Monitoring DNA damage in indigenous blue mussels (Mytilus edulis) sampled from coastal sites in Denmark

Mutation Research 585 (2005) 33–42

Monitoring DNA damage in indigenous blue mussels (Mytilus edulis) sampled from coastal sites in Denmark Jette Rank ∗ , Klara Jensen, Per Homann Jespersen Department of Environment, Technology and Social Studies, Building 11.2, Roskilde University, DK-4000 Roskilde, Denmark Received 16 March 2004; received in revised form 7 March 2005; accepted 8 April 2005 Available online 13 June 2005

Abstract Damage to DNA detected by use of the single-cell gel electrophoresis (comet) assay was monitored in blue mussels, Mytilus edulis, sampled from coastal waters in Denmark. Mussels from five locations in Køge Bay, an area receiving wastewater from many industries and municipalities, were collected five times during 1999 and six times in 2001. In 1999, both gill cells and haemolymph cells were examined, and sediments were sampled on three dates from the same five locations. In the autumn of 1999, mussels were also collected at six reference sites without known pollution. Results showed a significantly higher level of DNA damage in gill cells compared with haemolymph cells. Because of this, only gill cells were sampled for the monitoring in 2001. Levels of DNA damage, expressed as tail moments, were significantly higher for the mussels in Køge Bay when compared with levels of DNA damage in mussels from the non-polluted coastal areas. No clear seasonal variation was demonstrated. Analysis of the correlation between chromium, nickel, cadmium and mercury in sediments and tail moments in haemolymph and gill cells from the five sites showed a statistically significant positive correlation between tail moments and chromium, nickel and cadmium (P < 0.01). The overall conclusion was that the comet assay on blue mussels could be useful for screening of genotoxic pollution in marine waters. © 2005 Elsevier B.V. All rights reserved. Keywords: DNA damage; Comet assay; Mytilus edulis; Biomonitoring; Genotoxic chemicals

1. Introduction Concern for environmental and human health problems has resulted in a considerable interest in monitoring the effects of pollution in aquatic ecosystems. Of special concern are genotoxic chemicals. ∗

Corresponding author. Tel.: +45 4674 2071; fax: +45 4674 3041. E-mail address: [email protected] (J. Rank).

1383-5718/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.mrgentox.2005.04.008

Besides causing mutations and cancer they are also suspected of having the ability to affect the reproduction of many species and thereby cause enhanced instability in ecosystems under stress [1]. Against this background, many genotoxicity assays applied on marine invertebrates have been proposed for the detection of aquatic pollution [2]. Among the marine invertebrates, mussels are a good choice for the detection of pollution because they accumulate

34

J. Rank et al. / Mutation Research 585 (2005) 33–42

many chemical substances due to their great filtration capacity and their contact with sediments. Moreover, mussels have a low mobility, which ensures that detected adverse effects, most probably, are due to local pollution. For these reasons they are often used as bioindicators of various genotoxic endpoints. Monitoring of genotoxic effects can be carried out in situ by use of either indigenous or transplanted mussels. Mersh and Beauvais [3] transplanted caged zebra mussels, Dreissena polymorpha, to river sites receiving industrial effluents and showed that the frequencies of micronuclei in the haemocytes were significantly increased compared with the baseline level in the control group from a non-polluted area. In another study performed by Herbert and Zahn [4] marine mussels, Mytilus galloprovincialis, were sampled from polluted sites in the Northern Adriatic Sea. The alkaline unwinding assay was used in different experimental designs, which made it possible to show pollution-induced DNA strand breaks in gills and midgut. Native mussels, M. galloprovincialis, sampled from the same area, were used by Vukmirovic et al. [5] to show effects from environmental contamination. They used the alkaline filter-elution method and showed that single strand breaks and DNA crosslinks in haemolymph cells were both influenced by urban and industrial wastes. Also, Bolognesi et al. [6] used alkaline elution to determine the DNA damage in cells of indigenous mussels, M. galloprovincialis. The in situ assay showed statistically significant increases of DNA damage in the gill cells of mussels sampled at 11 polluted sites compared with mussels from a non-polluted site. The results of the alkaline elution assay were similar to results from the micronucleus assay applied on the gill cells from the same mussels. DNA adducts have also been used as indicators of genotoxic exposure. Green-lipped mussels, Perna viridis, were collected from a site in Hong Kong with low concentrations of polyaromatic hydrocarbons (PAHs) and transplanted to three polluted sites [7]. After 30 days of exposure in situ, DNA adducts in the gills were quantified and so were the PAH concentrations in the mussels. The study showed that the adduct levels were related to the PAH concentration and that the PAH concentrations varied highly between individual mussels. Another study performed by Ericson et al. [8] also used DNA adducts as indicators of genotoxic exposure. They collected blue mussels, M. edulis,

