumulation were distinguished: one corresponded to a quasi-linear development in time and applied to the gills and mantle, the second corresponded to a rapidly reached plateau and was observed to the visceral mass and foot. The visceral mass featured the highest level of accumulation (for all the data in the time-series, [U]visceral mass = 1.95 × [U]whole body, R2 = 0.92, n = 60), and governed the concentration measured for the whole body due to its high organ mass contribution to the total body mass (40 ± 2.9%). The foot showed a weak and constant accumulation (2.6 ± 1.4 ␮g g−1 , fw), however the results were close to the ICP-AES detection limit. Results concerning primarily the gills (10.8 ± 1 ␮g g−1 , fw) and mantle showed a significant increase in accumulation (3.5×, for gills; 4×, for mantle) between Days 2 and 42 (t-test, P < 0.05, n = 10), with a linear increase in the accumulation until Day 21 ([U]gills = 0.55 × exposure duration, R2 = 0.77, n = 20; [U]mantle =0.1 × exposure duration, R2 = 0.67, n = 20), and a plateau tendency from Day 21 to Day 42. Table 3 presents the measured subcellular U distributions (%) throughout the entire body after 21 and 42 days of exposure, as well as the distributions in the gills and visceral mass after 42 days of exposure. Whole body results after 42 days of exposure indicated that only 13 ± 5% of the uranium was accumulated in the cytosolic fraction, 41% was linked to membranes

(membranes + microsome), 16% was accumulated in lysosomial and mitochondria fraction and 30% was isolated during the first step of subcellular fractionation (nuclear and cellular debris). No significant differences (t-test, P < 0.05) were noted between the two exposure durations (21 and 42 days). However, a marked distribution difference was observed between the two organs studied: little uranium (4%) was found in the cytosol of gills as compared to the visceral mass (25%). Fig. 1B represents the relative burdens (%) in order to identify the target-organs during the U accumulation over the 42 days of exposure. This ratio is associated with a reduction in the variability as compared to the AF. Notable results include the linear increase (3.3–20%) of the gill contribution, which may explain the whole body accumulation level, and the decrease in the contribution of the visceral mass. 3.3. U depuration Fig. 1 compares the accumulation levels (␮g g−1 , fw) after 42 days of direct exposure and 42 days of direct exposure followed by 60 days without U exposure. It is noteworthy that during depuration phase, the U concentration in the water column was close to 5 ± 2 ␮g l−1 , and could lead to a weak waterborne exposure. The average mass (0.23±0.08 g, fw, n = 5, Table 2) of these bivalves was lower than that of other organisms. This relatively lengthy experimental study (102 days) led to a slight mass loss of C. fluminea, however no significant difference of relative mass of each organ was observed between the 42nd day of exposure and the end of the depuration period. A similar

102

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

Fig. 2. Relative U burdens (%) in the four organs of the bivalve C. fluminea, contribution (%) of insoluble fraction to the total accumulation (␮g g−1 , fw) measured in the soft body, the latter is indicated to the top of histograms of the main and complementary experiments (Mean ± 1 S.D., five replicates). Experiments were identified as explained in Section 2.2.

phenomena was observed after 30 days of decontamination, over the course of which only algae was introduced in the experimental system (Inza et al., 1998). No significant difference in the accumulation levels (t-test, P < 0.05, n = 10) between “42 day of exposure” and “42 day + 60 day of depuration” was observed relative to the gills, mantle and foot. However, accumulation levels in the visceral mass decreased (5.6 ␮g g−1 , fw; 0.3×) and led to a decrease in the accumulation levels of whole body (0.55×). 3.4. Influence of pH and of exposure modalities conditions on U accumulation and tissue distribution in C. fluminea Fig. 2 shows the relative U burden in the organs of the bivalve, the contribution (%) of insoluble fraction measured in the whole body relative to the U total concentration, and the level of bioaccumulation (␮g g−1 , fw) in the whole body as compared to the main experiment for the same exposure duration. For the low exposure level (100 ␮g l−1 ), a pH of 8.1 led to significantly lower levels of bioaccumulation (0.07×) in organisms than in the case of a pH of 7. However, as for pH 7, a plateau tendency was observed as of Day 14, and the visceral mass represented the target organ (82±6% of the total burden). During both experiments (“500; 7; 14d” and “1500; 7; 7d”), the increase in exposure levels in water led to an increase in bioaccu-

