The Comet Assay as Biomarker of Heavy Metal Genotoxicity in Earthworms

The Comet Assay as Biomarker of Heavy Metal Genotoxicity in EarthwormsS.A.Reinecke,A.J.ReineckeDepartment of Zoology,Stellenbosch University,Private Bag X1,Matieland,7602,South AfricaReceived:12November 2002/Accepted:21July 2003Abstract.The ubiquitous occurring earthworm species,Eise-nia fetida,were exposed to nickel chloride to determinewhether the heavy metal Ni caused DNA damage,as measured by the comet (single cell gel electrophoresis)assay.Primary cell cultures of earthworm coelomocytes were exposed in vitro and whole animals either in spiked artificial soil water or in spiked cattle manure ets formed were scored using mean tail lengths as well as comparing percentages of damage in five different damage classes.The exposure con-centrations used for the in vitro exposure (2,6,and 12␮g/ml)caused the formation of comets of which the mean tail lengths differed significantly (p Ͻ0.05)from those of unexposed controls but not from each other.Coelomocytes from worms exposed in artificial soil water at concentrations of 0.0049,0.0078,0.0175,and 0.025mg/ml formed comets of which the mean tail lengths differed significantly (p Ͻ0.05)between the exposure groups with increasingly longer tail lengths with higher concentration (dose-related response).The tail length means of the comets of the three highest exposure concentra-tions also differed significantly (p Ͻ0.05)from the controls.No dose-related response was found between comet tail lengths of the three exposure concentrations (60,240,and 480mg/kg)used for the worms in the cattle manure substrates,but the mean tail lengths of comets from all three exposure groups differed significantly (p Ͼ0.05)from the controls.The comets formed in cells from animals exposed in artificial soil water and in cattle manure substrates,scored within damage classes,indicated a clear shift with increasing exposure concentrations from low to high damage.Our results indicated DNA single-strand breaks in soil invertebrate cells caused by exposure to a nickel compound,verifying previous findings for mammals which indicated that this heavy metal has genotoxic potential.These results therefore suggest that earthworms may be useful indicator organisms to assess the genotoxic risks of nickel pollution to terrestrial environments and that the comet assay is a useful tool to use as biomarker of genotoxic effects on invertebrates in soil.Upon exposure to environmental contaminants,molecular,bio-chemical,and/or physiological compensatory mechanisms may become operative in the exposed organisms.This can result in inhibition or facilitation of one or more physiological processes or functions and/or structural changes.Biomarkers are sensi-tive indicators that reveal these changes.They are therefore also indicators that toxicants have entered the organism (Ever-aarts 1995)and,as such,provide information about the bio-availability of the toxicant.Furthermore they indicate that the toxicant(s)have been distributed within the organism and have initiated a toxicological effect at certain critical targets.As measures of toxic effects in organisms at the below individual (molecular or cellular)level (Van Gestel and Van Brummelen 1996),they represent early or initial responses to environmen-tal perturbations and contamination.Therefore,they are gen-erally taken as being more sensitive than tests measuring ef-fects at higher levels of biological hierarchy,such as the organ,individual,or population (McCarthy and Shugart 1990).In contrast to the simple measurement of contaminants accumu-lating in body tissues,biomarkers can offer more complete and biologically relevant information on the potential impact of toxic pollutants on the health of organisms (Shugart et al.1992).Various biomarkers have been used effectively during the past two decades,but primarily as single tests for specific toxicants.In the field,however,chemical pollution occurs as complex mixtures.This makes prediction of the resulting ef-fects rather difficult and requires the study of multiple biolog-ical endpoints (Devaux et al.1998;Flammarion et al.2002).Some reports of multibiomarker studies in freshwater environ-ments have been published,e.g.,those of Machala et al.(1997),Devaux et al.(1998),and Flammarion et al.