- Email: [email protected]
Effects of chronic exposure to environmentally relevant concentrations of waterborne depleted uranium on the digestive tract of zebrafish, Danio rerio
Journal of Environmental Radioactivity 142 (2015) 45e53
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
Effects of chronic exposure to environmentally relevant concentrations of waterborne depleted uranium on the digestive tract of zebrafish, Danio rerio Starrlight Augustine a, Sandrine Pereira a, b, Magali Floriani a, Virginie Camilleri a, atrice Gagnaire a, *, Christelle Adam-Guillermin a Sebastiaan A.L.M. Kooijman c, Be a b c
Institut de Radioprotection et Sûret e Nucl eaire (IRSN), PRP-ENV, SERIS, LECO, Cadarache, Saint-Paul-lez-Durance 13115, France CRCL, UMR INSERM 1052, CNRS 5286 Equipe de Radiobiologie, Cheney A- 1er etage, 28 Rue Laennec, 69373 Lyon Cedex 08, France Department of Theoretical Biology, Vrije Universiteit, Amsterdam, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 June 2014 Received in revised form 2 January 2015 Accepted 4 January 2015 Available online
Uranium is a naturally occurring element, but activities linked to the nuclear fuel cycle can increase background levels in the surrounding waters. For this reason it is important to understand how this affects organisms residing in the water column. The objective of this study was to assess histopathological effects of uranium on the gut wall of a widely used model organism: zebrafish, Danio rerio. To this end we exposed zebrafish to 84 and 420 nM depleted uranium for over a month and then examined the histology of intestines of exposed individuals compared to controls. The gut wall of individuals exposed to 84 and 420 nM of uranium had large regions of degraded mucosa. Using transmission electron microscopy (TEM) coupled to energy-dispersive X-ray spectroscopy microanalysis (EDX) we found that uranium induced a decrease in the amount of calcium containing mitochondrial matrix granules per mitochondria. This is suggestive of perturbations to cellular metabolism and more specifically to cellular calcium homeostasis. TEM-EDX of the gut wall tissue further showed that some uranium was internalized in the nucleus of epithelial cells in the 420 nM treatment. Fluorescent in situ hybridization using specific probes to detect all eubacteria was performed on frozen sections of 6 individual fish in the 84 nM and 420 nM treatments. Bacterial colonization of the gut of individuals in the 420 nM seemed to differ from that of the controls and 84 nM individuals. We suggest that hostemicrobiota interactions are potentially disturbed in response to uranium induced stress. The damage induced by waterborne uranium to the gut wall did not seem to depend on the concentration of uranium in the media. We measure whole body residues of uranium at the end of the experiment and compute the mean dose rate absorbed for each condition. We discuss why effects might be uncoupled from external concentration and highlight that it is not so much the external concentration but the dynamics of internalization which are important players in the game. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Zebrafish Depleted uranium Intestinal epithelium integrity Histopathology Mitochondrial matrix granules
1. Introduction Uranium is ubiquitous in natural waters at concentrations ranging from 0.04 to 25 nM depending on the composition of surrounding rocks and up to 4.2 mM at the vicinity of uraniferous sites (Bonin and Blanc, 2001). In addition, there is an international
Nucle aire (IRSN), * Corresponding author. Institut de Radioprotection et Sûrete ^t 186, B.P. 3, 13115 Saint-Paul-lezPRP-ENV, SERIS, LECO, Centre de Cadarache, Ba Durance, France. Tel.: þ334 42199493. E-mail address: [email protected] (B. Gagnaire). http://dx.doi.org/10.1016/j.jenvrad.2015.01.002 0265-931X/© 2015 Elsevier Ltd. All rights reserved.
increase in demand for primary energy production of nuclear origin and uranium (U) is the main component of nuclear fuel. Activities such as U ore extraction (e.g. leaching) and storing of mine dam tailings can contribute to increasing background levels of U in surrounding waters (e.g. Fernandas et al., 1995). Depleted uranium (DU), the byproduct of nuclear enrichment of uranium, is commonly used in military (munitions, armour plating), aviation and research fields. This gives rise to societal concern as to what are the long term effects of increasing levels of such forms of uranium on resident organisms. Histological analysis of tissues of organisms exposed in situ to uranium mine effluent or soil can be sensitive
46
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
indicators of perturbations to overall organism health and survival (Lourenço et al., 2011; Marques et al., 2009). However, the simultaneous exposure to a mixture of compounds, unknown prior history of specimens and perhaps limited literature on the histology of the organism can make some results difficult to interpret (Kelly and Janz, 2009). Characterizing sublethal histological alterations induced by U, under carefully controlled laboratory conditions, on well-studied model organisms is an important complementary approach to in situ observations of histopathology. Zebrafish, Danio rerio is a small (3e5 cm) teleost species with a long history as model organism in the fields of ecotoxicology and development (Laale, 1977). Consequently, the histology and development of zebrafish have been extensively studied in the literature and so form the basis for characterizing physiological alterations induced by uranium. We performed a chronic exposure experiment where individual zebrafish were exposed to 84 and 420 nM waterborne DU for 37 days. The experimental protocol, as well as how the exposure impacted the growth and reproduction of the individuals has been published in a companion study by Augustine et al. (2012). We sacrificed a subsample of the organisms from the 37 d experiment of that study for histopathological analysis of the intestine. Radioactive elements such as DU exert a dual radiological and chemical toxicity depending on the isotopic composition (Domingo, 2001). However, environmental risk associated with DU is considered to be mainly of chemical origin (Mathews et al., 2009; BeaugelinSeiller et al., 2008). Nonetheless, we took care to assess whole body residues on another subsample of the organisms and used these values to compute the mean absorbed dose rate for each condition. The objective of this study was to assess if DU induces damage to the gut wall. In addition, we looked for an eventual perturbation of hostemicrobe interactions by qualitatively assessing total bacterial colonization in the intestinal lumen. Next, we discuss observed histological alterations in relation to some of the observed organism level effects on the same individuals which were documented in Augustine et al. (2012). Finally, we relate our findings to the absorbed dose rates and discuss radiological and chemical aspects of the toxicity of DU. 2. Material and methods 2.1. Experimental conditions and tissue sampling Wild type zebrafish were obtained from a commercial fish re, Lyon). Zebrafish were supplier (Elevage de la Grande Rivie housed in synthetic water as in Augustine et al. (2012) at 26 C, pH ¼ 6.5 (12:12 light:dark cycle). Fish were fed commercial fish food (Tetramin™, Germany) 2e3 times a day. Water was renewed manually on a daily basis. There were three conditions: control, 84 nMDU in water and 420 nMDU in water. Exposed fish were sampled after 37 d exposure to 84 and to 420 nMDU respectively. Exposure protocol and water sample analysis can be found in Augustine et al. (2012). Control fish were from the same supplier. They were kept in the same synthetic water and fed the same food as exposed individuals. Fish were maintained in well oxygenated 119 l plexiglass aquaria (20 fish per aquaria). We examined the gut wall of individuals from the 3 treatment conditions: 0 nM, 84 nM and 420 nM DU in the media. The three conditions will hereafter be referred to as the DU0, DU84 and DU420 treatments. Zebrafish were sacrificed within minutes by immersion in melting ice in accordance with the ethical guidelines displayed and used by the NIH intramural research program (http://oacu.od.nih. gov/ARAC/documents/Zebrafish.pdf). No food was introduced into
the system 24 h prior to sacrifice in order to reduce the amount of food and feces present in the digestive tract during sampling. 2.2. Uranium analyses in biological samples Uranium bioaccumulation was determined at the whole body level according to procedures described in Augustine et al. (2012). Briefly, adult fish were dried for 6 d at 60 C. Whole body dry tissue was placed in a small glass beaker and digested in 8 ml of 65% HNO3 (Sigma Aldrich). Samples were evaporated to incipient dryness on a sand tray at 120e150 C. The digestion process was followed by a 2 ml H2O2 attack and evaporation to incipient dryness followed by a 2 ml 65% HNO3 digestion and 24 h evaporation to incipient dryness. These last two steps were repeated three times. Digested mineral residues were suspended in 5 ml of 2% HNO3 (v/v) acidified ultrapure water for ICP-MS analyses. Uranium concentrations were measured after 2% (v/v) HNO3 acidification by means of inductively coupled plasma-mass spectrometry (ICP-MS; Agilent 7500 Cx, quantification limit 0.042 nM). The raw data can be found in Table S1 (online Supplementary material). 2.3. Dose rate calculation The total absorbed dose rate of an organism is estimated by summing the internal and external dose rates. U concentrations in the whole body were converted into internal dose rates using Dose Conversion Coefficients (DCCs, Gy/unit of time per Bq/unit of volume or mass) calculated using EDEN 2.2 software (Beaugelin-Seiller et al., 2004). The external dose rate of the zebrafish was found to be negligible (DCC value >1010 Gy/d per Bq/g). Input data depend on the shapes and the elementary composition of the biological target. They also depend on the composition of the exposure media and the radioactive source (defined as a combination of radionuclides). Zebrafish were assumed to be ellipsoids (3.6 cm long, 0.8 cm high and 0.4 cm wide). Their tissues were assumed to be composed of hydrogen, carbon, oxygen and nitrogen, contributing 10.2, 9.5, 77 and 2.3% of the total wet mass, respectively. The radioactive source was determined according to the U contents in the whole fish measured after 37 days of exposure, and the presence of U decay products at the corresponding sampling time was modelled using Nucleonica software (version 2014.10.22, Institute for Transuranic Elements, Karlsruhe, Germany). The selected isotopic list of uranium isotopes (234U, 235U, 238 U) and their corresponding daughters are presented in Table 1. The obtained DCCs were then used to convert uranium concentrations (nM/g of fish) to dose rates absorbed by fish (mGy/d), see Table 2. 2.4. Light and transmission electron microscopy The intestines of 3 fish per condition were prepared for light microscopy, transmission electron microscopy (TEM) coupled to
Table 1 Depleted uranium isotopic decays. Isotopic lists of uranium isotopes (234U, 235U and 238 U), their related daughters selected for the source compositions (obtained by Nucleonica, version 2014.10.22) and corresponding DCCs (Dose Conversion Coefficient, expressed in Gy/d per Bq/g) obtained by EDEN 2.2. Isotopes