mulation (22.8 and 16.6 ␮g g−1 , fw, respectively), as compared to the main experiment for the same exposure duration; however, this increase was not proportional to the exposure levels (AF500 ␮g/l, 14d = 45.6; AF1500 ␮g/l, 7d = 11; AF63 ␮g/l, 7d = AF63 ␮g/l, 14d = 160). A marked difference was noted regarding the response of the gills to the bioaccumulation levels; the increase in the exposure level led to an increase in their contribution to the accumulated uranium ([U]gills /[U]whole body = 3.4±0.6 and 2.3 ± 03, for both “500; 7; 14” and “1500; 7; 7” experiments, respectively). However, the relative burdens were of 49.1 ± 3 and 38.4 ± 7.3%, values which were higher than those obtained during the main experiment ([U]gills /[U]whole body = 0.47; 12.3% after 14 days of exposure). From the homogenate of organism whole bodies, accumulated uranium in the insoluble fraction was of 80 ± 7; 86 ± 1; and 85 ± 4% for “1500; 7; 7d”; “500; 7; 14d” and “100; 8.1; 30d”, respectively. No significant difference was observed despite the difference in exposure conditions and accumulation levels.

4. Discussion Although several studies have examined the relationship between aqueous uranium speciation,

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

bioavailability and induced biological effects, there is a lack of understanding regarding the interactions between uranium and organisms. This is particularly the case as regards invertebrates such as bivalves, which are widely used as bioindicators of aquatic pollution, given their ability to accumulate and bioconcentrate contaminants, such as trace metals. Within this context, the results of this study are discussed and compared to the bioaccumulation capacities of others metals. This set of experiments demonstrated the variability associated with the waterborne uranium transfer to the bivalve taking into account the variability of certain key environmental factors that define the exposure parameters, notably the pH, exposure duration and water concentration. 4.1. Individual behaviour and accumulation variability A high individual variation was observed relative to the uranium concentration measured in the soft bodies (“100; 7; 2d”, S.D./mean = 0.4, n = 5) during each experiment, as compared the variations observed based on other metals and/or other test organisms under similar experimental conditions; e.g. the S.D./mean (n = 5) was close to 0.05 for HgII, MMHg and Cd in the macrophyte Elodea densa or in the bivalve C. fluminea (Simon and Boudou, 2001). Such high variability has nonetheless been observed for uranium accumulation levels in the lake whitefish (Coregonus clupeaformis), fed a U diet under laboratory conditions (Cooley et al., 2000). This is also the case for fish (lake trout, lake whitefish, white suckers) during field exposure (Swanson, 1983). Under the experimental conditions described herein, both the ventilation flow rate and the valve movement activity of the bivalves could be considered to exhibit an high natural individual variation over the short exposure duration in spite of the 7 days devoted to the acclimatisation phase (Tran et al., 2001). The physiological state of the animals, which is strongly related to its respiratory and feeding demands, could explain the observed uranium concentration variability in the soft bodies. Recently, Tran et al. (2001) have reported that changes to the ventiloraty flow rate strongly modified the cadmium bioaccumulation process in C. fluminea. Moreover, animal starvation (no algae feeding during the exposure phase to control the chemical speciation

