Environmental Toxicology and Chemistry, Vol. 30, No. 1 pp. 220–225, 2011# 2010 SETAC

Printed in the USADOI: 10.1002/etc.381

EFFECTS OF DIETARY URANIUM ON REPRODUCTIVE ENDPOINTS—FECUNDITY,

SURVIVAL, REPRODUCTIVE SUCCESS—OF THE FISH DANIO RERIO

OLIVIER SIMON,* ELMINA MOTTIN, BENJAMIN GEFFROY, and THOMAS HINTON

Institut de Radioprotection et de Surete Nucleaire, 13115 Saint Paul-Lez-Durance, France

(Submitted 5 September 2010; Returned for Revision 24 March 2010; Accepted 2 September 2010)

* T(olivie

Pub(wileyo

Abstract—Exposure to metal-contaminated water has been shown to result in a number of reproductive abnormalities in adult and larvaefish, such as failure of oocyte maturation and teratogenic effects. Recently, dietary uptake of metals by fish has been recognized as acritical route of exposure, however, the mechanisms of metal uptake and toxicity are poorly understood and in need of furtherinvestigation. The objectives of the present study are to quantify uranium (U dietary transfers from spiked artificial diets) in Danio reriotissues and embryos, as well as establish its effect on reproduction and embryonic development. Uranium’s environmental prominence iscurrently increasing because of new mining and milling activities. Uranium concentrations range from 0.02mg/L in natural waters to2 mg/L. The focus of this study was to examine the trophic transfer and effects of U following exposure modalities (dose, exposureduration 1 to 20 d). Two different isotopes were used to distinguish between chemical and radioactivity toxicity of U. Results showedthat U trophic transfer was low (0.52%). Uranium tissue distributions showed that accumulation occurred in digestive organs (liver,digestive tract) following dietary exposure. High levels of U were measured in the gonads (female in particular, >20% of relativeburden). High U accumulation levels in eggs indicated maternal transfer of the contaminant. Moreover, U trophic exposure ledto a reduction in reproduction success as a function of U accumulated levels. High U exposure conditions strongly reduced the totalnumber of eggs (50%) and their viability at 10 d (reduction of the clutch number, low quality of eggs). Environ. Toxicol. Chem.2011;30:220–225. # 2010 SETAC

Keywords—Trophic transfer Reproductive endpoints Danio rerio Uranium

INTRODUCTION

Exposure to metals can cause a number of reproductiveabnormalities in adult and larvae fish. Metals, such as Cd,are known to affect oocyte maturation, reduce vitellogeninsynthesis, and cause ovulation and teratogenic effects [1,2].Such reproductive effects may have serious implications onnormal population dynamics and community structure. Themechanisms of uptake and toxicity to waterborne metals(Cu, Zn, Cd) have been reasonably well studied [3,4]. However,such mechanisms and toxicity are not as well characterized fordietary exposures [5,6]. Recently, the dietary (i.e., trophic)transfer of metals to fish was recognized as a critical exposureroute in need of further investigation [7–10]. Indeed, theconsumption of metal-contaminated food can pose a healthrisk to wildlife as well as to humans [9]. Some data on trophictransfers of metals exists, but the trophic transfer of uranium(U) in the fish Coregonus clupeaformis has been demonstrated[11]. Moreover, during vitellogenesis, many exogenous andendogenous materials are transferred from the mother fish tothe oocytes and, subsequently, to eggs and yolk sac larvae.Maternal transfer of bioaccumulated contaminants to offspringmay be an important and overlooked mechanism of impairedreproductive success that affects fish populations [12,13].

Uranium is a naturally occurring radioactive metal, interest-ing from an ecotoxicological perspective because it can haveboth chemical and radiological toxicity [11]. Uranium’s chemo-toxicity is known to accumulate in and damage the liver inC. clupeaformis and in the zebrafish, Danio rerio [11,14].

Uranium’s environmental prominence is currently increas-ing because of new mining and milling activities to support the

o whom correspondence may be addressedr.simon@irsn.fr).lished online 15 October 2010 in Wiley Online Librarynlinelibrary.com).

220

resurging commercial nuclear power industry (in response toenergy production needs with low carbon output). Such anthro-pogenic activities can increase environmental concentrationsof U [15,16], from typical levels of 0.02mg/L in natural watersto approximately 2 mg/L. Enhanced concentrations cause themetal to accumulate in biota. For example, exposure to lowlevels of aqueous U resulted in concentrations of 1 to 20mg/gfresh weight in the bivalve Corbicula fluminea [17] and inD. rerio [14].

