GENOTOXICITY OF ACUTE AND CHRONIC GAMMA-IRRADIATION ON ZEBRAFISHCELLS AND CONSEQUENCES FOR EMBRYO DEVELOPMENT

SANDRINE PEREIRA,* STEPHANIE BOURRACHOT, ISABELLE CAVALIE, DELPHINE PLAIRE, MORGAN DUTILLEUL,

RODOLPHE GILBIN, and CHRISTELLE ADAM-GUILLERMIN

Institut de Radioprotection et de Surete Nucleaire, DEI/SECRE/LRE, Cadarache, France

(Submitted 15 June 2011; Returned for Revision 11 July 2011; Accepted 5 September 2011)

Abstract—The effects of radiation on biological systems have been studied for many years, and it is now accepted that direct damage toDNA from radiation is the triggering event leading to biological effects. In the present study, DNA damage induced by acute or chronicirradiation was compared at the cellular (zebrafish [Danio rerio] cell line ZF4) and developmental (embryo) levels. Zebrafish ZF4 cellsand embryos (at 3 h postfertilization) were exposed within ranges of acute doses (0.3–2Gy/d) or chronic dose rates (0.1–0.75Gy/d).DNA damage was assessed by immunodetection of g-H2AX and DNA-PK (DNA double-strand breaks) and the alkaline cometassay (DNA single-strand breaks). Zebrafish embryo development and DNA damage were examined after 120 h. At low doses, chronicirradiation induced more residual DNA damage than acute irradiation, but embryo development was normal. From 0.3Gy, a hyper-radiosensitivity phenomenon compared to other species was shown for acute exposure with an increase of DNA damage, animpairment of hatching success, and larvae abnormalities. These results suggest a dose-dependent correlation between unrepairedDNA damage and abnormalities in embryo development, supporting the use of DNA repair proteins as predictive biomarkersof ionizing radiation exposure. This could have important implications for environmental protection. Environ. Toxicol. Chem.2011;30:2831–2837. # 2011 SETAC

Keywords—Gamma irradiations DNA repair g-H2AX Zebrafish Comet assay

INTRODUCTION

DNA double-strand breaks (DSBs) are considered the keylesions responsible for radiation-induced cell death and are lesseasily repaired than other DNA damage. It is now well knownthat mammalian cells possess two major mechanisms forrepairing DSBs: homologous recombination (HR) and non-homologous end-joining (NHEJ) [1].

Nonhomologous end-joining protects cells from the effectsof radiation-induced DNA damage and requires four partners:Ku70, Ku80, DNA-PK, and DNA ligase 4 [2,3]. Each ofthese genes has been knocked out in mice, resulting in severeradiosensitivity [4]. This repair pathway is conserved withinvertebrates. Putative orthologs for each of these were identifiedin the zebrafish (Danio rerio) EST database and studied [5,6].The presence of these genes, together with the functionalcharacterization of the Ku70 and Ku80 genes in the zebrafish,suggest that an intact functional NHEJ probably operates in thezebrafish and also in other fish species.

Double-strand breaks in DNA can be determined indirectlyfrom the nuclear foci formed by DSB-triggered phosphorylationof the variant histone H2AX (g-H2AX), an early NHEJ-depend-ent event that is easily quantifiable by immunofluorescence;a one-to-one correlation was found between DSBs and nuclearg-H2AX foci [7]. DNA-PK autophosphorylation (pDNA-PK)has been demonstrated to be required for rejoining DSBsby NHEJ [8] and can be used as an indicator of NHEJactivity through the quantification of the nuclear foci formed

by pDNA-PK. Indeed, persistent unrepaired DSB can lead to theformation of micronuclei [9].

Despite several studies on DNA damage and repair indifferent fish species [10–12], very few data are available onthe most deleterious ones, the DNA DSBs. Furthermore, little isknown about the effects of acute and chronic exposure toionizing radiation on DNA repair mechanisms in fish [13].Most studies examined the effects of irradiation exposure onsurvival and embryo development in fish exposed to radio-nuclides acutely or chronically [14–20]. Acute and chronicexposures do not lead to the same effects at the organism level:multiple development abnormalities were shown in acute irra-diated embryos in medaka (Oryzias latipes) [18,20], whereasfew morphologic abnormalities were observed on chronicirradiated embryos [21].

At the cellular level, some authors have assessed the effectsof cell irradiation on survival (embryonic ZEB_J2 zebrafishcells) [22] or DNA damage (mutant RIC1 medaka; zebrafishcells and embryos) [23,24].