103

of the dissolved U species) led to a strong reduction in valve movement activity (10–15%, Tran, personal communication), and favoured an increase in the variation of the individual response. The decrease in variation of the individual accumulation with increasing exposure duration observed during the main experiment could be explained by a temporal decrease in individual behaviour variability. Another explanation relates to the direct effect of metal exposure on physiological behaviour. The addition of the metal to the exposure medium, even at low concentrations, could modify the ecophysiological response. Actually, this early response of the bivalve is used for biomonitoring of environmental pollution (Kramer et al., 1989). Fournier et al. (2003) reported that the minimum sensitivity threshold, expressed as the aqueous uranium concentration necessary to induce valve closure in 50% of the bivalves, was of 12 ␮g l−1 total uranium at pH 5.5 after 300 min of exposure. Anandraj et al. (2002) showed that Cu2+ and Zn2+ exposure (50 ␮g l−1 ) had little effect on the filtration rate, but suppressed the oxygen uptake. However, they also indicated that the marine bivalve Perna perna filtration rate was modified by Hg2+ exposure. Tran et al. (2004) reported that during the controlled exposure of C. fluminea to 60 ␮g l−1 of uranium (pH 6.5), no change in valve movement activity was noted, but a significant decrease in ventiloraty flow rate was remarked. This could contribute to an increase in the accumulation variability level. Finally, the variations in U accumulation could be explained by the freshwater bivalve’s natural individual variability in Ca concentration. Indeed, Markich et al. (2001) re-confirmed that Ca tissue concentrations could explain variations noted in metal accumulations between individuals (Hyridella depressa and Velesunio ambiguus). 4.2. Accumulation within the whole body and influence of aqueous uranium speciation Accumulation within the whole body level is influenced by both the exposure level and the pH. For a given pH, the accumulation was dependant upon the aqueous U concentration (Fig. 2). The linear relationships between U accumulation in the soft bodies and aqueous uranium concentration (100; 500; 1500 ␮g l−1 ) were compared for identical exposure durations. Following direct exposure of C. fluminea

104

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

to uranium, the Accumulate Factor AF63 ␮g/l, 42d was equal to 160. An identical value was obtained after 15 days of exposure in water at pH 6.5, however, this value was associated with a different ionic composition, which led to a different U species distribution (UO2 (OH)2 aq = 90%; UO2 2+ = 0.27%) for the same U exposure levels (Tran et al., 2004). According to its mathematical definition, a decrease in the aqueous U concentration (1500–100 ␮g l−1 ) should lead to an increase in the AF factor. As such, the accumulation capacity of uranium was near saturation for the high exposure levels used (>500 ␮g l−1 ). The uranium AF was always inferior to those measured in C. fluminea following Cd exposure (AF20 ␮g/l, 7d = 240; AF30 ␮g/l, 15d = 300), MMHg exposure (AF4 ␮g/l, 7d = 840; AF0.025 ␮g/l, 40d = 53, 000) and Hg2+ exposure (AF0.3 ␮g/l, 40d = 4800) (Simon, 2000). Regulation of the exposure medium in terms of pH and ionic composition led to a further understanding of the influence of aqueous uranium speciation on bivalve bioaccumulation. Despite the low free-metal ion concentration in the medium associated with the high uranium carbonato-complexes concentration (Table 1), bivalves were capable of accumulating and incorporating the metal. At fixed pH, the bioaccumulation measured within the whole body was not dependant upon the exposure levels, and the total or free ion uranium concentrations did not allow the bioaccumulation levels to be predicted. During this experiment, the uranium bioaccumulation rate and level was not directly proportional to the [UO2 2+ ] species, suggesting that other chemical species could contribute to the bioaccumulation within bivalves. In a study involving the green alga Chlorella vulgaris, Franklin et al. (2000) reported that an increase in pH from 5.7 to 6.5 increased the uranium toxicity (EC50 from growth inhibition bioassays of 78 ␮g l−1 down to 44 ␮g l−1 U). Equivocal results of uranium toxicity were however obtained by other authors when pH was varied. Increases in pH led to a decrease in UO2 2+ due to the formation of hydrolysis products and carbonato-complexes; however, an increase in pH could also lead to a decrease in hydronium ions, that may compete with UO2 2+ for binding to key transport sites. In terms of bioavailability, these two processes are counteracting and cannot be distinguished, a fact that increases the difficulty associated with interpreting accumulation and toxicity data (Campbell, 1995).

