The Science of the Total Environment 308 (2003) 221–234

0048-9697/03/$ - see front matter � 2002 Elsevier Science B.V. All rights reserved.PII: S0048-9697Ž02.00616-2

Biomarker responses and PAH uptake in Mya truncata followingexposure to oil-contaminated sediment in an Arctic fjord

(Svalbard)

L. Camus *, S.R. Birkely , M.B. Jones , J.F. Børseth , B.E. Grøsvik , B. Gulliksen ,a,b, c,d b e e d,c

O.J. Lønne , F. Regoli , M.H. Depledgec f b

Akvamiljø as, Mekjarvik 12, N-4070 Randaberg, Norwaya

School of Biological Sciences (Plymouth Environmental Research Centre), University of Plymouth, Plymouth, Devon, PL4 8AA,b

UKThe University Courses on Svalbard, PB 156, N-9170 Longyearbyen, Norwayc

Norwegian College of Fishery Sciences, University of Tromsø, N-9037 Tromsø, Norwayd

RF-Rogaland Research, P.O. Box 2503, Ullandhaug, N-4004 Stavanger, Norwaye

Institute of Biology and Genetic, University of Ancona, Via Ranieri, Monte d’Ago, I-60100 Ancona, Italyf

Received 5 February 2002; accepted 7 November 2002

Abstract

Expanding industrial activity (notably oil and gas exploration) in the Arctic requires assessment of the potentialimpact of chemicals on marine organisms living in seawater at low temperature. The bivalve Mya truncata is commonin Svalbard fjord (Norway) where it experiences low temperature throughout the year. To measure the impact ofpolycyclic aromatic hydrocarbons (PAH) on M. truncata, the responses of three biomarkers wtotal oxyradicalscavenging capacity-assay (TOSC), plasma membrane stability of haemocytes and respiration ratesx were investigatedfrom bivalves exposed to sediment contaminated with a PAH mixture (crude oil). After two weeks of exposure tothe contaminated sediment, TOSC showed no change. The high TOSC value (4010"1339 unit mg protein) ofy1

Mya truncata (control group) is thought to protect biomolecules with a low turnover rate efficiently in a low foodavailability environment. In the exposed bivalves, the haemocyte cellular membranes were significantly destabilisedcompared with controls (P-0.05). Respiration rate of control and PAH-exposed individuals (0.055"0.020 mg O2

dw h ) was similar and relatively low as is typical for polar bivalves, reflecting a strategy to minimise energyy1 y1

expenditure to cope with 9 months of starvation. Bioaccumulation of PAH by M. truncata was also low, due probablyto a combination of low metabolic rate and reduced solubility of the oil compounds at low temperature. Data indicatedan uptake of mainly low molecular weight compounds (two and three ring molecules). A good correlation oflogBAF (bioaccumulation factor) and logK (octanolywater partitioning coefficient) was shown (r s0.87).2

lipid ow

Tissue sensitivity andyor functional differences (digestive gland vs. haemocytes), PAH uptake route (dietary vs.

*Corresponding author. Tel: q47-79-023-300; fax: q47-79-023-301.E-mail address: Lionel.camus@unis.no (L. Camus).

222 L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

gills), the low metabolic rate of M. truncata and the low environmental temperature (reducing the bioavailability ofPAH) are factors that help explain these findings.� 2002 Elsevier Science B.V. All rights reserved.

Keywords: Biomarkers; Bioaccumulation; PAH; Sediment; Arctic; Mya truncata

1. Introduction

As human activities (notably oil and gas explo-ration) increase in the European Arctic, reliableenvironmental monitoring tools are required toidentify the potential impact of toxic chemicals onArctic biota (AMAP, 1998). Biomarkers wdefinedas ‘biological responses to a chemical or chemicalsthat give a measure of exposure or toxic effect’(Peakall, 1994)x are relevant measures of the insitu impact of oil discharges (Aas et al., 2001);however, most biomarkers have been developed intemperate regions characterised by high variationsof seawater temperature (i.e. from 10 to 20 8C)and little is known about their potential value forassessing pollution impacts for marine organismsliving in polar waters. The Arctic marine ecosys-tem is characterised by a combination of lowstable temperature, and a single, very brief, butwell-timed maximum of dense primary productionin May, followed by a rapid decline (Weslawskiet al., 1988). Polar marine invertebrates possessspecific biological adaptations to live under suchconditions. Respiration is low, presumably to copewith the low food supply and 9 months of starva-tion (Peck et al., 1997; Ahn and Shim, 1998;Clarke and Johnston, 1999), and antioxidantdefences are high compared with temperate spe-cies, presumably to cope with increased oxidativestress (Regoli et al., 2000). In addition, cellmembrane composition of Arctic species differsfrom temperate ectotherms to maintain fluidity atlow temperature (Gillis and Ballantyne, 1999).The physical properties of petroleum hydrocarbonsbehave differently at low temperature. Crude oildoes not dissolve to the same extent at lowcompared with higher temperatures. Such differ-ences in dissolution have important consequencesfor the bioavailability of oil components at differ-ent temperatures (Neff, 1979). Consequently, the

biological adaptation of cold-water organisms,together with the altered oil behaviour at lowtemperature, may affect the typical temperate-derived biomarker responses. Earlier studies haveattempted to test this hypothesis by looking at theeffects of poly aromatic hydrocarbons (PAH) onfish (Christiansen et al., 1996; Wolkers et al.,1996) and amphipods (Aunaas et al., 1991), butlittle similar work has been carried out with Arcticbivalves.