Primary lesions of DNA such as strand breaks weremeasured in vitro with the comet assay in different zebrafishcell types (male gonad cells, hepatocytes, embryo cells)exposed to chronic external gamma rays after 24 h [25]. Asignificant increase of DNA alterations was observed fromthe first dose rate tested (1mGy/d) for gonad cells versus750mGy/d for hepatocytes. For larvae of the same speciesexposed for 24 h, genotoxic effects were also demonstratedfrom 29mGy/d using the comet assay [25]. Radiation-induceduntargeted germline mutations were also observed in themedaka for chronic exposure (dose rate of 68mGy/d for 45–153 d) [26].

For acute irradiations (doses ranging from 2–10Gy), somestudies relate to the transgenerational genomic instability bystudying mutations in germ cells of medaka [27]. Nevertheless,

Environmental Toxicology and Chemistry, Vol. 30, No. 12, pp. 2831–2837, 2011# 2011 SETAC

Printed in the USADOI: 10.1002/etc.695

All Supplemental Data may be found in the online version of this article.* To whom correspondence may be addressed

(sandrine.pereira@ymail.com).Published online 27 September 2011 in Wiley Online Library

(wileyonlinelibrary.com).

2831

none of these studies have checked the occurrence and impact ofradiation-induced-DNA DSB on embryo development.

Embryogenesis is a particularly radiosensitive stage of thevertebrate life cycle, and zebrafish embryos are ideal forevaluating genotoxic stress as well as radiation-related studies[18,25]. Indeed, it has been shown that the susceptibility of theembryo to lethal effect of radiation seems to diminish after 5 to7 h postfertilization (hpf), which would support the hypothesisthat events occurring at or before gastrulation are particularlysensitive [28]. Although the radiosensitive target is in thenucleus, the precise damage pathways have not been exploredat the molecular level [29]. Some authors suggested that non-homologous end-joining (NHEJ) components mediate DNArepair to promote survival of irradiated cells during embryo-genesis [5].

In the present study we compare the effects of acute andchronic exposure of embryonic zebrafish cells (ZF4) andembryos at the molecular and macroscopic levels. We choseto work with cultured cells, which were easier for screeningthe repair mechanisms and the cellular sensitivity, and tocompare the results with developmental studies to enhancetheir environmental reliance. The studied dose rates in chronicirradiation experiments were chosen around the existing bench-mark level of 10mGy/d recommended by the InternationalAtomic Energy Agency (1992). The total accumulated chronicdoses corresponded to the same order of magnitude as thechosen acute doses. DNA damage and repair after exposureto acute doses (0.3–2Gy) or chronic dose rates (0.1–0.75Gy/d)on cells were assessed by immunodetection of g-H2AX andDNA-PK (DSB repair) and the comet assay (DNA strandbreaks) 24 h postirradiation. Zebrafish embryos aged at 3 hpfwere also exposed to the same doses and dose rates and wereexamined at different developmental stages for morphologicabnormalities and cellular DNA damage. These results allowthe cellular radiosensitivity of zebrafish to be evaluated.

MATERIALS AND METHODS

Cell culture

Embryonic zebrafish fibroblasts (ZF4) were obtained fromthe American Type Culture Collection (CRL-2050). Cells wereroutinely cultured at 288C and 5% CO2 as monolayers withGlutamax Dulbecco’s modified Eagle’s minimum-Ham’s F12medium (Gibco), supplemented with 20% of inactivated fetalcalf serum, penicillin, and streptomycin. All experiments wereperformed with cells in the plateau phase of growth (95–99% ofcells in G0/G1) at passages 7, 8, and 10 to overcome any biasgenerated by cell cycle.

Immunofluorescence

The immunofluorescence protocol employed for DNArepair and signaling factors was described elsewhere [30,31].Anti-pH2AXser139 antibodies (Euromedex) were used at 1:800.Anti-pDNA-PKthr2609 (ab4194; Abcam) was used at 1:100.Incubations with antimouse FITC secondary antibodies wereperformed at 1:100. Slides were mounted in 40,60 diamidino-2-phenyl-indole (DAPI)-stained Vectashield (Abcys) and exam-ined with a Nikon fluorescence microscope. Two hundred to600 nuclei were analyzed by experiment. Nuclei staining withDAPI also permitted indirect evaluation of the yield of G1

cells (homogeneous DAPI staining) and micronuclei. Onehundred cells were counted for micronuclei analysis. In theseconditions the DNA-PK kinases that phosphorylate the H2AXhistones require some minutes postirradiation to be fully active.

Consequently, the earliest time postirradiation was fixed at10min, in agreement with previous reports [31].