Such a pH effect has already been demonstrated using several metals (Perceval et al., 2002). Our experimental conditions led to marked differences in the theoretical free ion concentrations, even if the distribution of solution species calculated are to be used with caution (Denison, 2002). However, it was impossible to link the bioaccumulation differences observed with those calculated for the free-metal ion concentrations. Other parameters (Ca2+ or H+ competitions) could explain the difference in bioaccumulation levels. As suggested by Markich and Jeffree, 1994, the Ca concentration protects against the uptake of metals (Pb, Mn, Cd, Co, Hg, Cu) (via an increase in the competitive binding of Ca on the Ca channels) and hence against their toxicity (Canesi et al., 2000; Perceval et al., 2002). However, the Ca concentration level is not the only parameter to consider. Wang and Fisher (1999) demonstrated that the Ca concentrations in the medium did not significantly influence the metal (Ag, Cd, Co, Se and Zn) uptake in two marine bivalves (Mytilus edulis, Macoma balthica), however they confirmed the key-role of Ca channels in metal transport in mussels. In the present experimental conditions, a change in pH from 7 to 8.1 did not alter the Ca or Mg concentrations in the medium, but contributed to the modification of the uranium carbonato-complexes speciation (Table 1), the competitive uptake between Ca and uranyl ion species and consequently the uranium uptake. It is noteworthy that the uranyl–carbonate complexes are involved in the U deposition mechanism in the mammalian kidney (Dounce, 1949), and could participate in the transport of U in tissues. The strong affinity between U and the mineralised tissues such as skeleton or scales for fish; also confirms this key role, since U could displace Ca from the skeletal hydroxyapatite crystal lattice (Cooley et al., 2000). 4.3. Accumulation in organs The results regarding uranium target organs and internal distribution presented in this article indicate the potential for induced biological effects and also define an efficiency in terms of metal quantity and accessibility during any trophic transfer to a predator (Brown and Brumelis, 1996; Munger et al., 1999; Wallace and Lopez, 1996). Moreover, it is important to discuss the contribution of two main organs (gills and visceral mass) as a function of direct exposure

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

conditions. In the main experiment, the visceral mass presented a high (2×, 42 d) and rapid ability to accumulate uranium as compared to the gills. After the second day of exposure, the bioaccumulation appeared to be stabilised. Such a response in the visceral mass has been observed following direct contamination by Cd, Hg II at 1 ␮g l−1 , but it is noteworthy that for these metals, the accumulation patterns are greatly linked to the exposure levels (Inza et al., 1998). U tissue distribution differed according to the exposure level, e.g. at high exposure levels, the gills represented the main target organ with the highest concentration levels (Fig. 2). Thus, the U accumulated in the visceral mass could have two different origins: from direct transfer from the gills or from the accumulation following trophic transfer from particulate-associated metal (Marigomez et al., 2002). However, it would be surprising to find a trophic transfer in our experimental conditions. The experimental design (C. fluminea were not fed, the water column was renewed every 2 days) favoured the direct exposure route. Finally, uranium accumulated in this organ was eliminated after a long depuration period (Fig. 1). Additional experiments should be developed to study the depuration kinetics. However, as regards the main experiment, the linear accumulation patterns observed for gills could contribute to the increase in contamination within the whole body (relative burden increase from 3.3 to 20% between Day 2 and Day 42). When the uranium concentration in water column reached the high exposure levels, the gills seemed to be the main target organ. The accumulation/sequestration/transfer mechanisms are outlined below. For all experiments (in terms of the results gathered at the end of the exposure period), despite the high variability of U accumulation, there was a linear relationship ([U]gills = 3.3 × [U]whole body , R2 = 0.74, n = 20) between the gill and whole body accumulations. As suggested by others authors, such a response is indicative of a direct exposure route. 4.4. Metal transport and release The U bioaccumulation depended upon a combination of sorption–absorption, excretion and storage processes. Literature data indicated that the respective contribution of these three mechanisms depended on the heavy metal studied.

