Monitoring of polychlorinated biphenyl contaminationand estrogenic activity in water, commercial feedand farmed seafood

Barbara Pinto & Sonia L. Garritano &

Renza Cristofani & Giancarlo Ortaggi &Antonella Giuliano & Renata Amodio-Cocchieri &Teresa Cirillo & Maria De Giusti &Antonio Boccia & Daniela Reali

Received: 7 June 2007 /Accepted: 26 September 2007 /Published online: 13 November 2007# Springer Science + Business Media B.V. 2007

Abstract We evaluated the concentration and conge-ner distribution of seven “target” polychlorinatedbiphenyls (PCBs) present in water collected indifferent aquaculture farms of the Mediterranean area,commercial feeds, and farmed seafood. PCBs werepresent in feed and in tissues of all the analysedorganisms at levels ranging from 1.96 ng g−1 to124.00 ng g−1 wet weight, and in 10.5% of the watersamples, at levels from under detection limit to

33.0 ng l−1 with total PCB concentrations signifi-cantly higher in samples from the Tyrrhenian Seathan the Adriatic Sea. PCB congener distribution intissues resembled that of feed, suggesting thatcommercial feed is an important source of PCBs.The estrogenicity of organic extracts of the sampleswas also evaluated by using an in vitro yeastreporter assay. Estrogenic activity higher than 10%of the activity induced by 10 nM 17 β-estradiolwas observed in 20.0% of seafood samples and

Environ Monit Assess (2008) 144:445–453DOI 10.1007/s10661-007-0007-6

B. Pinto (*) :R. Cristofani :D. RealiDepartment of Experimental Pathology,Medical Biotechnology, Infectivologyand Epidemiology, University of Pisa,via San Zeno 37, Pisa 56127, Italye-mail: b.pinto@med.unipi.it

R. Cristofanie-mail: r.cristofani@med.unipi.it

D. Realie-mail: d.reali@med.unipi.it

S. L. GarritanoIARC, 150 Cours Albert-Thomas,Lyon Cedex 08 69372, Francee-mail: garritanos@iarc.fr

G. Ortaggi :A. GiulianoDepartment of Chemistry, University of RomaLa Sapienza, P.le Aldo Moro, Roma 5–00185, Italy

G. Ortaggie-mail: giancarlo.ortaggi@uniromal.it

A. Giulianoe-mail: antog.73@virgilio.it

R. Amodio-Cocchieri : T. CirilloDepartment of Food Science, University of NaplesFederico II, Via Università 100, Portici,Naples 80055, Italy

R. Amodio-Cocchierie-mail: amodio@unina.it

T. Cirilloe-mail: tcirillo@unina.it

M. De Giusti :A. BocciaDepartment of Experimental Medicine and Pathology,University of Roma La Sapienza, Viale Regina Elena,Roma 324–00161, Italy

M. De Giustie-mail: maria.degiusti@uniromal.it

A. Bocciae-mail: antonio.boccia@uniromal.it

15.8% of water samples. Seafood and watersamples from the Tyrrhenian Sea were more fre-quently estrogenic than the Adriatic ones (16.45versus 4.08%). A significant correlation of totalPCB concentrations on biological activity wasobserved for sea bass and mussels from the AdriaticSea (p<0.045 and p<0.04, respectively), and for seabass of the Tyrrhenian Sea (p=0.05). These resultsindicate the need of an integral approach in theexposure assessment to potential toxic compounds forhuman via food.

Keywords Monitoring .Marine aquaculture .

Polychlorinated biphenyls . Commercial feed .

Aquatic environment . Estrogenic activity

Introduction

There is extensive literature documenting the bioaccu-mulation of persistent organic pollutants in the aquaticenvironment, but relatively little data are available onaquaculture products. In the last decade, the aquacul-ture industry has grown significantly in Europe due toan increase of the demand of fish products, and it isassuming an increasing role also in Italy.

Beyond to the unquestionable health benefitsderiving from the fish consumption, recent researchesindicate that the fish and fish products, both farmedand wild species, can represent for the consumer ameaningful source of assumption of environmentalpolluting agents such as dioxins, polychlorinatedbiphenils (PCBs) including dioxin-like PCBs (dl-PCBs), brominated flame retardants (BFRs), organo-chlorine pesticides (OCs) and methilmercury(Antunes and Gil 2004; Fernandez et al. 2004; Hiteset al. 2004; Miao et al. 2000; USEPA 2005), in parti-cular when they represent an important part of the diet(Grimvall et al. 1997; Judd et al. 2004; Smith andGangolli 2002).

The presence of these chemical contaminants infish products from the Mediterranean area designedfor human consumption has been well-cited in manyreports during the last 20 years (Bayarri et al. 2001;Naso et al. 2005; Perugini et al. 2006; Storelli et al.2003).