DCCs
234
6.56 6.43 6.56 7.53
U 230 Th 226 Ra 222 Rn
105 105 105 105
Isotopes
DCCs
235
6.12 1.53 6.92 1.12
U 231 Th 231 Pa 227 Ac
105 106 105 106
Isotopes
DCCs
238
5.78 6.40 4.96 6.56 3.03
U 234 Th 234m Pa 234 U 234 Pa
105 107 106 105 106
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53 Table 2 Body residues (nM/g), estimated activity concentration (Bq/g) and dose rates (mGy/ d) in whole zebrafish after 37 days of exposure to 0, 84 and 420 nM DU (99.8% 238U, 0.2% 235U, 0.0006% 234U). Please note that values are computed per gram wet weight. Condition
DU0 (n ¼ 5)
DU84 (n ¼ 11)
DU420 (n ¼ 10)
65 32
402 385
Whole body residues
nM/g Standard deviation
7 2
Estimated activity concentration
Bq/g Standard deviation
0.06 0.02
Calculated dose rate
mGy/d Standard deviation
1.8 0.5
0.57 0.29
17.3 8.6
3.53 3.37
107 102
energy-dispersive X-ray spectroscopy microanalysis (EDX). The same protocol and same equipment (unless specified otherwise in the text) were used as in Barillet et al. (2010). 2.4.1. Light microscopy For each of the 3 replicates, 6 cross sections (3 in the anterior part and 3 in the posterior part) were observed with blue toluidine
47
staining under a light microscope (DM750 Leica Microsystems GmbH, Wetzlar, Germany) and subsequently photographed (Leica camera ICC50; LAS EZ Software) at magnifications 40, 100 and 400 respectively. Alterations of the histology of the gut wall were examined. 2.4.2. Transmission electron microscopy and energy-dispersive Xray analysis For each of the 3 replicates, 2 random cross sections were observed using TEM. 10 regions between the base and tip of the intestinal folds and comprising the epithelium were captured at magnification 6000. At least 10 regions were randomly chosen and photographed at magnification 43,000 in each of the 6000 regions. In each of the 43,000 magnification photographs, the number of mitochondria and the number of electron dense mitochondrial matrix granules, MMG, were scored. We used a custom bat file to randomize the names of each of the photos from all three conditions and used the randomized names during the counting process. This was done to remove bias induced from knowing beforehand from which condition a given photo was from. The total number of mitochondria and MMGs were summed up for each of the two sections for each replicate. Results are presented as total observed MMG/mitochondria for each section of all replicates. Uranium precipitates were detected using an Energy Dispersive X-
Fig. 1. Ultrastructural defects of the digestive tract of Danio rerio exposed to waterborne depleted uranium for 37 days. The mucosa (1) comprises all of the cells layers between the muscularis (2) and the lumen (L). After 37 d of exposure to waterborne depleted uranium (DU) some histological changes to the gut wall can be observed: (BeC), (EeF) and (HeL). (AeC) Microphotographs of cross section of intestine taken at 100. (A) Cross section of intestine for control (U0); (BeC) Cross sections of intestine of an individual exposed to 84 and 420 nMDU respectively for 37 days (U84 and U420); Scale bars: 100 mm. Black boxes represent regions of (AeC) which are enlarged in (DeF). (DeF) Microphotographs taken at 400 for DU0, DU84 and DU420 respectively. (D) Cross section of an intestinal fold for DU0. The fold is comprised of mucosa (1); the outer layer of the intestine is comprised of the muscularis (2). (EeF) After 37 d of exposure to waterborne depleted uranium entire regions of the mucosa are observed to have degenerated (3). Large empty spaces are visible in the mucosa (4) as illustrated in (F). (GeH) Microphotographs taken at 400 for DU0, DU84 and DU420 respectively. Vacuolization of cells increases at tips of intestinal folds as indicated by the arrowhead in (G). Excessive vacuolization and even entire empty regions (4) in the tissue of exposed individuals (HeI). Scale bars for (DeI): 20 mm.
48
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
Fig. 2. The number of mitochondrial matrix granules (MMG) per mitochondria. L indicates lumen of intestine. (A) 400 toluidine blue plastic section of an intestinal fold. Black arrow designates the epithelium. The black box indicates typical region which is enlarged in (CeE). Scale bar: 20 mm. (B) Results of image analysis. Number of MMG per
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
ray microanalyser (EDAX Inc., Mahwah, USA) (see Barillet et al., 2010 for details). 2.5. Frozen sections and immunohistochemistry The intestines of 5e6 fish per condition were prepared for immunohistochemistry. Each intestine was dissected and fixed with paraformaldehyde 4%, sucrose in phosphate buffered saline solution (overnight at 4 C). The following day, intestines were cut in 2e3 sections and embedded in Tissue-Tek O.C.T™ before storing at 80 C. 12 mm transversal sections of the intestine were obtained with a cryostat (Leica CM 3050) and placed on superfrost slides (Superfrost Plus; Menzel). Slides were dried on a 42 C hotplate for several hours before storage at 80 C. 2.5.1. Detection of eubacteria Fluorescent in situ hybridization (FISH) was performed according to a protocol very slightly modified from Bates et al. (2006). Detection of general bacterial colonization of the digestive tract was performed using a mixture of oligonucleotide probes Eub338-I (GCTGCCTCCCGTAGGAGT), Eub338-II (GCTGCCACCCGTAGGTGT) and Eub338-III (GCTGCCACCCGTAGGTGT) modified with Cy3 fluorescein at 50 extremity (Eurogentec). Modifications of protocol were as follows: slides were incubated in a dark incubation chamber at 42 C and sections were post washed in SSC 1 (150 mM NaCl, 15 mM Sodium Citrate, pH ¼ 7). Slides were then mounted in 40 ,60 Diamidino-2-Phenyl-indole stained Vecashield (Abcys) and examined with a Nikon fluorescence microscope (Nikon Eclipse E600). FISH was performed on at least two sections for each embedded section of intestine. Presence or absence of colonization was visually recorded. Intense and less intense colonization was qualitatively specified depending on the amount of observed fluorescence. 2.5.2. Statistical analysis Differences in MMG per mitochondria and background levels of damaged mitochondria were analysed with ANOVA using Tukey's test as the post hoc test. 3. Results
49
3.2. Histological observations Histological observations showed that after 37 d of exposure to 84 nM and to 420 nM waterborne uranium, certain regions of the mucosa were degenerating (Fig. 1B, C, E, F). In addition large empty spaces at certain tips of intestinal folds (Fig. 1F, H, I) were observed. We did not find a clear dose-dependent response between individuals exposed to 84 and 420 nMDU since the above mentioned histological effects were observed in both exposure conditions. We observed an increased vacuolization of epithelium cells at the tips of intestinal folds in controls (Fig. 1G, arrowhead). The vacuolization of the apical end of folds increased between anterior and posterior regions of the digestive tract in all conditions (controls and exposed). Thus, the vacuolization of epithelial cells is not a sensitive endpoint for assessing DU induced perturbations to the gut wall of zebrafish. On a finer spatial scale, we observed that the amount of electron dense MMG decreased with exposure condition (Fig. 2B). EDX spectra of these granules showed that they are mainly composed of calcium and phosphorus (Fig. 2I). There was no difference in total amount of mitochondria scored between conditions. In addition, background levels of what appeared to be damaged mitochondria were similar between conditions. The EDX spectra analysis of gut wall tissue revealed DU precipitates in the nucleus of several enterocyte cells (Fig. 3AeC) in the DU420 condition. We found precipitates in the nucleus of enterocyte cells on different sections in two of the three replicates. Large amounts of phosphorus and to a lesser extent calcium were colocalized with the DU (Fig. 3D). While qualitatively assessing bacterial colonisation of the gut lumen we were able to distinguish between three conditions: (type I) dense clouds of rod shaped fluorescence (e.g. Fig. 4A), (type II) thinner wisp like clouds of rod shaped fluorescence (e.g. Fig. 4B) and (type III) dense packets of roundish fluorescence (e.g. Fig. 4C). We observed no fluorescence in 1 DU420 individual whereas fluorescent signals were detected in all DU0 and DU84 individuals. In the DU0 condition, 2 individuals had type I and 3 individuals type II colonisation. In the DU84 condition, 4 individuals had type I, 1 individual type II and one individual had both type II and type III. In the DU420 condition, 1 individual had type I, 3 individuals type III and the lumen of 1 individual hardly fluoresced (Table 3).
3.1. Whole body residues and related dose rates 4. Discussion Results showed a significant accumulation of DU in fish after 37 days of exposure (Table 2) higher than in the control (10 and 60, respectively at 84 and 420 nM). We refer the reader to Augustine et al. (2012) for the full analysis of this data including a detailed discussion on the toxico-kinetics of DU. The internal estimated activity concentration due to uranium isotopes and decay products increases with DU in water, according to the mass concentrations, but remains low (maximal value of 3.53 Bq/g for the 420 nM exposed group). The mean total internal dose rate due to DU and decay products varies from 1.75 to 107 mGy/d. The contribution of 238U and 234U represented 74% and 20% of the total internal dose rate, respectively, indicating that internal dose rate mainly originated from the U isotopes rather than from their related daughters.
The pH of the exposure medium directly impacts the chemical speciation of U and, as a consequence, bioavailability and toxic effects of U. Only few U species are likely to be bioavailable for aquatic organisms (Fortin et al., 2004). Therefore, to date, the modalities to analyse fish DU contamination have used soft water as the exposure medium (artificial water without phosphate and low pH [6.5]) (Barillet et al., 2010, 2011; Lerebours et al., 2009, 2010). However, these experimental conditions do not reflect the complexity of a natural environment. Therefore, the results presented here have to be considered in a laboratory experimental context. In the next two subsections we will first relate our results to other organism level effects observed on individuals from the same experiment before discussing radiological and chemical aspects of DU.
mitochondria observed in each cross section. DU0: controls; DU84 and DU420: individuals exposed to 84 and 420 nMDU respectively for 37 days. (CeE) Example of Transmission electron microscopy (TEM) micro photographs (6000) for the DU0, DU84 and DU420 treatments respectively. Each image is representative of an enlargement of region in black box shown in (A). MV shows the microvillus brush border of the epithelium; GB is a goblet cell; EV is an endocytosis vesicle. Scale bars: 2 mm. The black boxes indicate typical regions which are enlarged in order to count MMG/mitochondria. (FeG) TEM photograph (43,000) DU0, DU84 and DU420 treatments respectively. Each photo represents enlargements of regions in black boxes in (CeE) respectively. The number MMG (arrowhead) and total number of mitochondria are scored in each 43,000 photo. Scale bars: 500 nm. (I) Energy-dispersive X-ray spectroscopy microanalysis of one MMG.