105

Gills were considered to represent the main target organ in terms of bioaccumulation following direct exposure to Cd, Cu, MMHg (Dallinger et al., 1987). A link could be established between metallothioneins (MT) induction, or metal binding proteins, and accumulation. The aptitude of Cd to induce MT synthesis could explain the high concentration levels in bivalves (Couillard et al., 1993; Perceval et al., 2002; Roesijadi, 1992; Serra et al., 1995), and could help to determine the Cd routes in tissues during detoxification (Marigomez et al., 2002). Given the chemical properties of uranium, the theoretical lack of MT induction following waterborne exposure (Cooley et al., 2000) led to a weak U concentration in gills and supports the incorporation into hemocytes (Marigomez et al., 2002). The high proportion of uranium accumulated in the non-cytosolic fraction of gills (95% for the main experiment) confirmed the lack of cytosolic proteins present to bind the metal. It is of note that for the marine bivalve Ruditapes decussates, the cytosolic fraction of Cd or Cu depends upon the exposure duration and level. Using a high exposure level (250 ␮g l−1 ), a considerable fraction of Cd was recovered within the cytosol after a short time period (48 h), this fraction tended to decrease (23.6 ± 7.4%) with increasing exposure time, up to 168 h (Romeo and Gnassia-Barelli, 1995). Uranium distribution (cytosol/insoluble fraction) in the whole body did not appear to depend upon the exposure levels in the water column (Fig. 2). Hemocytes (in which metals are associated with cytosolic proteins or included within lysosomes) constitute the most likely system for metal transport between the mollusc tissues and the digestive gland (Marigomez et al., 2002). The latter organ, which participates in the mechanisms of homeostasis regulation, was able to accumulate particulate metal via food, which was then transferred to the lysosomes and residuals bodies (Le Pennec and Le Pennec, 2001; Marigomez et al., 2002; Serra et al., 1995). The high burdens (about 48%) in the visceral mass bound to the cellular debris (including metal rich-granules) and lysosomes, and should prove the existence of sequestration in the process of detoxification. To confirm the key-role of metal sequestration, Vesk and Byrne (1999) proposed the use of granule (accumulated in the mantle of the marine bivalve H. depressa) microanalysis as a new approach for monitoring metals levels in the environment. These

106

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

structures are known to (i) store Ca and P, to (ii) sequestrate metal and to (iii) participate in the detoxification mechanism (Marigomez et al., 2002; Roesijadi and Robinson, 1994; Simkiss, 1981). Furthermore, Chassard-Bouchaud (1983) has shown that the crayfish Pontastacus leptodactylus concentrates uranium within the lysosomal system. Moreover, she demonstrated that a specific form was involved in the metal sequestration and detoxification processes (the spherocrystals calcium phosphate spherules) in the marine crab Carcinus maenas. U accumulates in lysosomes of marine and freshwater mollusks, crustaceans and mammals. When precipitated as U phosphate microneedles, U damages subcellular structures, such as lysosomal membranes (Ribera et al., 1996). This elimination process could lead to high accumulation levels in the digestive gland (storage in cellular debris (36 ± 9%, Table 3), certainly composed of mineral concretions and lysosomes (12.5 ± 5.5%)). Finally, visceral mass could be referred to as a final distribution/reservoir, rather than an input tissue, as it plays a key role in metal uptake (Marigomez et al., 2002). Thus, the tissue accumulation levels depend on the metal detoxification route.