(2002),and most of them included,among others,a biomarker of genotoxicity.As similar studies have not been done for soil organisms,we conducted the present study to investigate the possibility of adding a simple genotoxic biomarker to the list of biomarkers presently used for the monitoring of the soil environment.More specifically the aim was to evaluate the applicability of a genotoxicity assay,i.e.,the comet assay or single cell gel electrophoresis test (SCGE),for assessing the level of DNA damage caused by heavy metals in soil invertebrates.Among the molecular compounds of the cell,DNA is an important target of environmental stress in both aquatic and terrestrial organisms (Frenzilli et al.2001).The loss of DNACorrespondence to:S.A.Reinecke;email:SAR@sun.ac.zaArch.Environ.Contam.Toxicol.46,208–215(2004)DOI:10.1007/s00244-003-2253-0A R C H I V E S O FE nvironmental C ontamination a n d T oxicology©2004Springer-Verlag New York Inc.integrity may determine the induction of mutations(Kadhim and Parry1984;De Flora et al.1991;Baily et al.1996)and other irreversible toxic effects such as the“genotoxic disease syndrome”of invertebrates(Kurelec1993).The level of strand breakage in DNA has been proposed as a sensitive indicator of genotoxicity and an effective biomarker in environmental bio-monitoring(Shugart1990;Balpaeme et al.1996).Strand breaks are a potential biomarker because they are common modifications that may be produced by a wide range of agents and mechanisms(Nacci and Nelson1992;Mitchelmore et al. 1998).They may be produced directly by chemicals such as H2O2(hydrogen peroxide)and other reactive oxygen species (ROS).The latter may result from endogenous metabolism or can be produced in excess from redox cycling or other free radical interactions associated with organic xenobiotics,me-tabolites,and transition metals.The fairly novel technique of alkaline single-cell electrophoresis provides a very sensitive method for detecting strand breaks(Fairbairn et al.1995).It is a promising technique that has been applied mainly to the study of DNA single-strand breaks induced by chemicals in humans and experimental mammals(Betti and Nigro1996).It is a rapid,simple,and sensitive method to detect strand breaks in all kinds of cells,but with two important requirements:thefirst being that the cells should be available as a single cell suspen-sion and the second,that freshly collected cells/tissues are used,instead of frozen or preserved material(Ostling and Johanson1984;Singh et al.1988;McKelvey-Martin et al. 1993;Fairbairn et al.1995;Tice1995;Tice et al.2002). The only records available of the comet assay being used to study genotoxic effects in soil invertebrates are from the work of Verschaeve and Gilles(1995).They studied the effects of dioxin,mitomycin C,and X-rays on earthworms.During the present study earthworms were also used as models as they are important non-target animals in soils,often affected by metal containing chemicals.Although the effects of many heavy metals on earthworms are well studied in terms of their bioac-cumulation and toxicity with regard to life-cycle parameters (Reinecke and Reinecke1996),less is known about their genotoxicity.While the range of toxic chemicals in the environment is large,heavy metals represent one of the most widespread and serious forms of environmental contamination(Pruski and Dixon2002).The list of metals whose compounds are known to be carcinogenic to humans include As,Be,Cd,Cr,and Ni. More candidates are emerging from current research,including Co,Cu,and Fe(Kasprzak and Buzard2000).Nickel(Ni)is an ubiquitous,naturally occurring element in soil and a recognised environmental and industrial pollutant(Wozniak and Blasiak 2002).Anthropogenic input,such as the burning of fossil fuels, wear of nickel-coated surfaces,and spreading of sewage sludge and manure(Scott-Fordsmand et al.1998),can also cause elevated concentrations in soils.Elevated levels of nickel con-centrations in the soil environment,whether caused by the presence of serpentine soils(Mesjasz-Przybylowicz et al. 2001)or by anthropogenic Ni discharges,may lead to accu-mulation of the metal in soil-dwelling organisms.This may have complications for important nontarget organisms,such as earthworms and their predators(Reinecke et al.2000). Apart from humans,studies on the environmental effects of Ni on other organisms are scarce.With regard to soil organ-isms,Hartenstein et al.(1980)and Neuhauser et al.(1985)showed Ni to be,among other metals,detrimental to growth of earthworms,and Scott-Fordsmand et al.