Gamma irradiations

An IBL 637 irradiator (137Cs, 1Gy/min) with 137Cs sourcespurchased from CERCA-LEA (Framatome ANP) was used toperform acute gamma irradiations. Incubation at 288C wasapplied immediately after irradiation.

Gamma rays were emitted over 24 h by either a solution of137Cs in a polystyrene tube (20 or 200MBq in HCl 0.1M) or as asolid 137Cs line source (1.85GBq). Nominal dose rates of 1, 10,100, and 750mGy/d were verified by thermoluminescent dos-imeter measurements and cells and eggs were irradiated accord-ing to the procedure of Gilbin et al. [32].

Egg production

For chronic exposure, zebrafish eggs were obtained fromadult fish bred in the laboratory under standardized conditions.The genitors were purchased from Aquasylva and used asgenitors from 4 to 12 months old. They were kept in a 40-Ltank (60 fish/tank) with a 12:12-h light:dark photoperiod at atemperature of 25� 18C. Water was manually renewed bychanging 50% of the total volume each week and by refillingthe evaporated water each day. Water composition wascharacteristic of a soft water (inmgL�1:Kþ¼ 5.94; Naþ¼7.46; Mg2þ¼ 4.73; Ca2þ¼ 11.6; Cl�¼ 32.6; NO�

3 ¼19.5; SO2�

4 ¼ 9.56; pH 6.5� 0.2). Embryos were obtained fromspawning of two males and one female. Egg viability wasconfirmed when the blastula stage was reached at 3 hpf withoutvisible anomalies; otherwise they were discarded.

For acute exposure, eggs had to be taken directly from a fishbreeder (elevage de la Grande Riviere, France) near the irra-diator location to allow the irradiation to be done before thegastrula stage.

Experimental design for egg exposure

Eggs used in these experiments were aged at 3 hpf. Theywere maintained in the synthetic soft water described above,which was changed daily. They were kept at 26� 18C, with aphotoperiod of 14:10-h light:dark.

For chronic exposure, five eggs were placed in 20-ml glassvials and five replicate vials were used by condition. For acuteexposure, eggs were randomly distributed individually intoeach compartment of a 25-well square Petri dish with tworeplicates per dose. The stages of embryo development wereobserved daily under a binocular microscope by observing thefollowing specific endpoints: blastula (3 hpf), gastrula (4 hpf),segmentation (10–24 hpf) and pharyngeal (24–48 hpf). At48 hpf the embryos started to hatch and the number of hatchedeggs were recorded until 120 hpf (Organisation for EconomicCo-operation and Development, Test Method 212). The hatch-ing rate was determined as a percentage of hatched eggs fromlive eggs at a given time. The median hatching time (HT50),which represents the time necessary for 50% of the eggs tohatch, was then calculated.

Eggs and larvae at 120 hpf were euthanized with 0.05% (w/v)MS-222 (Sigma-Aldrich). Photographs for anterior–posteriorlength measurement were acquired according the procedure ofFraysse et al. [33] using a binocular microscope with a connectedcamera (DS Camera Control Unit DS-U1, Nikon).

Alkaline comet assay

Cell isolation. The alkaline comet assay, detectingDNA strand breaks (single- and double-strand breaks, and

2832 Environ. Toxicol. Chem. 30, 2011 S. Pereira et al.

alkali-labile sites) was applied on ZF4 cells; ZF4 cells weretrypsined, rinsed in phosphate-buffered saline solution (PBS),and diluted in 5ml L15-HEPES.

Comet assay. Cells were counted on a Mallassez cell andtheir viability was assessed using Trypan blue. The wholeisolation and comet assay procedures were validated using aknown genotoxicant, hydrogen peroxide (see SupplementalData).

The alkaline comet assay was applied according to theprocedure of Devaux et al. [34]. One hundred nucleoids perslide were analyzed at�400 magnification under a fluorescencemicroscope (Nikon Eclipse E600) equipped with a 515 to560 nm excitation filter. Comet figures were analyzed by usingComet IV software (Perceptive Instruments).

Statistical analyses

All statistical analyses were implemented with the statisticalcomputing software R 2.10 (R Development Core Team, 2006).We performed the analyses with add-on packages, MASS(Modern Applied Statistics with S), and multcomp (MultipleComparisons Using R).

The occurrence of g-H2AX (pH2AX) and pDNA-PK foci ofeach condition were compared to each other with analysis ofvariance (ANOVA) using Tukey’s all-pair comparisons test.Error normality and variance homogeneity assumption wastested with the statistics of Shapiro-Wilk and Levene, respec-tively. When residual normality and/or variance homogeneityassumptions were not satisfied, square root or logarithmictransformations were applied. The same procedure was per-formed on median embryo lengths and on comet analysis.