Although data exists on the transfer and effects of U-con-taminated water on adult and larvae stages of D. rerio [18,19] (acommonly used biological model in ecotoxicology [20,21]), theeffects of U on reproduction and the developing embryo havenot yet been studied in fish. Furthermore, dietary exposureshave not been examined from the perspective of chemicalversus radiological toxicity. Therefore, the main objective ofthe present study was to assess reproductive effects of dietaryexposures to U in the zebrafish. Concentration-dependenttests typical of those conducted in the ecotoxicological scienceswere performed. Tissue distributions, as well as uptake andelimination kinetics, were also studied. The second objectivewas to distinguish between the effects of U chemotoxicity andradiotoxicity. An innovative use of two different isotopes of U(with radioactivity levels differing by four orders of magnitude)enabled the distinction of chemical versus radiotoxicity fromdietary uptake.

MATERIALS AND METHODS

Fish husbandry

Adult fish, purchased from Aquasylva, were acclimatized totap water (pH¼ 8.1; Naþ¼ 16.1; Mg2þ¼ 10.6; Ca2þ¼ 79.4;Kþ¼ 1.43; Cl-¼ 25; NO�

3 ¼ 3.4; SO4�2 ¼ 95.8; PO3�

4 ¼ 0.4 inmg/L), kept under a 14:10 h regime at 26� 18C, and fed 3/dwith commercial flake and pellet food for trout. Prior to the

Reproductive effects of dietary uranium in D. rerio Environ. Toxicol. Chem. 30, 2011 221

experiment, fish had been acclimatized to a diet of preparedtrout food. The reproductive ability of the fish under theseconditions was confirmed prior to the start of the experiments.The visible genital papilla of mature egg-bearing females wasused to help distinguish the sexes. Individual groups of genitors(sex ratio¼ 1:2 females:males) were placed separately in small(3-L) spawning aquaria to keep the fish from eating the newlyspawned eggs. Genitor grouping was considered ready forexperimental use when egg viability reached 80%.

Experimental studies

Contaminated food preparation. In an attempt to distinguishthe chemotoxic effects of uranium from its radiotoxicity,two kinds of contaminated diets containing either depleteduranium (DU) or 233U were used. Depleted U is 238U fromwhich the naturally occurring radioisotope 235U has beenlargely removed, and 233U is an artificial isotope of U producedby neutron irradiation of thorium-232. The two U isotopes haveidentical chemical characteristics, but differ in specific activ-ities (1.44� 104 and 3.57� 108 Bq/g, respectively), with233U being 30,000 times more radioactive than DU.

Uranium-contaminated fish food was made by grindingcommercial trout food in a blender and then adding aqueoussolutions of uranyl nitrate (DU, Sigma-Aldrich, 2.38 mg/L) or233U (Institute for Reference Materials and Measurements,1 mg/L) to produce nominal concentrations (control; 5, 50,and 500mg/g) in an agar solution (10 g/L). The food-doughwas dried and cut into pellets that were easily eaten by fish.Pellets were weighed in aliquots for daily rationing. Comparedto nominal concentrations, the measured mean (�SD; n¼ 3) Uconcentrations in pellets were 0.03� 0.009, 4.8� 2.5,58.2� 11.96, and 448� 79mg/g, respectively. Fish were auto-matically fed three times a day with the prepared food, corre-sponding to 5% of fish weight per day. This method accountedfor the significant difference in mass between males andfemales. Preliminary studies using feed rations that varied from1 to 20% of the fish body weight confirmed that a food massequal to 5% of the body weight was rapidly and entirelyconsumed by fish.

Experimental dietary exposures. Danio rerio were exposedto four dietary treatments: control; environmentally low (5mg233U/g, fresh wt); intermediate (50mg/g, fresh wt of DU and233U); and high (500mg DU/g, fresh wt) concentrations of U for20 d. Control, intermediate, and high DU concentrations werereplicated three times (10 males and five females per conditionper replicate). Low and middle 233U diets were replicated once(10 males and five females per condition).

During each exposure series, 10 males and five females wereheld separately in two 15-L tanks. A flow-through system, ata rate of 1 ml/min (100% daily water change), was used toreduce aqueous exposure and to assist feces removal. Uraniumconcentrations in the water of each tank were assayed daily.

Tissue distributions: Kinetics of uptake and depuration

Tissue distributions of U were determined in five femalesafter a single feeding of 50mg 233U/g. Females were sacrificed24 h after the meal, and the concentrations of U in varioustissues were determined.

A second experiment was conducted to determine thekinetics of U accumulation and depuration, as well as to furtherexamine tissue distributions. Danio rerio females were fedcontaminated trout pellets (500mg DU/g) for 20 d, followedby 10 d on uncontaminated, control feed. Ten females were

sampled after the initial 5 d of feeding; another 10 females weresampled after 20 d, and the remaining 10 fish after 30 d (the last10 d on a clean diet to determine elimination rates).