(1998)studied the effects of Ni on the lysosomal membrane integrity of cells in these organisms,using the neutral-red retention biomarker. Since Ni was shown to be one of the heavy metals carcino-genic to humans(Wozniak and Blasiak2002),we exposed earthworms to nickel chloride and investigated the potential use of the comet assay as a biomarker of genotoxicity.The primary aim was to determine whether exposure of earthworms to nickel chloride resulted in strand breaks in DNA which could be reliably measured with the comet assay.The second-ary aim was to determine whether these effects were dose-responsive.Biomarker tests are often used with different aims in mind. Organisms can,for example,be used as bioindicators which are then collected in thefield and their cells tested directly.Bio-assays,using polluted substrates and exposing organisms under laboratory conditions,could,however,also be done,which entails longer exposures in the soil medium.In trying to estab-lish bioavaiability of substances to be tested,different other exposure routes could also be used,such as in vitro exposures using isolated cells or exposure in vivo by direct uptake or through feeding.We therefore tested DNA damage using dif-ferent exposure routes to evaluate the suitability of the assay to detect changes.This was done after direct cell exposure(in vitro),absorption through direct contact with the epidermis (artificial soil water),and the more ecologically relevant method by exposure in a natural substrate where the toxicant could be absorbed via skin contact as well as by feeding.Due to the differences in bioavailability of the toxicant to the organisms and/or cells and the resulting detrimental effects (such as mortality of cells or experimental animals)in the different media,it was deemed necessary to use a wide range of different exposure concentrations and ranges for the differ-ent experiments.Materials and MethodsAnimalsEisenia fetida Savigny1826was chosen for this study because it is an internationally accepted model species for toxicity testing with a cosmopolitan distribution(Edwards1984).Synchronized cultures were obtained from the breeding stock in our laboratory. Collecting CoelomocytesEarthworm coelomocytes were obtained by using a slightly modified noninvasive technique,originally described by Eyambe et al.(1991). The animals were exposed to an irritating extrusionfluid to which they react strongly by secreting coelomicfluid containing the coelomo-cytes.This extrusionfluid was similar to that used by the above-mentioned authors but PBS was used in stead of the saline solution used by them.The worms were also not exposed to thefluid in Petri dishes but in Eppendorf tubes.Worms were placed seperately into the tubes,and1ml of the extrusionfluid was added to each tube and left for3min,whereafter they were removed,washed in water,and placed back into the substrate.The extrusionfluid containing cells were centrifuged and the supernatant removed.The cell pellet was sus-The Comet Assay for Heavy Metal Genotoxicity209pended in PBS and washed three times in PBS using microcentrifu-gation for3min at380g.The concentration of cells in thefinal suspension was determined using the trypan blue exclusion method and dilutions calculated to be used for the exposures(in vitro)and for the comet assay,described below.Exposure of Experimental Animals and CellsConcentrations of NiCl2Used.Preliminary exposures were done foreach of the three experiments described below to determine the con-centrations of the toxicant to be used.The exposure ranges and the reasons for using them are given with the description of each of the experiments.In Vitro.For the in vitro exposures,concentrations of Ni were prepared by adding the specific nickel chloride concentrations in PBS to the cell suspensions in volumes that amounted to the concentration and volume of cells needed.By exposing the cells to a range of concentrations and measuring cell viability with the trypan blue ex-clusion method,thefinal exposure concentrations were chosen.A lower cell viability was found at concentrations higher than15.0␮g/lNiCl2,and three exposure concentrations,i.e.,2.0,6.0,and12.0NiCl2␮g/l,were therefore used for thefinal experiment.Ten replicates of the exposures,as well as controls without any nickel present,were pre-pared and the containers with cells incubated for1h in the dark at4°C before the cell suspensions were used for the comet protocol.In Artificial Soil ing the protocol provided by Kiewiet and Ma(1991),distilled water,reconstituted as described below to resemble groundwater,was prepared.Exposure concentrations wereprepared by dissolving NiCl2in the groundwater.Concentrations werechosen by exposing earthworms to a wide range of concentrations of the toxicant.It was found that worms did not survive concentrations higher than0.25mg/l for the exposure time selected and a concentra-tion range between0.0049and0.1mg/l selected(i.e.,0.0049,0.025, 0.05,and0.1mg/l).When thefinal experiment was done it was found that the cells harvested from the worms at the two highest