The physiologic and sanitary impacts of somexenobiotics, particularly PCBs and DDTs, on com-mercially important species are also a point of

concern, notably the impacts of compounds whichmay bind the human estrogen and androgen receptors(hERs, ARs) and interfere with the sexual hormoneshomeostasis. The agonistic/antagonistic properties ofPCBs or PCBs mixtures are well-documented (Mooreet al. 1997; Bonefeld-Jørgensen et al. 2001; Laytonet al. 2002; Portigal et al. 2002; Tavolari et al. 2006).

PCBs are very persistent substances, and a widespectrum of injurious effects on humans are reported(Bonefeld-Jørgensen and Ayotte 2003; Dewailly andWeihe 2003; Grandjean et al. 2001; Svensson et al.1994). Of the 209 possible congeners, only a limitednumber are of toxicological interest. Certain PCBs areknown to cause toxic effects with the same mecha-nisms of the “dioxins” (dl-PCBS) as they bind thesame cell receptor (Ah). Epidemiological surveysdemonstrated a significant correlation between PCBsexposure and toxic effects in individuals consuminglarge amounts of contaminated fish products (Rylanderet al. 1995; Patadin et al. 1998; Persky et al. 2001;Schantz et al. 2001).

Several reports on presence of PCBs in aquaticorganisms have been published recently, showingbioaccumulation and a pattern of biomagnificationalong the food chain (Catalan et al. 2004; Serranoet al. 2003). For fish species and especially farmedfish, the main sources of contamination are representedby the feed and the aquatic environment where theylive. Some studies have shown that food is the majorcontributor for PCBs accumulation in cultivatedspecies (Loizeau et al. 2001; Antunes and Gil 2004),even if the concentrations of contaminants in fish mayvary depending on the chemical properties of thecompound, the fish species, the physiological state, thetype of farming (intensive, semi-intensive, extensive),farming technology, season, etc..

On the basis of the available data in the EU SCOOPdatabase, the mean dioxin-like-PCB (dl-PCB) contam-ination levels observed in Europe in the fish species are30 pg I-TEQ/g. The toxicological risk assessment forthe man as a consumer of fish products fromprofessional fishing or from aquaculture, falls withinthe wider matter of the food quality assurance and foodsafety that is for a long time the objective of the EFSA(European Food Safety Authority).

The aim of this work was to evaluate the level andcongener distribution of seven target PCBs (28, 52,101, 118, 138, 153 and 180) in the edible tissue ofcultivated European sea bass (Dicentrarchus labrax,

446 Environ Monit Assess (2008) 144:445–453

Linnaeus, 1758), gilt-head sea bream (Sparus aurata,Linnaeus, 1758), Mediterranean mussel (Mitylusgalloprovincialis) in relation to different factors whichcan influence the level of contamination, such ascommercial aquaculture feeds and farming water. Sincesome of the seven investigated congeners showed invitro estrogenicity in a previous study (Garritano et al.2006), the estrogenic activity of water, feed andseafood organic extracts was also evaluated.

Materials and methods

Sample collection

Sampling stations were located at four fish farmscharacterised by productive technology (two intensiveinland farms, one intensive in-shore farm, and oneextensive off-shore farm) (Table 1), bred species(European sea bass, gilt-head sea bream, Europeanmussels), productivity and geographic location(Tyrrhenian Sea and Adriatic Sea, Central andSouthern Italy). Samplings were repeated during threetime intervals over the year as follows: (a) winter(November, December, January and February); (b)spring and autumn (March, April, May and October);(c) summer (June, July, August and September). On thewhole, 127 samples for chemical analyses and endocrinedisrupting activity were collected: 57 water samples,both outside the plants and inside the ponds and cages;40 seafood samples; 30 commercial feed samples.

Analytical sample preparation

Fish were collected at the achievement of the commer-cial size before being put on the market. The length and

weight of each specimen were measured and recorded,the edible portion (200 g wet weight) of the organismswas selected, homogenised, and subsequently lyophi-lised.

Mussels were analysed by grouping individuals incommon samples (5 g wet weight). Feeds were milledthin, divided in 5 g aliquots and lyophilized.

Fat extraction

Fat was cold-extracted from lyophilised fish muscletissues and mussels with petroleum ether/acetone (1:1,v/v). The extract was passed through a glass tubepacked with anhydrous sodium sulphate and thenevaporated by rotavapor (40°C and low pressure) andthe lipid residue was weighed.

Chemical analysis

The cleanup of fat extracts (50 mg) was carried out onExtrelut-NT3/Extrelut-NT1 cartridges (Merck KGaA-Darmstadt, Germany) with the addition of 0.36 g ofC-18 Isolute 40–60 mesh (Merck KGaA-Darmstadt,Germany) and eluted with acetonitrile. The extractswere concentrated under vacuum at 40°C, cleaned upby column adsorption-chromatography on Florisil 60/100 mesh (Supelco Bellefonte, PA USA) activated at130°C for 2 h and eluted with 30 ml n-hexane addedin 5 ml aliquots. The eluate was concentrated to asmall volume (<1 ml) by evaporation at roomtemperature under a flow of N2, and 1 ml isooctanewas added as a keeper.