Reproductive effects after 20 d of exposure

Reproductive effects, following 20 d of exposure to fourdietary U treatments (control, 50mg/g fresh wt of DU, 50mg/gfresh wt 233U, and 500mg/g fresh wt DU), were determined bybreeding fish for 2 d in clean water (and clean diet). Eachbreeding replicate consisted of a breeding colony composed oftwo males and one female; with five replicate breeding coloniesused per dietary condition. Reproductive success (number offemales that spawned) was measured. Live and dead, opaqueeggs were counted after 24 h postfertilization to assess eggviability rate (through the optic observation of the heart rate).Fifty live eggs (5 replicates � 10 eggs) were kept in an incubator(268C) for 10 d to determine mortality and hatching success(median hatching time) [19].

Uranium detection and tissues measurement

Uranium concentrations in water were analyzed dailyby means of inductively coupled plasma–atomic emissionspectrometry (ICP-AES; Optima 4300DV, Perkin-Elmer) aftera 2% (v/v) HNO3 acidification step.

All fish were anesthetized with Tricaine mesylate (MS-222).The distribution of U in the fish was analyzed in nine organs(digestive tract, liver, gills, skin, bones, muscle, gonads, eggs,and remaining body). Each biological sample was digested ina polypropylene tube using nitric acid (1 ml, Merck, 65% at958C for 1 h 30 min, and at 1058C) and perchloric acid (0.5 ml,Merck, 33% at 1058C).

Two methods were used to assay U in biological samples,depending on the isotope of interest. Depleted uranium wasanalyzed by inductively coupled plasma–mass spectrometry(ICP-MS; PQ Excell, ThermoElectron with S-Option) orICP-AES according to detection limit (0.1 ng/L or 10mg/L,respectively). The 233U concentrations were determinedby alpha spectroscopy. Samples were placed in a liquidscintillation cocktail (Instagel, Packard Instruments) for anal-ysis (alpha particle detection limit was 30 mBq or 0.084 ng Uper sample; using a Quantulus 1220, Wallac, ultra low-levelliquid scintillation instrument).

Statistical analysis

All data are presented as mean� SE with significance takenat p< 0.05. All statistics were performed with SigmaStat, Systatsoftware. Accumulation levels and hatching rate werecompared using analysis of variance (ANOVA) followed bythe post hoc multicomparison Tukey test and Fisher test,respectively.

RESULTS

No adult fish mortality occurred during the experimentalperiod.

Accumulation

Whole body and dietary transfer ratios. Uranium accumu-lation was detectable in fish under all conditions. Uraniumconcentrations in fish ranged from 13 to 670 ng/g (in females)with increasing U concentrations in the food (5 to 500mg/g;Table 1). Dietary transfer ratios (DTR: total body burden of U[mg] in fish at the end of 20 d/total U load [mg] provided through

Table 1. Average uranium (U) concentration (ng/g, fresh wt), total burden (mg), and DTR (%) measured in females and males after 20 d of trophic exposureconditionsa

Exposureconditions in food(mg/g, dry wt) Sex

Mass of fish(g, fresh wt)

[U](ng/g, fresh wt)

Total Uprovided to fish

in food (mg)Total U body

burden in fish (mg)DTR(%)

233U 5 Male 0.27� 0.12 26� 9 1.35 0.007 0.525 Female 0.78� 0.19 13� 2 3.9 0.020 0.26

DU 50 Male 0.29� 0.11 126� 37 16 0.037 0.2350 Female 0.82� 0.17 126� 17 41 0.100 0.25

233U 50 Male 0.37� 0.09 153� 56 18.5 0.057 0.3150 Female 0.76� 0.22 116� 47 38 0.088 0.23

DU 500 Male 0.73� 0.16 365� 108 365 0.270 0.07500 Female 1.28� 0.3 670� 127 640 0.860 0.13

a DU¼ depleted uranium; DTR¼ dietary transfer ratios.

222 Environ. Toxicol. Chem. 30, 2011 O. Simon et al.

contaminated pellets � 100) suggests that the highest dietaryuptake (i.e., largest DTR of 0.52%) occurred in male fish thatwere fed a diet containing the lowest U concentration. Highexposure conditions led to a lower accumulation in females thanin males (p¼ 0.036), even though female mass and females aretypically 223% larger than males).