50
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
Fig. 3. Histology and transmission electron microscopy (TEM) of DU precipitates in the nucleus of an epithelial cell. TEM images of an adult Danio rerio epithelium after 37 d exposure to 420 nMDU. (A) TEM photo of epithelium. Mv indicates microvillus brush border of enterocyte cells and L shows the intestinal lumen. (B) Enlargement of black box in (A) with N indicating the nucleus. (C) Higher-power view of uranium precipitates in the nucleus (enlargement of region in black box in B). (D) Energy-dispersive X-ray spectroscopy microanalysis which indicates the presence of DU precipitates in the nucleus. The two highest unlabelled peaks correspond to osmium and copper.
Fig. 4. Representative micro-photographs (1000) of the three types of bacterial colonisation. (A) type I colonisation where the lumen is colonised by dense clouds of bacteria, (B) type II colonisation where rod-shaped bacteria are thinly dispersed across the lumen of the digestive tract and (C) type III colonisation where packets of bacteria with roundish shaped fluorescence were observed. Scale bars: 5 mm.
4.1. Histopathology of DU The tissue degeneracy (Fig. 1) observed in some parts of the mucosa are suggestive of an alteration of the correct functioning of the gut wall. One of the reasons we performed this study was to shed light on some of the organism level effects of DU recorded in Augustine et al. (2012) where DU was shown to impact investment in reproduction in a rather special way. Energy/mass which would have been invested into eggs was burned to cover somatic
maintenance costs either because these costs increased in response to stress or because assimilation efficiency went down. From the standpoint of an individual it is possible to reason in terms of fluxes of energy (or mass) allocated to different physiological processes such as assimilation, maintenance, growth or reproduction (Augustine et al., 2011). The physiological mechanism by which energy allocation is perturbed is difficult to assess and few studies make a mechanistic link between macroscopic effects on reproduction and alterations of cellular processes. We hoped to
Table 3 Number of zebrafish presenting different types of fluorescence (see legend of Fig. 4) in their gut lumen for the different conditions of 0, 84 and 420 nM of DU.
DU-0 (n ¼ 5) DU-84 (n ¼ 6)a DU-420 (n ¼ 5) a
Hardly any fluorescence
Dense clouds of rods (type I)
Rods thinly dispersed through lumen (type II)
Packets of bacteria with roundish shaped fluorescence (type III)
0 0 1
2 4 1
3 2 0
0 1 3
One individual had both type II and type III fluorescence which is why there are 7 individuals in total.
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
contribute to this question by studying histopathological effects of DU on the gut since this organ plays a key role in energy assimilation. A previous study on daphnids proposed a link between decreasing assimilation efficiency and ultrastructural effects on the gut wall (Massarin et al., 2011) which further motivated this attempt. We looked at the intensity of bacterial colonization in the intestinal lumen at the end of the 37 days of exposure and found differences between each condition. There is a fine line between symbiotic, commensual and pathogenetic effects of communities of bacteria and the cells of the host; in general the presence of bacteria is a benefit, but under stress the organism may no longer be able to defend itself against pathogens or alternatively the bacteria may be able to invade the hosts cells and cease being a benefit. The increase in bacterial colonization in the 84 nM condition and seemingly decrease in the 420 nM condition may be a consequence of changes to the resident bacterial community, although further more detailed studies are needed to confirm this. Nonetheless, the present study indicates that bacterial colonization of the intestinal lumen may well be sensitive to environmental stressors such as DU. It is not possible to evaluate how long it took for the mucosa to start degenerating and if in fact the phenomenon would intensify if we continued exposure. At this time-scale we found no dosedependent effect of uranium on the observed damage. This is not surprising since in Augustine et al. (2012) we showed that effects were closely linked to amount of uranium internalized (whole body residues), see next subsection for the full discussion. How exactly DU enters the individual is not yet known but internalized DU precipitates were also observed in the nucleus of zebrafish gill cells (Barillet et al., 2010). We assumed that, given our experimental conditions, uptake would be primarily via the gills (Augustine et al., 2012). However, published studies suggest that DU might also enter via the olfactory rosettes (Faucher et al., 2012; Lerebours et al., 2009). It is also possible the DU enters via the gut wall was since the epithelium is in contact with contaminated water which enters during each feeding event. Now that DU was found in epithelial cells, the importance of uptake through the gut wall and contributions of this uptake route to total internalized DU should also be addressed in future studies. Another issue is: what is the mechanism by which uranium induces the damage to the gut wall? The occurrence of DU precipitates in the nucleus can lead us to expect DU induced DNA damage. Previous studies have effectively shown DU to be a potent genotoxic compound (Barillet et al., 2011, 2005; Lerebours et al., 2013; Pereira et al., 2012). We did not find any remarkable alterations to the ultrastructure of epithelial or goblet cells during our TEM observations. During the process of counting mitochondria and MMG we recorded the number of what looked like damaged mitochondria and we did not find any significant difference between conditions (not shown). So it appears that outside of the large damaged regions, there are no noticeable tissue level effects. The most sensitive endpoint to uranium exposure turned out to be a decrease in electron dense calcium containing mitochondrial matrix granules. This invites us to question whether there is a link between this particular observation and the degradation of large regions of the mucosa. The role of MMG in mitochondria might be to regulate calcium metabolism. The literature suggests that the physiological functions for which mitochondria sequester Ca2þ may be to induct mitochondria permeability transition and perhaps induce apoptosis (Gunter et al., 2004; Jiang et al., 2010). So one possibility is that the regions of degraded mucosa represent regions of apoptotic cells. In an attempt to see if this might be the case we performed preliminary TUNEL assays on the frozen sections of the digestive tract (Augustine & Pereira, unpublished). We did not find large regions
51
full of apototic nuclei in the tissue of exposed individuals, although further experimentation must be done to confirm this. In addition we observed the nuclei under the light microscope at 1000 to see if the nuclei of disrupted cells in these regions showed any of the characteristic morphology of apototic cells, i.e. fragmentation of the nucleus and formation of apoptotic bodies. During necrosis the nucleus usually contracts and forms urchin like structures while the cytoplasm is dispersed in the extracellular matrix. Unfortunately, we were unable to clearly see by visual inspection whether we were observing apoptosis or necrosis based on this type of morphological criteria. To sum up, a first working hypothesis is that the decrease in MMG per mitochondria in cells comprising the gut wall is a physical manifestation of perturbations to mitochondrial metabolism. The fact that these MMG seem to be involved in calcium homeostasis lead us to the second hypothesis that DU might further be inducing effects on mitochondrial calcium metabolism. Several published studies suggest that mitochondrial metabolism is in some way impacted by DU (Al Kaddissi et al., 2011; Lerebours et al., 2010, 2009) and our results further corroborate this idea. 4.2. Radiological and chemical aspects of the toxicity of DU The DU exposure led to a total absorbed dose rate in the whole fish of 1.75, 17.3 and 107 mGy/d respectively for the control, 84 nM and 420 nM groups, with a major contribution of U isotopes, whereas the contribution of uranium's daughters was low (~5% of total dose rate, see Table 2). For similar experimental conditions, Barillet et al. (2007) calculated a dose rate of 22.3 mGy/d for similar internal concentrations of DU in (male) zebrafish, i.e. ~50 nM/g wet weight. A remarkable feature of the kinetics of DU internalization is that the inter-individual variability is very high and the problem is exacerbated the higher the exposure concentration, as can be seen from the standard deviations presented in Table 2. In the companion study, Augustine et al. (2012) discuss in detail how the energetics of the fish conditions the toxico-kinetics; and vice versa since DU ultimately affects energetics. Effects of DU over time can be approximated as being proportional to internalized amount and the kinetics of internalization was shown to depend on the condition of the individual (fat, skinny, pre-spawn, post-spawn) and the amount of internalized uranium seems to saturate regardless of the amount in the exposure media. In other words the concentration of DU in the exposure media will impact the timing of the effects but not necessarily the extent. To sum up, growth, reproductive output and bioaccumulation measured on a mature female zebrafish over the course of any experiment ranging from 10 to even 37 days like we performed here, will be heavily influenced by five things: (1) whether or not she is still growing, (2) her nutritional status, (3) how much she has stored for reproduction and (4) the amount of that which was stored for reproduction which is actually ripe oocytes ready to be spawned. Furthermore, DU is continuously released during each spawning event. Females which happen to spawn a lot over the course of the experiment thus depurate more and are less affected by the metal. It is hard to say if males might be easier to standardize since they do not store so much mass allocated to reproduction. Investment in reproduction and depuration of DU through gamete release is much harder to quantify in males. The current situation is that none of the OECD guideline for zebrafish toxicity testing suggests standardizing these 5 elements together in the design of any test using adult zebrafish. The results of the study by Augustine et al. (2012) show that until that is done, each mature female zebrafish will perform quite differently from that of all of the others in the experiment and mean values will come with high standard deviations.