Because of their capacity to accumulate organiccompounds, filter-feeding bivalves have been usedextensively for monitoring contaminant levels andeffects in temperate water (Baumard et al., 1999;Lowe and Fossato, 2000). Therefore, the Arcticbivalve Mya truncata (Myidae), a filter feederliving buried in the sediment, was selected for thisstudy. It is distributed widely in Svalbard fjord(Koszteyn et al., 1991), and represents the mainsource of food of bearded seals (Erignathus bar-batus) and walrus (Odobaenus rosmarus) (Fisherand Stewart, 1997), the major nutritional compo-nents of the diet of the indigenous people of theNorth (AMAP, 1998). Previously, M. truncata hasbeen used as a monitoring species in the Arctic tomeasure petroleum hydrocarbon levels (Humphreyet al., 1987) following an experimental oil spill inthe Canadian Arctic but, unfortunately, no biolog-ical responses were measured (Mageau et al.,1987). Its capacity to accumulate chemicals thatcould be readily transferred to the highest levelsof the food web, including man, make M. truncataa key species for monitoring the potential impactof industrial activities in the Arctic.

This study reports a small-scale field exposureof the Arctic clam, Mya truncata, to PAH-contam-inated sediment in an Arctic fjord (Svalbard). Theaim of the work was to measure oxidative stressin a cold-water adapted bivalve exposed to PAH.Metabolic rate, antioxidant defences and impact of

223L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

reactive oxygen species (ROS) on cell membraneswere measured as these three biological featuresare closely linked (Sohal and Weindruch, 1996;Abele et al., 1998; Regoli, 2000; Winzer et al.,2001). This approach, consisting of responseslocated at different levels of biological function(biochemical, cellular and physiological) isrequired to gain a holistic understanding of theimpact of PAH to M. truncata (Depledge, 1994).

2. Material and methods

2.1. Field programme

In September 2000, the field experiment wasestablished at Isfjorden at Svalbard (788139N and158399E) (Fig. 1). Mya truncata were hand sam-pled by SCUBA divers near the Norwegian settle-ment of Longyearbyen at the outer part ofAdventfjorden (Fig. 1). The bivalves were storedfor 24 h at 2 8C in large plastic tanks filled withseawater with permanent aeration. Grab samplesof sediment were taken from the research vesselFyF Jan Mayen of the University of Tromsø at thesame location as the bivalves were sampled. Thefield site was at the mouth of Adventfjorden at 5m depth near the shore to provide easy access tothe divers. Seawater temperature was 6.5 8C duringestablishment and 5.5 8C when sampling wasperformed, 14 days later.

2.2. Experimental design

The sediment was divided in two equal weightseach of 40 kg and placed into two identical boxes(30 cm deep, 60 cm long and 30 cm large); onelabelled ‘control’ and the other ‘treatment’. Crudeoil from a North Sea oil field (0.4 l) was mixedwith 40 kg of sediment (10 ml of oil per kg ofsediment). Individuals of M. truncata were handburied into the sediment (20 individuals per box).Each box was lowered to the sea bottom andanchored at 5 m depth by divers. The treatmentbox was placed downstream of the control boxand 10 m apart to prevent any risks of contami-nation. To reduce predation (i.e. seals), a grid wasplaced on top of each box. A red buoy, connectedto the boxes by a rope, allowed easy relocation of

the site for latter sampling (Figs. 1 and 2). Animalswere left in the boxes for two weeks. Sedimentsamples were taken to determine levels of PAH atthe start and at the end to monitor oil sedimentconcentration in both boxes. To achieve this, sur-face sediment was sampled and placed in alumin-ium paper, and stored in the freezer at y80 8C.After 14 days, bivalves were removed from theboxes, stored in seawater and taken to the labora-tory for biomarker and PAH level analyses. Forthe TOSC assay, digestive glands were dissectedout, frozen in liquid nitrogen and stored at y808C prior to analysis. Haemolymph was sampledand cell membrane stability measured on livinghaemocytes. Respiration was measured and twoindividuals per group were frozen at y80 8C forPAH analysis.

3. Analyses

3.1. TOSC

The method was based on Winston et al. (1998),except that buffers were adjusted for marinebivalves (Regoli et al., 2000). Digestive glands ofM. truncata were homogenised with a potter infour volumes of 100 mM KH PO buffer, pH 7.5,2 4

2.5% NaCl. The homogenate was centrifuged at100 000=g for 1 h, and cytosolic fractions werealiquoted and stored at y80 8C. Peroxyl radicalsare generated by the thermal homolysis of 2-29-azo-bis-(2 methyl-propionamidine)-dihydrochlori-de (ABAP) at 35 8C. Peroxyl radical can oxidizethe substrate a-keto-g-methiolbutiryc acid(KMBA) to ethylene gas which is measured withgas chromatography. The optimal assay conditionswere 0.2 mM KMBA, 20 mM ABAP in 100 mMKH PO buffer, pH 7.4. Reactions were carried2 4

out in 10 ml rubber septa sealed vials in a finalvolume of 1 ml. Ethylene production was meas-ured by gas-chromatographic analysis of 200 mltaken from the head space of the reaction vials.Ethylene formation was monitored for 96 min witha Hewlett Packard (HP 5890 series II) gas chro-matograph equipped with a supelco SPB-1 capil-lary column (30 m=0.32 mm=0.25 mm) and aflamme ionisation detector (FID). The oven, injec-tion and FID temperatures were 35, 160 and 220

224 L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

Fig. 1. Position of the clam sampling (star) and caging sites (open circle) at Isfjorden, Svalbard (788139N, 158399E).

8C, respectively, helium was the carrier gas (1 mlymin flow rate) and a split ratio 20:1 was used.The data acquisition system was run by the soft-

ware Millenium32 (Waters). Each analysis�

required the measurement of control (no antioxi-dant in the reaction vial) and sample reactions

225L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

Fig. 2. Experimental set up deployed by scuba divers in Adventfjorden (Svalbard).

(biological fluid in the vial). In the presence ofantioxidant, ethylene production from KMBA wasreduced quantitatively and higher antioxidant con-centrations resulted in longer periods in whichethylene formation was inhibited relative to con-trols. By plotting the absolute value of the differ-ence between the ethylene peak area obtained ateach time point for the sample and control reactionit is possible to visualise whether the oxyradicalscavenging capacity of the solution is changed.The area under the kinetic curve was calculatedmathematically from the integral of the equationthat best defines the experimental points for boththe control and sample reactions. TOSC is thenquantified according to Eq. (1):

TOSCs100y(IntSAyIntCA*100) (1)

where IntSA and IntCA are the integrated areasfrom the curve defining the sample and controlreactions, respectively.