In skin, bone, liver, gonads, and eggs. Tissue distributionsof U were investigated for all dietary exposure conditions.However, no significant differences among the tissues wereobserved as a function of dietary U concentrations, for females(p¼ 0.85) or males (p¼ 0.78). These analyses were carriedout after the fish had been on contaminated diets for 20 d,followed by 2 d of gut clearance on clean diets during thereproduction period. Figure 1 shows the relative U tissue burden(percentage U in tissue/total body burden of U) measured ineight organs of female fish fed low-contaminated pellets. Themass of the liver was not measured. The relative mass of tissuesremained constant between experimentations. The rest of thebody (25%), skin, and gonads (20%) contained the highestrelative burdens. The other tissues represented a comparativelylarge mass (from 23 to 40%) and contained organs such asthe kidney, spleen, and heart, which may explain the relativelyhigh U burden observed, particularly since the kidney isconsidered to be a target organ for U accumulation [18].Uranium accumulation also occurred in other organs (Fig. 1),such as bone, digestive tract (10%), muscle, gills, and liver(<10%). Although the liver is the main organ for detoxifying

0

10

20

30

40

50

Liver Gills Muscle Gonades Skin Bone DT RB

Urelativeburden(%)

Female

Male

U233 5 g/g

Fig. 1. Average uranium relative burden (%) in the main organs (liver, gills,muscle, gonads, skin, bone, digestive tract [DT], and rest of the body [RB]) offish Danio rerio after 20 contaminated meals at 5mg 233U/g.

contaminants, accumulation level was found not to be directlylinked to exposure. Mineralized tissues (bones, and skin withscales) were sites of high U accumulation.

No significant differences in tissue distributions of U wereobserved between males and females (Fig. 1), except for thegonads (p¼ 4.3� 10�5), where ovaries consistently had higherrelative concentrations. For example, U in testes represented6.5 and 1.8% of the total U body burden for the 5 and 50mg/g diets, whereas, U in the ovaries ranged from 17 to 26% of thetotal female’s body burden, increasing with dietary U (Table 2).Uranium accumulation in ovaries strongly correlated to wholebody accumulation in the female (Burden in female [ng]¼ 2.64 �Burden in ovary; n¼ 15; R2¼ 0.97). Females (n¼ 5 per exposuretime) fed high concentrations of DU (500mg/g) had a rapidaccumulation in the ovaries (31� 29% relative to whole bodyburden at 5 d) that did not differ significantly after an additional15 d on contaminated diets (27� 14%; Fig. 2). The mass of thegonads remained constant over time (5 d: 0.06� 0.02 g; 20d: 0.08� 0.03 g; 20 dþ 10 d: 0.85� 0.03 g). The relativeburden in the ovaries seemed to increase (50� 28% relativeto whole body burden; Fig. 2) following a 10-d depurationperiod when the fish were on clean diets. The clearance fromthe ovaries has implications to U transfer to eggs. Notably,transfer to eggs was detected, particularly from females on thehighest U diet (Table 2).

The temporal kinetics of uptake and clearance describedabove for the ovaries were not observed for other organs. Forexample, neither the gills, skin, nor bones had detectableU concentrations after just 5 d on clean diets (Fig. 2). Addi-tionally, the bones’ slow metabolic rate is suggested by theretention of U, even after 10 d of clean diet (Fig. 2). Relativeburdens of U in the digestive tract were high (30� 12%) after5 d, then decreased strongly at 20 d (3� 0.5%) as the whole

Table 2. Uranium (U) concentration (mg/g, fresh wt) and relative burden(%) in ovary and eggsa

Exposureconditions in food(mg/g and Bq/g)

Exposureduration (d) Ovary Eggs

[U](ng/g)

Relativeburden (%)

[U](ng/g)

233U 5 and 1,780 20 42.4� 51 17� 42 ND50 and 17,800 20 224� 187 23� 10 146� 158

(n¼ 3)DU 500 and 10 20 951� 625 26� 15 850� 160

a DU¼ depleted uranium; ND¼ no data.

0

20

40

60

80

100

Liver Gills DT Skin Bones Ovary RB

Relativeburden

(%)

5d20d20d+10d

<LD

Fig. 2. Average uranium relative burden (%) in the main organs (liver, gills,digestive tract [DT], skin, bones, ovary, and rest of the body [RB]) of femalefishDanio rerio after 5 and 20 contaminated meals, and 20 d followed by 10 dwith control pellets at 500mg DU/g.

Reproductive effects of dietary uranium in D. rerio Environ. Toxicol. Chem. 30, 2011 223

body burden increased. After 10 d of depuration on a clean diet,no U was detected in the dietary tract (Fig. 2). The initialdominance of U within the digestive tract was also observed ina complementary experiment in which fish (n¼ 5) were fedone meal containing 50mg 233U/g. Uranium concentrations inthe digestive tract were 50� 9% of the total U accumulated.

Depleted uranium accumulation in the liver was below thedetection limit for ICP-AES during the 5–20–30-d uptakeand elimination experiment (Fig. 2). In another experimentin which five fish were fed one meal containing 50mg 233U/g, accumulation was measured in the liver (10.7� 8% of totalbody burdens). Differences in the two experiments, relative tothe detection of U in the liver, are likely due to the respectivemeasurement methods and not physiological differences in thetwo U isotopes; alpha spectroscopy of 233U is a more sensitivemethod than ICP-AES method of detecting DU.

