Aquatic Toxicology 77(2006)136–142Exposure of spermatozoa to duroquinone may impair reproduction of the common carp (Cyprinus carpio )through oxidative stressBingsheng Zhou a ,b ,Wenhua Liu a ,WilliamaDesmond O’Toolea ,Paul m a ,Rudolf aCenter for Coastal Pollution and Conservation,City University of Hong Kong,Kowloon,Hong Kong SAR,PRChina b Institute of Hydrobiology,Chinese Academy ofSciences,Wuhan 430072,PR ChinaReceived 13October 2005;received in revised form 8November 2005;accepted 25November 2005AbstractToxicity of manywaterborne organic contaminants to aquatic organisms is mediated through oxidative damages resulting from the production of reactive oxygen species (ROS).Using duroquinone as a model ROS inducer,we carried out in vitro and in vivo experiments to test the hypothesis that reproduction in common carp (Cyprinus carpio )can be impaired through oxidative damage of their spermatozoa.In vitro exposure of fish spermatozoa to 0,12.5,25,50,100and 200M duroquinone for 2h showed a significant increase in the level of ROS in a dose-dependant manner.Sperm motility was significantly reduced in all exposure groups,but lipid peroxidation (LPO)and DNA strand break (measured by comet assay)were only enhanced at 50M and above.A significant decrease in subsequent hatching rate was recorded in all the exposure groups,despite fertilization rate was not affected.In the in vivo experiment,spermatozoa were collected 24and 72h after fish received intra-peritoneal injections of 1.0and 10mg kg −1body weight duroquinone.DNA damage was clearly evident in spermatozoa of all treatment groups after 72h exposure,and ROS was significantly enhanced in the high concentration group.LPO however,remained unchanged in both treatment groups.The overall results of both our in vitro and in vivo experiments demonstrated that duroquinone can induce ROS production in spermatozoa,which may impair sperm quality and subsequently reproductive success through oxidative stress.©2006Elsevier B.V .All rights reserved.A variety of pollutants,such as polynuclear aromatic hydro-carbons (PAHs),polychlorobiphenyls (PCBs),quinones,aro-matic nitro compounds,aromatic hydroxylamines,bipyridyls and transition metals are known to enhance production of reac-tive oxygen species (ROS)(Livingstone,2001).ROS can attack a variety of biological macromolecules such as DNA,pro-teins and lipids,causing oxidative damages (Kelly et al.,1998).In mammals,oxidative damage has been implicated in patho-physiological symptoms including cell death,tissue injury,car-cinogenesis and aging (Kehrer,1993;Kelly et al.,1998).Fur-thermore,ROS have been extensively demonstrated to play an essential role in the pathogenesis of many reproductive processes in human beings (review by Agarwal et al.,2003).Because the∗Corresponding author.Tel.:+852********;fax:+852********.E-mail address:bhrswu@.hk (R.S.S.Wu).they are very susceptible to oxidative damage when attacked by ROS,leading to lipid oxidation and therefore changes in mem-brane fluidity and the integrity of DNA in the spermatozoan nucleus.Consequently,increased levels of ROS are believed to be involved in male infertility due to lipid peroxidation,affects sperm mobility,and oxidative stress-induced DNA damage that may accelerate the process of germ cell apoptosis leading to a decline in sperm counts (Ong et al.,2002;Agarwal et al.,2003;Baker and Aitken,2004).Production of ROS in fish in response to insult of nants has also been shown to cause oxidative damage to DNA,proteins,and other macromolecules,and has been posed as one of the underlying mechanisms leading to carcinogenesis (Di Giulio et al.,1989;Lemaire and 1997;Livingstone et al.,2000;Livingstone,2001;Oakes et 2004).In fish,metabolism of xenobiotics (e.g.,PAH)compounds such as quinones that can generate ROS via 0166-445X/$–see front matter ©2006Elsevier B.V .All rights reserved.doi:10.1016/j.aquatox.2005.11.006B.Zhou et al./Aquatic Toxicology 77(2006)136–142cycling (Peters et al.,1996).Recent research has suggested that oxidative damage might be associated with reproductiveimpairment in fish.Forexample,studies have demonstrated that oxidative stress,as quantified by an increase in 2-thiobarbituric acid reactive substances (TBARS)and lipid peroxides (LPO),is associated with reproductive impairment of the white sucker (Catostomus commersoni )exposed to