Hatching rates were compared to control using pairwise chi-square tests for each measurement time. To keep a global alpharisk at a 5% level, the p value of the pairwise chi-square testswere adjusted according to the Holm method. Median hatchingtimes were obtained using the Hill equation, which is charac-terized by two parameters, the Hill number (nH) and the 50%effect concentration (EC50) with the probability function writ-ten as:

f ðxÞ ¼ xnH

xnH þ ECnH50

The Excel macro REGTOX, based on the Marquardt algo-rithm, was used to calculate ECs ([35]; http://www.normalesup.org/�vindimian/DOC_en_web/doc_en_2.htm).

The confidence intervals of the parameters were estimatedwith a nonparametric bootstrap simulation.

RESULTS

Chronic irradiation leads to DNA damage

DNA damage and repair were analyzed on ZF4 cells after24 h of exposure to gamma rays at 1, 10, 100, and 750mGy/d(Fig. 1).

As a first step, we examined the occurrence of g-H2AX focion gamma-irradiated ZF4 cells. This exposure led to theproduction of significantly higher numbers of g-H2AX focifrom the dose rate of 100mGy/d (ANOVA: F3,12¼ 404.2,p< 0.001, Tukey test: p< 0.001 for all different letters) com-pared with the background values in control cells (such as24.2� 0.6 vs. 1� 1 g-H2AX foci per cell, respectively, for750mGy/d and the control) (Fig. 1A). These results suggest theproduction of a significant number of persistent DSB after 24 hof chronic irradiation, suggesting that the induced DSBs were

not easily reparable. To investigate whether the chronic irra-diation decreases the DSB-repair rate we assessed under thesame conditions the presence of pDNA-PK foci. These foci arelower in number than g-H2AX foci for doses >100mGy/d(3.5� 0.5 pDNA-PK foci vs. 8.5� 1.5 g-H2AX foci per cell at100mGy/d) (ANOVA: F3,12¼ 41.87, p< 0.001, Tukey test:p< 0.001 for all different letters) (Fig. 1A). A known conse-quence of the propagation of persistent DSBs throughout thecell cycle is the formation of micronuclei, which are responsiblefor mitotic death [36]. At 750mGy/d, no DNA-PK fociwere observed that show a defect in DNA repair mechanisms.At the same time, an important increase of micronuclei numberwas also observed (Fig. 1A). DNA single-strand breaks werealso observed in ZF4 cells using the comet assay at 750mGy/d,which is consistent with the impairment of DNA repairmechanisms (Fig. 1B).

Chronic irradiation leads to developmental effects

In this test the embryo mortality never exceeded 20%. Themain effect observed between the different conditions (1, 10,100, and 1,000mGy/d) was an acceleration of the hatchingprocess in eggs exposed to gamma irradiation (Fig. 2A). TheHT50 was calculated for each condition. Exposure to externalgamma irradiation causes a decrease of HT50, significant at 10and 1,000mGy/d (based on nonoverlapping confidence inter-vals) by 12 to 17%, respectively (Fig. 2B). Nevertheless, despite

Mic

ronu

clei

num

ber

(/10

00 c

ells

)

Mea

n ta

il le

ngth

010203040506070

0 1 10 100 750

B

A

Dose rate (mGy/d)

Dose rate (mGy/d)

a a

b

c

a a ba

a a

a

a

b

0

5

10

15

20

25

30

35

control 10 100 7500

10

20

30

40

50

60Number of H2AX foci per cellNumber of DNA-PK foci per cellMicronuclei number (/1000 cells)

a

b

b

c

Num

ber

of H

2AX

or

DN

A-P

K f

oci p

er c

ell

Fig. 1. DNA damage and repair in chronic 24 h gamma-irradiated ZF4 cellsand embryos. (A)g-H2AXandDNA-PKfoci andmicronuclei number inZF4cells (different letters¼ < 0.001). (B) DNAdamage in ZF4 cells assessed bythe comet assay (different letters¼ < 0.05).

Genotoxicity of gamma-irradiation on zebrafish embryos Environ. Toxicol. Chem. 30, 2011 2833

an accelerated hatching for higher doses, no morphologicalabnormalities and/or increased mortality were observed ongamma-irradiated embryos.