from three polluted harbours at the Icelandic southwestern coast and a reference site. The DNA adducts were analysed in gills and digestive glands by use of the 32 P-postlabelling assay. The mussels from the polluted harbours showed significantly higher levels of DNA adducts than mussels from the reference site, and the gills allowed more sensitive detection than the digestive glands. Transplanted mussels from the reference site did not show higher levels of DNA adducts after 6 weeks in situ exposure at the most polluted site. The comet assay, first introduced on mammalian cells in the late 1980s [9], is now commonly used on different species of mussels. Pavlica et al. [10] exposed zebra mussels, D. polymorpha, in situ for 1 month in the river Sana at 11 km downstream from the Zagreb municipal wastewater outlet. The increased level of DNA damage in the haemocytes was similar to the level of DNA damage caused by 7 days experimental exposure to 100 ␮g/l pentachlorophenol. Steinert et al. [11] evaluated blue mussels, M. edulis, in San Diego Bay, where mussels were transplanted from clean sites to polluted sites and exposed for 12 and 32 days. The comet assay was carried out on haemocytes and sperm nuclei and showed that the levels of DNA damage (tail length) had a good correlation with time of exposure as well as the measured contamination of, respectively, sediments and mussel tissue. In 1998, our laboratory performed a pilot monitoring study in Køge Bay south of Copenhagen in Denmark [12]. Blue mussels, M. edulis, were sampled from four polluted sites during the summer and haemolymph cells were analysed by the comet assay. These results showed increased levels of DNA damage compared with mussels from non-polluted sites in Denmark. The present study is an extension of the pilot study and had three main objectives: (1) to compare the sensitivity of gill cells and haemocytes; (2) to examine the seasonal variation of the DNA damage; (3) to investigate correlations between DNA damage and heavy metals in sediments from which the mussels were collected.

2. Materials and methods 2.1. Sampling of the mussels Blue mussels, M. edulis, were sampled from five sites in Køge Bay located south of Copenhagen in

J. Rank et al. / Mutation Research 585 (2005) 33–42

35

Fig. 1. Map of Denmark (grey) and surrounding area showing the five locations in Køge Bay where the blue mussels, Mytilus edulis, were collected in 1999 and 2001. The numbers refer to the non-polluted locations (1, Havnsø; 2, Ordrup; 3, Holbæk; 4, Gershøj; 5, Gilleleje; 6, Hornbæk) used for baseline studies in 1999.

the northeastern part of the Baltic sea (Fig. 1). Mussels were collected from stone dikes in shallow waters near the shore on five dates during 1999 and six dates in 2001. Køge Bay was chosen because it receives wastewater from industries as well as municipalities in a highly industrialised area. At all sampling sites there were plenty of mussels. However, although only the biggest mussels were selected their size was relatively small, ranging from 19.4 to 43.0 mm. Mussels sampled in the autumn of 1999 from six unpolluted sites at the north coast of Zealand (Fig. 1) ranged in size from 40 to 60 mm. The small size of the Køge Bay mussels can be explained by stress from instability of the salinity. Køge Bay receives water from the brackish Baltic sea and from the more saline Kattegat (Fig. 1) and the salinity will, therefore, change depending on wind and weather. In the present study, the salinity was measured in a range of 7–23‰. Salinity of the non-polluted sites (Fig. 1) ranged from 12 to 20‰, but was stable for each individual site. Mussels were transferred to the laboratory and acclimatized for 2 days at 6 ◦ C with a 7 h light and 17 h dark cycle in aerated seawater from the sampling site of origin before the comet assay was performed.