5. Conclusion The present set of data, obtained using the freshwater clam C. fluminea exposed to realistic aqueous uranium concentrations, emphasised that the direct U uptake rate was less important than in the case of Cd or Hg. Results indicated a marked difference of U distribution in organs as a function of exposure levels and duration (gills were favored in the case of high exposure levels, whereas the visceral mass presented higher accumulation levels at environmentally relevant U concentrations). The high accumulation in the visceral mass could be due to its detoxification role following transfer through the gills. As such, these results suggested that the main toxicity of uranium should potentially take place in the digestive gland. These experiments did not allow any link to be established between the free-metal ion concentrations and the bioaccumulation capacities, though a significant pH effect was observed on the bioaccumulation rate (14×; from pH 8.1 to 7). Further experiments should be carried out to study the role of the Ca species present in the water

column. As regards uranium toxicity, the link between the cellular Ca disruption and U toxicity deserves to be investigated. Indeed, the toxicity of uranium could be directly linked with the quantity of non-essential metals accumulated or indirectly with the disruptions to the intracellular Ca concentration maintained during homeostasis (Delamarre and Truchet, 1984; Romeo and Gnassia-Barelli, 1995). Further experiments are also needed to determine the efficiency of the detoxification processes likely to be induced by the presence of uranium. Finally, the next step would be to evaluate the risk of exposure to uranium linked with the ingestion of contaminated prey. The information gathered during this study should help to improve the ecological risk assessment of freshwater sites affected by the presence of uranium.

Acknowledgements The authors wish to D. Tran, C. Adam and F. Denison for any discussions and help to this study. We are also grateful to V. Camilleri for uranium measurement. All experiments presented in this paper complied with the current laws in France, where they were performed. This work was a part of the ENVIRHOM research program funded by the Institute for Radioprotection and Nuclear Safety.

References Anandraj, A., Marshall, D.J., Gregory, M.A., McClurg, T.P., 2002. Metal accumulation, filtration and O2 uptake rates in mussel Perna perna (Mollusca: Bivalvia) exposed to Hg2+ , Cu2+ and Zn2+ . Comp. Biochem. Physiol. 132, 355–363. Betcher, R.N., Gascoyne, M., Brown, D., 1988. Uranium in groundwaters of southeastern Manitoba, Canada. Can. J. Earth. Sci. 25, 2089–2103. Brown, D.H., Brumelis, G., 1996. A biomonitoring method using the cellular distribution of metals in moss. Sci. Total Environ. 187, 153–161. Campbell, P.G.C., 1995. Interactions between trace metals and aquatic organisms: a critique of the free-ion activity model. In: Tessier, A., Turner, D.R. (Eds.), Metal Speciation and Bioavailability in Aquatic Systems. Wiley, New York, pp. 45–102. Canesi, L., Ciacci, C., Gallo, G., 2000. Hg2+ and Cu2+ with antagonist-mediated Ca2+ signaling in isolated Mytilus digestive gland cells. Aquat. Toxicol. 49, 1–11.