Thus, a sample that displays no oxyradicalscavenging capacity would give an area equal tothe control (IntSAyIntCAs1) and a resultingTOSCs0. On the other hand, as IntSAyIntCAgoes to 0 the hypothetical TOSC value approaches100. Because the area obtained with the sample isrelated to that of the control, the obtained TOSCvalues are not affected by small variations ininstrument sensitivity, reagents or other assay con-ditions. The specific TOSC value was calculatedby dividing the experimental TOSC by the con-centration of protein used for the assay. Data areexpressed as TOSC unit per mg protein.

3.2. Haemocyte cellular membrane stability

The method measures the stability of the cellmembranes of the haemocytes and was reportedby Grøsvik et al. (2000) and Camus et al. (2000).It has been optimised further by Camus et al.

226 L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

(2002). In summary, ethidium homodimer-1(EthD-1) is marketed through Molecular ProbesInc. as a probe to detect dead cells preferably incombination with the viability probe calcein AM.EthD-1 is taken up over the membranes andintercalates with DNA and undergoes approxi-mately a 40-fold enhancement upon binding. How-ever, EthD-1 is also taken up by healthy cells overtime, but to a lesser extent. Therefore, an increasein uptake of EthD-1 in a cell population of cellsfrom treated organisms compared to cells fromuntreated indicates an effect on cell membranestability (Grøsvik et al., 2000; Van Winkle et al.,1999; Jackson et al., 1997). Good correlation withthe neutral red retention assay developed by Loweet al. (1995) in Mytilus sp. was shown (Camus etal., 2000), indicating the reliability of this assayas a biomarker of membrane stability.

Haemolymph, sampled with a hypodermicsyringe from the adductor muscle of each individ-ual, was mixed with physiological saline solution(4.77 gyl Hepes, 25.48 gyl NaCl, 13.06 gylMgSO , 0.75 gyl KCl and 1.47 gyl CaCl , pH4 2

7.4) in a 1:1 ratio. 200 ml of cell suspension wasadded per well of a dark microplate and mixedwith 50 ml of Ethidium homodimer-1 (EthD-1)(Molecular Probes, Eugene, OR, USA). Four rep-licates (4 wells of the microplate) were used perindividual. Incubation lasted 5 min in the dark atroom temperature. Fluorescence of EthD-1 wasread at ex544yem612. Four wells of the microplatewere filled with 200 ml of saline solution and 50ml of Ethidium homodimer-1 (EthD-1), the fluo-rescence was measured and subtracted to the sam-ple fluorescence data for background fluorescencecorrection. No washing of the cells was performed.Membrane destabilisation data were normalised toprotein concentration of the haemolymph andexpressed as EthD-1 fluorescence unit per mgprotein.

3.3. Respiration

A closed respirometer was used in this study.Five bivalves were placed individually into glassvials (480 ml), sealed with a screwed Teflon lidand incubated for 4 h at 5.5 8C. Six empty vials,containing seawater but no bivalves, were incubat-

ed as controls and the oxygen concentration mon-itored. The oxygen concentration in each vial wasmeasured by injecting a seawater sample takenfrom the vials into a respiration chamber (100 ml)connected to a polarographic Clark-type oxygensensor (Strathkelvin microcell MC100� andmicrocathode oxygen electrode SI130) maintainedat 5.5 8C by a recirculating refrigeration bath. ThePO was calibrated to oxygen-saturated sea water2

(10.03 mg l O , 36‰ salinity). Soft body tissuey12

was dried at 100 8C for 24 h and respiration datawere normalised to dry weight body mass. Dataare expressed as mg O per gram dry weight per2

hour.

3.4. Protein determination

Total protein concentrations were determinedspectrophotometrically by the method of Bradford(1976).

3.5. Extraction of sediment and tissue samples forGCyMS-selective ion monitoring analysis

Whole body tissue of Mya truncata was cut intofine pieces by means of scissors until a homoge-nous sample was reached. The following methodwas applied to bivalve tissue and sediment. Eightdeuterated PAH as quantitation internal standards(QIS) covering the molecular size of interest(naphthalene-d8, phenanthrene-d10, dibenzothio-phene-d8, fluoranthene-d10, pyrene-d10, chrysene-d12, benzo(a)pyrene-d12 and dibenzo(a,h)anthracene-d14) were added to homogenised tissueor sediment (5–10 g) and boiled for 2 h in a 10%(wyv) solution of potassium hydroxide in metha-nol to achieve saponification. The digest wasfiltered and extracted repeatedly three times withcyclohexane. Cleanup of extracts was performedby solid phase extraction using 3 ml tubes contain-ing 0.5 g of normal phase packing (SupelcleanLC-Si, Supelco INC, Bellefonte, PA, USA). Fur-ther extraction was performed with a mixture ofdichloromethaneycyclohexane (1:3). Recoveryinternal standard (RIS) mixture was added andsamples were stored in capped vials until GCyMS-analysis.

227L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

3.6. GCyMS-selective ion monitoring analysis

PAH analysis of all samples was performed byGas Chromatography (HP5890, Hewlett Packard,USA) connected to a Mass Spectrometer (FinniganSSQ7000, USA) and analysed in selected ionmode (GCyMS-SIM). The GC was equipped witha CP-SIL 8CB fused silica column (Chrompack,50=0.25 mm i.d., film thickness 0.25 m). Injectorand detector temperatures were both 300 8C. Thecolumn was held at 50 8C for 1 min, ramped at25 8C min to 120 8C, 3 8C min to 320 8Cy1 y1

and held for 17 min at 320 8C. Helium was usedas carrier gas, with a flow rate of 0.6 ml min aty1

50 8C. The M. truncata tissue samples and sedi-ment were analysed for 26 parent PAH compounds(including dibenzothiophenes) and expressed,respectively, as mg kg lipid and mg kg wety1 y1

sediment. The sum of all analysed PAHs aredesignated total PAH (TPAH). Lipid concentra-tions were measured in samples according to themethod of (Folch et al., 1957).