2.2. Physical and chemical parameters Salinity, temperature, oxygen concentration and pH were measured for each site at every sampling time. Sediment samples were collected at the same locations on the last three sampling dates in 1999 and analysed for chromium, nickel, cadmium and mercury (Table 1), which are all heavy metals supposed to be genotoxic. The sediments were frozen directly after collection and kept at –20 ◦ C until the analysis of the heavy metals was performed. The sediments were dried during 20 h at 105 ◦ C and 0.5 g was digested in concentrated HNO3 for 30 min in a microwave oven. Two samples per sediment were filtered and diluted with distilled water to a volume of 25 ml. A Varian spectrophotometer, SpectraA 250 Plus, was used for the flame atomic absorption analysis. 2.3. Preparation of gill cells and haemocytes A pooled sample was composed of either 0.1 ml haemolymph or the gills from 4 to 10 mussels, depending on the size. In 1999, only one sample was analysed per date and for each site. Haemocytes and gill cells

36

J. Rank et al. / Mutation Research 585 (2005) 33–42

Table 1 Heavy-metal concentrations and the amount of organic matter in sediments from the five sites measured on three sampling dates in 1999 Site

Date

Cr (mg/kg)

Ni (mg/kg)

Cd (mg/kg)

Hg (␮g/kg)

Organic matter (g/kg)

Bøgeskov

27 July 6 September 6 December

5.7 4.8 5.1

1.3 0.8 2.1

0.33 0.03 0.27

5.2 4.1 11.0

9.24 8.44 21.8

Strøby

27 July 6 September 6 December

3.4 4.5 1.9

1.0 1.0 0.6

0.21 ND 0.07

3.8 3.0 2.8

4.80 4.65 5.55

Køge

27 July 6 September 6 December

3.3 2.9 1.9

0.7 0.5 0.3

0.03 0.06 0.06

2.1 2.5 2.8

6.37 6.58 6.11

Mosede

27 July 6 September 6 December

2.9 3.2 2.6

1.2 0.8 0.6

0.11 ND 0.05

3.8 2.7 3.3

4.61 5.82 5.91

Hundige

27 July 6 September 6 December

1.6 2.4 2.2

0.2 0.2 0.2

ND ND 0.04

1.7 0.9 2.3

3.88 2.75 3.33

All values are calculated per kg dry matter of sediment. ND: not detectable.

were examined for each day, except for 23 June, where only gill cells were analysed. In 2001, two replicates were analysed per date and for each site, but only the gill cells were analysed. The preparation of the cell suspensions was carried out under yellow light to avoid UV damage of the DNA. A volume of 0.1 ml haemolymph was pulled out from the posterior end of each mussel with a 1-ml sterile syringe by forcing the needle into the space between the two shells. The haemolymph was centrifuged at 3500 rpm (1123 × g) for 3 min. The supernatant was removed and the cells in the pellet were ready for the comet assay. The anterior and posterior adductor muscles were cut with a scalpel, and the gills were removed and placed in 2 ml cold (4 ◦ C) calcium- and magnesiumfree saline (CMFS) consisting of 20 mM HEPES, 200 mM NaCl, 12.5 mM KCl, 5 mM EDTA, according to the procedure described by Wilson et al. [13]. The gills were washed four times in 2 ml CMFS by moving them from one test tube to another. In the last tube, the gill tissue was minced for 1 min with a pair of scissors. The gill cell suspension was poured into a 25-ml Erlenmeyer flask with 8 ml CMFS and shaken at 60 rotations per min for 1 h at 1 ◦ C in the dark. After filtration of the suspension through a 20␮m sieve and centrifugation at 2000 rpm (443 × g) for 5 min, 8 ml of the supernatant was removed and the