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108 Chassard-Bouchaud, C., 1983. Cellular and subcellular localization of Uranium in the crab Carcinus maenas: a microanalytical study. Mar. Pollut. Bull. 14 (4), 133–136. Colle, C., Garnier-Laplace, J., Roussel-Debet, S., Adam, C., Baudin, J.P., 2001. Comportement de l’uranium dans l’environnement. In: Métivier, H. (Ed.), l’uranium de l’environnement à l’homme. EDP Sciences, Les Ulis, pp. 187–211. Cooley, H.M., Evans, R.E., Klaverklamp, J.F., 2000. Toxicology of dietary uranium in lake whitefish (Coregonus clupeaformis). Aquat. Toxicol. 48, 495–515. Couillard, Y., Campbell, P.G.C., Tessier, A., 1993. Response of metallothionein concentrations in a freshwater (Anadonta grandis) along an environmental cadmium gradient. Limnol. Oceanogr. 38, 299–313. Dallinger, R., Prosi, F., Segner, H., Back, H., 1987. Contaminated food and uptake of heavy metals by fish: a review and a proposal for further research. Oecologia 73, 91–98. Delamarre, P., Truchet, M., 1984. Comparative study of mercury and cadmium exposure in a freshwater fish (Brachydanio rerio). Histological microanalysis of Cd-induced mineral concretions. Vie Milieu 34, 79–86. Denison, F., 2002. Application of chemical speciation modelling to uranium toxicity and bioavailability studies: compilation of a coherent database for simple experimental systems and an investigation of the effect of database uncertainty on model predictions. SERLAB-02-41, Technical. Dounce, A.L., 1949. The mechanism of action of uranium compounds in the animal body. In: Voegtlin, C., Hodge, H.C. (Eds.), Pharmacology and Toxicology of Uranium Compounds, Part I & II. Mc Graw-Hill Book Company, New York, pp. 951–991. Fournier, E., Tran, D., Garnier-Laplace, J., 2003. Valve closure response to uranium exposure for a freshwater bivalve Corbicula fluminea: quantification of the influence of pH. Environ. Toxicol. Chem., submitted for publication. Franklin, N.M., Stauber, J.L., Markich, S.J., Lim, R.P., 2000. pH-dependent toxicity of copper and uranium to a tropical freshwater alga (Chorella sp.). Aquat. Toxicol. 48, 275–289. Inza, B., Ribeyre, F., Boudou, A., 1998. Dynamics of cadmium and mercury compounds (inorganic mercury or methylmercury): uptake and depuration in Corbicula fluminea. Effects of temperature and pH. Aquat. Toxicol. 43, 273–285. Kramer, K.J.M., Jener, H.A., De Zwart, D., 1989. The valve movement response of mussels: a tool in biological monitoring. Hydrobiologia 189, 433–443. Khune, W.W., Caldwell, C.A., Gould, W.R., Fresquez, P.R., Finger, S., 2002. Effects of depleted uranium on the health and survival of Ceriodaphnia Dubia and Hyalella Azteca. Environ. Toxicol. Chem. 21, 2198–2203. Le Pennec, G., Le Pennec, M., 2001. Evaluation of the toxicity of chemical compounds using digestive acini of the bivalve mollusc Pecten maximus L. maintained alive in vitro. Aquat. Toxicol. 53, 1–7. Marigomez, I., Soto, M., Cajaraville, M.P., Angulo, E., Giamberi, L., 2002. Cellular and subcellular distribution of metals in mollusks. Microsc. Res. Tech. 56, 358–392.