52
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
The dose rate absorbed by each individual depends on how much is internalized, so the potential radiological toxicity of depleted uranium is highly dependent upon toxico-kinetics. Since the internalized amount seems to saturate it might also be that radiological toxicity might come into play at a later stage in the organisms life as radiological effects become more important the longer the fish carries its burden of DU around. The chemical toxicity could potentially be diminished by precipitating DU and storing it in an inactive form in vacuoles for example. Such a mechanism might not counter radiological effects on neighbouring tissues. Further research is needed to address this issue. We provide rough estimates of mean (with standard deviations) of the dose rate absorbed per condition. But the way this relates to actual exposure media is a more complex story. What we learn from this experiment is that some fish in the DU84 condition received the same dose as some individuals in the DU420 condition and vice-versa. Since the 5 elements mentioned in the previous paragraphs were not standardized, it is more relevant to relate the results to the physiological status of the fish than directly to the concentration in the exposure media. We have done this in Augustine et al. (2012) where we coupled a bioenergetics model for zebrafish with a toxico-kinetic module specifying uptake and depuration of uranium. Overall, the internal dose rate absorbed in the DU84 and DU420 conditions is generally higher than that of the controls as well as that of natural background dose-rates for aquatic species reported in Hosseini et al. (2010). However, the mean dose rate absorbed remains lower than the radiological benchmark which was reported to be equal to 240 mGy/ d at the ecosystem level (Garnier-Laplace et al., 2006). Histopathological effects are thus observed at dose rates below the radiological benchmark, which indicates that this benchmark does not protect fish from adverse effects of depleted uranium once it bioaccumulates in high enough levels in the body. To the contrary, the Predicted No Effect Concentration (PNEC) of 1.26 nM (INERIS, 2008; Beaugelin-Seiller et al., 2015) would be more protective, since effects were seen for exposure concentrations above that limit. And again, a toxico-kinetic module taking into account the ambient water chemistry (which affects uptake) as well as elimination (which is affected by the physiology of the fish) is needed to assess this. Previous studies showed that oxidative stress biomarkers such as catalase, SOD and GPX, which are known to be impacted by ionizing radiation, responded differently in animals contaminated to DU or 233U, while other biomarkers modulated by a number of chemicals like acetylcholinesterase presented no difference between DU and 233U (Barillet et al., 2007; Al Kaddissi et al., 2011). However, few significant differences in biological responses were detected, despite huge differences (3 to 4 orders of magnitude) in the total dose rates for DU and 233U. These findings suggest that the radiotoxicity was low compared to the chemical toxicity under these exposure conditions. Finally, Mathews et al. (2009) studied the chemical and radiological risks induced by chronic exposure to uranium in freshwater ecosystems. These authors concluded that the risks to the environment from uranium's chemical toxicity generally outweigh those from radiological toxicity, regardless of the source term considered. 5. Conclusions In this study we showed that chronic exposure to waterborne uranium incurs damage to the gut wall of zebrafish. These results observed seemed to be more linked to chemical toxicity of uranium rather than radiotoxicity. Even more significant is the fact that the tissular damage incurred at 84 nM was similar to that incurred at
420 nM e we discussed why this strengthens arguments in favour of incorporating toxicokinetics in any type of risk assessment scenario analysis. Acknowledgements This work is part of the ENVIRHOM research program supported by the Institute for Radioprotection and Nuclear Safety and the ^te d'Azur region. We thank Alban Carlat for Provence Alpes Co assisting in dissections. We are grateful to Karine Faucher for helping with tissue preparation of controls. Finally, we would like to thank Nick Beresford for helpful suggestions concerning the discussion on radiological aspects of the toxicity of DU. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jenvrad.2015.01.002. References Al Kaddissi, S., Legeay, A., Gonzalez, P., Floriani, M., Camilleri, V., Gilbin, R., Simon, O., 2011. Effects of uranium uptake on transcriptional responses, histological structures and survival rate of the crayfish Procambarus clarkii. Ecotoxicol. Environ. Saf. 74, 1800e1807. Augustine, S., Gagnaire, B., Adam-Guillermin, C., Kooijman, S.A.L.M., 2012. Effects of uranium on the metabolism of zebrafish, Danio rerio. Aquat. Toxicol. 118e119, 9e26. Augustine, S., Gagnaire, B., Floriani, M., Adam-Guillermin, C., Kooijman, S.A.L.M., 2011. Developmental energetics of zebrafish, Danio rerio. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 159, 275e283. Barillet, S., Adam, C., Palluel, O., Devaux, A., 2007. Bioaccumulation, oxidative stress, and neurotoxicity in Danio rerio exposed to different isotopic compositions of uranium. Environ. Toxicol. Chem. 26, 497e505. Barillet, S., Adam-Guillermin, C., Palluel, O., Porcher, J.M., Devaux, A., 2011. Uranium bioaccumulation and biological disorders induced in zebrafish (Danio rerio) after a depleted uranium waterborne exposure. Environ. Pollut. 159, 495e502. Barillet, S., Buet, A., Adam, C., Devaux, A., 2005. Does uranium exposure induce genotoxicity in the teleostean Danio rerio? Radioprotection 40, 175e181. Barillet, S., Larno, V., Floriani, M., Devaux, A., Adam-Guillermin, C., 2010. Ultrastructural effects on gill, muscle, and gonadal tissues induced in zebrafish (Danio rerio) by a waterborne uranium exposure. Aquat. Toxicol. 100, 295e302. Bates, J.M., Mittge, E., Kuhlman, J., Baden, K.N., Cheesman, S.E., Guillemin, K., 2006. Distinct signals from the microbiota promote different aspects of zebrafish gut differentiation. Dev. Biol. 297, 374e386. Beaugelin-Seiller, K., Garnier-Laplace, J., Gilbin, R., Adam, C., 2008. Uranium chemical and radiological risk assessment for freshwater ecosystems receiving ore mining releases: principles, equations and parameters. In: AIP Conference Proceedings, Buzios, Rio de Janeiro, pp. 333e336. Beaugelin-Seiller, K., Jasserand, F., Garnier-Laplace, J., Gariel, J.C., 2004. EDEN: software to calculate the dose rate of energy for the non-human biota, due to the presence of radionuclides in the environment. Environ. Stud. 11, 87e96. vrier, L., 2015. Is Beaugelin-Seiller, K., Simon, O., Gilbin, R., Garnier-Laplace, J., Fe enough information available to derive an overall EQS for uranium in French freshwaters, according to European Guidance? Uranium e Past and Future Challenges, pp. 55e62. Bonin, B., Blanc, P. L., 2001. L'uranium dans le milieu naturel, des origines jusqu' a la tivier. Les Ulis, EDP mine. L'uranium de l'environnement a l'homme. H. Me Sciences: 7e41. Domingo, J.L., 2001. Reproductive and developmental toxicity of natural and depleted uranium: a review. Reprod. Toxicol. 15, 603e609. Faucher, K., Floriani, M., Gilbin, R., Adam-Guillermin, C., 2012. Uranium-induced sensory alterations in the zebrafish Danio rerio. Aquat. Toxicol. 124e125, 94e105. Fernandas, H.M., Veiga, L.H.S., Franklin, M.R., Prado, V.C.S., Taddei, J.F., 1995. Environmental impact assessment of uranium mining and milling facilities: a study case at the poços de caldas uranium mining and milling site, Brazil. J. Geochem. Explor. 52, 161e173. Fortin, C., Dutel, L., Garnier-Laplace, J., 2004. Uranium complexation and uptake by a green alga in relation to chemical speciation: the importance of the free uranyl ion. Environ. Toxicol. Chem. 23, 974e981. Garnier-Laplace, J., Della-Vedova, C., Gilbin, R., Copplestone, D., Hingston, J., Ciffroy, P., 2006. First derivation of predicted-no-effect values for freshwater and terrestrial ecosystems exposed to radioactive substances. Environ. Sci. Technol. 40, 6498e6505. Gunter, T.E., Yule, D.I., Gunter, K.K., Eliseev, R.A., Salter, J.D., 2004. Calcium and mitochondria. FEBS Lett. 567, 96e102.
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53 Hosseini, A., Beresford, N.A., Brown, J.E., Jones, D.G., Phaneuf, M., Thørring, H., Yankovich, T., 2010. Background dose-rates to reference animals and plants arising from exposure to naturally occurring radionuclides in aquatic environments. J. Radiol. Prot. 30, 235e264. INERIS, 2008. Uranium. Toxicological and Environmental Data Sheets of Chemicals. http://www.ineris.fr/substances/fr/substance/cas/7440-61-1. Jiang, L., Allagnat, F., Nguidjoe, E., Kamagate, A., Pachera, N., Vanderwinden, J.M., Brini, M., Carafoli, E., Eizirik, D.L., Cardozo, A.K., Herchuelz, A., 2010. Plasma membrane Ca2þ-ATPase overexpression depletes both mitochondrial and endoplasmic reticulum Ca2þ stores and triggers apoptosis in insulin-secreting BRIN-BD11 cells. J. Biol. Chem. 285, 30634e30643. Kelly, J.M., Janz, D.M., 2009. Assessment of oxidative stress and histopathology in juvenile northern pike (Esoxlucius) inhabiting lakes downstream of a uranium mill. Aquat. Toxicol. 92, 240e249. Laale, H.W., 1977. The biology and use of zebrafish, Brachydanio rerio in fisheries research. A literature review. J. Fish Biol. 10, 121-122, IN121eIN122, 123e173. Lerebours, A., Adam-Guillermin, C., Brethes, D., Frelon, S., Floriani, M., Camilleri, V., Garnier-Laplace, J., Bourdineaud, J.P., 2010. Mitochondrial energetic metabolism perturbations in skeletal muscles and brain of zebrafish (Danio rerio) exposed to low concentrations of waterborne uranium. Aquat. Toxicol. 100, 66e74. Lerebours, A., Cambier, S., Hislop, L., Adam-Guillermin, C., Bourdineaud, J.P., 2013. Genotoxic effects of exposure to waterborne uranium, dietary methylmercury and hyperoxia in zebrafish assessed by the quantitative RAPD-PCR method. Mutat. Res. Genet. Toxicol. Environ. Mutagen. 755, 55e60.
53
Lerebours, A., Gonzalez, P., Adam, C., Camilleri, V., Bourdineaud, J.P., GarnierLaplace, J., 2009. Comparative analysis of gene expression in brain, liver, skeletal muscles, and gills of zebrafish (Danio rerio) exposed to environmentally relevant waterborne uranium concentrations. Environ. Toxicol. Chem. 28, 1271e1278. Lourenço, J., Silva, A., Carvalho, F., Oliveira, J., Malta, M., Mendo, S., Gonçalves, F., Pereira, R., 2011. Histopathological changes in the earthworm Eisenia andrei associated with the exposure to metals and radionuclides. Chemosphere 85, 1630e1634. Marques, S.M., Antunes, S.C., Pissarra, H., Pereira, M.L., Gonçalves, F., Pereira, R., 2009. Histopathological changes and erythrocytic nuclear abnormalities in Iberian green frogs (Ranaperezi Seoane) from a uranium mine pond. Aquat. Toxicol. 91, 187e195. Massarin, S., Beaudouin, R., Zeman, F., Floriani, M., Gilbin, R., Alonzo, F., Pery, A.R.R., 2011. Biology-based modeling to analyze uranium toxicity data on Daphnia magna in a multigeneration study. Environ. Sci. Technol. 45, 4151e4158. Mathews, T., Beaugelin-Seiller, K., Garnier-Laplace, J., Gilbin, R., Adam, C., DellaVedova, C., 2009. A probabilistic assessment of the chemical and radiological risks of chronic exposure to uranium in freshwater ecosystems. Environ. Sci. Technol. 43, 6684e6690. , I., Garnier-Laplace, J., AdamPereira, S., Camilleri, V., Floriani, M., Cavalie Guillermin, C., 2012. Genotoxicity of uranium contamination in embryonic zebrafish cells. Aquat. Toxicol. 109, 11e16.