pulp and paper mill efflu-ent,although the relationship between gonad size,gonadal oxidative stress,and plasma steroids remained unclear (Oakes et al.,2003;Oakes and Van der Kraak,2003).DNA dam-age,as manifested by DNA strand break,has been correlated with the mutagenicity and carcinogenicity of environmental pol-lutants (Kurelec,1993),and associated with reduced growth,abnormal development and reduced survival of embryos,lar-vae and adults of bivalves and urchins (Steinert,1999;Lesser and Barry,2003).Indeed,DNA strand breaks has been com-monly employed as a biomarker for measuring genotoxic-ity,mutagenicity and carcinogenicity of a range of chemi-cals in mammalian cells (McKelvey-Martin et al.,1993)as well as in aquatic animals (Mitchelmore and Chipman,1998;Kamer and Rinkevich,2002;Lee and Steinert,2003;Siu et al.,2004).Gametes of many aquatic animals are discharged into water and therefore directly exposed to pollutants prior to fertilization.In addition,exposure of parental adults to toxicants in water may affect also gamete quality (Rurangwa et al.,2004),which may subsequently affect fertilisation success,hatching success and larval survival (Kime and Nash,1999).Indeed,sperm quality is closely related to fertilization success,hatching of embryos and subsequent survival of offspring (Au et al.,2002,2003),and sperm motility is often employed as a key criterion to assess or predict sperm quality in fish and sea urchins (Kime and Nash,1999;Au et al.,2002).The computer-assisted sperm analysis system (CASA),which was first developed for human applications,has been successfully employed to study environ-mental stress in fish sperm (Kime and Nash,1999).Most studies on effects of toxicants on fish reproduction have focused ondisruption of reproductive hormones and vitellogenin,gonadal-somatic index,gonad histopathology,and fecundity,fertiliza-tion,hatching success,and larvae survival (Kime and Nash,1999).To our knowledge,effect of toxicants on fish reproduc-tion through ROS induction and sperm quality has not been documented thus far.In the light of the potential link between oxidative stress and reproduction in feral fish (Oakes and Van der Kraak,2003;Oakes et al.,2003),the impact of oxidativestress on fish reproduction warrants further study.Duroquinone is able to induce ROS production through the NADPH-dependent redox cycling in hepatic microsomes of the flounder Platichthys flesus (Lemaire and Livingston,1997).Using duroquinone as a model ROS inducer,we carried out in vitro and in vivo experiments to test the hypothesis that reproduction in the common carp (Cyprinus carpio )can be impaired through oxidative damage of their spermatozoa.The dose-dependent responses of DNA strand breaks and LPO to ROS production were also investigated,and the observed oxida-tive damages were further success and hatching rate. 2.Materials and methods 2.1.Chemicals All chemicals used were of analytical grade and purchased from Sigma (St.Louis,MO,USA).Duroquinone and dichlo-rofluorescin diacetate (DCF-DA)were dissolved in distilled and stock The common carp (C.carpio )which has a global distri-bution and been used extensively in toxicity assessments was employed as a model species in the present study.One-year-oldcarp (80–100g)were provided by the experimental farm of the Department of Agriculture,Fisheries,and Conservation,HongKong SAR.Prior to experimentation,fish were acclimated infully aerated,dechlorinated freshwater for at least 2weeks in300-l fiberglass tanks (20.5±0.5◦C under a 12:12light–dark cycle).The water was recirculated through filters with activated charcoal and sand.Experimental fish were fed daily with a com-mercial food (J.W.Vitra,35%protein)at 2%of body weight.For the in vivo experiment,the male carps were randomly sep-arated into two treatment groups and a control group.The treat-ment groups received a single intraperitoneally (ip)injection of either 1mg kg −1(n =13)or 10mg kg −1(n =11)of duroquinone dissolved in corn oil containing 4%of acetonitrile.The control group (n =12)received only corn oil with 4%acetonitrile.At 6and 54h after injection,fish were induced to spawn by another ip injections of 2mg kg −1of common carp pituitary extract (courtesy of the Institute of Hydrobiology,Chinese Academy of Sciences,prepared by homogenizing the carp pituitary gland in 0.9%NaCl followed by centrifugation