Feeds were extracted as for fish. The extraction fromwater samples was carried out by adding 1 l sample to2 l separatory funnel with dichloromethane, vigorouslyshaking for 3 min, and purifying extracts on Fluorisil

Table 1 Plants description and geographic location

Plants Number of ponds Extent Species Annual productivity

Farming A Intensive inland (Tyrrhenian Sea) 34 ponds 18,400 m3 Gilthead sea bream c.a. 2,000 quintalsSea bass

Farming C Intensive inland (Adriatic Sea) 35 ponds 25,000 m3 Gilthead sea bream c.a. 500 tonsSea bassMussels

Farming G Intensive inshore (Tyrrhenian Sea) 56 cages 110,000 m3 Gilthead sea bream c.a. 1,200 tonsSea bassMussels

Farming T Extensive offshore (Adriatic Sea) 50 rows 1 km2 Mussels c.a. 750 tons

Environ Monit Assess (2008) 144:445–453 447

PR 60/100 mesh. Seven PCB congeners (IUPACnn. 28, 52, 101, 118, 153, 138 and 180) were detectedin all samples according to the analytical method ofItalian Public Health laboratories (ISS 2002).

For PCBs quantification, purified samples were in-jected into a capillary gas chromatographer HEWLETT-PACKARD 3396 Series III, equipped with an electroncapture detector (GC-ECD) and a capillary columnHP-5 (Crosslinked 5% PHME Siloxane), length 30 m,d.i. 0.32 mm, film thickness 0.25 µm. The analysis wascarried out with Temperature program 60°C for 2 min,increasing of 10°C/min to 170°C stay for 2 min, in-creasing of 2°C/min to 210°C, increasing for 10°C/minto 260°C.

The results were confirmed by GC-MS. The internalstandard solution (PCB 209) was added to the extractbefore injecting. The evaluation of PCB concentrationsin the samples was carried out by comparison with acalibration curve obtained by a pool of the seven ICESPCB congeners: PCB 28 (2,4,4′ tri-chlorobiphenyls),PCB 52 (2,2′,5,5′ tetra-chlorobiphenyls), PCB 101(2,4,5,2′,5′ penta-chlorobiphenyls), PCB 118(2,4,5,3′,4′ penta-chlorobiphenyls), PCB 138 (2,2′3,4,4′,5′ hexa-chlorobiphenyls) PCB 153 (2,2′,4.4′,5,5′hexa-chlorobiphenyls) and PCB 180 (2,2′,3,4,4′,5,5′hepta-chlorobiphenyls). All of the compounds(95–99% pure) were purchased from Dr. Ehrenstorfer(GmbH, Augsburg, Germany). In the analyticalconditions applied, the detection limits in fish (onwet weight) and feed were: 0.2 ng g−1 wet weight(w wt) for the PCBs nn 180; 0.25 for PCB 138;0.3 ng g−1 (w wt) for the PCB n 153; 0.45 for PCB118; 0.5 ng g−1 (w wt) for the PCBs 101: 0.60 forPCB 28 and 0.66 ng g−1 (w wt) for the PCB 52; inwater samples, 0.08 ng l−1 for PCB 180, 138 and 153;0.09 ng l−1 for PCB 118 and 101; 0.10 ng l−1 for PCB28 and 52.

The mean recovery obtained by PCB standardspiked samples was 70±7% for fish, mussels andfeed. For water samples the mean recovery was 76±4%. Total PCB levels were calculated as the sum (∑)of all the seven determined congeners.

Yeast bioassay

Estrogenic activity of organic extracts from water,fish and feed was tested by Saccharomyces cerevisiaeyeast strain RMY326 ER-ERE (His3 Leu2-3, 112trp1-1 ura3-52/hER-TRP1-2µ[pG/ER(G)], ERE-CYC-

LacZ-URA3-2 µ[pUCΔSS-ERE], HIS-3CEN/ARS[pRS423]) which carries the human estrogen receptoralpha (hERa) and a Xenopus laevis vitellogeninestrogen-responsive element (ERE) linked to a report-er gene lacZ encoding for the enzyme β-galactosi-dase. Plasmid [pG/ER(G)] was used as the yeastexpression vector for ERα and [PUCΔSS-ERE] as itsβ-galactosidase reporter plasmid (Liu and Picard1998; Liu et al. 1999). The used method has beenpreviously described (Pinto et al. 2005; Garritanoet al. 2006).

Yeast cultures were incubated at 28°C for 7 h bycontinuously shaking on an orbital shaker (210 rpm)in 1 ml of selective medium, then diluted in freshmedium to an optical density ≤0.1 (OD600 nm) andincubated at 28°C for 17 h (overnight) in the absenceor presence of 17 β-estradiol (positive control),vehicle (negative control) and organic extracts.Dimethylsulphoxide (DMSO) was used as solvent.Solutions of extracts were evaporated under a gentleflow of nitrogen and the pellet was resuspended in10 µl DMSO. The samples were added to the yeastculture so that the concentration of DMSO did notexceed 1% (v/v).