Contents lists available at ScienceDirect
Journal of Environmental Radioactivity journal homepage: www.elsevier.com/locate/jenvrad
Effects of chronic exposure to environmentally relevant concentrations of waterborne depleted uranium on the digestive tract of zebrafish, Danio rerio Starrlight Augustine a, Sandrine Pereira a, b, Magali Floriani a, Virginie Camilleri a, atrice Gagnaire a, *, Christelle Adam-Guillermin a Sebastiaan A.L.M. Kooijman c, Be a b c
Institut de Radioprotection et Sûret e Nucl eaire (IRSN), PRP-ENV, SERIS, LECO, Cadarache, Saint-Paul-lez-Durance 13115, France CRCL, UMR INSERM 1052, CNRS 5286 Equipe de Radiobiologie, Cheney A- 1er etage, 28 Rue Laennec, 69373 Lyon Cedex 08, France Department of Theoretical Biology, Vrije Universiteit, Amsterdam, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 June 2014 Received in revised form 2 January 2015 Accepted 4 January 2015 Available online
Uranium is a naturally occurring element, but activities linked to the nuclear fuel cycle can increase background levels in the surrounding waters. For this reason it is important to understand how this affects organisms residing in the water column. The objective of this study was to assess histopathological effects of uranium on the gut wall of a widely used model organism: zebrafish, Danio rerio. To this end we exposed zebrafish to 84 and 420 nM depleted uranium for over a month and then examined the histology of intestines of exposed individuals compared to controls. The gut wall of individuals exposed to 84 and 420 nM of uranium had large regions of degraded mucosa. Using transmission electron microscopy (TEM) coupled to energy-dispersive X-ray spectroscopy microanalysis (EDX) we found that uranium induced a decrease in the amount of calcium containing mitochondrial matrix granules per mitochondria. This is suggestive of perturbations to cellular metabolism and more specifically to cellular calcium homeostasis. TEM-EDX of the gut wall tissue further showed that some uranium was internalized in the nucleus of epithelial cells in the 420 nM treatment. Fluorescent in situ hybridization using specific probes to detect all eubacteria was performed on frozen sections of 6 individual fish in the 84 nM and 420 nM treatments. Bacterial colonization of the gut of individuals in the 420 nM seemed to differ from that of the controls and 84 nM individuals. We suggest that hostemicrobiota interactions are potentially disturbed in response to uranium induced stress. The damage induced by waterborne uranium to the gut wall did not seem to depend on the concentration of uranium in the media. We measure whole body residues of uranium at the end of the experiment and compute the mean dose rate absorbed for each condition. We discuss why effects might be uncoupled from external concentration and highlight that it is not so much the external concentration but the dynamics of internalization which are important players in the game. © 2015 Elsevier Ltd. All rights reserved.
Keywords: Zebrafish Depleted uranium Intestinal epithelium integrity Histopathology Mitochondrial matrix granules
1. Introduction Uranium is ubiquitous in natural waters at concentrations ranging from 0.04 to 25 nM depending on the composition of surrounding rocks and up to 4.2 mM at the vicinity of uraniferous sites (Bonin and Blanc, 2001). In addition, there is an international
Nucle aire (IRSN), * Corresponding author. Institut de Radioprotection et Sûrete ^t 186, B.P. 3, 13115 Saint-Paul-lezPRP-ENV, SERIS, LECO, Centre de Cadarache, Ba Durance, France. Tel.: þ334 42199493. E-mail address: [email protected] (B. Gagnaire). http://dx.doi.org/10.1016/j.jenvrad.2015.01.002 0265-931X/© 2015 Elsevier Ltd. All rights reserved.
increase in demand for primary energy production of nuclear origin and uranium (U) is the main component of nuclear fuel. Activities such as U ore extraction (e.g. leaching) and storing of mine dam tailings can contribute to increasing background levels of U in surrounding waters (e.g. Fernandas et al., 1995). Depleted uranium (DU), the byproduct of nuclear enrichment of uranium, is commonly used in military (munitions, armour plating), aviation and research fields. This gives rise to societal concern as to what are the long term effects of increasing levels of such forms of uranium on resident organisms. Histological analysis of tissues of organisms exposed in situ to uranium mine effluent or soil can be sensitive
46
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
indicators of perturbations to overall organism health and survival (Lourenço et al., 2011; Marques et al., 2009). However, the simultaneous exposure to a mixture of compounds, unknown prior history of specimens and perhaps limited literature on the histology of the organism can make some results difficult to interpret (Kelly and Janz, 2009). Characterizing sublethal histological alterations induced by U, under carefully controlled laboratory conditions, on well-studied model organisms is an important complementary approach to in situ observations of histopathology. Zebrafish, Danio rerio is a small (3e5 cm) teleost species with a long history as model organism in the fields of ecotoxicology and development (Laale, 1977). Consequently, the histology and development of zebrafish have been extensively studied in the literature and so form the basis for characterizing physiological alterations induced by uranium. We performed a chronic exposure experiment where individual zebrafish were exposed to 84 and 420 nM waterborne DU for 37 days. The experimental protocol, as well as how the exposure impacted the growth and reproduction of the individuals has been published in a companion study by Augustine et al. (2012). We sacrificed a subsample of the organisms from the 37 d experiment of that study for histopathological analysis of the intestine. Radioactive elements such as DU exert a dual radiological and chemical toxicity depending on the isotopic composition (Domingo, 2001). However, environmental risk associated with DU is considered to be mainly of chemical origin (Mathews et al., 2009; BeaugelinSeiller et al., 2008). Nonetheless, we took care to assess whole body residues on another subsample of the organisms and used these values to compute the mean absorbed dose rate for each condition. The objective of this study was to assess if DU induces damage to the gut wall. In addition, we looked for an eventual perturbation of hostemicrobe interactions by qualitatively assessing total bacterial colonization in the intestinal lumen. Next, we discuss observed histological alterations in relation to some of the observed organism level effects on the same individuals which were documented in Augustine et al. (2012). Finally, we relate our findings to the absorbed dose rates and discuss radiological and chemical aspects of the toxicity of DU. 2. Material and methods 2.1. Experimental conditions and tissue sampling Wild type zebrafish were obtained from a commercial fish re, Lyon). Zebrafish were supplier (Elevage de la Grande Rivie housed in synthetic water as in Augustine et al. (2012) at 26 C, pH ¼ 6.5 (12:12 light:dark cycle). Fish were fed commercial fish food (Tetramin™, Germany) 2e3 times a day. Water was renewed manually on a daily basis. There were three conditions: control, 84 nMDU in water and 420 nMDU in water. Exposed fish were sampled after 37 d exposure to 84 and to 420 nMDU respectively. Exposure protocol and water sample analysis can be found in Augustine et al. (2012). Control fish were from the same supplier. They were kept in the same synthetic water and fed the same food as exposed individuals. Fish were maintained in well oxygenated 119 l plexiglass aquaria (20 fish per aquaria). We examined the gut wall of individuals from the 3 treatment conditions: 0 nM, 84 nM and 420 nM DU in the media. The three conditions will hereafter be referred to as the DU0, DU84 and DU420 treatments. Zebrafish were sacrificed within minutes by immersion in melting ice in accordance with the ethical guidelines displayed and used by the NIH intramural research program (http://oacu.od.nih. gov/ARAC/documents/Zebrafish.pdf). No food was introduced into
the system 24 h prior to sacrifice in order to reduce the amount of food and feces present in the digestive tract during sampling. 2.2. Uranium analyses in biological samples Uranium bioaccumulation was determined at the whole body level according to procedures described in Augustine et al. (2012). Briefly, adult fish were dried for 6 d at 60 C. Whole body dry tissue was placed in a small glass beaker and digested in 8 ml of 65% HNO3 (Sigma Aldrich). Samples were evaporated to incipient dryness on a sand tray at 120e150 C. The digestion process was followed by a 2 ml H2O2 attack and evaporation to incipient dryness followed by a 2 ml 65% HNO3 digestion and 24 h evaporation to incipient dryness. These last two steps were repeated three times. Digested mineral residues were suspended in 5 ml of 2% HNO3 (v/v) acidified ultrapure water for ICP-MS analyses. Uranium concentrations were measured after 2% (v/v) HNO3 acidification by means of inductively coupled plasma-mass spectrometry (ICP-MS; Agilent 7500 Cx, quantification limit 0.042 nM). The raw data can be found in Table S1 (online Supplementary material). 2.3. Dose rate calculation The total absorbed dose rate of an organism is estimated by summing the internal and external dose rates. U concentrations in the whole body were converted into internal dose rates using Dose Conversion Coefficients (DCCs, Gy/unit of time per Bq/unit of volume or mass) calculated using EDEN 2.2 software (Beaugelin-Seiller et al., 2004). The external dose rate of the zebrafish was found to be negligible (DCC value >1010 Gy/d per Bq/g). Input data depend on the shapes and the elementary composition of the biological target. They also depend on the composition of the exposure media and the radioactive source (defined as a combination of radionuclides). Zebrafish were assumed to be ellipsoids (3.6 cm long, 0.8 cm high and 0.4 cm wide). Their tissues were assumed to be composed of hydrogen, carbon, oxygen and nitrogen, contributing 10.2, 9.5, 77 and 2.3% of the total wet mass, respectively. The radioactive source was determined according to the U contents in the whole fish measured after 37 days of exposure, and the presence of U decay products at the corresponding sampling time was modelled using Nucleonica software (version 2014.10.22, Institute for Transuranic Elements, Karlsruhe, Germany). The selected isotopic list of uranium isotopes (234U, 235U, 238 U) and their corresponding daughters are presented in Table 1. The obtained DCCs were then used to convert uranium concentrations (nM/g of fish) to dose rates absorbed by fish (mGy/d), see Table 2. 2.4. Light and transmission electron microscopy The intestines of 3 fish per condition were prepared for light microscopy, transmission electron microscopy (TEM) coupled to
Table 1 Depleted uranium isotopic decays. Isotopic lists of uranium isotopes (234U, 235U and 238 U), their related daughters selected for the source compositions (obtained by Nucleonica, version 2014.10.22) and corresponding DCCs (Dose Conversion Coefficient, expressed in Gy/d per Bq/g) obtained by EDEN 2.2. Isotopes
DCCs
234
6.56 6.43 6.56 7.53
U 230 Th 226 Ra 222 Rn