at 3000×g for 10min).Eighteen hours after the pituitary injection,the abdomen of fish was gently squeezed and semen collected and stored immedi-ately on ice for later analyses.For the in vitro experiment,the semen obtained as described above from four untreated males was pooled.The pooled sample was then diluted with an immobilization medium (200mM KCl,30mM Tris,pH 8.0,osmolality >400mosmol kg −1)(Perchec et al.,1995)to obtain a sperm density of 3×107cells ml −1.Thesperm sub-samples (n =4)were then exposed for 2h to final con-centrations of 0,12.5,25,50,100and 200M of duroquinone dissolved in acetonitrile.2.3.ROS measurementROS in sperm was described in Pottset al.(2000).The semen samples were diluted with the immobi-lization medium to obtain a sperm density of 3×107cells ml −1.Dichlorofluorescin (DCF)was added to the diluted semen sam-ples.The samples were then incubated at 4◦C for 0.5h and centrifuged at 3000×g at 4◦C for 5min.The pellets were rinsed with the immobilization buffer and centrifuged again.ROS was measured using a fluorescence spectrophotometer (excitation:485nm;emission:538nm).The result was expressed as the intensity of fluorescence 硫代巴比妥酸白色吸盘鱼 ，纸浆和造纸厂污水138 B.Zhou et al./Aquatic Toxicology 77(2006)136–1422.4.Sperm motilityThe carp sperm is typically non-motile in the semen since the semen has high sperm density (4×109to 30×109spermatozoa ml −1)and high viscosity.In fact,teleost sperm becomes motile only after being discharged and diluted in water (Perchec et al.,1995).After the sperm was in vitro exposed to the toxicant in the immobilization medium,sperm motility was assessed following a two-step dilution.This was immediately followed by a second five-fold dilution in an acti-vation medium with low osmolarity (5mM KCl,45mM NaCl and 30mM Tris,pH 8.0,<160mosmol kg −1)(Perchec et al.,1995).Sperm motility was analysed using a CRISMAS system (Image House,Copenhagen)following the procedures described by Wu et al.(2003).Briefly,the swimming behaviour of sperm was recorded for 2s using a JVCCCD color camera mounted on a microscope at 200×magnification.Video pictures taken at 32frames s −1were digitized and stored in a computer.By fol-lowing the movement of an individual sperm during 2s,the mean curviline velocity (VCL),mean straight-line velocity (VSL)and angular path velocity (V AP)were computed.VCL was calcu-lated by summing the distances of the linear path between the position of a sperm in all successive frames and then dividing by the interval during which the sperm was tracked.VSL and V AP were calculated from the straight-line path and curvilinear path (m s −1)between the first and the last position of a sperm,et assayComet assay wasstrand breaks in the in vitro exposurethe method described in Siu et al.(2004).conductedunder dim yellow lightto fromUV radi-ation (Singh et al.,1988).were diluted with the immobilization density of 1×106cells ml −1.Five sample was combined with 500l of 1%low melt agarose in calcium and magnesium free phosphate buffered saline at 42◦C.Fifty micro-liters of the mixture was immediately added to each of the two sample areas of a Comet Slide TM purchased from Trivigen,Inc.(Gaithersburg,MD,USA)and allowed to solidify at 4◦C in the dark for 10min.The slides were immersed in fresh lysis buffer (2.5M NaCl,100mM EDTA,10mM Tris,1%Triton X-100,1%DMSO,pH 10)for at least 1h at 4◦C in a Coplin jar.After that,the slides were drained well,aligned in two rows in a horizontal electrophoresis tray and immersed in an alka-line electrophoresis buffer (300mM NaOH and 1mM EDTA,pH >13)for 45min to denature the DNA.Electrophoresis was carried out in the same buffer at 1V cm −1and 300mA for 30min.The slides were drained well,dehydrated by dipping into absolute ethanol for 5min and air-dried for storage.Prior to the analysis,50l of 1%SYBR ®Green I staining solution (Molecular Probe,Eugene,OR,USA)was added to each agarose spot of the slide and analysed with an Axiovert 100M confocal microscope (Zeiss,Germany)at 200×magnification.A total of 100cells (50cells from each spot)were scored for each sampleand the captured images were analysed using VisComet Ver.1.5image analysis software (Impuls,Germany).Tail length (mea-sured from the middle of the head to the end of the tail)and tail DNA content (tail %DNA)were measured.Olive Tail Moment (tail length ×tail DNA content)was calculated using the for-mula below:M Tail Olive =(|CG Tail −CG Head |)×%DNA Tailwhere M Tail Olive is the Olive Tail Moment,CG Tail the centre of gravity of the tail,CG Head the centre of gravity of the head and %DNA Tail is the percent of migrated DNA in the tail compared to the 2004)was determined.2.6.