The estrogenic activity of PCB standards waspreviously tested (Garritano et al. 2006).

The β-galactosidase activity of the samples wasexpressed as a percentage of the activity obtainedwith 10 nM 17 β-estradiol (E2) (Gong et al. 2003;Pinto et al. 2005).

Statistical analysis

Association between categorical variables was testedby chi-square test or Fisher exact test, when appro-priate. A cut-off of β-gal activity higher than 10% ofthe activity elicited by 10 nM E2 was used todichotomize estrogenic activity values.

Spearman correlation coefficient was used to assessthe relationship among quantitative variables. Log-trasformed total PCBs values and arcsin-transformedβ-gal activity values were used in parametricanalyses. General Linear Models (GLM) were usedto test mean differences in total PCB and β-gal ac-tivity in the classes determined by “sea” and “type.”

Bonferroni’s test was used in the post-hoc analysis.The effect of total PCBs on β-gal activity wasestimated by simple linear regression, separately forthe two seas and the three seafood types (gilthead

448 Environ Monit Assess (2008) 144:445–453

seabream, sea bass, and mussels). SAS software 8.2version was used to perform statistical computations.

Results and discussion

PCBs assessment in water samples, feed and tissues

PCBs were present in 10.5% of the water samples, atlevels from <d.l. to 33.0 ng l−1 and in tissues of all theanalysed organisms and in commercial feed (100% ofthe samples) at levels ranging from 1.96 ng g−1 to124.00 ng g−1 wet weight (Table 2).

PCB 138 and 153 showed the highest concen-trations (followed in descending order by the 118,101, 28, 52, 180 in seafood tissues and by 118, 101,52, 180, 28 in feed, respectively), representing morethan 25% of total PCBs content, probably due to theirhigh lipophilicity and persistence in the aquaticecosystem. PCB 28 was the dominant congener inwater samples (Table 3).

Bioconcentration of PCBs in aquatic organisms isdemonstrated to be proportional to the degree ofchlorination (Gray 2002; Naso et al. 2005). The lesschlorinated homologues are usually more rapidlymetabolised and eliminated than the higher chlorinated

Table 3 Individual PCB congener concentrations (median values and range) determined in seafood (ng g−1 w.wt.), feed (ng g−1),water (ng l−1)

Sample PCB 28 PCB 52 PCB 101 PCB 118 PCB 138 PCB 153 PCB 180

G. seabream <d.l. 1.22 2.11 3.67 4.20 3.30 0.97(<d.l.–18.00) (<d.l.–7.40) (<d.l.–8.80) (0.3–11.50) (0.56–16.00) (<d.l.–14.40) (<d.l.–6.00)

Sea bass <d.l. 1.61 2.39 3.77 4.63 4.22 0.69(<d.l.–18.00) (<d.l.–10.00) (<d.l.–10.57) (<d.l.–11.29) (1.25–24.85) (0.95–27.00) (<d.l.–6.21)

Mussels <d.l. <d.l. 2.31 1.00 3.03 1.99 <d.l.(<d.l.–1.44) (<d.l.–0.41) (<d.l.–11.88) (<d.l.–5.65) (1.20–50.00) (0.61–60.00) (<d.l.–8.00)

Feed <d.l. 0.05 1.36 1.55 2.06 1.67 0.40(<d.l.–1.50) (<d.l.–21.00) (<d.l.–6.30) (<d.l.–8.78) (<d.l.–8.60) (<d.l.–10.00) (<d.l.–3.03)

Water <d.l. <d.l. <d.l. <d.l. <d.l. <d.l. <d.l.(<d.l.–33.00) (<d.l.–3.94) (<d.l.–0.70) (<d.l.–0.34) (<d.l.–0.34) (<d.l.–2.70) (<d.l.)

d.l. Detection limit

Table 2 Total PCB concentrations (median values and range) in seafood (ng g−1 w.wt.), feed (ng g−1), water (ng l−1) from differentfarming plants

Sample Plant G Plant A Plant C Plant T(T)a (T)a (A)b (A)b

Fish 40.10 15.78 9.05 –10.20–72.00 3.52–28.52 4.95–89.69

G. seabream 42.00 13.52 9.99 –10.20–72.00 3.52–18.06 8.26–43.44

Sea bass 40.00 19.74 6.04 –12.63–46.00 5.58–28.52 4.95–89.69

Mussels 20.78 – 4.58 8.377.48–124.00 3.29–55.94 5.45–58.96

Feed 9.85 6.58 11.63 –2.00–33.60 1.96–16.68 2.83–29.00

Water <d.l <d.l. <d.l. –<d.l.–33.00 <d.l. <d.l.

d.l. Detection limita Tyrrhenian Seab Adriatic Sea

Environ Monit Assess (2008) 144:445–453 449

congeners because of the presence of more unsubsti-tuted ring positions on their biphenyl rings availablefor the metabolic attack, and this could explain the lowfrequency of contamination by PCB 28 and 52observed in fish tissues. Moreover, these non-coplanar,low chlorinated congeners have a lower logKow andtherefore, a lower propensity to leave the aqueousenvironment for organic compartments (Naso et al.2005). In water collected at the farming plant G,which were the only water samples to have measur-able concentrations of PCBs, the tri- and tetrachlor-obiphenyls (PCB 28 and 52) showed the highestconcentration levels.