exposure concentrations(0.05and0.1mg/l)had a very low viability,as shown with the trypan blue method,and two more concentrations in between 0.0049and0.025were added to obtain a range of at least four concentrations.Thefinal range was therefore0.0,0.0049,0.0087, 0.0175,and0.025mg/l.Clitellate worms from a synchronized control culture were weighed to select a group of homogenic biomass.They were then placed on moistfilter paper in Petri dishes for48h in the dark at20°C to depurate their gut contents.The artificial soil water with the prepared Ni concentrations was aerated,using an aquarium pump and aerators for6h before introducing the worms.The worms were exposed in glass beakers in the dark at20°C for48h and the water continuously aerated.Two replicates offive worms per500ml of water were used for each concentration and the control.In Substrate.Substrates were prepared by dissolving the amount of NiCl2in distilled water and mixing it thoroughly into200g of washed, dried cattle manure in a180ϫ110ϫ85-mm plastic container to attain a moisture content of75%.The prepared substrates were left for48h in a climate-controlled room at20°C and70%RH to stabilize.Two replicates were prepared for each exposure concentration and the control and exposure concentrations used were60,240,and480mg/kgNiCl2.These exposure concentrations were chosen by doing a rangefinding test using life-cycle parameters such as growth and reproduc-tion as criteria and limiting the exposure range to concentrations below the point where these parameters were negatively affected.Clitellate earthworms were collected from control cultures and10worms added per200g substrate and the container was closed with a perforated lid and covered with black plastic.The worms were exposed to these substrates for4weeks in the climate-controlled room.They were removed and treated as described above to collect the coelomocytes, and the comet assay was performed.No food was added during the exposure period,as it has been shown that the amount of cattle manure substrate used,provided sufficient food for the4-week exposure period (Reinecke and Reinecke1996).The Comet AssayWe followed the procedure as described originally by Singh et al. (1988),with a few minor modifications.In this assay,cells are embedded in an agarose gel and lysed by detergents.Subsequently,the gel is placed in an alkaline electrophore-sis buffer for DNA unwinding.The presence of strand breaks within the nuclear DNA increases the electrophoresis mobility of the DNA fragments.During electrophoresis these single-stranded DNA frag-ments move away from the damaged nuclei within the gel and form a comet-like pattern.After staining with ethidium bromide,“comet tails”and the remaining nuclear DNA can be visualized using a fluorescence microscope.The modifications to the method of Singh et al.(1988)were the following:Normal microscope slides,not fully frosted slides,were used.These slides,covered with thefirst agarose layer,were,however, dried in an oven at60°C for30min to enable the adherence of the gel layer to the slides;cell lysing was only done for1h instead of overnight.Scoring was done under a Leitz Diaplanfluorescence microscope equipped with a Ploemopak2.3(excitationfilter,515–650nm;barrier filter,580nm).The scoring was done by randomly scanning and measuring100comets per slide.The comets selected for scoring were of uniform nuclear size.This was deemed very important due to the variety of cell types which occurs in the earthworm coelomicfluid and which are layered onto the gels.The cell type selected for this assay was the neutrophilic coelomocytes.The scoring was done by visual inspection under the microscope and measurements of the head and tail lengths of comets were made with an eyepiece micrometer and accorded a numeric value with regard to the following damage classes: largely undamaged—no tail visible(class1);low damage—tail length not more than30␮m and with lowfluorescence and head still round and brightlyfluorescent(class2);medium damage—tail length be-tween30and50␮m and head and tail about equally brightlyfluores-cent(class3);high damage—tail length between50and70␮m and bright and head small and weaklyfluorescent(class4);and extreme damage—tail length more than70␮m and head not a round unit anymore(class5).Comets where the head had disintegrated fully with only the tail visible was deemed to be apoptotic and were not counted. For each treatment100comets per slide of three slides per treatment were scored.Positive and negative controls were always included.For the positive control the slides were exposed to UV radiation in an Ultralum UVC-508Crosslinker at120,000␮J for1.5min directly after layering of cells and before lysis.The negative control cells were exposed to the solvent(Milli-Q water)without any chemicals added. Statistical AnalysisSignificant differences between the results of the different treatment groups were determined using one-way ANOVA and Tukey’s all pairwise multiple comparison.Where the normality failed,the Kruskal–Wallis one-way test was used,in the case of significance, followed by Dunn’s post hoc analysis.The level of significance was pϽ0.05.