cells were resuspended in the remaining 2 ml. The suspension was centrifuged at 3500 rpm (1123 × g) for 3 min, the supernatant removed and the cells in the pellet were ready for the comet assay. 2.4. Comet assay procedure In the present study, the method used was mainly based on the protocol described by Wilson et al. [13]. Cells were embedded in two layers of agarose on frosted slides. The agarose solutions used for the gill cells were prepared in Kenny’s salt solution (200 mM NaCl, 9 mM KCl, 0.7 mM K2 HPO4 , 2 mM NaHCO3 , pH 7.9). For the haemolymph cells a phosphatebuffered salt solution (137 mM NaCl, 2.7 mM KCl, 1.75 mM KH2 PO4 , 8 mM Na2 HPO4 , pH 7.35) was used. Normal melting-point (NMP) agarose (0.8%) was melted in a microwave oven for 45 s and kept in a water bath at 55 ◦ C for 15 min, and from this solution 100 ␮l were transferred to frosted slides and allowed to gel for 15 min. Low melting-point (LMP) agarose (0.65%), was melted in a microwave oven for 45 s and kept in a water bath at 37 ◦ C for 15 min, and 125 ␮l were added to the pellet of cells and mixed on a Whirl mixer. From this suspension 100 ␮l were transferred to the slides precoated with NMP agarose. A cover glass was placed on the gel and the slides were cooled for 15 min on ice to make the gel solid. The cover glass was removed

J. Rank et al. / Mutation Research 585 (2005) 33–42

37

and the slides were placed in the lysis solution (2.5 M NaCl, 0.1 M EDTA, 10 mM TRIS, 1% Triton X-100, pH 10) for at least 1 h in a refrigerator. Excess solution was removed by rinsing the slides for 10 s in the electrophoresis solution (0.3 M NaOH, 1 mM EDTA). Slides were then placed in the electrophoresis chamber and covered with the cold (4 ◦ C) electrophoresis solution. Alkaline unwinding (pH > 13) was carried out for 15 min followed by 5 min of electrophoresis at 300 mA and 25 V. The pH was brought to neutrality by rinsing the slides two times for 5 min in 0.4 M TRIS. All work was carried out in yellow light to avoid UV damage of the DNA. 2.5. Scoring of DNA damage using image analysis The DNA in the nuclei was stained with 85 ␮l ethidium bromide (20 ␮g/ml). After at least 5 min of staining, the slides were examined by use of a fluorescence microscope (Dialux, 50× oil-immersion objective). Slides were scored blind, as recommended by Albertini et al. [14]. On each slide, 50 randomly chosen cell nuclei were analysed automatically using software from Kinetic Imaging (Komet 3.1). To avoid interference from apoptotic and necrotic cells, highly damaged nuclei were not scored. The DNA strand breaks are expressed in the unit tail moment, TM (tail length in ␮m multiplied by DNA% in the tail). 2.6. Statistical analyses The unit tail moment (TM) does not follow an obvious statistical distribution. The empirical distribution, shown in Fig. 2 reveals a left-skewed distribution. In a study of Bauer et al. [15], a chi-square distribution was used to describe similar data. However, this did not work satisfactorily with our data. Instead, we used a transformation to obtain normally distributed data in order to be able to use statistical procedures that require normality. It was shown that the fourth root of the tail moment (TM’) fits the normal distribution quite well. Thus, TM’ (the fourth root of TM) was used throughout the statistical analyses. The 4802 measurements of tail moment were analysed in a three-way analysis of variance (ANOVA) design with location, date and tissue as independent variables, and with all interactions. Analysis of correlations between TM and the chemical and physical parameters measured in

Fig. 2. The empirical distribution of the standardised residual for TM with a normal-distribution curve.

sediments was carried out for the three last sampling dates in 1999.

3. Results 3.1. Replicates In 2001, DNA damage in gill cells were analysed in two replicates. Even though a slight decrease in the variance within each sample could be observed with increasing number of mussels, statistical analysis showed that in 27 out of 29 comparisons no significant (5%) difference in tail moments (TM) could be observed between the means of the two replicates. Based on these results, the duplicate measurements were considered as one group. 3.2. Gill cells versus haemocytes In 1999, tail moments were measured in both gill cells and haemocytes, as shown in Fig. 3. Statistical analysis showed that tail moments in gill cells were significantly higher than in haemocytes (P < 0.0005). Given these results, it was decided to measure DNA damage in gill cells only in the monitoring work of 2001.