(LPO)Thiobarbituric acid reactive substances (TBARS)were mea-sured as the product of lipid peroxidation (Potts et al.,2000).The semen sample was diluted with the immobilization medium to obtain a sperm density of 107to 108cells ml −1.The diluted sample was centrifuged at 3000×g at 4◦C for 5min and resus-pended in the immobilization buffer,and 1%(v/v)thiobarbituric acid and 2.4M ice-cold trichloroacetic acid were added to pre-cipitate proteins.The mixture was then incubated at 100◦C in a water bath for 0.5h,cooled and centrifuged at 3000×g for 10min to remove the precipitate.The absorbance of the super-natant was measured at 535nm and the value was compared with that of the TBARS standard curve.The content of TBARS was expressed as nM per 108cells.2.7.Fertilization rate,hatching rate and larval survivorshipEggs from four females without duroquinone treatment were obtained,using the same artificial spawning procedures described for semen collection.Groups of 60–80eggs were ran-domly segregated from the pooled eggs and transferred to Petri dishes.One milliliter of the semen sample after in vitro expo-sure to duroquinone (described in Section 2.2),was added to the dish to achieve a sperm to egg ratio of 105:1(spermato-zoa:egg).Ten milliliters of water was then added to the Petri dish to mobilize the sperm to initiate fertilization (which was expected to be completed by about 10min).After 2h,fertil-ization success was determined by identifying the numbers of the two-cell stage embryos and unfertilized eggs under a stereo microscope.Embryos were kept in Petri dishes and incubated atroom temperature (27–28◦C)for another 4days,and water was replaced daily.Fertilization rate and hatching rate were calcu-lated as the respective percentage of fertilized eggs and embryos in the egg samples.2.8.Statistical analysisNormality and the homogeneity of variance of all data were first tested by the Kolmogorov test and the Bartlett test,respectively.Data obtained from the Comet assay were log-transformed.For the in vitro study,one-way analysis of variance (ANOV A)was used to examine the difference among individual groups for all parameters except for ROS production,which wasB.Zhou et al./Aquatic Toxicology 77(2006)136–142139analysed using two-way ANOV A.Where significant differences were revealed by the ANOV A,pair-wise comparisons were made between different concentrations and the control group,using a Dunnett’s test.For the in vivo study,two-way ANOV A was used to examine the effects of concentration and duration of exposure on each parameter.Where significant effects were detected,pair-wise comparisons were made using Tukey test.The P level used in all the statistical analyses was set at 0.05.3.ResultsIn the in vitro experiment,ROS production in exposed sperm exhibited a good dose-dependent relationship to duro-quinone after 2h (Fig.1).ROS was significantly increased upon exposure to 12.5M duroquinone (P <0.05)and above,and ROS production was three-fold higher than the con-trol value when exposed to 200M duroquinone.In the in vivo experiment however,no significant increase in ROS production was found after 24h exposure (relative fluo-rescence per 107sperm in the control,1and 10mg kg −1treatment groups =1182.75±186.20,1321.95±109.36and 1452.06±87.14,respectively)(mean ±S.E.M.,n =4).A sig-nificant increase in ROS was evident at the higher concen-tration (10mg kg −1)treatment (1479.44±61.39fluorescence per 107sperm)after 72h in vivo exposure (P <0.01)when compared with the control (1109.56±57.54fluorescence per 107sperm);whereas no significant difference could be found between the control and the lower concentration treatment (1256.19±100.25fluorescence per 107sperm,1mg kg −1),as well as between the two treatment concentrations.Notably,the measured in vivo ROS was lower than those measured in the in vitro experiment.The level of TBARS is indicative of the extent of LPO.In the in vitro experiment,the measured TBARS was typically around 0.32nmol 108sperm.Levels of TBARS remained unchanged at lower