3.7. Calculation and quality assurance of GCyMSanalyses

Calibration standards for all non-alkylated PAHcompounds were prepared in seven different con-centrations (10–2000 mg l ) and response factory1

curves were calculated. For calculation of alkylatedPAH, the response factor curves for the respectivenon-alkylated homologes were used. The reproduc-ibility of the response factors for three standardscovering the concentration range was checked foreach series of samples analysed, and the calculatedconcentrations were not allowed to exceed 20%deviation from the expected concentrations. RISand QIS were added to the calibration standardsin concentrations of approximately 500 mg l ofy1

the individual deuterated compounds. One proce-dure blank and one control sample were extractedfor every series of samples (ns12). The controlsused were either a certified reference material(SRM2974, NIST, Gaithersburg, MD, USA) or anunexposed sample spiked with an appropriateamount of a certified mixture of PAH (Dr Eher-enstorfer Reference Materials, Augsburg, Germa-ny). The recovery of PAH in the samples was

calculated by comparing the relative signal inten-sity of RIS and QIS in samples and calibrationstandards (defined to be 100%). Extraction recov-ery for 2-ring PAH compounds was at least 30%in sediment samples and at least 40% in tissuesamples. Extraction recovery for larger PAH com-pounds was at least 50% in both sediment andtissue samples.

3.8. Statistical analysis

The analyses were carried out with JMP, ver.(3.2.6), SAS Institute Inc., Cary, NC, USA. Statis-tical analyses were run after testing for normaldistribution and homogenous variances. ParametricDunnett’s test was used at a significance level ofP-0.05. Values of TOSC, EthD-1 fluorescenceand respiration were plotted as Box and Whiskerplots. These indicate the range of values and themedian. The horizontal bars that denote the upperand lower limits of the boxes include 95% of thedata points; and 50% of the data points areencompassed by the upper and lower limits of theboxes.

4. Results

4.1. TOSC

No significant difference (P)0.05) in TOSCvalues was measured in exposed and controlbivalves (Fig. 3). Interestingly, variability of thecontrol group was twice that of the exposed group.The specific mean TOSC value of M. truncata inthe control group was 4014"1339 unit mgy1

protein (mean"S.D.).

4.2. Membrane stability

A significant increase in uptake of EthD-1 intothe haemocytes of M. truncata was measured inthe exposed group (P-0.05) (Fig. 4). In theexposed group, the measured fluorescence wastwice that of the control. Thus, the cellular mem-branes of haemocytes of M. truncata exposed tooil demonstrate increased permeability to EthD-1,indicating destabilisation of the membranes.

228 L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

Fig. 3. Total Oxyradical Scavenging Capacity (TOSC) valuestowards peroxyl radicals in the digestive gland of Mya truncatafollowing two weeks of field deployment wcontrol (ns6);exposedsoil-contaminated sediment (ns6)x.

Fig. 4. Stability of the membrane of the haemocytes of Myatruncata measured as fluorescence units of Ethidium homodi-mer-1 (EthD-1) after two weeks exposure to control (ns7)and oil contaminated sediment (ns11). Significant effect isindicated by a star (P-0.05).

Fig. 5. Respiration of Mya truncata measured as uptake of mgof oxygen per g of dry weight body mass per hour (ns5)following two week exposure to control and oil contaminatedsediment.

4.3. Respiration

No significant differences were observed in theoxygen uptake of M. truncata in either group (Fig.5). Nevertheless, there was a decrease in themedian of the exposed group, indicating adepressed respiration in some individuals. Thebasal (September) oxygen uptake of M. truncataat 5.5 8C was 0.055"0.020 mg O g dw hy1 y12

(mean"S.D.).

4.4. PAH in sediment

The total concentration of PAH in the oil-contaminated sediment at day 0 was 940.43 mgg wet sediment; two and three ring moleculesy1

represent 83 and 16% of TPAH, respectively. Aftertwo weeks at sea, the remaining total PAH amountin the surface sediment of the exposed group was431.67 mg g wet sediment. The four and fivey1

ring molecules remained adsorbed in the sedimentafter two weeks while the concentrations of thetwo and three ring compounds decreased by 50%(Fig. 6).

4.5. PAH uptake in M. truncata

The total amount of PAH in the clam tissue(200.08 mg g lipid) was ten times higher thany1

control clams after two weeks of exposure tocontaminated sediment. Nevertheless, the PAH lev-el in the clam tissue was two times less than theconcentration of oil in the sediment after twoweeks. Only two and three ring molecules were

229L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

Fig. 6. Concentration of PAHs according to the number of rings in sediment (SCssediment control; SOssediment with oil; valuesare in mg g wet sediment) and in tissue of M. truncata (MCsMya control; MOsMya oil exposed; values are in mg g lipid)y1 y1

after a two week exposure.

detected (Fig. 6) in the exposed M. truncata; 93%of TPAH were two ring compounds.

4.6. Relationships between bioaccumulation factorand logKow

The bioaccumulation factors (BAFs; ratio oilcompounds in bivalve tissue and compound levelsin sediment) were calculated on the lipid weightof the different samples. The BAFs were lowerthan 0.6, indicating that M. truncata poorly accu-mulated PAH from the sediment. As large mole-cules (five to six rings) were not detected in thepresent experiment, they were not included in thecalculations when logBAF was plotted againstlogK (hydrophobycity of PAH expressed by theow

octanolywater partition coefficients), a significantpolynomial of degree two relationship was found

(r s0.87, P-0.05), indicating that lipid-based2

bioaccumulation factors increase with hydropho-bicity and reach a plateau for logK above 4.5ow

(Fig. 7).