38

J. Rank et al. / Mutation Research 585 (2005) 33–42

Fig. 3. DNA damage shown as tail moment for: (A) haemolymph cells; and (B) gill cells from blue mussels, Mytilus edulis, sampled in 1999. At 23 June no analysis was made for the haemolymph cells. The dotted line shows the baseline of DNA damage obtained from mussels sampled in the autumn of 1999 at non-polluted sites.

3.3. Seasonal variations In spite of missing data from 23 June for the haemolymph cells (Fig. 3), a seasonal variation in tail moments was demonstrated in samples collected in 1999. The highest tail moments were observed in mussels sampled in the summer month (June and July) and the lowest values were seen in December. The highest level of DNA damage, 19.4 TM, was observed in gill cells of mussels sampled in June at the location Køge. A seasonal pattern in tail moments was not observed in 2001 (Fig. 4), and statistical analyses could not support the hypothesis of a seasonal variation, which indicates that the findings in 1999 could be fortuitous. In general, the DNA damage in 2001 was lower than that observed in 1999, with most tail-moment values below 5, indicating that the level of genotoxic pressure in Køge Bay was lower in 2001 compared with 1999.

3.4. Reference sites and baseline levels of DNA damage In the pilot study from 1998 [12], the location of Strøby was supposed to have the lowest level of DNA damage, because it was situated away from the outlets of municipal and industrial wastewaters. Results showed that this was not the case, and for the subsequent work the location Bøgeskov was included in the study, as this site was located in an area without outlets from industries or cities. However, results showed that mussels from Bøgeskov also had high levels of DNA damage (Figs. 3 and 4). Statistical analysis of all data from 1999 and 2001 showed that DNA damage at the two locations, Bøgeskov and Strøby, was significantly higher (P < 0.05) than for the three other sites, and that DNA damage for the locations decreased in the following order: Strøby > Bøgeskov > Mosede > Hundige > Køge

J. Rank et al. / Mutation Research 585 (2005) 33–42

39

Fig. 4. DNA damage shown as tail moment for gill cells from blue mussels, Mytilus edulis, sampled in 2001. The dotted line shows the baseline of DNA damage obtained from mussels sampled in the autumn of 1999 at non-polluted sites.

(Fig. 5). Therefore, it is not recommended only to use the knowledge of wastewater outlets to coastal areas as an indicator of genotoxic pollution. In an attempt to find the baseline levels of DNA damage, we analysed mussels sampled in the autumn of 1999 from six sites in recreational coastal areas without industries and far from big cities (Fig. 1). The DNA damage was found to be low for both haemolymph (TM, 1.23 ± 0.23) and gill cells (1.38 ± 0.92). In the same period most of the observed tail moments in mussels from Køge Bay (Fig. 3) were higher than for these reference sites. Presuming that these data can be used as a general baseline of DNA-damage, most results from 2001 are above the baseline (Fig. 4), indicating that indigenous mussels in the monitoring area of Køge Bay were exposed to a higher level of genotoxic agents.

3.5. Physical and chemical data Multiple regression analysis including all measurement series from 1999 to 2001, with TM as dependent variable and the four water parameters as independent variables showed that higher pH (P < 0.0005) and oxygen concentration (P = 0.005) resulted in an increase in DNA damage, whereas increasing salinity had a reducing effect (P < 0.0005). Temperature also showed a positive effect, but just above the 5% significance level. These findings show that water parameters have some influence. However, the cause for the correlations is not obvious and need to be examined further. For three sampling dates in 1999, the amount of organic matter and the concentrations of chromium, nickel, cadmium and mercury in sediments were analysed for correlation with the tail moments. Positive and significant (P < 0.01) correlations were shown with chromium, nickel and cadmium. However, as all the sediment parameters correlate with each other, the cause–effect relationships of single parameters cannot be determined by correlations alone. A regression analysis showed that the sediment parameters together had a very significant (P < 0.0005) effect on the level of DNA-damage. 3.6. Size of the mussels

Fig. 5. Estimated marginal means of tail moments expressed as TM’ (the fourth root of the tail moment) shown for the different locations.