Dose-dependent accumulation of DNA damage

DNA damage and repair kinetics were analyzed on zebrafishcells after 0.3, 1, and 2Gy gamma irradiation exposure (Fig. 3).It has been shown that in human control cells acute irradiation of1Gy induced 40� 4 g-H2AX foci per cell per Gy 10min afterirradiation and 4� 2 residual g-H2AX foci 24 h postirradiation(see Supplemental Data) [31]. For ZF4 cells irradiated at 1Gy(Fig. 3B), we observed 24 h postirradiation a significant amountof delayed unrepaired DSB (12� 0.5 g-H2AX foci per cell perGy; p< 0.05) and an increase of micronuclei number in cells.The same results were observed for 2Gy irradiation (data notshown). For lower doses, (0.3Gy) (Fig. 3A), the number ofg-H2AX foci scored 10min after irradiation was found to behigher (21� 2 g-H2AX foci per cell per Gy; p< 0.05) than thatassessed in the human controls at the same dose (N. Foray,INSERM U1052, Lyon, France, personal communication)and very few residual g-H2AX foci (as in control cells) were

scored 4 h after irradiation (Fig. 3A). For the two doses, anincrease of micronuclei number was observed (Fig. 3A,B).

All these results, an accelerated DNA repair for lower dosesand accumulation of DSB for higher doses (from 0.3Gy),suggest a high radiosensitivity of embryonic zebrafish cellsas observed in radiosensitive human cells [31].

Acute irradiation leads to abnormal embryo development

Embryos were irradiated at 3 hpf with 0.3, 1, or 2Gy. Asignificant decrease in hatching success of exposed eggs wasobserved. For controls, hatching began at 48 hpf and endedat 80 hpf, whereas 2Gy-exposed eggs began hatching later(at 72 hpf) and also finished also later (at 120 hpf). Furthermore,the values of HT50 calculated for each condition stronglysupports an increase of hatching time and a decrease of hatchingsuccess (Fig. 4A). Together, the body length of zebrafishprolarvae decreased with increasing dose (Fig. 4B, ANOVA:F3,33¼ 12.16, p< 0.001, Tukey test: p< 0.05 for all differentletters). At 120 hpf the decrease of body length came with anincrease of abnormalities (Fig. 4C) and 28.6% of 2Gy-exposedlarvae presented morphologic abnormalities (see SupplementalData). The observed abnormalities were tail atrophia and trunkaxis malformations.

DISCUSSION

In the present study we observed that DNA damage inducedby chronic irradiation may have less of an impact on embryodevelopment than DNA damage induced by acute radiation.The results obtained on acute irradiation suggest a dose-dependent correlation between DNA damage accumulation(unrepaired DSBs) and abnormalities in embryo development.

Chronic irradiation clearly leads to NHEJ impairment(Fig. 1A), produced by a decrease of DNA-PK foci numberfor dose rates>0.1Gy/d and an increase of micronuclei number(Fig. 1B). After 24h, we observed more micronuclei in chronicirradiated cells than acute irradiated cells for the same doses(56 vs 30 for 1,000 nuclei, respectively) (Figs. 1A, 3B).

However, even if DSBs and micronuclei were observedin chronic radiated ZF4 cells, no embryo abnormalities wereobserved for any dose rates. This could be explained by theactivation of another DNA repair mechanism such as homo-logous recombination, which could supply a dysfunction in theNHEJ repair mechanism as observed in different types ofirradiated human cells [36]. Moreover, because the productionof micronuclei is the stigma of the main mitotic cell death, wecan hypothesize that in chronic irradiated embryos the balancebetween the elimination of abnormal cells with micronucleiand cell renewal embryonic cells would be sufficient for viableembryonic development. Thus, these DSBs may not havedetrimental effects on embryo development, because the sus-ceptibility of the embryo to lethal effects of radiation appears todiminish after 5 to 7 hpf [28]. Nevertheless, we cannot excludethe presence in embryos of other DNA damage (e.g., muta-tions), which potentially could have an impact on the repro-duction success or be transmitted to other generations. This wasobserved in medaka for chronic exposure [26], where DNAmutations were observed in exposed medaka and mutationswere transmitted to offspring from exposed parents.

Acute radiation from 0.3Gy leads to abnormalities.Unrepaired DSBs may have subsisted in different organ cells(inducing micronuclei as observed in ZF4 cells, Fig. 3) andcould be linked to the observed abnormalities. At an irradiationdose of 1Gy, we observed a plateau of repair mechanisms from

0

20

40

60

80

100

24 48 66 72 96Time (hpf)

01 mGy/d

10 mGy/d

100 mGy/d

1000 mGy/d

Hat

chin

g (

%)

40

45

50

55

60

65

70

75

0 1 10 100 1000

dose rate (mGy/d)

HT

50

B

A

aa

a

b

b

Fig. 2. Exposure to external chronic radiation causes an acceleration ofhatching. hpf¼ hours postfertilization. (A)Hatching (%)of eggs exposed to arangeof external gamma irradiation (mean�SE,n¼ 5). (B)Effect of gammairradiation on the external median hatching time (HT50) (differentletters¼ < 0.05).