105 105 105 105
Isotopes
DCCs
235
6.12 1.53 6.92 1.12
U 231 Th 231 Pa 227 Ac
105 106 105 106
Isotopes
DCCs
238
5.78 6.40 4.96 6.56 3.03
U 234 Th 234m Pa 234 U 234 Pa
105 107 106 105 106
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53 Table 2 Body residues (nM/g), estimated activity concentration (Bq/g) and dose rates (mGy/ d) in whole zebrafish after 37 days of exposure to 0, 84 and 420 nM DU (99.8% 238U, 0.2% 235U, 0.0006% 234U). Please note that values are computed per gram wet weight. Condition
DU0 (n ¼ 5)
DU84 (n ¼ 11)
DU420 (n ¼ 10)
65 32
402 385
Whole body residues
nM/g Standard deviation
7 2
Estimated activity concentration
Bq/g Standard deviation
0.06 0.02
Calculated dose rate
mGy/d Standard deviation
1.8 0.5
0.57 0.29
17.3 8.6
3.53 3.37
107 102
energy-dispersive X-ray spectroscopy microanalysis (EDX). The same protocol and same equipment (unless specified otherwise in the text) were used as in Barillet et al. (2010). 2.4.1. Light microscopy For each of the 3 replicates, 6 cross sections (3 in the anterior part and 3 in the posterior part) were observed with blue toluidine
47
staining under a light microscope (DM750 Leica Microsystems GmbH, Wetzlar, Germany) and subsequently photographed (Leica camera ICC50; LAS EZ Software) at magnifications 40, 100 and 400 respectively. Alterations of the histology of the gut wall were examined. 2.4.2. Transmission electron microscopy and energy-dispersive Xray analysis For each of the 3 replicates, 2 random cross sections were observed using TEM. 10 regions between the base and tip of the intestinal folds and comprising the epithelium were captured at magnification 6000. At least 10 regions were randomly chosen and photographed at magnification 43,000 in each of the 6000 regions. In each of the 43,000 magnification photographs, the number of mitochondria and the number of electron dense mitochondrial matrix granules, MMG, were scored. We used a custom bat file to randomize the names of each of the photos from all three conditions and used the randomized names during the counting process. This was done to remove bias induced from knowing beforehand from which condition a given photo was from. The total number of mitochondria and MMGs were summed up for each of the two sections for each replicate. Results are presented as total observed MMG/mitochondria for each section of all replicates. Uranium precipitates were detected using an Energy Dispersive X-
Fig. 1. Ultrastructural defects of the digestive tract of Danio rerio exposed to waterborne depleted uranium for 37 days. The mucosa (1) comprises all of the cells layers between the muscularis (2) and the lumen (L). After 37 d of exposure to waterborne depleted uranium (DU) some histological changes to the gut wall can be observed: (BeC), (EeF) and (HeL). (AeC) Microphotographs of cross section of intestine taken at 100. (A) Cross section of intestine for control (U0); (BeC) Cross sections of intestine of an individual exposed to 84 and 420 nMDU respectively for 37 days (U84 and U420); Scale bars: 100 mm. Black boxes represent regions of (AeC) which are enlarged in (DeF). (DeF) Microphotographs taken at 400 for DU0, DU84 and DU420 respectively. (D) Cross section of an intestinal fold for DU0. The fold is comprised of mucosa (1); the outer layer of the intestine is comprised of the muscularis (2). (EeF) After 37 d of exposure to waterborne depleted uranium entire regions of the mucosa are observed to have degenerated (3). Large empty spaces are visible in the mucosa (4) as illustrated in (F). (GeH) Microphotographs taken at 400 for DU0, DU84 and DU420 respectively. Vacuolization of cells increases at tips of intestinal folds as indicated by the arrowhead in (G). Excessive vacuolization and even entire empty regions (4) in the tissue of exposed individuals (HeI). Scale bars for (DeI): 20 mm.
48
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
Fig. 2. The number of mitochondrial matrix granules (MMG) per mitochondria. L indicates lumen of intestine. (A) 400 toluidine blue plastic section of an intestinal fold. Black arrow designates the epithelium. The black box indicates typical region which is enlarged in (CeE). Scale bar: 20 mm. (B) Results of image analysis. Number of MMG per
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
ray microanalyser (EDAX Inc., Mahwah, USA) (see Barillet et al., 2010 for details). 2.5. Frozen sections and immunohistochemistry The intestines of 5e6 fish per condition were prepared for immunohistochemistry. Each intestine was dissected and fixed with paraformaldehyde 4%, sucrose in phosphate buffered saline solution (overnight at 4 C). The following day, intestines were cut in 2e3 sections and embedded in Tissue-Tek O.C.T™ before storing at 80 C. 12 mm transversal sections of the intestine were obtained with a cryostat (Leica CM 3050) and placed on superfrost slides (Superfrost Plus; Menzel). Slides were dried on a 42 C hotplate for several hours before storage at 80 C. 2.5.1. Detection of eubacteria Fluorescent in situ hybridization (FISH) was performed according to a protocol very slightly modified from Bates et al. (2006). Detection of general bacterial colonization of the digestive tract was performed using a mixture of oligonucleotide probes Eub338-I (GCTGCCTCCCGTAGGAGT), Eub338-II (GCTGCCACCCGTAGGTGT) and Eub338-III (GCTGCCACCCGTAGGTGT) modified with Cy3 fluorescein at 50 extremity (Eurogentec). Modifications of protocol were as follows: slides were incubated in a dark incubation chamber at 42 C and sections were post washed in SSC 1 (150 mM NaCl, 15 mM Sodium Citrate, pH ¼ 7). Slides were then mounted in 40 ,60 Diamidino-2-Phenyl-indole stained Vecashield (Abcys) and examined with a Nikon fluorescence microscope (Nikon Eclipse E600). FISH was performed on at least two sections for each embedded section of intestine. Presence or absence of colonization was visually recorded. Intense and less intense colonization was qualitatively specified depending on the amount of observed fluorescence. 2.5.2. Statistical analysis Differences in MMG per mitochondria and background levels of damaged mitochondria were analysed with ANOVA using Tukey's test as the post hoc test. 3. Results
49
3.2. Histological observations Histological observations showed that after 37 d of exposure to 84 nM and to 420 nM waterborne uranium, certain regions of the mucosa were degenerating (Fig. 1B, C, E, F). In addition large empty spaces at certain tips of intestinal folds (Fig. 1F, H, I) were observed. We did not find a clear dose-dependent response between individuals exposed to 84 and 420 nMDU since the above mentioned histological effects were observed in both exposure conditions. We observed an increased vacuolization of epithelium cells at the tips of intestinal folds in controls (Fig. 1G, arrowhead). The vacuolization of the apical end of folds increased between anterior and posterior regions of the digestive tract in all conditions (controls and exposed). Thus, the vacuolization of epithelial cells is not a sensitive endpoint for assessing DU induced perturbations to the gut wall of zebrafish. On a finer spatial scale, we observed that the amount of electron dense MMG decreased with exposure condition (Fig. 2B). EDX spectra of these granules showed that they are mainly composed of calcium and phosphorus (Fig. 2I). There was no difference in total amount of mitochondria scored between conditions. In addition, background levels of what appeared to be damaged mitochondria were similar between conditions. The EDX spectra analysis of gut wall tissue revealed DU precipitates in the nucleus of several enterocyte cells (Fig. 3AeC) in the DU420 condition. We found precipitates in the nucleus of enterocyte cells on different sections in two of the three replicates. Large amounts of phosphorus and to a lesser extent calcium were colocalized with the DU (Fig. 3D). While qualitatively assessing bacterial colonisation of the gut lumen we were able to distinguish between three conditions: (type I) dense clouds of rod shaped fluorescence (e.g. Fig. 4A), (type II) thinner wisp like clouds of rod shaped fluorescence (e.g. Fig. 4B) and (type III) dense packets of roundish fluorescence (e.g. Fig. 4C). We observed no fluorescence in 1 DU420 individual whereas fluorescent signals were detected in all DU0 and DU84 individuals. In the DU0 condition, 2 individuals had type I and 3 individuals type II colonisation. In the DU84 condition, 4 individuals had type I, 1 individual type II and one individual had both type II and type III. In the DU420 condition, 1 individual had type I, 3 individuals type III and the lumen of 1 individual hardly fluoresced (Table 3).
3.1. Whole body residues and related dose rates 4. Discussion Results showed a significant accumulation of DU in fish after 37 days of exposure (Table 2) higher than in the control (10 and 60, respectively at 84 and 420 nM). We refer the reader to Augustine et al. (2012) for the full analysis of this data including a detailed discussion on the toxico-kinetics of DU. The internal estimated activity concentration due to uranium isotopes and decay products increases with DU in water, according to the mass concentrations, but remains low (maximal value of 3.53 Bq/g for the 420 nM exposed group). The mean total internal dose rate due to DU and decay products varies from 1.75 to 107 mGy/d. The contribution of 238U and 234U represented 74% and 20% of the total internal dose rate, respectively, indicating that internal dose rate mainly originated from the U isotopes rather than from their related daughters.
The pH of the exposure medium directly impacts the chemical speciation of U and, as a consequence, bioavailability and toxic effects of U. Only few U species are likely to be bioavailable for aquatic organisms (Fortin et al., 2004). Therefore, to date, the modalities to analyse fish DU contamination have used soft water as the exposure medium (artificial water without phosphate and low pH [6.5]) (Barillet et al., 2010, 2011; Lerebours et al., 2009, 2010). However, these experimental conditions do not reflect the complexity of a natural environment. Therefore, the results presented here have to be considered in a laboratory experimental context. In the next two subsections we will first relate our results to other organism level effects observed on individuals from the same experiment before discussing radiological and chemical aspects of DU.
mitochondria observed in each cross section. DU0: controls; DU84 and DU420: individuals exposed to 84 and 420 nMDU respectively for 37 days. (CeE) Example of Transmission electron microscopy (TEM) micro photographs (6000) for the DU0, DU84 and DU420 treatments respectively. Each image is representative of an enlargement of region in black box shown in (A). MV shows the microvillus brush border of the epithelium; GB is a goblet cell; EV is an endocytosis vesicle. Scale bars: 2 mm. The black boxes indicate typical regions which are enlarged in order to count MMG/mitochondria. (FeG) TEM photograph (43,000) DU0, DU84 and DU420 treatments respectively. Each photo represents enlargements of regions in black boxes in (CeE) respectively. The number MMG (arrowhead) and total number of mitochondria are scored in each 43,000 photo. Scale bars: 500 nm. (I) Energy-dispersive X-ray spectroscopy microanalysis of one MMG.