concentrations (i.e.12.5and 25M)ofduroquinone,Fig.1.Dose-dependent induction of reactive oxygen species (ROS)in the sper-matozoa after exposure to various concentration of duroquinone (0,12.5,25,50,100and 200M)for 2h.The measured ROS is expressed as fluorescence inten-sity of 107sperm (mean ±S.E.M.,n =4).Different letters denote significant difference between datasets.Fig.2.Lipid peroxidation (LPO)in the sperm after exposure to duroquinone at concentrations of 12.5,25,50,100and 200M.The TBARS concentration is expressed as nM 10−8cells (mean ±S.E.M.,n =4).Different letters denote significant difference between data sets.while the levels increased significantly when the sperm were exposed to higher concentrations (50,100and 200M)of duro-quinone for 2h (Fig.2).In the in vivo experiment,no significant induction of TBARS could be found in both treatment groups after 24h in vivo exposure (control:0.16±0.01nmol TBARS 108sperm;1mg kg −1:0.20±0.04nmol TBARS 108sperm;10mg kg −1:0.22±0.03nmol TBARS 108sperm;P >0.05)as well as 72h in vivo exposure (control:0.20±0.04nmol TBARS 108sperm;1mg kg −1:0.22±0.03nmol TBARS 108sperm;10mg kg −1:0.26±0.04TBARS 108sperm;P >0.05)(mean ±S.E.M.,n =4).The measured TBARS in sperm after in vivo exposure was ca.50%lower than those in in vitro expo-sures.The level of DNA damage,expressed as the mean Olive Tail Moment,is shown in Fig.3.DNA damage increased signifi-cantly when sperm were exposed to concentration over 25M of duroquinone for 2h,as compared with the control (P <0.001).A good dose–response relationship was found between the level of DNA damage and the concentration of duroquinone (R 2=0.9916;P <0.0001).Fig.3.Oliver Tail Moment of the spermatozoa after exposure to different concentrations of duroquinone for 2h (mean ±S.E.M.,n =3).Different letters denote significant difference between datasets.140 B.Zhou et al./Aquatic Toxicology 77(2006)136–142Fig.4.Sperm motility after in vitro exposure to different concentrations of duroquinone for 2h.The velocity is expressed as m s −1;VCL,mean curvi-linear velocity;VSL,mean straight-line velocity;V AP,angular path velocity (mean ±S.E.M.,n =4).Different letters denote significant difference between data sets.Despite all three parameters of sperm motility,VCL (mean curvilinear velocity),VSL (mean straight-line velocity);and V AP (angular path velocity)decreased significantly in all treat-ment groups as compared with the control (P <0.01)(Fig.4),a dose-dependent reduction in these three sperm motility parame-ters was not observable when sperm was exposed to an increas-ing concentrations of duroquinone.In the control group,in vitro fertilization rate was 98.29±1.05%(mean ±S.E.M.,n =4).Despite the sperm had been exposed to different concentration of duroquinone,almost all eggs were fertilized (fertilization rate ranged from 99.3to 98.8%)and no significant difference in fertilization rates could be found between the control and the various treatment groups (mean ±S.E.M.,n =4).However,a significant decrease in hatching rate was recorded in all the treat-ment groups (P <0.001)(Fig.5),and a good dose-dependent response was also clearly evident.All embryos from the 100and 200M treatment groups failed tohatch.Fig.5.Hatching rate after in vitro exposure to different concentrations of duroquinone for 2h (mean ±S.E.M.,n =4).Different letters denote significant difference between data sets.4.DiscussionTo our knowledge,this is the first in vitro and in vivo study to investigate ROS production in fish sperm exposed to duro-quinone and relating this to oxidative damage and reproductive impairment.Results of the present study suggest that fish repro-duction can be impaired by duroquinone-mediated oxidative damages.Although the exact mechanism of ROS induction in carp spermatozoa is not known,two independent mechanisms have been shown to be responsible for ROS generation in mam-The first system involves an enzymatic system located in the sperm plasma membrane that utilizes NAD(P)H as substrate,and the second system involves the mitochondrial electron transport chain which could redox cycle xenobiotics and produce ROS and oxidative damages (Palmeira and Wallace,1997;Vernet et al.,2004).A recent study has further identified the presence of cytochrome P450-reductase in epididymal sperm preparations (Baker et al.,2004),which may also metabolize xenobiotics