2834 Environ. Toxicol. Chem. 30, 2011 S. Pereira et al.

24 h postirradiation (data not shown), which could suggest adefect in DNA repair mechanisms associated with a highnumber of residual DSBs in cells (Fig. 3B).

At the cellular level, we observed different DNA damageand repair after chronic or acute gamma irradiation exposure.First, at the same exposure time, the quantity of energy depos-ited to induce breaks in DNA is less important for chronic thanacute irradiation, which is shown by the presence of a highnumber of DSB immediately after acute exposure (Fig. 3)compared to what could have been observed for chronic irra-diation. Second, after 24 h the number of g-H2AX foci obtainedversus total dose is higher for acute versus chronic exposure:acute 1Gy-irradiated ZF4 cells displayed 12 foci versus 24.2foci for 0.75Gy-chronic irradiated cells. These last cells alsodisplayed a higher number of micronuclei. These results

showed that chronic irradiation inhibits DNA repair mecha-nisms (decrease in DNA-PK foci number) with a higher toxicityof DSB that may induce cell death as shown by the observedformation of micronuclei.

On the contrary, we observed that the DSB repair process isquasi-completed 4 h postirradiation for acute dose rates<0.3Gy (Fig. 3A). From acute doses >0.3Gy, we observedunrepaired DSB at 4 h postirradiation that may accumulate dueto NHEJ repair pathway impairment (12� 0.5g-H2AX foci at24 h postirradiation). These unrepaired DSBs may be moredeleterious, as they may bemisrepaired, leading to the abnormaldevelopment observed.

Zebrafish embryo cells seem to be more radiosensitive thanother species (see Supplemental Data): at 1Gy, ZF4 cells have ahigher number of g-H2AX foci than human or mice cells

Fig. 3. Double-strand break repair is impaired at doses up to 0.3Gy in g-irradiatedZF4 cells. (A,C) Kinetics ofg-H2AX foci disappearance on 0.3Gy and 1Gyg-irradiated ZF4 cells, respectively. Each data plot represents the mean�SE (n¼ 6) of at least six independent experiments. Micronuclei number was assessed on1,000 cells stained with DAPI. (B,D) Representative localization of g-H2AX on 0.3 and 1Gy gamma-irradiated ZF4 cells, respectively, at each repair time.Conventional microscopy, DAPI staining. Scale bar¼ 5 mm.

Genotoxicity of gamma-irradiation on zebrafish embryos Environ. Toxicol. Chem. 30, 2011 2835

[37,38]. Furthermore, it has been shown that some radiosensi-tive cells elicit accelerated DSB repair processes 2 to 8 hpostirradiation [36], which is consistent with DSB repairkinetics observed for 0.3 DSB that were repaired at 4 h in0.3Gy irradiated ZF4 cell (Fig. 3).

CONCLUSIONS

The results of the present study provide the first cluesof zebrafish radiosensitivity and demonstrate the detectionof DNA DSBs in zebrafish as an explanation of the causal

relationship between irradiation-induced DNA damage anddevelopmental abnormalities.

Further studies will be necessary to investigate the use oftools such as DNA repair proteins, micronuclei, and g-H2AXdetection as predictive biomarkers. This could have importantimplications for environmental biomonitoring in relation toionizing radiation exposure.

SUPPLEMENTAL DATA

Fig. S1. (400KB PDF).

Fig. 4. Exposure to external acute radiation causes a delay in hatching, an impairment of hatching success, and larvae abnormalities. hpf¼ hours postfertilization.(A) Effect of g-irradiation on the median hatching time (HT50) (different letters¼ < 0.05). (B) The body length of zebrafish prolarvae decreased with increasingdose irradiation (different letters¼ < 0.01). (C) Abnormalities observed on embryos exposed from 0.3Gy to 2Gy. Conventional microscopy. Scale bar¼ 1mm.