50
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
Fig. 3. Histology and transmission electron microscopy (TEM) of DU precipitates in the nucleus of an epithelial cell. TEM images of an adult Danio rerio epithelium after 37 d exposure to 420 nMDU. (A) TEM photo of epithelium. Mv indicates microvillus brush border of enterocyte cells and L shows the intestinal lumen. (B) Enlargement of black box in (A) with N indicating the nucleus. (C) Higher-power view of uranium precipitates in the nucleus (enlargement of region in black box in B). (D) Energy-dispersive X-ray spectroscopy microanalysis which indicates the presence of DU precipitates in the nucleus. The two highest unlabelled peaks correspond to osmium and copper.
Fig. 4. Representative micro-photographs (1000) of the three types of bacterial colonisation. (A) type I colonisation where the lumen is colonised by dense clouds of bacteria, (B) type II colonisation where rod-shaped bacteria are thinly dispersed across the lumen of the digestive tract and (C) type III colonisation where packets of bacteria with roundish shaped fluorescence were observed. Scale bars: 5 mm.
4.1. Histopathology of DU The tissue degeneracy (Fig. 1) observed in some parts of the mucosa are suggestive of an alteration of the correct functioning of the gut wall. One of the reasons we performed this study was to shed light on some of the organism level effects of DU recorded in Augustine et al. (2012) where DU was shown to impact investment in reproduction in a rather special way. Energy/mass which would have been invested into eggs was burned to cover somatic
maintenance costs either because these costs increased in response to stress or because assimilation efficiency went down. From the standpoint of an individual it is possible to reason in terms of fluxes of energy (or mass) allocated to different physiological processes such as assimilation, maintenance, growth or reproduction (Augustine et al., 2011). The physiological mechanism by which energy allocation is perturbed is difficult to assess and few studies make a mechanistic link between macroscopic effects on reproduction and alterations of cellular processes. We hoped to
Table 3 Number of zebrafish presenting different types of fluorescence (see legend of Fig. 4) in their gut lumen for the different conditions of 0, 84 and 420 nM of DU.
DU-0 (n ¼ 5) DU-84 (n ¼ 6)a DU-420 (n ¼ 5) a
Hardly any fluorescence
Dense clouds of rods (type I)
Rods thinly dispersed through lumen (type II)
Packets of bacteria with roundish shaped fluorescence (type III)
0 0 1
2 4 1
3 2 0
0 1 3
One individual had both type II and type III fluorescence which is why there are 7 individuals in total.
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
contribute to this question by studying histopathological effects of DU on the gut since this organ plays a key role in energy assimilation. A previous study on daphnids proposed a link between decreasing assimilation efficiency and ultrastructural effects on the gut wall (Massarin et al., 2011) which further motivated this attempt. We looked at the intensity of bacterial colonization in the intestinal lumen at the end of the 37 days of exposure and found differences between each condition. There is a fine line between symbiotic, commensual and pathogenetic effects of communities of bacteria and the cells of the host; in general the presence of bacteria is a benefit, but under stress the organism may no longer be able to defend itself against pathogens or alternatively the bacteria may be able to invade the hosts cells and cease being a benefit. The increase in bacterial colonization in the 84 nM condition and seemingly decrease in the 420 nM condition may be a consequence of changes to the resident bacterial community, although further more detailed studies are needed to confirm this. Nonetheless, the present study indicates that bacterial colonization of the intestinal lumen may well be sensitive to environmental stressors such as DU. It is not possible to evaluate how long it took for the mucosa to start degenerating and if in fact the phenomenon would intensify if we continued exposure. At this time-scale we found no dosedependent effect of uranium on the observed damage. This is not surprising since in Augustine et al. (2012) we showed that effects were closely linked to amount of uranium internalized (whole body residues), see next subsection for the full discussion. How exactly DU enters the individual is not yet known but internalized DU precipitates were also observed in the nucleus of zebrafish gill cells (Barillet et al., 2010). We assumed that, given our experimental conditions, uptake would be primarily via the gills (Augustine et al., 2012). However, published studies suggest that DU might also enter via the olfactory rosettes (Faucher et al., 2012; Lerebours et al., 2009). It is also possible the DU enters via the gut wall was since the epithelium is in contact with contaminated water which enters during each feeding event. Now that DU was found in epithelial cells, the importance of uptake through the gut wall and contributions of this uptake route to total internalized DU should also be addressed in future studies. Another issue is: what is the mechanism by which uranium induces the damage to the gut wall? The occurrence of DU precipitates in the nucleus can lead us to expect DU induced DNA damage. Previous studies have effectively shown DU to be a potent genotoxic compound (Barillet et al., 2011, 2005; Lerebours et al., 2013; Pereira et al., 2012). We did not find any remarkable alterations to the ultrastructure of epithelial or goblet cells during our TEM observations. During the process of counting mitochondria and MMG we recorded the number of what looked like damaged mitochondria and we did not find any significant difference between conditions (not shown). So it appears that outside of the large damaged regions, there are no noticeable tissue level effects. The most sensitive endpoint to uranium exposure turned out to be a decrease in electron dense calcium containing mitochondrial matrix granules. This invites us to question whether there is a link between this particular observation and the degradation of large regions of the mucosa. The role of MMG in mitochondria might be to regulate calcium metabolism. The literature suggests that the physiological functions for which mitochondria sequester Ca2þ may be to induct mitochondria permeability transition and perhaps induce apoptosis (Gunter et al., 2004; Jiang et al., 2010). So one possibility is that the regions of degraded mucosa represent regions of apoptotic cells. In an attempt to see if this might be the case we performed preliminary TUNEL assays on the frozen sections of the digestive tract (Augustine & Pereira, unpublished). We did not find large regions
51
full of apototic nuclei in the tissue of exposed individuals, although further experimentation must be done to confirm this. In addition we observed the nuclei under the light microscope at 1000 to see if the nuclei of disrupted cells in these regions showed any of the characteristic morphology of apototic cells, i.e. fragmentation of the nucleus and formation of apoptotic bodies. During necrosis the nucleus usually contracts and forms urchin like structures while the cytoplasm is dispersed in the extracellular matrix. Unfortunately, we were unable to clearly see by visual inspection whether we were observing apoptosis or necrosis based on this type of morphological criteria. To sum up, a first working hypothesis is that the decrease in MMG per mitochondria in cells comprising the gut wall is a physical manifestation of perturbations to mitochondrial metabolism. The fact that these MMG seem to be involved in calcium homeostasis lead us to the second hypothesis that DU might further be inducing effects on mitochondrial calcium metabolism. Several published studies suggest that mitochondrial metabolism is in some way impacted by DU (Al Kaddissi et al., 2011; Lerebours et al., 2010, 2009) and our results further corroborate this idea. 4.2. Radiological and chemical aspects of the toxicity of DU The DU exposure led to a total absorbed dose rate in the whole fish of 1.75, 17.3 and 107 mGy/d respectively for the control, 84 nM and 420 nM groups, with a major contribution of U isotopes, whereas the contribution of uranium's daughters was low (~5% of total dose rate, see Table 2). For similar experimental conditions, Barillet et al. (2007) calculated a dose rate of 22.3 mGy/d for similar internal concentrations of DU in (male) zebrafish, i.e. ~50 nM/g wet weight. A remarkable feature of the kinetics of DU internalization is that the inter-individual variability is very high and the problem is exacerbated the higher the exposure concentration, as can be seen from the standard deviations presented in Table 2. In the companion study, Augustine et al. (2012) discuss in detail how the energetics of the fish conditions the toxico-kinetics; and vice versa since DU ultimately affects energetics. Effects of DU over time can be approximated as being proportional to internalized amount and the kinetics of internalization was shown to depend on the condition of the individual (fat, skinny, pre-spawn, post-spawn) and the amount of internalized uranium seems to saturate regardless of the amount in the exposure media. In other words the concentration of DU in the exposure media will impact the timing of the effects but not necessarily the extent. To sum up, growth, reproductive output and bioaccumulation measured on a mature female zebrafish over the course of any experiment ranging from 10 to even 37 days like we performed here, will be heavily influenced by five things: (1) whether or not she is still growing, (2) her nutritional status, (3) how much she has stored for reproduction and (4) the amount of that which was stored for reproduction which is actually ripe oocytes ready to be spawned. Furthermore, DU is continuously released during each spawning event. Females which happen to spawn a lot over the course of the experiment thus depurate more and are less affected by the metal. It is hard to say if males might be easier to standardize since they do not store so much mass allocated to reproduction. Investment in reproduction and depuration of DU through gamete release is much harder to quantify in males. The current situation is that none of the OECD guideline for zebrafish toxicity testing suggests standardizing these 5 elements together in the design of any test using adult zebrafish. The results of the study by Augustine et al. (2012) show that until that is done, each mature female zebrafish will perform quite differently from that of all of the others in the experiment and mean values will come with high standard deviations.