Reproductive effects

Reproduction success (number of clutch, number andviability of eggs). Reproductive effects were tested in 50females [15 females from each of three DU dietary treatments(control, 50mg DU/g, and 500mg DU/g) and 5 females from233U dietary treatment (50mg 233U/g)]. All females were fedwith contaminated pellets for 20 d. Only three of the fivefemales on the 50mg 233U/g diet spawned (Table 3). However,

Table 3. Reproduction success expressed by the number of clutch, the total egg numviability (%� SE), and the larval mortality

Exposure conditions[U] in diet(mg/g and Bq/g)

No. of femaleswith clutch Total no.

Fecundityb

(mean� SD; ranno. of females

Control 8 of 15 5,576 372� 441; 0–126DU 50 and 1 8 of 15 5,516 368� 396; 0–983

500 and 10 6 of 15 2,828 180� 334; 0–988233U 50 and 17,800 3 of 5 429 86� 107; 0–226

a U¼ uranium; DU¼ depleted uranium.b Mean fecundity was based on total number of females, including those that proc Combined impact¼mean fecundity � (mean % egg viability/100) � (mean % la

this percentage of females with unsuccessful clutches was notsignificantly different from control fish (Fisher test, p¼ 1),where only 8 of the 15 females produced eggs. Forty percentof the females exposed to the highest DU diet produced eggs(Table 3).

Of the clutches that were successful, the number of eggs perfemale in the control group was 1.4 times lower than forthe females on the highest U diet conditions. Consequently,the total number of eggs spawned was half that for fish on thehigh U diet versus control fish. Equally important, mean eggviability was higher in control fish: Six of the eight clutcheshad an egg viability of greater than 80% (indicative of good eggquality), whereas only one clutch among the females exposedto U had similar viability.

Hatching time and larvae mortality after 10 s exposure. Theonset of hatching was monitored every day. The mean timerequired for 50% of the eggs to hatch was 76.8� 8.1 h in controlfish. This dropped to 66.3� 9 h in fish exposed to 50mg DU/g,and 69.3� 12 h in females exposed to 500mg DU/g. This trend,however, was not statistically different between treatments(Fisher test, p¼ 0.9).

Survival of developing larvae was monitored daily until 10-dpostfertilization. Mean survival of larvae from control fishwas 77� 8% (Table 3), whereas mean survival decreasedsignificantly (<25%) for all larvae produced from parents ondietary U treatments (Table 3). A combined impact factor wascalculated that incorporated fecundity, egg viability, and larvaesurvival (Table 3). When combined, these factors showed all Udietary treatments caused a reduction in the number of livinglarvae (at 10-d postfertilization) of approximately one order ofmagnitude compared to controls.

Data in Table 3 suggests that the mean viability of eggs wasinfluenced by high chemical concentrations of U (500mg DU/g)or from high radioactivity concentrations (17,800 Bq/g of 233U).Instead, the overall combined impact appears to be influencedmore by reduced fecundity and effects in early survival of thedeveloping larvae (Table 3). Fecundity seemed to be differentbetween DU 50 and 233U treatments of 50 Ug/g (respectivemeans of 368 and 86), which implies that the radioactivity didhave an impact. Further experiments are required to confirmthis observation.

DISCUSSION

In the present study, the primary pathway of fish contam-ination was through the diet. Average U concentrations in waterover 20 d of exposure were below ICP-AES detection limits(10mg/L) for the low and middle U dietary treatments. The500mg/g U concentration in pellets resulted in tank water

ber, the fecundity (number of eggs per female, average, min; max), the eggs(%) at 10-d postfertilization (dpf)a

Eggs Larvae

ge;)

Meanviability (%)

No.of females withviability >80%

Survival(%) at 10 dpf

Combinedimpactc

1; 15 66� 36 5 77� 8 190� 12; 15 57� 23 1 13� 23 27� 21; 15 62� 30 1 16� 32 19� 33; 5 59� 29 1 21� 26 11� 8

duced no eggs.rvae survival/100).

224 Environ. Toxicol. Chem. 30, 2011 O. Simon et al.

having 17� 6.5mgU/L (n¼ 38). Thus, some nondietary expo-sure may have occurred at the highest treatment level. However,a high pH (8.1) and alkaline conditions in the tap water wouldstrongly decrease the proportion of bioavailable UO2þ

2

and UO2OHþ ions in the water [22,23], reducing the probabilityof uptake of U from the water.