2836 Environ. Toxicol. Chem. 30, 2011 S. Pereira et al.

Acknowledgement—The work reported in the present study receivedfinancial support from Electricite De France Grants and ENVIRHOM IRSNprogram. S. Pereira is a postdoctoral fellow funded by the HEMI-BREAKSproject (Agence Nationale de la Recherche) and by the Institute forRadioprotection andNuclear Safety (ENVIRHOM research program). Someof the irradiation experiments were performed at the INSERMU836, Lyon,France (special thanks to Nicolas Foray and Adeline Granzotto) and at theInstitut de Radioprotection et de Surete Nucleaire, DRPH, SRBE, LRTOXFontenay-aux-Roses, France (special thanks to Jean-Marc Bertho, AgnesFrancois and Isabelle Dublineau).

REFERENCES

1. Smith GC, Jackson SP. 1999. The DNA-dependent protein kinase.GeneDev 13:916–934.

2. Dudas A, Chovanec M. 2004. DNA double-strand break repair byhomologous recombination. Mutat Res 566:131–167.

3. KobayashiH,SimmonsLA,YuanDS,BroughtonWJ,WalkerGC.2008.Multiple Ku orthologues mediate DNA non-homologous end-joining inthe free-living form and during chronic infection of Sinorhizobiummeliloti. Mol Microbiol 67:350–363.

4. CollisSJ,DeWeeseTL, JeggoPA,ParkerAR.2005.The life anddeathofDNA-PK. Oncogene 24:949–961.

5. Bladen CL, Lam WK, Dynan WS, Kozlowski DJ. 2005. DNA damageresponse andKu80 function in the vertebrate embryo.Nucleic Acids Res33:3002–3010.

6. Bladen CL, Navarre S, Dynan WS, Kozlowski DJ. 2007. Expression ofthe Ku70 subunit (XRCC6) and protection from low dose ionizingradiation during zebrafish embryogenesis. Neurosci Lett 422:97–102.

7. Rothkamm K, Lobrich M. 2003. Evidence for a lack of DNA double-strand break repair in human cells exposed to very low x-ray doses.ProcNatl Acad Sci U S A 100:5057–5062.

8. Chan DW, Chen BP, Prithivirajsingh S, Kurimasa A, Story MD, Qin J,Chen DJ. 2002. Autophosphorylation of the DNA-dependent proteinkinase catalytic subunit is required for rejoining of DNA double-strandbreaks. Gene Dev 16:2333–2338.

9. Foray N, Arlett CF, Malaise EP. 1999. Underestimation of the smallresidual damage when measuring DNA double-strand breaks (DSB): Isthe repair of radiation-induced DSB complete? Int J Radiat Biol75:1589–1595.

10. Kosmehl T, Hallare AV, Braunbeck T, Hollert H. 2008. DNA damageinduced by genotoxicants in zebrafish (Danio rerio) embryos aftercontact exposure to freeze-dried sediment and sediment extracts fromLaguna Lake (The Philippines) as measured by the comet assay.MutatRes 650:1–14.

11. Sandrini JZ, Bianchini A, Trindade GS, Nery LE, Marins LF. 2009.Reactive oxygen species generation and expression of DNA repair-related genes after copper exposure in zebrafish (Danio rerio)ZFL cells.Aquat Toxicol 95:285–291.

12. Cambier S, Gonzalez P, Durrieu G, Bourdineau JP. 2010. Cadmium-induced genotoxicity in zebrafish at environmentally relevant doses.Ecotoxicol Environ Saf 73:312–319.

13. Adam C. 2007. Genotoxicity of radionuclides in aquatic organisms.IRSN 2007-81. Institut de Radioprotection et de Surete Nucleaire,Cadarache, France.

14. McAleerMF,DavidsonC,DavidsonWR,YentzerB,FarberSA,RodeckU, Dicker AP. 2005. Novel use of zebrafish as a vertebrate model toscreen radiation protectors and sensitizers. Int J Radiat Oncol Biol Phys61:10–13.

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

16. Knowles JF. 1999. Long-term irradiation of a marine fish, the plaicePleuronectes platessa: An assessment of the effectson size andcomposition of the testes and of possible genotoxic changes in peripheralerythrocytes. Int J Radiat Biol 75:773–782.

17. Geiger GA, Parker SE, Beothy AP, Tucker JA, Mullins MC, Kao GD.2006. Zebrafish as a ‘‘biosensor’’? Effects of ionizing radiation andamifostine on embryonic viability and development. Cancer Res66:8172–8181.

18. Yasuda T, Aoki K, Matsumoto A, Maruyama K, Hyodo-Taguchi Y,Fushiki S, Ishikawa Y. 2006. Radiation-induced brain cell death can beobserved in living medaka embryos. J Radiat Res (Tokyo) 47:295–303.