Sea bass and gilt-head sea bream showed not astatistically significant difference in PCBs (median =19.74 and 16.04, respectively), while mussel samplesshowed a wider range of contamination (3.29–124.00 ng g−1 wet weight). The higher valuesdetected in mussels may be related to differences inmetabolism but also ascribed to their feeding habits;mussels are filtering organisms, and hence mayconcentrate large amounts of environmental contam-inants into their tissues.

Comparison between the two seas and for farming plant

Based on GLM approach to two factors analysis ofvariance (sea and type), total PCB concentrationswere significantly higher (more than the double) infish and mussels from the Tyrrhenian Sea than theAdriatic Sea (p<0.002). Higher differences werefound in sea bass in respect of mussels (p<0.001).

The observed differences between the two seasseem to be mainly attributable to samples collected atthe intensive in-shore plant G, whose marine waterwas as well the only to reveal detectable levels of

PCBs. Gilthead sea bream and sea bass from inlandfarms (A and C) were generally less polluted fromPCBs than those coming from the in-shore sea-farm G(p<0.02).

Individual congener distribution also differeddepending on the geographic location of farming plants,with the species collected in the Adriatic Sea mostlycontaminated with penta- and hexachlorobiphenyls,whereas samples from the Tyrrhenian Sea showed anunvarying pattern of contamination (Fig. 1).

Tissue and feed PCB concentrations correlation

In general, the overall PCB congener distribution infish tissues resembled that of feed (Fig. 2). Thecongeners 138 and 153 were the most frequentlydetected PCBs both in fish tissues and in feed (97%up to 100% of tissues samples and 97% of feedsamples, respectively.

In cultivated organisms, the relative proportions ofPCB congeners, determined as the concentration ratiosbetween individual congeners and total PCB, were

0

20

40

60

80

100

28 52 101 118 138 153 180

Congener

Po

sitiv

e s

am

ple

s (%

)

Fig. 1 Prevalence of positive samples for individual PCBcongeners in fish and mussels from the Adriatic Sea (filledbars) and the Tyrrhenian Sea (empty bars)

0

20

40

60

80

100

28 52 101 118 138 153 180

Congener

Pos

itive

sam

ples

(%

)

Fig. 2 Distribution patterns of PCB congeners in fish (filledbars) and feed (empty bars) samples

0

5

10

15

20

25

30

28 52 101 118 138 153 180

Congener

% to

tal P

CB

Fig. 3 PCB congener patters in fish (filled bars) and feed(empty bars) samples. The relative proportions of PCBcongeners were determined as the concentration ratios betweenindividual congeners and total PCB content

450 Environ Monit Assess (2008) 144:445–453

similar to those measured in diet feed (Fig. 3) (r=0.96,p<0.0005, n=7) and this is in accordance withfindings of Antunes and Gil (2004).

These results suggest that commercial diet is animportant source of PCBs in farmed species but aloneit does not explain the observed differences in totalPCB concentration levels of organisms from the twoseas. In fact, total concentrations of PCBs incommercial feed used in the different farming plantswere similar. Therefore, the uptake of xenobioticsfrom the environmental water also plays an importantrole in bioaccumulation of several classes of chemicalsubstances.

Estrogenic activity

The β-gal activity induction of the seven PCB stand-ards was assessed in a previous work (Garritano et al.2006). PCB 118, which belongs to the class of 12PCBs identified by WHO as a “dioxin-like” PCB, washighly estrogenic in the yeast assay (β-gal activity =85.88±19.79 of E2, concentration 5 µg ml−1). At0.05 µgml−1 PCB138 and 153 elicited weak estrogenicactivity (20.95±7.82 and 26.37±5.63, respectively).

Organic extracts from fish tissues (gilt-head seabream and sea bass) and water elicited an ER-mediatedresponse higher than 10% E2 in 20.0 and 15.8% ofanalysed samples, respectively. On the whole, of the127 samples processed only 11.8% of them wereestrogenic in the yeast estrogen assay (range 10.03–48.0% E2). Extracts from mussels and feed samplesdid not show any appreciable induction of β-galactivity and for many of them (30.0 and 20.0%,respectively) it was not possible to measure theinduction of enzymatic activity because of theircytotoxic effect on the yeast cells.