Dietary transfer ratios (ratio of metal concentration inorganisms to metal concentration in food) were higher in malesthan in females, but remained less than 0.52%. Thus, thelow DTR, as proposed by Mathews and Fisher [10], confirmedthat U biomagnification, as observed for other metals (Am, Cd,Mn, and Co) is unlikely. The DTR derived from spiked pelletedfood was less than DTR values derived from fishing eatingnaturally contaminated prey. Dietary transfer ratios ranged from1 to 13% after dietary U exposure between C. fluminea and thecrayfish Orconectes limosus [17]. Recent investigations suggestthat metals stored in natural food may be in a form that ismore available than spiked commercial food as used in thepresent study [24]. Indeed, the trophic availability of the metalscould depend on the metal accumulation in specific subcellularcomponents of the prey [25]. However, the mechanisms areunclear and do not seem to be operative in all cases.For example, no difference was observed in the trophic transferefficiency of Cd between a diet of waterborne exposed chiro-nomids and a diet of manufactured pellet food [8]. Furtherexperiments are therefore needed to compare the resultsobtained in our artificially contaminated U diets to those fromnaturally contaminated prey.

Relatively high levels of U in the gonads (particularly in theovaries) were observed. Fish exposed to heavy metals (Cu, Cd,Pb) often show elevated levels of metals in the gonads [1,26,27].As proposed by Cooley and Klaverkamp [11], gonadal matura-tion could favor U accumulation by binding to the vitellogenin.Vitellogenin is an egg yolk precursor protein and is a glyco-lipo-phosphoprotein normally synthesized in the liver. The Ubinding ability to this protein could be explained by the highaffinity of UO2þ

2 to the phosphate group [28].The high U accumulation levels in eggs furthermore

indicated maternal transfer of the contaminant. Metal transfersfrom females to eggs have been shown for Cd and Hg [2,29,30]for the cuttlefish Sepia officinalis after trophic transfer [31].

The developing fish embryo can also be adversely affectedby direct metal accumulation from the water. After 2 d of directexposure (250mg DU/L), the embryo and chorion of eggsaccumulated close to 2mg/g dry weight and 10 mg/g dry weight,respectively. These exposures caused significant declines inhatching success [19]. In the future, it will be important to knowthe U distribution in chorion and in embryos to better assessthe U maternal transfer efficiency. Uranium transfer throughmaternal contamination could potentially cause an enhancedresistance in the progeny. For example, in Nile tilapia (Oreo-chromis niloticus), the first generation of progeny (F1) froma parent exposed to a metal appeared more resistant to heavymetals than those not exposed [26].

The number of eggs produced per gravid female (closeto 700, Table 3) in control animals was higher than thoseobserved in the literature. Typically, the number of eggsper female is a function of the female’s mass. An averageproduction for a colony (one female, 0.37� 0.1 gþ two males)varies between 50 and 150 eggs. Moreover, the egg outputis cyclical with peaks every 3–4 d [32]. In the presentstudy, experimental design (colony: two malesþ one female,48 h as reproduction period), quality of adult food, andionic composition of water could have favored an enhanced

reproductive capacity. Regardless of the cause, the same exper-imental design and husbandry conditions were used for alltreatments in the present study.

All U treatments in the present study resulted in a combinedimpact (fecundityþ egg viabilityþ larvae survival) that ulti-mately reduced the number of surviving larvae fish. Fecundityseemed to be different between DU 50 and 233U 50 treatments(respective means of 368 and 86). The combined impact seemsto have been the result of both chemical and radiotoxic mech-anisms (Table 3). Identical results have been observed formetals, such as As, and MeHg after contaminated dietconsumption [5,9]. The reduction of fecundity could be linkedto a direct effect of U in females, but the U may interfere withthe spawning behavior of either the male or female fish.Methylmercury and lead have been found to alter spawningbehavior in male fathead minnows [9,33,34].

In the present study, U exposure led to a mortality rate of10-d postfertilization, comparable to those obtained after directexposure of eggs [19]. It is noteworthy that the accumulationlevels in eggs after trophic transfer (Table 2; 0.85mg/g) wereclose to half those measured after high direct exposure level ofeggs (250mg/L); experiments based on mature adult stage afterdirect exposure showed significant U effect (decrease of theeggs production/fecundity and on the ability to spawn).

Finally, taking into account the reproductive strategies ofthis oviparous species (i.e., small, numerous eggs; small paren-tal investment per egg; rapid development), the level of impactobserved from the U treatments would probably not havedetrimental consequences on the population, but would leadto a decrease in the population level. Uranium exposure couldaffect the dynamics of population and, consequently, the struc-ture and function of the ecosystem as observed after directexposure. The life history strategies of individual species,however, are known to significantly influence populationresponses to contaminants. Thus, the level of impact observedin the present study might be quite different in other species offish with different life history characteristics.

Hatching rate was considered to be a good endpoint forreproductive-toxicity assessment. Delay in hatching can have aslarge an impact on some populations as increased mortality orreduced fecundity [35]. A delay in the start of hatching wasobserved after direct exposure of D. rerio eggs (Cd, U) [19,20].However, dietary exposure did not show this effect; the dis-tribution of U in eggs, chorion, and embryo was necessary toconfirm the relationship between U accumulation in eggs andeffects.