and produce ROS.The marked increase in ROS production found in this study indicates that carp sperm can metabolize duroquinone.Thus,it may be possible that duroquinone,as a redox cycler (Lemaire and Livingston,1997),may be metabolized by cytochrome P450reductase or the mito-chondrial electron transfer chain in carp sperm,leading to ROS production.In the in vitro study,significant increases of sperm ROS lev-els were readily observable even in the lowest concentration (12.5M)(Fig.1),indicating that spermatozoa are sensitive to duroquinone.However,in the in vivo exposure,no signif-icant induction of ROS could be measured after 24h expo-sure in both the low (1mg kg −1)and higher concentrations (10mg kg −1)of duroquinone.Significant increase in ROS could only be observed after 72h exposure to 10mg kg −1,indicating (a)in vivo induction of ROS is time dependent,and (b)in vitro induction of ROS is more rapid than in vivo induction,prob-ably because sperms were in direct contact with duroquinone.Notably,the levels of ROS measured in the in vitro experiment were also higher than those in the in vivo experiment.Oxidative damage would occur when antioxidant capacity is overwhelmed by ROS.However,no data are available for antioxidant capac-ity in fish semen/sperm thus far.In mammalian semen,enzymes such as glutathione peroxidase,glutathione reductase and super-oxide dismutase,as well as antioxidants such as glutathione (GSH),Vitamin C and Vitamin E present in semen can coun-teract H 2O 2and other ROS (Vernet et al.,2004).However,the majority of cytoplasma in spermatozoa was lost during their final stages of spermatogenesis,leading to a low activity of most of the cytoplasmic defence enzymes (Tramer et al.,1998).For example,significant decreases in GPx (−2.4x)and SOD (−1.7x)activities,as well as the amount of Vitamin E (−3.8x),were recorded in rat epididymal spermatozoa undergo the maturation process as they move from caput to cauda (Tramer et al.,1998).Consequently,the lowering of antioxidant capacity may make sperm particularly susceptible to ROS damage (Donnelly et al.,2000).Lipid peroxidation has been used as an indicator of oxida-tive damage in fish tissues (Kelly et al.,1998).Thomas andB.Zhou et al./Aquatic Toxicology77(2006)136–142141Wofford(1993)showed that lipid oxidation,as measured by mamondialdehyde(MDA),significantly increased in liver and ovarian homogenates after Atlantic croaker(Micropogonias undulatus)were exposed to a PCB mixture(Aroclor1254). Recently,Oakes and Van der Kraak(2003)reported oxida-tive stress(quantified by TBARS)in liver and gonad of white sucker(mersoni)exposed to pulp and paper mill effluent. In this study,however,LPO level only increased significantly when sperm were exposed to≥50M of duroquinone for2h (Fig.2)despite marked increase of ROS level could be found readily in lower concentrations(Fig.1).Similar to mammalian sperm,fish sperm also contain a high level of polyunsaturated fatty acids(PUFA),which are particularly susceptible to ROS induced lipid peroxidation(Vernet et al.,2004).The fact that a higher LPO level was found in the in vitro experiment as compared with the in vivo exposure suggest a higher antiox-idative capacity in vivo.However,it should be cautioned that although TBARS is commonly used as an indicator of lipid peroxidation,thiobarbituric acid can react with a number of other oxidation products,e.g.4-hydroxynoneal,other unsatu-rated aldehydes,and endoperoxides.Consequently,the TBARS test is non-specific,and only provides a crude measure of lipid peroxidation.Comet assay has been extensively used to detect DNA lesions in somatic cells from aquatic animals,e.g.blood cells,hepa-tocytes,gill cells(Mitchelmore and Chipman,1998;Lee and Steinert,2003)although it would be difficult to correlate the observed genotoxicity with mutation and reproductive impair-ments.This study represents thefirst successful attempt to apply this assay to detect DNA damage infish sperm and relate this to reproduction.DNA damage,as indicated by comet assay, showed a clear dose-dependent response to concentration of duroquinone and associated with enhanced ROS production. Mature spermatozoa have a low DNA repairing ability and also a weak antioxidant capacity(Van Loon et al.,1991),and are therefore more prone to oxidative stress caused by ROS.For in vivo exposure,significant