An interesting observation was that the size of the mussels decreased gradually during both years of monitoring, starting with an average size for all locations of 38.9 mm at the first sampling in 1999 and ending

40

J. Rank et al. / Mutation Research 585 (2005) 33–42

with an average size of 22.5 mm at the last sampling in 2001. This indicates that the aquatic environment in Køge Bay during these years has changed in a way that gives the mussels very bad conditions for growth. The reason for such a change could either be a decrease in available food or a general increase of non-genotoxic pollutants.

4. Discussion Many types of mussel cell have been used for the detection of DNA damage in the comet assay. Mitchelmore et al. [16] and Large et al. [17] used digestive gland cells, Steinert et al. [11] and Pavlica et al. [10] successfully used haemocytes in their studies, and Wilson et al. [13] obtained good results with gill cells. Moreover, studies using digestive gland cells have shown positive results with benzo(a)pyrene, leading to the conclusion that these cells can be considered able to transform this chemical to the diolepoxide thought to be responsible for the DNA damage [18]. Results of the present study are similar to those of a study by Mitchelmore et al. [19], who found higher levels of DNA damage in gill cells compared with haemolymph and digestive gland cells from the mussel Tapes semidecussatus, after chronic exposure to polluted sediments. Therefore, we recommend using gill cells in monitoring studies. Seasonal variation in DNA damage could simply be due to effects of temperature as shown in a study by Buschini et al. [20], where they kept zebra mussels, D. polymorpha, at four different temperatures. Analyses of haemolymph cells in the comet assay showed a clear positive correlation of TM with temperature. Shaw et al. [21] showed seasonal variations of DNA damage in indigenous mussels, M. edulis. Mussels from a reference site and also mussels transplanted from the reference site to a polluted site had a 1.6-fold higher DNA damage (comet assay) in August than in December. Moreover, a monitoring study from May to September on populations of Paris guadrifolia in Lithuania and Norway showed that the highest level of chromosome aberrations in the root-tip cells was found in June and July [22]. The influence of sunlight could be an important explanation of seasonal variation. In a study by Steinert et al. [23], levels of DNA single strand breaks in blue mussels were higher for sunlight-exposed

mussels compared with mussels protected from sunlight. It has not been possible to obtain knowledge of the seasonal variation of genotoxic chemicals in the wastewater outlets. Chemical analysis of genotoxic substances in outlets from industries and municipalities is very difficult because large numbers of genotoxic chemicals are commonly used and discharged into the marine environment. Because it is impossible to measure all types of chemical in the marine environment, it is important to use other indicators of the pollution level. Distance from wastewater outlets could be a good indicator, at least in rivers [10]. In the present study, the measurements of DNA damage showed that this was not the case for coastal waters. We found that DNA damage was high at places where we expected to find low levels. However, as the sediment concentrations of three out of four heavy metals showed a positive correlation with DNA damage we propose to use some of these as indicators of genotoxic pollution. Primary sedimentation or movements of sediments are factors that can influence the pollution levels and are difficult to predict. In shallow coastal waters, as used in the present study, weather conditions such as a storm can cause turbulence of sediments, and streams can transport the pollutants from their outlet to locations far away, where sedimentation can take place. Sediment sampled 3 days after one of the most powerful storms in Denmark (3 December 1999) showed a dramatic increase in organic matter and mercury concentrations (Table 1) for the site Bøgeskov, indicating that this area has a high degree of sedimentation. Because many chemicals bind to sedimentary organic particles, events such as the above-mentioned December storm can influence the pollution levels and thereby the genotoxic exposure of the mussels. Knowledge of baseline DNA damage is also very important. However, very few coastal areas are without anthropogenic impact, and sediments and biota are often contaminated with chemicals to some degree. In our study, we were able to find coastal areas without industrial pollution resulting in very low levels of DNA damage in the mussels. The age of the mussels can also influence DNA damage. Levels of DNA strand breaks in blood cells of dab (Limanda limanda) detected with the comet assay were significantly higher in adults than in juvenile fish [24]. The difference was explained by a higher capacity of