Commercial contaminated food spiked with U led to accu-mulation in all tissues examined, with particularly high con-centrations observed in the ovaries. Effects on egg viability andlarval survival were observed for the lower DU exposureconditions. The no-observed- adverse-effect level is thereforelower than 50mg/g in diets, and close to field bioaccumulationlevels. Effects following 233U conditions were observed, indi-cating that radiotoxicity should be studied.

The use of contaminated organisms (daphnids or chirono-mids) as prey in further experiments would permit the compar-ison of natural versus manufactured diet in the trophic transfer.Currently, risk assessment studies have mainly been based onexposures of test animals to dissolved metals; this data suggeststhe importance of considering the contribution of dietary met-als. Trophic exposure in controlled conditions, as in the presentstudy, are easy to reproduce and interpret in direct-exposurestandardized tests. Such exposure modalities could feasibly beadded to guidelines for establishing water quality.

Reproductive effects of dietary uranium in D. rerio Environ. Toxicol. Chem. 30, 2011 225

Acknowledgement—This work is part of the Radioprotection de l’Environ-nement a l’homme (ENVIRHOM) research program supported by theInstitute for Radioprotection and Nuclear Safety in France. We especiallythank Virginie Camilleri, Sandrine Frelon, and Marcel Morello for technicalassistance with the ICP-AES/MS and scintillation liquid alpha measure-ments. The authors thank Katherine Harper for her assistance in proofreadingthe manuscript.

REFERENCES

1. Thompson J, Bannigan J. 2008. Toxic effects on the reproductive systemand the embryo. Reprod Toxicol 25:304–315.

2. Sellin MK, Kolok AS. 2006. Cadmium exposures during earlydevelopment: Do they lead to reproductive impairment in fatheadminnows. Environ Toxicol Chem 25:2957–2963.

3. Olsson PE, Kling P, Hogstrand C. 1998. Mechanisms of heavy metalaccumulation and toxicity in fish. In Langston WJ, Bebianno MJ, edsMetal Metabolism in Aquatic Environments. Chapman & Hall, London,UK, pp 321–350.

4. McGeer JC, Szebedinszky C, McDonald G, Wood CM. 2000. Effects ofchronic sublethal exposure to waterborneCu, Cd or Zn in rainbowtrout 2:Tissue specific metal accumulation. Aquat Toxicol 50:245–256.

5. Boyle D, Brix KV, Amlund H, Lundebye AK, Hogstrand C, Bury NR.2008. Natural arsenic contaminated diets perturb reproduction in fish.Environ Sci Technol 42:5354–5360.

6. Clearwater SJ, Farag AM, Meyer JS. 2002. Bioavailability and toxicityof dietborne copper and zinc in fish. Comp BiochemPhysiol C 132:269–313.

7. Lapointe D, Couture P. 2009. Influence of the route of exposure on theaccumulation and subcellular distribution of nickel and thallium injuvenile fathead minnows Pimephales promelas. Arch Environ ContamToxicol 57:571–580.

8. Bechard KM, Gillis PL, Wood CM. 2008. Trophic transfer of Cdfrom larval chironomids (Chironomus riparius) exposed via sedimentor waterborne routes, to zebrafish (Danio rerio): Tissue-specific andsubcellular comparisons. Aquat Toxicol (Amst) 90:310–321.

9. Hammerschmidt CR, Sandheinrich MB, Wiener JG, Rada RG. 2002.Effects of dietary methylmercury on reproduction of fathead minnows.Environ Sci Technol 36:877–883.

10. Mathews T, Fisher NS. 2008. Trophic transfer of seven trace metals in afour-step marine food chain. Mar Ecol Prog Ser 367:23–33.

11. Cooley HM, Klaverkamp JF. 2000. Accumulation and distribution ofdietary uranium in lake whitefish (Coregonus clupeaformis). AquatToxicol (Amst) 48:477–494.

12. Muscatello JR, Janz DM. 2009. Assessment of larval deformities andselenium accumulation in northern pike (Esox lucius) and white sucker(Casostomus commersoni) exposed to metal mining effluent. EnvironToxicol Chem 28:609–618.

13. Hammerschmidt CR, Sandheinrich MB. 2005. Maternal diet duringoogenesis is the major source of methylmercury in fish embryos.EnvironSci Technol 39:3580–3584.

14. Lerebours A, Gonzalez P, Adam C, Camilleri V, Bourdineaud JP,Garnier-Laplace J. 2009. Comparative analysis of gene expression inbrain, liver skeletal muscles, and gills of zebrafish (Danio rerio) exposedto environmentally relevant waterborne uranium concentrations.Environ Toxicol Chem 28:1271–1278.