Taken together, water and seafood samples from theTyrrhenian Sea were more frequently positive thanAdriatic ones (22.03 versus 5.26%). In particular, thesampling area of the Tyrrhenian Sea where is locatedthe farming plant G seems to significantly influence thequality of stabling water and of seafood, as in this sitethe highest frequency of positive samples (38.46% ofseafood samples and 46.67% of water samples; p=0.033) beyond the highest value of estrogenic activityfor water (48.0% E2, corresponding to 81.4 ng/l E2-equivalents) was recorded. High values of enzymaticinduction were also detected in sea bass and water frominland plants (up to 33.34% of the β-gal activity of E2).

When data were processed on the whole, theoverall total PCB content and β-gal activity did notshow any significant correlation (r=0.003; p=0.76).

When considering data stratified for geographiclocation and farming plant, a significant regression oftotal PCB concentrations on biological activity wasobserved for sea bass (p<0.045) and mussels (p<0.04)from the Adriatic Sea. In Tyrrhenian Sea a similarcorrelation (p=0.05) was observed for sea bass.

Environmental and public health

This study determined the concentrations and distri-bution of seven “target” polychlorinated biphenyls(PCBs) recommended by the European Union asindicators of PCB contaminations, in environmentalmedia (feed and water) and some commonly con-sumed marine cultured organisms collected in fourfarming plants. Moreover, it evaluated the ability oforganic extracts of the samples to exert estrogeniceffects by binding to the nuclear human estrogenreceptor alpha in an in vitro yeast estrogen assay.

In recent years scientific commissions, the NationalHealth Service (ISS) and the National Institute ofResearch for food and Nutrition (INRAN), in collab-oration with the Food Safety and Veterinary PublicHealth Service, aquaculture farms and category asso-ciations, promoted an increased fish consumption indiet, in order to mainly prevent the risk of cardiovas-cular disease. The health benefits of eating fish foodderiving from their high nutritional value and theircontent in omega-3 polyunsaturated fatty acids havebeen well documented (Kris-Etherton et al. 2003;Foran et al. 2005; He et al. 2004). However, bothfarmed and wild fish and seafood have been shown toaccumulate a variety of toxic pollutants just withinthe lipid component, some of which associated withadverse health effects and endocrine disorders.

In commercial feed and in water a mixture ofseveral environmental compounds other than PCBsincluding other persistent bioaccumulative contami-nants (POPs) able to bind to the estrogen receptormay be present which have been shown to haveestrogenic activity both on the yeast test and on otherin vitro and in vivo assays. Therefore, the observedestrogenic activity might be the result of the inter-actions between different contaminants with agonisticand/or antagonistic properties.

Environ Monit Assess (2008) 144:445–453 451

The risk of endocrine disruption due to the presenceof PCBs (particularly of dioxin-like ones) and otherPOPs, in animal tissues has undergone research in recentyears. This risk is supposed to exist for humans, toobeing them at the top of the food chain and therefore ofthe bio-accumulation process. The human health effectsof exposure to chemical pollutants present in fish tissuesare a function both of the toxicological properties ofcontaminants which can have additive, synergistic orantagonistic effects in vivo, and their concentration infish tissues, and fish consumption rates.

These results indicate the need of an integralapproach in the exposure assessment to potentialtoxic compounds occurring in environmental matricesand food which takes into account the analyticaldetermination of chemical compounds as well as theevaluation of a biological activity which can interferewith metabolic pathways in organisms.

This kind of approach may hence provide a furtherresource for preliminary estimation of the risk forhumans to introduce with diet, in this case seafood, amixture of compounds, some of which also able todisrupt the endocrine system, particularly the estro-genic pathway.

Acknowledgements This study was part of a multidisciplinaryResearch Project on Marine Environment and AquacultureProducts: Environmental Protection and Quality supported by theItalian Ministry of the Environment (Project Prot. N SvS/C7/8735).

Project leader: A. BocciaCollaborating groupsProject coordinator and sampling manager: De Giusti MChemical analysis: Amodio-Cocchieri R, Ortaggi G, Cirillo T,

Giuliano AEndocrine disruptors: Pinto B, Garritano S, Reali DMicrobiological analysis: De Giusti M, Grasso GM, DeVito E,

Tufi D, Del Cimmuto AGenetically Modified Organisms in feed: Romano-Spica V,

Ricciardi GStatistical analysis: Cristofani R

References

Antunes, P., & Gil, O. (2004). PCB and DDT contamination incultivated and wild sea bass from Ria de Aveiro, Portugal.Chemosphere, 54, 1503–1507.

Bayarri, S., Baldassarri, L. T., Iacovella, N., Ferrara, F., & diDomenico, A. (2001). PCDDs, PCDFs, PCBs and DDE inedible marine species from the Adriatic Sea. Chemo-sphere, 43, 601–610.

Bonefeld-Jørgensen, E. C., Andersen, H. R., Rasmussen, T. H.,& Vinggaard, A. M. (2001). Effect of highly bioaccumu-lated polychlorinated biphenyl congeners on estrogen andandrogen receptor activity. Toxicology, 158, 141–153.