(SigmaStat,Version2.0,Jandel Scientific).210S.A.Reinecke and A.J.ReineckeResults and DiscussionThe results of comet tail lengths are given in Figs.1,2,and4. The percentages of comets in the different damage classes,as explained in the methods,are presented in Figs.3and5.For discussions of other methods of measuring and quantification of DNA damage and the rational behind them,see the reviews by McKelvey-Martin et al.(1993)and Fairbairn et al.(1995). The results of the mean tail lengths,as a measurement of DNA damage,using the in vitro exposure,are shown in Fig.1. Significant differences(pϽ0.05)between the positive and the negative controls clearly indicated the effectiveness and integ-rity of the method.The responses to the three exposure con-centrations used differed statistically from both controls but not from each other.This indicates that the presence of nickel chloride reduced the DNA integrity in the cells at very low concentrations.No dose-related response was,however,evi-dent for the three exposure concentrations and it might there-fore be necessary to repeat this experiment,including more exposure concentrations at smaller increments as those used for this experiment.The results of this experiment did,however, indicate that it is feasible to use earthworm coelomocytes in a primary cell culture,to do a rapid screening for the genotoxic effects of the metal.In vitro studies have the potential to provide experimental results with high statistical power(Lovell and Thomas1997).It is,however,necessary to give careful consideration to the need for appropriate replication and repeat experiments,particularly if the level of the dose–response relation is more relevant for biological interpretation than a simple positive result.Using the cell as the only unit in the experimental design can lead to misinterpretation(Lovell and Thomas1997),especially since the low concentration ranges to which cells are exposed are several orders of magnitude lower than those which give results in whole animals.Also,failure to include interanimal variability in this design can lead to incorrect conclusions.In vivo studies,as done in the present study,should therefore also be included to make more realistic deductions from the results. In the experiment where earthworms were exposed directly to a solution of nickel chloride in artificial soil water,a dose-related response(increasing number of DNA strand breaks,as indicated by tail length,with increasing concentration)was evident(Fig.2).In this experiment significant differences(pϽ0.05)between the positive and the negative controls were also found.The effect of the lowest exposure concentration(0.0049 mg/l)was not significantly different from that of the negative control.This concentration did not seem to cause more strand breaks than that which occurs normally.The range of exposure concentrations used in this experiment seems to encompass the range within which the animals can still survive for the expo-sure period.Cells of animals exposed to concentrations higher than0.025mg/l had a very low viability(as measured with the trypan blue exclusion method)and results of DNA breaks could therefore not be relied on to give accurate indications of genotoxic damage.At an exposure concentration of0.05mg/l the animals did not survive the exposure period.In Fig.3a comparison is given of the percentage of comets in the different damage classes at the different exposure concentrations.Here, the very high percentage of extremely damaged cell nuclei at the0.025mg/kg NiCl2exposure concentration can be seen,and it can be deduced that the rate of DNA repair would have had to be very high to compensate for such damage.In Fig.3a clear shift can also be seen,with the majority of comets falling in the lower damage classes,at the lower exposure concentra-tions,to the majority in the higher damage classes at the higher concentrations.This is also indicative of a dose-related re-sponse.Zhuang et al.