19. Tsyusko O, Glenn T, Yi Y, Joice G, Jones K, Aizawa K, Coughlin D,Zimbrick J, Hinton T. 2011. Differential genetic responses to ionizingirradiation in individual families of Japanese medaka, Oryzias latipes.Mutat Res 718:18–23.

20. Kuhne WW, Gersey BB, Wilkins R, Wu H, Wender SA, George V,Dynan WS. 2009. Biological effects of high-energy neutrons measuredin vivo using a vertebrate model. Radiat Res 172:473–480.

21. Hyodo-Taguchi Y, Etoh H. 1993. Vertebral malformations in medaka(teleost fish) after exposure to tritiated water in the embryonic stage.Radiat Res 135:400–404.

22. Ryan LA, Seymour CB, O’Neill-MehlenbacherA,Mothersill CE. 2008.Radiation-induced adaptive response in fish cell lines. J EnvironRadioact 99:739–747.

23. Hidaka M, Oda S, Kuwahara Y, Fukumoto M, Mitani H. 2010.Cell lines derived from a medaka radiation-sensitive mutant havedefects in DNA double-strand break responses. J Radiat Res (Tokyo)51:165–171.

24. Adam C, Larno V, Giraudo M, Barillet S, Gania Y, Devaux A. 2007.Genotoxicite des radionucleides chez les organismes aquatiques: Etat del’art et resultats preliminaires. In ARET-Actualites. Numero special,Effets des toxiqueschimiques / radioactifs sur levivant.Associationpourla Recherche en Toxicologie, Paris, France.

25. Jarvis RB, Knowles JF. 2003. DNA damage in zebrafish larvae inducedby exposure to low dose rate [gamma]-radiation: Detection by thealkaline comet assay. Mutat Res-Gen Tox En 541:63–69.

26. Tsyusko O, Yi Y, Coughlin D, Main D, Podolsky R, Hinton TG, GlennTC. 2007.Radiation-induced untargeted germlinemutations in Japanesemedaka. Comp Biochem Phys C 145:103–110.

27. ShimadaA,ShimaA. 2001.High incidenceofmosaicmutations inducedby irradiating paternal germ cells of the medaka fish, Oryzias latipes.Mutat Res 495:33–42.

28. WalkerC, StreisingerG. 1983. Induction ofmutations by gamma-rays inpregonial germ cells of zebrafish embryos. Genetics 103:125–136.

29. WileyLM,RaabeOG,KhanR,StraumeT.1994.Radiosensitive target inthe earlymouse embryo exposed to very low doses of ionizing radiation.Mutat Res 309:83–92.

30. Foray N, Marot D, Gabriel A, Randrianarison V, Carr AM, PerricaudetM, Ashworth A, Jeggo P. 2003. A subset of ATM- and ATR-dependentphosphorylation events requires the BRCA1 protein.EMBO J 22:2860–2871.

31. Joubert A, Zimmerman KM, Bencokova Z, Gastaldo J, Renier W,ChavaudraN, FavaudonV,Arlett C, ForayN. 2008.DNAdouble-strandbreak repair defects in syndromes associated with acute radiationresponse: At least two different assays to predict intrinsic radio-sensitivity? Int J Radiat Biol 84:107–125.

32. Gilbin R, Alonzo F, Garnier-Laplace J. 2008. Effects of chronic externalgamma irradiation on growth and reproductive success of Daphniamagna. J Environ Radioactiv 99:134–145.

33. 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.

34. DevauxA,PesonenM,MonodG. 1997.Alkaline comet assay in rainbowtrout hepatocytes. Toxicol In Vitro 11:71–79.

35. Vindimian E. 2003. REGTOX: Macro ExcelTM for dose-responsemodelling, France.

36. Joubert A, Foray N. 2006. Repair of radiation-induced DNA double-strand breaks in human cells: History, progress and controversies. InLandseer BR, ed, New Research on DNA Repair. Nova Science,Hauppauge, NY, USA, pp 273–294.

37. Joubert A,BistonMC,BoudouC, Ravanat JL, BrochardT, CharvetAM,Esteve F, Balosso J, Foray N. 2005. Irradiation in presence of iodinatedcontrast agent results in radiosensitization of endothelial cells:Consequences for computed tomography therapy. Int J Radiat Oncol62:1486–1496.

38. Foray N, Charvet AM, Duchemin D, Favaudon V, Lavalette D.2005. The repair rate of radiation-induced DNA damage: A stochasticinterpretation based on the gamma function. J Theor Biol 236:448–458.

Genotoxicity of gamma-irradiation on zebrafish embryos Environ. Toxicol. Chem. 30, 2011 2837