52
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53
The dose rate absorbed by each individual depends on how much is internalized, so the potential radiological toxicity of depleted uranium is highly dependent upon toxico-kinetics. Since the internalized amount seems to saturate it might also be that radiological toxicity might come into play at a later stage in the organisms life as radiological effects become more important the longer the fish carries its burden of DU around. The chemical toxicity could potentially be diminished by precipitating DU and storing it in an inactive form in vacuoles for example. Such a mechanism might not counter radiological effects on neighbouring tissues. Further research is needed to address this issue. We provide rough estimates of mean (with standard deviations) of the dose rate absorbed per condition. But the way this relates to actual exposure media is a more complex story. What we learn from this experiment is that some fish in the DU84 condition received the same dose as some individuals in the DU420 condition and vice-versa. Since the 5 elements mentioned in the previous paragraphs were not standardized, it is more relevant to relate the results to the physiological status of the fish than directly to the concentration in the exposure media. We have done this in Augustine et al. (2012) where we coupled a bioenergetics model for zebrafish with a toxico-kinetic module specifying uptake and depuration of uranium. Overall, the internal dose rate absorbed in the DU84 and DU420 conditions is generally higher than that of the controls as well as that of natural background dose-rates for aquatic species reported in Hosseini et al. (2010). However, the mean dose rate absorbed remains lower than the radiological benchmark which was reported to be equal to 240 mGy/ d at the ecosystem level (Garnier-Laplace et al., 2006). Histopathological effects are thus observed at dose rates below the radiological benchmark, which indicates that this benchmark does not protect fish from adverse effects of depleted uranium once it bioaccumulates in high enough levels in the body. To the contrary, the Predicted No Effect Concentration (PNEC) of 1.26 nM (INERIS, 2008; Beaugelin-Seiller et al., 2015) would be more protective, since effects were seen for exposure concentrations above that limit. And again, a toxico-kinetic module taking into account the ambient water chemistry (which affects uptake) as well as elimination (which is affected by the physiology of the fish) is needed to assess this. Previous studies showed that oxidative stress biomarkers such as catalase, SOD and GPX, which are known to be impacted by ionizing radiation, responded differently in animals contaminated to DU or 233U, while other biomarkers modulated by a number of chemicals like acetylcholinesterase presented no difference between DU and 233U (Barillet et al., 2007; Al Kaddissi et al., 2011). However, few significant differences in biological responses were detected, despite huge differences (3 to 4 orders of magnitude) in the total dose rates for DU and 233U. These findings suggest that the radiotoxicity was low compared to the chemical toxicity under these exposure conditions. Finally, Mathews et al. (2009) studied the chemical and radiological risks induced by chronic exposure to uranium in freshwater ecosystems. These authors concluded that the risks to the environment from uranium's chemical toxicity generally outweigh those from radiological toxicity, regardless of the source term considered. 5. Conclusions In this study we showed that chronic exposure to waterborne uranium incurs damage to the gut wall of zebrafish. These results observed seemed to be more linked to chemical toxicity of uranium rather than radiotoxicity. Even more significant is the fact that the tissular damage incurred at 84 nM was similar to that incurred at
420 nM e we discussed why this strengthens arguments in favour of incorporating toxicokinetics in any type of risk assessment scenario analysis. Acknowledgements This work is part of the ENVIRHOM research program supported by the Institute for Radioprotection and Nuclear Safety and the ^te d'Azur region. We thank Alban Carlat for Provence Alpes Co assisting in dissections. We are grateful to Karine Faucher for helping with tissue preparation of controls. Finally, we would like to thank Nick Beresford for helpful suggestions concerning the discussion on radiological aspects of the toxicity of DU. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jenvrad.2015.01.002. References Al Kaddissi, S., Legeay, A., Gonzalez, P., Floriani, M., Camilleri, V., Gilbin, R., Simon, O., 2011. Effects of uranium uptake on transcriptional responses, histological structures and survival rate of the crayfish Procambarus clarkii. Ecotoxicol. Environ. Saf. 74, 1800e1807. Augustine, S., Gagnaire, B., Adam-Guillermin, C., Kooijman, S.A.L.M., 2012. Effects of uranium on the metabolism of zebrafish, Danio rerio. Aquat. Toxicol. 118e119, 9e26. Augustine, S., Gagnaire, B., Floriani, M., Adam-Guillermin, C., Kooijman, S.A.L.M., 2011. Developmental energetics of zebrafish, Danio rerio. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 159, 275e283. Barillet, S., Adam, C., Palluel, O., Devaux, A., 2007. Bioaccumulation, oxidative stress, and neurotoxicity in Danio rerio exposed to different isotopic compositions of uranium. Environ. Toxicol. Chem. 26, 497e505. Barillet, S., Adam-Guillermin, C., Palluel, O., Porcher, J.M., Devaux, A., 2011. Uranium bioaccumulation and biological disorders induced in zebrafish (Danio rerio) after a depleted uranium waterborne exposure. Environ. Pollut. 159, 495e502. Barillet, S., Buet, A., Adam, C., Devaux, A., 2005. Does uranium exposure induce genotoxicity in the teleostean Danio rerio? Radioprotection 40, 175e181. Barillet, S., Larno, V., Floriani, M., Devaux, A., Adam-Guillermin, C., 2010. Ultrastructural effects on gill, muscle, and gonadal tissues induced in zebrafish (Danio rerio) by a waterborne uranium exposure. Aquat. Toxicol. 100, 295e302. Bates, J.M., Mittge, E., Kuhlman, J., Baden, K.N., Cheesman, S.E., Guillemin, K., 2006. Distinct signals from the microbiota promote different aspects of zebrafish gut differentiation. Dev. Biol. 297, 374e386. Beaugelin-Seiller, K., Garnier-Laplace, J., Gilbin, R., Adam, C., 2008. Uranium chemical and radiological risk assessment for freshwater ecosystems receiving ore mining releases: principles, equations and parameters. In: AIP Conference Proceedings, Buzios, Rio de Janeiro, pp. 333e336. Beaugelin-Seiller, K., Jasserand, F., Garnier-Laplace, J., Gariel, J.C., 2004. EDEN: software to calculate the dose rate of energy for the non-human biota, due to the presence of radionuclides in the environment. Environ. Stud. 11, 87e96. vrier, L., 2015. Is Beaugelin-Seiller, K., Simon, O., Gilbin, R., Garnier-Laplace, J., Fe enough information available to derive an overall EQS for uranium in French freshwaters, according to European Guidance? Uranium e Past and Future Challenges, pp. 55e62. Bonin, B., Blanc, P. L., 2001. L'uranium dans le milieu naturel, des origines jusqu' a la tivier. Les Ulis, EDP mine. L'uranium de l'environnement a l'homme. H. Me Sciences: 7e41. Domingo, J.L., 2001. Reproductive and developmental toxicity of natural and depleted uranium: a review. Reprod. Toxicol. 15, 603e609. Faucher, K., Floriani, M., Gilbin, R., Adam-Guillermin, C., 2012. Uranium-induced sensory alterations in the zebrafish Danio rerio. Aquat. Toxicol. 124e125, 94e105. Fernandas, H.M., Veiga, L.H.S., Franklin, M.R., Prado, V.C.S., Taddei, J.F., 1995. Environmental impact assessment of uranium mining and milling facilities: a study case at the poços de caldas uranium mining and milling site, Brazil. J. Geochem. Explor. 52, 161e173. Fortin, C., Dutel, L., Garnier-Laplace, J., 2004. Uranium complexation and uptake by a green alga in relation to chemical speciation: the importance of the free uranyl ion. Environ. Toxicol. Chem. 23, 974e981. Garnier-Laplace, J., Della-Vedova, C., Gilbin, R., Copplestone, D., Hingston, J., Ciffroy, P., 2006. First derivation of predicted-no-effect values for freshwater and terrestrial ecosystems exposed to radioactive substances. Environ. Sci. Technol. 40, 6498e6505. Gunter, T.E., Yule, D.I., Gunter, K.K., Eliseev, R.A., Salter, J.D., 2004. Calcium and mitochondria. FEBS Lett. 567, 96e102.
S. Augustine et al. / Journal of Environmental Radioactivity 142 (2015) 45e53 Hosseini, A., Beresford, N.A., Brown, J.E., Jones, D.G., Phaneuf, M., Thørring, H., Yankovich, T., 2010. Background dose-rates to reference animals and plants arising from exposure to naturally occurring radionuclides in aquatic environments. J. Radiol. Prot. 30, 235e264. INERIS, 2008. Uranium. Toxicological and Environmental Data Sheets of Chemicals. http://www.ineris.fr/substances/fr/substance/cas/7440-61-1. Jiang, L., Allagnat, F., Nguidjoe, E., Kamagate, A., Pachera, N., Vanderwinden, J.M., Brini, M., Carafoli, E., Eizirik, D.L., Cardozo, A.K., Herchuelz, A., 2010. Plasma membrane Ca2þ-ATPase overexpression depletes both mitochondrial and endoplasmic reticulum Ca2þ stores and triggers apoptosis in insulin-secreting BRIN-BD11 cells. J. Biol. Chem. 285, 30634e30643. Kelly, J.M., Janz, D.M., 2009. Assessment of oxidative stress and histopathology in juvenile northern pike (Esoxlucius) inhabiting lakes downstream of a uranium mill. Aquat. Toxicol. 92, 240e249. Laale, H.W., 1977. The biology and use of zebrafish, Brachydanio rerio in fisheries research. A literature review. J. Fish Biol. 10, 121-122, IN121eIN122, 123e173. Lerebours, A., Adam-Guillermin, C., Brethes, D., Frelon, S., Floriani, M., Camilleri, V., Garnier-Laplace, J., Bourdineaud, J.P., 2010. Mitochondrial energetic metabolism perturbations in skeletal muscles and brain of zebrafish (Danio rerio) exposed to low concentrations of waterborne uranium. Aquat. Toxicol. 100, 66e74. Lerebours, A., Cambier, S., Hislop, L., Adam-Guillermin, C., Bourdineaud, J.P., 2013. Genotoxic effects of exposure to waterborne uranium, dietary methylmercury and hyperoxia in zebrafish assessed by the quantitative RAPD-PCR method. Mutat. Res. Genet. Toxicol. Environ. Mutagen. 755, 55e60.
53
Lerebours, A., Gonzalez, P., Adam, C., Camilleri, V., Bourdineaud, J.P., GarnierLaplace, J., 2009. Comparative analysis of gene expression in brain, liver, skeletal muscles, and gills of zebrafish (Danio rerio) exposed to environmentally relevant waterborne uranium concentrations. Environ. Toxicol. Chem. 28, 1271e1278. Lourenço, J., Silva, A., Carvalho, F., Oliveira, J., Malta, M., Mendo, S., Gonçalves, F., Pereira, R., 2011. Histopathological changes in the earthworm Eisenia andrei associated with the exposure to metals and radionuclides. Chemosphere 85, 1630e1634. Marques, S.M., Antunes, S.C., Pissarra, H., Pereira, M.L., Gonçalves, F., Pereira, R., 2009. Histopathological changes and erythrocytic nuclear abnormalities in Iberian green frogs (Ranaperezi Seoane) from a uranium mine pond. Aquat. Toxicol. 91, 187e195. Massarin, S., Beaudouin, R., Zeman, F., Floriani, M., Gilbin, R., Alonzo, F., Pery, A.R.R., 2011. Biology-based modeling to analyze uranium toxicity data on Daphnia magna in a multigeneration study. Environ. Sci. Technol. 45, 4151e4158. Mathews, T., Beaugelin-Seiller, K., Garnier-Laplace, J., Gilbin, R., Adam, C., DellaVedova, C., 2009. A probabilistic assessment of the chemical and radiological risks of chronic exposure to uranium in freshwater ecosystems. Environ. Sci. Technol. 43, 6684e6690. , I., Garnier-Laplace, J., AdamPereira, S., Camilleri, V., Floriani, M., Cavalie Guillermin, C., 2012. Genotoxicity of uranium contamination in embryonic zebrafish cells. Aquat. Toxicol. 109, 11e16.