(1996),working on mammalian cell lines,also found a dose-related response between Ni exposure and comet tail length.DNA strand breaks were evident as a result of exposure to Ni spiked substrates as shown in Fig.4.Significant differences (pϽ0.05)between that of the negative control and all the exposure concentrations were evident,however,insufficient exposures precluded the establishment of a dose-related re-sponse reliably.Again,as with the in vitro experiment,a dose-related response might be evident if more concentrations with smaller increments could be included.In Fig.5,where the damage classes are plotted,a shift from less to more comets from the lower to higher classes between the60and the240 mg/kg NiCl2exposure concentrations can be seen.At the highest concentration used in this experiment(480mg/kg), compared to240mg/kg exposure concentration,a shift back to the lower damage classes occurred.Measurement of cell via-bility indicated a lower number of viable cells at the latter concentration than in the others,which could indicate that,with high concentration exposures over longer periods of time, cytotoxicity could obscure genotoxicity.According to Hart-mann et al.(2001)it is important to measure cytotoxicity when screening for genotoxicity with the comet assay,as the occur-rence of cells with completely fragmented chromatin could lead to false-positive test results.Tice et al.(2000)also indi-cated that it has not been well established whether cytotoxicity influences DNA migration in the comet assay and whether it can influence the test results.Dead or dying cells can undergo rapid DNA fragmentation,which should be expected to in-crease DNA migration in the comet assay.It is therefore mandatory to perform concurrent viability tests to control for excess cytotoxicity.The comet assay provides an advantage over other strand break assays because measurements are made on individual cells.Scoring these cells on slides provides an independent measure of the toxicity of a test compound.Dead cells can be identified by their distinct morphology compared to cells ex-hibiting DNA damage(Fairbairn et al.1995;Speit et al.1996). The presence of apoptotic cells can also have an effect on the results of the comet assay,since the onset of apoptotic frag-mentation might appear as comets(Henderson et al.1998; Roser et al.2001).Apoptotic cells can,however,be distin-guished on the slides and therefore compensated for in the methods or in the calculations.Nickel is known to be potentially carcinogenic for mamma-lian cells.Our results provide evidence of its genotoxicity and potential for carcinogenicity in earthworms.According to Hartwig(1995)and Kasprzak and Buzard(2000)the mode of action is still not well understood and direct genotoxic effects are rather weak and/or restricted to comparatively high con-centrations.Two types of action seem to be predominant in the way in which heavy metals cause damage.First,it could happen by the induction of oxidative species(ROS).It is postulated that the effects of Ni can result from the generation of oxygen radicals in the interaction between metals and pro-The Comet Assay for Heavy Metal Genotoxicity211teins.These radicals can subsequently interact with DNA,inducing damage to its bases,causing DNA strand breaks and probably DNA–protein cross-links (Kasprzak et al.1992;Misra et al.1993).Research on the specific mechanism is ongoing (Dally and Hartwig 1997;Hartwig 1998;Takahashi et al.2000)but is not within the scope of the present paper.A second possible damaging action could happen when the metal interacts with DNA repair processes.Current evidence as de-scribed by Hartwig (1998)suggests that DNA repair systems are very sensitive targets for nickel(II),leading to a diminished removal of endogenous DNA lesions and of DNA damage induced by environmental agents,which in turn may increase the risk of tumor formation.It has been suggested that the underlying mechanism might have to involve the ability of the metal to compete with magnesium ions or to displace zinc ions in zinc finger structures of DNA repair enzymes (Hartwig 1998).According to Kasprzak and Buzard (2000)a distinction must be made between Ni and the tentative carcinogens,Co,et tail lengths.Result of in vitro exposure of Eisenia fetida to NiCl 2:mean (␮m)Ϯstandard deviation (SD;error bars).Asterisks indicate statistical differences.Different numbers of asterisks indicate statistically significant differences;same number,differences not statisticallysignificantet tail lengths.Result of exposure of Eisenia fetida to NiCl 2in artificial soil water.Mean (␮m)Ϯstandard deviation (SD;error bars).Asterisks indicate statistical differences.Different numbers of asterisks indicate statistically significant differences;same number,differences not statistically significant212S.A.Reinecke and A.J.Reinecke。