5. Discussion

The metabolic rate of M. truncata falls withinthe range measured for other Arctic bivalves. Forexample, 50–57 mg O g dw h was the rangey1 y1

2

reported for Clinocardium ciliatum at 0 8C(Schmid, 1996), and between 63.7 and 94 mg O2

g dw h for Chlamys islandicus (0.5 8C) andy1 y1

Hiatella arctica (3.5 8C) (Camus, 2002). Themetabolic rate of Arctic bivalves is low comparedwith temperate equivalents (Dame, 1972; Vahl,1973; Bayne and Scullard, 1977; Navarro andWinter, 1982). The oxygen consumption rate of

230 L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

Fig. 7. Relationship between the bioaccumulation factor (BAFlipid) and the octanolywater partition coefficient (K ) plotted on aow

log–log scale in Mya truncata exposed for two weeks to crude oil contaminated sediment. The adjustement (polynomial of degree2) was significant at the threshold level of P-0.05.

M. truncata (Arctic) at 5 8C was four times lowerthan that for M. arenaria (temperate) at 10 8C(Camus, 2002). Reduced metabolism of polarbivalves is considered an important mechanism forenergy conservation in cold water where food isin short supply (Clarke, 1991; Peck et al., 1997,2000; Ahn and Shim, 1998). Such low metabolicrates enable Arctic organisms to survive longperiods of no food supply and facilitate the transferof more energy into growth when food is availableduring the brief phytoplankton bloom.

In suspension feeding bivalves, the main routeof PAH uptake is via filtering activity and, there-fore, uptake of PAH from sediment occurs onlyafter the chemical has desorbed into interstitial oroverlying water (Foster et al., 1987; Narbonne etal., 1999; Baumard et al., 1999). As observed byNarbonne et al. (1999), at the end of the presentexperiment, 50% of the initial sediment burden oftwo and three ring hydrocarbon compounds weredesorbed and transferred to the water compartment.Chemical tissue analysis revealed that some ofthese compounds were bioaccumulated by M. trun-cata. The logarithmic shape of the logBAF vs.lipid

logK curve confirms that uptake of the PAHow

molecules is based upon their degree of lipophil-

icity and size. The shape of the curve, similar tothe one reported for Mytilus edulis and for semi-permeable membranes in general, is typical oforganisms that have a poor capability to metabolisePAH (Baussant et al., 2001).

No difference in oxygen uptake was shownbetween exposed and non-exposed M. truncata,indicating that respiration rate was not affected byoil exposure. Exposure to PAH increases the ratesof oxygen consumption of other bivalves speciessuch as Venus verrucosa (Axiak and George,1987), Mytilus edulis (Widdows et al., 1982), Myaarenaria (Fong, 1976; Stainken, 1978) and Maco-ma balthica (Steckoll et al., 1980). It has beensuggested that this increased basal metabolic rateis due to inhibition of membrane-bound respiratoryenzymes (Steckoll et al., 1980). Hence, the PAHuptake did not affect the respiratory processes inM. truncata.

The basal TOSC value measured in M. truncata(4000 unit mg protein) is elevated comparedy1

with temperate species and this appears to be ageneral trend for polar bivalve species (e.g. 2800unit mg protein for Chlamys islandica and 2000y1

for Hiatella arctica; Camus et al., 2002; Camus,2002; Regoli et al., 2000). Regoli and Winston

231L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

(1998) reported 600 unit mg protein in they1

temperate mussel Mytilus edulis and 700 for thetemperate starfish Leptasterias epichlora. It hasbeen suggested that this higher capacity to scav-enge ROS in Arctic species supports the view thatthe marine ecosystem is highly prooxidant (Abeleet al., 1998; Colella et al., 2000; Regoli et al.,2000) due to increase in UV-B radiation associatedto the depletion of the ozone hole (Yocis et al.,2000; Qian et al., 2001). Recently, however,Camus (2002) hypothesised that polar marineorganisms require a high TOSC to protect efficient-ly biomolecules, characterised by a low turnoverrate (Whiteley et al., 1996), from natural oxidationin an environment where food availability is low.

Considering the uptake of PAH and that metab-olism of PAH enhances the production of ROS(Livingstone et al., 1990; Winston and Di Giulio,1991), a change in TOSC value was expected inexposed bivalves. This change has been measuredby others (Regoli, 2000; Winzer et al., 2001;Camus et al., 2002) but was not observed in thepresent study. The low metabolic rate of M. trun-cata, measured in this study, may result in a lowROS production caused by the metabolism of PAH.Furthermore, the logarithmic shape of thelogBAF vs. logK indicates a low PAH meta-lipid ow

bolic capability of M. truncata. The high level ofantioxidant defence measured in this experimentprovides M. truncata with the capability to dealwith moderate variations in ROS generation result-ing in no difference of TOSC between treatments.Depledge and Lundebye (1996) emphasised theneed to consider the variability of a biomarkerresponse to provide additional useful insight intothe chemical impact of pollutant. TOSC valuevariability was reduced in the exposed group. Thedecrease in variability of the exposed group mayreflect a homogeneous induction of the antioxida-tive system in each individual.

Oxidative damages occur when ROS productionexceeds the total ROS scavenging capacity of anorganism (Winston and Di Giulio, 1991). Delete-rious consequences can arise and, one of these islipid membrane peroxidation (Viarengo et al.,1989; Regoli, 2000; Winzer et al., 2001). Theobservation of the destabilised haemocyte mem-branes in the present study is supported by other

demonstrations of haemocytes being the main siteof defence against foreign substances in marinebivalves (Lowe et al., 1995; Carballal et al., 1997;Dyrynda et al., 1998). Using the neutral redretention method, Lowe et al. (1995) providedclear evidence of contaminant-induced lysosomalmembrane damage in the haemocytes of Mytilusgalloprovincialis. Lysosomes, subcellular organ-elles that contain various enzymes (i.e. acid hydro-lases), are sites of intense ROS production(Winston et al., 1996); the latter are claimed to bea cytotoxic mechanism against invaders and for-eign compounds (Pipe, 1992; Van Der Knaap etal., 1993). Another role of lysosomes is to accu-mulate contaminants that damage cells. The pro-duction of ROS can compromise the functionalintegrity of the lysosomal membrane followinglipid peroxidation (Viarengo et al., 1989; Winstonet al., 1996) resulting in release of acid hydrolasesin the cytosol and further damage and disruptionto the cell. Furthermore, under conditions of per-sistent xenobiotic challenge, oxidant stress mightfar exceed the normal antioxidant defences inbivalve haemocytes thereby, not preventing lipidperoxidation (Winston et al., 1996). The fluores-cent assay used in the present study showed goodcorrelation with the neutral red retention methodin haemocytes of Mytilus edulis (Camus et al.,2000), revealing that the destabilised plasmamembrane of the haemocytes, measured in exposedM. truncata, could be a consequence of decreasedlysosomal membrane stability.