Bonefeld-Jørgensen, E. C., & Ayotte, P. (2003). Toxicologicalproperties of persistent organic pollutants and relatedhealth effects of concern for the Arctic populations. InAMAP 2003. AMAP Assessment 2002: Human health inthe Arctic, Chapter 6. Arctic Monitoring and AssessmentProgramme (AMAP), Oslo, Norway 2003:xiv+137pp.Retrieved from: http://www.amap.no HH_CO6.pdf ISBN82-7971-016-7.

Catalan, J., Ventura, M., Vives, I., & Grimalt, J. O. (2004). Theroles of food and water in the bioaccumulation oforganochlorine compounds in high mountain lake fish.Environmental Science and Technology, 38, 4269–4275.

Dewailly, E., & Weihe, P. (2003). The effect of Arctic pollutionon population health. In AMAP 2003. AMAPAssessment2002: Human health in the Arctic, Chapter 9. ArcticMonitoring and Assessment Programme (AMAP), Oslo,Norway 2003:xiv + 137pp. Retrieved from http://www.amap.no HH_CO9.pdf ISBN 82-7971-016-7.

Fernandez, M. A., Gomara, B., Bordajandi, L. R., Herrero, L.,Abad, E., & Abalos, M., et al. (2004). Dietary intakes ofpolychlorinated dibenzo-p-dioxins, dibenzofurans and di-oxin-like polychlorinated biphenyls in Spain. Food Addi-tives and Contaminants, 21, 983–991.

Foran, J. A., Carpenter, D. O., Hamilton, M. C., Knuth, B. A.,& Schwager, S. J. (2005). Risk-based consumption advicefor farmed Atlantic and wild Pacific salmon contaminatedwith dioxin-like compounds. Environmental Health Per-spectives, 113, 552–556.

Garritano, S., Pinto, B., Calderisi, M., Cirillo, T., Amodio-Cocchieri, R., & Reali, D. (2006). Estrogen-like activity ofseafood related to environmental chemical contaminants.Environmental Health: A Global Access Science Source,5, 9. Retrieved March 30, 2006, from http://www.ehjournal.net/contents/5/1/9.

Gray, J. S. (2002). Biomagnification in marine systems: Theperspective of an ecologist. Marine Pollution Bulletin, 45,46–52.

Gong, Y., Chin, H. S., Lim, L. S., Loy, C. J., Obbard, J. P., &Yong, E. L. (2003). Clustering of sex hormone disruptorsin Singapore’s marine environment. Environmental HealthPerspectives, 111, 1448–1453.

Grandjean, P., Weihe, P., Burse, V. W., Needham, L. L., Storr-Hansen, E., Heinzow, B., et al. (2001). Neurobehaviouraldeficits associated with PCB in 7-year-old childrenprenatally exposed to seafood neurotoxicants. Neurotox-icology and Teratology, 23, 305–317.

Grimvall, E., Rylander, L., Nilsson-Ehle, P., Nilsson, U.,Strömberg, U., Hagmar, L., et al. (1997). Monitoring ofpolychlorinated biphenyls in human blood plasma: Meth-odological developments and influence of age, lactation,and fish consumption. Archives of Environmental Con-tamination and Toxicology, 32, 329–336.

He, K., Song, Y., Daviglus, M. L., Liu, K., Van Horn, L., Dyer, A.R., et al. (2004). Accumulated evidence on fish consumptionand coronary heart disease mortality. Circulation, 109,2705–2711.

452 Environ Monit Assess (2008) 144:445–453

Hites, R. A., Foran, J. A., Carpenter, D. O., Hamilton, M. C.,Knuth, B. A., & Schwager, S. J. (2004). Global assess-ment of organic contaminants in farmed salmon. Science,303, 226–229.

ISS Istituto Superiore di Sanità. (2002). Istisan report 02/38ISSN 1123-3117, ISS, Roma; 2002. Electronic versionretrieving from http://www.iss.it/binary/publ/publi/0238.1109329216.pdf.

Judd, N., Griffith, W. C., & Faustman, E. M. (2004).Contribution of PCB exposure from fish consumption tototal dioxin-like dietary exposure. Regulatory Toxicologyand Pharmacology, 40, 125–135.

Kris-Etherton, P. M., Harris, W. S., & Appel, L. J. (2003). Fishconsumption, fish oil, omega-3 fatty acids, and cardiovas-cular disease. Arteriosclerosis, Thrombosis, and VascularBiology, 23, 20–30.

Layton, A. C., Sanseverino, J., Gregory, B. W., Easter, J. P.,Sayler, G. S., & Schultz, T. W. (2002). In vitro estrogenreceptor binding of PCBs: Measured activity anddetection of Hydroxylated metabolites in a recombinantyeast assay. Toxicology and Applied Pharmacology, 180,157–163.

Liu, J. W., Jeannin, E., & Picard, D. (1999). The anti-estrogenhydroxytamoxifen is a potent antagonist in a novel yeastsystem. Biological Chemistry, 380, 1341–1345.