DNA damage was observed only in the10mg kg−1treatment,and this result is in agreement with enhanced ROS induction in this group.High concentrations of uric acid,an important antioxidant in seminal plasma were reported in rainbow trout(Oncorhynchus mykiss),yellow perch (Percaflavescens)and carp(C.carpio),which can potentially protectfish spermatozoa against oxidative damage(Ciereszko et al.,1999).Interestingly,a good correlation between ROS pro-duction and DNA damage,as well as a good dose-dependent relationship between the ROS production and hatching rate, were found in this study.The results appear to suggest that DNA damage in sperm may affect subsequent hatching success, despite fertilization rate was not affected.Au et al.(2002)demonstrated a good relationship between sperm motility and sperm quality,as measured by subsequent fertilization success and hatching success in the sea urchin Anthocidaris crassipina.Conceivably,fertilization ability may depend on the capacity of sperm to move fast enough tofind the egg in water after their releases.As such,pollutants affecting sperm velocity may significantly reduce fertilization success.All three parameters of sperm motility,VCL,VSL and V AP,were significantly reduced after in vitro exposure to duroquinone, future experiments should be carried out to determine whether in vivo exposure could also decrease sperm motility and subse-quently fertilization rate.ROS production and the resultant lipid and DNA damages have been implicated in detrimental effects on reproduction (Agarwal et al.,2003).In mammals,ROS itself has been shown to inhibit sperm movement via the depletion of intracellular ATP (Armstrong et al.,1999).Studies have also indicated that oxida-tive damage to sperm plasma membrane by ROS may impair sperm function,leading to male infertility(Agarwal et al.,2003). In addition,sperm DNA quality is vital for correct conveyance of genetic material to the next generation,and negative correla-tions have been recorded between the percentage of spermatozoa with fragmented DNA and the decline in sperm counts associ-ated with male infertility(Sun et al.,1997;Shen et al.,1999; Barroso et al.,2000).Conceivably,the marked increase in sperm ROS production associated with DNA damage and significant decrease in sperm motility may in turn impairfish reproduc-tion.In this study,however,no change in fertilization rate could be found in the treatment groups exposed to duroquinone,and our result is consistent with that reported in Donnelly et al. (2000),who found human sperm with damaged DNA did not alter subsequent in vitro fertilization rate.In agreement with previousfindings,in which increasing the concentration of4-nitroquinoline-N-oxide caused a proportional decrease in the percentage of grass shrimp embryos hatching from the egg sacs and a proportional increase in DNA tail moments(Lee and Steinert,2003),our results also showed a significant increase of DNA tail moments associated with decrease in hatching rate after in vitro exposure.Taken together,the results of the present study clearly demon-strated that duroquinone can induce ROS production in sperma-tozoa both in vivo and in vitro,which results in oxidative dam-age in DNA and lipid peroxidation.These oxidative responses were associated with reproductive impairment,as measured by decreases in sperm motility and hatching success.It may be concluded that xenobiotic-induced ROS production and corre-sponding oxidative damage may be one of the major causes damagingfish reproduction processes.From a more practical point of view,the use offish sperm for in vivo or in vitro assays may provide a quick,cost effective means for routine monitor-ing and toxicological assessment of paring with another non-invasive method,such as the use of blood cells from fish for in vitro and in vivo exposure(Nacci et al.,1996;Andrade de et al.,2004),the oxidative stress on spermatozoa can be more directly related to reproductive parameters such as subsequent motility impairment and hatching success. AcknowledgementsWe would like to thank Michael Chiang for assisting infish sperm motility measurement and Department of Agriculture, Fisheries,and Conservation,HKSAR,for providing the com-mon carp.This research was partially supported by a grant from the Hong Kong Research Grants Council(Project No.:AoE/P-04/04)to RSSW.。