Previous studies have investigated the relation-ship between oxyradical toxicity and membranealterations. Reduced capability to scavenge oxy-radicals in mussels (Mytilus galloprovincialis) wasassociated with destabilisation of lysosomal mem-branes (Regoli, 2000) and similar findings werereported in the Arctic scallop Chlamys islandicus(Camus et al., 2002). Increased lipid peroxidationwas measured in flounder hepatocytes followingoxidative stress while TOSC increased (Winzer etal., 2001). In M. truncata, antioxidant defences ofdigestive gland tissue were not overwhelmed byROS production (as indicated by the stable TOSCvalue), nevertheless, destabilised haemocyte cel-lular membranes were measured. This lack ofcorrelation may reflect functional differences andy

232 L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

or different PAH route of uptake. Dissolved PAHsare the main compounds that were taken up by M.truncata, therefore, these PAH molecules enteredthe haemolymph system via the gills and impacteddirectly the haemocytes. Digestive gland biochem-ical properties are affected mainly by the dietaryuptake of PAH associated with ingested foodparticles. The link between oxyradical toxicity andmembrane destabilisation, and our data suggeststhat the tissue sensitivity andyor functional differ-ences have to be taken into consideration to showevidence of a relationship between TOSC andmembrane destabilisation. Viarengo et al. (1994,1995) suggested that the low level of unsaturatedlipids observed in polar bivalves confer a betterprotection to the membranes against oxidativestress. Nevertheless, in this experiment, haemocytecell membranes were the most sensitive tissue ofMya truncata to PAH exposure.

6. Conclusions

The study showed that PAHs were taken up byM. truncata and resulted in destabilisation of thehaemocyte membranes but did not affect the totaloxyradical scavenging capacity of the digestivegland. Tissue sensitivity and functional differences(digestive gland vs. haemocytes), PAH uptakeroute (dietary vs. gills), the typical low metabolicrate of M. truncata and low temperature reducingthe bioavailability of PAH, are factors that explainthese findings. Impaired function of the cellularhaemocyte membranes can have consequences onthe phagocytosis of micro-organisms (Grundy etal., 1996). Given the important role of haemocytesin cellular immunity of bivalves (Dyrynda et al.,1998), PAH-induced cell injury will contribute toa loss of immune protection and finally to carci-nogenesis in M. truncata as observed in M. aren-aria (Van Beneden, 1997).

Acknowledgments

We thank the crew of the research vessel of theUniversity of Tromsø, FyF Jan Mayen and RuppertKrapp (University of Kiel) for helping with thesampling. This research was funded by a MarieCurie grant through the European programme

Training and Mobility of Researchers contract noERBFMBICT983185.

References

Aas E, Beyer J, Jonsson G, Reichert WL, Andersen OK.Evidence of uptake, biotransformation and DNA binding ofpolyaromatic hydrocarbons in Atlantic cod and corkwingwrasse caught in the vicinity of an aluminium works. MarEnviron Res 2001;52:213 –229.

Abele D, Burlando B, Viarengo A, Portner HO. Exposure to¨elevated temperatures and hydrogen peroxide elicits oxida-tive stress and antioxidant response in the Antarctic intertidallimpet Nacella concinna. Comp Biochem Physiol1998;120B(2):425 –435.

Ahn IY, Shim JH. Summer metabolism of the Antarctic clam,Laternula elliptica (King and Broderip) in Maxwell Bay,King George Island and its implications. J Exp Mar BiolEcol 1998;224(2):253 –264.

AMAP (1998), AMAP Assessment Report: Arctic PollutionIssues. Arctic Monitoring and Assessment Programme(AMAP), Oslo, Norway. xiiq 859 pp.

Aunaas T, Olsen A, Zachariassen KE. The effects of oil andoil dispersants on the amphipod Gammarus oceanicus fromArctic waters. Polar Res 1991;10(2):619 –630.

Axiak V, George J. Bioenergetic responses of the marinebivalve Venus verrucosa on long-term exposure to petroleumhydrocarbons. Mar Environ Res 1987;23:33 –47.

Baumard P, Budzinski H, Garrigues P, Narbonne JF, BurgeotT, Michel X, Bellocq J. Polycyclic hydrocarbons (PAH)burden of mussels (Mytilus sp.) in different marine environ-ment in relation with sediment PAH contamination, andbioavailability. Mar Environ Res 1999;47:415 –439.

Baussant T, Sanni S, Jonsson G, Skadsheim A, Børseth JF.Bioaccumulation of polycyclic aromatic compounds: 1. Bio-concentration in two marine species and in semipermeablemembrane devices during chronic exposure to dispersedcrude oil. Environ Toxicol Chem 2001;20(6):1175 –1184.

Bayne BL, Scullard C. Rates of nitrogen excretion by speciesof Mytilus (Bivalvia: Mollusca). J Marine Biol Assoc UK1977;57:355 –369.

Bradford MM. A rapid and sensitive method for the quantita-tion of microgram quantities of protein utilising the principleof protein-dye binding. Anal Biochem 1976;72:248 –254.

Camus, L. (2002), Biomarkers relevant to oil and gas industrialactivities in low temperature marine ecosystems. PhD thesis.University of Plymouth, 240 p.

Camus L, Grøsvik BE, Børseth JF, Depledge MH. Stability oflysosomal and cell membranes in haemocytes of the com-mon mussel (Mytilus edulis): effect of low temperatures.Mar Environ Res 2000;50:325 –329.

Camus L, Jones MB, Borseth JF, Grosvik BE, Regoli F,Depledge MH. Total oxyradical scavenging capacity andcell membrane stability of haemocytes of the Arctic scallop,Chlamys islandicus, following benzo(a)pyrene exposure.Mar Environ Res 2002;54:425 –430.