J. Rank et al. / Mutation Research 585 (2005) 33–42

biotransformation in the adults than in the juveniles. In our study, we found a correlation between mussel size and DNA strand breaks, which could be due to a higher capacity of biotransformation in the larger mussels compared with the smaller ones. However, this hypothesis needs further examination of the chemicals present in the environment because not all compounds need bio-activation to become genotoxic. To find causality between DNA damage in indigenous mussels and specific chemicals is very difficult. Even if there is a significant correlation, as we found for chromium, nickel and cadmium, it could be a coincidence. Taking into consideration that many chemicals are present in both water and sediments and that they can interact with the DNA in many different ways, it is a very complex scenario to analyse. However, Steinert et al. [11] used the comet assay on cells from mussels transplanted from a reference site to polluted locations and found significant correlations between DNA damage and the pollution levels of heavy metals, PAHs, PCBs and chlorinated pesticides in both sediments and mussel tissue. It was also shown that DNA damage in mussel cells responded rapidly to the contaminants. The conclusion of the present study is that screening of DNA damage in marine mussels from locations suspected to be polluted seems to be useful for a general classification of genotoxic marine sites. However, further work is needed to understand the dynamics of DNA damage in marine organisms.

Acknowledgments The authors wish to thank Lykke Enøe for assistance with the chemical analysis of the heavy metals and JanOle Nielsen for help with the fieldwork. The Danish Natural Science Research Council supported the work.

References [1] F.E. W¨urgler, P.G.N. Kramers, Environmental effects of genotoxins (eco-genotoxicology), Mutagenesis 7 (1992) 321–327. [2] D.R. Dixon, J.T. Wilson, Genetics and marine pollution, Hydrobiologia 420 (2000) 29–43. [3] J. Mersch, M.-N. Beauvais, The micronucleus assay in the zebra mussels, Dreissena polymorpha, to in situ monitor genotoxicity in freshwater environments, Mutat. Res. 393 (1997) 141–149.