Liu, J. W., & Picard, D. (1998). Bioactive steroids ascontaminants of the common carbon source galactose.FEMS Microbiological Letters, 159, 167–171.

Loizeau, V., Abarnou, A., Cugier, P., Jaouen-Madoulet, A., LeGuellec, A. M., & Menesguen, A. (2001). A model ofPCB bioaccumulation in the sea bass food web from theSeine Estuary (eastern English Channel). Marine PollutionBulletin, 43, 242–55.

Miao, X.-S., Swenson, C., Woodward, L. A., & Li, Q. X.(2000). Distribution of polychlorinated biphenyls inmarine species from French Frigate Shoals, North PacificOcean. The Science of the Total Environment, 257, 17–28.

Moore, M., Mustain, M., Daniel, K., Chen, I., Safe, S.,Zacharewski, T., et al. (1997). Antiestrogenic activity ofhydroxylated polychlorinated biphenyl congeners identi-fied in human serum. Toxicology and Applied Pharmacol-ogy, 142, 160–168.

Naso, B., Perrone, D., Ferrante, M. C., Bilancione, M., &Lucisano, A. (2005). Persistent organic pollutants in edilemarine species from the Gulf of Naples, Southern Italy.The Science of the Total Environment, 343, 83–95.

Patadin, S., Koopman-Esseboom, C., de Ridder, M. A.,Weisglas-Kuperus, N., & Sauer, P. J. (1998). Effects ofenvironmental exposure to polychlorinated and biphenylsand dioxin on birth size and growth in Dutch children.Pediatric Research, 44, 538–545.

Persky, V., Turjk, M., Anderson, H. A., Hanrahan, L. P., Falk,C., Steenport, D. N., et al. (2001). The effects of PCB

exposure and fish consumption on endogenous hormones.Environmental Health Perspectives, 109, 1275–1283.

Perugini, M., Giammarino, A., Olivieri, V., Nardo, W., &Amorena, M. (2006). Assessment of edile marine speciesin the Adriatic Sea for contamination from polychlorinatedbiphenyls and organochlorine insecticides. Journal ofFood Protection, 69, 1144–1149.

Pinto, B., Garritano, S., & Reali, D. (2005). Occurrence ofestrogen-like substances in the marine environment of theNorthern Mediterranean Sea. Marine Pollution Bulletin,50, 1681–1685.

Portigal, C. L., Cowell, S. P., Fedoruk, M. N., Butler, C. M.,Rennie, P. S., & Nelson, C. C. (2002). Polychlorinatedbiphenyls interfere with androgen-induced transcriptionalactivation and hormone binding. Toxicology and AppliedPharmacology, 179, 185–194.

Rylander, L., Stromberg, U., & Hagmar, L. (1995). Decreasedbirth weight among infants borne to women with a highdietary intake of fish contaminated with persistent organ-ochlorine compounds. Scandinavian Journal of Work andEnvironmental Health, 21, 368–375.

Schantz, S. L., Gasior, D. M., Polverejan, E., McCaffrey, R. J.,Sweeney, A. M., Humphrey, H. E., et al. (2001). Impair-ments of memory and learning in older adults exposed topolychlorinated biphenyls via consumption of Great Lakesfish. Environmental Health Perspectives, 109, 605–611.

Serrano, R., Simal-Julian, A., Pitarch, E., Hernandez, F., Varo,I., & Navarro, J. C. (2003). Biomagnification study onorganochlorine compounds in marine aquaculture: The seabass (Dicentrarchus labrax) as a model. EnvironmentalScience and Technology, 37, 3375–3381.

Smith, A. G., & Gangolli, S. D. (2002). Organochlorinechemicals in seafood: Occurrence and health concerns.Food and Chemical Toxicology, 40, 767–779.

Storelli, M. M., Giacominelli-Stuffler, R., D’Addabbo, R., &Marcotrigiano, G. O. (2003). Health risk of coplanarpolychlorinated biphenyl congeners in edible fish from theMediterranean Sea. Journal of Food Protection, 66, 2176–2179.

Svensson, B. G., Hallberg, T., Schultz, A., & Hagmar, L.(1994). Parameters of immunological competence insubjects with high consumption of fish contaminated withpersistent organochlorine compounds. Archives of Occu-pational and Environmental Health, 65, 351–358.

Tavolari, S., Bucci, L., Tomasi, V., & Guarnieri, T. (2006). Selectedpolychlorobiphenyls congeners bind to estrogen receptoralpha in human umbilical vascular endothelian (HUVE) cellsmodulating angiogenesis. Toxicology, 218, 67–74.

USEPA United Nations Environmental Protection Agency,Office of Water (2005). Update 2005: National Listingof Fish and Wildlife Advisories. EPA-823-F-05-004.Retrieved September, 2005, from http://www.epa.gov/waterscience/fish.

Environ Monit Assess (2008) 144:445–453 453