233L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

Carballal MJ, Lopez C, Azevedo C, Villalba A. Enzymesinvolved in defence functions of hemocytes of musselMytilus galloprovincialis. J Invertebrate Pathol 1997;70(2):96 –105.

Christiansen JS, Dalmo RA, Ingebrigtsen K. Xenobiotic excre-tion in fish with aglomerular kidneys. Mar Ecol Prog Ser1996;136(1–3):303 –304.

Clarke A. What is cold adaptation and how should we measureit? Am Zool 1991;31(1):81 –92.

Clarke A, Johnston NM. Scaling of metabolic rate with bodymass and temperature in teleost fish. J Anim Ecol 1999;68(5):893 –905.

Colella A, Patamia M, Galtieri A, Giardina B. Cold adaptationand oxidative metabolism of Antarctic fish. Ital J Zool2000;67:33 –36.

Dame RF. The ecological energetics of growth, respiration andassimilation in the intertidal American oyster Crassostreavirginica. Mar Biol 1972;17:243 –250.

Depledge MH. The rational basis for the use of biomarkers asecotoxicological tools. In: Fossi MC, Leonzio C, editors.Non-destructive biomarkers in vertebrates. Boca Raton,Florida: Lewis Publishers, 1994. p. 271 –295.

Depledge MH, Lundebye AK. Physiological monitoring ofcontaminant effects in the individual rock crabs, Hemigrap-sus Edwardsi: the ecotoxicological significance of variabilityin response. Comp Biochem Physiol 1996;113C(2):277 –282.

Dyrynda EA, Pipe RK, Burt GR, Ratcliffe NA. Modulationsin the immune defences of mussels (Mytilus edulis) fromcontaminated sites in the UK. Aquat Toxicol 1998;42(3):169 –185.

Fisher KI, Stewart REA. Summer foods of Atlantic walrus,Odobenus rosmarus rosmarus, in northern Foxe basin,northwest territories. Can J Zool 1997;75(7):1166 –1175.

Folch J, Lees M, Stanley GHS. A simple method for theisolation and purification of total lipids from animal tissue.J Biol Chem 1957;226:497 –509.

Fong WC. Uptake and retention of Kuwait crude oil and itsefects on oxygen uptake by the soft-shell clam, Mya aren-aria. J Fish Res Board Can 1976;33:2774 –2780.

Foster GD, Baksi SM, Means JC. Bioaccumulation of traceorganic contaminants from sediment by baltic clams (Maco-ma balthica) and soft-shell clams (Mya arenaria). EnvironToxicol Chem 1987;6(12):969 –976.

Gillis TE, Ballantyne JS. Mitochondrial membrane composi-tion of two arctic marine bivalve mollusks, Serripes groen-landicus and Mya truncata. Lipids 1999;34(1):53 –57.

Grundy MM, Moore MN, Howell SM, Ratcliffe NA. Phago-cytic reduction and effects on lysosomal membranes bypolycyclic aromatic hydrocarbons, in haemocytes of Mytilusedulis. Aquat Toxicol 1996;34(4):273 –290.

Grøsvik BE, Bjørnstad A, Camus L, Børseth JF, AndersenOK. Membrane destabilisation of haemocytes measured byuptake of fluorescent probes. Mar Environ Res 2000;50:432 –433.

Humphrey B, Boehm PD, Hamilton MC. The fate of chemi-cally dispersed and untreated crude oil in Arctic benthicbiota. Arctic 40 suppl 1987;1:149 –161.

Jackson WF, Huebner JM, Rusch NJ. Enzymatic isolation andcharacterisation of single vascular smooth muscle cells fromcremasteric arterioles. Microcirc Lond 1997;4(1):35 –50.

Koszteyn J, Kwasniewski S, Rozycki O, Weslawski JM. Atlasof the marine fauna of Southern Spitsbergen. In: KlekowskiRZ, Weslawski JM, editors. Invertebrates Part 1, vol. 2PolishAcademy of Sciences Institute of Ecology, Institute ofOceanology, 1991.

Livingstone DR, Martinez PG, Michel X, Narbonne JF, OharaS, Ribera D, Winston GW. Oxyradical production as apollution-mediated mechanism of toxicity in the commonmussel, Mytilus edulis (L.), and other mollusks. Funct Ecol1990;4(3):415 –424.

Lowe DM, Fossato VU. The influence of environmentalcontaminants on lysosomal activity in the digestive cells ofmussels (Mytilus galloprovincialis) from the VeniceLagoon. Aquat Toxicol 2000;48:75 –85.

Lowe DM, Fossato VU, Depledge MH. Contaminant-inducedlysosomal membrane damage in blood cells of musselsMytilus galloprovincialis from the Venice Lagoon: an invitro study. Mar Ecol Prog Ser 1995;129:189 –196.

Mageau C, Engelhardt FR, Gilfillan ES, Boehm PD. Effectsof short-term exposure to dispersed oil in Arctic inverte-brates. Arct 40 suppl 1987;1:162 –171.

Narbonne JF, Djomo JE, Ribera D, Ferrier V, Garrigues P.Accumulation kinetics of polycyclic aromatic hydrocarbonsadsorbed to sediments by the mollusk Corbicula fluminea.Ecotoxicol Environ Saf 1999;42:1 –8.

Navarro JM, Winter JE. Ingestion rate, assimilation efficiencyand energy balance in Mytilus chilensis in relation to bodysize and different algal concentrations. Mar Biol 1982;67:255 –266.

Neff JM. Biological effects of PAH in the aquatic environment.London: Applied Science Publishers Ltd, 1979.

Peakall DB. the role of biomarkers in environmental assess-ment.1. Introduction. Ecotoxicology 1994;3(3):157 –160.

Peck LS, Brockington S, Brey TA. Growth and metabolism inthe Antarctic brachiopod Liothyrella uva. PhilosopicalTransactions of the Royal Society of London Series B1997;352(1355):851 –858.