15. Colle C, Garnier-Laplace J, Roussel-Debet S, Adam C, Baudin JP. 2001.Comportement de l’uranium dans l’environnement. In Metivier H, ed,L’uranium de l’environnement a l’homme. EDP Sciences, Les Ulis,France, pp 187–211.

16. Bonin B, Blanc PL. 2001. L’uranium dans le milieu naturel, des originesa la mine. In Metivier H, ed, L’uranium de l’environnement a l’homme.EDP Sciences, Les Ulis, France, pp 8–41.

17. Simon O, Garnier-Laplace J. 2005. Laboratory and field assessment ofuranium trophic transfer efficiency in the crayfish Orconectes limosusfed the bivalve C. fluminea. Aquat Toxicol (Amst) 74:372–383.

18. Barillet S, Adam C, Palluel O, Devaux A. 2007. Bioaccumulation,oxidative stress, and neurotoxicity in Danio rerio exposed to differentisotopic compositions of uranium. Environ Toxicol Chem 26:497–505.

19. Bourrachot S, Simon O, Gilbin R. 2008. The effects of waterborneuranium on the hatching success, development, and survival of earlylife stages of zebrafish (Danio rerio). Aquat Toxicol (Amst) 90:29–36.

20. Fraysse B, Mons R, Garric J. 2006. Development of a zebrafish 4-dayembryo-larval bioassay to assess toxicity of chemicals. EcotoxicolEnviron Saf 63:253–267.

21. Nagel R. 2002. DarT: The embryo test with the zebrafish D. rerio—Ageneral model in ecotoxicology and toxicology.ALTEX:Alternativen zuTierexperimenten 19:38–48.

22. Fortin C, Denison FH, Garnier-Laplace J. 2007. Metal-phytoplanktoninteractions: Modelling the effect of competing ions (Hþ, Ca2þ, andMg2þ) on uranium uptake. Environ Toxicol Chem 26:242–248.

23. Zeman FA, Gilbin R, Alonzo F, Lecomte-Pradines C, Garnier-Laplace J,Aliaume C. 2008. Effects of waterborne uranium on survival, growth,reproduction and physiological processes of the freshwater cladoceranDaphnia magna. Aquat Toxicol (Amst) 86:370–378.

24. Ng T, Wood CM. 2008. Trophic transfer and dietary toxicity of Cdfrom the oligochaete to the rainbow trout. Aquat Toxicol (Amst) 87:47–59.

25. Wallace WG, Luoma SN. 2003. Subcellular compartmentalization of Cdand Zn in two bivalves. II. Significance of trophically available metal(TAM). Mar Ecol Prog Ser 257:125–137.

26. Lourdes M, Cuvin-Aralar A, Aralar EV. 1995. Resistance to a heavymetal mixture in Oreochromis niloticus progenies from parentschronically exposed to the same metals. Chemosphere 30:953–963.

27. Dural M, Goksu MZL, Ozak AA. 2007. Investigation of heavy metallevels in economically important fish species captured from the TuziaLagoon. Food Chem 102:415–421.

28. Michon J, Frelon S, Garnier C, Coppin F. 2009. Determinations ofuranium(VI) binding properties with some metalloproteins (transferrin,albumin, metallothionein and ferritin) by fluorescence quenching.J Fluoresc 20:581–590.

29. Jezierska B, Lugowska K, Witeska M. 2008. The effects of heavy metalson embryonic development of fish. Fish Physiol Biochem 35:625–640.

30. Hammerschmidt CR, Sandheinrich MB. 2006. Maternal diet duringoogenesis is the major source of methylmercury in fish embryos.EnvironSci Technol 39:3580–3584.

31. Lacoue-Labarthe T, Warnau M, Oberhansli F, Teyssie JL, Jeffre R,Bustamante P. 2008. First experiments on the maternal transfer of metalsin the cuttlefish Sepia officinalis. Mar Pollut Bull 57:826–831.

32. Paull GC, Van Look KJW, Santos EM, Filby AL, Gray DM, Nash JP,Tyler CR. 2008. Variability in measures of reproductive success inlaboratory-kept colonies of zebrafish and implications for studiesaddressing population-level effects of environmental chemicals. AquatToxicol (Amst) 87:115–126.

33. Weber DN. 1993. Exposure to sublethal levels of waterborne leadalters reproductive behavior patterns in fathead minnows (Pimephalespromelas). Neurotoxicology 14:347–358.

34. Sandheinrich MB, Miller KM. 2006. Effects of dietary methylmercuryon reproductive behavior of fathead minnows (Pimephales promelas).Environ Toxicol Chem 25:3053–3057.

35. Alonzo F, Hertel-Aas T, Gilek M, Gilbin R, Oughton DH, Garnier-Laplace J. 2008. Modelling the propagation of effects of chronicexposure to ionizing radiation from individuals to populations. JEnvironRadioact 99:1464–1473.