41

[4] A. Herbert, R.K. Zahn, Monitoring DNA damage in Mytilus galloprovincialis and other aquatic animals, Z. Angew. Zool. 1 (1990) 13–33. [5] M. Vukmirovi´c, N. Bihari, R.K. Zahn, W.E.G. M¨uller, R. Batel, DNA damage in marine mussel Mytilus galloprovincialis as a biomarker of environmental contamination, Mar. Ecol. Prog. Ser. 109 (1994) 165–171. [6] C. Bolognesi, R. Rabboni, P. Roggieri, Genotoxicity biomarkers in M. Galloprovincialis as indicators of marine pollutants, Comp. Biochem. Physiol. 2 (1996) 319–323. [7] L. Xu, G.J. Zheng, P.K.S. Lam, B.J. Richardson, Relationship between tissue concentrations of polycyclic aromatic hydrocarbons and DNA adducts in green-lipped mussels (Perna viridis), Ecotoxicology 8 (1999) 73–82. [8] G. Ericson, H. Skarph´edinsd´ottir, L.D. Zuanna, J. Svavarsson, DNA adducts as indicators of genotoxic exposure in the indigenous and transplanted mussels, Mytilus edulis L. from Icelandic coastal sites, Mutat. Res. 516 (2002) 91–99. [9] N.P. Singh, M.T. McCoy, R.R. Tice, E.L. Schneider, A simple technique for quantification of low levels of DNA damage in individual cells, Exp. Cell. Res. 175 (1988) 184–191. [10] M. Pavlica, G.I.V. Klobuˇcar, N. Mojaˇs, R. Erben, D. Papeˇs, Detection of DNA damage in haemocytes of zebra mussel using comet assay, Mutat. Res. 490 (2001) 209–214. [11] S.A. Steinert, R. Streib-Montee, J.M. Leather, D.B. Chadwick, DNA damage in mussels at sites in San Diego Bay, Mutat. Res. 399 (1998) 65–85. [12] J. Rank, K. Jensen, Comet assay on gill cells and haemocytes from the blue mussel, Mytilus edulis, Ecotox. Environ. Safe. 54 (2002) 323–329. [13] J.T. Wilson, P.L. Pascoe, J.M. Parry, D.R. Dixon, Evaluation of the comet assay as a method for the detection of DNA damage in the cells of a marine invertebrate, Mytilus edulis L. (Molusca: Pelecypoda), Mutat. Res. 399 (1998) 87–95. [14] R.J. Albertini, D. Anderson, G.R. Douglas, L. Hagmar, K. Hemminki, F. Merlo, A.T. Natarejan, H. Norppa, D.E.G. Shuker, R. Tice, M.D. Waters, A. Aitio, IPCS guidelines for the monitoring of genotoxic effects of carcinogens in humans, Mutat. Res. 463 (2000) 111–172. [15] E. Bauer, R.-D. Recknagel, U. Fiedler, L. Wollweber, C. Bock, K.O. Greulich, The distribution of the tail moments in single-cell gel electrophoresis (comet assay) obeys a chisquare (␹2 ) not a gaussian distribution, Mutat. Res. 398 (1998) 101–110. [16] C.L. Mitchelmore, C. Birmelin, D.R. Livingstone, J.K. Chipman, Detection of DNA strand breaks in isolated mussel (Mytilus edulis L.) digestive gland cells using the “Comet” assay, Ecotoxicol. Environ. Safe. 41 (1998) 51–58. [17] A.T. Large, J.P. Shaw, L.D. Peters, A.D. McIntosh, L. Webster, A. Mally, J.K. Chipman, Different levels of mussel (Mytilus edulis) DNA strand breaks following chronic field and acute laboratory exposure to polycyclic aromatic hydrocarbons, Mar. Environ. Res. 54 (2002) 493–497. [18] B.M. Coughlan, M.G.J. Hartl, S.J. O’Reilly, D. Sheehan, C. Morthersill, F.N.A.M. van Pelt, J. O’Halloran, N.M. O’Brien, Detecting genotoxicity using the Comet assay following chronic exposure of Manila clam Tapes semidecussatus to polluted

42

J. Rank et al. / Mutation Research 585 (2005) 33–42

estuarine sediments, Mar. Pollut. Bull. 44 (2002) 1359– 1365. [19] C.L. Mitchelmore, C. Birmelin, J.K. Chipman, D.R. Livingstone, Evidence for cytochrome P-450 catalysis and free radical involvement in the production of DNA strand breaks by benzo(a)pyrene and nitroaromatics in mussel (Mytilus edulis L.) digestive gland cells, Aquat. Toxicol. 41 (1998) 193–212. [20] A. Buschini, P. Carboni, A. Martino, P. Poli, C. Rossi, Effects of temperature on baseline and genotoxicant-induced DNA damage in haemocytes of Dreissena polymorpha, Mutat. Res. 537 (2003) 81–92. [21] J.P. Shaw, A.T. Large, J.K. Chipman, D.R. Livingstone, L.D. Peters, Seasonal variation in mussel Mytilus edulis digestive gland cytochrome P4501A- and 2E-immunoidentified protein

levels and DNA strand breaks (Comet assay), Mar. Environ. Res. 50 (2000) 405–409. [22] J.R. Lazutka, A. Stapulionyt´e, D.K. Bjerketvedt, A. Odland, Seasonal variation in the frequency of abnormal anaphases and mitotic index values in wild populations of herb-Paris (Paris quadrifolia L., Trilliacea): implications for genetic monitoring, Mutat. Res. 534 (2003) 113–122. [23] S.A. Steinert, R. Streib-Montee, M.P. Sastre, Influence of sunlight on DNA damage in mussels exposed to polycyclic aromatic hydrocarbons, Mar. Environ. Res. 46 (1998) 55–358. [24] F. Akcha, F.V. Hubert, A. Pfhol-Leszkowics, Potential value of the comet assay and DNA adduct measurement in dab (Limanda limanda) for assessment of in situ exposure to genotoxic compounds, Mutat. Res. 534 (2003) 21–32.