Peck LS, Colman JG, Murray AWA. Growth and tissue masscycles in the infaunal bivalve Yoldia eightsi at Signy Island,Antarctica. Polar Biol 2000;23(6):420 –428.

Pipe RK. Generation of reactive oxygen metabolites by thehaemocytes of the mussel Mytilus edulis. Dev Comp Immu-nol 1992;16:111 –122.

Qian JG, Mopper K, Kieber DJ. Photochemical production ofthe hydroxyl radical in Antarctic waters. Deep-Sea Res PartI-Oceanographic Res Pap 2001;48(3):741 –759.

Regoli F. Total oxyradical scavenging capacity (TOSC) inpolluted and translocated mussels: a predictive biomarkerof oxidative stress. Aquat Toxicol 2000;50(4):351 –361.

Regoli F, Nigro M, Bompadre S, Winston GW. Total oxidantscavenging capacity (TOSC) of microsomal and cytosolic

234 L. Camus et al. / The Science of the Total Environment 308 (2003) 221–234

fractions from Antarctic, Arctic and Mediterranean scallops:differentiation between three potent oxidants. Aquat Toxicol2000;49(1–2):13 –25.

Regoli F, Winston GW. Applications of a new method formeasuring the total oxyradical scavenging capacity inmarine invertebrates. Mar Environ Res 1998;46(1–5):439 –442.

Schmid MK. In: Rieman F, editor. On the distribution andoxygen consumption of ecologically important benthic ani-mals in the waters around Svalbard (Arctic). Berichte ZurPolarforschung. Bremerhaven: Alfred-Wegner-Institut, 1996.

Sohal SR, Weindruch R. Oxidative stress, caloric restrictionand aging. Science 1996;273:59 –63.

Stainken DM. Effects of uptake and discharge of petroleumhydrocarbons on the respiration of the soft-shell clam Myaarenaria. J Fish Res Board Can 1978;35:637 –642.

Steckoll MS, Clement LE, Shaw DG. Sublethal effects ofchronic oil exposure on the intertidal clam Macoma balthica.Mar Biol 1980;57:51 –60.

Vahl O. Pumping and oxygen consumption rates of Mytilusedulis L. of different sizes. Ophelia 1973;12:45 –52.

Van Beneden R. Environmental effects and aquatic organisms:investigations of molecular mechanisms of carcinogenesis.Environ Health Perspect 1997;105(3):669 –674.

Van Der Knaap WPW, Adema CM, Sminia T. Invertebrateblood-cells-morphological and functional-aspects of thehemocytes in the pond snail Lymnaea stagnalis. CompHaemotol Int 1993;3(1):20 –26.

Van Winkle LS, Johson ZA, Nishio SJ, Bronn CD, PlopperCG. Early events in naphthalene-induced acute Clara celltoxicity–Comparison of membrane permeability and ultras-tructure. Am J Respir Cell Mol Biol 1999;21(1):44 –53.

Viarengo A, Pertica M, Canesi L, Mazzucotelli A, OrunesuM, Bouquegneau JM. Purification and biochemical charac-terisation of a lysosomal copper rich thionein like proteininvolved in metal detoxification in the digestive gland ofmussels. Comp Biochem Physiol C 1989;93C(2):389 –395.

Viarengo A, Accomando R, Roma G, Benatti U, Damonte G,Orunesu M. Differences in lipid-composition of cell mem-branes from Antarctic and mediterranean scallops. Comp

Biochem Physiol B-Biochem Mol Biol 1994;109(4):579 –584.

Viarengo A, Canesi L, Martinez P, Peters L, Livingstone D.Prooxidant processes and antioxidant defence systems in thetissues of the Antarctic scallop (Adamussium colbecki)compared with the Mediterranean scallop (Pecten jacobeus).Comp Biochem Physiol 1995;111B(1):119 –126.

Weslawski JM, Zajaczkowski M, Kwasniewski S, Jezierski J,Moskal W. Seasonality in an Arctic fjord ecosystem: Horn-sund, Spitsbergen. Polar Res 1988;6:185 –189.

Whiteley NM, Taylor EW, ElHaj AJ. A comparison of themetabolic cost of protein synthesis in stenothermal andeurythermal isopod crustaceans. Am J Physiol-Regul Inte-grative Comp Physiol 1996;40(5):1295 –1303.

Widdows J, Bakke T, Bayne BL, Donkin P, Livingstone DR,Lowe DM, Moore MN, Evans SV, Moore SL. Responses ofMytilus edulis to exposure to WAF of North Sea oil. MarBiol 1982;67:15 –31.

Winston GW, Di Giulio RT. Prooxidant and antioxidant mech-anisms in aquatic organisms. Aquat Toxicol 1991;19:137 –191.

Winston GW, Moore MN, Kirchin MA, Soverchia C. Produc-tion of reactive oxygen species by hemocytes from themarine mussel, Mytilus edulis: Lysosomal localisation andeffect of xenobiotics. Comp Biochem Physiol C-PharmacolyToxicol Endocrinol 1996;113(2):221 –229.

Winston GW, Regoli F, Dugas AJ, Fong JH, Blanchard KA.A rapid gas chromatographic assay for determining oxyrad-ical scavenging capacity of antioxidants and biologicalfluids. Free Radical Biol Med 1998;24(3):480 –493.

Winzer K, Winston GW, Becker W, Van Noorden CJF, Kohler¨A. Sex-related responses to oxidative stress in primarycultured hepatocytes of European flounder (Platichthys fle-sus L.). Aquat Toxicol 2001;52:143 –155.

Wolkers J, Jorgensen EH, Nijmeijer SM, Witkamp RF. Time-dependent induction of two distinct hepatic cytochromeP4501A catalytic activities at low temperatures in Arcticchar (Salvelinus alpinus) after oral exposure to ben-zo(a)pyrene. Aquat Toxicol 1996;35(2):127 –138.

Yocis BH, Kieber DJ, Mopper K. Photochemical productionof hydrogen peroxide in Antarctic Waters. Deep-Sea ResPart I- Oceanographic Res Pap 2000;47(6):1077 –1099.