107

Markich, S.J., Jeffree, R.A., 1994. Absorption of divalent trace metals as analogues of calcium by Australian freshwater bivalves: an explanation of how water hardness reduces metal toxicity. Aquat. Toxicol. 29, 257–290. Markich, S.J., Brown, P.L., Jeffree, R.A., 2001. Divalent metal accumulation in freshwater bivalves: an inverse relationship with metal phosphate solubility. Sci. Total Environ. 275, 27–41. Miller, A.C., Stewart, M., Brooks, K., Shi, L., Page, N., 2002. Depleted uranium-catalyzed oxidative DNA damage: absence of significant alpha particle decay. J. Inorg. Biochem. 91, 246– 252. Munger, C., Hare, L., Craig, A., Charest, P.M., 1999. Influence of exposure time on the distribution of cadmium within the caldoceran Ceriodaphnia dubia. Aquat. Toxicol. 44, 195–200. Paquet, F., 1993. Etude expérimentale des cinétiques de l’américium-241 chez le homard Homarus gammarus. Analyse des mécanismes d’accumulation et de détoxification au niveau subcellulaire. Thesis. Rapport CEA R-562, 401 p. Perceval, O., Pinel-Alloul, B., Methot, G., Couillard, Y., Giguère, A., Campbell, P.C.G., Hare, L., 2002. Cadmium accumulation and metallothionein synthesis in freshwater bivalves (Pyganodon grandis): relative influence of the metal exposure gradient versus limnological varaibility. Environ. Pollut. 118, 5–17. Ribera, D., Labrot, F., Tisnerat, G., Narbonne, J.F., 1996. Uranium in the environment: occurrence, transfer, and biological effects. Rev. Environ. Contam. Toxicol. 146, 53–89. Roesijadi, G., 1992. Metallothioneins in metal regulation and toxicity in aquatic animals. Aquat. Toxicol. 22, 81–114. Roesijadi, G., Robinson, W.E., 1994. Metal regulation in aquatic animals: mechanisms of uptake, accumulation, and release. In: Malins, D.C., Ostrander, G.K. (Eds.), Aquatic Toxicology. CRC Press, Boca Raton, FL, pp. 387–420. Romeo, M., Gnassia-Barelli, M., 1995. Metal distribution in different tissues and in subcellular fractions of the Mediterranean clam Ruditapes decussatus treated with cadmium, copper or zinc. Comp. Biochem. Physiol. 111, 457–463. Serra, R., Carpene, E., Marcantonio, A.C., Isani, G., 1995. Cadmium accumulation and Cd-binding proteins in the bivalve Scapharca inaequivalvis. Comp. Biochem. Physiol. 111, 165– 174. Simkiss, K., 1981. Cellular discrimination processes in metal accumulation cells. J. Exp. Biol. 94, 317–327. Simon, O., 2000. Etude des voies de contamination—directe et trophique—des organismes aquatiques continentaux par les métaux—mercure, cadmium, zinc. Approches expérimentales au laboratoire et in situ. Thesis. Université Bordeaux 1, no. 2244, 171 p. Simon, O., Boudou, A., 2001. Direct and trophic contamination of the herbivorous carp Ctenopharyngodon idella by inorganic mercury and methylmercury. Ecotoxicol. Environ. Saf. 50, 48– 59. Swanson, S.M., 1983. Levels of 226 Ra, 210 Pb, and total U in fish near a Saskatchewan uranium mine and mill. Health Phys. 45, 67–80. Tran, D., Boudou, A., Massabuau, J.C., 2001. How water oxygenation levels influences cadmium accumulation pattern

108

O. Simon, J. Garnier-Laplace / Aquatic Toxicology 68 (2004) 95–108

in the Asiatic clam Corbicula fluminea: a laboratory and field study. Environ. Toxicol. Chem. 20, 2073–2080. Tran, D., Massabuau, J.C., Garnier-Laplace, J., 2004. CO2 effect on uranium accumulation process on the freshwater clams Corbicula fluminea. Environ. Toxicol. Chem. 23, 739– 747. Waite, D.T., Joshi, S.R., Sommerstad, H., 1988. The effect of uranium mine tailing on radionuclide concentration in Langley bay, Saskatchewan, Canada. Arch. Environ. Contam. Toxicol. 17, 373–380. Van der Lee, J., de Windt, L., 2000. CHESS Tutorial and cookbook. Update for version 2.5 Users Manual Nr. LHm/RD/00/13, Ecole des Mines de Paris, Fontainebleau, France.

Vesk, P.A., Byrne, M., 1999. Metal levels in tissue granules of the freshwater bivalve Hyridella depressa (Unionida) for biomonitoring: the importance of cryopreparation. Sci. Total Environ. 225, 219–229. Wallace, W.G., Lopez, G.R., 1996. Relation between subcellular cadmium distribution in prey and cadmium trophic transfer to a predator. Estuaries 19, 923–930. Wang, W.X., Fisher, N.S., 1999. Effects of calcium and metabolic inhibitors on trace element uptake in two marine bivalves. J. Exp. Mar. Biol. Ecol. 236, 149–164. WHO. 2001. Depleted uranium, sources, exposure and health effects. WHO/SDE/PHE/01.1, World Health Organization, Geneva, Switzerland.